Development of urinary impairment after spinal cord injury courses with altered urethral serotonin signalling in the female mice

Abstract

Spinal Cord Injury (SCI) often leads to urinary incontinence. While alterations in the bladder have been considered the primary cause of post-SCI urinary dysfunction, evidence suggests that urethral afferents may influence post-SCI bladder activity. These afferents respond to locally produced serotonin (5-HT), and 5-HT signalling may likely be affected by SCI. Here, we investigated the role of urethral 5-HT on post-SCI bladder reflex activity using female C57BL/6 mice (wild-type -WT- and animals lacking urethral 5-HT (tryptophan hydroxylase 1 -tph1- null -tph1^−/−). In WT animals, bladder overactivity and increased 5-HT⁺ urethral cells were observed at one and four weeks post-SCI (1w and 4w, respectively), while, in contrast, tph1^−/− mice showed some preservation of bladder function. Both genotypes developed urethral smooth muscle atrophy 4w post-SCI, but fibrosis of striated muscle was detected only in tph1^−/− 4w SCI mice. Sensory and cholinergic fibres were upregulated only in WT SCI mice, suggesting 5-HT depletion may block their expansion. Pharmacological blockade of 5-HT receptors did not prevented urinary impairment but affected SCI-induced bladder dysfunction. These observations indicate that urethral 5-HT contributes to SCI-induced urinary impairment and urethral tissue reorganization, offering new insights into the urethrovesical reflex and potentially pinpointing new therapeutic targets.

Introduction

Appropriate urine storage and release are dependent on the coordinated contraction and relaxation of the lower urinary tract (LUT) organs- the urinary bladder (the urine reservoir), and the urethra, considered the bladder outlet and encompassing the urethral passage and sphincters¹. The synchronised activity between the bladder and the urethra is controlled by intricate neuronal circuits, including sensory afferent pathways, and autonomic and motor circuits that mediate the correct communication between the LUT and supraspinal structures. These circuits are severely disturbed by spinal cord injury (SCI), leading to urinary impairment^1,2. Following an initial post-SCI period of little or no bladder reflex activity³, an alternative micturition reflex arises at the lumbosacral spinal cord functioning in the absence of supraspinal input^2,4. Consequently, post-SCI micturition becomes involuntary, and the presence of neurogenic detrusor overactivity (NDO) gives rise to strong, frequent and involuntary detrusor contractions, often concurrent with detrusor-sphincter dyssynergia (DSD)⁵, resulting in urinary incontinence. Moreover, due to sustained, dangerously high intravesical pressures, there is considerable risk of damage to the upper urinary tract, development of autonomic dysreflexia, and recurrent urinary infections, worsening functional outcomes and patients’ quality of life^2,6.

The bladder has traditionally been seen as the main player in micturition and the central focus of urinary dysfunction after neurologic injury. Thus, bladder alterations in response to SCI are well documented^7–9, and pharmacological interventions are mainly focused on modulating bladder activity⁴. However, recent data obtained in various animal models and human patients have shown that the urethra, previously seen as a passive LUT component, is an active participant in micturition. In fact, urethral afferent input, generated by urine flow along the urethra during voiding, triggers a facilitatory ureterovesical feedback that increases voiding efficiency^10–15. This reflex is suppressed by urethral anaesthesia, suggesting this effect must originate from urethral flow receptors^12,13,16. Moreover, urethral electrical stimulation triggers detrusor contraction^17–20, evoking inhibitory or excitatory bladder reflexes depending on the location and frequency of the stimuli^20,21.

As urethral nerve fibres present in the lamina propria or muscle layers do not reach the luminal surface, from where stimuli originate, the perception of sensory input in the urethra is likely mediated by specialised epithelial cells, intermingled with other lining cells^22–25. These specialised cells are unique and present neuron-like properties, including the expression of sensory receptors and the ability to release neuromediators when stimulated^12,26, and, thus, are often designated by paraneuronal cells. This group of cells include serotoninergic paraneurones, which are mainly located along the urethral sphincter region, in the vicinity of sensory and cholinergic nerves²⁷. The importance of serotonin (5-HT) in the regulation of an urethrovesical reflex has been shown by Coelho and coworkers. They demonstrated that, under isovolumetric cystometry, the instillation of a 5-HT solution into the urethral channel triggers robust bladder-reflex contractions. On the other hand, under normal cystometry, i.e. with an open urethral outlet, 5-HT administration decreases frequency, while strongly enhancing the amplitude of bladder contractions. Altogether, these observations support the existence of a serotonergic urethrovesical reflex that operates through complex neuronal mechanisms rather than a simple translocation of neuromediators between adjacent organs²⁷.

Changes in the urethra in cases of bladder dysfunction, as after SCI, have been seldom addressed, although there is abundant information about SCI-induced changes in bladder structure and function^4,7,8,28,29. This matter was explored in a recent study, which described significant urethral tissue rearrangement in response to SCI, including smooth muscle atrophy and fibrosis in the external urethral sphincter, accompanied by loss of sensory and autonomic neuronal fibres³⁰. Here, to expand this previous study and further deepen our understanding of the effects of SCI on LUT function, we focused on the role of urethral 5-HT and further investigated the effects of SCI on this organ.

Results

Tph1^-/- mice display attenuated urinary dysfunction in chronic SCI

To evaluate bladder function after SCI, animals were deeply anaesthetised with urethane and underwent cystometries before euthanasia. Urodynamic recordings from WT intact (WT INT) animals were indicative of normal bladder function, with high-amplitude and low-frequency bladder contractions and normal intravesical pressures (Fig. 1A, G-J; Table 1). Intact animals lacking peripheral serotonin (tph1^−/− INT) presented a similar pattern of bladder function (Fig. 1B, G-J; Table 1).

Fig. 1.

Effects of SCI and peripheral serotonin on urinary dysfunction. (A-F) Bladder function was assessed by cystometry under urethane anaesthesia. Representative cystometrograms depicting the voiding function of (A) Wild-type (WT) intact (INT), (B) tph1^−/− INT, (C) WT one week after injury (1w SCI), (D) tph1^−/− 1w SCI, (E) WT four weeks after injury (4w SCI), and (F) tph1^−/− 4w SCI mice; (G-J) Analysis of urodynamic parameters of (G) frequency of bladder contractions, (H) amplitude of bladder contractions, (I) basal pressure, and (J) peak pressure. INT animals from both genotypes presented typical bladder function (A-B, G-J), while both WT and tph1^−/− SCI mice showed signs of urinary dysfunction at 1w SCI (C-D, G-J). Four weeks after injury, urinary dysfunction was maintained in both genotypes, but tph1^−/−mice presented attenuated signs of urinary dysfunction, compared to WT 4w SCI animals (F, G-J). Graphs represent mean ± standard deviation, and p < 0.05 was considered statistically significant. ****p < 0.0001 vs. WT INT; **p < 0.001 vs. tph1^−/− INT; ****p < 0.0001 vs. tph1^−/− INT; #p < 0.05 vs. WT 1w SCI; ## p < 0.01 vs. WT 1w SCI; $p < 0.05 vs. WT 4w SCI; $$p < 0.01 vs. WT 4w SCI. Two-way ANOVA followed by Tukey´s post-hoc multiple comparison test.

Table 1.

Urodynamic parameters from WT and tph1^−/−intact or spinal cord-injured mice. Abbreviations: WT, Wyld-Type; tph1^−/−, homozygote Tryptophan hydroxylase 1 deficient mice; SCI, spinal cord injury. ****p ≤ 0.0001 vs. WT INT, #p ≤ 0.05 vs. WT 1w SCI, ##p ≤ 0.01 vs. WT 1w SCI; ****p ≤ 0.05 vs. tph1^−/− INT; $p ≤ 0.05 vs. WT 4w SCI; $$≤0.01 vs. WT 4w SCI. Data are presented as mean ± standard deviation.

	Frequency of bladder reflex contractions (contractions per min)	Voiding amplitude (cm H2O)	Basal pressure (cm H2O)	Peak pressure (cm H2O)
WT INT	0.4 ± 0.12	37.5 ± 1.7	8.54 ± 0.85	46.03 ± 2.3
tph1−/− INT	0.4 ± 0.15	34.81 ± 6.57	11.35 ± 6.73	46.02 ± 4.77
WT 1w SCI	1.80 ± 0.30	14.96 ± 4.41 **** vs. WT INT	22.56 ± 6.16	36.98 ± 3.54
tph1−/− 1w SCI	1.34 ± 0.36	10.81 ± 1.85 **** vs. tph1−/− INT	26.94 ± 3.6	37.64 ± 3.60
WT 4w SCI	3.65 ± 2.27 **** vs. WT INT # vs. WT 1w SCI	13.87 ± 2.22 **** vs. WT INT	48.42 ± 24.82 **** vs. WT INT ## vs. WT 1w SCI	57.11 ± 22.1 ## vs. WT 1w SCI
tph1−/− 4w SCI	1.96 ± 0.83 ** vs. tph1−/− INT $ vs. WT 4w SCI	14.15 ± 10.8 **** vs. tph1−/− INT	24.64 ± 0.85 $$ vs. WT 4w SCI	41.01 ± 12.66 $$ vs. WT 4w SCI

One-week post-injury, bladder function was severely disrupted, with signs of bladder dysfunction in both genotypes. Compared to WT INT animals, WT 1w SCI mice showed significantly decreased amplitude of bladder reflex contractions (Fig. 1C; H; Table 1; p < 0.001 vs. WT INT). The frequency of bladder reflex contractions presented an increased tendency, as well as the basal pressures. Peak pressures remained similar (Fig. 1C, G, I-J; Table 1). Likewise, tph1^−/− 1w SCI mice presented decreased amplitude of bladder contractions when compared to tph1^−/− INT animals (Fig. 1D-H; Table 1; p < 0.001 vs. tph1^−/− INT), as well as a tendency for increased frequency and basal pressures that did not reach statistical significance (Fig. 1D, G, I-J; Table 1).

Four weeks post-SCI, the bladder reflex activity of WT mice worsened. When compared to intact mice, the frequency of bladder contractions increased (Fig. 1E, G; Table 1; p < 0.001 vs. WT INT), as well as the basal (Fig. 1E, I; Table 1; p < 0.001 vs. WT INT) and peak pressures (Figure E, J; Table 1; p < 0.01 vs. WT 1w SCI), causing a decrease in the amplitude of bladder contractions (Fig. 1E, H; Table 1; p < 0.0001 vs. WT INT). At four weeks after SCI, tph1 ^−/− mice presented a similar bladder behaviour to that observed at 1 week after SCI, except for a significantly increased frequency of bladder contractions in comparison with spinal intact tph1 ^−/− animals (Fig. 1E-F, G; Table 1; p < 0.01 vs. tph1 ^−/− INT). When compared to WT counterparts at the same time-point, tph1^−/− 4w SCI animals presented a significantly decreased frequency of bladder contractions (Fig. 1E-G; Table 1; p < 0.05 vs. WT INT), basal (Fig. 1E-F, I; Table 1; p < 0.01 vs. WT 4w SCI) and peak pressures (Fig. 1E-F, J; Table 1; p < 0.01 vs. WT 4w SCI), suggesting an influence of 5-HT absence on NDO phenotype 4 weeks post-injury.

Development of SCI-induced urinary dysfunction courses with upregulation of urethral serotonergic paraneurons

To study the presence of serotonergic paraneurons in the urethral lining of WT and tph1^−/− mice following SCI, an immunofluorescence staining against 5-HT was performed. In WT mice, immunoreactive cells were slender and elongated in shape, located in the deeper layers of the epithelium and exhibiting thin processes towards the luminal surface or lamina propria (Fig. 2A-C). The number of 5-HT⁺ cells increased in response to SCI in a time-dependent manner. This increase was already apparent one week after injury (WT 1w SCI), but only reached statistical significance four weeks later, when the number of 5-HT⁺ cells tripled in comparison to intact WT mice (Fig. 2B; p < 0.01 vs. WT INT). These serotonergic cells were located near sensory peptidergic neuronal fibres, labelled with anti-calcitonin gene-related peptide (CGRP) antibody (Fig. 2D-F). No 5-HT⁺ cells were detected in urethral sections from tph1^−/− mice, whether they were spinal intact or submitted to SCI (Fig. 2G-I).

Fig. 2.

Effects of SCI on the peripheral and central 5-HT production. The presence of serotonin (5-HT) positive cells in the urethral epithelium was assessed by immunostaining against 5-HT. Central concentration of 5-HT was measured by HPLC in the lumbosacral spinal cord (L5-S1 spinal cord segments). (A-C) Representative images of 5-HT staining in WT (A) INT, (B) 1w SCI and (C) 4w SCI animals. (D-F) Magnification of urethral double-immunostaining against 5-HT and calcitonin gene-related peptide (CGRP) (represented by dashed boxes in A-C) in WT (D) INT, (E) 1w SCI and (F) 4w SCI, showing the proximity of 5-HT⁺ cells to sensory peptidergic fibres; (G-I) Immunostaining against 5-HT in the urethral epithelium of tph1^−/− (G) INT, (H) 1w SCI and (I) 4w SCI mice confirming the absence of 5-HT⁺ cells in tph1^−/− mice. (J) Quantification of 5-HT⁺ cells in WT animals, showing a significant increase in the number of 5-HT-stained paraneurons at 4w SCI. (K) HPLC quantification of 5-HT concentration in the lumbosacral spinal cord in WT and tph1^−/− INT, 1w SCI and 4 w SCI animals showed a significant decrease of this neurotransmitter 1 and 4 weeks after injury. Graphs represent mean ± standard deviation, and p < 0.05 was considered statistically significant. **p < 0.01 vs. WT INT; ****p < 0.0001 vs. WT INT; ****p < 0.0001 vs. tph1^−/− INT. One (graph J) or two-way ANOVA (graph K) followed by Tukey´s post-hoc multiple comparison test. Scale bars correspond to 50 μm.

Central expression of 5-HT was determined by HPLC in the lumbosacral spinal cord (L5-S1 spinal cord segments). After SCI, 5-HT levels significantly decreased in both WT and tph1^−/− mice, both at 1 week and 4 weeks post-SCI, demonstrating the disruption of central 5-HT pathways induced by SCI (Fig. 2K; p < 0.0001 vs. INT for both genotypes).

Alterations in urethral innervation in response to SCI are dependent on peripheral 5-HT synthesis

As before²⁷, the distribution of sensory and cholinergic fibres, in the vicinity of cell processes of serotonergic urethral paraneurons, was investigated by immunohistochemistry in the urethra. Sensory innervation was assessed by CGRP immunostaining, an established marker of peptidergic sensory fibres. In the presence of peripheral 5-HT (WT mice), SCI induced a time-dependent increase in CGRP expression in the internal urethral sphincter (IUS) region, particularly four weeks after injury (WT 4w SCI) (Fig. 3A, C; p < 0.01 vs. WT INT). No time-dependent alterations in CGRP expression were detected in the external urethral sphincter (EUS) (Fig. 3A, D). In the absence of peripheral 5-HT expression (tph1^−/− mice), no changes were detected in the urethral (IUS and EUS) sensory innervation after SCI, irrespective of the timepoint after spinal trauma (Fig. 3B-D).

Fig. 3.

Influence of peripheral serotonin on urethral tissue rearrangement after SCI. (A-D) Sensory innervation assessment by calcitonin gene-related peptide (CGRP) staining in the proximal urethra. Representative images of (A) WT 4w SCI and (B) tph1^−/− 4w SCI mice CGRP staining. Densitometry quantification of sections of the (C) internal urethral sphincter (IUS) showed significantly decreased CGRP⁺ fibres 4 weeks after injury in tph1^−/− mice, compared to WT animals. No time-dependent changes in CGRP⁺ fibres were seen in the (D) external urethral sphincter (EUS), independently of the genotype. (E-H) Parasympathetic innervation assessment by vesicular transporter of acetylcholine (VAChT) staining in the proximal urethra. Representative images of (E) WT 4w SCI and (F) tph1^−/− 4w SCI mice VAChT staining. Quantification in the (G) IUS showed decreased VAChT sprouting 4 weeks after injury in tph1^−/−mice, compared to WT at the same timepoint. No time-dependent changes in VAChT⁺ fibres were seen in the (H) external urethral sphincter (EUS), independently of the genotype. (I-K) Smooth muscle cells’ integrity assessment by smooth muscle actin (SMA) immunoreaction in the urethra. Representative images of (I) WT 4w SCI and (J) tph1^−/− 4w SCI mice SMA staining. (K) Quantification showed significantly decreased SMA expression 4 weeks after SCI, regardless of the genotype. (L-N) Collagen deposition in EUS cells was assessed by Sirius Red (SR) Staining. Representative images of (L) WT 4w SCI and (N) tph1^−/− 4w SCI mice SR staining. (N) In WT animals, collagen areas remained unchanged by SCI progression, but in the absence of peripheral 5-HT, EUS fibrosis tended to increase in a time-dependent manner, reaching statistical significance at 4w SCI when compared to intact. Graphs represent mean ± standard deviation, and p < 0.05 was considered statistically significant. *p < 0.05 vs. INT WT; **p < 0.01 vs. INT WT; ##p < 0.01 vs. WT 1w SCI; *p < 0.05 vs. tph1^−/− INT; ###p < 0.001 vs. tph1^−/− 1w SCI; $p < 0.05 vs. WT 4w SCI; $$$p < 0.001 vs. WT 4w SCI. Two-way ANOVA followed by Tukey´s post-hoc multiple comparison test. Scale bars correspond to 50 μm.

Parasympathetic innervation was assessed by immunostaining urethral sections against vesicular transporter of acetylcholine (VAChT), an established marker of cholinergic parasympathetic fibres. VAChT⁺ fibres were significantly increased in WT mice submitted to SCI at 4 weeks after injury (WT 4w SCI). This was evident at the IUS but not at the EUS region (Fig. 3E, G-H; p < 0.01 vs. WT INT; p < 0.01 vs. WT 1w SCI). In the absence of peripheral 5-HT expression (tph1^−/− mice), no time-dependent changes in the cholinergic innervation of the IUS and EUS after SCI were observed.

SCI-induced smooth muscle atrophy of the IUS does not depend on peripheral 5-HT

The integrity of the IUS, known to be affected after SCI³⁰, was evaluated by immunostaining of smooth muscle actin (SMA). The abundant expression of SMA found in intact animals significantly decreased four weeks after SCI both in WT (Fig. 3I, K; p < 0.05 vs. WT INT; p < 0.01 vs. WT 1w SCI) and tph1^−/− mice (Fig. 3J, K; p < 0.05 vs. tph1^−/− INT; p < 0.001 vs. tph1^−/− 1w SCI), demonstrating signs of IUS atrophy independent of peripheral 5-HT.

Striated muscle fibrosis is dependent on peripheral 5-HT synthesis

The collagen content of the EUS was measured by Sirius-Red staining to assess fibrosis. In WT animals, collagen areas remained unchanged by SCI progression, with a slight but not significant decrease at 1w SCI, which returned to baseline values at 4w SCI (Fig. 3L, N; p < 0.01 vs. WT 1w SCI). On the other hand, tph1^−/− mice displayed a time-dependent increase in EUS collagen content. At 4w SCI, collagen levels were significantly higher in tph1^−/− mice compared to intact counterparts (Fig. 3L-M; p < 0.05 vs. INT tph1^−/−). Although not statistically significant, EUS collagen content in tph1^−/− 4w SCI appeared to increase, compared to WT animals at the same timepoint.

Treatment with 5-HT receptor antagonists altered urinary function but did not prevent SCI-induced NDO emergence in WT mice

To test whether modulation of urethral 5-HT signalling could improve urinary function following spinal injury, SCI WT mice were daily treated with Ritanserin and Ondansetron, antagonists of 5-HT₂ and 5-HT₃ receptors, respectively, and known to have an effect on bladder activity^24,27. Ritanserin and Ondansetron were intraperitoneally administered separately (1 mg/kg) or in combination (1 mg/kg + 1 mg/kg), and an additional group of animals received the vehicle solution (absolute ethanol + H₂O). Animals were submitted to daily removal of urine by abdominal compression until the end of the 28-day protocol and the volume of urine was recorded every 3 days. This record showed persistently high urine volumes in all experimental groups (Fig. 4A). Ten days post-SCI, all animal groups presented a reduction in urine volumes, slightly more evident in vehicle-treated animals, followed by an increase at later time-points. At day 22, Ritanserin-treated animals showed a drop in urine volume and ended the protocol, at 28 days post-SCI, with the lowest volume of retained urine. At day 22, Ondansetron- and Ritanserin + Ondansetron-treated animals appeared to be able to retain higher urine volumes compared to vehicle- and Ritanserin-treated mice. However, no statistically significant differences were seen between treatments at any time point (Fig. 4A). Importantly, as animals were daily assessed, it was observed that in animals treated with 5-HT receptor antagonists, some urine samples were cloudy and, in repeated occasions, it was necessary to perform anaesthetised catheterisation for urine removal in Ritanserin-treated mice. In the remaining treated groups, LUT infections were less frequent and tended to resolve quickly. In any case suspect of urinary tract infection and following in-house veterinary advice, animals received antibiotics until urine samples returned to normal.

Fig. 4.

Effects of serotonin receptor antagonist treatment in WT 4w SCI mice. (A) Measurement of urine volume retention in 4w SCI mice receiving vehicle, Ritanserin (RT), Ondansetron (ON) or Ritanserin + Ondansetron (RT + ON). Urine volumes were measured every three days during daily manual bladder emptying. Ritanserin treatment presented a tendency for decreased urinary retention around day 20-post SCI with no statistical significance. Two-way ANOVA repeated measures analysis followed by Tukey´s post-hoc comparison test. (B-E) Representative cystometrograms depicting bladder contractility in mice submitted to SCI and daily-treated with (B) Vehicle (ethanol + H₂O), (C) Ritanserin (1 mg/Kg), (D) Ondansetron (1 mg/Kg) or (E) Ritanserin (1 mg/Kg) combined with Ondansetron (1 mg/Kg), for 28 days. (F-I) Analysis of urodynamic parameters of (F) frequency of expulsive bladder contractions, (G) amplitude of bladder contractions, (H) basal pressure, and (I) peak pressure. Ritanserin administration alone induced significantly increased amplitudes of voiding contractions, compared to the other groups. One-way ANOVA followed by Tukey´s post-hoc comparison test (**p < 0.01 vs. vehicle, ON and RT + ON).

After 4 weeks of daily pharmacological treatment, animals were anaesthetized with urethane and underwent cystometries to evaluate bladder function. The frequency and peak pressure of bladder reflex contractions were similar between groups, although the lowest basal pressure was recorded in animals receiving Ritanserin, albeit without statistical significance (Fig. 4B-I). Urodynamic recordings showed the amplitude of bladder contractions significantly increased in Ritaserin-treated mice, compared to other treated groups (Fig. 4B-E, G; p < 0.01 vs. vehicle; p < 0.01 vs. ON; p < 0.05 vs. RT + ON).

Discussion

The present study focused on the importance of urethral 5-HT in bladder function after SCI. Experiments were performed in a mouse model of complete spinal cord transection at T8/T9 level, and the use of tph1^−/− mice, which lack peripheral serotonin, allowed for a deeper understanding of the role of urethral 5-HT in LUT function.

Our results show that the bladder function of all animals, irrespective of the genotype, was affected by SCI. One week post-injury, both WT and tph1^−/− mice exhibited decreased amplitude of bladder reflex contractions, accompanied by a slight, but non-significant increase in their frequency. Four weeks after injury, bladder dysfunction in WT animals was further exacerbated, with increased frequency, basal and peak pressure of bladder contractions, compared to spinal intact WT mice. In tph1 ^−/− SCI mice, bladder function was less affected. Compared to their WT 4w SCI counterparts, tph1^−/− mice presented decreased frequency of bladder contraction and basal and peak pressures, suggesting a less intense urinary dysfunction after the spinal injury became chronic. One can hypothesize that the lack of peripheral 5-HT attenuated NDO development, indicating this monoamine is involved in NDO pathophysiology, likely by modulating cholinergic and sensory fibre maladaptive sprouting. Accordingly, it has been demonstrated that 5-HT has an excitatory action on both cholinergic and sensory neurons, which amplify bladder contractility by increasing afferent drive^31–34. As cystometric observations suggest a role for 5-HT in SCI-induced bladder dysfunction, we assessed the serotonergic population of urethral cells lining the lumen. As before, 5-HT positive cells were located in the deeper layers of the urethral epithelium, in proximity to peptidergic fibres present in the lamina propria, in parallel with what was described by Coelho and colleagues in the rat²⁷. Urethral 5-HT⁺ cells were increased in WT mice 4 weeks post-SCI. While we were not able to quantify 5-HT concentration in the urethra by HPLC due to technical issues, this likely reflects increased levels of this neurotransmitter, accompanying the development of NDO-like features found in cystometries. High 5-HT levels likely resulted in increased excitability of bladder afferents, which is considered a key mechanism in SCI-induced urinary dysfunction^35,36. Heightened bladder reflex activity in SCI WT animals may be mediated via activation of 5-HT receptors^31–34, as described in animal models of interstitial cystitis/bladder pain syndrome While we tried to evaluated the expression levels of 5-HT receptors in the urethra by RT-qPCR, the limited tissue yields obtained did not allow it. This was likely due to low concentration and poor-quality RNA.

Peripheral 5-HT corresponds to 98% of total body 5-HT, with its majority being produced by enterochromaffin cells in the intestinal mucosa³⁷. In non-pathological conditions, peripheral 5-HT plays a minor role in peristalsis and motility, with a prominent intervention of central-derived 5-HT³⁸. Only under inflammatory conditions does peripheral 5-HT assume a more prominent role, promoting a cascade of events that culminate in exaggerated gastrointestinal activity and inflammatory symptoms^39,40. One could speculate that a similar response may occur in the LUT, as intact animals seemed to have little dependency on peripheral 5-HT to maintain efficient urinary function. After SCI, as descending supraspinal tracts were severed, lumbosacral 5-HT was markedly reduced. This could have triggered a compensatory mechanism that increased peripheral 5-HT synthesis, ultimately leading to the hyperexcitability of urethral afferents. In the absence of supraspinal input, this results in uninhibited bladder contractions and urine leakage that further contribute to Neurogenic Detrusor Overactivity (NDO). This mechanism likely operates in a positive feedback loop, where each episode of urine leakage further activates urethral paraneurons, amplifies local 5-HT release, and in turn promotes additional incontinence events (Fig. 5).

Fig. 5.

Schematic representation of the proposed mechanism beyond the urethral 5-HT signalling loop after SCI. Under physiological conditions, urine flow through the urethra activates serotonergic paraneurons located in the urethral mucosa, inducing serotonergic local release. This urethral serotonin activates 5-HT₂ and 5-HT₃ receptors expressed by urethral sensory nerves, causing their excitation. The input generated reaches supraspinal centres via the lumbosacral spinal cord, and a descending efferent input is generated towards the LUT, resulting in enhanced voiding efficiency.

Following spinal cord injury, supraspinal modulation of lower urinary tract function. The number of urethral serotonergic paraneurons embedded in the urethral lining increases, leading to exaggerated local 5-HT release, increased afferent excitability and heightened sensory input. Once reaching the lumbosacral spinal cord, the sensory input is integrated into a spinally confined reflex that emerges through maladaptive neuroplasticity after SCI. Without supraspinal involvement, this leads to strong uninhibited bladder contractions - Neurogenic Detrusor Overactivity (NDO)- and urine leakage. This mechanism likely operates in a positive feedback manner, in which urine leakage episodes further stimulate urethral paraneurons and perpetuate excess 5-HT release and recurrent incontinence episodes.

The influence of peripheral 5-HT in sensory and cholinergic fibre rearrangement in the urethra after SCI was also studied by immunohistochemistry. Contrary to WT mice, which exhibited increased urethral expression of sensory (CGRP) and cholinergic (VAChT) markers in the IUS after SCI, levels of CGRP and VAChT in tph1^−/− 4w SCI mice were similar to those observed in uninjured mice, indicating that the absence of peripheral 5-HT blocks SCI-induced expansion of sensory and cholinergic innervation in the IUS. As SCI leads to increased 5-HT expression in the urethral epithelium of WT mice, it may be possible that high local 5-HT levels could have a direct effect on sensory and cholinergic fibres, promoting CGRP and VAChT upregulation, as the lack of peripheral 5-HT prevented such SCI-induced upregulation. While a direct effect of 5-HT on the expression of these neuronal markers is yet to be demonstrated, some studies have suggested that 5-HT can induce direct upregulation of cholinergic and sensory activity^41,42. Therefore, the absence of the monoamine in tph1-deficient SCI mice attenuates cholinergic and sensory markers and the responsiveness of these nerve fibres, potentially impairing neurotransmitter release.

Of note, increased IUS levels of CGRP and VAChT expression in WT 4w SCI mice contrasts with our previous observations in the female rat, where SCI coursed with denervation of the same fibres³⁰. This is probably related to differences in injury progression between rats and mice. Indeed, our urodynamic data suggest discrepancies between mouse and rat bladder reflex activity, as mice presented signs of bladder contractions at 1 week post-injury, in contrast with the SCI rat, whose bladder is mostly areflexic at this time point^30,43. These differences may reflect species-related differences in LUT innervation and function and possibly result in different time courses of functional recovery in response to SCI. Rats have a strong adrenergic tone supplying the IUS, maintaining a tighter urethral closure during urine storage⁴⁴. During voiding, the EUS presents bursting-like behavior, resulting in pulsatile expulsive flow⁴⁵, due to a robust somatic innervation via the pudendal nerve^45,46. In mice, on the other hand, voiding does not require EUS pumping and occurs through prolonged urethral relaxation between periods of high-tonic EUS activity during voiding bladder contractions^45,47. While there is no data comparing EUS innervation in both species, one could speculate that the sympathetic and motor neurons supplying the urethral sphincter may be weaker in mice. After SCI in mice, this EUS relaxation period is not long enough to achieve efficient voiding⁴⁵, resulting in intermittent and inefficient voiding leading to high amounts of urinary retention and the necessity to perform abdominal compression for urine removal well beyond the period of spinal shock⁴⁵.

Interestingly, while there was an expansion of sensory and cholinergic innervation in the IUS of WT but not in transgenic animals, levels of SMA were similarly decreased, suggesting smooth muscle atrophy in both genotypes. This suggests that urothelial 5-HT may be the main driver behind increased CGRP and VAChT upregulation in the IUS instead of mediators produced by smooth muscle cells, including nerve growth factor, which has a positive effect on sensory fibres⁴⁸. While poorly explored in the context of SCI-induced urinary impairment, IUS atrophy probably has a facilitatory role in NDO emergence. In healthy conditions, the IUS provides tonic sympathetic tone that maintains urethral resistance and sphincter closure during urine storage^1,49. Smooth muscle atrophy likely weakens urethral tone and leads to increased frequency of leaking episodes. As the urine passage through the urethra is known to increase urethral afferent excitability and initiate a positive feedback leading to enhanced bladder contractility^12,25, IUS atrophy could function as the first step of a positive loop that further increases bladder contractility and contributes to NDO. It is also plausible that hyperexcitability of urethral afferents further sensitize NDO-driven C-fibers contributing to its maintenance. IUS atrophy can also explain the emergence of DSD, as the loss of outlet resistance and impaired sensory input can desynchronize bladder and sphincter contraction, promoting simultaneous detrusor contractions against an inadequately relaxing sphincter during voiding.

Contrary to the rat, where SCI coursed with EUS fibrosis³⁰, no evidence of increased collagen content was seen in the EUS of WT 4w SCI mice. Nevertheless, this was seen in their tph1 ^−/− counterparts, suggesting a pro-fibrotic outcome in the absence of peripheral 5-HT. This is in contrast with other studies demonstrating decreased collagen deposition in models of fibrotic syndromes in the lung, liver, and skin in spinal intact tph1^−/− mice, suggesting a role of the peripheral monoamine in collagen deposition and wound healing^50–53. The reason for this discrepancy can only be speculated at present, but may reflect an SCI-induced peripheral immune dysregulation⁵⁴ that also occurs in the bladder⁵⁵. As 5-HT regulates immune responses⁵⁶, an unbalanced fibrotic response may have taken place in the IUS of tph1 ^−/− SCI mice.

Since the absence of peripheral 5-HT in tph1^−/− 4w SCI led to improved bladder function, we investigated whether pharmacological blockade of 5-HT receptors involved in urethral serotonergic signalling^24,27 could attenuate NDO symptoms in WT SCI mice. Potent and long-lasting antagonists of 5-HT₂ (Ritanserin) and 5-HT₃ (Ondansetron Hydrochloride) receptors were intraperitoneally delivered once a day for 28 days, starting on injury day. The blockade of 5-HT₂ receptor with Ritanserin lowered basal pressure and increased the amplitude of bladder contractions, potentially reflecting improved storage function after treatment. As the administration of 5-HT to the urethra is known to induce strong bladder contractions²⁷, blockade of the 5-HT₂ receptor may have reduced bladder reflex activity. In addition, while we were not able to demonstrate the effects on the activity of the sphincter, it is possible that Ritanserin could improve storage by impacting sphincter activity and promoting continence, pending on further studies to titrate the most appropriate dosage of Ritanserin. For clinical translation, one might consider a combination of Ritanserin, to make the bladder a more effective reservoir, with clean intermittent catheterization to promote timely and periodic urine removal would mitigate risks of urinary retention and infections. Accordingly, some mice presented suggestive signs of potential urinary infections, and, upon veterinary advice, animals were maintained on antibiotic treatment.

We also tested the effects of a 5-HT₃ antagonist, Ondansetron Hydrochloride, as 5-HT₃ receptors are known to be upregulated after SCI^57,58. Surprisingly, Ondansetron administration did not affect bladder function as previously seen in rats²⁷, with no alterations in urodynamic parameters or urine residual volumes, compared to vehicle-treated SCI mice. Interestingly, the combined treatment with Ondansetron and Ritanserin did not change bladder function in SCI mice, suggesting that 5-HT₃ blockade counteracted the beneficial effects of Ritanserin. Together, these results may indicate that in our SCI mouse model, 5-HT₂, but not 5-HT₃, may be an interesting target to modulate bladder reflex activity after SCI. However, it should be recalled that drugs were delivered intraperitoneally, making it difficult to discern the main site of action.

An important limitation of the present study is that tph1^−/− mice have a global deficiency of tph1. Hence, peripheral 5-HT was absent from all peripheral tissues, and not only from the urethra. While it is not possible to fully exclude the possibility that the observed results could reflect systemic changes, one should recall that mediators produced by paraneurons primarily have a local paracrine action, acting on the nearby cells^26,59. Our data demonstrating the presence of 5-HT–producing paraneurons in WT animals, which are absent in tph1^/ mice, suggests that urethral 5-HT is synthesized locally rather than supplied by enteric or blood-circulating 5-HT. In addition, past observations showed that the systemic depletion of 5-HT in non-injured tph1^/ immediately induced an increase in bladder efficiency when serotonin was applied locally into the urethral lumen²⁷. This further supports the idea that local availability of 5-HT at urethral sensory sites, rather than systemic peripheral 5-HT, is the critical determinant of urethrovesical reflex function.

Conclusions

This study builds on prior work on urethral involvement in SCI-urinary dysfunction and on the importance of the 5-HT urethrovesical reflex. We found that SCI induced urinary impairment in WT animals that was partially prevented in tph1^−/− mice. In wild-type mice, bladder dysfunction was accompanied by a time-dependent increase in serotonergic paraneurons in the urethral epithelium, with expansion of sensory and cholinergic nerve fibres in the sphincter. The upregulation of CGRP and VAChT-positive fibres was absent in tph1^−/− mice. In both genotypes, there was muscle atrophy of the IUS, but EUS fibrosis was only present in tph1^−/− SCI animals. Blocking the 5-HT₂ but not the 5-HT₃ serotonin receptor improved bladder reflex activity. These data suggest that modulation of peripheral 5-HT may be used as a future therapeutic tool for NDO management.

Materials and methods

Animals and drugs

In-house-bred female C57BL/6 mice (aged 4–5 months, weighing 20–25 g) were maintained under a 12 h light/dark schedule and controlled temperature and air humidity, with ad libitum access to food and water. Wild-type (WT) and homozygous Tph1-deficient mice (tph1^−/−) were divided into three groups for each genotype: intact (INT), left to recover one week (1w SCI) or four weeks after spinal cord transection (4w SCI) (n = 4/5 animals per group). Tph1^−/−mice were kindly provided by Prof. Michael Bader (Max-Delbrück-Center for Molecular Medicine, Berlin-Buch, Germany). Tph1 ablation was achieved through homologous recombination by the insertion of selection cassettes (neomycin and thymidine kinase) into the endogenous locus, yielding an isoform-specific disruption that abolishes peripheral TPH1 expression⁶⁰. Additional groups of WT female mice were submitted to SCI and treated for 4 weeks with antagonists of 5-HT receptors: Ritanserin (1 mg/kg), Ondansetron (1 mg/kg), or a combination of the two (n=−4-5 per group). Experimental procedures were carried out following the European Communities Council Directive 2010/63/EU, and institutional regulations from the Faculty of Medicine of Porto, Portugal (Ethical approval reference ORBEA_121 2022/2 709; 27 September 2022). The ARRIVE guidelines have been carefully followed during the course of this study.

Induction of spinal cord injury

Animals were submitted to a laminectomy between the T7-T10 vertebrae under deep anaesthesia (medetomidine (1 mg/Kg) and ketamine (75 mg/Kg) by intraperitoneal (IP) injection). The exposed spinal cord (T8/T9) was completely sectioned with a scalpel, and a small piece of sterile haemostatic sponge was placed between the retracted ends of the cord. The surgical wound was closed in two layers and anaesthesia was reverted (atipamezole (1 mg/Kg) by IP injection). After surgeries, animals were placed under post-surgical observation and received 5% saline-glucose (0.5 ml, subcutaneous) to compensate for blood loss and dehydration. During post-operative care, they received subcutaneous antibiotics (enrofloxacin; 5 mg/Kg) and buprenorphine (0.05 mg/Kg, twice daily). To avoid urinary retention, urine was manually drained by abdominal compression during the entire extension of the protocol.

Cystometry under urethane anaesthesia and euthanasia

Bladder reflex activity was evaluated by cystometry under anaesthesia before euthanasia in all animal groups. Following deep anaesthesia with subcutaneous urethane (1.2 g/Kg), a suprapubic skin incision was made, and muscle bundles were separated for bladder exposure. A 23-gauge syringe needle, attached to a polyethylene tube linked to a pressure transducer, was inserted into the bladder dome, and sterile saline was infused for 45 min at a rate of 1.6 ml/h. Bladder contractions were recorded by a pressure transducer connected to the needle. Animals were maintained on a heating plate during the procedure to conserve body temperature at 37 °C. After bladder recordings, animals received an IP injection of sodium pentobarbital (65 mg/g) for euthanasia. Experimental groups were divided and processed for the collection of fresh (n = 4–5) or fixed tissue (n = 4–6).

Tissue perfusion and immunohistochemistry

After cystometries, animals were terminally anaesthetized and the lumbosacral spinal cord was collected and immediately frozen at −80 °C. The urethras were dissected, fixed by immersion in 4% paraformaldehyde (PFA) for 6 h, and cryoprotected in 30% sucrose in phosphate buffer with 0.1% sodium azide for at least 24 h. Collected urethras were cut into 12 μm-thick longitudinal sections in a Leica cryostat and collected in Superfrost Plus slides. Slides were stored at −20 °C until further processing.

Alternate urethral sections were thawed, washed in phosphate-buffered saline (PBS) and in PBS containing 0.3% Triton 100 (PBST), and subsequently blocked with 10% normal horse serum (NHS) in PBST for 2 h. Tissue sections were incubated for 72 h at 4 °C with primary antibodies (Table 2) in 2% NHS in PBST. After several washes with PBST, sections were incubated with suitable Alexa™ fluorochrome-labelled secondary antibody in 2% PBST (Invitrogen - ThermoFisher Scientific, Porto, Portugal) for 1 h at room temperature (Table 2). After subsequent washing, sections were mounted using an anti-fade mounting medium (Slowfade^® Gold Life Technologies) and observed with an epifluorescence microscope (Axioimager Z1, Ziss Z1 from Zeiss) using the AxioVision 4.6 software.

Table 2.

Primary and secondary antibodies used in immunohistochemistry.

Primary antibodies	Dilution	Host species	Manufacturer	Catalog #
SMA	1:1000	rabbit	Abcam	Ab124964
CGRP	1:2000	sheep	Enzo Life Sciences	BML-CA1137-0100
VAChT	1:1000	rabbit	Synaptic Systems	sysy139103
5-HT	1: 5000	rabbit	Cell Signalling	ABIN61783

Secondary antibodies	Dilution	Host species	Manufacturer	Catalog #
Rabbit/Alexa 488	1: 1000	Donkey	ThermoFisher	A21206
Sheep/Alexa 568	1: 1000	Goat	ThermoFisher	A21099

HPLC

Quantification of 5-HT in the L5-S1 spinal cord was performed by HPLC, as previously described²⁷. Stored fresh tissue at −80 °C was immersed in ice-cold perchloric acid for 2 h before analysis. Samples were filtered on Costar Spin-X microfilter tubes, and 50 µL of the eluate was injected into an HPLC. The lower limit for detection was 500 fmol.

Sirius red staining

Frozen urethral sections were left to dry at room temperature for 1 h before staining with Picro-sirius red solution for 90 min. The sections were then washed 2 times with 0,5% acidified water and dehydrated in three changes of 100% ethanol and two benzene baths. Sections were mounted with Entellan definitive mounting medium.

Administration of Pharmacological antagonists of 5-HT receptors (5-HT₂ and 5-HT₃)

To test whether pharmacological modulation of urethral 5-HT signalling pathways could result in ameliorated urinary function, Ritanserin (Cat. No. 1955. Tocris Bristol, UK), a potent, non-selective and long-lasting 5-HT₂ receptor antagonist, and Ondansetron hydrochloride (Cat. No. 2891. Tocris Bristol, UK), a selective 5-HT₃ receptor antagonist, were administered to additional groups of WT mice. Per the manufacturer’s instructions, drugs were reconstituted in ethanol (Ritanserin) and distilled H₂O (Ondansetron Hydrochloride) and stock solutions were stored at −80 °C. Female WT mice underwent spinal cord transection surgeries at the T8/T9 spinal level, as described above, and were divided into four experimental groups: (1) Ritanserin (RT) (1 mg/Kg/day); (2) Ondansetron (ON) (1 mg/Kg/day); (3) Ritanserin + Ondansetron (RT + ON) (1 mg/Kg/day + 1 mg/Kg/day); and (4) Vehicle (ethanol + H₂O in the same concentration as used for drugs dilutions). Aliquots of Ritanserin and Ondansetron’s stock solutions were diluted in saline and delivered via intraperitoneal injection, starting on surgery day and spanning 4 weeks post-SCI. Drugs and dosages were chosen according to previous studies^24,27. Retained urine volumes were recorded every three days upon daily bladder compression in all groups, since SCI induction and throughout the protocol. After 4 weeks of treatment, all animals underwent cystometries as described in 5.3.

Data analysis

Cystometrograms were analysed using LabScribe software (World Precision Instruments, Hertfordshire, UK). Animals suffering from bladder overflow induced by anaesthesia (identified by the loss of bladder contractility in the presence of high intravesical pressures exceeding the typical threshold pressures for reflex voiding (superior to 40 cm H₂0) and a progressive increase of these pressures without any voiding episode (up to 50 cm H₂0), were immediately euthanized and excluded of the analysis to ensure that only reflex-driven voiding cycles were considered.

For immunohistochemistry quantifications, three to five longitudinal urethral sections per animal were considered. For 5-HT staining, the number of positive cells in all the sections was quantified. For CGRP and VAChT staining, results were quantified by densitometry using in Fiji, as before³⁰. The results are presented as an average of the values measured in the posterior (closer to the vagina) and anterior side of the urethral wall. Data was collected near the sphincter region of the proximal urethra. Fibrosis levels were quantified as the area of collagen red-stained fibres compared to the total tissue area. All measurements were performed on raw files. For exemplificative images, small adjustments in luminosity and/or contrast might have been made. Data was statistically analysed using one or two-way ANOVA followed by Tukey’s multiple comparison post hoc test using GraphPad Prism 9 software. Results are represented as mean ± SD, and p < 0.05 was considered statistically significant.

Author contributions

AF: Conceptualisation, methodology, formal analysis, original draft, writing, and editingSSC, AA: MethodologyCR: Methodology, formal analysis, and editingCDR: Conceptualisation, methodology, formal analysis, writing, and editingAll authors read and approved the final manuscript.

Data availability

Data will be made available upon reasonable request to the corresponding author.

Declarations

Competing interests

The authors declare no competing interests.