Determining integrity of bladder innervation and smooth muscle function one year after lower spinal root transection in canines

Abstract

Aim:

To assess bladder smooth muscle function and innervation after long-term lower spinal root transection in canines.

Methods:

Thirteen female mixed-breed hound dogs underwent bladder decentralization, which included transection of all sacral dorsal and ventral roots caudal to L7 and hypogastric nerves, bilaterally (n=3); all sacral roots and hypogastric nerves plus transection of L7 dorsal roots, bilaterally (n=4); or a sham operation (n=6). At a year after initial surgery, bladder function was assessed in vivo by stimulation of the pelvic plexus. The bladder was harvested for ex vivo smooth muscle contractility studies. Remaining bladder was evaluated for nerve morphology immunohistochemically using neuronal marker PGP9.5, apoptotic activity using terminal deoxynucleotidyl transferase dUTP nick end labeling, and histopathology using a hematoxylin and eosin stain.

Results:

Sacral root decentralization did not reduce maximum strength of pelvic plexus stimulation-induced bladder contraction, although long-term sacral dorsal and ventral root plus L7 dorsal root transection significantly decreased contraction strength. Electric field stimulation-induced contractions of the detrusor from all decentralized animals were preserved, compared to controls. Viable nerves and intramural ganglia were visualized in the bladder wall, regardless of group. There was no difference in amount of apoptosis in bladder smooth muscle between groups.

Conclusion:

Bladder smooth muscle cells maintain their function after long-term bladder decentralization. While pelvic plexus-induced bladder contractions were less robust at one year after lower spinal root transection, the absence of atrophy and preservation of at least some nerve activity may allow for successful surgical reinnervation after long-term injury.

INTRODUCTION

Neurogenic bladder is a disorder of the lower urinary tract caused by the disruption of the neurocircuitry that regulates bladder function.¹ If left untreated, it can lead to complications such as bladder and kidney stones, chronic urinary tract infections, and depression due to the social consequences of incontinence.² Urodynamic dysfunction frequently occurs in patients with spinal cord injury³ or neurological disorders, such as spina bifida⁴ and multiple sclerosis⁵. A 2012 survey showed that patients who sustain spinal cord injuries prioritize recovery of bladder function over other faculties.⁶ The most common type of neurogenic bladder in patients with sacral spinal cord dysfunction is a lower motor neuron neurogenic bladder, which is characterized by detrusor areflexia and external urethral sphincter (EUS) denervation.² As a consequence, patients often experience urine stasis that requires frequent clean intermittent catheterization.⁷ Given the undesirable ramifications of urinary incontinence, finding ways to combat the disabilities that result from lower spinal cord dysfunction should be of utmost importance to the scientific research community.

Our lab ultimately aims to develop an approach to surgically reinnervate the bladder for treatment of lower motor neuron lesion-induced bladder dysfunction in a canine model; however, we must first establish the impact that the induced long-term injury has on the integrity of the bladder. While skeletal muscle degeneration following nerve injury has been well investigated, comparably little is known about the effects of nerve root injury on smooth muscle. Confirmation of bladder smooth muscle function after long-term neuronal injury is critical to the success of surgical reinnervation due to the likelihood that patients would delay undergoing an invasive surgery until all non-surgical therapies are exhausted. In this study, we utilize a long-term lower spinal root injury canine model to study the impact of sacral decentralization, or a marked decrease in neuronal input, on bladder smooth muscle and intramural neuronal function.

MATERIALS AND METHODS

Animals

Studies were conducted with Temple University IACUC approval and in compliance with NIH, USDA and AAALAC guidelines. Thirteen female mixed-breed hound dogs, acquired at 6–8 months of age weighing 18–25 kg (Marshall BioResources, North Rose, NY) were used in this study. Dogs were housed in groups of three and exposed to a 12-hr light/dark cycle. Animals were randomly assigned to a 12-month survival after surgical sacral root transection (n=3), 12-month survival after surgical sacral dorsal and ventral root plus L7 dorsal root transection (n=4), or a sham operation (n=6). Credé’s maneuver was performed on all experimental animals twice daily.

Surgical Spinal Root Injury

Immediately prior to surgery, animals received 20 mg/kg IV dose of cefazolin with re-dosing every four hours until completion of the procedure. Antibiotic prophylaxis included 30 mg/kg cephalexin PO twice a day for five days following the surgery. Animals received 6 mg/kg IV of propofol to allow for placement of endotracheal tube before placement on 0.5–4% mean alveolar concentration of isoflurane with oxygen.

Following initial anesthesia, catheterization was performed by passing a double balloon Foley catheter through the urethra and into the bladder. External pressure transducers were interfaced with the bladder lumen port to monitor intravesical pressure and the distal balloon to monitor urethral sphincter pressure. An additional catheter with balloons was placed into the rectum and measured both rectal and anal sphincter pressure, the former serving as a substitution for abdominal pressure. Normal saline cystometrograms were recorded with an infusion rate of 30 mL/min.

Animals underwent a laminectomy of L6-S3 vertebrae to expose the lower spinal cord and spinal roots. Roots were stimulated with a current of 1–3 mAmp, a frequency of 20-Hz, and a duration of 0.2 msec using a monopolar electrode to identify L7, S1, S2, and S3 roots prior to transection. As previously described^{8, 9} three animals underwent surgical transection of all dorsal and ventral sacral roots caudal to L7, bilaterally, and hypogastric nerves, bilaterally. The dorsal root of L7 was also transected, bilaterally, in addition to the sacral root and hypogastric nerve transections, in another a subset of animals (n=4). Five to ten mm were removed from each transected root or nerve to ensure complete separation. The loose ends of sacral roots were ligated with silk sutures. Hypogastric nerves were accessed via abdominal surgery. Sham-operated controls underwent lumbosacral laminectomy, nerve root identification via electrical stimulation without root transection, and abdominal opening with identification of hypogastric nerves. All animals underwent tail amputation at the end of the procedure to prevent self-mutilation of the now decentralized tail.

In vivo Bladder Functional Electrical Stimulation

Immediately prior to euthanasia one year after transection, nerves within the pelvic plexus of the bladder were stimulated with either a monopolar or bipolar electrode with a current of 0.5 mAmp-10 mAmp, a frequency of 20-Hz, and a duration of 0.2 msec. Bladders were stimulated bilaterally and the side that produced the largest contraction was reported. Bladder capacity was determined during three successive filling cystometrograms using normal saline at 30 mL/minute. The bladder was then fully emptied and filled again to about half of the bladder capacity. Changes in pressures were continuously recorded with external pressure transducers interfaced with the PowerLab® multichannel data acquisition system and LabChart® software (ADInstruments, Colorado Springs, CO). Strength of nerve-evoked bladder contractions after pelvic plexus stimulation were derived from differences between the resting baseline pressure and the peak pressure obtained during continuous stimulation.

Ex vivo Nerve-evoked Stimulation in Bladder Smooth Muscle

The dorsal aspect of the bladder was harvested at euthanasia and placed in HTK organ preservation media, composed of 15 mM NaCl, 9 mM KCl, 1 mM potassium hydrogen 2-ketoglutarate, 4 mM MgCl₂, 18 mM histidine NaCl, 2 mM tryptophan, 30 mM mannitol, and 0.015 mM CaCl₂ and kept on ice overnight. Strips of smooth muscle were isolated from bladders and denuded of the mucosal layers. They were then suspended in muscle baths between platinum plates located approximately 1 cm apart in 10 mL of oxygenated Tyrode’s solution, composed of 125 mM NaCl, 2.7 mM KCl, 0.4 mM NaH₂PO₄,1.8 mM CaCl₂, 0.5 mM MgCl₂, 23.8 mM NaHCO₃, and 5.6 mM glucose, kept at 37 °C, stretched to ~2 g tension, and allowed to accommodate to the bath for at least 30 min before testing. Strips were treated with iso-osmolar Tyrode’s solution with 120 mM potassium chloride (KCl) solution and the force of contraction that was generated was measured. Electric field stimulation (EFS) was delivered to the strips at 12 V, with a pulse duration of 1 ms, until a maximum contraction was obtained using varying frequencies (2 Hz, 5 Hz, 12 Hz, 20 Hz, and 30 Hz) using a Grass S88 stimulator (Natus Neurology Inc., Warwich, RI) interfaced with a Stimu-Splitter II (Med-Lab Instruments, Loveland, CO) power amplifier and LabChart® software (ADInstruments, Colorado Springs, CO). Frequency response curves were generated. Muscle strips were then treated with 1 µM tetrodotoxin (TTX) and allowed to incubate for 15 min before repeating EFS.

Tissue Collection, PGP9.5 Immunohistochemistry, and TUNEL Assay

At euthanasia, full-thickness bladder tissue was collected, fixed in 4% paraformaldehyde for 4 h, and equilibrated in 10% sucrose in phosphate buffer overnight, following by 30% sucrose in phosphate buffer for approximately 6 h. Tissues were embedded in OCT Compound (Scigen, Gardena, CA) and stored at −80°C until processing. Tissues were then cryosectioned into 14 µm sections. Subsets of sections were permeabilized with 0.1% Triton-X, blocked using 4% goat serum, and stained using 1° anti-PGP9.5 at 5 µg/ml (Abcam, Cambridge, MA) and 2° Cy™3 Goat anti-Mouse at 2.5 µg/ml (Jackson IR, West Grove, PA) to visualize nerves. Additional sections were processed using ApopTag® Fluorescein Direct In Situ Apoptosis Detection Kit (Millipore, Burlington, MA) according to the manufacturer’s directions, which detects apoptotic cells using terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL). Slides were then stained with 0.1 µg/ml 4′,6-diamidino-2-phenylindole, dihydrochloride (DAPI) in PBS for 15 min before coverslipping with 80% glycerol in PBS. Slides were imaged using a Nikon Eclipse E1000 upright microscope (Nikon, Melville, NY) equipped with an EXi Aqua bio-imaging camera (QImaging, Surrey, BC, Canada). Adjacent sections were stained with hematoxylin and eosin and were grossly assessed in a blinded fashion for presence of histopathology. A minimum of three sections were examined per stain and per animal.

Data Analysis

Bladder capacities were measured using the inflection point of the generated cystometrograms. If an inflection point was not apparent due to root transection, capacities were determined using the volume that generated 60 cmH₂O, a criterion selected based on exhibited signs of discomfort during awake cystometry.

Data were analyzed using either two-tailed unpaired t-tests or ANOVA with Tukey’s post hoc analysis using GraphPad Prism 7.0 (La Jolla, CA). A p-value of < 0.05 was considered statistically significant. All data are expressed as mean and standard error (SEM).

RESULTS

Behavioral Observations following Surgery

Although one of the surgical groups underwent an extensive decentralization that included the dorsal root of L7, bilaterally, all animals retained hind limb function. Animals that underwent sacral ventral and dorsal root plus dorsal root of L7 consistently showed evidence of overflow incontinence such as frequent urine dribbling. A culture-confirmed UTI was treated in one animal from the group that included L7 dorsal root transection.

In vivo Bladder Contractility

All animals had similar bladder capacities at initial surgery prior to decentralization of 61.15 ± 14.18 mL for sham animals, 55.20 ± 4.80 mL for animals that underwent transection of ventral and dorsal sacral roots and the hypogastric nerves, and 62.68 ± 6.96 mL for animals that underwent transection of sacral roots plus the dorsal roots of L7 and hypogastric nerves. There was no difference in the bladder capacity between sham animals (87.50 ± 29.62 mL), animals that received the sacral root transection (146.0 ± 66.85 mL; p = 0.5748) and animals that received sacral root plus L7 dorsal root transection (67.03 ± 14.97 mL; p = 0. 0.6719). There was no significant difference between the resting bladder pressure of sham-operated animals (25.32 ± 15.32 cmH₂O), animals that received sacral decentralization (11.35 ± 3.249 cmH₂O; p=0.5774), and animals that received sacral and L7 dorsal root decentralization (18.16 ± 4.465 cmH₂O; p=0.6696). Maximum detrusor contraction after pelvic plexus stimulation was recorded during the terminal surgery. While year-long sacral root injury did not decrease the strength of pelvic plexus stimulation-induced bladder contractions in vivo (10.2 ± 2.8 cmH₂O), compared to sham controls (23.3 ± 4.3 cmH₂O), sacral decentralization with additional L7 dorsal root transection significantly decreased the induced contraction (5.1 ± 1.3 cmH₂O; p<0.05; Figure 1). Despite the diminished strength of contraction, all animals exhibited a maximum strength of contraction > 0 cmH₂O.

Figure 1. Bladder pressure via stimulation of the pelvic plexus decreases after long-term lower spinal root injury in vivo.

Each point represents the maximal contraction yielded from stimulation of either left or right pelvic plexus in a single animal, whichever produced the strongest bladder contraction. Animals that received sacral dorsal and ventral root transection (S1-S3; dark gray square, n=3) did not significantly decrease in maximal bladder contraction after pelvic plexus stimulation, compared to sham-operated controls (black circle, n=5). Animals that additionally received L7 dorsal root transection (S1-S3 ventral and dorsal + L7 dorsal; light gray triangle, n=4) showed a significant decrease. Despite the decrease, note that no animals exhibited a complete loss of pelvic plexus stimulation-induced bladder contractions. Data was analyzed using a one-way ANOVA followed by Tukey’s post hoc analysis.

Ex vivo Bladder Contractility Experiment

Within 24 h of the terminal surgery, at one year after the root transection, bladder smooth muscle contractility and intramural nerve function was tested ex vivo. In contrast to the in vivo experiment, bladder smooth muscle strips isolated from animals that underwent long-term injury exhibited equally robust contractions to electric field stimulation at all tested frequencies compared to strips from sham control animals (Figure 2). Treatment with 1 µM TTX blocked EFS-induced contractions across groups (Figure 2).

Figure 2. Ex vivo nerve-evoked bladder smooth muscle contractility changes at one year after lower spinal root injury.

Frequency-response curves were generated by exposing muscle strips to varying frequencies while keeping voltage (12 V) and pulse duration (1 ms) consistent. Solid curves represent pre-drug responses and dashed curves represent response to electric field stimulation after treatment with 1 μM tetrodotoxin (TTX). Sham data are represented by blue circles (pre-drug: N=6, n=125; TTX: N=6, n=24), sacral decentralization by red squares (pre-drug: N=3, n=48; TTX: N=3, n=6), and sacral + L7 dorsal root decentralization by green triangles (pre-drug: N=4, n=108; TTX: N=4, n=17). N = number of animals within treatment group; n = number of strips tested within treatment group. Mean ± SEM per group (generated by mean per dog within group) is shown.

PGP9.5 Immunohistochemistry

Nerve density appeared qualitatively similar across groups. Intramural ganglia were identified in both decentralized and sham groups (Figure 3). Intramural ganglia containing large rounded neuronal cell bodies were also identified within the bladder smooth muscle layers in each decentralized group and in the sham operated group (Figure 3). Hematoxylin and eosin staining showed no presence of pyknotic neuronal cell bodies.

Figure 3. Intramural ganglia and nerves are present within bladder wall at one year after sacral root + L7 dorsal root transection.

Full thickness bladder tissue stained for PGP9.5 (neuronal marker) and DAPI. A. PGP9.5⁺ nerves within smooth muscle of the bladder in sham-operated animal. B. PGP9.5⁺ intramural ganglion in sham with HE of the same ganglion. C. PGP9.5⁺ nerves within smooth muscle of the bladder in sacral + L7 dorsal root transection animal. D. PGP9.5⁺ intramural ganglion in sacral + L7 dorsal root transection animal. Scale bar represents 100 microns.

Apoptosis Assay

No difference in apoptosis within smooth muscle of the bladder was observed after staining with TUNEL (0 smooth muscle cells/area for all groups). Apoptotic cells were observed in the bladder urothelium (the inner epithelial layer) of each transected group and control group. Positive staining in the urothelium served as an internal positive control for the assay.

H&E Assessment of Bladder Wall

The smooth muscle uniformly appeared intact without signs of cell death across all tissues. The urothelium frequently looked disturbed in samples from sacral dorsal and ventral + L7 dorsal root transection animals, with variations in size and quantity of epithelial cells (Figure 4). An increase in blood vessels in the suburothelial space was observed in animals that received both the sacral transection and the sacral dorsal and ventral + L7 dorsal root transection (Figure 4, noted by asterisk). While inflammatory infiltrates were visible in all samples, lymphocytic clumping was noted in two out of four animals that received sacral dorsal and + L7 dorsal root transection (not shown).

Figure 4. HE-stained bladder tissue.

Full-thickness histological sections of the detrusor were stained to assess tissue morphology. Visible blood vessels are marked by an asterisk. u – urothelium; m – smooth muscle; ct – connective tissue.

DISCUSSION

While some studies have addressed urodynamic changes after spinal cord injury in patients¹⁰ and the short-term consequences of decentralization on bladder function in animal models¹¹, no study to date has evaluated the effects of long-term lower spinal root injury on bladder function in vivo or on the individual components that contribute bladder function within the same model. In this study, we combined in vivo urodynamic tests with ex vivo assessment of isolated smooth muscle fascicles from the same animals to determine the extent to which smooth muscle dysfunction contributes to lower motor neuron neurogenic bladder in a canine model.

The extensive decentralization in this study, which included transection of all sacral roots plus the dorsal root of L7, as well as the hypogastric nerves bilaterally, was selected based on a previous study that identified the L7 dorsal root as one of the major sources of sensory neuron cell bodies innervating the bladder.¹² Because these animals are part of a larger reinnervation study, we needed to ensure that the bladder was decentralized as much as possible to better evaluate the success of reinnervation.

Preservation of smooth muscle contractility may indicate that detrusor smooth muscle integrity does not depend on external neuronal input. However, it is more likely that other neuronal inputs (originating from sites other than the sacral spinal cord) may be maintaining the detrusor smooth muscle, such as the network of intramural ganglia that act upon smooth muscle. The robust response to electric field stimulation seen in the smooth muscle strips regardless of treatment group indicates that the intramural neurocircuitry is still viable because the response to EFS is action potential-dependent, which was confirmed by the inhibition of contractions following treatment with sodium channel blocker TTX.

Nerve density between each group looks grossly similar, although density can vary from animal to animal as a result variations in tissue processing. We conducted a TUNEL assay to evaluate apoptosis occurring within the bladder tissue. As expected, there was evidence of apoptosis in the urothelium (the inner epithelial layer) of the bladder due to the rate of cell turnover as an immunological response to environmental factors.¹³ The lack of TUNEL labeling in the smooth muscle of samples from the injury group confirms that the decentralization procedure did not induce cell death within the smooth muscle of the bladder. Future studies will address the effects of decentralization on urothelial integrity.

Pelvic plexus-induced bladder contractions significantly decreased by one year after decentralization that included the transection of all sacral roots, the dorsal roots of L7, and the hypogastric nerves, bilaterally, confirming a previous finding that the L7 dorsal root plays an important role in bladder sensation.¹² Despite the significant decrease in the strength of bladder contraction, the preservation of smooth muscle function and at least some nerve activity may encourage successful surgical reinnervation after long-term lower spinal root injury.