Multiparity affects conduction properties of pelvic floor nerves in rabbits

Abstract Introduction Women often develop pelvic floor dysfunction due to damage to the pelvic musculature during childbirth; however, the effect on pelvic floor nerves function is less understood. This study used adult rabbits to evaluate the electrophysiological and histological characteristics of the bulbospongiosus (Bsn) and pubococcygeus nerves (Pcn) in multiparity. Methods Compound nerve action potentials (CNAP) were compared between age‐matched nulliparous and multiparous animals and associated to the histological characteristics of myelinated axons from the Bsn and Pcn nerves. The extensor digitorum longus nerve (EDLn) was used as negative control. Data were analyzed with unpaired two‐tailed Student's t test or Mann–Whitney U test to determine significant differences between groups. Results The onset and peak latencies, duration, and conduction velocity of the motor fibers in these pelvic nerves were not significantly different between nulliparous and multiparous animals. However, the peak‐to‐peak amplitude and area of the CNAP in both Bsn and Pcn were reduced in multiparous rabbits. Histology showed a higher percentage of axons with myelin disorganization caused by multiparity in these pelvic nerves. Together, the data indicate a reduction in the number of functional pelvic axons due to multiparity. As expected, no effect of parity was observed in the EDLn controls. Conclusions Present findings demonstrated that multiparity affects myelination and consequently conduction properties in the small pelvic floor nerves.


| INTRODUC TI ON
Pelvic floor muscles play a critical role in reproductive and excretory processes. Women are more prone than men to the onset of pelvic floor dysfunctions as the pelvic floor is more drastically impacted by pregnancy. Childbirth is a well-recognized risk factor in developing pelvic organ prolapses and stress urinary incontinence (SUI). In both cases, there is compelling evidence indicating that the pelvic floor muscles are injured during pregnancy and parturition. Indeed, the weakening of pelvic floor muscles is considered the leading cause of SUI in women and might be related to impaired function of their motor innervation as the pudendal and perineal compound muscle action potential is reportedly affected (Olsen, Ross, & Stansfield, 2003). Furthermore, pudendal neuropathy has been associated with pelvic floor disorders including SUI in parous women (Sangwan et al., 1996;Snooks, Barnes, & Swash, 1984;Snooks, Swash, & Mathers, 1990), and a link between SUI and nerve injuries affecting the urethral viscerosomatic reflex has been reported (Aguiar et al., 2013). Consequently, a number of therapies to manage SUI, ranging from Kegel exercises to electrical stimulation of sacral and pudendal nerves, have been proposed as potential therapies to improve the function of pelvic floor musculature and its innervation (Elser, 2012).
In female rabbits, the Bsn originates at the ischiorectal fossa, from the second branch of the pudendal nerve, and extends caudomedially toward the pubic bone, while the Pcn originates from the sacral S3 and S4 branches (Cruz et al., 2017). The anatomical location of Bsn and Pcn makes them susceptible to injuries due to pregnancy and childbirth, as indicated by the reportedly altered viscerosomatic reflexes, muscle injury, and impaired muscle contractility in multiparous rabbits (López-García et al., 2016;López-Juárez et al., 2018;Martínez-Gómez et al., 2011). These alterations could be directly or indirectly related to deficiencies in the functional and anatomical properties of the pelvic and perineal innervation. This has been supported by a report that multiparity affects ganglia of the pelvic plexus located in the vicinity of Bsn and Pcn in rabbits . Based on this information, we hypothesized that multiparity directly damages the pelvic floor nerves, and evaluated the electrophysiological and histological characteristics of Bsn and Pcn to determine the effect of multiparity on the compound nerve action potential (CNAP) parameters, and axon myelination in nulliparous and multiparous rabbits. The study confirmed the damage of the pelvic nerves due to multiparity.
Six-month-old nulliparous female rabbits mated with sexually experienced males and copulated 24 hr after the first three deliveries (Martínez-Gómez et al., 2011). Thus, rabbits were pregnant and lactating for 20 days when pups were weaned. On the day of the fourth delivery, neonate pups were sacrificed to avoid lactation and to allow multiparas reaching a hormonal status quite similar to nulliparas as determined by the estradiol levels in serum (López-García et al., 2013). Nulliparous animals were sacrificed when reached the age above mentioned and multiparous at postpartum day 20, with an urethane overdose (i.p.).

| CNAP recordings
The Bsn and Pcn on the right side were dissected in six animals per group as recently described (Cruz et al., 2017). The animals were anesthetized with urethane (Sigma Chemical, USA; 0.9 g/kg; 20% in distiller water; i.p.) and fixed in a dorsal supine position, and the pelvic and perineal muscles were exposed. The Bsn and Pcn segments were dissected proximal to their muscle insertion (Figure 1a-c). As control, a segment of the right extensor digitorum longus nerve (EDLn), a motor nerve located in the popliteal region of the leg, was exposed lateral to gastrocnemius muscle and dissected (Figure 1a, inset).
The nerves were stimulated by wrapping silver wire electrodes connected to a constant current stimulator (Digitimer DS3) at one end and recording from suction electrodes made with glass capillaries filled with the Krebs saline solution and polished tips. The CNAP were evoked by the application of square current pulses at variable intensities trigger by a Grass stimulator (S48). The evoked responses were recorded by drawing two nerve segments into the suction electrodes connected to a grass amplifier (Grass 7P511) and to an oscilloscope (Tektronix TDS2024C) to visualize and record data. The stimulation train was ascendant with current graduate pulses (µA), 0.05 ms pulse duration at 1 Hz frequency. The activation threshold (1 × T) was defined as the minimum current that evoked a visible CNAP. Afterward, the current was gradually increased until reach the maximal CNAP responses (approximately 6 × T). It is important to note that the inversion of the polarity did not change any CNAP of the nerves.
The recorded CNAP were used to measure the maximal amplitude (A), considered as the voltage (mV) at the peak value of the CNAP; the peak latency (pl), calculated time elapsed from the stimulus artifact to the CNAP peak and the onset latency (ol), the time elapsed from the stimulus to the beginning of the CNAP multiplied by A and by 2; the CNAP area, calculated from [(pl-ol)*A]*2; the CNAP duration, calculated from (pl-ol)*2; and the peak conduction velocity, which was calculated dividing the length of the nerve section by the elapsed time from the stimulus to the CNAP-component peak.

| Nerve histology
An additional cohort of 6 animals per group was used to evaluate the histology of the Bsn, Pcn, and EDLn. The nerves were dissected and immediately fixed in Karnovsky buffer (200 mM sodium cacodylate, 25% glutaraldehyde, 1% paraformaldehyde, pH 7.3) during 24 hr at 4°C. The tissue was then immersed into 100 mM sodium cacodylate, pH 7.4, until embedded in epon resin. The tissue was incubated in 1% sodium tetroxide diluted in Zetterquist´s buffer (5% veronal acetate, pH 7.3-7.5) for 1 hr, washed three times with distilled water, and dehydrated with ascending concentrations of ethanol (70%, 80%, 90%, and 100%) 10 min each, at room temperature. The tissue was then washed two times for 15 min in a propylene oxide solution. Finally, nerve segments were immersed in epon-acetonitrile mix solution 1:1 (v/v) during 1 hr, rinsed in epon-acetonitrile mix solution 2:1 (v/v) for 1 hr, sank in pure epon for 16 hr in constant rotation, and transferred to inclusion templates at 60°C for 24 hr for resin polymerization.
The tissue was cut in 1 μm semithin transverse sections in an ultramicrotome (Leica Ultracut UCT) and stained with toluidine blue. Slides were analyzed using a light microscopy (Olympus CX31), and photomicrographs taken with an attached digital camera (Nikon DS-Ri1). Axons that did not show a well-defined circular form of the myelin sheet, that had myelin invaginations, myelin interruption, and/or apparent delamination of the myelin sheaths, were identified as "axons with myelin disorganization" and counted. The digital images of the nerves were divided into quadrants with a grid placed at the center of each fascicle. In the right superior quadrant, the total number of myelinated axons either normal or with myelin disorganization were counted and reported as percentage of the total number of axons in that quadrant.

| Statistical analyses
Data are presented as the mean ± standard error of the mean (SEM), unless otherwise stated. The normality of data was assessed using a Kolmogorov-Smirnov test, and the means for each group analyzed with an unpaired two-tailed Student's t test or Mann-Whitney U test.
p < 0.05 values were considered statistical significant. Statistical analysis was performed using the GraphPad Prism 6.0e software (GraphPad).

| Evoked CNAP area and amplitude were reduced in the pelvic floor nerves of multiparous animals
The evoked CNAP in the Bsn and Pcn (Figure 2a,b), but not of the EDLn of multiparous animals was reduced compared to nulliparous rabbits ( Figure 2c). Values for the onset and peak latencies, duration, and peak conduction velocity for Bsn, Pcn, and EDLn were not significantly different in nulliparous and multiparous rabbits (Tables   1-3 (Figure 4a). The Pcn was also significantly reduced in these animals at 5 × T: U = 0, p = 0.002, and 6 × T t = 2.44, p = 0.035 (Figure 4b). Multiparity did not affect the CNAP area when measured in the EDLn (Figure 4c).

| Multiparity affected myelinated axons of the Bsn and Pcn but not of EDLn
Evaluation of toluidine blue-stained semithin sections showed remarkable differences in the appearance of myelinated axons between nulliparous and multiparous rabbits (Figure 5a

| D ISCUSS I ON
Present findings demonstrate that multiparity significantly decreased the amplitude and area of the CNAP evoked in the pelvic Bsn and Pcn motor nerves. In contrast, the onset and peak latencies, duration, and conduction velocity of the motor fibers in these pelvic nerves were not significantly different between nulliparous and multiparous animals. None of variables measured from the CNAP evoked in the EDLn were affected by multiparity. This result was consistent with a higher percentage of axons with myelin disorganization in Bsn and Pcn from multiparous animals.
Twenty days after the fourth delivery, CNAP parameters as the latencies to the onset and to the peak evoked in Bsn and Pcn of nulliparas and multiparas did not differ. These results contrast with those from clinical studies reporting an increase in the pudendal nerve motor terminal latency (PNMTL) in patients with EAS and/ or EUS dysfunction after parturition (Aguiar et al., 2013;Sangwan et al., 1996;Snooks et al., 1984Snooks et al., , 1990Tetzschner, Sørensen, & Lose, 1997). However, the different methodological approaches between ours and the last studies offer a possible explanation to this discrepancy. The present study used rabbit nerves ex vivo, and the clinical studies evaluated women with fecal or incontinence urinary of varied duration or parous women on 12 weeks postpartum.
Furthermore, the nerves evaluated in both studies were different; the pudendal nerve was examined clinically, whereas the Bsn and Pcn were examined in multiparous rabbits 20 days after the fourth delivery. It is also possible that a more drastic impair of the EAS was involved in the PNMTL cases (Snooks et al., 1990).
In this study, we reported conduction velocities for both Bsn and Pcn higher than 10 m/sec, which is in agreement with values estimated for myelinated fibers of the pelvic nerve of rats (Nakayama, Noda, & Hotta, 1998). In the study conducted for Nakayama and colleagues, the shorter nerve segment used could have impede to discriminate the fast large myelinated fibers, which is supported F I G U R E 5 Multiparity increases the occurrence of axons with myelin disorganization in perineal and pelvic nerves of rabbits. Representative photomicrographs from Bsn (a-c), Pcn (d-f), and EDLn (g-i) transverse sections stained with Toluidine blue from nulliparas (N) and multiparas (M). Data are medians ±minimal to maximal values for Bsn (j) and Pcn (k). Symbols indicate abnormal myelinated fibers: myelin disruption (asterisk), invagination (ampersand), and apparent separation of the myelin sheaths (arrow). Scale bar, 10 μm herein from data gathered from EDLn (Li & Shi, 2007). We noted that multiparity does not affect the CNAP-peak conduction velocities of Bsn and Pcn in rabbits, suggesting that multiparity does not significantly affect the conduction velocity of the majority of myelinated fibers (e.g., fiber diameter, internode distance, nodal area, etc. (Waxman, 1980)), as has been reported for myelinated fibers in the pelvic nerves of aging rats (Nakayama et al., 1998).
Multiparity decreases the amplitude and area of CNAP of the Bsn and Pcn, to 40% and 70% observed in nulliparous animals, respectively. Considering that this value represents the extent in which summed action potentials are propagated along the nerve; this finding suggests that partial damage occurred to some axons in these pelvic nerves due to multiparity. This notion is consistent with the observed myelin disorganization in the toluidine blue-stained semithin sections of both nerves. Altogether, these data support the interpretation that the observed focal demyelination in these nerves caused by multiparity impairs the conduction of the depolarization signal in the damaged axons, resulting in functional gaps between nodes of Ranvier (Robinson, 2000). Certainly, the ultrastructural evaluation of the Bsn and Pcn using electron microscopy is warranted to further define the specific types of myelin damaged that occur as a result of parturition. It is known that pudendal nerve crush injury during vaginal distention at delivery results in abnormal nerve fascicles around the urethra, reducing the leak point pressure in female rats (Damaser et al., 2003). Overall, the impaired propaga- Clinical and basic science studies support that labor trauma induces damage to pelvic floor nerves, which leads to the onset of pelvic floor disorders including fecal and urinary incontinence (Aguiar et al., 2013;Snooks et al., 1984Snooks et al., , 1990Tetzschner et al., 1997). This is commonly accepted as one of the frequent causes of EAS and EUS disabilities (Olsen et al., 2003). Functional and histological evidence showed in this report indicates that multiparity in rabbits partially impairs the function of Bsn and Pcn, persisting for 20 days after delivery. This damage seems correlated to the disorganized reflex myographic activity of the Bsm and Pcm that has been reported in animals with impaired urodynamic function (Martínez-Gómez et al., 2011). To our knowledge, this is the first study reporting the direct nerve injury to small nerves of pelvic floor, generated as a consequence of multiparity. Further studies, however, are needed to address the possible extent and duration of recovery of these damaged myelinated axons in Bsn and Pcn.
An important limitation of this study is the lack of information on the possible effects of parturition on other axon types including medium and slow conducting B-and C-fibers. The relatively low current used in this study were below the activation threshold for these fibers. In rabbits, the Bsm assists in voiding while the In summary, our present findings indicate that multiparity reduces the CNAP amplitude and area of the Bsn and Pcn in rabbits.
These changes were associated with a high percentage of abnormal myelinated axons, and underlie the importance of the pelvic floor neuromuscular damage, as a clinical target for the treatment of the dysfunctional pelvic floor, including SUI.