Phosphorylation of spinal signaling-regulated kinases by acute uterine cervical distension in rats

Introduction 

Although labor pain is common and severe, our understanding of the neurobiological and molecular basis of this type of pain is limited because of lack of a suitable experimental model.1 Pain during the first stage of labor has been described as visceral pain and is mediated mainly by cervical distension.2 Recently, an acute uterine cervical distension (UCD) model in the lightly anesthetized rat has been used in several reports to study the physiology of acute visceral nociception from the uterine cervix.3, 4, 5, 6, 7, 8 UCD produces a stimulus-dependent increase in afferent nerve and reflex abdominal muscle. Furthermore, these UCD-induced responses can be inhibited by intravenous or intrathecal opioids.3, 4, 5, 6 In addition, local infiltration of the cervix with lidocaine can inhibit expression of the immediate early gene product cFos in the spinal cord induced by UCD,8 further suggesting that UCD is a truly noxious stimulus and reflects noxious input from the uterine cervix rather than traction on other visceral or somatic structures by deformation of the tissue during distension. Therefore, this acute UCD model can be used as a visceral model related to labor pain.

The extracellular signal-regulated kinases 1 and 2 (ERK 1/2) are two isoforms of the mitogen-activated protein kinases, which can be activated (phosphorylated) following various stimuli and transduce extracellular stimuli into intracellular post-translational and transcriptional responses.9 In the spinal dorsal horn, ERK 1/2 have been shown to be specifically and rapidly activated after somatic noxious stimulation, and ERK inhibitors prevent or reduce pain behavior induced by these stimuli,10, 11, 12, 13 suggesting a role for spinal ERK in somatic pain. More recently, studies also showed some noxious visceral stimulations, for example intracolonic application of mustard oil and capsaicin,14 gastric distension15 and colorectal distension,16 induce ERK phosphorylation in spinal dorsal horn neurons, indicating that spinal ERK may also be involved in visceral pain processing. However, there has been no study examining spinal ERK 1/2 activation in UCD-induced visceral pain. In this study, we tested the hypothesis that UCD stimulus can activate spinal ERK 1/2 by immunohistochemical assessment of phosphorylated-ERK 1/2 (pERK 1/2) in the spinal dorsal horn neurons after UCD stimulation.

Methods 

All animal experimental procedures were approved by the Medicine Committee on Animal Research of our institute and were conducted in accordance with the recommendations of the International Society for the Study of Pain Ethical Guidelines.17

A total of 62 adult female Sprague-Dawley rats weighing 190-230g, purchased from Shanghai Slac Laboratory Animal Co. Ltd., Shanghai, China, were used. The rats were housed in 12-h light/dark cycles and had free access to standard food and tap water at the ambient temperature of 24-28°C. Food, but not water, was withheld for 12h before the study.

Uterine cervical distension was performed as described previously3 with slight modifications. Animals were anesthetized with intraperitoneal chloral hydrate 300mg/kg and breathed spontaneously. The right carotid artery was catheterized for continuous monitoring of mean arterial blood pressure and heart rate with Medlab-U/4C biological signal collection system (Nanjing Medease Science and Technology Co. Ltd., Nanjing, China). The right jugular vein was also catheterized for fluid administration. A small low midline laparotomy was then performed to expose the uterus, and two fine metal rods (22-gauge needles, Becton Dickinson Medical Devices Co. Ltd., Shanghai, China) were inserted through both uterine cervical canals. The rods entered from the uterus and left through the vaginal wall, both as near as possible to the cervix. With one rod fixed to a metal stand by a silk suture, UCD was applied to the cervix by hanging various standard weights on another rod through a silk suture and pulley. The sham operation group included rats with only an abdominal incision and two metal rods inserted through cervical canals but without distension (0g). All exposed wounds were infiltrated with about 1mL 1% lidocaine, without epinephrine, before incision. Rectal temperature was monitored continuously and maintained at 37-38°C by means of a circulating-water heating pad and heat lamp.

The visceromotor response induced by UCD was recorded by quantifying electromyographic (EMG) activity with a concentric needle electrode inserted into the right rectus abdominis superior to the inguinal ligament. Through the Medlab-U/4C biological signal collection system, EMG activity was continuously amplified (5000 gain), filtered (highpass, 500-5000Hz), digitized and stored for an off-line analysis. After a 40- to 60-min stabilization period and once a light plane of anesthesia was achieved, evidenced by mean arterial pressure ⩾100mmHg, the study started. Preliminary experiments showed that a stable UCD-evoked EMG response could usually be obtained at that time. Rats were subjected to UCD stimulus (25, 50, 75 and 100g, respectively, for10 s) or no distension (sham group, 0g) (n=6 for each distension force) following a 5-min basal EMG recording. Each UCD force was only applied once for each individual rat. The raw EMG data over the 10 s just before (baseline) and during UCD were rectified and quantified by calculating the area under the curve. For data analysis, the percentage of area under the curve of distension to baseline was used as the responses to UCD for each animal. In addition, mean arterial pressure and heart rate were averaged over the corresponding time periods, and the percent of pre-distension basal value was used as the cardiovascular response to UCD. The relevant data in the sham group also were calculated.

Ten minutes after UCD (60min after anesthesia for sham rats), saline followed by 4% paraformaldehyde was perfused transcardially at 4°C. The spinal cord was dissected out by laminectomy and divided into cervical (C5-8), thoracic (T5-8), thoracolumbar (T12-L2) and lumbosacral (L6-S1) segments. In a separate experiment, another 32 rats were used for observing the time course of spinal ERK 1/2 phosphorylation after UCD stimulus. In these rats, the thoracolumbar segments were removed at 10, 30, 60 and 120min, respectively, after 75g UCD (n=5 for each time point) or no distension (0g, n=3 for each corresponding time point) for ERK 1/2 immunohistochemistry. The distension force and spinal segment chosen in the time course study were based on preliminary experiments. The spinal segments were fixed in 4% paraformaldehyde and imbedded in paraffin, and cut transversally into 4-μm thick sections interspaced at 300μm.

For each segment, according to a computer-generated random digits table, three nonadjacent sections were randomly selected from 10 sections numbered from 1 to 10 processed for pERK 1/2 immunohistochemical staining. Immunohistochemical staining was performed using DAKO EnVision™+ System, Peroxidase kit (DakoCytomation, Glostrup, Denmark). Briefly, each section was deparaffinized, rehydrated and incubated with fresh hydrogen peroxide 3mL/L in methanol for 30min at room temperature and then washed in PBS. Sections were incubated with monoclonal mouse anti-pERK 1/2 (E4) at 1:100 dilution (Santa Cruz Biotechnology, Inc., Santa Cruz, California, USA) overnight at 4°C, washed three times in PBS, followed by incubation with secondary antibody for 30min at room temperature. Then 3,3′-diaminobenzidine tetrachloride was used for color development and the sections were counterstained with hematoxylin, which resulted in brown-colored precipitates at the antigen site. Primary antibody was substituted by PBS for negative control.

Under light-field microscopy at ×100, the number of pERK 1/2-immunoreactive dorsal horn neurons under identical view per section was counted bilaterally and expressed as neurons/section by an assistant blinded to the treatment. The numbers of pERK 1/2-IR neurons of the three sections from each spinal segment were averaged for each animal.

Results 

A single intraperitoneal chloral hydrate bolus combined with local infiltration of lidocaine resulted in a stable anesthetic state until the termination of the experiments. No additional dose of anesthetics was necessary.

Effect of UCD on visceromotor response: In the absence of UCD stimulus (sham, 0g), spontaneous EMG activity was generally absent in the lightly anesthetized rat. In contrast, UCD resulted in a stimulus-dependent increase in EMG activity. The evoked EMG responses increased, respectively, by 1.8-fold at 25g, 2.5-fold at 50g, 4.6-fold at 75g, and 4.8-fold at 100g over the sham rats (P<0.05 for each distension force). UCD of 75 and 100g produced a significant increase in EMG responses (P<0.05) compared to UCD of 25g, but with no difference between 75 and 100g (Fig. 1 A and B). UCD failed to exert significant effects on mean arterial pressure and heart rate compared to the sham group, even with greater distension forces of 75 and 100g (data not shown).

View Large Image Download to PowerPoint

Fig. 1. Electromyographic (EMG) response to uterine cervical distension (UCD). (A) Representative tracings show EMG activities recorded from the rectus abdominis after each UCD stimulus. (B) A summary of stimulus-dependent changes in EMG response to UCD. Each symbol represents the mean±SEM for six rats. ∗P < 0.05 vs. sham (0g). #P<0.05 vs. 25g UCD.

Effect of UCD on spinal neuronal ERK 1/2 phosphorylation: In the non-distended sham rats, a few pERK 1/2-IR neurons were observed bilaterally in spinal dorsal horn with no differences among spinal segments. UCD resulted in a clear stimulus-dependent increase in the number of pERK-IR neurons in T12-L2 segments. pERK staining was predominantly localized to the cytoplasm of neurons in the deep dorsal horn (laminae III-VII) and the central canal (lamina X), although a few labeled nuclei and glial cells were also observed (Fig. 2 A-E). The numbers of pERK 1/2-IR neurons in T12-L2 segments induced by UCD from 25g to 100g were significantly increased compared with the sham group (by an average increase of 2.6-fold at 25g, 5.3-fold at 50g, 8.1-fold at 75g, and 8.0-fold at 100g; P<0.05 for each) and with other spinal segments. The maximal activation observed was in the 75- and 100-g groups, but the difference between 75 and 100g was not significant (Fig. 3). This corresponded with the EMG response to UCD. In contrast, although UCD produced a slight (non-significant) increase in the number of pERK 1/2-IR neurons in lumbosacral segments, there was no significant evidence of ERK 1/2 activation after UCD in the cervical, thoracic and lumbosacral segments compared to sham (Fig. 3). Therefore, ERK 1/2 activation after UCD stimulation was specifically localized to the thoracolumbar spinal cord.

View Large Image Download to PowerPoint

Fig. 2. Phosphorylated-ERK 1/2 (pERK 1/2) expression in the spinal dorsal horn neurons induced by uterine cervical distension (UCD). (A-E) Photomicrographs of pERK 1/2 labeling in the thoracolumbar (T12-L2) segment 10min after UCD of 0 (A), 25 (B), 50 (C), 75 (D) and 100g (E). Scale bars, 100μm.

View Large Image Download to PowerPoint

Fig. 3. Numbers of pERK 1/2--immunoreactive (IR) neurons in cervical (C5-8), thoracic (T5-8), thoracolumbar (T12-L2) and lumbosacral (L6-S1) segments 10min after various UCD forces. Each symbol represents the mean±SEM for six rats. ∗P < 0.05 vs. sham (0g). #P<0.05 vs. 25g UCD. +P<0.05 vs. cervical, thoracic and lumbosacral segments. ++P<0.05 vs. cervical segment.

Time course of post UCD ERK 1/2 phosphorylation: Fig. 4 shows the time course of the numbers of pERK 1/2-IR neurons in T12-L2 segments in response to 75-g UCD stimulus or no distension (0g). The activation of spinal ERK induced by 75-g UCD was evident at 10min, peaked at 60min with a slow decline for 120min but remained significant.

View Large Image Download to PowerPoint

Fig. 4. Time course of phosphorylated-ERK 1/2-immunoreactive (pERK 1/2-IR) neurons number in the thoracolumbar (T12-L2) spinal segment after 75g uterine cervical distension (UCD) (closed squares) or sham (0g) (open squares). Each symbol represents the mean±SEM for five rats in 75g UCD and three rats in sham at each time point. ∗P<0.05 vs. sham. #P<0.05 vs. 60min.

Discussion 

The present study shows that a single 10-s UCD in the lightly anesthetized rat produces phosphorylation of ERK 1/2 selectively in the thoracolumbar spinal dorsal horn neurons known to receive uterine cervical afferent input by the hypogastric nerve. This activation of spinal ERK 1/2 corresponds well with the evoked EMG response to UCD. Thus, UCD-induced visceral nociception may involve spinal ERK 1/2 activation.

For technical reasons, we are unable to perform UCD in awake rats, and thus unable to observe behavioral responses to cervical distension. However, contractile response of the abdominal muscles that can be quantified using EMG recordings is a typically measured variable that correlates with distension-induced visceral pain, even in the lightly anesthetized rat.15, 18 Indeed, as observed previously,3, 4, 5, 6 we found that UCD triggered a stimulus-dependent increase in EMG activity in the rectus abdominis, supporting distension of uterine cervix as noxious. Blood pressure and heart rate are also sometimes used as measures of visceral nociception. However, in contrast to the conscious model, in the anesthetized rat, the distension-mediated increase in cardiovascular indices was either blunted or more commonly, converted into stimulus-dependent reductions.19 Similarly, the cardiovascular responses were also remarkably variable depending on the anesthetic used. For example, with pentobarbital and chloralose anesthesia, UCD did not affect blood pressure and heart rate.3 In contrast, UCD evoked an increase in mean arterial pressure, but not in heart rate in rats with halothane anesthesia at about 0.5-0.7%, a level that allowed a reflex response of rectus abdominis muscles to UCD but prevented purposeful escape behavior.4, 5, 6 Thus, it is not surprising that both mean arterial pressure and heart rate were unaffected by UCD in the present study. This result implies that cardiovascular response to UCD can be less specific and sensitive than the EMG response, and is not typical of changes accompanying visceral pain.

Activation of ERK in the spinal cord has a substantial role in somatic nociception.10, 11, 12, 13 Recent studies have shown that ERK phosphorylation in the spinal dorsal horn neurons might also participate in some visceral pain processing.14, 15, 16 However, the role of spinal ERK 1/2 in response to acute noxious UCD has not been previously described. The present study clearly demonstrated that UCD can rapidly induce activation of spinal ERK 1/2 in a stimulus-dependent manner. The pERK 1/2-IR neurons were specifically localized to T12-L2 segments, which is consistent with uterine cervical innervation by the hypogastric nerve.3 Furthermore, ERK 1/2 activation corresponded well with the incidence of the EMG response to UCD. Therefore, we believe that activation of spinal ERK 1/2 is, at least in part, responsible for nociceptive transmission from the uterine cervix in the spinal cord.

ERK is activated by membrane depolarization and calcium influx.20 Presumably, the ERK 1/2 phosphorylation in T12-L2 segments results from primary afferent activation induced by UCD, which produce action potentials and associated calcium transfer in spinal neurons by releasing neurotransmitters and neurotrophins, for example glutamate and brain-derived neurotrophic factor.11, 13 However, the effect of UCD on pERK expression was greater than on EMG activity in the present study, also indicating that the former may be more sensitive than the later in detecting noxious UCD stimulation. In addition, most pERK 1/2-IR neurons were primarily localized to the deep dorsal horn and central canal neurons, unlike the highly localized, superficial dorsal horn pERK 1/2 expression observed with somatic10, 11, 12, 13 and other visceral stimuli such as colorectal and gastric distension,15, 16 but similar to a previous study in which UCD increased spinal cFos expression mostly in the deep dorsal horn and central canal regions.8 Although yet to be confirmed, this discrepancy is likely to reflect that afferents from different viscera terminate in different spinal dorsal cord laminae.

Previous studies in rats indicate that the uterine cervix is dually innervated by both the hypogastric and pelvic nerves to enter the spinal cord through the T12-L3 and L6-S1 dorsal roots.21 However, behavioral studies as well as epidural anesthesia and paracervical blockade in human parturients favor the hypogastric nerve as relevant to labor pain.2 In the present study, although ERK 1/2 activation in lumbosacral segments slightly increased after UCD, a significant ERK 1/2 activation was observed only in T12-L2 segments. This is consistent with a previous study of UCD resulting in increases in activity of single afferents in the hypogastric nerve in rats,3 and further confirms that nociception from the uterine cervix is transmitted mostly via the hypogastric rather than pelvic nerves to the spinal cord.

ERK produces both short-term functional changes by post-transcriptional processing and long-term adaptive changes by increasing gene transcription. In acute pain conditions, the involvement of ERK in nociceptive processing is likely to result from post-translational regulation.10, 16 ERK can directly modulate some key membrane receptors and ion channels. For example the A-type potassium channel enhances excitability of neurons in the dorsal horn.22, 23 In contrast, ERK participates in chronic pain possibly via transcriptional regulation.12, 14 ERK can translocate into the nucleus and induce transcription of genes, such as prodynorphin and NK-1,11 both of which have been suggested to be involved in pain mechanisms. In the present study, spinal ERK 1/2 was significantly activated 10min after UCD stimulus, although it is entirely possible that ERK 1/2 activation occurred more rapidly. Other studies showed that activation of ERK in the dorsal horn was detected at 2min after various stimuli.15, 24 Given the rapid time frame of ERK 1/2 activation, ERK contributes to UCD-induced visceral pain by post-translational regulation.

Considering that activation of spinal ERK 1/2 peaked at 60min, with a slow decline but remained significant for 120min after UCD, a transcriptional role for ERK in UCD-induced nociceptive transmission is also possible. This may be possible with labor pain, since labor persists for several hours. At present, it is not clear whether spinal ERK activation contributes to labor pain by both post-translational changes and inducing transcription of genes, or the relative contribution of each as labor progresses. Future studies are needed to elucidate the exact mechanism by which spinal ERK acts in labor pain.

Under the current experimental conditions, that is acute, brief, and single distension, the UCD model is not a model of labor pain. It does, however, offer a simple animal model to study acute nociception from the uterine cervix. As the first step, the study using this UCD model will probably advance our understanding of labor pain more than other animal models of visceral noxious stimulation. In addition, although the female rats used in this study were at random points in the estrous cycle, UCD induced similar EMG activity changes regardless of estrogen treatment in ovariectomized rats,4, 6 allowing us to ignore influences of the estrus cycle.

In conclusion, this study demonstrates that noxious UCD can rapidly induce activation of the ERK in spinal dorsal horn neurons in rats and suggests a potential involvement of ERK activation in acute visceral pain arising from the uterine cervix. The spinal ERK pathway may therefore be a potential target for novel pharmacological intervention for this form of visceral pain, including pain during the first stage of labor.