Long-Term Visceral Hypersensitivity Following Induction of Chemical Colitis is Primarily Reduced via Kappa Opiate Pathways - A Comparative Study of Different Opioidergic Subtypes

Irritable bowel syndrome (IBS) is a gastrointestinal syndrome characterized by chronic abdominal pain and altered bowel habits in the absence of any organic cause. Although IBS represents 25 to 50 percent of all referrals to gastroenterologists [1] and is the second most frequent cause for work absenteeism after common cold [2]. It remains a chronic condition without any generally accepted and successful therapy for all IBS patients. Clinical studies have traditionally brought attention to colonic motility as potential cause for the symptoms associated with IBS. One study described 3cpm (contractions per minute) in the unstimulated colon in patients with IBS compared to 6cpm in controls [3]. Other studies have described selective hypersensitization of visceral afferent nerves in the gut in IBS patients based upon the observation that many patients with IBS have lower tolerance for rectal balloon distension than controls [4,5]. Interestingly, epidemiological studies have revealed in IBS patients an increased prevalence of persistent ARTICLE INFO ABSTRACT


Introduction
Irritable bowel syndrome (IBS) is a gastrointestinal syndrome characterized by chronic abdominal pain and altered bowel habits in the absence of any organic cause. Although IBS represents 25 to 50 percent of all referrals to gastroenterologists [1] and is the second most frequent cause for work absenteeism after common cold [2]. It remains a chronic condition without any generally accepted and successful therapy for all IBS patients. Clinical studies have traditionally brought attention to colonic motility as potential cause for the symptoms associated with IBS. One study described 3cpm (contractions per minute) in the unstimulated colon in patients with IBS compared to 6cpm in controls [3]. Other studies have described selective hypersensitization of visceral afferent nerves in the gut in IBS patients based upon the observation that many patients with IBS have lower tolerance for rectal balloon distension than controls [4,5]. Interestingly, epidemiological studies have revealed in IBS patients an increased prevalence of persistent alterations in visceral nociception following acute gastroenteritis.
The extent to which these observations might contribute to the pathogenesis of IBS in the general population remains unclear, as only a relatively limited number of patients have been studied.
Since IBS generally presents as a complex of symptoms, treatment is based on the predominant symptom and subtype (constipation vs. diarrhea vs. mixed vs. unclassified). Lifestyle and dietary modification are of cornerstone importance and considered first line treatment. Further, in patients with IBS and abdominal pain, antispasmodics and antidepressants, anxiolytics and antibiotics are used [6,7]. Limited evidence suggests that antienkephalinase agents, somatostatin analogues, alpha (2) receptor agonists, selective serotonin reuptake inhibitors, probiotics and herbal treatments may be useful in IBS patients [8]. However, some of the serotonin receptor agonists/antagonists used for causal therapy in IBS have been removed from the market, due to severe side effects such as ischemic colitis and serious cardiovascular complications. Hence at the time being, research is still in progress for a specific causal therapy with as little unwanted side effects as possible. Nociceptive mechanisms of potential interest in this context are opioidergic pathways, the glutamate receptor axis (e.g. NMDA receptor antagonists) as well as a combination of these two pathways. In the study of Burton and Gebhart the pain threshold to CRD in rats with or without acute inflammation of the colon was elevated after administration of the κ-opiate receptor agonist (ORA) U50.488 [9]. Others found that the analgesic effect of fentanyl can be potentiated by a co-application of ketamine [10].
Endogenous opioids are distributed throughout the whole organism and produce a diversity of biologic effects such as analgesia, motoric or stress responses. Opioids bind to three different receptor subtypes, the κ-, µ-and δ-opioid receptors (OR). κ-OR are mainly found in the tunica muscularis as well as in the myenteric plexus (Auerbach) where they have an impact on analgesia and motility. µ-OR of the gastrointestinal tract are primarily located in the tunica mucosa, submucosa, as well as in the submucosal plexus (Meissner). Finally, the δ-OR are found in the submucosal and myenteric plexus and the peripheral nerve fibers [11]. Given the hypomotile effect of opioids, their use in treatment of IBS may be accompanied by problematic unwanted effects such as constipation because of diminished intestinal passage and secretion [12,13]. Further gastrointestinal side effects may include emesis and nausea, but also reduced motility of the gallbladder and elevated tonus of the sphincter oddi. Thus, it would be desirable to identify an opioidergic substance with maximal analgesic potency and minimal side effects. This holds especially true because nausea, vomiting and constipation often represent symptoms of IBS patients. In our study, we focused on two main objectives. First, to reproduce the known long-term effect of a chemically induced colitis on visceral hypersensitivity to tonic distension in Lewis rats in an animal model with phasic colorectal distensions as stimulus in order to simulate the hyperalgesic status of IBS patients following acute gastrointestinal infection. Secondly, to study the impact of different opioid receptor agonists and the NMDA antagonist ketamine on long-term visceral hypersensitivity following induction of TNB colitis. We aimed to compare the clinically used µ-ORA's fentanyl and morphine, the δ-ORA SNC80 as well as the NMDA antagonist ketamine with the κ-ORA U50.488.

Animals
As laboratory animals male Lewis rats (Harlan, Horst, Netherlands) weighing between 200 and 320g are used. They housed individually in temperature-controlled rooms with a 12h dark-light cycle. Access to water and food was free. Before induction of the TNB colitis as well as before implantation of the electrodes, the animals were fasted for 24 hours. The state department of veterinary services approved the study.

Drugs
Every drug is given in a 1ml solution. When a combination of two drugs were used, each drug is given in a 0.5ml solution to reach the targeted volume of 1ml. The dose of each drug was chosen based upon the literature. For the κ-ORA 1S,2S-U50.488 (18mg/kg BW i.p. [14], Sigma-Aldrich Chemie, Buchs (SG), Switzerland) H 2 O dist.

Implantation of EMG Electrodes
To quantify the visceromotor response (VMR) to phasic colorectal distension (CRD), the activity of the abdominal wall musculature was recorded by electromyography (EMG). For this purpose, we implanted small-sized electrodes into the lateral abdominal wall of the animals following median laparotomy under combined anesthesia with ketamin hydro-chloride (1ml/kg BW i.p., Ketamin 10%®, Sanofi-Ceva, Duesseldorf, Germany) and xylazin hydrochloride (0,05ml/animal i.p., Xylazin 2%®, Sanofi-Ceva, Duesseldorf, Germany). As bipolar electrodes, teflon-insulated silver wires (Cooner Wire Co., Chatsworth, CA/USA) were used, passing through the abdominal wall musculature. After surgical intervention, animals could recover for one week. After this week, animals were put together in groups of five with food and water ad libidum.

Timetable
First forty male Lewis rats were divided into two groups of 20 animals, a TNB colitis (test) and a saline group (control). Five weeks after induction of a TNB colitis or colorectal injection of saline, the EMG electrodes were implanted. After a recovery period of one-week, phasic CRD experiments were started. The VMR to CRD was studied with eight test substances or their combinations (saline, U50.488, fentanyl, morphine, SNC80, ketamine, ketamine/ fentanyl or ketamine/mor-phine). In the process of the testing, the animals were given a recovery period of three days after each series of stimulation ( Figure 1). The order of the injected substances was identical in all animals and every substance was used in each animal.
Therefore, the sequence of administration was not randomised.  Standard Instruments, Karlsruhe, Deutschland). The VMR to CRDs in the rat has been extensively characterized by Ness [20]. Under slight anesthesia with isoflurane, the lubricant-coated balloon (sonography lubricant, central pharmacy of the university hospital of Berne, Switzerland) was inserted into the colorectum, positioned about 1cm proximal of the anus and fixed to the tail with adhesive tape. Afterwards, the animals were placed in a Boreman cage and the EMG electrodes connected with the EMG recording system. The study protocol consisted of phasic CRDs (every minute 40mmHg for a period of 30sec). Before the application of a test substance, we performed three CRD's as basic measurement (measurement A).

Measurement of the Visceromotor Response
The application of the test substance was followed by 15 CRD's. The first three and the last three CRD's were denoted measurement B and C, respectively. The first 3 CRDs and the subsequent generated 15 CRDs were 11 min apart. The EMG was recorded during the whole trial, starting 1 min before the first CRD and ending 1 min after the last CRD ( Figure 2). The drugs were administered intraperitoneal, either 1 min (Ketamine) after measurement A or 1min before measurement B (saline, 1S,2S-U50.488, £Fentanyl, morphine and SNC80).

Statistical Analyses
To evaluate the effect of a chemically induced colitis on visceral nociception, the difference in EMG activity (in µV) over time, resulting in an area under the curve (µV*time), between period C and B in the test group (TNB) versus the control group (saline) was compared.
In this first part of the experimental protocol, a possible adaptation in the course of time was analysed. Therefore, only the first and the last three distensions were studied. To quantify the impact of the different ORA's and ketamine, the differences in EMG activity between the measurement C and A were compared. Since ketamine is hypothesized to act only on activated pain pathways, the last three measurements of the repetitive distensions were evaluated.
This protocol was consistently applied to all the substances to compare the results. Therefore, a negative value represents a decrease of VMR and thus a decrease in visceral sensory function.
A mean value for the three consecutive distensions was calculated.
Due to the non-normal distribution, the Wilcoxon rank sum test as well as the Wilcoxon signed rank test for statistical analysis were used. All values are given as the median of the area under the curve.
A p-value of ≤ 0.05 was considered significant.

Influence of Ketamine on VMR in colitis Animals
The comparison of VMR to CRD during measurement A and C did not show any significant reduction after application of ketamine compared to saline (85 vs. 183; p=0.376) (Figure 4).

Influence of Ketamine Combined with µ-ORA on VMR in colitis Animals
The combination of ketamine with fentanyl caused a significantly

Influence of TNB Colitis on VMR to CRD
In our series of phasic CRD tests, it was possible to reproduce the already known influence of TNB colitis on the VMR to tonic distensions. Contrary to our earlier experiments, in the present study we modified our experimental set-up and used phasic distensions as stimulus. In the control animals, we observed a reduction of the EMG response during the 15 phasic CRD's after an i.p. application of saline. This finding can be explained by adaptation to the distension stimulus in healthy animals. This mechanism appears to fail in TNB-pretreated animals. In these latter animals, VMR to CRD raises significantly during the series of distensions.
Similarly, to our results, among others, Zhou et al. described an increase in sensitivity and hyperalgesia following the induction of a TNB colitis in the rat model [21]. This was tested by using mechanic and thermal stimulation of the hind-paws. Zhou et al. concluded that this somatic hypersensitivity reflects tonic impulse input from the inflamed colon along with central sensitization.  [24]. Other reports suggested that subcutaneous, but not central administration of κ-ORAs inhibited the VMR to CRDs [25].

Influence of Different ORA's on VMR in Colitis Animals
In their model, Larsson et al. used mice and distensions were performed with increasing as well as repeated steady distensions lasting 10s and with 5 min intervals. U-69.593 was used as κ-ORA.

Kitamura et al. showed that intravenous administration of U50.488
produced strong visceral antinociception in the un inflamed colon [26]. This effect was completely antagonized by naloxone. It has been shown that κ-, but not µ-and δ-ORAs, given intra-arterial attenuated the response to the noxious stimulus. For this purpose, Su et al. performed recordings of mechanosensitive pelvic nerve afferent fibres of the decentralized S1 dorsal root in anesthetized rats. This antinociceptive activity of κ-ORAs such as U50.488 during inflammation is partly mediated by NO [27]. The peripheral injection of U50.488H was enhanced by the injection of the NOreleasing agent FK409 and reversed by the administration of the NO scavenger PTIO [28]. Another group demonstrated that the peripheral antinociception induced by the κ-ORA bremazocine was mediated by L-arginine/NO/cGMP activation [29]. The long-term persistence of the analgesic effect after induction of chemical colitis could be explained by the theory that the corresponding receptors are either not localized or restricted to the site of inflammation or they are not affected by the colitis. Other groups suggest a peripheral upregulation of κ-opioid receptors in or following colonic inflammation [30][31][32].
µ-ORA Fentanyl: Following an intraperitoneal application of fentanyl, we observed a decreased pain inhibitory effect on visceral nociception compared to morphine. We are not able to explain why fentanyl leads to a lesser reduction in pain than morphine in our setup. Theoretically, the use of both ORA at different concentrations could have changed this finding. µ-ORA Morphine: In our study, the administration of morphine was associated with a significant pain reduction to CRD in TNBtreated animals. Galantine et al. discovered that the µ-opioid morphine is highly active as analgesic against peritoneovisceral pain states in rats [33]. Similar effects have been described earlier, where systemic injections of morphine dose-dependently inhibited the VMR to CRD in rats with uninflamed or inflamed colons [22]. δ-ORA SNC80: In our study, application of the δ-ORA SNC80 reduced the EMG response to the visceral stimulus. Although this effect did not reach statistical significance in our study due to the largely dispersed data, it supports earlier findings by Sengupta et al. who described in a behavioural study that the δ-ORA SNC80 attenuated the responses of pelvic nerve afferent fibers in an uninflamed as well as in an inflamed colon but with significantly greater potency in the inflamed colon [27]. The greater potency of κ-ORAs in the TNB-inflamed condition suggests a peripheral upregulation of κ-OR's in colonic inflammation. The existence of distinct subtypes of δ-opioid receptors, δ 1 and δ 2 , has been proposed by several research groups [32,34]. This hypothesis has been supported by the introduction of specific agonists and antagonists for both receptor subtypes. Bilsky et al. [35] reported that antinociception in response to intracerebroventricular application of SNC80 was fully antagonized by the DALCE/Cys-DELT combination but only partially blocked when these δ 1 -and δ 2receptor antagonists were separately given [36]. Therefore, these authors postulated that SNC80 is an unspecific δ-agonist. Pacheco et al. described the peripheral occurrence of δ 1 -and δ 2 -receptors and their antinociceptive function as result of the activation of the L-arginine/NO/cGMP pathway [37].

Influence of Ketamine on VMR in Colitis Animals
In our study, the intraperitoneal application of ketamine as non-competitive antagonist of the NMDA receptor [38,39] did not lead to any statistically significant decrease of the VMR to CRD. One possible explanation could be that the VMR in our pain

Colitis Animals
Nadeson et al. showed in their work the possibility to potentiate the impact of intrathecal fentanyl by additional administration of ketamine and to reduce the amount of fentanyl used by a coapplication of ketamine [10]. In our study, we were not able to increase the antinociceptive effect of fentanyl by adding ketamine when both drugs were given by a non-spinal way. This combination has also been studied by several other research groups with mainly contradictory results [41][42][43]. The reason for these confounding results remains unclear. In our experiments, we were not able to further reduce the VMR to CRD by co-application of ketamine. This discordance of findings might be a result of the different models.

Hirota et al. postulated a competitive interaction of ketamine with
the µ-and κ-OR as functional µ-and κ-OR-antagonist that causes a non-opioidergic analgesic mechanism [44].

Limitations
Variability of the data can at least partially be explained by the difference in behaviour of the animals even before the beginning of the measurements. Some were calm, while others were already agitated resulting in higher baseline electromyographic activity and EMG measurements. Some animals did not survive the entire protocol. Thus, the total number of animals in each test series differed between 11 and 15. Another potential weak point of the present study is the lack of dose-response curves for the different substances which were tested. The respective concentrations for each test substance were determined by literature search. The aim of a follow-up study must be to confirm the conclusions of our study by including dose-response curves. In this context, both the absolute and relative expression of opiate receptor subtypes in the colon should be quantified in colitis and control animals. Finally, all substances were given in the same order in every animal, therefore no randomisation was performed. Thus, an adaption of the test animals to the CRD stimulus can not be excluded.
The irritable bowel syndrome is a very common disease.
However, even in the year 2019, it cannot be treated satisfyingly. Postinfectious hyperalgesia may play a crucial role in the development of IBS at least in a subpopulation of IBS patients. The pathways of how to counteract this altered post-infectious visceral nociception have not been fully characterised yet. In earlier studies, we were able to demonstrate in Lewis rats that a chemically induced inflammation of the colon mucosa causes long-lasting visceral hyperalgesia in response to tonic visceral stimuli [45][46][47]. In our present series of tests, we were able to confirm these findings in an experimental set-up with phasic distensions. We also demonstrated that, at the applied doses, the κ-opioid receptor agonist U50.488 has the most potent analgesic effect on viscero-sensory function to CRD in Lewis rats following induction of a chemical colitis compared to the µ-ORA fentanyl and morphine, the δ-ORA SNC80 and the NMDA antagonist ketamine. Furthermore, we could show that the µ-ORA morphine also possesses a positive effect on the visceral hyperalgesia in the TNB-colitis rats. In our study, the application of ketamine alone or a combination of ketamine and the ORA's missed to reduce or further reduce the VMR compared to the application of opioids alone. To conclude, our experiments demonstrate a strong analgesic effect of the selective κ-ORA U50.488 on long-term nociception in response to colorectal stimulation in a chemically induced colitis model. This is particularly noteworthy given the fact that this opiate receptor agonist unfolds its analgesic effect without the unwanted side effects of other opiate subtypes.