Changes in beta-adrenergic neurotransmission during postoperative ileus in rat circular jejunal muscle

Authors


  • Parts of this work were presented at the Digestive Disease Week in Chicago, IL, May 17–22. 2011.

Address for Correspondence
Michael S. Kasparek, Department of Surgery, Ludwig-Maximilians-University, Marchioninistr. 15, 81377 Munich, Germany.
Tel: +49 89/7095 6561; fax: +49 89/7095 8894;
e-mail: michael.kasparek@med.uni-muenchen.de

Abstract

Background  To explore postoperative changes in β-adrenergic neurotransmission that participate in pathophysiology of postoperative ileus.

Methods  Contractile activity of circular jejunal muscle strips was studied. Groups (n = 6/group) were: naïve controls, sham controls 1 and 7 days after laparotomy, and rats 12 h, 1, 3, and 7 days after laparotomy with standardized small bowel manipulation (postoperative ileus). Dose-responses to the β-agonist isoprenaline (3 × 10−10–10−7 mol L−1) were studied in presence/absence of tetrodotoxin (global neural blockade; 10−6 mol L−1), N6-(1-iminoethyl)-l-lysine (inhibition of inducible nitric oxide synthesis; 10−4 mol L−1), nimesulide (cyclooxygenase-2 inhibition; 10−5 mol L−1), or propranolol (β-blockade; 5 × 10−6 mol L−1). Histochemistry for inflammatory cells and intestinal transit were studied.

Key Results  Intramural inflammation and delayed transit (postoperative ileus) occurred only in ileus groups. The inhibitory potential of isoprenaline decreased in all postoperative groups including sham (P < 0.05). Tetrodotoxin enhanced isoprenaline-induced inhibition in ileus and sham groups (P < 0.05). N6-(1-iminoethyl)-l-lysine and nimesulide decreased isoprenaline-induced inhibition in ileus groups 12 h, 1, and 7 days, and in sham controls 7 days postoperatively (P < 0.05). Propranolol prevented isoprenaline effects in all groups (P < 0.05).

Conclusions & Inferences  Inhibitory effects of isoprenaline on contractile activity were decreased for 7 days postoperatively. Changes in β-adrenergic neurotransmission do not induce postoperative ileus and appear to be caused by anesthesia and laparotomy rather than bowel manipulation.

Abbreviations:
anova

analysis of variance

AUC

area under the contractile curve

BL

baseline

COX

cyclooxygenase

CSA

cross-sectional area

FITC

fluorescein isothiocyanate

IC50

half maximal inhibitory concentration

iNOS

inducible nitric oxide synthase

KRB

Krebs-Ringer`s bicarbonate buffer

Lo

optimal length

L-NIL

N6-(1-iminoethyl)-l-lysine

MPO

myeloperoxidase

NC

naïve controls

NDS

normal donkey serum

NO

nitric oxide

NS

not significant

SC1d

sham controls 1 day postoperatively

SEM

standard error of the mean

P1d

animals 1 day after induction of postoperative ileus

PBT

phosphate-buffered saline with Triton X

TTX

tetrodotoxin

Introduction

Despite considerable clinical and basic science research in postoperative gastrointestinal motility, postoperative ileus remains poorly understood and is associated with increased patient morbidity and health care costs.1,2 Current thinking is that postoperative ileus is caused mainly by reflex inhibition of contractile activity via sympathetic nerves in the early phase (about the first 3 h postoperatively), although in the later phase hours and days after the operative procedure, an inflammatory response develops within the gut wall, initiating the release of mediators, like nitric oxide (NO) and prostaglandins that subsequently inhibit contractile activity.3 Furthermore, there is a growing understanding of the intense interaction of neuronal pathways and intramural inflammation during postoperative ileus. For instance, intramural inflammation can activate central inhibitory reflex pathways aggravating postoperative ileus4; in contrast, activation of the vagal anti-inflammatory pathway can decrease intramural inflammation and ameliorate postoperative ileus.5–8

The distribution of adrenergic receptors in the gastrointestinal tract varies widely between different species, different segments of bowel (e.g. stomach, jejunum, ileum, colon), and different cell types (e.g. smooth muscle, neurons, glia, inflammatory cells). Beta-receptors have been identified on presynaptic, noradrenergic terminals of the sympathetic nervous system,9 on neurons of the enteric nervous system,10,11 and on smooth muscle cells of the circular and longitudinal muscle layer of rat small intestine.11–13 The source of endogenous catecholamines acting on these receptors is either from the sympathetic nervous system releasing catecholamines locally or from systemic release from the adrenal glands, because the enteric nervous system itself has no adrenergic nerves and is unable to produce and release endogenous catecholamines. Moreover, sympathetic nerves are all extrinsic nerves and synapse mainly in the enteric nervous system; indeed, there is virtually no input from the sympathetic nervous system directly to the gut smooth muscle.14

In the past, postoperative ileus was mainly characterized in vitro by an impaired procontractile response to a cholinergic stimulus.15 Although a variety of neurotransmitters participate in coordination of smooth muscle contractile activity physiologically,14,16 changes in neurotransmission from other neurotransmitters during postoperative ileus have only rarely been studied. While functional studies by others demonstrated that stimulation of adrenergic receptors inhibits contractile activity,17–19 we studied postoperative changes in β-adrenergic neurotransmission on the contractile activity in rat small intestine. We hypothesized that β-adrenergic changes in neurotransmission occur during postoperative ileus and participate in its pathophysiology. These changes could be related to perioperative alterations in sympathetic tone and/or systemic catecholamine levels; such changes would be of particular interest, because specific agonists and antagonists for β-1, β-2, and β-3 receptors are now available, allowing pharmacologic interventions without the risk of detrimental cardiac or respiratory side effects.

Materials and methods

Preparation of animals

Procedures and animal care were performed according to the guidelines of the German Animal Welfare Act and were approved by the local Institutional Review Board (Regierung Oberbayern). Three month-old, male, Sprague-Dawley rats (Charles River, Sulzfeld, Germany) were used. Animals were maintained under a 12 h/12 h light/dark cycle with free access to normal rat chow and water until 12 h preoperatively when rat chow was withdrawn. Body weight was measured in all animals prior to operation (postoperative groups) and before tissue harvest and is expressed in [%] change compared to preoperative body weight.

Experimental groups

Combined effects of anesthesia, laparotomy, and small intestinal manipulation were studied using a well-established model to induce postoperative ileus.20,21 In brief, animals underwent anesthesia with 2% isoflurane (Abbott, Wiesbaden, Germany), after which a 3 cm midline laparotomy was performed, the entire small intestine was exteriorized on a moist gauze, and the small bowel from the Ligament of Treitz to the ileocecal junction was manipulated extensively with two, moist, cotton applicators. Thereafter, the intestine was returned into the abdominal cavity, and the abdominal wall was closed using a running suture. Animals recovered from anesthesia on a heating pad before they were returned to their cages, where they had free access to water and rat chow until 12 h prior to tissue harvest. Tissue harvest was performed under anesthesia with isoflurane, and animals were killed by transection of the abdominal aorta and exsanguination at the end of the procedure. In the ileus groups, tissue was harvested at four different time points: After 12 h, and at 1, 3, and 7 days after induction of postoperative ileus (groups: P12h, P1d, P3d, P7d; n = 6 per group).

Because bowel manipulation is a key factor in the induction of postoperative ileus,15,20 we used sham controls (SC) undergoing the above-mentioned procedure without intestinal exteriorization and manipulation to study the combined effects of anesthesia and laparotomy alone. To determine short- and long-term effects of sham laparotomy, SC were studied 1 day (SC1d) and 7 days (SC7d) after anesthesia and laparotomy (n = 6 per group). Non-operated rats without prior anesthesia or operation served as naïve controls (NC; n = 6).

Measuring gastrointestinal transit and tissue harvest

Intestinal transit was studied in all animals by gavage of 2 mL semisolid charcoal solution (10% charcoal and 10% gum arabicum in distilled water; Sigma-Aldrich, Steinheim, Germany) under a short anesthesia with isoflurane. Thirty minutes later, rats were anesthetized again by inhalation of isoflurane, and the abdominal cavity was opened. The entire small intestine from the Ligament of Treitz to the ileocecal junction was removed, and the distance from the oral end to the aboral end of the charcoal front as well as the overall length of the small intestine were measured in cm. Results are expressed as [%] of small intestinal length passed by charcoal at the time of tissue harvest.

Recording mechanical activity

After tissue harvest, a 10 cm jejunal segment beginning 10 cm distal to the ligament of Treitz was flushed and stored in chilled, modified Krebs-Ringer`s bicarbonate buffer (KRB; NaCl 116.4, KCl 4.7, CaCl2 2.5, MgSO4 1.17, KH2PO4 1.2, NaHCO3 23.8, glucose 11.1 and calcium disodium versenate 0.26; concentrations in mmol L−1) preoxygenated with 95% O2 and 5% CO2 (Linde, Pullach, Germany). The segment was opened along the mesenteric border and pinned flat in a Petri dish filled with chilled, preoxygenated KRB. Under microscopic exposure (Zeiss, Jena, Germany), the mucosa and submucosa were carefully removed, and standardized muscle strips (10 × 3 mm) were cut transversely in the direction of the circular muscle layer. A silk loop was tied to each end of the muscle strip, and the loops were attached to a fixed glass hook on one end and the other end to a noncompliant force transducer measuring isometric force. This attachment was immersed in a 10 mL organ chamber (Radnoti Glass Technology, Monrovia, CA, USA) filled with warm (37.5 °C), preoxygenated KRB.

Contractile activity was saved digitally on a computer using the interface PowerLab 8/30 Data Acquisition System (ADInstruments Inc., Colorado Springs, CO, USA). Dedicated software (LabChart 7; ADInstruments Inc.) was used for online visual control of raw tracings as well as for further analysis of contractility studies later on.

Experimental design

After an equilibration period of 10 min, all but 2% of muscle strips developed spontaneous contractile activity. Strips without spontaneous contractile activity were excluded from the experiment. Muscle strips were then stretched in 5 min intervals with intervening washouts of the bath solution until reaching optimal length (Lo), where further stretching did not change amplitude or frequency of contractions. After 10 min of equilibration at Lo, spontaneous contractile activity was measured for 3 min in all muscle strips. Thereafter, each experimental condition was studied in two strips per rat.

Response to the non-specific β-agonist isoprenaline on spontaneous contractile activity and after precontraction

The nonspecific β-receptor agonist isoprenaline was used to determine whether or not postoperative changes in β-adrenergic neurotransmission occur, and cumulative dose-responses to isoprenaline (3 × 10−10–10−7 mol L−1) were studied in two muscle strips per rat in all groups. After establishing stable contractile activity at Lo, the effect of isoprenaline was studied on spontaneous contractile activity. After washout of the bath solution and restitution of contractile activity, all further dose-response experiments with these two muscle strips were performed after precontraction with bethanechol (3 × 10−6 mol L−1) for 10 min to achieve stable and reproducible conditions after washout and repeated administration of the β-receptor agonist. First, dose-dependent effects of isoprenaline were studied on stimulated contractile activity (after precontraction). Thereafter, different substances were used to determine the involvement of different mechanisms in changes in β-adrenergic neurotransmission during postoperative ileus. Immediately after the next washout, 10−4 mol L−1 N6-(1-iminoethyl)-l-lysine dihydrochloride (L-NIL) and 10−5 mol L−1 nimesulide were both added simultaneously into the organ bath. These drugs were used to inhibit inducible NO synthase (iNOS) and cyclooxygenase-2 (COX-2), respectively, enzymes that are known to be upregulated during postoperative ileus and which produce NO and prostaglandins, respectively, which inhibit contractile activity.22–24 Dose-responses to isoprenaline were studied 10 min after preincubation with L-NIL and nimesulide followed by 10 min of precontraction. After a final washout, propranolol (5 × 10−6 mol L−1) was added into the organ chamber. Ten minutes later, muscle strips were precontracted for 10 min, and dose responses to isoprenaline were studied to determine the β-receptor-specific effect of isoprenaline.

Two different muscle strips per rat were used to study whether isoprenaline effects were mediated via β-receptors on the enteric nervous system or the smooth muscle by using tetrodotoxin (TTX) to prevent the depolarization-dependent release of neurotransmitters from the enteric nervous system. Therefore, after establishing stable spontaneous contractile activity, 10−6 mol L−1 TTX was added into the bath solution, and dose-responses to isoprenaline were studied 10 min later.

Histochemistry and immunohistochemistry

A 4 cm segment of jejunum immediately proximal to the segment used for contractility studies was flushed with chilled KRB. After removing the mucosa and submucosa, the tissue was stretched 150% in length and 250% in width and was fixed in 100% ethanol for 10 min, followed by two washing steps in 0.01 mol L−1 phosphate buffered saline with 0.3% Triton X (PBT). All histologic studies were performed with these whole mounts. To stain for myeloperoxidase (MPO)-positive cells (neutrophils and monocytes), fixed whole mounts were incubated for 10 min in a 1 : 10 dilution of 10 mL KRB, 10 mg Hanker-Yates-reagent, and 100 μl hydrogen peroxide.25 Mast cells were stained by overnight incubation with fluorescein isothiocyanate (FITC)-avidin (1 : 10 in 10% normal horse serum) at 4 °C followed by three 15 min washouts in PBT.15 Staining for macrophages was performed by preincubation of whole mounts in 5% normal donkey serum (NDS in PBT for 2 h at room temperature) followed by incubation with the primary antibody (ED2, mouse anti-rat CD 163 1 : 100 in PBT with 2.5% NDS) over night at 4 °C. After three washouts in PBT, specimens were incubated with the secondary antibody (donkey anti-mouse 1 : 500 in PBT) for 4 h at room temperature. After staining, whole mounts were coverslipped and sealed with quick-hardening Eukitt medium (Kindler, Freiburg, Germany); specimens were examined with light microscopy (BX41; Olympus, Essex, UK) or fluorescent microscopy (Axiophot; Zeiss, Feldbach, Switzerland) as appropriate. Inflammatory cells were counted in 15, randomly chosen areas. Myeloperoxidase-positive cells were counted at 200-fold magnification; mast cells and macrophages were counted at 400-fold magnification. Data are expressed as number of cells per square millimeter.

Data analysis

Integral, frequency, and amplitude of contractions were calculated using dedicated software (LabChart 7; ADInstruments Inc.). Spontaneous contractile activity was measured for 3 min for each strip at Lo, and the integral was expressed as (g mm−2 s−1) with mm2 reflecting cross-sectional area (CSA) that was calculated by the following equation after measuring length and wet weight of each muscle strip at the end of the experiment: CSA (mm2) = tissue wet weight (mg)/tissue length (mm) × tissue density (mg mm−3). Tissue density was defined as 1.03 mg mm−3 according to previous reports.26,27 Effects of cumulative doses of isoprenaline were quantified in 3 min intervals immediately after administration of each dose and were compared with baseline activity 3 min before the first dose of isoprenaline was administered. For the experiments involving precontraction, muscle strips were incubated with bethanechol for 10 min; baseline contractile activity was measured during the last 3 min of precontraction immediately before the first dose of isoprenaline was administered. Because conditions were studied in two strips per rat, spontaneous contractile activity and responses were averaged between both strips. Drug responses are presented as [%] change from baseline activity (defined as 0%), which represents contractile activity 3 min before the first administration of isoprenaline. Negative values represent decreases in mechanical activity.

Statistical comparisons

Data are summarized as mean ± standard error of the mean. Analysis of variance (anova), repeated measures anova, and post hoc Student′s t-tests or two-sample t-tests were used, which ever was appropriate. A Bonferroni correction was applied when evaluating statistical significance of multiple t-tests, and anova on Ranks was used when data were not distributed normally. Although raw data of dose-response curves were used for statistical comparison of groups and treatments by anova, half maximal inhibitory concentrations (IC50), representing the negative log of molar value resulting in 50% inhibition of contractile activity were calculated from pooled data of each group to simplify data presentation and to facilitate interpretation.

Drugs

Isoprenaline, bethanechol, L-NIL, nimesulide, propranolol, NDS, and FITC-Avidin were purchased from Sigma-Aldrich, ED2 antibody from Abd Serotec (Düsseldorf, Germany), secondary antibody donkey anti mouse from Millipore (Temecula, CA, USA), tetrodotoxin from Tocris (Ellisville, MO, USA), and Hanker-Yates reagent from Polysciences (Warrington, PA, USA).

Results

Postoperative body weight

Body weight decreased early postoperatively in animals after induction of postoperative ileus as well as in SC when compared to their preoperative bodyweight (P12h, P1d, SC1d; each P < 0.05; Fig. 1); all other animals showed, however, an increase of body weight 7 days postoperatively (P < 0.05). By 3 days after ileus induction, body weight of the P3d was comparable to the preoperative weight (P = NS).

Figure 1.

 Postoperative body weight decreased early postoperatively when compared to preoperative body weight, whereas animals 7 days postoperatively showed an increase of body weight. *P < 0.05 vs preoperative body weight. N/A, not applicable; NC, naïve controls; P12h, P1d, P3d, P7d: 12 h, 1d, 3d, and 7d after induction of postoperative ileus; SC1d, SC7d: 1d and 7d after sham operation.

Intestinal transit

In NC, charcoal solution passed to 54 ± 5% of the small intestinal length. Sham operation had no effect on small bowel transit when measured at 1 (SC1d) and 7 (SC7d) days postoperatively (P = NS). In contrast, transit was delayed in all groups after induction of postoperative ileus up to postoperative day 7 with a more pronounced effect in early postoperative groups (all P < 0.05; Fig. 2A).

Figure 2.

 (A) Small intestinal transit (% of small bowel length passed by charcoal marker) was delayed in all groups after induction of postoperative ileus, although sham operation had no effect on gastrointestinal transit. (B) Histochemistry and immunohistochemistry: MPO-positive cells and mast cells were increased 12 h after ileus induction with a subsequent decrease to almost normal counts at day 7. In contrast, macrophage counts increased gradually and were still increased at day 7. Sham operation did not cause any changes in cell counts 1 and 7 days postoperatively. *P < 0.05 vs NC. NC, naïve controls; P12h, P1d, P3d, P7d: 12 h, 1d, 3d, and 7d after induction of postoperative ileus; SC1d, SC7d: 1d and 7d after sham operation.

Histochemistry and immunohistochemistry

Myeloperoxidase-positive cells and mast cells were increased 12 h after ileus induction with the greatest values at 12 h (46-fold increase) and 1 day (55-fold increase), respectively (both P < 0.05, Fig. 2B). Thereafter, these cell counts decreased and reached baseline levels 7 days postoperatively (P7d) for MPO-positive cells, whereas mast cell counts remained slightly increased at day 7. In contrast, macrophage counts increased progressively up to day 7 (fourfold increase; P < 0.05). Sham controls (who underwent no bowel manipulation) did not show any changes in any of these cell counts at 1 and 7 days after sham operation (P = NS).

Spontaneous contractile activity

Spontaneous contractile activity differed between groups. When compared to NC, spontaneous activity in both the ileus and SC groups was increased when measured early (P12h, SC1d) and late (P7d, SC7d) (P < 0.05; Fig. 3). In contrast, 1 (P1d) and 3 days (P3d) after ileus induction, spontaneous contractile activity was comparable to NC (P = NS).

Figure 3.

 Spontaneous contractile activity was increased early and late after induction of postoperative ileus and in sham controls. *P < 0.05 vs NC. NC, naïve controls; P12h, P1d, P3d, P7d: 12 h, 1d, 3d, and 7d after induction of postoperative ileus; SC1d, SC7d: 1d and 7d after sham operation.

Response to non-specific β-agonist isoprenaline

Isoprenaline inhibited spontaneous contractile activity dose-dependently in all groups at all times (P < 0.05; Fig. 4A). The isoprenaline-induced inhibition of the area under the contractile curve (integral) originated from a profound inhibition of the amplitude of contractions, while the frequency of contractions remained unaffected (Fig. 4B). Because this pattern was similar in all groups, data are shown only for NC.

Figure 4.

 (A) Effect of isoprenaline on contractile activity in experimental groups (representative raw tracings). NC, naïve controls; P12h, P1d, P3d, P7d: 12 h, 1d, 3d, and 7d after induction of postoperative ileus; SC1d, SC7d: 1d and 7d after sham operation. (B) Effect of isoprenaline on area under the contractile curve (AUC), amplitude, and frequency of contractions in NC. Isoprenaline decreased the amplitude of contractions without changing the frequency of contractions, thereby causing a decrease in the area under the contractile curve (P < 0.05). *P < 0.05 for dose-dependent inhibition. BL, baseline.

When compared to NC, however, the inhibitory effect of isoprenaline on spontaneous activity was attenuated in all postoperative groups both after induction of ileus as well as after sham laparotomy (P < 0.05; Fig. 5A). The decrease in the inhibitory potential of isoprenaline was most pronounced late after ileus induction in P7d. Tetrodotoxin had no effect on the response to isoprenaline in NC; in contrast, TTX enhanced the isoprenaline-induced inhibition in all groups after ileus induction as well as after sham laparotomy (all P < 0.05; see Fig. 5B). In the presence of TTX, the effect of isoprenaline was still decreased compared to NC in all groups except in P1d.

Figure 5.

 Effect of isoprenaline on spontaneous contractile activity. (A) Contractile activity was inhibited by isoprenaline in all groups. This inhibitory potential was attenuated in all postoperative groups, but the inhibitory effect was attenuated most strongly in P7d. (B) Tetrodotoxin attenuated these postoperative changes in all groups (see table). *P < 0.05 vs NC; + P < 0.05 vs all seven other groups after ileus induction; §P < 0.05 vs control response without TTX. BL, baseline; NC, naïve controls; P12h, P1d, P3d, P7d: 12 h, 1d, 3d, and 7d after induction of postoperative ileus; SC1d, SC7d: 1d and 7d after sham operation.

All other experiments were carried out after precontraction of muscle strips with bethanechol. Precontraction enhanced isoprenaline-induced inhibition only in P7d (IC50: 7.78 ± 0.1 vs 8.02 ± 0.2; P < 0.05), although precontraction had no effect on isoprenaline responses in all other groups. In contrast, after precontraction, the decreased effects of isoprenaline in postoperative groups (compared to NC), as was seen on spontaneous contractile activity, was observed only in P12h, P1d, SC1d, and SC7d. When isoprenaline effects after precontraction were studied in the presence of both the iNOS inhibitor L-NIL and the COX-2 inhibitor nimesulide, the inhibitory potential of isoprenaline was decreased compared to the responses without L-NIL and nimesulide in P12h, P1d, P7d, and SC7d (all P < 0.05; Fig. 6A). The β-receptor antagonist propranolol prevented isoprenaline-induced inhibition in all groups, but there was a remaining inhibition ranging from −15 ± 4 to −41 ± 6% at the maximum dose of isoprenaline (10−7 mol L−1); however, significant inhibition of contractile activity persisted in P12h, P1d, and P3d (all P < 0.05; Fig. 6B).

Figure 6.

 (A) Effect of isporenaline after inhibition of i-NOS and COX-2 with L-NIL and nimesulide in precontracted muscle strips. (B) L-NIL and nimesulide decreased isoprenaline-induced inhibition in all postoperative groups except in P3d and SC1d. *P < 0.05 vs NC, §P < 0.05 vs control response without L-Nil and nimesulide. (C) The effect of isporenaline in the presence of propranolol in precontracted muscle strips. The isoprenaline effect was decreased in all groups with a persisting inhibition in P12h, P1d, and P3d. *P < 0.05 for dose-dependent inhibition. BL, baseline; NC, naïve controls; P12h, P1d, P3d, P7d: 12 h, 1d, 3d, and 7d after induction of postoperative ileus; SC1d, SC7d: 1d and 7d after sham operation.

Discussion

The aim of our study was to delineate postoperative changes in β-adrenergic neurotransmission that might participate in the pathophysiology of postoperative ileus in a rat model. Our key finding was that abdominal surgery appears to cause a long-lasting desensitization to isoprenaline (β-agonist) that appears to increase up to day 7 after surgery. The changes we observed were independent of the operative procedure performed as well as independent of the ileus-inducing bowel manipulation and not clearly associated with typical features of postoperative ileus such as intramural inflammation and delayed gastrointestinal transit; therefore, the changes we observed did not appear to be specific for postoperative ileus, but rather seem to depend on effects of the anesthesia and laparotomy.

These long-lasting changes in β-adrenergic neurotransmission shown in our study might be the result of two different mechanisms: first, the desensitization to β-adrenergic stimuli that we observed may be caused by increased systemic catecholamines released from the adrenal glands or second, by a long-lasting change in sympathetic tone after abdominal surgery. Considering the first possible explanation, abdominal surgery causes an increase in plasma catecholamine levels lasting from hours up to several days after the operation.28–33 Moreover, catecholamines may impair gastrointestinal motility as it is known from other diseases like pheochromocytoma,34 and from in vivo studies of α- and β-agonists on contractile activity.35 Furthermore, there is some evidence that β-receptor blockade might shorten postoperative ileus in humans.36–38 Whether the postoperative increase in catecholamines causes a down-regulation of β-receptors or their down-stream signaling pathways to explain this apparent desensitization, has not been determined yet.

The other cause for the long-lasting changes in β-adrenergic neurotransmission we observed might be a long-lasting alteration in sympathetic tone. Others have described sympathetic mechanisms that appear to play an important role in early postoperative ileus within a few hours after the surgical procedure.39,40 Moreover, it has been known since the beginning of the last century that postoperative ileus can be prevented in part by sympathectomy or splanchnic anesthesia.41–50 Different nociceptive stimuli activate distinct inhibitory neural pathways. Although the skin incision and the laparotomy cause a brief inhibition of contractile activity via a spinal reflex mediated by adrenergic pathways,39,51 intestinal manipulation activates central neuronal stuctures (e.g. supraoptic nucleus and hypothalamus) by release of corticotrophin-releasing factor. This process leads to activation of spinal efferents that cause inhibition of contractile activity by non-adrenergic neurotransmitters, such as NO or vasoactive intestinal polypeptide.39,52,53 In our study, SC and animals after bowel manipulation share the nociceptive stimulus of skin incision and laparotomy with inhibition of gastrointestinal motility via adrenergic pathways, which may initiate the observed long-lasting alterations in β-adrenergic neurotransmission. Recent studies have shown long-lasting changes in sympathetic innervation in humans. Amar et al. demonstrated that both thoracic and abdominal surgery cause an impairment of heart rate variability, which serves as a marker for increased sympathetic activity that lasts at least up to postoperative day 654 and is accompanied by a down-regulation of β-receptors and cyclic adenosine monophosphate production in blood lymphoctes.55 This observation demonstrates that abdominal surgery has a long-lasting effect on the autonomous nervous system and that these changes can be associated with alterations in adrenergic receptor expression and activity of down-stream signaling pathways in effector cells of the sympathetic nervous system.

We also performed experiments to: first, explore potential changes in the function of the enteric nervous system; second, identify alterations in pacemaker activity; and third, study whether the inflammatory mediators NO and prostaglandins participate in postoperative changes of the isoprenaline effect.

We used TTX to differentiate the effects of isoprenaline on enteric nerves vs direct effects on smooth muscle cells. Our data suggest that in unoperated animals, the inhibitory effect of isoprenaline is mediated primarily via β-receptors located on smooth muscle cells to which, under in vivo conditions, systemically released catecholamins might bind. In contrast, the inhibitory potential of isoprenaline in postoperative animals was decreased, and TTX restored the isoprenaline sensitivity at least in part. It is unlikely that TTX prevented presynaptic, pro-contractile β-receptor-mediated effects; such effects in non-sphincteric regions of the gastrointestinal tract have —to the best of our knowledge —not yet been described. More likely, there is a tonic procontractile stimulation from the enteric nervous system that appears to be more pronounced in the postoperative period (as evidenced by the increase in spontaneous contractile activity in four of our groups), such that blockade of this procontractile stimulus might subsequently enhance and restore the inhibitory potential of isoprenaline. We observed a comparable phenomenon in a previous study, in which TTX increased the inhibitory potential of hydrogen sulfide, another endogenous inhibitory agent.56 However, these postoperative changes in the function of the enteric nervous system are not specific for postoperative ileus, as they were observed also in SC.

We also aimed to determine whether or not the pacemakers of contractile activity, the Interstitial Cells of Cajal are involved in the inhibitory effect of isoprenaline. We analyzed not only area under the contractile curve but also the amplitude and frequency of contractions. We observed that the frequency of contractions was unaffected, whereas isoprenaline dose-dependently decreased the amplitude of contractions in all groups. Although others have suggested that pacemaker activity of Interstitial Cells of Cajal can be inhibited by α-adrenergic agonists, we conclude from our data that stimulation of β-receptors has no effect on pacemaker function of these cells and that changes in the pacing activity of these cells are not responsible for postoperative alterations of the isoprenaline effect.57

We used selective inhibitors of iNOS and COX-2 to study potential effects of NO and prostaglandines that are released during the intramural inflammatory response and delay gastrointestinal transit in postoperative ileus.22,24,58 N6-(1-iminoethyl)-l-lysine and nimesulide decreased the isoprenaline-induced inhibition early (P12h and P1d) and late after ileus induction (P7d). This finding suggests that NO and prostaglandins released from their inducible enzymes enhance inhibitory β-adrenergic effects in the postoperative period. Surprisingly, a similar effect was also seen in SC 7 days postoperatively, in which neither intramural inflammation nor delayed gastrointestinal transit was present. However, even after blocking iNOS and COX-2, the isoprenaline effect remained reduced compared to NC in P12h, P1d, and SC7d. Therefore, it is very likely that effects of NO and prostaglandins interact with or are even overwhelmed by postoperative changes in presynaptic neurotransmission, like that mentioned above; however, we cannot comment further on this potential interaction, because we did not perform studies using TTX, L-NIL, and nimesulide at the same time.

In conclusion, we demonstrated that abdominal surgery causes long-lasting changes in β-adrenergic neurotransmission that are rather caused by alterations in tonic procontractile stimulation from the enteric nervous system than by inflammatory mediators such as NO and prostaglandins; however, these changes do not appear to be specific for postoperative ileus. Nevertheless, these alterations might participate in other changes of postoperative bowel function, e.g. abdominal discomfort, cramps, and bloating, that are not necessarily associated with delayed gastrointestinal transit, but still impact on patients′ well-being.

Acknowledgment

We thank Prof. Michael G. Sarr, MD for his valuable input to this manuscript.

Funding

This work was supported by a grant from the Deutsche Forschungsgemeinschaft, Germany, (KA 2329/5-1 to MSK).

Disclosure

No competing interests declared.

Author Contribution

BG and PB performed the research; BG and MSK designed the research study; BG and MSK analyzed the data; BG, MHM, MEK, and MSK wrote the paper.