Pathophysiology of Postoperative Ileus: from Bench to Bedside

Transcription

Pathophysiology of Postoperative
Ileus: from Bench to Bedside
F
Frans Olivier The
Pathofysiology of postoperative ileus:
from bench to bedside
Thesis University of Amsterdam
© 2008 Frans O. The, Amsterdam, the Netherlands
All rights reserved. No part of this publication may be reproduced or transmitted in any form or by
any means, electronic or mechanical, including photocopy, recording or any information storage
and retrievel system, without written permission of the author.
The research discribed in this thesis was carried out by the department of Gastroenterology and
Hepatology, Academic Medical Center, Amsterdam, the Netherlands and was supported by the
Technology Foundation STW, applied science division of NWO and the technology program of the
ministry of Economic Affairs (NWO-STW, grant nr AKG.5727).
Edited by: R.A. de Leeuw, idEAct®, Amsterdam, the Netherlands
Printed by: Buijten & Schipperheijn, Amsterdam, the Netherlands
Pathophysiology of Postoperative
Ileus: from Bench to Bedside
ACADEMISCH PROEFSCHRIFT
ter verkrijging van de graad van doctor
aan de Universiteit van Amsterdam
op gezag van de Rector Magnificus
prof. dr. D.C. van den Boom
ten overstaan van een door het college voor promoties ingestelde
commissie, in het openbaar te verdedigen in de Agnietenkapel
op dinsdag 5 februari 2008, te 10.00 uur door
Frans Olivier The
geboren te Groningen
P
Promotiecommissie:
Promotor: Prof. dr. G.E.E. Boeckxstaens
Co-promotor: Dr. W.J. de Jonge
Overige leden: Prof. dr. J.C. Kalff
Prof. dr. D. Grundy
Prof. dr. J.F.W.M. Bartelsman
Prof. dr. R.M. Buijs
Prof. dr. M.W. Hollmann
Prof. dr. W.A. Bemelman
Prof. dr. M.P.M. Burger
Faculteit der Geneeskunde
V
Voor mijn ouders
en Willemijn
Chapter 1
8
General Introduction
Table of contents
Chapter 2
16
Postoperative Ileus Is Maintained by Intestinal Immune
Infiltrates That Activate Inhibitory Neural Pathways in Mice
Gastroenterology 2003; 125: 1137-1147
Chapter 3
44
The ICAM-1 antisense oligonucleotide ISIS-3082 prevents the
development of postoperative ileus in mice
British Journal of pharmacology 2005; 146: 252-258
Chapter 4
64
Chapter 5
94
The vagal anti-inflammatory pathway attenuates intestinal
macrophage activation and inflammation by nicotinic
acetylcholine receptor mediated activation of Jak-2/Stat-3.
Nature Immunology 2005; 6: 844-851
Activation of the Cholinergic Anti-Inflammatory Pathway
Ameliorates Postoperative Ileus in Mice
Gastroenterolgy 2007; 133: 1219-1228
Chapter 6
Central activation of the cholinergic anti-inflammatory pathway shortens postoperative ileus in mice
Submitted for publication
118
Chapter 7
Mast Cell Degranulation During Abdominal Surgery Initiates
Postoperative Ileus in Mice
136
Chapter 8
Intestinal handling induced mast cell activation and
inflammation in human postoperative ileus
Gut 2008; 57: 33-40
166
192
Chapter 9
Mast Cell Stabilization as Treatment of Postoperative Ileus:
a Pilot Study
Submitted for publication
216
Chapter 10
Summery and conclusions
228
Chapter 11
samenvatting en conclusies
dankwoord
colour figures
1
1
Chapter
1
General introduction and aim
of this thesis
P
General introduction and aim of this
thesis
Postoperative ileus is a transient motility disorder characterized by impaired gastrointestinal
propulsion in the absence of any mechanical obstruction1. Every abdominal surgical
procedure is followed by some degree of hypomotility and gastrointestinal dysfunction2. The
patient endures nausea, vomiting, abdominal cramping and does not tolerate oral food or
fluid intake3. Besides this considerable discomfort experienced by patients, postoperative
ileus is also an important risk factor for complications such as aspiration pneumonia or
wound dehiscence and subsequently prolongs the duration of hospital admission4, 5. In the
US the annual expenses related to post-operative ileus exceed 1 billion dollars, reflecting
its socio-economical impact3.
Post-operative ileus is still considered inevitable2 and preventative therapeutic strategies
are lacking. In addition, (symptomatic) treatment options have barely improved over the
last decade6. In general, patients are deprived from oral food or fluids until first peristalsis
(a surgeon’s symphony) occurs. Upon this first empirical hallmark oral fluids are cautiously
reintroduced followed by gradual extension of oral intake. Nasogastric decompression
introduced by Wagensteen in 19317 was one of the first and only alleviating therapeutic
interventions and is still the most commonly used strategy combined with iv fluids and
nothing by mouth. Unfortunately, this approach only relieves symptoms and does not
shorten let alone prevent post-operative ileus.
Post-surgical disturbances in gastrointestinal propulsion have been described as early as
the late 19th century when Bayliss and Starling discovered that splanchnic denervation
improves contractility of the gut after laparotomy8. Since then, numerous studies have been
conducted attempting to identify the exact pathophysiological mechanism. Most of these
studies have focused on (autonomous) neurogenic and (stress) hormonal factors1, 9. It is
generally believed that opening of the abdominal cavity and manipulation of the intestines
during surgery activates both somatic and visceral nerve fibers triggering inhibitory neural
pathways10, 11. These inhibitory reflexes are now generally thought to be responsible for the
post-surgical delay in gastrointestinal propulsion12-15. As a consequence, many prokinetic
10
overcome this neurogenic inhibitory pathway16-20. However, this strategy has proven rather
ineffective in most clinical trials17, 21. Most likely, this approach has failed, as it is indeed
ineffective to step on the gas without removing the brake. Moreover, postoperative ileus
Chapter 1
drugs have been evaluated to stimulate gastrointestinal motor activity and as such to
usually lasts several days, a fact that cannot be explained by activation of visceral nerve
fibers during or immediately after surgery alone. Indeed, once the abdomen is closed,
stimulation of mechano- or pain receptors ceases and other mechanisms should come into
play.
Recently Kalff et al. have shown that in rodents, handling of intestinal loops during
abdominal surgery triggers a mild inflammatory response22 This inflammation is restricted
to the muscularis propria and leads to impaired muscle contractility and subsequent
delayed intestinal transit. This reduction in neuromuscular function develops 4 to 6 hours
after surgery and lasts for more than 24 hours in rodents, most likely explaining why postoperative ileus can last for several days. Postoperative ileus however, is not restricted to
the small intestine but involves the entire gastrointestinal tract2. One possible explanation
could be that this local inflammation triggers neural pathways affecting the entire gut.
Several studies have indeed shown that epidural infusion of anesthetics shortens ileus23
indirectly suggesting the involvement of a spinal inhibitory neural pathway. In chapter 2 we
evaluated this hypothesis in a mouse model of postoperative ileus.
If inflammation induced by surgical handling is indeed an important pathophysiological
mechanism, more insight in the players and mediators involved is crucial for the development
of drugs interfering with this pathway. One of the eminent events in any inflammatory
response is the extravasation of immune cells from the circulation into the targeted area,
i.e the area of the intestine that has been manipulated. One of the first events leading to
extravasation of leukocytes is the upregulation of adhesion molecules, such as Leukocyte
Function-associated Antigen-1 (LFA-1) and InterCellular Adhesion Molecule-1 (ICAM-1) 24,
. Agents that interfere with this process reduce inflammation and might therefore represent
interesting tools to shorten post-operative ileus. We tested the potency of antibodies and
anti-sense oligonucleotides targeting ICAM-1 to prevent the influx of inflammatory cells
into the manipulated area and as such shorten postoperative ileus (chapter 3).
11
25
Although it seems obvious that manipulation of the intestine is the trigger of the inflammatory
response and therefore should be minimized, it still remains crucial to identify the
mechanism leading to the upregulation of ICAM-1 and other adhesion molecules. Kalff et al
demonstrated that manipulation of the intestine leads to activation of resident macrophages,
a key event in the attraction of leukocytes26. Interestingly, Borovikova et al. reported that
the activation of macrophages by endotoxin can be reduced by vagus nerve stimulation
in a sepsis model27. They demonstrated that this effect is mediated by acetylcholine, the
neurotransmitter released by the vagus nerve, interacting with the alpha7 nicotinic receptor
on the macrophage28. Nicotine indeed dampened macrophage activation by LPS in vitro
leading to a reduction in the release of pro-inflammatory cytokines. Especially as the
gastrointestinal tract is under strict control of the vagus nerve, we explored whether the
anti-inflammatory properties of vagus nerve stimulation also apply to the gastrointestinal
tract (chapter 4), and could represent a powerful tool to reduce inflammation induced by
intestinal manipulation. In chapters 4, 5 and 6, we studied the effect of peripheral and
central activation of the vagus nerve and identified the intracellular signal transduction
pathway mediating the anti-inflammatory effect of nicotine receptor activation in the
macrophages.
Although interference with macrophage function is certainly an interesting therapeutic
approach, an even more preferable strategy would be to prevent macrophage activation
during surgery. The exact mechanisms involved are far from elucidated and subject of
ongoing studies, but one of the most likely triggers is undoubtedly the influx of bacteria.
Schwartz et al. indeed showed that intestinal manipulation correlates with a transient barrier
dysfunction which results in fluorescent micro-sphere translocation29. These micro-spheres,
mimicking luminal bacteria, can be found in mesenteric lymph vessels and monocytes
recruited to the handled gut wall. Based on these findings, we reasoned that this brief
increase in intestinal permeability results from mast cell activation. Intense stimulation
of afferent nerve fibers indeed leads to local release of Calcitonin Gene-Related Peptide
(CGRP) and substance P30, mast cell activation31, and attraction of inflammatory cells,
a mechanism known as neurogenic inflammation32. As mast cells play a central role in
this process and are known to increase mucosal permeability33, 34, we investigated their
possible role in postoperative ileus in chapter 7.
12
increased considerably creating many opportunities to improve the current treatment of
postoperative ileus. It should be emphasized though that these conclusions are based
on animal studies, and therefore not automatically apply to the human situation. For this
Chapter 1
Based on the studies described in Chapters 2 and 7, the insight in the pathogenesis has
reason, we designed a series of studies evaluating our hypothesis in man. In chapter 8 we
focused on mast cell degranulation, pro-inflammatory mediator release and subsequent
neutrophil influx in response to surgical bowel handling. We compared the extent of mast
cell activation and inflammation during a conventional laparotomy with that of a minimal
invasive surgical procedure. In addition, in-vivo intestinal leukocyte recruitment was
visualized using leukocyte-SPECT scans and post-operative recovery was evaluated in
open and minimal invasive surgical patients.
Finally, in chapter 9 we conducted a randomized double-blind proof of principle study
evaluating the role of mast cell stabilization in the treatment of post-operative ileus in
patients. In summary, the present thesis focuses on the pathogenesis of postoperative
ileus, an iatrogenic disorder with a significant morbidity and economic impact. We have
demonstrated that in contrast to earlier believes, postoperative ileus is a local inflammatory
disorder. We identified the cells of the innate immune system that are involved and evaluated
new therapeutic approaches and their mechanism of action. Finally, the animal data were
translated to the human situation and a first step to clinical application was undertaken.
13
Reference List
1. Livingston EH, Passaro EP, Jr. Postoperative ileus. Dig.Dis.Sci. 1990;35:121-132.
2. Miedema BW, Johnson JO. Methods for decreasing postoperative gut dysmotility. Lancet Oncol. 2003;4:365-372.
3. Prasad M, Matthews JB. Deflating postoperative ileus. Gastroenterology 1999;117:489-492.
4. Collins TC, Daley J, Henderson WH, Khuri SF. Risk factors for prolonged length of stay after
major elective surgery. Ann.Surg. 1999;230:251-259.
5. Longo WE, Virgo KS, Johnson FE, Oprian CA, Vernava AM, Wade TP, Phelan MA, Henderson
WG, Daley J, Khuri SF. Risk factors for morbidity and mortality after colectomy for colon cancer. Dis.Colon Rectum 2000;43:83-91.
6. Luckey A, Livingston E, Tache Y. Mechanisms and treatment of postoperative ileus. Arch.Surg.
2003;138:206-214.
7. Wangensteen OH. The Early Diagnosis of Acute Intestinal Obstruction with comments on pathology and treatment. J.Surg.Obst.& Gyn. 1932;40:1-17.
8. Bayliss WM, Starling EH. The movements and innervations of the small intestine. J.Physiol
(Lond). 1899;24:99-143.
9. Person B, Wexner SD. The management of postoperative ileus. Curr.Probl.Surg. 2006;43:6-65.
10. De Winter BY, Boeckxstaens GE, De Man JG, Moreels TG, Herman AG, Pelckmans PA.
Effect of adrenergic and nitrergic blockade on experimental ileus in rats. Br.J.Pharmacol.
1997;120:464-468.
11. Boeckxstaens GE, Hirsch DP, Kodde A, Moojen TM, Blackshaw A, Tytgat GN, Blommaart PJ.
Activation of an adrenergic and vagally-mediated NANC pathway in surgery-induced fundic
relaxation in the rat. Neurogastroenterol.Motil. 1999;11:467-474.
12. Bauer AJ, Boeckxstaens GE. Mechanisms of postoperative ileus. Neurogastroenterol.Motil.
2004;16 Suppl 2:54-60.
13. Tache Y, Monnikes H, Bonaz B, Rivier J. Role of CRF in stress-related alterations of gastric
and colonic motor function. Ann N Y Acad Sci 1993;697:233-43.
14. Barquist E, Bonaz B, Martinez V, Rivier J, Zinner MJ, Tache Y. Neuronal pathways involved in
abdominal surgery-induced gastric ileus in rats. Am.J.Physiol 1996;270:R888-R894.
15. Plourde V, Wong HC, Walsh JH, Raybould HE, Tache Y. CGRP antagonists and capsaicin on
celiac ganglia partly prevent postoperative gastric ileus. Peptides 1993;14:1225-1229.
16. Seta ML, Kale-Pradhan PB. Efficacy of metoclopramide in postoperative ileus after exploratory
laparotomy. Pharmacotherapy 2001;21:1181-1186.
17. Bonacini M, Quiason S, Reynolds M, Gaddis M, Pemberton B, Smith O. Effect of intravenous
erythromycin on postoperative ileus. Am.J.Gastroenterol. 1993;88:208-211.
18. Brown TA, McDonald J, Williard W. A prospective, randomized, double-blinded, placebo-controlled trial of cisapride after colorectal surgery. Am.J.Surg. 1999;177:399-401.
19. Hallerback B, Bergman B, Bong H, Ekstrom P, Glise H, Lundgren K, Risberg O. Cisapride in
the treatment of post-operative ileus. Aliment.Pharmacol.Ther. 1991;5:503-511.
20. Jepsen S, Klaerke A, Nielsen PH, Simonsen O. Negative effect of Metoclopramide in postoperative adynamic ileus. A prospective, randomized, double blind study. Br.J.Surg. 1986;73:290291.
21. Bungard TJ, Kale-Pradhan PB. Prokinetic agents for the treatment of postoperative ileus in
adults: a review of the literature. Pharmacotherapy 1999;19:416-423.
22. Kalff JC, Carlos TM, Schraut WH, Billiar TR, Simmons RL, Bauer AJ. Surgically induced leukocytic infiltrates within the rat intestinal muscularis mediate postoperative ileus. Gastroenterology 1999;117:378-387.
23. Kehlet H, Holte K. Review of postoperative ileus. Am.J.Surg. 2001;182:3S-10S.
24. Smith CW, Marlin SD, Rothlein R, Toman C, Anderson DC. Cooperative interactions of LFA-1
and Mac-1 with intercellular adhesion molecule-1 in facilitating adherence and transendothelial
migration of human neutrophils in vitro. J.Clin.Invest 1989;83:2008-2017.
25. Issekutz AC, Rowter D, Springer TA. Role of ICAM-1 and ICAM-2 and alternate CD11/CD18
14
Chapter 1
ligands in neutrophil transendothelial migration. J.Leukoc.Biol. 1999;65:117-126.
26. Kalff JC, Schraut WH, Simmons RL, Bauer AJ. Surgical manipulation of the gut elicits an intestinal muscularis inflammatory response resulting in postsurgical ileus. Ann.Surg. 1998;228:652663.
27. Borovikova LV, Ivanova S, Zhang M, Yang H, Botchkina GI, Watkins LR, Wang H, Abumrad N,
Eaton JW, Tracey KJ. Vagus nerve stimulation attenuates the systemic inflammatory response
to endotoxin. Nature 2000;405:458-462.
28. Wang H, Yu M, Ochani M, Amella CA, Tanovic M, Susarla S, Li JH, Wang H, Yang H, Ulloa L,
Al Abed Y, Czura CJ, Tracey KJ. Nicotinic acetylcholine receptor alpha7 subunit is an essential
regulator of inflammation. Nature 2003;421:384-388.
29. Schwarz NT, Beer-Stolz D, Simmons RL, Bauer AJ. Pathogenesis of paralytic ileus: intestinal
manipulation opens a transient pathway between the intestinal lumen and the leukocytic infiltrate of the jejunal muscularis. Ann.Surg. 2002;235:31-40.
30. Sharkey KA. Substance P and calcitonin gene-related peptide (CGRP) in gastrointestinal
inflammation. Ann N Y Acad Sci 1992;664:425-42.
31. Suzuki R, Furuno T, McKay DM, Wolvers D, Teshima R, Nakanishi M, Bienenstock J. Direct
neurite-mast cell communication in vitro occurs via the neuropeptide substance P. J.Immunol.
1999;163:2410-2415.
32. Foreman JC. Substance P and calcitonin gene-related peptide: effects on mast cells and in human skin. Int Arch Allergy Appl Immunol 1987;82:366-71.
33. Kanwar S, Kubes P. Mast cells contribute to ischemia-reperfusion-induced granulocyte infiltration and intestinal dysfunction. Am.J.Physiol 1994;267:G316-G321.
34. Berin MC, Kiliaan AJ, Yang PC, Groot JA, Kitamura Y, Perdue MH. The influence of mast cells
on pathways of transepithelial antigen transport in rat intestine. J.Immunol. 1998;161:25612566.
15
2
2
Chapter
2
Postoperative Ileus Is Maintained by Intestinal Immune
Infiltrates That Activate Inhibitory Neural Pathways in Mice
Wouter J. de Jonge,
René M. van den Wijngaard,
Frans O. The,
Merel-Linde ter Beek,
Roelof J. Bennink,
Guido N. J. Tytgat,
Ruud M. Buijs,
Pieter H. Reitsma,
Sander J. van Deventer,
Guy E. Boeckxstaens
Abstract
Background & Aims: Postoperative ileus after abdominal surgery largely contributes to
patient morbidity and prolongs hospitalization. We aimed to study its pathophysiology in
a murine model by determining gastric emptying after manipulation of the small intestine.
Methods: Gastric emptying was determined at 6, 12, 24, and 48 hours after abdominal
surgery by using scintigraphic imaging. Intestinal or gastric inflammation was assessed
by immune-histochemical staining and measurement of tissue myeloperoxidase activity.
Neuromuscular function of gastric and intestinal muscle strips was determined in organ
baths. Results: Intestinal manipulation resulted in delayed gastric emptying up to 48 hours
after surgery; gastric half-emptying time 24 hours after surgery increased from 16.0 ± 4.4
minutes after control laparotomy to 35.6 ± 5.4 minutes after intestinal manipulation. The
sustained delay in gastric emptying was associated with the appearance of leukocyte
infiltrates in the muscularis of the manipulated intestine, but not in untouched stomach or
colon. The delay in postoperative gastric emptying was prevented by inhibition of intestinal
leukocyte recruitment. In addition, postoperative neural blockade with hexamethonium (1
mg/kg intraperitoneally) or guanethidine (50 mg/kg intraperitoneally) normalized gastric
emptying without affecting small-intestinal transit. The appearance of intestinal infiltrates after
intestinal manipulation was associated with increased c-fos protein expression in sensory
neurons in the lumbar spinal cord. Conclusions: Sustained postoperative gastroparesis
after intestinal manipulation is mediated by an inhibitory enterogastric neural pathway that
is triggered by inflammatory infiltrates recruited to the intestinal muscularis. These findings
show new targets to shorten the duration of postoperative ileus pharmacologically.
18
P
Postoperative ileus is characterized by a transient hypomotility of the gastrointestinal
tract that occurs after essentially every abdominal operation.1 It is a major contributor to
postoperative discomfort and results in prolonged hospitalization and increased patient
morbidity2 The pathophysiology of postoperative ileus is unclear, and as a result, current
Chapter 2
Background
treatment is limited to supportive procedures—such as nasogastric suction, early
postoperative feeding,3,4 and minimal use of opioid analgesics— that are known to intensify
ileus.5,6 Earlier pharmacological means of accelerating postoperative intestinal motility,
for instance, by antiadrenergic7 or cholinergic8 agents or by inhibiting peripheral opioid
effects on gastrointestinal transit,5 have had limited success.4,6,9 Therefore, more insight
into the mechanism mediating postoperative ileus is required for the development of new
pharmacological strategies to treat postoperative ileus.
Most previous experimental animal studies have focused on the pathophysiology of
instant hypomotility during or directly after abdominal surgery.10-13 This early component
of postoperative ileus results from the activation of mechanoreceptors, nociceptors, or
both by bowel manipulation during surgery. The subsequent stimulation of afferent fibers
triggers both spinal and supraspinal reflexes, inhibiting gastrointestinal motility and causing
an acute generalized postoperative ileus.10 However, because mechanical activation
mechanism cannot explain the prolonged nature of postoperative ileus. In previous reports,
it has been shown that the sustained phase of postoperative intestinal hypomotility due to
bowel handling results from inflammarory, rather than neuronal, mechanisms.14 Previously,
it has been shown that intestinal handling during abdominal surgery led to an impaired in
vitro contractility and a delayed transit of the manipulated small intestine. The latter resulted
from activation of resident macrophages and the subsequent establishment of a neutrophilic
infiltrate in the muscularis of the small intestine after bowel handling.14 Although this
phenomenon can account for the impaired propulsive motility of the small intestine, it does
not explain the hypomotility of the entire gastrointestinal tract, as observed in postoperative
ileus.15 It should also be emphasized that in human postoperative ileus, small-intestinal
motility recovers within 12 hours after surgery, whereas gastric and colonic motility remain
19
Neuro-Immune Interactions Maintain Postoperative Ileus
of mechanoreceptors and nociceptors ceases shortly after closure of the wound, this
disturbed for 3–5 days.1,6,15 Therefore, mechanisms other than local inflammation determine
the long-term hypomotility of untouched parts of the gastrointestinal tract.
In this study, our aim was to show in a murine model for postoperative ileus that leukocyte
infiltrates recruited in the intestinal muscularis by selective small-intestinal manipulation
affect the motility of parts of the gastrointestinal tract, distant from the site of manipulation,
by triggering an inhibitory neural pathway.
20
Animals
Mice (female BALB/c; Harlan Nederland, Horst, The Netherlands) were kept under
environmentally controlled conditions (lights on from 8:00 AM to 8:00 PM; water and rodent
nonpurified diet ad libitum; 20°C–22°C; 55% humidity). Mice were used at 8–12 weeks of
Chapter 2
Materials and Methods
age. Animal experiments were performed in accordance with the guidelines of the Ethical
Animal Research Committee of the University of Amsterdam.
Surgical Procedures
Mice were used at 6–10 weeks of age. After an overnight fast, mice were anesthetized
by an intraperitoneal (IP) injection of a mixture of ketamine (100 mg/kg) and xylazine
(20 mg/kg). Surgery was performed under sterile conditions. Mice (10–12 per treatment
group) underwent control surgery of only laparotomy or of laparotomy followed by intestinal
manipulation. The surgery was performed as follows. A midline abdominal incision was
made, and the peritoneum was opened over the linea alba. The small bowel was carefully
exteriorized, layered on a sterile moist gauze pad, and manipulated from the distal duodenum
to the cecum for 5 minutes by using sterile moist cotton applicators. Contact or stretch on
the stomach or colon was strictly avoided. After the surgical procedure, the abdomen was
closed by a continuous 2-layer suture (Mersilene 6-0 silk; Ethicon, Somerville, NJ). After
4 hours, mice were completely recovered from anesthesia. At 6, 12, 24, and 48 hours after
surgery, the gastric emptying rate was measured with gastric scintigraphy (see below).
Thereafter, mice were quickly anesthetized and killed by cervical dislocation, and the
stomach and small intestine were removed for histological analysis.
Treatments
Monoclonal antibodies against intracellular cell adhesion molecule-1 (anti-CD54 [ICAM-1];
immunoglobulin [Ig]G2b; clone YN1/1.7; 4.5 mg/kg)16 and lymphocyte function–associated
antigen-1 (CD11a [LFA-1]; IgG2a;H154.163; 2.3 mg/kg)16 were dissolved in dialyzed saline
(0.9% sodium chloride) and given by IP injection 1 hour before surgery. Identical quantities
21
closure, mice were allowed to recover for 4 hours in a heated (32°C) recovery cage. After
of nonspecific isotypematched IgGs were administered as controls. Hexamethonium (1
mg/kg) or guanethidine (50 mg/kg) was dissolved in sterile 0.9% sodium chloride and
administered by a single IP injection. Hexamethonium was administered 10 minutes, and
guanethidine 1 hour before the onset of gastric emptying tests.
Gastric Emptying and Transit
To determine the gastric emptying rate of a noncaloric semiliquid test meal, mice were
orally administered 0.1 mL of a 30 mg/ml methylcellulose solution containing 10 MBq of
technetium-99m (99mTc)-Albures (Nycomed-Amersham, Eindhoven, The Netherlands)
(albumin microcolloid) in water. Caloric solid test meals were prepared by baking 4 mL
of egg yolk mixed with 1 mL of water containing 400 MBq of 99mTc-Albures. Mice were
offered 100 mg of the baked egg yolk, which was consumed within 1 minute. Immediately
after the administration (semiliquid) or consumption (solid) of the test meal, mice were
scanned with a gamma camera set at 140 keV with 20% energy windows, fitted with a
pinhole collimator equipped with a 3-mm tungsten insert. A series of static images of the
entire abdominal region were obtained by scanning for 30 seconds at 16-minute intervals.
Static images were obtained at 1, 16, 32, 48, 64, 80, 96 (semiliquid), and 112 minutes
(solid) after administration of the test meal. The scanning frequency applied (once every 16
minutes) elicited no delay in gastric emptying because of handling stress.17 Static images
were analyzed by using Hermes computer software (Hermes, Stockholm, Sweden). To
determine the gastric emptying rate, a region of interest (ROI) was drawn around the gastric
and total abdominal region in each image obtained. Gastric emptying was measured by
determining the percentage of activity present in the gastric ROI, compared with the total
abdominal ROI, for each image. Subsequently, the gastric half-emptying time (t1⁄2) and
gastric retention at 64 minutes (Ret64) were determined for each individual mouse by
using DataFit software (version 6.1; Oakdale Engineering, Oakdale, PA). To this end, the
modified power exponential function y(t) - 1 - (1 - ekt)b was used, where y(t) is the fractional
meal retention at time t, k is the gastric emptying rate in minutes, and b is the extrapolated
y-intercept from the terminal portion of the curve. For determination of gastrointestinal
transit at 24 hours after surgery, animals were killed at 80 minutes after consumption of
the solid test meal. The abdomen was opened and the stomach clamped. Stomach, small
intestine, cecum, and colon were carefully exteriorized, and small intestine was divided into
22
6 fragments of equal length. The amount of
99m
Tc present in the stomach, small-intestinal
fragments, cecum, and colon was subsequently counted in a gamma counter. The geometric
∑(% radioactivity per segment x segment number)/100
Immunohistochemistry
Chapter 2
center was calculated from each experimental group according to the following formula:
Immunohistochemistry was performed as follows: after rehydration, endogenous peroxidase
activity was eliminated by incubating sections in 150 mmol/L of sodium chloride, pH
7.4, and 50% methanol, containing 3% (wt/vol) H2O2. Nonspecific protein-binding sites
were blocked by incubation for 30 minutes in TENG-T buffer (10 mmol/L Tris, 5 mmol/L
ethylenediaminetetraacetic acid [EDTA], 150 mmol/L sodium chloride, 0.25% gelatin, and
0.05% Tween-20, pH 8.0). Serial sections were incubated overnight with an appropriate
dilution of rat monoclonal antibodies raised against LFA-1, CD3, and CD4. Binding of the
primary antibodies was visualized with 3-amino-9-ethyl carbazole (Sigma, St. Louis, MO)
as a substrate, dissolved in sodium acetate buffer (pH 5.0) to which 0.01% H2O2 was
added.
C-fos immunohistochemistry was performed according to Bonaz et al.,18 with modifications.
Mice were anesthetized with a mixture of fentanyl citrate/fluanisone (Hypnorm; Janssen,
Beerse, Belgium) and midazolam (Dormicum; Roche, Mijdrecht, The Netherlands) at either
with 8 mL of a 0.9% NaCl solution, followed by 50 mL of 4% paraformaldehyde in phosphate
buffer (0.1 mol/L; pH 7.4). After perfusion, the spinal cord was rapidly removed, postfixed
overnight in the same fixative at 4°C, and cryoprotected for 24 hours in 30% sucrose
solution containing 0.05% sodium azide. After fixation, part of the lumbar spinal cord (L1
to L6) was embedded in Tissue-Tek (Sakura Finetek Inc., Torrance, CA). Fortymicrometer
transversal sections were cryostat-cut, and freefloating sections were incubated overnight
at 4°C with the primary polyclonal sheep antibody (0.3 μg/mL; Sigma Genosys, St.
Louis, MO) in 0.25% gelatin and 0.5% Triton X-100 in Tris-buffered saline (TBS; pH 7.4).
Sections were washed in TBS and incubated with biotinylated anti-sheep antiserum (Vector
Laboratories, Burlingame, CA) for 1.5 hours at room temperature. After washing in TBS,
23
90 minutes or 24 hours after surgery. Mice were then transcardially perfused (1.6 mL/min)
sections were processed for avidin– biotin–peroxidase (Vectorstain; Vector Laboratories),
and peroxidase was visualized by using diaminobenzidine in 0.02% nickel sulphate in TBS
as the chromogen. For quantification of the number of c-fos–expressing neurons, positive
nuclei in 30 sections were counted per lumbar spinal cord analyzed (n = 3 per treatment
group).
Muscularis Whole-Mount
Preparation Whole mounts of ileal segments were prepared as previously described,14
with slight modifications. In short, ileal segments (1–6 cm distal from the cecum) were
quickly excised, and mesentery was removed. Intestinal segments were cut open along the
mesentery border, fecal content was washed out in ice-cold phosphate-buffered saline, and
segments were pinned flat in a glass dish filled with preoxygenated Krebs–Ringer solution
(pH 7.4). Mucosa was removed, and the remaining full-thickness sheet of muscularis
externa was fixed for 10 minutes in 100% ethanol. Muscularis preparations were stored in
70% ethanol at 4°C until analysis.
Myeloperoxidase Activity Assay
Tissue myeloperoxidase (MPO) activity was determined as follows: either full-thickness ileal
segments or isolated ileal muscularis was blotted dry, weighed, and homogenized in a 20x
volume of a 20 mmol/L potassium phosphate buffer (pH 7.4). The suspension was centrifuged
(8000g for 20 minutes at 4°C), and the pellet was taken up in 1 mL of a 50 mmol/L potassium
phosphate buffer (pH 6.0) containing 0.5% hexadecyltrimethylammoniumbromide and 10
mmol/L EDTA and stored in 0.1-mL aliquots at -70°C until analysis. Fifty microliters of the
appropriate dilutions of the tissue homogenate was added to 445 µL of assay mixture,
which contained 0.2 mg/mL tetramethylbenzidine in 50 mg of potassium phosphate buffer
(pH 6.0), 0.5% hexadecyltrimethylammoniumbromide, and 10 mmol/L EDTA. The reaction
was started by adding 5 µL of 30 mmol/L H2O2 to the assay mixture, and the mixture was
incubated for 3 minutes at 37°C. After 3 minutes, 30 L of a 300 µg/mL catalase solution
was added to each tube, and tubes were placed on ice for 3 minutes. The reaction was
ended by adding 2 mL of 0.2 mol/L glacial acetic acid and incubating at 37°C for 3 minutes.
Absorbance was read at 655 nm. One unit of MPO activity was defined as the quantity of
MPO activity required to convert 1 µmol of H2O2 to H2O per minute at 25°C by using purified
MPO activity as a standard (Sigma), and activity was given in units per gram of tissue.
24
In Vitro Contractility Measurements
Stomach and ileum were quickly excised and cut open, and fecal content was flushed with
dish. After removal of the mucosa, longitudinal muscle strips (approximately 10 x 5 mm) of
the gastric fundus and antrum, circular muscle strips (approximately 0.7 x 5 mm) from the
antrum, and circular muscle strips of the ileum (approximately 1.0 x 5.0 mm) were mounted
in organ baths (25 mL) filled with Krebs–Ringer solution (pH 7.4), maintained at 37°C,
Chapter 2
ice-cold Krebs–Ringer solution (pH 7.4). Tissues were pinned down flat on a dissecting
and continuously aerated with a mixture of 5% CO2 and 95% oxygen. One end of each
muscle strip was anchored to a glass rod and placed between 2 platinum electrodes. The
other end was connected to a strain gauge transducer (type GM2/GM3; Scaime, Juvigny,
France) for continuous recording of isometric tension. Recording and analysis of muscle
contractions were performed with Acknowledge software (Biopac Systems Inc., Goleta,
CA). The gastric and ileal muscle strips were brought to their optimal point of length-tension
relationship by using 3 µmol/L acetylcholine and were then allowed to equilibrate for at
least 60 minutes before experimentation. Neurally mediated contractions of the muscle
strips of both the gastric fundus and the antrum were induced by means of electrical field
stimulation (0.5–16 Hz; 1- and 2-ms pulse duration; 10-second pulse trains). Responses
were always measured at the top of the contractile peak. In a second series of experiments,
contractions were evoked by the muscarinic receptor agonist carbachol (0.1 nmol/L to 3
µmol/L) and prostaglandin F2α (0.1 nmol/L to 3 µmol/L). Between the responses to the
different contractile receptor agonists, tissues were washed 4 times with an interval of
Contractions were calculated in grams of contraction per gram of tissue dry weight.
Drugs and Solutions
Acetylcholine, carbachol, prostaglandin F2α, hexamethonium, and guanethidine were
obtained from Sigma. A Krebs–Ringer solution was used that contained 118.3 mmol/L
NaCl, 4.7 mmol/L KCl, 1.2 mmol/L MgSO4 , 1.2 mmol/L KH2PO4 , 2.5 mmol/L CaCl2 ,
25 mmol/L NaHCO3 , 0.026 mmol/L EDTA, and 11.1 mmol/L glucose. Dr Y. van Kooyk,
Free University Amsterdam, kindly provided antibodies against ICAM-1 and LFA-1. Rat
monoclonal antibodies against CD3ε, CD4, and LFA-1 were purchased from Phar-Mingen
(San Diego, CA).
25
15 minutes. At the end of each experiment, muscle strips were blotted dry and weighed.
Results
Intestinal Manipulation Generates a Sustained Gastroparesis
At 6, 12, 24, and 48 hours after laparotomy or laparotomy combined with intestinal
manipulation, gastric emptying of a noncaloric semiliquid test meal was measured by
scintigraphic imaging. Examples of such an abdominal scan series of mice that underwent
laparotomy intestinal manipulation are presented in Figure 1. The anesthetics used during
abdominal surgery (ketamine 100 mg/kg and xylazine 20 mg/kg) did not alter postoperative
(>6 hours) gastric emptying.17 Also, as shown in Figure 1B and C, laparotomy alone had no
effect on the rate of gastric emptying at any time after surgery. After intestinal manipulation,
however, gastric emptying was significantly delayed (Figure 1). The delay was especially
pronounced 6 hours after surgery; intestinal manipulation increased Ret64 by 2.5-fold
compared with laparotomy only (Figure 1B). The (t1⁄2) was increased 3-fold (Figure 1B).
Gastric emptying after intestinal manipulation remained significantly delayed at 12 and 24
hours after surgery (Figure 1B), although the animals were fully recovered from surgery
at these time points. At 48 hours after surgery, Ret64 and t1⁄2 in intestinal manipulation–
treated mice had recovered to normal (Figure 1B). Similar results were obtained by using
a caloric solid test meal (Figure 1C). At 24 hours after surgery, gastric emptying of a caloric
solid test meal was delayed to an extent similar to that of the semiliquid test meal: intestinal
manipulation increased the t1⁄2 2.5-fold compared with laparotomy (Figure 1C).
A
st
L
st
IM
t=0
t=16
t=32
t=48
26
t=64
t=80 min
Chapter 2
*
60
*
40
IM t½ (min)
IM Ret 64 (%)
*
*
*
20
L T½ (min)
L Ret64 (%)
30
p<0.05
C
0
10
20
30
time after surgery (hrs)
60
10
20
L
IM
T½ solid (min)
0
T½ liquid (min)
*
p<0.05
IM Ret64 (%) / IM T½ (min)
80
40
50
Figure 1. Gastric emptying is delayed after abdominal
surgery. (A) A representative series of planar scintigraphic scans of mice that underwent laparotomy (L)
or intestinal manipulation (IM) is shown. The position of
the stomach is indicated (st) with a dotted circle. From
these scans, gastric emptying could be repetitively assessed for each mouse individually by determining the
amount of radioactivity present in the gastric region
compared with the total abdominal region. Note the
difference in radioactivity in the intestinal region between L and IM mice (arrows) at t=80 minutes. (B)
Half-emptying time (t1⁄2; open symbols) and gastric
retention after 64 minutes (Ret64; filled symbols) as a
function of time after L (squares) or IM (circles). Intestinal manipulation, performed at t =0 hours, resulted in
a significant (P <0.05) increase in t1⁄2 and Ret64 compared with laparotomy at t = 6, 12, and 24 hours after
surgery. Similar results were obtained with use of a
caloric, solid test meal; t1⁄2 was significantly increased
after intestinal manipulation compared with mice that
underwent L only (C). *Significant difference from L
with 1-way analysis of variance, followed by Dunnett’s
multiple comparison test. Data represent mean SEM
of 8–15 mice.
27
B
Intestinal Manipulation Recruits Leukocytes Into Intestinal
Muscularis
The delayed gastric emptying at 12, 24, and 48 hours after intestinal manipulation coincided
with an enhanced activity of the neutrophil indicator MPO in transmural ileal homogenates
(Figure 2). At 24 and 48 hours after surgery, intestinal manipulation, but not laparotomy
alone, resulted in a significant (P <0.05) increase in MPO activity measured in homogenates
of ileal tissue (Figure 2) or in ileal homogenates from which the mucosa was stripped off
(Figure 3). No increase in MPO activity was observed at earlier time points after surgery
(Figure 2). Histological analysis of transverse sections of ileal tissue indeed showed the
presence of LFA-1+ leukocytes in the ileal muscularis 24 hours after intestinal manipulation
(Figure 4B), but not after laparotomy alone (Figure 4A). Further immunohistochemical
staining showed that these leukocytes were MPO+, but CD3- and CD4- (data not shown).
Examination of the presence of inflammatory cells containing MPO activity in whole-mount
preparations (Figure 4C–F) and in isolated ileal muscularis tissue (Figure 3) confirmed the
presence of leukocyte infiltrates in the muscularis of manipulated ileum only (Figure 4C and
D). It is important to note that no increased presence of LFA-1+ leukocytes was found in
the muscularis of gastric antrum (Figure 4G and H) or in colonic tissue (data not shown) at
MPO activity (U/g ileal tissue)
any time point after surgery.
16
L
IM
8
0
* p<0.05
* p<0.05
* p<0.05
6
12
24
48 hrs PO
28
Figure 2. Ileal myeloperoxidase
(MPO) activity was selectively increased at 12, 24, and 48 hours
after surgery with intestinal manipulation (IM). MPO activity was determined in whole homogenates
of ileum isolated 6, 12, 24, and 48
hours after surgery as indicated.
MPO activity was significantly increased 12, 24, and 48 hours after laparotomy with IM (gray bars)
compared with laparotomy only (L;
white bars). *Significant difference
from L for each time point with a
Student t test (P< 0.05). Data represent mean SEM of 6–8 mice.
Occurrence of Delayed Gastroparesis Depends on Intestinal
Leukocyte Influx
intestinal manipulation mice received a preoperative bolus with monoclonal blocking
antibodies against ICAM-1 and LFA-1 to prevent leukocyte recruitment during the
postoperative period. Analysis of MPO-containing leukocytes in ileal muscularis (Figure
4E) or MPO activity in ileal muscularis homogenates (Figure 3) at 24 hours after intestinal
Chapter 2
To evaluate the role of the small-intestinal infiltrate in the development of gastroparesis,
manipulation showed that antibody treatment inhibited the leukocyte recruitment down to
30% (P <0.05) of untreated ileal segments. Prevention of the postoperative inflammatory
infiltrate did not affect the delay in gastric emptying 6 hours after surgery but normalized
gastric emptying 24 hours after intestinal manipulation (Figure 5). This effect was seen
with a noncaloric liquid, as well as with a caloric solid test meal (Figure 5B). Treatment with
identical quantities of isotype-matched control IgG did not affect leukocyte recruitment or
the observed postoperative delay in gastric emptying. These observations show that the
later phase of postoperative gastric ileus is mediated by an intestinal inflammatory infiltrate.
The antibody regimen could not prevent gastroparesis 6 hours after surgery, which is in line
3
* p<0.05
* p<0.05
2
1
0
L
IM
IM +MAb IM +hex
29
Figure 3. Intestinal manipulation results in
an increase in MPO activity measured in ileal muscularis. MPO activity was measured
in homogenates of ileal muscularis tissue
isolated 24 hours after surgery. Laparotomy
(L) with intestinal manipulation (IM) was associated with significantly increased MPO
activity in ileal muscularis tissue compared
with L alone. Treatment with ICAM-1– and
LFA-1–blocking antibodies before IM prevented the increase in MPO activity (IM_
ab). Treatment with hexamethonium did
not affect the increased MPO activity found
24 hours after IM (IM+hex). *Significant difference from L with 1-way analysis of variance (P<0.05) followed by Dunnett’s multiple comparison test. Data represent mean
SEM of 5–8 mice.
MPO activity (U/g ileal muscularis)
with the observation that the intestinal MPO activity was not increased at this time point.
Figure 4. (see fullcolor chapter 11) Focal leukocyte infiltrates after intestinal manipulation in the ileal
muscularis tissue. (A and B) Transverse sections of the ileal intestinal muscularis 24 hours after
laparotomy (A) and intestinal manipulation (B) were stained with mouse-specific monoclonal rat antibodies against LFA-1 (CD11a). Note the presence of LFA-1+ leukocytes in the ileal muscularis after
(B) intestinal manipulation (arrows), but not after (A) laparotomy. Sections were counterstained with
hematoxylin. MPO activity–containing leukocytes were visualized in whole mounts of ileal muscularis
tissue (C–F) isolated 24 hours after surgery. Intestinal manipulation (D), but not laparotomy (C), was
associated with a focal influx of MPO-containing leukocytes. Preoperative treatment of the mice with
monoclonal rat-blocking antibodies against ICAM-1 (CD54), combined with rat monoclonal antibodies
against LFA-1, prevented leukocyte influx (E). Postoperative treatment with hexamethonium did not
affect the presence of MPO-staining cells 24 hours after laparotomy with intestinal manipulation (F
). (G and H) Transverse sections of gastric antrum stained with monoclonal antibody against LFA-1.
Note the lack of LFA-1_ cells in the antral muscularis after laparotomy (G), as well as laparotomy with
intestinal manipulation (H). Sections were counterstained with hematoxylin. Bar is 75 mm (A, B, G,
and H) or 0.6 mm (C, D, E, and F ).
30
Chapter 2
Postoperative Inflammatory Infiltrates in the Intestinal Muscularis
Activate Spinal Afferent Neurons and Result in Gastric Ileus
Next, we investigated whether the small-intestinal infiltrate induced gastroparesis by
activation of an inhibitory neural pathway. To evaluate afferent neurotransmission in this
context, we measured the induction of the immediate-early gene c-fos within the spinal cord
24 hours after laparotomy or laparotomy with intestinal manipulation. Intestinal manipulation
significantly (P < 0.05) increased the number of nuclei expressing c-fos protein in the lumbar
dorsal horn of the spinal cord compared with laparotomy alone (Figure 6A and B). Most
positively labeled nuclei were found in laminae I of the lumbar dorsal horn. Treatment with
neutralizing antibodies against ICAM-1 and LFA-1 before intestinal manipulation prevented
the increase in spinal c-fos expression (Figure 6A and B), showing that intestinal leukocyte
infiltrates mediate spinal afferent activation. Treatment with control IgG antibodies did not
To further examine whether the sustained phase of delayed gastric emptying after intestinal
manipulation was neurally mediated, mice were treated either with hexamethonium, an
antagonist of nicotinic receptors (1 mg/kg, 10 minutes before gastric scintigraphy), or with
guanethidine, an adrenergic blocker (50 mg/kg, 1 hour before gastric scintigraphy) at 24
hours after abdominal surgery. These treatments did not affect gastric emptying (t1⁄2 or
Ret64) in control mice that underwent control laparotomy (data not shown). Furthermore,
the treatment with hexamethonium (Figures 3 and 4F) or guanethidine (not shown) did not
affect the leukocyte recruitment in the ileal muscularis after intestinal manipulation at 24
hours. After intestinal manipulation, however, treatment with these neural blockers either
partially (6 hours after surgery) or completely (24 hours after surgery) prevented the delay
in gastric emptying, compared with treatment with vehicle control (Figure 5A and B).
31
prevent increased c-fos expression after intestinal manipulation.
A
6 hr PO
60
100
20
60
100
20
60
100
IM
IM +hex
IM +gua
60
60
p<0.05
100
20
Time (min)
60
10
20
L
IM
IM +MAb IM +hex IM +gua
32
100
T½ solid (min)
30
p<0.05
B
60
p<0.05
p<0.05
20
p<0.05
20
20
T½ liquid (min)
IM
IM MAb
L
24 hr PO
p<0.05
relative gastric content (%)
100
Hexamethonium Ameliorates Postoperative Gastric Emptying, But Not Intestinal
Transit
acceleration of intestinal transit, we evaluated the effects of hexamethonium on intestinal
transit. Figure 7 shows that, in mice that underwent intestinal manipulation, the radiolabeled
test meal accumulates in the stomach, and that the small-intestinal transit is delayed
compared with control mice that underwent laparotomy. As indicated in Figure 7, intestinal
Chapter 2
Because normalization of gastric emptying could also be secondary to improvement or
manipulation and vehicle (saline) treatment led to a significant decrease of the geometric
center (P < 0.05). Postoperative treatment with hexamethonium prevented this surgeryinduced delay in gastric emptying but did not prevent the delay in small-intestinal transit.
Consequently, the geometric center was not different from that in mice that underwent
intestinal manipulation and received saline (Figure 7). The finding that hexamethonium
treatment normalizes gastric emptying even though intestinal transit is still delayed implies
that the delayed gastric emptying is not secondary to a functional obstruction of the small
intestine. To further evaluate the effect of hexamethonium on the delay in intestinal transit
induced by manipulation, we tested the in vitro contractility of intestinal circular muscle
strips. As shown in Figure 8, intestinal manipulation led to an impaired contractile activity of
circular muscle in response to carbachol. The addition of hexamethonium (3 x 10-5 mol/L)
Figure 5. Gastroparesis after intestinal manipulation (IM) is prevented by blocking leukocyte infiltration or neural blockade by hexamethonium or guanethidine treatment. Gastric emptying, determined
by scintigraphic imaging of the abdomen after oral administration of a semiliquid noncaloric meal at 6
and 24 hours (A) after IM, was compared with laparotomy alone (L). Values in (A) are given as relative gastric content compared with the total abdominal region. Corresponding t1⁄2 (B) with semiliquid
noncaloric (gray bars) and caloric solid (white bars) test meals were significantly (P < 0.05) increased
at 6 and 24 hours after IM, compared with L. Preoperative treatment with anti–ICAM-1 and anti–
LFA-1 antibodies (IM+MAb) normalized the t1⁄2 of semiliquid and solid test meals (B) at 24 hours after
surgery. Postoperative injections of hexamethonium (IM+hex) or guanethidine (IM+gua) normalized
t1⁄2 at 6 and 24 hours (B). Values are averages SEM of 8–12 mice per treatment group. Significant
differences (P < 0.05), determined by 1-way analysis of variance with treatment groups as variants,
are indicated.
33
did not reverse the impaired contractile response (Figure 8).
A
L
IM
Meancfoscount per section
B
IM + MAb
20
* p<0.05
10
Figure 6. Expression of c-fos in
the spinal cord 24 hours after intestinal manipulation. (A) c-fos–
labeled nuclei in the left and right
hemispheres of the lumbar dorsal
0
horn of mice 24 hours after control
L
IM
IM + Mab
laparotomy (L), intestinal manipulation (IM), or IM with pretreatment
with neutralizing antibodies against ICAM-1 and LFA-1 (IM + MAb). Images are representative of 3
mice examined in each group. The number of nuclei labeled per section was significantly increased
after IM (B) compared with control. Pretreatment of mice with neutralizing antibodies against ICAM-1
and LFA-1 prevented increased c-fos expression after intestinal manipulation. Significant differences
(P < 0.05), determined by 1-way analysis of variance with treatment group as variants, are indicated.
Values are averages SEM of 3 mice per treatment group.
34
Neuromuscular Properties of Gastric Fundus and Antrum Are Not
Affected by Intestinal Manipulation
neuromuscular function, the in vitro contractility of isolated muscle strips from gastric
fundus and antrum was investigated in organ baths. In Figure 9, the isomeric contractile
responses to increasing concentrations of the muscarinic receptor agonist carbachol
(0.1 nmol/L to 3 mmol/L) or prostaglandin F2a (0.1 nmol/L to 3 µmol/L) were determined
Chapter 2
To exclude the possibility that the delayed gastric emptying resulted from impaired local
from longitudinal (Figure 9A and B) or circular (Figure 9C) muscle strips isolated from
gastric fundus (Figure 9A) and antrum (Figure 9B and C). Intestinal manipulation did not
F
7
affect the dose-dependent contractile response to stimulation of gastric muscle strips with
prostaglandin F2α or carbachol, compared with mice that underwent laparotomy alone.
GC ± SEM
L + saline
4.2 ± 0.3*
2.6 ± 0.3
IM+ saline
IM+ hexamethonium3.0 ± 0.3
40
30
20
10
stomach
1
2
3
4
5
6
cecum
colon
small intestine (fragment nr)
Figure 7. Postoperative hexamethonium treatment accelerates postoperative gastric emptying, but
not intestinal transit. Transit was measured as a percentage distribution of the nonabsorbable 99mTcAlbures (albumin microcolloid) over the gastrointestinal tract after oral intake of a caloric solid test
meal. Stomach and 6 equal segments of small bowel, cecum, and colon were isolated 80 minutes
after oral ingestion of the caloric test meal (baked egg yolk), and radioactivity was counted in each
segment. In mice that underwent intestinal manipulation (IM) and received vehicle (saline) (dark
gray bars), the distribution of radioactivity indicates a delayed gastric emptying and an impaired
small-intestinal transit time compared with control mice that underwent only laparotomy (L; black
bars). The geometric center (GC) was significantly lower (P < 0.05; 1-way analysis of variance) in
mice that received IM + saline. Postoperative treatment with hexamethonium prevented the surgeryinduced delay in gastric emptying (IM + hexamethonium; light gray bars), but not intestinal transit.
Consequently, the geometric center was not different from that in mice that underwent IM + saline.
The impaired intestinal transit after manipulation is highlighted by a higher percentage of radioactivity found in intestinal fragments 1 and 2 in manipulated intestine compared with L and by the lower
percentage of radioactivity in fragments 5 and 6 (indicated by the dotted boxes). Numbers shown are
averages SEM of 8 mice per group.
35
% of total radioactivity
50
In addition, contractions evoked by nerve stimulation (0.5–16 Hz; 1-ms pulse duration; 10second pulse trains) in gastric fundus (Figure 9A) and antrum (Figure 9B and C) from mice
that underwent intestinal manipulation were not significantly different from contractions in
those that underwent control laparotomy.
A
0.12
+ saline
contraction (g/g tissue/mm)2
0.08
*
0.04
0
9
B
8
7
-log[carbachol]
*
6
+ 3*10-5M hexamethonium
0.12
0.08
*
*
*
0.04
0
*
9
8
7
-log[carbachol]
6
Figure 8. Ileal circular muscle carbachol dose–response curves 24 hours after laparotomy or intestinal manipulation. (A) intestinal manipulation (open squares) significantly suppresses contraction to
higher doses of carbachol compared with laparotomy (open circles). (B) The addition of hexamethonium (3 x10-5 mol/L) to the organ bath did not reverse the impaired contractility of mice that underwent
intestinal manipulation (filled squares). Values are mean SEM of 6 mice. Contractions are expressed
in grams of contraction per gram of tissue per square millimeter. *Significant differences (P < 0.05)
after unpaired Student t tests.
36
2
1
1
1
0
0
(Hz) 1
(ms) 1
2
2
1
4
1
8 16 16
1 1 2
B
1
0
(Hz) 1
(ms) 1
2
1
4
1
8 16 16
1 1 2
0
9
8
7
6
-log[carbachol]
2
2
1
1
0
9
8
7
6
-log[carbachol]
0
Chapter 2
2
2
L
IM
9
8
7
-log[PGF2α ]
6
L
IM
9
8
7
6
-log[PGF2α ]
Figure 9. In vitro gastric contractility of mice that underwent intestinal manipulation was not altered.
Lack of effect of intestinal manipulation on in vitro contractility of longitudinal muscle strips of gastric
fundus (A) and antrum (B) or circular muscle strips of the antrum (C) on different receptor agonists
and electric field stimulation is shown. Dose–response curves after electrical pulse stimulation (left),
carbachol (middle), or prostaglandin F2α (right) are shown. There was no difference in the neuromotor responses of mice that underwent laparotomy (filled symbols) or intestinal manipulation (open
symbols). Contractions are expressed in grams of contraction per gram of tissue per square millimeter. Values shown are means SEM (n= 6 - 7). No significant differences (P < 0.05) were found after
1-way analysis of variance followed by a Dunnett’s multiple comparison test.
37
A
Discussion
Postoperative ileus is associated with vomiting, bloating, nausea, and abdominal pain
and contributes considerably to postoperative patient morbidity. In addition, it has a
major economic effect due to prolonged hospitalization and increased costs of health
care. The annual economic cost resulting from the occurrence of postoperative ileus in
the U.S. population has been estimated to be $750,000,000,2 and this may even be a
gross underestimation, because drug costs and indirect costs were not measured. Until
now, treatment of postoperative ileus has been rather disappointing, mainly because
of a lack of pathophysiological insight. Here we provide data clarifying the underlying
mechanisms of the sustained phase of postoperative ileus. First, we confirmed that bowel
manipulation induces the local influx of inflammatory cells. 14 Subsequently, we showed that
the recruitment of this muscular infiltrate is associated with the activation of an inhibitory
adrenergic neural pathway that leads to prolonged postoperative gastroparesis. Our data
suggest that this mechanism is responsible for the generalized hypomotility observed in
postoperative ileus.
Most previous studies have evaluated only the acute effects of abdominal surgery on
gastrointestinal motility.10,11,19,20 However, we show here that, in mice, intestinal manipulation,
but not laparotomy alone, delays gastric emptying up to 48 hours after surgery. Two phases
can be distinguished in the period of postoperative gastric hypomotility: a first acute phase
that is not related to any inflammatory event and a second, later onset, and more sustained
phase that is temporally associated with a leukocyte influx into the intestinal muscularis.
Abundant evidence has been reported indicating that the mechanism underlying the
first, acute phase is a neurally mediated phenomenon: chemical neural blockade with
capsaicin,20,21 hexamethonium,10 or adrenergic antagonists12 reduced the rate of postoperative
ileus in animal models. In addition, surgical procedures that interrupt neural input to the
investigated gastrointestinal region, such as vagotomy or splanchnectomy,10 prevented or
reduced the postoperative hypomotility. Furthermore, studies evaluating neuronal c-fos
expression showed that both spinal and supraspinal pathways synapsing in the brainstem
are activated during abdominal surgery.22 The inhibitory efferent pathways involved have
been shown to be adrenergic and nonadrenergic noncholinergic in nature.10,11,19
38
In this study, we confirmed that the acute phase of postoperative ileus is mediated by a
blocker guanethidine improved the manipulation-induced delayed gastric emptying. The
observation that guanethidine only partially normalized the gastric emptying after intestinal
manipulation is in concert with the involvement of a nonadrenergic mechanism in the
efferent pathway mediating this phenomenon.10,11 These findings clearly indicate that bowel
Chapter 2
neural inhibitory mechanism: the nicotinic antagonist hexamethonium and the adrenergic
manipulation activates neural pathways, most likely via activation of mechanoreceptors or
nociceptors. However, mechanisms other than mechanical activation of these receptors
must be involved after closure of the abdomen to explain for the prolonged phase of
postoperative ileus, which lasts up to 24 hours, as observed in this study.
In this respect, Kalff et al.14 previously described that intestinal manipulation initiated
the up-regulation of ICAM-1 and LFA-1 and the subsequent recruitment of leukocytes
into the intestinal muscularis, leading to impaired contractility of circular muscle strips
of jejunum. It was suggested that these functional changes in the intestinal muscularis
resulting from a local inflammatory response were directly responsible for the sustained
paralysis of the gastrointestinal tract. In this study, we showed that the occurrence of an
inflammatory infiltrate was confined to the manipulated small intestine and was absent
in the non-manipulated stomach or colon. In addition, although the in vitro contractility of
ileal circular muscle strips was impaired after intestinal manipulation (compare with Kalff
finding shows that the delayed gastric emptying 24 hours after intestinal manipulation is not
due to impaired gastric neuromuscular function related to inflammation.
Instead, our results provide evidence that gastric ileus is the result of activation of an
inhibitory adrenergic neural pathway triggered by manipulation-induced leukocyte infiltrates
in the intestinal muscularis. This evidence is based on 2 main findings. First, the neuronal
blockers guanethidine and hexamethonium normalized postoperative gastric emptying.
Second, we confirmed23 that the occurrence of muscular infiltrates was associated with
the
activation of c-fos expression in spinal sensory neurons. Furthermore, blockade
of manipulationinduced intestinal leukocyte recruitment by treatment with neutralizing
antibodies against LFA-1 and its main cellular ligand, ICAM-1,24 prevented postoperative
39
et al.14), that of gastric muscle strips was unaffected by intestinal manipulation. The latter
activation of spinal neurons and normalized gastric emptying. These findings indicate that
the activation of t the adrenergic inhibitory pathway is most probably maintained by the
leukocyte infiltrate in the small-intestinal muscularis. The finding that ICAM-1 treatment
did not normalize the delay in gastric emptying 6 hours after surgery further corroborates
this notion, because no infiltrate was yet present at that time. What specific cell population,
leukocyte-derived mediator, or afferent nerve receptor is responsible for the neuro-immune
interaction leading to the activation of the adrenergic pathway remains to be established.
Alternatively, impaired gastric emptying may simply be secondary to stasis of chyme in
the intestine. The intestinal malfunction resulting from the manipulation-induced muscular
inflammation could theoretically back up the emptying of the stomach. However, we
showed that hexamethonium did normalize gastric emptying even though intestinal
transit remained delayed, making this possibility less likely. The independent modulation
of gastric emptying and intestinal transit is in agreement with previous reports.25,26 The
finding that hexamethonium normalized only gastric emptying and not intestinal transit
does not imply that the inhibitory neural input is confined to the stomach. Rather, the delay
in intestinal transit being resistant to hexamethonium can be explained by the local effect of
manipulation-induced muscular inflammation on intestinal motility.14 Indeed, we found that
hexamethonium did not prevent the occurrence of the infiltrate and had no effect on the
impaired in vitro contractility of the manipulated small intestine. To what extent the inhibitory
neural input contributes to the impaired intestinal transit cannot be determined from our
experiments.
Finally, intestinal inflammation could affect gastric motility via enhanced release of circulating
inflammatory mediators from the site of inflammation, such as the cytokines interleukin1β, tumor necrosis factor-α, or interleukin-627; prostaglandins28; bradykinin; or mediators
released by activated mast cells that potentially may affect gastric motility. However, in
our current study, hexamethonium or guanethidine administered 24 hours after surgery
could prevent gastroparesis, which implies that neuronal activity, rather than circulating
mediators, determines the delay in gastric emptying.
Several pathophysiological mechanisms may explain the inflammatory events observed in
surgically manipulated bowel tissue. Mechanical manipulation of the bowel during surgery
40
leads to intense activation of nerve fibers in the gut wall. This may result in local release
of substances with potent proinflammatory properties, such as substance P29 or calcitonin
In addition, recruitment of leukocytes may also be initiated via the release of proinflammatory
mediators by activated resident intestinal muscularis macrophages14 or mast cells. The
latter are known to be activated by neurally released substance P,30 and massive mast
cell activation has been described in response to manipulation of the gut.31 These leads,
Chapter 2
gene-related peptide,29 which can potentially induce neurogenic inflammation.
together with our current data, suggest that the anti-inflammatory effects of mast cell
stabilization may be instrumental in shortening the duration of postoperative ileus.
We conclude that postoperative ileus is a neurally mediated disorder that consists of
an early phase, which results from the triggering of afferents by activation of mechanoreceptors, nociceptors, or both after bowel manipulation or trauma, and a second,
prolonged, phase, in which an adrenergic inhibitory pathway is triggered by a local infiltrate.
In the rat, incremental degrees of surgical intestinal manipulation and trauma have been
shown to be proportional to the increase in recruitment of leukocyte infiltrates and the
severity of intestinal paralysis.32 This positive correlation may also explain the relation
between the extent, site, and length of intra-abdominal manipulation duration and the
severity of postoperative ileus found in human studies.6 These findings indicate that to
accelerate resumption of postoperative gastrointestinal motility and patient recovery, bowel
manipulation and the consequent recruitment of leukocytes should be kept minimal during
targets in reducing the duration and severity of postoperative ileus pharmacologically by
inhibiting postoperative recruitment of leukocytes to the intestinal wall, for instance, by using
blocking antibodies33 or antisense nucleotides against ICAM-1.34 Shortening postoperative
ileus is clinically and socioeconomically highly desired, and we anticipate that temporal
perioperative prevention of the influx of inflammatory cells may evolve as a new approach
to reduce postoperative patient morbidity.
41
abdominal surgery, i.e., during laparoscopy. However, our study also shows important new
Reference List
1. Prasad M, Matthews JB. Deflating postoperative ileus. Gastroenterology 1999;117:489–492.
2. Livingston EH, Passaro EP Jr. Postoperative ileus. Dig Dis Sci 1990;35:121–132.
3. Bradshaw BG, Liu SS, Thirlby RC. Standardized perioperative care protocols and reduced
length of stay after colon surgery. J Am Coll Surg 1998;186:501–506.
4. Steinbrook RA. An opioid antagonist for postoperative ileus. N Engl J Med 2001;345:988–989.
5. Taguchi A, Sharma N, Saleem RM, Sessler DI, Carpenter RL, Seyedsadr M, Kurz A. Selective
postoperative inhibition of gastrointestinal opioid receptors. N Engl J Med 2001;345:935–940.
6. Holte K, Kehlet H. Postoperative ileus: a preventable event. Br J Surg 2000;87:1480–1493.
7. Smith J, Kelly KA, Weinshilboum RM. Pathophysiology of postoperative ileus. Arch Surg
1977;112:203–209.
8. Ruwart MJ, Klepper MS, Rush BD. Carbachol stimulation of gastrointestinal transit in the postoperative ileus rat. J Surg Res 1979;26:18–26.
9. Resnick J, Greenwald DA, Brandt LJ. Delayed gastric emptying and postoperative ileus after
nongastric abdominal surgery: partII. Am J Gastroenterol 1997;92:934–940.
10. Boeckxstaens GE, Hirsch DP, Kodde A, Moojen TM, Blackshaw A,Tytgat GN, Blommaart PJ.
Activation of an adrenergic and vagallymediated NANC pathway in surgery-induced fundic
relaxation in the rat. Neurogastroenterol Motil 1999;11:467–474.
11. Boeckxstaens GE, Hollmann M, Heisterkamp SH, Robberecht P, de Jonge WJ, van Den
Wijngaard RM, Tytgat GN, Blommaart PJ. Evidence for VIP(1)/PACAP receptors in the afferent
pathway mediating surgery-induced fundic relaxation in the rat. Br J Pharmacol 2000;131:705–
710.
Effect of adrenergic and nitrergic blockade on experimental ileus in rats. Br J Pharmacol
1997;120:464–468.
13. Kalff JC, Buchholz BM, Eskandari MK, Hierholzer C, Schraut WH, Simmons RL, Bauer AJ. Biphasic response to gut manipulation and temporal correlation of cellular infiltrates and muscle
dysfunction in rat. Surgery 1999;126:498–509.
14. Kalff JC, Carlos TM, Schraut WH, Billiar TR, Simmons RL, Bauer AJ. Surgically induced leukocytic infiltrates within the rat intestinal muscularis mediate postoperative ileus. Gastroenterology 1999; 117:378–387.
15. Kehlet H, Holte K. Review of postoperative ileus. Am J Surg 2001;182:S3–S10.
16. Lub M, van Kooyk Y, Figdor CG. Competition between lymphocyte function-associated antigen
1 (CD11a/CD18) and Mac-1 (CD11b/CD18) for binding to intercellular adhesion molecule-1
(CD54). J Leukoc Biol 1996;59:648–655.
17. Bennink RJ, De Jonge WJ, Symonds EL, Van Den Wijngaard RM, Spijkerboer AL, Benninga
MA, Boeckxstaens GE. Validation of gastric-emptying scintigraphy of solids and liquids in mice
usingdedicated animal pinhole scintigraphy. J Nucl Med 2003;44:1099–1104.
18. Bonaz B, Plourde V, Tache Y. Abdominal surgery induces Fos immunoreactivity in the rat brain.
J Comp Neurol 1994;349:212–222.
19. De Winter BY, Robberecht P, Boeckxstaens GE, De Man JG, Moreels TG, Herman AG, Pelckmans PA. Role of VIP1/PACAP receptors in postoperative ileus in rats. Br J Pharmacol 1998;
124:1181–1186.
abdominal surgery-induced gastric ileus in rats. Am J Physiol 1996;270:R888–R894.
21. Holzer P, Lippe IT, Amann R. Participation of capsaicin-sensitive afferent neurons in gastric motor inhibition caused by laparotomy and intraperitoneal acid. Neuroscience 1992;48:715–722.
22. Bonaz B, Tache Y. Corticotropin-releasing factor and systemic capsaicin-sensitive afferents
are involved in abdominal surgeryinduced Fos expression in the paraventricular nucleus of the
hypothalamus. Brain Res 1997;748:12–20.
42
Chapter 2
23. Kreiss C, Birder LA, Kiss S, VanBibber MM, Bauer AJ. COX-2 dependent inflammation increases spinal Fos expression during rodent postoperative ileus. Gut 2003;52:527–534.
24. Marlin SD, Springer TA. Purified intercellular adhesion molecule-1 (ICAM-1) is a ligand for
lymphocyte function-associated antigen 1 (LFA-1). Cell 1987;51:813–819.
25. Freeman ME, Cheng G, Hocking MP. Role of alpha- and betacalcitonin gene-related peptide in
postoperative small bowel ileus. J Gastrointest Surg 1999;3:39–43.
26. Tanila H, Kauppila T, Taira T. Inhibition of intestinal motility and reversal of postlaparotomy ileus
by selective alpha 2-adrenergic drugs in the rat. Gastroenterology 1993;104:819–824.
27. Collins SM. The immunomodulation of enteric neuromuscular function: implications for motility
and inflammatory disorders. Gastroenterology 1996;111:1683–1699.
28. Schwarz NT, Kalff JC, Turler A, Engel BM, Watkins SC, Billiar TR, Bauer AJ. Prostanoid production via COX-2 as a causative mechanism of rodent postoperative ileus. Gastroenterology
2001; 121:1354–1371.
inflammation. Ann N Y Acad Sci 1992; 664:425–442.
neurite-mast cell communication in vitro occurs via the neuropeptide substance P. J Immunol
1999;163: 2410–2415.
31. Moriwaki K, Fujii K, Yuge O. Protein exudation induced by manipulation of the intestines and
mesentery during laparotomy in rat. A study of the mechanism of “third space” loss. In Vivo
1997; 11:325–327.
32. Kalff JC, Schraut WH, Simmons RL, Bauer AJ. Surgical manipulation of the gut elicits an
intestinal muscularis inflammatory response resulting in postsurgical ileus. Ann Surg 1998;228:
652–663.
33. Kavanaugh AF, Schulze-Koops H, Davis LS, Lipsky PE. Repeat treatment of rheumatoid arthritis patients with a murine antiintercellular adhesion molecule 1 monoclonal antibody. Arthritis
Rheum 1997;40:849–853.
34. Bennett CF, Kornbrust D, Henry S, Stecker K, Howard R, Cooper S, Dutson S, Hall W, Jacoby
HI. An ICAM-1 antisense oligonucleotide prevents and reverses dextran sulfate sodium-induced colitis in mice. J Pharmacol Exp Ther 1997;280:988–1000.
43
3
3
Chapter
3
The ICAM-1 antisense oligonucleotide
vents
the
ISIS-3082
development
preof
postoperative ileus in mice
British Journal of pharmacology 2005; 146: 252-258
Frans O. The,
Wouter J. de Jonge,
Roel J. Bennink,
Rene M. van den Wijngaard
Guy E. Boeckxstaens
Abstract
Background & Aims: Intestinal manipulation (IM) during abdominal surgery triggers the
influx of inflammatory cells, leading to postoperative ileus. Prevention of this local muscle
inflammation, using intercellular adhesion molecule-1 (ICAM-1) and leukocyte functionassociated antigen-1-specific antibodies, has been shown to shorten postoperative ileus.
However, the therapeutic use of antibodies has considerable disadvantages. The aim of the
current study was to evaluate the effect of ISIS-3082, a mouse-specific ICAM-1 antisense
oligonucleotide, on postoperative ileus in mice Methods: Mice underwent a laparotomy
or a laparotomy combined with IM after treatment with ICAM-1 antibodies, 0.1–10 mgkg-1
ISIS-3082, saline or ISIS-8997 (scrambled control antisense oligonucleotides, 1 and 3
mg kg-1). At 24 h after surgery, gastric emptying of a
99m
TC labelled semi-liquid meal was
determined using scintigraphy. Intestinal inflammation was assessed by myeloperoxidase
(MPO) activity in ileal muscle whole mounts. Results: IM significantly reduced gastric
emptying compared to laparotomy. Pretreatment with ISIS-3082 (0.1–1 mg kg-1) as well
as ICAM-1 antibodies (10 mg/kg-1), but not ISIS-8997 or saline, improved gastric emptying
in a dose-dependent manner. This effect diminished with higher doses of ISIS-3082 (3–10
mgkg-1). Similarly, ISIS-3082 (0.1–1 mgkg-1) and ICAM-1 antibodies, but not ISIS-8997
or higher doses of ISIS-3082 (3–10mg kg-1), reduced manipulation-induced inflammation.
Immunohistochemistry showed reduction of ICAM-1 expression with ISIS-3082 only.
Conclusion: ISIS-3082 pretreatment prevents postoperative ileus in mice by reduction of
manipulation-induced local intestinal muscle inflammation. Our data suggest that targeting
ICAM-1 using antisense oligonucleotides may represent a new therapeutic approach to the
prevention of postoperative ileus.
46
P
Background
Postoperative ileus is characterised by a generalised hypomotility of the gastrointestinal tract,
and is observed after almost every abdominal surgical procedure1. Although self-limiting,
leading to extra costs of between 750 million and 1 billion US dollars1, 2. Mainly due to a lack
of pathophysiological insight, treatment is limited to supportive and conservative measures
such as no oral feeding and intravenous (i.v.) fluids3.
Chapter 3
postoperative ileus is responsible for increased morbidity and prolonged hospitalisation,
Acute studies have convincingly shown that a laparotomy, but especially handling of the
intestine, inhibits gastrointestinal motility by activation of spinal and supraspinal inhibitory
pathways4-9. Recently, it became clear that manipulation of the intestine also triggers
the influx of inflammatory cells. This process becomes prominent several hours after
abdominal surgery and is now accepted to play a crucial role in the prolonged inhibition of
gastrointestinal motility10-12. This local inflammation not only leads to impaired contractility
of the diseased intestinal segment but also triggers an adrenergic inhibitory neural pathway,
explaining the more generalised aspect of postoperative ileus10-13. Leukocyte function-
associated antigen-1 (LFA-1) and its ligand intercellular adhesion molecule-1 (ICAM-1) are
two adhesion molecules that are crucial in the process of transmigration and recruitment of
strongly upregulated in response to inflammatory stimuli, including intestinal manipulation
(IM)11, 16, 17.
An important role for ICAM-1 in the development of the inflammatory infiltrate mediating
postoperative ileus is suggested by the observation that administration of a combination
of blocking antibodies to LFA-1 and ICAM-1 prior to abdominal surgery prevented the
recruitment of inflammatory cells in manipulated tissue and postoperative ileus11, 12. Although
it has not been studied whether blockade of only one of these adhesion molecules has
a similar effect, these data indicate that ICAM-1 may be an important target to prevent
postoperative ileus. However, the use of antibodies as therapeutic strategy in humans
still has considerable downsides, such as the formation of neutralizing antibodies or the
development of hypersensitivity reactions18,19.
47
ICAM-1 Antisense Oligonucleotides Prevent Postoperative Ileus
leukocytes14, 15. ICAM-1, normally only moderately expressed on vascular endothelium, is
Antisense oligonucleotides are 15–25-base long oligomers designed to hybridise to the
specific mRNA encoding for the target protein. As such, it prevents the translation of
mRNA, thereby downregulating the expression of the respective protein20, 21. ISIS-3082
is a murine ICAM-1-specific antisense oligonucleotide with anti-inflammatory properties
in experimental models of colitis, and a human-specific form, ISIS-2302 (alicaforsen), is
currently being tested in a clinical trial to evaluate this drug as potential new treatment in
patients with inflammatory bowel disease22, 23. In the present study, we investigated the
efficacy of the antisense oligonucleotide ISIS-3082 to shorten postoperative ileus in our
experimental mouse model.
48
Laboratory Animals
Female Balb/C mice (Harlan Nederland, Horst, The Netherlands), 12–15 weeks old, were
rodent nonpurified diet ad libitum; temperature 20-22°C; 55% humidity). All experiments
were performed after approval of the Ethical Animal Research Committee of the University
of Amsterdam and according to their guidelines.
Chapter 3
kept under environmentally controlled conditions (light on from 08:00 till 20:00 h; water and
Surgical Procedures: Abdominal Surgery
Mice were anaesthetised by intraperitoneal (i.p.) injection of 10 ml/kg of an anaesthetic
solution containing 0.078 mg/ml fentanyl citrate, 2.5 mg/ml fluanisone (Hypnorm;
Janssen, Beerse, Belgium) and 1.25mg/ml midazolam (Dormicum; Roche, Mijdrecht,
The Netherlands). Surgery was performed under sterile conditions. Mice underwent a
laparotomy, or a laparotomy followed by small IM, as described previously12.
In short, a midline incision was made and the peritoneal cavity was opened along the
linea alba. The small intestine was carefully exteriorised from the distal duodenum until the
cecum and gently manipulated for 5 min using sterile moist cotton applicators. Contact or
the abdomen was closed using a two-layer continuous suture (Mercilene Softsilk 6-0). Mice
recovered from surgery in a temperature-controlled cage at 32°C with free access to water,
but not to food. At 24 h after surgery, gastric emptying was measured. Thereafter, mice
were anaesthetised and killed by cervical dislocation. The small intestine was removed,
flushed in ice-cold phosphate-buffered saline (PBS), and snap frozen in liquid nitrogen or
fixed in ethanol for further analysis.
Drug preparation and treatment
ICAM-1 antibody (anti-CD54; IgG2b; clone YN1/1.7)24 was kindly provided by Professor Y.
van Kooyk (Department of Molecular Cell Biology & Immunology, VU University Medical
Center, Amsterdam, The Netherlands). Antibodies were dissolved in sterile 0.9% NaCl and
injected i.p. 1 h prior to the surgical intervention in a dose of 10 mg/kg12.
49
stretch of stomach or colon was strictly avoided. After repositioning of the intestinal loops,
ICAM-1 antisense oligonucleotide (ISIS-3082) and its scrambled control oligonucleotide
(ISIS-8997) were kindly provided by Dr Frank Bennett (ISIS-Pharmaceuticals, Carlsbad, CA,
U.S.A.). The specific sequences of the oligonucleotides used in this study were: ISIS-3082,
50-TGCATCCCCCAGGCCACCAT-30 and ISIS-8997, 50-CAGCCATGGTTCCCCCCAAC30. The final concentration of the oligonucleotide was determined using spectrometry
(Nanodrop ND-1000, Nanodrop Technologies Inc., Wilmington, DE, U.S.A.). ISIS-3082,
ISIS-8997 or their vehicle (sterile 0.9% NaCl) was injected subcutaneously (s.c.) once daily
starting 6 days prior to the surgical procedure. As intracellular localisation of the drug is
only achieved after 24 h, the onset of action of antisense oligonucleotide is not instant25.
Therefore, ISIS-3082 or ISIS-8997 was administered by s.c. injection once a day for 6 days
to achieve a steady-state concentration (approximately five half-lives) prior to surgery26.
ISIS-3082 was administered in a pharmacological range of 0.1, 0.3, 1.0, 3.0 or 10mg/kg,
which has been shown to be effective in DSS-colitis27. As the most effective dose of ISIS3082 was 1mg/kg, the control oligonucleotide, ISIS-8997, was tested in the same dose, as
well as a higher dose of 3mg/kg.
Measurement of gastric emptying
As previously described, gastric emptying rate was determined after gavage of a semi-liquid,
noncaloric test meal (0.1 ml of 3% methylcellulose solution containing 10 MegaBecquerel
(MBq) of 99mTc-Albures. Mice were scanned using a gamma camera set at 140 keV28. The
entire abdominal region was scanned for 30 s, immediately and 80 min after gavage. During
the scanning period, mice were conscious and manually restrained. The static images
obtained were analysed using Hermes computer software (Hermes, Stockholm, Sweden).
Gastric retention was calculated by determining the percentage of activity present in the
gastric region of interest compared to the total abdominal region of interest.
Whole-mount preparation
Ileal segments (4–6 cm proximal of cecum) were quickly excised. The mesentery was
removed from the intestine, which was cut open along its border. Faecal content was
washed out in ice-cold PBS, after which tissue segments were fixed in 100% ethanol for 10
min. Fixed preparations were kept in 70% ethanol at 41°C until further analysis.
50
Before final analysis, segments were stretched 1.5 times to their original size and pinned
down on a glass dish filled with 70% ethanol, after which the mucosa was carefully
removed.
Fixed preparations were rehydrated by incubation in 50% ETOH and PBS, pH 7.4, for 5
min. To visualise MPO-positive cells, preparations were incubated for 10 min with 3-amino9-ethyl carbazole (Sigma, St Louis, MO, U.S.A.) as substrate and dissolved in sodium
acetate buffer (pH 5.0), to which 0.01% H2O2 was added12.
Chapter 3
Assessment of leukocyte infiltration of the intestinal muscle
Immunohistochemistry
Immunohistochemical staining for ICAM-1 was performed on acetone fixed transverse
ileal segments. Endogenous peroxidase activity was eliminated by incubation of segments
in methanol containing 0.3% H2O2. Nonspecific protein-binding sites were blocked by
incubation in PBS, pH 7.4, containing 10% of normal goat serum for 10 min. Sections were
incubated overnight with biotinylated hamster anti-mouse ICAM-1 antibodies (Pharmingen,
San Diego, CA, U.S.A.) (dilution 1 : 1000). Next, sections were incubated with ABComplex/
HRP (DAKOCytomation, Glostrup, Denmark) for 30 min. HRP was visualised using
SigmaFast DAB (Sigma-Aldrich, St Louis, MO, U.S.A.), incubating 5 min, and contra-
Statistical analysis
A sample size of eight animals was used for each treatment group. Statistical analysis was
performed using SPSS 12.02 software for Windows. The data were expressed as mean
± s.e.m. Owing to the sample size, data were considered nonparametrically distributed.
The nonparametric Kruskal-Wallis test was used to analyse the cohort of independent
variables. If the difference between the multiple variables was statistically significant, the
Mann–Whitney test was performed to compare the individual treatment groups, identifying
the specific statistical differences. P<0.05 was considered statistically significant.
51
stained with 2% methyl green for 2 min.
Results
Effect of IM on gastric emptying and local intestinal muscle
inflammation 24 h after abdominal surgery
At 24 h after abdominal surgery, IM resulted in a significant increase of gastric retention
80 min after gavage of a noncaloric test meal, compared to a laparotomy (Figure 1). The
observed delay in gastric emptying after IM coincided with a profound local intestinal muscle
inflammatory cell influx compared to laparotomy (Figure 2).
Effect of ICAM-1 antisense oligonucleotide (ISIS-3082) pretreatment on gastric
emptying and intestinal muscle inflammation 24 h after abdominal surgery
Pretreatment with ISIS-3082 (0.1–1 mg/kg) reduced gastric retention in a dose-dependent
manner, restoring gastric emptying 24 h after IM at a dosage of 1mg/kg (Figure 3). This
effect was not observed with higher dosages (3– 10 mg/kg) (Figure 3). Moreover, ISIS3082 did not affect gastric emptying 24 h after a laparotomy in the absence of IM in mice
treated with 1mg/kg, compared to their vehicle control. In contrast, 1 and 3mg/kg ISIS8997, the scrambled control antisense oligonucleotides, did not improve gastric retention
24 h after IM (Figure 3).
0.5
*
Gastric retention
(% of total)
0.4
0.3
0.2
**
0.1
IM
kg
IC
AM
Ab
IM
10
m
g/
L
0.0
52
Figure 1 Effect of IM on gastric retention 24 h after abdominal surgery compared
to laparotomy only (L), or IM
after treatment with ICAM-1
antibodies (anti-CD54 IgG2b
clone 1/1.7). Each individual
group consisted of eight animals. Data are mean ± s.e.m.
gastric retention 80 min after
gavage of semi-liquid test
meal; *P<0.05 compared to L
control; **P<0.05 compared to
IM control.
The number of MPO-positive cells in muscle whole mounts diminished dose-dependently
(0.1–1 mg/kg) in mice treated with ISIS-3082 (Figure 4). Higher doses (3–10 mg/kg),
however, did not elicit reduction of the cellular infiltrate, nor did the scrambled control
antisense oligonucleotide (1 and 3mg/kg). To evaluate whether administration of high
10 mg/kg on animals who only underwent laparotomy. 10 mg/kg of ISIS-3082 did not show
an increase in MPO positive cells after laparotomy (Figure 4). Similar to 1mg/kg ISIS3082, administration of ICAM-1-specific antibodies (10 mg/kg i.p.) 1 h before IM resolved
the impaired gastric emptying observed 24 h after surgery, and significantly reduced the
Chapter 3
doses of ISIS-3082 had a local pro-inflammatory effect27, 29, we also studied the effect of
manipulation-induced leukocyte influx (Figures 1, 2, 5a–f).
Small-intestinal ICAM-1 expression Figure 6 shows the immunohistochemical staining for
*
200
**
100
10
m
g/
k
g
IC
AM
Ab
IM
IM
0
L
Intestinal Muscle Inflammation
( MPO-pos./mm -2 )
300
53
Figure 2 Effect of IM, laparotomy only (L) or IM pretreated
with ICAM-1 antibodies (antiCD54 IgG2b clone 1/1.7) on local inflammatory cell influx 24 h
after abdominal surgery. Each
individual group consisted of
eight animals. Data are mean ±
s.e.m. number of MPO-positive
cellsmm-2; *P<0.05 compared
to L control; **P<0.05 compared to IM control.
ICAM-1 on transverse ileal tissue segments to assess the in situ effect.
30
Gastric retention
(% of total)
*
20
10
**
-1
Sa
lin
0. Sa e L
1m lin
0. g e IM
3 kg
m
1. g k 1IM
0
m g -1
3. g k IM
0
IS
g
IS 1 mg -1
0
-8 .
k IM
IS 99 0 m g -1
IS 7 1 g
kg IM
-8
.
0
99
-1
7 mg IM
3.
0 kg -1
m
IM
g
1 kg m 1
g IM
kg
L
0
Figure 3 Effect of ISIS-3082, ISIS-8997 or vehicle on gastric retention 24 h after laparotomy (L) or
laparotomy with IM. Each individual group consisted of eight animals. Data are mean ± s.e.m. gastric retention 80 min after gavage of semi-liquid test meal; *P<0.05 compared to L control; **P<0.05
compared to vehicle IM control.
54
*
Chapter 3
Intestinal Muscle Inflammation
( MPO-pos. mm -2)
300
200
**
Figure 4 Effect of ISIS-3082, ISIS-8997 or vehicle on local inflammatory cell influx 24 h after laparotomy (L) or laparotomy with IM. Each individual group consisted of eight animals. Data are mean
± s.e.m. number of MPO-positive cells/mm-2; *P<0.05 compared to L control; **P<0.05 compared to
vehicle IM control.
55
0.
Sa
-1
L
0
**
lin
Sa e L
1
l
m ine
g
IM
0.
3 kg 1
m
1. g k IM
0
gm
1
3. g k IM
g0
IS
1
m
IS
10 g
I
-8
kg M
.
0
IS 99
IS 7 mg 1IM
1
-8
k
99 .0 m g -1
7
3. g k IM
0
g -1
m
IM
10 g
.0 kg m 1I
g M
kg
100
5a
5b
5c
5d
5e
5f
Figure 5 MPO staining of muscle whole mounts from mice that underwent a laparotomy after pretreatment with saline (a), or a laparotomy with intestinal manipulation after pretreatment with saline
(b), ICAM-1 antibodies (10mg/kg) (c), 1mg/kg ISIS-3082 (d), 10 mg/kg ISIS-3082 (e) or 1mg/kg ISIS8997 (f). Magnification x20; insertion x65.
56
Chapter 3
57
Figure 6 (see fullcolor chapter 11) ICAM-1 staining of ileal transverse segments from mice pretreated
with saline that underwent a laparotomy (a), and from mice that underwent a laparotomy with IM after
pretreatment with saline (b), 1mg/kg ISIS-3082 (c), 10 mg/kg ISIS-3082 (d) or 1mg/kg ISIS-8997 (e).
Note the increased ICAM-1 expression in the densely vascularised submucosa, but also in the blood
vessels, visible in the muscularis propria after IM (arrow heads). Only pretreatment with 1mg/kg ISIS3082 reduces the ICAM-1 expression (c).
Discussion
In the present study, we show that both ICAM-1 antibodies and the antisense oligonucleotide
ISIS-3082, targeted against ICAM-1, attenuate postoperative ileus by reducing manipulationinduced inflammation. These findings illustrate the importance of ICAM-1 in the pathogenesis
of postoperative
ileus, and suggest that ISIS-3082 may represent a potential new pharmacological approach
to prevent postoperative ileus.
Postoperative ileus complicates abdominal surgical intervention and causes prolonged
hospitalisation1. With regard to its pathophysiology, it has been shown that intestinal
handling during abdominal surgery activates mast cells and resident macrophages, initiating
the recruitment of neutrophils into the intestinal muscle layer10-12, 30. This local infiltrate of
leukocytes is now recognised as a crucial player in postoperative ileus, as it has been
shown to activate inhibitory neural pathways that lead to a generalised hypomotility of the
gastrointestinal tract12, 13. Although it is not known to what extent the same mechanism is
responsible for the development of postoperative ileus in patients, Kalff et al.31 observed an
increase in mRNA expression in the human intestine for several proinflammatory proteins
like LFA-1, iNOS, IL-6 and TNF-a after abdominal surgery.
Upregulation of adhesion molecules such as LFA-1 and ICAM-1 are necessary for the
extravasation of leukocytes. Here, we show that ICAM-1 expression is clearly increased
after IM, being most profound in the vasculature between the submucosal and the muscle
layers, but also in the muscularis propria. This observation confirms that manipulation of
the small intestine indeed increases the expression of ICAM-1, facilitating local infiltration of
inflammatory cells 10,11. Previous studies demonstrated that pretreatment with a combination
of antibodies against LFA-1 and ICAM-1 prevented postoperative ileus by blocking of this
manipulation-induced infiltrate12. In the present study, we show that pretreatment with
antibodies targeted to ICAM-1 alone also results in a reduction of inflammatory cell influx
and the prevention of delayed gastric emptying. These results illustrate that ICAM-1 is an
important target to prevent postoperative ileus.
58
The use of antisense oligonucleotides is a novel approach to block the synthesis of regulatory
peptides. These 15–25-baselong oligomers hybridise to the specific mRNA, preventing its
translation, thereby downregulating the expression of the respective protein20, 21. ISIS-3082
is a mouse-specific ICAM-1 antisense oligonucleotide, which has been shown to be effective
32
. We used ISIS-3082 to study its anti-inflammatory effects in our experimental model for
postoperative ileus. Similar to the ICAM-1 antibody (anti-CD54 IgG2b clone YN1/1.7)24, 33,
ISIS-3082 reduces the IM-induced inflammatory cell influx and improves gastric emptying
in a dose-dependent manner, with a maximum effect at 1mg/kg, restoring delayed gastric
Chapter 3
in experimental murine models for heart allograft rejection and inflammatory bowel disease27,
emptying. As the nonsense control oligonucleotide (ISIS 8997) in a dose of 1 as well 3mg/kg
did not have these effects, a sequence unspecific effect of the phosphorothioate backbone
can be excluded. Therefore, we conclude that the anti-inflammatory effect of ISIS-3082
observed results from a sequence-specific reduction in ICAM-1 mRNA translation and
protein expression. The latter is supported by the immunohistochemical staining showing a
reduction of ICAM-1 expression by ISIS-3082, but not by ISIS-8997 or saline.
In the pharmacological range tested, the anti-inflammatory effect of ISIS-3082 diminished
in higher doses (3 and 10 mg/kg). Bennett et al.27 observed a similar dosedependent effect
in a study evaluating ISIS-3082 in a DSS colitis model. The lack of effect of higher dosages
in antisense oligonucleotides like ISIS-308227, 29, 34. However, ICAM-1 expression was not
reduced in the presence of the local muscle inflammation, making this possibility less
likely. A more plausible explanation might be the biphasic response of ribonuclease H
activity on phosphorothioate antisense oligonucleotide concentration. Low concentrations
of phosphorothioate oligonucleotides increase ribonuclease H activity, whereas high
concentrations have the opposite effect, leading to increased stability of the antisensebound mRNA35. The latter leads to decreased breakdown of ISIS-3082-bound (ICAM-1specific) mRNA by ribonuclease H, and a diminished effect on ICAM-1 protein synthesis.
At present, treatment of postoperative ileus consists of supportive measures such as nothing
by mouth, nasogastric suction, i.v. fluids, and the use of prokinetic and antiemetic drugs.
Unfortunately, this approach has been rather disappointing3, 36. Based on the current data,
pretreatment of patients with antibodies or antisense oligonucleotides targeted against
59
may be explained by the pro-inflammatory properties of the phosphorothioate backbones
ICAM-1 are possible new preventive strategies to shorten postoperative ileus. One of the
risks of using antibody treatment is the potential formation of neutralising antibodies18, 19.
Antisense oligonucleotides could represent an alternative to antibody treatment. The human
equivalent of ISIS-3082 (ISIS-2302) is currently being tested in a clinical trial as a putative
new treatment for inflammatory bowel disease. Based on the bell-shaped dose–response
curve, it should be emphasised that the therapeutic range is narrow, compromising its
clinical use. In addition, one should also consider that leukocyte recruitment to traumatised
tissues is needed for healing of the surgical wound. Both ISIS-3082 and ISIS-2302 have
been extensively tested in several, also surgeryinvolving, models, disorders and clinical
trials. None of theses studies reported impairment of wound healing or other postsurgical
complications32, 37, 38. In conclusion, ICAM-1 antisense pretreatment prevented postoperative
ileus in mice by reduction of manipulationinduced intestinal muscle inflammation. Our
data encourage further clinical evaluation of ICAM-1 antisense oligonucleotides as tool to
prevent postoperative ileus.
60
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
Prasad M, Matthews JB. Deflating postoperative ileus. Gastroenterology 1999;117:489-492.
Livingston EH, Passaro EP, Jr. Postoperative ileus. Dig.Dis.Sci. 1990;35:121-132.
Kehlet H, Holte K. Review of postoperative ileus. Am.J.Surg. 2001;182:3S-10S.
Plourde V, Wong HC, Walsh JH, Raybould HE, Tache Y. CGRP antagonists and capsaicin on
Barquist E, Bonaz B, Martinez V, Rivier J, Zinner MJ, Tache Y. Neuronal pathways involved in
De Winter BY, Boeckxstaens GE, De Man JG, Moreels TG, Herman AG, Pelckmans PA.
Effects of mu- and kappa-opioid receptors on postoperative ileus in rats. Eur.J.Pharmacol.
1997;339:63-67.
De Winter BY, Boeckxstaens GE, De Man JG, Moreels TG, Herman AG, Pelckmans PA.
1997;120:464-468.
Boeckxstaens GE, Hirsch DP, Kodde A, Moojen TM, Blackshaw A, Tytgat GN, Blommaart PJ.
Boeckxstaens GE, Hollmann M, Heisterkamp SH, Robberecht P, de Jonge WJ, van den Wijngaard RM, Tytgat GN, Blommaart PJ. Evidence for VIP(1)/PACAP receptors in the afferent
pathway mediating surgery-induced fundic relaxation in the rat. Br.J.Pharmacol. 2000;131:705710.
Kalff JC, Buchholz BM, Eskandari MK, Hierholzer C, Schraut WH, Simmons RL, Bauer AJ. Biphasic response to gut manipulation and temporal correlation of cellular infiltrates and muscle
dysfunction in rat. Surgery 1999;126:498-509.
Kalff JC, Carlos TM, Schraut WH, Billiar TR, Simmons RL, Bauer AJ. Surgically induced leukocytic infiltrates within the rat intestinal muscularis mediate postoperative ileus. Gastroenterology 1999;117:378-387.
de Jonge WJ, van den Wijngaard RM, The FO, ter Beek ML, Bennink RJ, Tytgat GN, Buijs
RM, Reitsma PH, van Deventer SJ, Boeckxstaens GE. Postoperative ileus is maintained by
intestinal immune infiltrates that activate inhibitory neural pathways in mice. Gastroenterology
2003;125:1137-1147.
Kreiss C, Birder LA, Kiss S, VanBibber MM, Bauer AJ. COX-2 dependent inflammation increases spinal Fos expression during rodent postoperative ileus. Gut 2003;52:527-534.
Smith CW, Marlin SD, Rothlein R, Toman C, Anderson DC. Cooperative interactions of LFA-1
and Mac-1 with intercellular adhesion molecule-1 in facilitating adherence and transendothelial
migration of human neutrophils in vitro. J.Clin.Invest 1989;83:2008-2017.
Issekutz AC, Rowter D, Springer TA. Role of ICAM-1 and ICAM-2 and alternate CD11/CD18
ligands in neutrophil transendothelial migration. J.Leukoc.Biol. 1999;65:117-126.
Dustin ML, Rothlein R, Bhan AK, Dinarello CA, Springer TA. Induction by IL 1 and interferongamma: tissue distribution, biochemistry, and function of a natural adherence molecule
(ICAM-1). J.Immunol. 1986;137:245-254.
Rothlein R, Dustin ML, Marlin SD, Springer TA. A human intercellular adhesion molecule
(ICAM-1) distinct from LFA-1. J.Immunol. 1986;137:1270-1274.
Shawler DL, Bartholomew RM, Smith LM, Dillman RO. Human immune response to multiple
injections of murine monoclonal IgG. J.Immunol. 1985;135:1530-1535.
LoBuglio AF, Wheeler RH, Trang J, Haynes A, Rogers K, Harvey EB, Sun L, Ghrayeb J, Khazaeli MB. Mouse/human chimeric monoclonal antibody in man: kinetics and immune response.
Proc.Natl.Acad.Sci.U.S.A 1989;86:4220-4224.
Crooke ST. Therapeutic applications of oligonucleotides. Annu.Rev.Pharmacol.Toxicol.
1992;32:329-376.
Stein CA, Cheng YC. Antisense oligonucleotides as therapeutic agents--is the bullet really
magical? Science 1993;261:1004-1012.
61
1.
2.
3.
4.
Chapter 3
Reference List
22. Miner P, Wedel M, Bane B, Bradley J. An enema formulation of alicaforsen, an antisense
inhibitor of intercellular adhesion molecule-1, in the treatment of chronic, unremitting pouchitis.
Aliment.Pharmacol.Ther. 2004;19:281-286.
23. van Deventer SJ, Tami JA, Wedel MK. A randomised, controlled, double blind, escalating dose
study of alicaforsen enema in active ulcerative colitis. Gut 2004;53:1646-1651.
24. Lub M, van Kooyk Y, Figdor CG. Competition between lymphocyte function-associated antigen
1 (CD11a/CD18) and Mac-1 (CD11b/CD18) for binding to intercellular adhesion molecule-1
(CD54). J.Leukoc.Biol. 1996;59:648-655.
25. Butler M, Stecker K, Bennett CF. Cellular distribution of phosphorothioate oligodeoxynucleotides in normal rodent tissues. Lab Invest 1997;77:379-388.
26. Crooke ST, Graham MJ, Zuckerman JE, Brooks D, Conklin BS, Cummins LL, Greig MJ,
Guinosso CJ, Kornbrust D, Manoharan M, Sasmor HM, Schleich T, Tivel KL, Griffey RH.
Pharmacokinetic properties of several novel oligonucleotide analogs in mice. J.Pharmacol.Exp.
Ther. 1996;277:923-937.
27. Bennett CF, Kornbrust D, Henry S, Stecker K, Howard R, Cooper S, Dutson S, Hall W, Jacoby
HI. An ICAM-1 antisense oligonucleotide prevents and reverses dextran sulfate sodium-induced colitis in mice. J.Pharmacol.Exp.Ther. 1997;280:988-1000.
28. Bennink RJ, de Jonge WJ, Symonds EL, van den Wijngaard RM, Spijkerboer AL, Benninga
using dedicated animal pinhole scintigraphy. J.Nucl.Med. 2003;44:1099-1104.
29. Pisetsky DS, Reich CF. Stimulation of murine lymphocyte proliferation by a phosphorothioate
oligonucleotide with antisense activity for herpes simplex virus. Life Sci. 1994;54:101-107.
30. de Jonge WJ, The FO, van der CD, Bennink RJ, Reitsma PH, van Deventer SJ, van den
Wijngaard RM, Boeckxstaens GE. Mast cell degranulation during abdominal surgery initiates
postoperative ileus in mice. Gastroenterology 2004;127:535-545.
31. Kalff JC, Turler A, Schwarz NT, Schraut WH, Lee KK, Tweardy DJ, Billiar TR, Simmons RL,
Bauer AJ. Intra-abdominal activation of a local inflammatory response within the human muscularis externa during laparotomy. Ann.Surg. 2003;237:301-315.
32. Stepkowski SM, Tu Y, Condon TP, Bennett CF. Blocking of heart allograft rejection by intercellular adhesion molecule-1 antisense oligonucleotides alone or in combination with other immunosuppressive modalities. J.Immunol. 1994;153:5336-5346.
33. Pruijt JF, van Kooyk Y, Figdor CG, Lindley IJ, Willemze R, Fibbe WE. Anti-LFA-1 blocking antibodies prevent mobilization of hematopoietic progenitor cells induced by interleukin-8. Blood
1998;91:4099-4105.
34. Zhao Q, Temsamani J, Iadarola PL, Jiang Z, Agrawal S. Effect of different chemically modified
oligodeoxynucleotides on immune stimulation. Biochem.Pharmacol. 1996;51:173-182.
35. Gao WY, Han FS, Storm C, Egan W, Cheng YC. Phosphorothioate oligonucleotides are inhibitors of human DNA polymerases and RNase H: implications for antisense technology. Mol.
Pharmacol. 1992;41:223-229.
36. Luckey A, Livingston E, Tache Y. Mechanisms and treatment of postoperative ileus. Arch.Surg.
2003;138:206-214.
37. Kahan BD, Stepkowski S, Kilic M, Katz SM, Van Buren CT, Welsh MS, Tami JA, Shanahan
WR, Jr. Phase I and phase II safety and efficacy trial of intercellular adhesion molecule-1 antisense oligodeoxynucleotide (ISIS 2302) for the prevention of acute allograft rejection. Transplantation 2004;78:858-863.
38. Chen W, Langer RM, Janczewska S, Furian L, Geary R, Qu X, Wang M, Verani R, Condon T,
Stecker K, Bennett CF, Stepkowski SM. Methoxyethyl-modified intercellular adhesion molecule-1 antisense phosphorothiateoligonucleotides inhibit allograft rejection, ischemic-reperfusion injury, and cyclosporine-induced nephrotoxicity. Transplantation 2005;79:401-408.
62
63
Chapter 3
4
4
Chapter
4
The vagal anti-inflammatory
pathway attenuates intestinal
macrophage activation and inflammation by nicotinic acetyl-
choline receptor mediated activation of Jak-2/Stat-3.
Nature Immunology 2005; 6: 844-851
Wouter J de Jonge,
Esmerij P van der Zanden,
Frans O The,
Maarten F Bijlsma,
David J van Westerloo,
Roelof J Bennink,
Hans-Rudolf Berthoud,
Satoshi Uematsu,
Shizuo Akira,
Rene M van den Wijngaard
Guy E Boeckxstaens
Abstract
Acetylcholine released by efferent vagus nerves inhibits macrophage activation. Here
we show that the anti-inflammatory action of nicotinic receptor activation in peritoneal
macrophages was associated with activation of the transcription factor STAT3. STAT3 was
phosphorylated by the tyrosine kinase Jak2 that was recruited to the α7 subunit of the
nicotinic acetylcholine receptor. The anti-inflammatory effect of nicotine required the ability
of phosphorylated STAT3 to bind and transactivate its DNA response elements. In a mouse
model of intestinal manipulation, stimulation of the vagus nerve ameliorated surgery-induced
inflammation and postoperative ileus by activating STAT3 in intestinal macrophages. We
conclude that the vagal anti-inflammatory pathway acts by α7 subunit−mediated Jak2STAT3 activation.
66
T
Background
The innate immune response has been increasingly recognized as being under substantial
neuronal control1. For example, acetylcholine or nicotine effectively attenuates the activation
of macrophages2. This so-called ‘cholinergic anti-inflammatory pathway’ is characterized
by a nicotine dose−dependent decrease in the production of proinflammatory mediators,
1β), IL-6 and IL-18 2, by macrophages stimulated with endotoxin. Consistently, stimulation of
the efferent vagus nerve dampens macrophage activation in rodent models of endotoxemia
and shock1, 2. Two nicotinic acetylcholine receptor (nAChR) subtypes are involved in the
nicotine-induced decrease in proinflammatory cytokine production by stimulated human
Chapter 4
including high-mobility group box 1 proteins3, tumor necrosis factor (TNF), interleukin 1β (IL-
and mouse macrophages: the α7 homopentamer expressed by monocyte-derived
human and mouse macrophages4, and the α4β2 heteropentamer expressed by alveolar
macrophages5. Activation of the α7 homopentamer nAChR inhibits transactivational activity
deactivating effect of acetylcholine on macrophages has remained unknown.
Here we evaluated the involvement of the transcription factor STAT3 in this process,
because STAT3 is a potential negative regulator of inflammatory responses6, 7. STAT3 and
the tyrosine kinase Jak2, which phosphorylates STAT3, are required for both IL-6 receptor
(IL-6R) and IL-10R signaling. IL-6 contributes to the progression of many inflammatory
diseases, whereas IL-10 is an anti-inflammatory cytokine that suppresses the activation of
macrophages. IL-6R signaling is inhibited by the Src homology 2 domain protein SOCS3,
whose expression is induced by STAT3 activation8, 9. SOCS3 binds to the glycoprotein 130
(gp130) subunit of the IL-6R, leading to inhibited activation of STAT3 by IL-6R ligands8, 9.
Consistent with that finding, in LPS-stimulated macrophages deficient in SOCS3, IL-6R
ligands induce a sustained STAT3 activation, which leads to the reduced production of
proinflammatory cytokines such as TNF10.
Here we demonstrate that nicotine exerts its anti-inflammatory effect on peritoneal
macrophages via Jak2 and STAT3 signaling in vitro and in vivo. In isolated peritoneal
macrophages, nicotine activated nAChRs, leading to phosphorylation of STAT3 via Jak2.
67
Vagal Anti-inflammatory Pathway Mediated through Nicotinic Jak-2/Stat-3 Activation
of the transcription factor NF-κB p65 3. However, the subcellular mechanism explaining the
Jak2 was recruited to the α7 subunit of the nAChR and was phosphorylated after nicotine
binding. We further studied the effect of cholinergic inhibition of macrophage activity in vivo
on the occurrence of post-surgical intestinal inflammation in a mouse model of postoperative
ileus11, 12. Postoperative ileus is characterized by general hypomotility of the gastrointestinal
tract and delayed gastric emptying13 and is a pathological condition commonly noted after
abdominal surgery with intestinal manipulation. This condition is the result of inflammation
of the intestinal muscularis due to activation of resident macrophages14, 15 that are triggered
by bowel manipulation12. We show here that perioperative stimulation of the vagus nerve
prevented manipulation-induced inflammation of the intestinal muscularis externa and
ameliorated postoperative ileus. The effectiveness of stimulation of the vagus nerve in
reducing intestinal inflammation depended on STAT3 activation in macrophages in the
intestinal muscularis. Hence, our data demonstrate the molecular pathway responsible for
cholinergic inhibition of macrophage activation and suggest that stimulation of the vagus
nerve or administration of cholinergic agents may be effective anti-inflammatory therapy for
the treatment of postoperative ileus and other inflammatory diseases.
68
Methods
Reagents and antibodies.
Nicotine, hexamethonium, α-bungarotoxin, methyllycaconitine citrate, d-tubocurarin,
dihydro-β-erythroidine, AG 490, cycloheximide, actinomycin-D and rat monoclonal
antibody to β2 nAChR subunit (anti-β2) were from Sigma-Aldrich. Polyclonal rabbit antipolyclonal anti-actin, rabbit polyclonal anti-STAT1 and rabbit polyclonal anti-STAT3 were
from Santa Cruz Biotechnology; and rabbit polyclonal anti−phosphorylated STAT1 and
anti-phosphorylated STAT3 were from Cell Signaling Technology. ELISA kits for IL-6, IL-10,
MIP-1α, MIP-2 and TNF were from R&D Systems.
Chapter 4
Jak2, anti−phosphorylated Jak2, anti-SOCS3 and anti-α7 were obtained from Abcam; goat
Cell culture and transient transfection.
Resident peritoneal macrophages were collected from BALB/c mice by flushing of the
of heparin. Peritoneal cells were plated at a density of 1 x 106 cells/cm2 in RPMI medium
supplemented with 10% FCS, and macrophages were left to adhere for 2 h in a humidified
atmosphere at 37 °C with 5% CO2. Cells were washed and the remaining macrophages were
left for 16−20 h. Subsequently, cells were preincubated with the appropriate concentration of
nicotine for 15 min, followed by challenge for 3 h with LPS (1−100 ng/ml). NAChR blockers
were added 30 min before nicotine, and no toxicity was noted after 4 h of incubation
with any blocker, as assessed by the trypan blue exclusion test. Cycloheximide (10 µg/
ml) and actinomycin-D (5 µg/ml) were added 5 min before nicotine. Cells were lysed for
immunoblots 30 min after exposure to nicotine and/or LPS. Peritoneal macrophages were
transfected with the Effectine reagent (Qiagen) according to the manufacturer's instructions.
A cytomegalovirus-driven Renilla luciferase reporter plasmid was cotransfected to allow
assessment of transfection efficiency. The pCAGGS-neo expression vectors encoding wildtype hemagglutinin-tagged STAT3 or the dominant negative mutant hemagglutinin-tagged
STAT3D17 were provided by I. Touw (Erasmus University, Rotterdam, The Netherlands) and
T. Hirano (Osaka University, Osaka, Japan). In hemagglutinin-tagged STAT3D, glutamic
acids 434 and 435 were replaced by alanines17. After transfection, cells were selected for
16 h with neomycin (2.0 mg/ml; Sigma-Aldrich), were washed and were treated with nicotine
69
peritoneal cavity with 5 ml of ice-cold Hank’s balanced salt solution containing 10 U/ml
and LPS 24 h after transfection. Transfection was verified by immunoblot with horseradish
peroxidase−tagged rabbit polyclonal anti-hemagglutinin (Abcam). For small interfering RNA
transfection, cells were transfected with a small interfering RNA oligonucleotide specific to
SOCS3 (ID 160220; Ambion) using RNAiFect (Qiagen) according to the manufacturer’s
instructions. A fluorescein isothiocyanate−labeled control random RNA oligonucleotide
(Ambion) was cotransfected to optimize transfection efficiency.
Immunoblots.
Cells were scraped in 50 µl of ice-cold lysis buffer containing 150 mM NaCl, 0.5% Triton
X-100, 5 mM EDTA and 0.1% SDS. Samples were 'taken up' in 50 µl sample buffer (125
mM Tris-HCl, pH 6.8, 2% SDS, 10% β-mercaptoethanol, 10% glycerol and 0.5 mg/ml of
bromophenol blue), were separated by SDS-PAGE and were blotted onto polyvinyldifluoride
membranes (Millipore). Membranes were blocked in 0.1% Tween-20 in Tris-buffered saline
containing 5% nonfat dry milk and were incubated overnight with the appropriate antibodies
in 1% BSA and 0.1% Tween-20 in Tris-buffered saline. Horseradish peroxidase−conjungated
secondary antibodies were visualized with Lumilite plus (Boehringer-Mannheim).
Immunoprecipitation.
Peritoneal macrophages at a density of 1 x 106 per cm2 were preincubated for 30 min
with 1 µM nicotine and 100 µM AG 490, were scraped in lysis buffer (20 mM Tris-HCl,
pH 7.6, 2.5 mM EDTA, 1 mM EGTA, 1% Triton X-100, 0.5% sodiumdeoxycholate, 10%
glycerol, 1 mM Na3VO4, 50 mM NaF, 1 µg/ml of aprotinin, 1 µg/ml of leupeptin and 1 mM
phenylmethyl sulfonyl fluoride), were sonicated for 10 s and were centrifuged at 4 °C for
20 min at 14,000g. Lysates preabsorbed to 20 µl protein A−protein G (Sigma-Aldrich)
were incubated overnight with the appropriate antibodies and were immunoprecipitated
with 40 µl protein A−protein G. Alternatively, the TrueBlot system (eBioscience)
was used for immunoprecipitation according to the manufacturer’s instructions.
Immunoprecipitates were recovered by centrifugation, were washed in ice-cold wash
buffer (0.1% Triton X-100 and 1 mM phenylmethyl sulfonyl fluoride in Tris-buffered saline)
and were ‘taken up’ in sample buffer (125 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol
and 0.5 mg/ml of bromophenol blue), followed by immunoblot as described above.
70
Surgical procedures.
Mice (female BALB/c) were used at 15−20 weeks of age. IL-6- and IL-10-deficient mice and
their respective C57BL/6 wild-type counterparts were obtained from Jackson Laboratories.
LysM-Cre Stat3fl/fl and Stat3fl/fl mice were maintained at Osaka University (Osaka, Japan).
Abdominal surgery with intestinal manipulation was done as described elsewhere11. Mice
(n = 10−12) were assigned to the following four groups: control surgery of laparotomy
only; laparotomy followed by intestinal manipulation combined with sham preparation of
or intestinal manipulation in combination with electrical stimulation of the vagus nerve.
Intestinal manipulation consisted of 5 min of manipulation of the distal duodenum to the
cecum with sterile moist cotton applicators. At 3 or 24 h after surgery, mice were killed
Chapter 4
the cervical area; laparotomy in combination with electrical stimulation of the vagus nerve;
by cervical dislocation. The small intestine was removed, flushed and fixed in ice-cold
100% ethanol for the preparation of whole mounts. Small intestinal muscularis strips were
prepared by pinning of freshly isolated intestinal segments in ice-cold PBS and removal
of mucosa facing upward. Muscle strips were ‘snap-frozen’ in liquid nitrogen and were
set by the Animal Ethics Committee of the University of Amsterdam (Amsterdam, The
Netherlands).
Electric stimulation of the vagus nerve.
Stimulation of the vagus nerve was essentially done as described2. The left cervical nerve
was prepared free from the carotid artery and was ligated with 6-0 silk suture. The distal
part of the ligated nerve trunk was placed in a bipolar platinum electrode unit. In some
experiments, the vagus nerve was transected and the distal part was stimulated. Voltage
stimuli (5 Hz for 2 ms at 1 or 5 V) were applied for 5 min before and for 15 min after the
intestinal manipulation protocol described above. For sham stimulation of the vagus nerve,
in control mice the cervical skin was opened and was covered by moist gaze for 20 min.
Local blockade of nicotinic receptors in the ileum was done as follows: in anesthetized
mice (n = 7), a midline laparotomy incision was made and 6 cm of ileum proximal to the
cecum was carefully externalized and placed in a sterile preheated tube. The segment
was incubated for 20 min with a preheated (37 °C) solution of hexamethonium (100 µM
71
stored at − 80 °C until analysis. All animal experiments were in compliance with guidelines
in 0.9% NaCl) or vehicle. The temperature of the intestinal tissue was monitored with a
thermal probe. Leakage of hexamethonium solution into the peritoneal cavity was strictly
avoided. After incubation, the hexamethonium solution was removed and the ileal segment
was washed three times with 0.9% NaCl and was included in the manipulation protocol.
Measurement of gastric emptying.
Gastric emptying of a semiliquid, noncaloric ‘test meal’ (0.5% methylcellulose) containing
10 MBq 99mTc was assessed by scintigraphic imaging as described37.
Quantification of leukocyte accumulation at the intestinal muscularis.
Myeloperoxidase activity in ileal muscularis tissue was assayed as a measure of leukocyte
infiltration as described11, 23. Whole mounts of ethanol-fixed ileal muscularis were prepared
and stained for myeloperoxidase activity as described11, 23.
RT-PCR.
Total RNA from tissue was isolated with Trizol (Invitrogen), treated with DNase and reverse-transcribed. The resulting cDNA (0.5 ng) was subjected to 40 cycles of Light Cycler
PCR (FastStart DNA Master SYBR Green; Roche). The primers used were as follows:
TNF antisense, 5’-AAAGCATGAT CCGCGACGT-3’, and sense, 5’-TGCCACAAGCAGGAATGAGAA-3’; MIP-2 antisense, 5’-AGTGAACTGCGCTGTCAATGC-3’, and sense,
5’-GCAAACTTTTTGACCGCCCT-3’; SOCS3 antisense, 5’-ACCTTTCTTATCCGCGACAG-3’, and sense, 5’-TGCACCAGCTTGAGTACACAG-3’; and glyceraldehyde phosphodehydrogenase (GAPDH) antisense, 5’- ATGTGTCCGTCGTGGATCTGA-3’, and sense,
5’-ATGCCTGCTTCACCACCTTCT-3’. PCR products were quantified with a linear regression method using the Log(fluorescence) per cycle number38 and data are expressed as
the percentage of GAPDH transcripts for each sample. For qualification, the resulting PCR
products were separated by 2.5% agarose gel electrophoresis and analyzed by ethidium
bromide staining.
Immunohistochemistry.
For double-labeling of macrophages and cholinergic fibers, Sprague-Dawley rats
(300−350 g; Harlan Industries) were anesthetized with pentobarbital sodium (90 mg/
kg intraperitoneally) and were perfused transcardially with heparinized saline (20 U/
72
ml) followed by ice-cold 4% phosphate-buffered paraformaldehyde, pH 7.4. Gastric and
intestinal tissue were extracted and were postfixed for a minimum of 2 h in the same
fixative. Tissue was cryoprotected overnight in 18% sucrose and 0.05% sodium azide in
0.01 M PBS. Flat sections 20 µm in thickness and cross-sections 25 µm in thickness of
the corpus and mid ileum were cut on a cryostat and were processed in PBS. Sections
were pretreated with 0.5% sodium borohydride in PBS and were subsequently blocked in
donkey normal serum. Monoclonal mouse anti−rat CD163 (ED2; Serotec) and polyclonal
0.05% sodium azide in PBS with 0.5% Triton X-100 and were incubated for 20 h at 20 °C
or for 48 h at 4 °C. Secondary antibodies used were indocarbocyanine-conjugated donkey
anti-mouse (Jackson ImmunoResearch) for ED2 and carbocyanine-conjugated donkey
Chapter 4
goat anti−vesicular acetylcholine transporter (Chemicon) were diluted in 0.1% gelatin and
anti-goat (Jackson Immuno-Research) for vesicular acetylcholine transporter in 0.05%
sodium azide in PBS with 0.5% Triton X-100. Sections were mounted in 100% glycerol with
the addition of 5% N-propyl gallate as an antifade agent.
Alexa 546−labeled dextran particles (molecular weight, 10,000; Molecular Probes) 24 h
before surgery. At 1 h after surgery, anesthetized mice were perfused with 10 ml of ice-cold
0.9% NaCl containing 1 mM Na3VO4, followed by 20 ml of ice-cold 4% formaldehyde solution,
pH 7.4. Intestinal tissue was isolated, fixed overnight in 4% formaldehyde, dehydrated
and embedded in paraffin. Sections 6 µm in thickness were cut and were immunostained
with polyclonal rabbit anti−phosphorylated STAT3 (Cell Signaling Technologies) and biotinlabeled anti-rabbit according to the manufacturer’s instructions. Biotin was visualized with
3-amino-9-ethyl carbazole (Sigma) as a chromogen, followed by counterstaining with
hematoxylin. Alternatively, Alexa 488−streptavidin (Molecular Probes) with 4,6-diamidino2-phenylindole nuclear counterstain was used for analysis by confocal microscopy.
Statistics.
Statistical analysis of the results was performed by variance followed by Dunnett’s posthoc test or nonparametric Mann-Whitney U tests with SPSS. A probability value (P) of less
than 0.05 was considered significant.
73
In vivo labeling of mouse phagocytes was achieved by intraperitoneal injection of 20 µg
Results
Nicotine activates STAT3 in macrophages
To study the cellular response of macrophages to nicotinic receptor activation, we isolated
peritoneal macrophages from mice and investigated the effect of nicotine on LPS-induced
cytokine production. Nicotine reduced the LPS-induced release of TNF, MIP-2 and IL-6 but
not IL-10 in a dose-dependent way (Fig. 1a), consistent with published reports on the antiinflammatory effect of nicotine on human and mouse monocyte-derived macrophages2, 4, 5.
Given the crucial function of STAT3 in anti-inflammatory responses6, 7, we hypothesized
that activation of STAT3 and its gp130-binding regulatory protein SOCS3 may be involved
in the anti-inflammatory effect of nicotine. Consistent with that hypothesis, we found that
nicotine treatment activated STAT3 as well as SOCS3 in resting and LPS-stimulated
primary peritoneal macrophages in a dose- and time-dependent way (Fig. 1b,c). Nicotine
activated STAT3 directly, as phosphorylation of STAT3 was not affected by the protein
synthesis inhibitors actinomycin D and cycloheximide (Fig. 1d). In contrast, interferon-γinduced STAT1 activation was not effected by nicotine (Fig. 1e). Thus, nicotine reduced
the production of proinflammatory cytokines and activated STAT3 as well as SOCS3 in
stimulated macrophages.
Cytokine (ng/ml)
A
TNF
0.8
MIP-2
6
IL-6
3
0.4
0.3
0.6
4
2
0.2
0.4
2
1
0.1
0
100 10 102 103
0
100 10 102 103
0
100 10 102 103
Nicotine (nM)
74
IL-10
0
100 10 102 103
B
C
No LPS
0
101 102 103
0
10
10
0
PY-STAT3
10
(
PY-STAT3
STAT3
STAT3
Time (min)
SOCS3
Nicotine ( 100nM)
10
60
ot e30 40
00 M
Chapter 4
20
S
STAT3
D Vehicle
–
+
Act-D
–
+
CHX
–
+
E
Nicotine (nM) 0
PY-STAT3
PY-STAT1
STAT3
STAT1
SOCS3
0
Time (mi
PY-STAT3
Actin
nicotine
(100nM)
60
IFN- 100 ng/ml
Vehicle
10
102
103
0
10
102
103
Actin
Actin
Figure 1. Nicotine attenuates peritoneal macrophage activation and induces phosphorylation of
STAT3 and SOCS3 expression.
(a) ELISA of TNF, MIP-2, IL-6 and IL-10 in the supernatants of peritoneal macrophages stimulated
with 100 ng/ml of LPS in vitro in the presence of nicotine (dose, horizontal axes). Data represent
mean ± s.e.m. of four independent experiments in triplicate. (b) Immunoblots for phosphotyrosineSTAT3 (PY-STAT3), STAT3 and SOCS3 in cell lysates of peritoneal macrophages stimulated with
1 ng/ml of LPS (right) or no LPS (left) in the presence of nicotine (concentration, above lanes).
Blot is representative of five independent experiments. (c) Immunoblot of phosphorylated STAT3
(PY-STAT3) and STAT3 in cell lysates of peritoneal macrophages stimulated with 100 nM nicotine
(time, above lanes). Blot is one representative of three independent experiments. (d) Immunoblot of
phosphorylated STAT3 (PY-STAT3), STAT3 and SOCS3 in cell lysates of peritoneal macrophages
pretreated with vehicle, actinomycin-D (Act-D) or cycloheximide (CHX) and incubated with saline (-)
or 100 nM nicotine (+). Blot is representative of three independent experiments. (e) Immunoblot of
phosphorylated STAT1 (PY-STAT1) and STAT1 in peritoneal macrophages incubated with nicotine
(concentration, above lanes) and stimulated with 100 ng/ml of interferon-γ (IFN-γ). Actin, loading
control.
75
nicotine (nM)
Time (min)
LPS (1 ng/ml)
g2
1
3
Vehicle
10 20 hi
30 40
Deactivation by nicotine requires STAT3 transactivation
We next sought to determine whether the anti-inflammatory effect of nicotine depended on
nuclear transactivation of phosphorylated STAT3. We overexpressed a dominant negative
form of STAT3 (STAT3D) in primary peritoneal macrophages. Dimerized STAT3D is altered
in its ability to bind DNA response elements and induce transcription of target genes16,
17
. Nicotine failed to reduce LPS-induced TNF release in LPS-stimulated macrophages
transfected with STAT3D but not those transfected with the STAT3 wild-type construct (Fig.
2a). Thus, the nicotine-induced inhibition of TNF release is dependent on STAT3 DNA
transactivation.
To evaluate whether SOCS3 expression is crucial to the nicotinic anti-inflammatory effect,
we abrogated SOCS3 expression in peritoneal macrophages with SOCS3-specific small
interfering RNA (Fig. 2b). SOCS3 expression was substantially decreased in response to
nicotine (less than 10% of that expressed in control transfected cells), whereas STAT3
activation was not affected (transfection efficiency was more than 90%; Fig. 2b). In
macrophages with reduced SOCS3, however, nicotine was still able to decrease endotoxininduced production of IL-6 (data not shown) and TNF in a concentration-dependent way,
although the reduction was less pronounced than that in control transfected cells (Fig.
2c). Thus, blockade of STAT3 transactivation counteracted the anti-inflammatory effects
of nicotine, whereas blockade of SOCS3 expression did not. These results indicate that
SOCS3 expression is not strictly required for the reduction in macrophage TNF release by
nicotine.
Figure 2. Inhibition of macrophage activation by nicotine requires transactivation of STAT3 but not
SOCS3 expression.
(a) TNF in the supernatants of peritoneal macrophages transiently transfected with dominant negative STAT3D, wild-type STAT3 (STAT3 WT)17 or empty vector (Vector), then incubated with nicotine
and stimulated with 10 ng/ml of endotoxin. Values are expressed as the percent of TNF released
without the addition of nicotine for each group. Data are mean s.e.m. of three independent experiments done in duplicate. *, P < 0.05 (one-way analysis of variance followed by Dunnett’s multiple
comparison test). (b) Immunoblot for phosphorylated STAT3 (PY-STAT3), STAT3 and SOCS3 in
peritoneal macrophages transiently transfected with control oligonucleotide or SOCS3-specific small
interfering RNA (siRNA), then incubated with 100 nM nicotine. Blot is representative of three independent experiments. (c) TNF in the culture supernatants of peritoneal macrophages transfected with
control oligonucleotide or SOCS3 siRNA, then preincubated with nicotine and stimulated with 10 ng/
ml of LPS. Data are presented as percentage of TNF produced without addition of nicotine for each
treatment group and are the mean ± s.e.m. of three independent experiments done in duplicate.
76
B
100
60
40
20
0
Nicotine (nM) 0
STAT3D
STAT3WT
Vector
PY-STAT3
SOCS3
0 10-1 100 101 102 103
STAT3
102 102
Chapter 4
*
SOCS3
80
Control
siRNA
Control
Percentage decrease in TNF
A
Percentage decrease in TNF
C
100
80
60
40
Control
SOCS3
20
0
0
100
101
Nicotine (nM)
102
77
Nicotine (nM)
STAT3 phosphorylation depends on α7 nAChR activation
To determine whether STAT3 activation by nicotine was mediated by nAChR, we
pretreated cells with nAChR antagonists. The nonselective antagonists hexamethonium
and d-tubocurarine prevented the STAT3 phosphorylation induced by nicotine (Fig. 3a).
In addition, the α7 nAChR−selective antagonists α-bungarotoxin and methyllycaconitine
blocked the nicotine-induced STAT3 activation (Fig. 3a). A prominent function for the α7
receptor in nicotine-induced deactivation of macrophages corroborates published reports
on human and mouse monocyte-derived macrophage cultures3, 4. The selective non-α7
nAChRv antagonist dihydro-β-erythroidine did not affect nicotine-induced STAT3 activation
(data not shown).
Blocking nAChR also counteracted the attenuation of proinflammatory mediator release by
nicotine in activated macrophages. Hexamethonium, d-tubocurarine and methyllycaconitine
prevented the reduction in endotoxin-induced release of IL-6 (Fig. 3b) and MIP-2 (data not
shown) by nicotine in a dose-dependent way. Hexamethonium (effective dose leading to
50% inhibition (ED50), 6.46 ± 2.90 nM) was more potent than methyllycaconitine (ED50, 24.0
± 3.4 nM) and was far more potent than d-tubocurarine (ED50, 0.80 ± 0.23 µM) in attenuating
the inhibition of IL-6 release (Fig. 3b). The high ED50 for d-tubocurarine is probably due to
its low affinity for α7 nAChRs18 and is in line with its modest inhibitory effect on STAT3
activation by nicotine (Fig. 3a). In addition to methyllycaconitine, α-bungarotoxin abolished
IL-6 reduction by nicotine. However, exposure of the cells to α-bungarotoxin decreased
IL-6 production in the presence and absence of nicotine, which compromised adequate
determination of its ED50 (data not shown). Thus, STAT3 activation is dependent on the
fig 3
activation of nAChRs by nicotine, most likely exclusive through activation of the α7 nAChR
subunit
A
Bgt
(1 mg/ml)
d-TC
(1 µM)
Hexa
(1 µM)
MLA
(1 µM)
Nicotine (nM) 0 101 102 103 0 101 102 103 0 101 102 103 0 101 102 103 0 101 102 103
PY-STAT3
STAT3
Actin
78
100
80
60
40
20
0
0
100
101
102 103
nAchR blocker (nM)
104
The macrophage α7 nAChR recruits Jak2
STAT3 phosphorylation normally requires activity of the cytoplasmic tyrosine kinase Jak2 8.
Therefore, we investigated whether STAT3 phosphorylation depended on Jak2 activity and
whether nAChRs expressed on macrophages recruit Jak2. Phosphorylation of STAT3 after
nicotine treatment of peritoneal macrophages was effectively blocked by AG 490, a selective
inhibitor of Jak2 phosphorylation19, 20 (Fig. 4a). In agreement with that finding, nicotine failed
to reduce IL-6 release by LPS-stimulated peritoneal macrophages treated with AG 490
(data not shown). Binding studies have distinguished two main categories of nAChRs
based on their affinity for either α-bungarotoxin (α7-containing homopentamers) or nicotine
(α4β2 pentamers)18. Because our blocking studies suggested involvement of the α7 nAChR
subtype, we analyzed putative associations of α7 with Jak220 by immunoprecipitation (Fig.
e Jon
4b). The α7 (56-kilodalton)21 receptor was expressed in primary peritoneal macrophage
lysates. Immunoprecipitation of Jak2 from peritoneal macrophage cell lysates showed a
weak association of Jak2 with the α7 receptor after culture in the absence of nicotine. To
investigate whether Jak2 is recruited to the nAChR and is phosphorylated after binding of
its ligand, we preincubated cells with nicotine. Nicotine exposure increased the amount of
α7 nAChR detected in Jak2 and phosphorylated Jak2 immunoprecipitates (Fig. 4b).
79
Chapter 4
IL-6 inhibition (%)
120
B
Figure 3. STAT3 phosphorylation by
nicotine is prevented by α7-selective
nAChR antagonists.
(a,b) Peritoneal macrophages were
pretreated with the nAChR blockers
d-tubocurarin (d-TC), α-bungarotoxin
(αBgt), hexamethonium (Hexa) or
α-methyllycaconitine (MLA) and were
incubated with nicotine (concentration,
above lanes). Lysates were collected
for immunoblot of phosphorylated
STAT3 (PY-STAT3), STAT3 and actin
(a) and IL-6 was measured in supernatants (b). (a) Blots are representative
of three independent experiments. (b)
Filled squares, hexamethonium; open
squares, methyllycaconitine; open
circles, d-tubocurarine). Data are presented as the percentage of inhibition
of IL-6 release measured without the
addition of an nAChR blocker and represent mean values ± s.e.m. of three
independent experiments done in triplicate.
e
A
Nicotine (nM)
vehicle (1%EtOH)
+AG 490 (10 M)
0
0
101
102
103
101
102
103
+AG 490 (100 M)
0
101
102
103
PY-STAT3
STAT3
1
2
No lysate
IB:
IP: 7
Lysate
B
3
4
5
–
–
+
–
+
+
Nicotine (1 M)
AG490 (100 M)
Jak2
PY-Jak2
7
Igh
IP: Jak2
7
Jak2
Igh
IP: PY-Jak2
7
PY-Jak2
Igh
Figure 4. Nicotine-induced STAT3 phosphorylation
occurs through activation of Jak2 that is recruited to
the α7 nAChR subunit. (a) Immunoblot of phosphorylated STAT3 (PY-STAT3) and STAT3 in peritoneal
macrophages incubated with AG 490 (concentrations,
above blots). Blot is representative of three independent experiments. (b) Immunoblots of peritoneal
macrophages treated with 1 M nicotine (lanes 4 and
5) or with 1 μM nicotine plus 100 μM AG 490 (lane 5).
Cell lysates were immunoprecipitated (IP) with antiα7 (top), anti-Jak2 (middle) or anti−phosphorylated
Jak2 (PY-Jak2; bottom), followed by immunoblot (IB;
antibodies, left margin). Lane 2, coprecipitate in the
absence of lysate (negative control). IgH, immunoglobulin heavy chain. Blots are representative of four
independent experiments.
80
To further demonstrate that Jak2 is phosphorylated after nAChR activation, we pretreated
cells with the Jak2 phosphorylation blocker AG 490 before adding nicotine. Cells treated
with AG 490 had reduced phosphorylated Jak2 in α7 immunoprecipitates, whereas Jak2
recruitment to the α7 receptor was not affected (Fig. 4b). The latter finding demonstrates
that Jak2 is recruited and phosphorylated after nicotine binding.
Stimulation of the vagus nerve ameliorates inflammation
inflammation in vivo. We assessed the effect of stimulation of the vagus nerve on the
inflammation that follows intestinal manipulation in our mouse model11, because this immune
response is associated with the activation of macrophages12, 22. We electrically stimulated
Chapter 4
We next evaluated whether activation of nAChR on macrophages would attenuate intestinal
the left cervical vagus nerve during intestinal manipulation surgery and investigated the
effects on muscular inflammation and gastric emptying 24 h later (Fig. 5). Consistent with
published findings11, 23, intestinal manipulation of mice resulted in a delayed gastric emptying
compared with that of mice that underwent only laparotomy, indicative of the development
43.0 ± 6.7% for intestinal manipulation). However, stimulation of the vagus nerve prevented
the intestinal manipulation−induced gastroparesis 24 h after surgery (gastric retention,
25.2% ± 3.2%; Fig. 5). Notably, stimulation of the vagus nerve in itself may alter gastric
emptying during the vagus stimulation protocol24. However, we found that stimulation of
the vagus nerve did not affect basal gastric emptying 24 h after surgery (gastric retention,
15.7% ± 3.6%; Fig. 5). The last finding demonstrates that normalization of gastric emptying
after stimulation of the vagus nerve was not a direct effect on gastric motility but resulted
from reduced inflammation of the manipulated bowel segment11.
We next analyzed muscularis tissue for granulocytic infiltrates by measuring myeloperoxidase
activity in muscularis tissue homogenates and quantifying cellular infiltrates (Fig. 6). The
intestinal manipulation−induced inflammation of the muscularis externa in mice that received
stimulation of the vagus nerve was reduced in a voltage-dependent way compared with
that of mice that received intestinal manipulation plus sham stimulation. Prior vagotomy
of the proximal end of the stimulated vagus nerve did not affect these results (data not
shown), indicating that the anti-inflammatory effect of stimulation of the vagus nerve was
not dependent on the activation of central nuclei, which confirms published reports2.
81
of postoperative ileus11 (gastric retention after 60 min, 14.5% ± 2.7% for laparotomy and
We next incubated intestinal segments with the nicotinic receptor blocker hexamethonium
before intestinal manipulation combined with stimulation of the vagus nerve. In intestinal
segments treated with hexamethonium, stimulation of the vagus nerve failed to prevent
inflammation, in contrast to incubation with vehicle (Fig. 6), demonstrating that the
anti-inflammatory effect of vagus stimulation acted through local activation of nicotinic
Relative gastric content (%)
receptors.
100
80
60
40
*
20
0
0
20
40
60
Time after oral gavage (min)
82
80
Figure 5. Perioperative electrical
stimulation of the left cervical vagus nerve prevents gastroparesis
24 h after surgery with intestinal
manipulation in mice.
Gastric emptying curves of a semiliquid ‘test meal’ are for mice that
underwent surgery with intestinal manipulation (filled circles),
control laparotomy surgery (gray
triangles), stimulation of the vagus nerve plus control laparotomy
surgery (gray diamonds) or stimulation of the vagus nerve plus surgery with intestinal manipulation
(filled squares). Values are means
± s.e.m.; n = 8−10. *, P < 0.05 for
gastric retention at 60 min, surgery
with intestinal manipulation versus
surgery with intestinal manipulation
plus stimulation of the vagus nerve
(Mann-Whitney U test).
Chapter 4
83
Figure 6. (see fullcolor chapter 11) Vagal nerve stimulation reduces recruitment of inflammatory
infiltrates to the intestinal muscularis by activating peripheral nicotinic acetylcholine receptors. MPO
activity measured in intestinal muscularis tissue homogenates isolated 24 h after surgery with IM.
VNS with 5V, but not 1V, -stimulus prevents the increased muscularis MPO activity elicited by IM.
Asterisks indicate significant differences in MPO activity in intestinal muscularis tissue from L control
and IM VNS5V determined by one-way ANOVA followed by Dunnett’s multiple comparison test. Data
represent mean ± SEM of 10-15 mice (a). MPO-activity containing cells were stained in whole mount
preparations of intestinal muscularis (b and c) prepared 24 hrs post-operatively . Mice underwent
IM with sham VNS (IM Sham), or IM combined with VNS using 1, or 5 V pulses (IM VNS1V, and IM
VNS5V) (b). Mice were pretreated with hexamethonium (100 M; Hexa) or vehicle and underwent
Laparotomy (L) with VNS (L VNS5V) or IM with VNS5V (e). MPO-positive cells were counted in five
consecutieve microscopic fields of whole mount preparations of the indicated groups. Asterisks indicate significant differences (P<0.05) from (left graph) L control and (right graph) IM VNS5Vgroups
using one-way ANOVA followed by Dunnett’s multiple comparison test. Data represent mean ± SEM
of 5-8 mice.
Stimulation of the vagus nerve activates STAT3 in vivo
To further investigate whether macrophages mediated the anti-inflammatory effect of
stimulation of the vagus nerve, we analyzed the expression of transcripts of macrophagederived inflammatory mediators in muscularis tissue 3 h after surgery. Stimulation of the
vagus nerve reduced the expression of Cxcl2 mRNA (Fig. 7a,b) and Ccl3 mRNA (data
not shown) but did not notably alter the expression of Tnf transcripts in muscularis tissue,
confirming earlier reports2, 4. However, when we analyzed peritoneal lavage fluid for the
presence of macrophage inflammatory mediators 3 h after intestinal manipulation, we found
that stimulation of the vagus nerve significantly reduced the secretion of TNF, IL-6, MIP-2
(Fig. 7c) and MIP-1α (data not shown) in the peritoneal cavity. This reduction was not due
to enhanced expression of IL-10, as stimulation of the vagus nerve was similarly potent in
reducing intestinal manipulation−induced inflammation in IL-10-deficient mice (Fig. 7d).
Moreover, the peritoneal IL-10 in wild-type mice did not reach the limit of detection (31 pg/
ml) at 1, 3 or 6 h after intestinal manipulation (data not shown). Expression of Socs3 (Fig.
7a) but not Socs1 (data not shown) was increased in muscularis tissue after stimulation of
the vagus nerve even in mice that underwent this stimulation without manipulation of the
bowel.
(bp)
A
Tnf
-123
Cxcl2
-107
Socs3
-121
Gapdh
-132
No
RT
Relative
mRNA expression
B
L
sham
L
VNS
IM
VNS
IM
sham
*
4
Tnf
20
2
0
*
Cxcl2
10
L
VNS
IM
sham
IM
VNS
0
*
*
5
10
L
sham
Socs3
L
sham
L
VNS
84
IM
sham
IM
VNS
0
L
sham
L
VNS
IM
sham
IM
VNS
R l
e
N
x e io
f
Cxcl2
Socs3 activation
Given the
of acetylcholine, cholinergic
regulation of macrophage
4 short half-life
most likely requires that cholinergic nerves be in close proximity to intestinal macrophages.
To investigate
this, we immunohistochemically double-labeled vesicular acetylcholine
*
2
1
transporter−positive vagal efferent fibers and macrophages in rat intestinal musclaris tissue.
1,600
800
800
20
0
*
MIP-2
L
sham
L
VNS
D
IM
sham
IM
VNS
Il10 –/–
IL-6 (pg/ml)
600
400
0
L
sham
L
VNS
*
IM
sham
IM
VNS
0
ND
L
sham
Il10 –/–
800
*
300
0
*
IL-6
Chapter 4
*
TNF
L
VNS
IM
sham
IM
VNS
*
*
300
ND
L
sham
L
VNS
IM
sham
IM
VNS
0
L
sham
L
VNS
IM
sham
IM
VNS
Figure 7. Vagal stimulation reduces intestinal manipulation-induced proinflammatory mediator expression and release in vivo, independent of IL-10 production.
(a,b) Real-time PCR for macrophage proinflammatory mediators (a, left margin; b, above graphs) of
RNA isolated from intestinal muscularis strips prepared 3 h after the following procedures: control
laparotomy surgery plus sham stimulation of the vagus nerve (L sham); control laparotomy surgery
plus stimulation of the vagus nerve with 5-V pulses (L VNS); surgery with intestinal manipulation plus
sham stimulation of the vagus nerve (IM sham); or surgery with intestinal manipulation plus stimulation of the vagus nerve with 5-V pulses (IM VNS). (a) No RT, no reverse transcriptase added to
reaction (to control for nonspecific amplification); bp, base pairs. (b) Quantification38 of data and normalization of results to the expression of GAPDH. (c) Release of macrophage proinflammatory mediators into peritoneal lavage fluid obtained 3 h after treatment of mice with the procedures described
in a,b. (d) IL-6 in peritoneal cavities (left) and myeloperoxidase-positive cells intestinal muscularis
tissues (right) of IL-10-deficient mice (open bars) and their wild-type counterparts (filled bars) after
treatment with the procedures described in a,b. Right, myeloperoxidase-positive cells were quantified
in whole-mount preparations of intestinal muscularis tissue isolated 24 h after the procedures. *, P <
0.05, compared with the respective control laparotomy surgery group (one-way ANOVA followed by
Dunnett’s multiple comparison test (b,c) or Mann Whitney U test (d)). Data represent mean ± s.e.m.
of five to eight mice. ND, not detectable.
85
40
Cell density (cells/mm2)
Cytokine (pg/ml)
C
t
Macrophages were in close proximity to nerve terminals in the myenteric plexus in the
ileum (Fig. 8a) and circular muscle of gastric corpus (data not shown). Hence, acetylcholine
released from efferent nerve terminals could easily reach macrophages in the nanomolar
concentration range.
To verify that the enhanced SOCS3 expression reflected increased STAT3 activation in vivo,
we immunohistochemically analyzed intestinal tissues for the presence of phosphorylated
STAT3 in mice that underwent control laparotomy surgery, intestinal manipulation alone
or intestinal manipulation plus stimulation of the vagus nerve (Fig. 8b,c). We found
phosphorylated STAT3−positive nuclei in mice that underwent control laparotomy (Fig. 8b).
Intestinal manipulation resulted in the appearance of phosphorylated STAT3−positive cells
adhering to the serosal site of the bowel wall, most probably granulocytes and monocytes
recruited to the peritoneal compartment as a result of tissue trauma inflicted by the intestinal
manipulation procedure. However, when stimulation of the vagus nerve was applied, we
noted phosphorylated STAT3−positive nuclei in cells between longitudinal and circular
muscle layers surrounding the myenteric plexus. To identify the cellular source of the
phosphorylated STAT3−positive nuclei, we labeled tissue phagocytes in vivo by pretreating
mice with Alexa 546−labeled dextran particles (molecular weight, 10,000). This procedure
labels F4/80 antigen−positive macrophages populating the intestinal muscularis25. Most
of phosphorylated STAT3−positive nuclei in intestinal tissue of mice that had undergone
stimulation of the vagus nerve localized together with cells that had taken up Alexa
546−labeled dextran particles, indicating that these phosphorylated STAT3−positive nuclei
represented macrophages (Fig. 8c). These observations corroborate our in vitro findings
on the function of STAT3 in the cholinergic inhibition of tissue macrophages and are in
line with our proposed function of the network of resident intestinal macrophages26 as the
inflammatory cells targeted by stimulation of the vagus nerve. To further demonstrate that
the cholinergic anti-inflammatory pathway critically depends on STAT3 activation in vivo, we
studied the inflammatory response to intestinal manipulation in mice specifically deficient
in STAT3 in their myeloid cell lineage (called ‘LysM-Stat3fl/-’ mice here). LysM-Stat3fl/- mice
lack STAT3 in their macrophages and granulocytes6. In Stat3fl/+ control mice as well as in
LysM-Stat3fl/- mice, intestinal manipulation led to increased peritoneal IL-6 (Fig. 9a) as well
as massive inflammatory infiltrates in the manipulated muscularis tissue (Fig. 9b). Notably,
however, stimulation of the vagus nerve reduced peritoneal IL-6 and intestinal inflammation
86
in Stat3fl/+ control mice but failed to do so in LysM-Stat3fl/- mice. These data support the
87
Figure 8. Stimulation of the vagus nerve activates STAT3 in intestinal macrophages in muscularis
tissue. (see fullcolor chapter 11)
Cholinergic nerve fibers are in close anatomical apposition to macrophages in small intestine. (a)
Confocal microscopy of macrophages (ED2; red) and cholinergic nerve fibers (vesicular acetylcholine transporter; green) around the myenteric plexus of rat ileum. Arrows indicate close anatomical appositions of varicose cholinergic nerve fibers and macrophages at the perimeter of myenteric
ganglia and the tertiary plexus outside the ganglia (arrowheads). Scale bar, 10 μm. (b) Mouse ileum
sections stained for phosphorylated STAT3 1 h after control laparotomy surgery (L sham), intestinal
manipulation (IM sham) or intestinal manipulation combined with stimulation of the vagus nerve (IM
VNS). Transverse section of a complete ileal villus of a control
mouse (control laparotomy). SM, submucosa; CM, circular
muscle layer; LM, longitudinal muscle layer; MP, myenteric
plexus. Arrowheads indicate phosphorylated STAT3−positive
nuclei. Scale bar, 20 μm (40 μm for left image). (c) Phosphorylated STAT3−positive nuclei (green) in mouse ileum 1
h after intestinal manipulation plus stimulation of the vagus
nerve, visualized by confocal microscopy. Arrowheads indicate colocalization of phosphorylated STAT3 nuclei (PYSTAT3; green) with phagocytes prelabeled by prior injection
of Alexa 546−labeled dextran particles (red). Nuclear counterstain is 4,6-diamidino-2-phenylindole (DaPi; blue). Inset,
enlarged macrophage showing dextran particles and STAT3
immunoreactivity. Scale bar, 20 μm (10 μm for boxed area).
Experiments are representative of three independent incubations in three mice per group.
Chapter 4
critical function of STAT3 activation in the cholinergic anti-inflammatory pathway in vivo.
Discussion
The cholinergic anti-inflammatory pathway represents a physiological system for controlling
macrophage activation and inflammation in sepsis models1. Its working mechanism
ultimately involves the prevention of NF-κB p65 activity3 after α7 nAChR activation4, but
the exact cellular mechanism has remained unclear. Here we have demonstrated that
nicotine acts on macrophages via the recruitment of Jak2 to the α7 nAChR and activation
of Jak2, thereby initiating the anti-inflammatory STAT3 and SOCS3 signaling cascade.
Notably, recruitment of Jak2 to the α7 nAChR subunit has also been described in neuronal
PC12 cells exposed to nicotine, as part of a neuroprotective mechanism against β-amyloidinduced apoptosis20. Our results in resident peritoneal macrophages were consistent with
our in vivo data, as we found activation of STAT3 in intestinal macrophages in response
to stimulation of the vagus nerve in mice, which indicates activation of STAT3 induced by
acetylcholine derived from vagal efferents.
IL-6 (pg/ml)
LysM-STAT3fl/–
600
B
*
* *
300
0
L
sham
IM
sham
Cell density(cells/mm2)
A
1,000
IM
VNS
LysM-STAT3fl/– *
*
*
500
0
L
sham
IM
sham
IM
VNS
Figure 9. Stimulation of the vagus nerve fails to reduce inflammation in LysM-Stat3fl/- mice.
IL-6 was measured in peritoneal lavage fluid (a) and myeloperoxidase-positive inflammatory infiltrates were quantified in muscularis tissues (b) of Stat3fl/+ control mice (filled bars) or LysM-Stat3fl/mice (open bars) treated with control laparotomy surgery (L sham), intestinal manipulation (IM sham)
or intestinal manipulation plus stimulation of the vagus nerve (IM VNS). (a) Peritoneal lavage fluid
was collected 3 h after the procedures. (b) Infiltrates were quantified in whole-mount preparations
24 h after the surgical procedures. Data are mean ± s.e.m.; n = 3−4. *, P < 0.05, compared with the
respective control laparotomy surgery group (Mann-Whitney U test).
88
Activation of the STAT3 cascade after nAChR ligation is fully consistent with the observed
inhibition of proinflammatory cytokine release by macrophages, because STAT3 is a
negative regulator of the inflammatory response6, 27. In our studies, the anti-inflammatory
effect of nicotine on macrophages required DNA binding and transactivation of STAT3, as
nicotine failed to inhibit TNF production in macrophages overexpressing STAT3 altered
in its in DNA-binding capacity17. Likewise, activation of STAT3 is required for the anti-
and proliferation28. In addition, STAT3 phosphorylation is required for IL-6-induced growth
arrest and differentiation29.
Chapter 4
inflammatory properties of IL-108,28 and the IL-10-induced attenuation of cytokine production
SOCS3 specifically disables STAT3 phosphorylation via IL-6R but does not interfere with
IL-10R signaling9,
. Conditional knockout mice specifically lacking SOCS3 in their
10, 30
macrophages (LysM-Socs3fl/-) show resistance to endotoxemia, explained by the antiinflammatory effect of sustained STAT3 activation through IL-6R ligands10. Regardless
not contribute to the anti-inflammatory effect of nAChR activation, as blockade of SOCS3
expression did not prevent the anti-inflammatory action of nicotine. Hence, the antiinflammatory effect of cholinergic activation in macrophages rests mainly on enhanced
STAT3 rather than SOCS3 activation.
We have shown that STAT3 was activated by nicotine directly and that involvement of
enhanced signaling via IL-10R here was unlikely, as we found the macrophage deactivation
induced by stimulation of the vagus nerve to be similarly effective in IL-10-deficient mice.
Moreover, nicotine-induced STAT3 activation could be prevented by nAChR blockers.
Our observations suggest that the molecular route exerting the anti-inflammatory effect
of nAChR activation mimics the signaling pathway of IL-10R without the requirement of
IL-10 itself. That hypothesis is supported by our finding and those of another study2 that,
consistent with the action of IL-1031, nicotine does not alter TNF mRNA expression but
decreases the release of TNF protein. Furthermore, LysM-Stat3fl/- mice have a phenotype
resembling that of IL-10-deficient mice6. Nicotine-induced inhibition of the release of highmobility group box 1 in mouse RAW264.7 macrophages is associated with inhibition of
NF-κB p65 transcriptional activity3. Our finding that nicotine repressed macrophage activity
89
of that finding, our results have indicated that the enhanced expression of SOCS3 did
via STAT3 may very well explain that observation, as IL-10−STAT328 signaling blocks NFκB DNA-binding32, 33, possibly through direct interaction of dimerized STAT3 with the p65
subunit34.
We have shown here that recruitment of inflammatory infiltrates induced by bowel
manipulation and the resulting symptoms of postoperative ileus were reduced substantially
by stimulation of the vagus nerve. Our results have shown strict cholinergic control of
macrophage activation in vivo, which may be substantiated by the observation that
cholinergic (vesicular acetylcholine transporter−positive) nerve fibers are situated in close
proximity to resident macrophages in intestinal myenteric plexus. At first glance, our data
may seem contradictory to the outcome of earlier attempts to treat postoperative ileus
using cholinergic agents such as neostigmine, which had only limited success35. That lack
of efficacy could be explained by the fact that the inflammatory process had already been
fully accomplished by the time these agents were administered, leaving the activation of
inhibitory neural pathways11 unaffected. Our results indicate that nicotinic receptor activation
before or during surgery prevents postoperative intestinal inflammation and will certainly be
a promising strategy for treating postoperative ileus. Notably, vagus nerve stimulators are
clinically approved devices for the treatment of epilepsy and depression36. In conclusion,
we have shown here that inhibition of macrophage activation via the cholinergic antiinflammatory pathway is brought about via Jak2-STAT3 signaling. Our data may aid in
further development of therapeutic strategies for modifying the cholinergic anti-inflammatory
pathway to treat various inflammatory conditions.
90
91
1. Tracey, K.J. The inflammatory reflex. Nature 420, 853–859 (2002). 2. Borovikova, L.V. et al. Vagus nerve stimulation attenuates the systemic inflammatory response
to endotoxin. Nature 405, 458–462 (2000). 3. Wang, H. et al. Cholinergic agonists inhibit HMGB1 release and improve survival in experimental sepsis. Nat. Med. 10, 1216–1221 (2004). 4. Wang, H. et al. Nicotinic acetylcholine receptor α7 subunit is an essential regulator of inflammation. Nature 421, 384–388 (2003). 5. Matsunaga, K., Klein, T.W., Friedman, H. & Yamamoto, Y. Involvement of nicotinic acetylcholine
receptors in suppression of antimicrobial activity and cytokine responses of alveolar macrophages to Legionella pneumophila infection by nicotine. J. Immunol. 167, 6518–6524 (2001). 6. Takeda, K. et al. Enhanced Th1 activity and development of chronic enterocolitis in mice devoid
of Stat3 in macrophages and neutrophils. Immunity 10, 39–49 (1999).
7. Welte, T. et al. STAT3 deletion during hematopoiesis causes Crohn’s disease-like pathogenesis and lethality: a critical role of STAT3 in innate immunity. Proc. Natl. Acad. Sci. USA 100,
1879–1884 (2003). 8. Levy, D.E. & Lee, C.K. What does Stat3 do? J. Clin. Invest. 109, 1143–1148 (2002).
9. Kubo, M., Hanada, T. & Yoshimura, A. Suppressors of cytokine signaling and immunity. Nat.
Immunol. 4, 1169–1176 (2003). 10. Yasukawa, H. et al. IL-6 induces an anti-inflammatory response in the absence of SOCS3 in
macrophages. Nat. Immunol. 4, 551–556 (2003). 11. de Jonge, W.J. et al. Postoperative ileus is maintained by intestinal immune infiltrates that activate inhibitory neural pathways in mice. Gastroenterology 125, 1137–1147 (2003). 12. Kalff, J.C. et al. Intra-abdominal activation of a local inflammatory response within the human
muscularis externa during laparotomy. Ann. Surg. 237, 301–315 (2003). 13. Livingston, E.H. & Passaro, E.P., Jr. Postoperative ileus. Dig. Dis. Sci. 35, 121–132 (1990). 14. Bauer, A.J. & Boeckxstaens, G.E. Mechanisms of postoperative ileus. Neurogastroenterol.
Motil. 16, 54–60 (2004). 15. Mikkelsen, H.B., Mirsky, R., Jessen, K.R. & Thuneberg, L. Macrophage-like cells in muscularis
externa of mouse small intestine: immunohistochemical localization of F4/80, M1/70, and Iaantigen. Cell Tissue Res. 252, 301–306 (1988). 16. Shimozaki, K., Nakajima, K., Hirano, T. & Nagata, S. Involvement of STAT3 in the granulocyte
colony-stimulating factor-induced differentiation of myeloid cells. J. Biol. Chem. 272, 25184–
25189 (1997). 17. de Koning, J.P. et al. STAT3-mediated differentiation and survival and of myeloid cells in response to granulocyte colony-stimulating factor: role for the cyclin-dependent kinase inhibitor
p27(Kip1). Oncogene 19, 3290–3298 (2000). 18. Lukas, R.J. et al. International Union of Pharmacology. XX. Current status of the nomenclature
for nicotinic acetylcholine receptors and their subunits. Pharmacol. Rev. 51, 397–401 (1999). 19. Meydan, N. et al. Inhibition of acute lymphoblastic leukaemia by a Jak-2 inhibitor. Nature 379,
645–648 (1996). 20. Shaw, S., Bencherif, M. & Marrero, M.B. Janus kinase 2, an early target of alpha 7 nicotinic
acetylcholine receptor-mediated neuroprotection against A β-(1–42) amyloid. J. Biol. Chem.
277, 44920–44924 (2002). 21. Drisdel, R.C. & Green, W.N. Neuronal α-bungarotoxin receptors are alpha7 subunit homomers.
J. Neurosci. 20, 133–139 (2000). 22. Kalff, J.C. et al. Surgically induced leukocytic infiltrates within the rat intestinal muscularis mediate postoperative ileus. Gastroenterology 117, 378–387 (1999). 23. de Jonge, W.J. et al. Mast cell degranulation during abdominal surgery initiates postoperative
ileus in mice. Gastroenterology 127, 535–545 (2004).
24. Takahashi, T. & Owyang, C. Vagal control of nitric oxide and vasoactive intestinal polypeptide
release in the regulation of gastric relaxation in rat. J. Physiol. (Lond.) 484, 481–492 (1995). Chapter 4
Reference List
25. Ozaki, H. et al. Isolation and characterization of resident macrophages from the smooth muscle
layers of murine small intestine. Neurogastroenterol. Motil. 16, 39–51 (2004). 26. Mikkelsen, H.B. Macrophages in the external muscle layers of mammalian intestines. Histol.
Histopathol. 10, 719–736 (1995). 27. Wang, T. et al. Regulation of the innate and adaptive immune responses by Stat-3 signaling in
tumor cells. Nat. Med. 10, 48–54 (2004). 28. Williams, L.M., Ricchetti, G., Sarma, U., Smallie, T. & Foxwell, B.M. Interleukin-10 suppression
of myeloid cell activation - a continuing puzzle. Immunology 113, 281–292 (2004 .
29. Akira, S. et al. Molecular cloning of APRF, a novel IFN-stimulated gene factor 3 p91-related
transcription factor involved in the gp130-mediated signaling pathway. Cell 77, 63–71 (1994). 30. Lang, R. et al. SOCS3 regulates the plasticity of gp130 signaling. Nat. Immunol. 4, 546–550
(2003).
31. Kontoyiannis, D. et al. Interleukin-10 targets p38 MAPK to modulate ARE-dependent TNF
mRNA translation and limit intestinal pathology. EMBO J. 20, 3760–3770 (2001). 32. Wang, P., Wu, P., Siegel, M.I., Egan, R.W. & Billah, M.M. Interleukin (IL)-10 inhibits nuclear factor κB (NF κB) activation in human monocytes. IL-10 and IL-4 suppress cytokine synthesis by
different mechanisms. J. Biol. Chem. 270, 9558–9563 (1995). 33. Schottelius, A.J., Mayo, M.W., Sartor, R.B. & Baldwin, A.S., Jr. Interleukin-10 signaling blocks
inhibitor of κB kinase activity and nuclear factor κB DNA binding. J. Biol. Chem. 274, 31868–
31874 (1999). 34. Yu, Z., Zhang, W. & Kone, B.C. Signal transducers and activators of transcription 3 (STAT3)
inhibits transcription of the inducible nitric oxide synthase gene by interacting with nuclear factor κB. Biochem. J. 367, 97–105 (2002). 35. Longo, W.E. & Vernava, A.M., III. Prokinetic agents for lower gastrointestinal motility disorders.
Dis. Colon Rectum 36, 696–708 (1993). 36. George, M.S. et al. Vagus nerve stimulation. A potential therapy for resistant depression? Psychiatr. Clin. North Am. 23, 757–783 (2000).
37. Bennink, R.J. et al. Validation of gastric-emptying scintigraphy of solids and liquids in mice using dedicated animal pinhole scintigraphy. J. Nucl. Med. 44, 1099–1104 (2003). 38. Ramakers, C., Ruijter, J.M., Deprez, R.H. & Moorman, A.F. Assumption-free analysis of quantitative real-time polymerase chain reaction (PCR) data. Neurosci. Lett. 339, 62–66 (2003). 92
93
Chapter 4
5
5
Chapter
5
Activation of the Cholinergic
Anti-Inflammatory
Ameliorates
Pathway
Postoperative
Ileus in Mice
Gastroenterolgy 2007; 133: 1219-1228
Frans O. The,
Guy E. Boeckxstaens,
Susanne A. Snoek,
Jenna L. Cash,
Roelof J. Bennink,
Gregory J. Larosa,
David R. Greaves,
Wouter J. de Jonge
Abstract
Background & Aims: We previously showed that intestinal inflammation is reduced by
electrical stimulation of the efferent vagus nerve, which prevents postoperative ileus in
mice. We propose that this cholinergic anti-inflammatory pathway is mediated via alpha7
nicotinic acetylcholine receptors expressed on macrophages. The aim of this study was to
evaluate pharmacologic activation of the cholinergic anti-inflammatory pathway in a mouse
model for postoperative ileus using the alpha7 nicotinic acetylcholine receptor-agonist ARR17779. Methods: Mice were pretreated with vehicle, nicotine, or AR-R17779 20 minutes
before a laparotomy (L) or intestinal manipulation (IM). Twenty-four hours thereafter gastric
emptying was determined using scintigraphy and intestinal muscle inflammation was
quantified. Nuclear factor-κB transcriptional activity and cytokine production was assayed
in peritoneal macrophages. Results: Twenty-four hours after surgery IM led to a delayed
gastric emptying compared with L (gastric retention: L + saline 14% ± 4% vs IM + saline
38% ± 10%, P = 0.04). Pretreatment with AR-R17779 prevented delayed gastric emptying
(IM + AR-R17779 15% ± 4%, P = 0.03). IM elicited inflammatory cell recruitment (L + saline
50 ± 8 vs IM + saline 434 ± 71 cells/mm2, P = 0.001) which was reduced by AR-R17779
pretreatment (IM + AR-R17779 231 ± 32 cells/mm2, P = 0.04). An equimolar dose of nicotine
was not tolerated. Subdiaphragmal vagotomy did not affect the anti-inflammatory properties
of AR-R17779. In peritoneal macrophages, both nicotinic agonists reduced nuclear factor
κB transcriptional activity and proinflammatory cytokine production, with nicotine being
more effective than AR-R17779. Conclusions: AR-R17779 treatment potently prevents
postoperative ileus, whereas toxicity limits nicotine administration to ineffective doses.
Our data further imply that nicotinic inhibition of macrophage activation may involve other
receptors in addition to alpha7 nicotinic acetylcholine receptor.
96
P
Background
Postoperative ileus (POI) is characterized by impaired propulsive function of the entire
gastrointestinal tract after abdominal surgery.1 Although normal peristalsis is restored after
3–5 days, POI inflicts patient discomfort (eg, nausea, vomiting, abdominal pain), accounts
for a considerable increase in morbidity, and prolongs hospitalization.2 The additional
annual health care expenses related to POI in the United States are estimated to exceed
Research in rodent models for this pathologic condition has revealed that handling of
the intestine during abdominal surgical procedures initiates a biphasic response. Initially,
spinal and supraspinal inhibitory pathways become activated, enhancing central release
of corticotrophin- releasing factor.3,4 This sympathetic stress response results in the
Chapter 5
1 billion dollars.1,2
instantaneous impairment of gastric emptying, lasting up to 3 hours.4 An inflammatory
response of the muscularis propria mediates the delay in gastrointestinal transit observed
up to 24 hours thereafter and represents the prolonged phase in POI.5–7 The importance
of this induced intestinal leukocyte infiltration in the pathogenesis of POI is stressed by the
observation that prevention of this inflammation ameliorates POI.5 Intercellular adhesion
molecule 1 targeting antibodies or antisense oligonucleotides prevent extravasation of
reduced POI.6 Activation of macrophages that reside in the intestinal muscle layer have
been implicated to play an important role in the initiation of the manipulation-induced muscle
inflammation.7 Recently, the vagus nerve has been put forward to represent an inhibitory
feedback mechanism that negatively regulates innate immune responses.8,9 Enhanced
efferent vagal nerve output has been shown to reduce inflammatory responses in rodent
models for sepsis, ischemia/ reperfusion, pancreatitis, and POI.10–13 Its anti-inflammatory
potency most likely involves activation of the nicotinic acetylcholine receptors (nAChRs)
on immune cells such as macrophages.10,13,14 The cellular pathways of nicotinic inhibition
of macrophage activation involves the anti-inflammatory Janus kinase 2 (Jak2)/signal
transducer and activator of transcription 3 (STAT3) signaling pathway10 and inhibition of
nuclear factor κB (NF-κB) signaling.15
97
Peripheral Nicotinic Agonists Ameliorate Postoperative Ileus
leukocytes to the intestinal muscle layer and normalize gastric emptying, indicating a
Previous studies have indicated that nicotine has anti-inflammatory properties. Ghia et al16
recently showed an important role for cholinergic inflammatory control in 2 experimental
colitis models. Chemical as well as surgical blockade of vagal nerve signaling results in a
significant increase of inflammation. Conversely, nicotine treatment resulted in reduction
of the inflammatory response, independent of vagal nerve activity. However, even though
some clinical studies evaluating the role of nicotine in inflammatory bowel disease show
improvement compared with placebo, results generally are disappointing and administration
provokes significant toxic adverse events.17 Given the purported role of alpha7 nAChRs in
mediating the cholinergic anti-inflammatory pathway, 8,10,14 specific alpha7 nAChR agonists
may have higher therapeutic potential than general nicotinic agonists. Because of the
growing interest in manipulation of central nAChRs to treat neuropsychologic disorders
such as Alzheimer’s disease, attention deficit hyperactivity disorder, and schizophrenia,
several of such agonists have been developed in the past decade.18 The nAChR agonist
AR-R17779, a spirooxazolidinone, has a high affinity for the alpha7 receptor subtype19 and
potently activates peripheral as well as central alpha7 nAChRs.19–22
In the present study, we show that the anti-inflammatory efficacy attained with electrical
vagal nerve stimulation can be mimicked by AR-R17779. Pretreatment with AR-R17779
ameliorates POI and reduces the manipulation-induced inflammatory response, although
a similar treatment with nicotine is ineffective. On the other hand, although both nicotinic
agonists reduced activation of peritoneal macrophages, nicotine was more potent in
reducing cytokine release and NF-κB activation as compared with AR-R17779.
98
Reagents and Antibodies
Nicotine used ([-]-nicotine) and Zymosan A from S. cerevisiae, was from Sigma-Aldrich
(Zwijndrecht, the Netherlands). Antibodies against nAChR alpha7 were obtained from
Abcam (Cambridge, UK), anti–STAT-3 (PY705) from Cell Signaling Technology (Beverly,
MD), and goat polyclonal anti–β-actin, rabbit polyclonal anti–STAT-3 were from Santa Cruz
Biotechnology, Inc (Santa Cruz, CA). Rat monoclonal anti-F4/80 was obtained from Serotec
derived chemokins (KC), tumor necrosis factor (TNF), and RANTES were purchased from
R&D Systems (Minneapolis, MN).
Animals
Chapter 5
(Oxford, UK). The enzyme-linked immunosorbent assays for interleukin-6, Keratinocyl-
Female Balb/C mice (Harlan Nederland, Horst, The Netherlands), 12–15 weeks, were kept
under environmentally controlled conditions (light on from 8:00 AM until 8:00 PM; water
and rodent nonpurified diet ad libitum; temperature, 20°C–22°C; humidity, 55%). LysMCre
and STAT-3flox/flox mice23 were kindly made available by Dr S. Uematsu and Professor S.
Akira (Osaka University, Osaka, Japan). All experiments were performed according to the
Study Protocol
Mice were assigned randomly to 1 of 7 treatment groups (ie, sham [stimulation], 5 V electrical
vagus nerve stimulation, vehicle [saline], nicotine at 0.9 or 23.0 µmol/kg or AR-R17779 at
0.09, 0.9, or 23.0 µmol/kg). The assigned therapy was administered via intraperitoneal
injection 20 minutes before the surgical procedure was performed as described in the next
section. A subgroup of animals underwent a subdiaphragmal vagotomy 30 minutes before
treatment with saline or AR-R17779.
Surgical Procedures
Mice were anesthetized by intraperitoneal injection of a mixture of Fentanyl Citrate/
Fluanisone (Hypnorm; Janssen, Beerse, Belgium) and Midazolam (Dormicum; Roche,
Mijdrecht, The Netherlands). The surgical procedure was performed under sterile conditions.
99
guidelines of the Ethical Animal Research Committee of the University of Amsterdam.
Mice underwent a laparotomy (L) or small intestinal manipulation (IM) as described
previously.5 In short, the small intestine was exteriorized carefully and manipulated gently
for 5 minutes using sterile, moist cotton applicators. After repositioning of the intestinal
loops, the abdomen was closed using a 2-layer continuous suture (Syneture Sofsilk 6-0).
Mice recovered from surgery in a temperature-controlled cage at 32°C with free access to
water, but not food. Twenty-four hours after surgery, gastric emptying was measured by a
scintigraphic method23 and mice were killed by cervical dislocation. The small intestine was
removed, flushed in ice-cold phosphate-buffered saline (PBS), and snap-frozen in liquid
nitrogen or fixed in ethanol 100% for 10 minutes and stored in ethanol 70% at 4°C until
further analysis.
Electrical Vagal Nerve Stimulation
Electrical vagal nerve stimulation (EVNS) was performed as described previously.10 To
minimize cardiovascular responses, the left cervical branch was stimulated, avoiding
sinoatrial-induced bradycardia. Five-volt stimuli with a frequency of 5 Hz, 5 ms10 were
applied for 5 minutes before and 15 minutes after abdominal surgery. For sham stimulation,
a cervical midline incision was made, after which the wound was covered with sterile, moist
gauzes for 20 minutes.
Subdiaphragmal Vagotomy
A midline incision was made under general anesthesia, after which a retractor was placed.
Under microscopic view, both vagal nerve trunks were cut, distal from the diaphragm but
proximal to the division of the hepatic branch. During this procedure, the intraperitoneal
organs were protected and kept moist using sterile gauzes drenched in 0.9% NaCl. Any
palpation or manipulation of the small intestine was specifically avoided. The abdomen
was closed using a 2-layer continuous suture (Syneture Sofsilk 6-0). Animals were kept in
a temperaturecontrolled cage at 32°C for 30 minutes, after which they entered the study
protocol.
Measurement of Gastric Emptying
Gastric emptying was determined as described previously.24 After gavage of a semiliquid,
noncaloric test meal (0.1 mL of 3% methylcellulose solution containing 10 MBq of
99m
Tc-
Albures), mice were scanned using a gamma camera set at 140 keV.24 The entire abdominal
100
region was scanned for 30 seconds, immediately and 80 minutes after gavage.6,24 During
the scanning period mice were conscious and restrained manually. The static images
obtained were analyzed using Hermes computer software (Hermes, Stockholm, Sweden).
gastric region of interest compared with the total abdominal region of interest.6,24
Whole-Mount Preparation
As previously described,5 the mucosa was separated carefully from the muscle layer. Fixed
preparations were rehydrated by incubation in 50% ethanol and PBS, pH 7.4, for 5 minutes.
3-amino-9-ethyl carbazole (Sigma, St. Louis, MO) as a substrate and dissolved in sodium
acetate buffer (pH 5.0) to which 0.01% H2O2 was added.5 To quantify the extent of intestinal
muscle inflammation, the number of myeloperoxidase-positive cells in 3 randomly chosen
1-mm2 fields were counted and expressed as the number of myeloperoxidase-positive cells
Chapter 5
To visualize myeloperoxidasepositive cells preparations were incubated for 10 minutes with
per mm2.
Cell Culture and Immunohistochemistry
Resident peritoneal macrophages were harvested by flushing the peritoneal cavity with
5 mL of Hank’s balanced salt solution containing 10 U/mL heparin. Peritoneal cells were
plated in Opti-Mem I medium (Gibco, Carlsbad, CA), supplemented with 10 mmol/L
were left to adhere for 2 hours in a humidified atmosphere at 37°C with 5% CO2. Cells were
washed and adhering cells were left for 16–20 hours. Subsequently, cells were stimulated
with lipopolysaccharide (LPS) (Escherichia coli 100 ng/mL; Sigma-Aldrich) and interferon-γ
(10 ng/mL) in the presence of indicated concentrations of nicotinic agonist for 3 hours, or
lysed 30 minutes after nicotinic agonist/LPS/interferon-γ exposure for immunoblotting, as
described.10 For confocal microscopy, macrophages were left to adhere for 16–20 hours
on glass slides (Nunc, Rochester, NY) in RPMI medium supplemented with 10% fetal calf
serum. Cells were washed 5 times with ice-cold Hank’s balanced salt solution containing 1
mmol/L Na3VO4, and fixed in ice-cold 4% phosphate-buffered (pH 7.4) paraformaldehyde
for 1 hour. After washing with ice-cold PBS pH 7.4, cells were stained with appropriate
antibodies at 4°C for 16–20 hours. Antibodies were visualized using anti-rat Alexa546labeled secondary antibodies and biotinlabeled anti-rabbit antibodies, followed by Alexa
101
L-glutamine, 100 U/mL penicillin, and 100 µg/mL gentamycingentamycin. Macrophages
488-streptavidin (Molecular Probes). Sections were mounted in glycerol mounting medium
to which DAPI (10 µg/mL; Molecular Probes) nuclear counter stain was added.
Immunoblotting
As described,10 cells were scraped in 50 µL of ice-cold lysis buffer containing 150 mmol/L
NaCl, 0.5% Triton X-100, 5 mmol/L ethylenediaminetetraacetic acid, 0.1% sodium dodecyl
sulfate, 0.5% deoxycholate, 10% glycerol, 1 mmol/L Na3VO4, 50 mmol/L NaF, 1 µg/mL
aprotinin, 1 µg/mL leupeptin, and 1 mmol/L phenylmethylsulfonyl fluoride. Samples were
suspended in 50µL sample buffer (125 mmol/L Tris-HCl, pH 6.8, 2% sodium dodecyl
sulfate, 10% β-mercaptoethanol, 10% glycerol, and 0.5 mg/mL bromophenol blue),
loaded onto sodium dodecyl sulfate–polyacrylamide gel electrophoresis gels, and blotted
onto polyvinylidene difluoride membranes (Millipore). Membranes were blocked in Trisbuffered saline/0.1% Tween-20 containing 5% nonfat dry milk and incubated overnight with
appropriate antibodies in Tris-buffered saline/0.1% Tween-20/1% bovine serum albumin.
Horseradish-peroxidase–conjugated secondary antibodies were visualized using Lumilite
plus (Boehringer-Mannheim, Germany).
Reverse-Transcription Polymerase Chain Reaction
Total RNA from tissue was isolated using Trizol (Invitrogen, Carlsbad, CA), treated with DNase,
and reverse transcribed. The resulting complementary DNA (0.5ng) was subjected to Light
Cycler polymerase chain reaction (CYBR Green Fast start polymerase; Roche, Mannheim,
Germany) for 40 cycles. Primers used were TNFα forward 5’ GACAAGGCTGCCCCGACTA
3’;
reverse
5’AGGAGGTTGACTTTCT
CCTGGTATG
3’,
and
HPRT
forward
5’GACCGGTCCCGTCATGC 3’; reverse 5’ TCATA-ACCTGGTTCATCATCGC 3’; RANTES
forward 5’ GACACCACTCCCTGCTGCT 3’, reverse 5’ GAAATACTCCTTGACGTGGGCA
3’.
NF-κB Activity Assay
Immortalized splenic macrophages Mf4/425 (a kind gift from Professor M. P. Peppelenbosch,
University of Groningen, Groningen, the Netherlands) were co-transfected with NFκB luciferase and cytomegalovirus renilla luciferase reporter constructs (Clontech,
MountainView, CA) using Jet PEI (PolyTransfection), according to the manufacturer’s
instructions. Briefly, 0.5 µg per 106 cells of constructs NF-κB–luc and 5 ng cytomegalovirus
102
Renilla Luciferase was suspended in 75 µL of 150 mmol/L sterile NaCl solution. Also,
1.6 µL of Jet PEI solutions was suspended in 75 µL of 150 mmol/L sterile NaCl solution.
The Jet PEI/NaCl solution then was added to the DNA/NaCl solution and incubated at
room temperature for 30 minutes and 150 µL of the DNA/Jet PEI was added to the cells.
The transfection was allowed to proceed for 16 hours, and the medium was refreshed.
Twenty-four hours after transfection, cells were pretreated with nicotinic agonists at the
concentration indicated for 1 hour, and subsequently stimulated with zymosan (5 particles
per cell) for 6 hours. After treatment, the medium was removed; the cells were washed
3 times with ice-cold PBS, the cells were lysed with Passive Lysis Buffer supplied in the
activity according to the manufacturer’s instructions.
Statistics
Chapter 5
Dual Luciferase Reporter Assay Kit (Promega), and the lysate was assayed for luciferase
Statistical analysis was performed with the use of SPSS 12.02 software for Windows
(SPSS, Inc, Chicago, IL). Data were analyzed using the nonparametric Mann–Whitney
U test for independent samples. The Friedman’s 2-way analysis of variance was used to
explore multiple dependent value assays (ie, the in vitro inflammatory protein response).
If the Friedman’s analysis was significant, individual values compared with the 0-nmol/L
concentration were tested with a Mann–Whitney U test. P values less than 0.05 were
considered statistically significant and results were depicted as mean ± SEM.
103
Results
Electrical Vagus Nerve Stimulation and AR-R17779 Prevent POI
Consistent with our previous results,5,10 gastric emptying was impaired significantly in mice
subjected to IM when compared with L alone (Figure 1A). First, we confirmed that activation
of the cholinergic anti-inflammatory pathway by electrical stimulation of the vagus nerve
during surgery ameliorates POI. The postoperative delay in gastric emptying resulting from
IM was reduced significantly if the surgical procedure was combined with a 5-V electrical
stimulation of the left cervical vagus nerve for 20 minutes (Figure 1A).
A
relative
gastric content (%)
60
50
MPO-pos.
cells/mm 2
*
40
30
##
20
**
**
L IM
AR-R
0.2
L IM
AR-R
5.0
10
surgery
treatment
dose (mg/kg)
B
#
600
L IM
sham
L IM
EVNS
L IM
saline
-
#
L IM
nicotine
0.4
L IM
AR-R
0.02
*
500
400
##
300
**
200
100
surgery
treatment
dose (mg/kg)
L IM
sham
L IM
EVNS
L IM
saline
-
104
L IM
nicotine
0.4
L IM
AR-R
0.02
L IM
AR-R
0.2
L IM
AR-R
5.0
i
Gastric retention
measured 24 hours after electrical vagal nerve stimulation in L control
30
e
mice indicated that electrical vagal nerve stimulation by itself did not influence postoperative
**
**
t
gastric emptying 24 hours after the procedure (Figure 1A sham vs EVNS).
To explore the potential of pharmacologic nAChR activation of the cholinergic antiL IM ofL nAChR
urgery
IM next
L Iinvestigated whether
L IM
Ladministration
M
M
L agonist
M
inflammatory
pathway, we
dose ( or
g/ alpha7
g)
0 2
0.4
2
5 of ARnicotine
nAChR agonist AR-R17779 would
ameliorate
POI.
Administration
R17779
in
a dose of 0.2 or #5 mg/kg restored gastric emptying to the levels seen with L alone
B
600
M O o
c s
failed to improve gastric emptying (Figure 1A), although nicotine administration at this dose
0
range has been
reported to already cause significant behavioral changes in rats.20 Mice
treated with
* 10.6
300higher doses of nicotine (up to a dose equimolar to 5 mg/kg AR-R17779;
mg/kg nicotine) clearly developed behavioral agitation within seconds after administration,
Chapter 5
(Figure 1A). In contrast, nicotine at a dose equimolar to 0.2 mg/kg AR-R17779 (0.4 mg/kg)
illustrating marked neurotoxicity, in agreement with other studies.26 Therefore, no further
1 were conducted using nicotine at doses higher than 0.4 mg/kg.
experiments
C
600
#
#*
400
**
300
200
100
surgery
sham/VGX
treatment
L IM
sham
saline
L IM
VGX
saline
L IM
VGX
AR-R 5.0
Figure 1. POI is ameliorated after treatment with alpha7-selective agonists AR-R17779. (A) Gastric
retention 80 minutes after gavage of a semiliquid test meal in mice that had undergone the indicated
treatment and L or IM 24 hours previously. (B) Quantitative analysis of IM-induced inflammatory cell
recruitment 24 hours after indicated treatment and surgery. (C) Quantitative analysis of manipulationinduced inflammatory cell recruitment 24 hours after indicated treatment and surgery in VGX or
sham-vagotomized animals. Data shown are mean values ± SEM of 6–8 mice. L, ; IM, #P < 0.05
vs sham laparotomy. ##vs sham IM. *P < 0.05 vs saline laparotomy. **P < 0.05 vs saline IM.
105
MPO-pos.
cells/mm 2
500
Manipulation-Induced Inflammation Is Decreased by Electrical Vagal Nerve
Stimulation and Pretreatment With AR-R17779
We established previously that delayed gastric emptying results from an intestinal muscle
inflammation inflicted by the surgical bowel handling.5 Because electrical vagal nerve
stimulation and AR-R17779 also acts on neuronal receptors we investigated whether
the improved gastric emptying results from inhibition of the inflammatory response in
manipulated muscle tissue. IM initiated a marked recruitment of inflammatory cells to the
muscle layer of the handled small-bowel segment (Figure 1B). However, if IM surgery was
combined with electrical vagal nerve stimulation, the number of inflammatory cells recruited
to the muscle layer was reduced significantly (Figure 1B).
Next, we investigated whether preoperative treatment with nicotinic agonists would attain
a similar anti-inflammatory response. Figure 1B shows that IM results in a significant influx
of leukocytes into the small intestine 24 hours after surgery, whereas pretreatment with
AR-R17779 (5 mg/kg) significantly reduced the number of inflammatory cells infiltrating
the intestinal muscle segment in response to IM. Pretreatment with nicotine 0.4 mg/kg
A
80
**
60
100
*
*
**
**
40
20
0
0
1
10
Nicotinic agonist (nM)
100
KC % decrease
IL6 % decrease
TNF % decrease
100
80
**
60
1
10
20
0
1
*
*
AR-R17779
**
**
Nicotine
** **
40
10
20
0
*
60
0
100
40
0
80
100
106
100
% ec
ec
or its equimolar dose of AR-R17779 (0.2 mg/kg) failed to reduce the number
* of recruited
*
inflammatory cells significantly (Figure 1B).
0
0
AR-R17779
Pretreatment
Reduces
Intestinal Muscle
Inflammation
Independent
of
0
1
10
100
0
1
1
100
Vagal Nerve
Signaling
icotinic
agon
One potential mechanism for the observed anti-inflammatory effects of AR-R17779 could
be activation
of central nAChRs and subsequently increased vagal efferent activity. To
00
RR
79
e
investigate whether the effect achieved with AR-R17779 depends on enhanced vagal nerve
re
ico i in mice that had undergone
signaling, we tested whether AR-R17779
** 5 mg/kg was effective
C
not elicit an intestinal muscle inflammation (L sham vs L VGX, Figure 1C). Importantly, the
potency of AR-R17779 to reduce myeloperoxidase-positive infiltrate recruitment was not
0
100
affected by VGX (Figure 1C),
indicating
that the anti-inflammatory effect of AR-R17779 is
Chapter 5
d
a subdiaphragmal bilateral vagotomy (VGX) before treatment and IM or L. VGX in itself did
independent of vagal activity.
80
60
*
40
*
20
**
0
Vehicle
20
LPS
LPS
LPS
0
1
Nicotinet Nicotine
i (
MLA
100 RANTES
80
*
60
40
20
0
Vehicle
LPS
LPS
LPS
Nicotine Nicotine
MLA
)
KC
d cre se
Figure 2. Nicotinic agonists reduce
cytokine and chemokine production in macrophages. (A) TNF,
RR 7 7
interleukin-6, or KC release from primary peritoneal macrophages stimulated with LPS (100 ng/mL)
100 of nicotine (
co AR-R17779
ine*
in the
)N or
( )AR-R1
at the779
indicated concentration. Data shown are
** presence
**
*
80
mean percentages
compared with vehicle/endotoxin treatment baseline concentration ± SEM of 5–7
i
* < 0.05* vs 0 nmol/L nicotine
mice measured
or AR-R17779 baseline concentration.
60 in duplicate. *P
**P < 0.01 vs 0 nmol/L nicotine or AR-R17779 baseline concentration. (B) Nicotine suppresses the
up-regulation of inflammatory mediator transcripts by activated macrophages via alpha7 nAChR.
1
0
Peritoneal
macrophages were pretreated with vehicle (media), nicotine (80 nmol/L), or nicotine (80
i
i t ( )0
0
1
10(MLA;
10 5mmol/L) for 1 hour and then stimulated with LPS (100 ng/mL)
nmol/L) + methyllycaconitine
Nicotinic
agonist (nM)expression was normalized to hypoxanthine phosphoribosyltransfor 15 hours. Cytokine
transcript
ferase messenger RNA. Data shown are mean ± SEM from 4 independent experiments using cells
from different donors. Asterisks indicate significant differences (P < 0.05) relative to LPS treated
F
NTES
samples.
80
107
TNF
Normalised expression (%)
*
100
I
Normalised expression (%)
B
Macrophage Activation Is Modulated by Nicotine and AR-R17779
We hypothesized that AR-R17779 would exert its anti-inflammatory effect via activation
of peripheral nAChRs on macrophages because nicotinic agonists have been shown
previously to dose-dependently inhibit release of proinflammatory cytokines and chemokines
by macrophages stimulated with endotoxin.10,14 To assess the potency of AR-R17779 to
reduce inflammatory mediator release in vitro, peritoneal macrophages were stimulated
with LPS and interferon-γ in the presence of nicotine or AR-R17779 in a 0–1000 nmol/L
concentration range. As shown in Figure 2A, nicotine as well as AR-R17779 reduced TNF
and KC production in LPS activated macrophages, whereas interleukin-6 (and RANTES,
not shown) was reduced significantly only by nicotine. In conjunction, nicotine inhibited
transcription of TNF and RANTES, an effect that was blocked by a selective alpha7 nAChR
antagonist methyllycaconitine (MLA) (Figure 2B).
Nicotinic agonists have been shown previously to reduce pro-inflammatory cytokine
production via inhibition of NF-κB activation.8 The modest effect of AR-R17779 on proinflammatory mediator production prompted us to explore the potency of AR-R17779 to
reduce NF-κB transcriptional activity. To this end, we investigated the effect of nicotine and
AR-R17779 on NF-κB activation induced by zymosan particles in a reporter assay using
the immortalized splenic macrophage cell line Mf4/4,25 transiently transfected with a κB
responsive element linked to luciferase gene. As shown in Figure 3A, NF-κB transcriptional
activity was induced by zymosan particles. When cells were pretreated with nicotine or
AR-R17779, NF-κB transcriptional activity was reduced. Notably, however, although
nicotine reduced activity to background levels, AR-R17779 failed to reduce NF-κB activity
completely, even at concentrations as high as 10 µmol/L.
We previously reported that nicotinic stimulation of nAChRs on peritoneal macrophages
leads to activation of the Jak2/STAT3 pathway, diminishing its pro-inflammatory cytokine
release.10 We next investigated whether AR-R17779 activated similar pathways in
peritoneal macrophages. Both AR-R17779 and nicotine induced the phosphorylation of
STAT3 in peritoneal macrophages (Figure 4). Immunoblot analysis of cell lysates from
peritoneal macrophages confirmed this observation, showing that nicotine and AR-R17779
led to a dose-dependent increase in STAT3 phosphorylation, although AR-R17779 was less
effective compared with nicotine (Figure 4A). The results of the immunoblot were confirmed
108
by immunofluorescent staining of peritoneal macrophages with a phospho-STAT3–specific
antibody (Figure 4B). We earlier reported that in our model for POI, the anti-inflammatory
effect of vagal nerve activation depends on STAT3 activation. We subsequently investigated
whether the anti-inflammatory properties observed with AR-R17779 functionally depend on
STAT3 expression. To this end, we harvested peritoneal macrophages from LysM-Cre/
Stat3flox/- conditional knock-out mice that specifically lack STAT3 in their macrophages and
neutrophils.23 AR-R17779 elicited a dose-dependent reduction in TNF, interleukin-6, and
KC release in endotoxinstimulated peritoneal macrophages from unaffected Stat3flox/flox
control mice, but failed to reduce the release of these cytokines and chemokines in Stat3-
Chapter 5
deficient peritoneal macrophages (Figure 5).
0,6
0,6
*
*
0,4
*
0,2
0,0
vehicle
Nicotine
0.1 μM
Nicotine
1.0 μM
AR-R17779
1.0 μM
AR-R17779
10 μM
Figure 3. Nicotinic agonists reduce NF-kB transcriptional activity in macrophages. Macrophages (immortalized splenocytes Mf4/4) transiently transfected with a κB firefly luciferase reporter were treated
with nicotine or AR-R17779 at the indicated concentrations, and then stimulated with medium ( ) or
zymosan (5 particles per cell), respectively; ( ). Cells were co-transfected with a cytomegalovirus–
renilla luciferase construct to normalize for transfection efficiency. Shown are normalized means ±
SEM of 3–4 independent experiments in duplicate. Asterisks indicate significant differences (P < .05)
of LPS vs vehicle.
109
NF- B activity
(RLU luciferase/renilla)
0,8
Figure 4. (see fullcolor chapter 11) Nicotinic agonists induce STAT3 activation in peritoneal macrophages. (A) Immunoblots showing a dose-dependent increase of phosphorylated STAT3 in peritoneal macrophages treated with nicotine (0–100 nmol/L) or AR-R17779 (0–1000 nmol/L). Blots shown
are representative of 3 independent experiments. (B) Confocal images of peritoneal macrophages
attached to glass slides and stimulated with LPS (10 ng/mL) with the addition of either vehicle: ARR17779 (100 nmol/L), or nicotine (100 nmol/L). Treatment of cells with AR-R17779 (middle) or nicotine
(lower) enhances nuclear staining of phosphorylated STAT3 in F4/80 (red)-positive macrophages.
110
A
STAT3flox/flox
LysMCre/STAT3flox/
75
**
50
*
25
0
B
10
100
AR-R17779 (nM)
1000
STAT3flox/flox
LysMCre/STAT3flox/
IL-6
% decrease
100
75
50
**
25
0
0
10
100
AR-R17779 (nM)
**
1000
C
STAT3flox/flox
LysMCre/STAT3flox/
100
KC
% decrease
Chapter 5
0
75
50
**
**
100
1000
25
0
0
10
AR-R17779 (nM)
Figure 5. (A) TNF, (B) interleukin-6, and (C) KC release by peritoneal macrophages harvested from
STAT3flox/flox( ) controls or LysMCre/STAT3flox/flox( ) mice and stimulated with LPS (100 ng/mL)
in the presence of nicotine or AR-R17779 (0–1000 nmol/L). Data shown are mean percentages compared with vehicle/LPS treatment baseline concentration ± SEM of 6 assays. *P < 0.05 vs 0 nmol/L
AR-R17779 baseline concentration. **P < 0.01 vs 0 nmol/L AR-R17779 baseline concentration.
111
TNF-alpha
% decrease
100
Discussion
The cholinergic anti-inflammatory pathway is a now well-established mechanism to control macrophage
activation.8,9 Electrical vagal nerve stimulation has been shown to dampen inflammatory responses
via enhanced efferent vagal output.10,13 However, pharmacologic activation of this pathway might be
a more feasible therapeutic strategy to treat a wide range of inflammatory disorders. Here, we have
shown that presurgical systemic administration of the alpha7 nAChR agonist AR-R17779 is effective
in ameliorating POI through the reduction of manipulation-induced inflammation.
The intestinal muscle inflammation resulting from peri-operative bowel handling now is accepted
widely to play an important role in the pathogenesis of prolonged POI.5,7 This self-limiting disturbance
of normal gastrointestinal propulsion inflicts considerable patient discomfort, morbidity, and is a major
cause of prolonged hospitalization.1 Resident intestinal macrophages located between the circular
and longitudinal muscle layer, in close contact with myenteric cholinergic nerve fibers,10 have been
shown to play an important role in the initiation of the manipulation-induced inflammatory response.7
Inhibitory strategies specifically targeting this macrophage population may have potential in the
treatment of POI.
Electrical stimulation of the left cervical vagus nerve reduces manipulation-induced small intestinal
inflammation and prevents the development of POI, consistent with our previous results.10 The
attenuation of macrophage activation achieved by electrical vagal nerve stimulation is mediated by
alpha7 nAChR–dependent STAT3 signaling in intestinal macrophages.10 Therefore, we hypothesized
that pharmacologic interaction with the alpha7 nAChR may embody a noninvasive and attractive
alternative to electrical vagal nerve stimulation. Our results show that systemic single-dose
administration of nicotine in a tolerable dose (although known to already provoke significant toxic
effects20,26) fails to reduce inflammation or improve postoperative gastric emptying. Increasing the
dose revealed striking adverse events (eg, clonic seizure), excluding further experimentation. This
observation is in line with previous nicotine toxicity studies performed in mice26 and the numerous
side effects observed in clinical studies conducted with nicotine in inflammatory bowel disease.27,28 In
contrast, the alpha7-selective agonist AR-R17779 was well tolerated and was effective in reducing
the manipulation-induced inflammation and normalized the gastric emptying rate. The reason why
nicotine fails to reduce inflammation or POI most likely rests on the 5-fold lower affinity and 35,000-
112
fold lower selectivity for the alpha7 nAChR of this compound compared with AR-R17779.19 The
absence of an inflammation-dampening response after nicotine treatment in vivo also results, in part,
from the dosing frequency. Ghia et al16 only observed an anti-inflammatory effect after 5 consecutive
days of nicotine administration in drinking water in experimental colitis. Furthermore, their data also
suggested a nicotinic anti-inflammatory effect that was independent of vagal nerve integrity in line
with our current results with AR-R17779.
AR-R17779 does not pass the blood-brain barrier easily, as shown by previous pharmacokinetic
studies. At 30 minutes after intravenous administration of 30 mg/kg AR-R17779 (a dose 6 times
median effective concentration value for alpha7 nAChR activation is 27 µmol/L.20 Concurrently, VGX
in our present study did not affect the anti-inflammatory potency of AR-R17779. Although AR-R17779
also activates central nAChRs, these data indirectly suggest a peripheral site of action. However,
Chapter 5
higher than the dose used in the current study) only 7 µmol/L was present in the brain, although the
cerebroventricular organs lack a proper blood-brain barrier and thereby could represent a gateway
to the central nervous system for substances such as AR-R17779. Therefore, a centrally mediated
mechanism, for instance triggering the HPA-axis, cannot be ruled out completely at this time. Microglia,
the central nervous system macrophage, also express alpha7 nAChRs29 and might represent an
alternative central target for AR-R17779. Indeed, stimulation of these receptors results in a reduction
in their LPS-induced TNF release and subsequent neuroinflammation.29 Unexpectedly, VGX did not
lead to a significantly higher inflammatory reaction compared with nonvagotomized animals. Cervical
local inflammation,11,13,30 which has led to the speculation that the cholinergic anti-inflammatory
pathway represents a regulatory mechanism.8,9 On the other hand, Bernik et al31 did not observe a
worsened inflammation after vagotomy in septic shock, nor did Luyer et al32 in hemorrhagic shock. It
may well be that the vagotomy procedure itself elicits the release of acetylcholine, altering the outcome
within a certain time frame. Similar to this study by Luyer et al,32 the interval between vagotomy and
inflammatory insult in our current study was less than 1 hour (45 and 50 min, respectively). In a
different study from our laboratory, subdiaphragmal vagotomy performed more then 1 hour before
intestinal manipulation did elicit a significant increase of intestinal inflammation (unpublished data).
However, the exact explanation remains to be elucidated.
Further analysis of resident peritoneal macrophages showed that exposure to nicotine or ARR17779 activates STAT3 signaling and reduces NF-κB activation. These results are in agreement
113
vagotomy enhances the inflammatory response in models for endotoxemia, pancreatitis, and acute
with previous findings,10,15 and consistent with a model in which the cholinergic anti-inflammatory
pathway is dependent on Jak2/STAT3 signaling downstream from alpha7 nAChR activation through
AR-R17779.10 Hence, a plausible mechanism for the cellular effect of nicotine would be that proteins
in the STAT3 and NF-κB signaling pathways interact to mount an anti-inflammatory response
(ie, as described for p65/c-rel and phosphorylated STAT3).33 This hypothesis is currently under
investigation.
Despite its ability to reduce manipulation-induced inflammatory responses in vivo, AR-R17779 was
less potent in reducing macrophage NF-κB activation and pro-inflammatory mediator release as
compared with nicotine. This implies that the in vivo effects of AR-R17779 in reducing inflammation
may not rest exclusively on the modulation of macrophage activation. Various other cell types involved
in the innate inflammatory response, such as endothelial cells,34 and dendritic cells (unpublished
observation),35 have been shown to express the alpha7 nAChR and may be targeted by AR-R17779.
This is supported further by a recent report that in a rat model of endotoxemia another alpha7 nAChR
agonist, GTS-21, was found to ameliorate endotoxin-induced immune responses by a mechanism
independent of macrophage TNF and MIP2 release.36 In addition, our data indicate that nicotinic
receptors other than alpha7 nAChR may be involved in the effects of macrophage activation
pathways and function in vitro, such as NF-κB activation and cytokine production, consistent with
findings that macrophages express several subtypes of nAChRs.35 In conclusion, we show here that
a single, preoperative dose of the alpha7 agonist AR-R17779 matches the anti-inflammatory potency
of electrical vagal nerve stimulation. AR-R17779 prevents POI in mice and reduces the manipulationinduced intestinal muscle inflammation. This alpha7 selective agonist binds to its receptor on
macrophages, activating the cholinergic anti-inflammatory pathway in a vagal efferent–independent
manner. Our data encourage further clinical exploration of alpha7-selective agonists such as ARR17779 as putative treatment for POI and various other inflammatory disorders involving the innate
immune system.
114
115
1. Prasad M, Matthews JB. Deflating postoperative ileus. Gastroenterology 1999;117:489ñ492.
2. Livingston EH, Passaro EP Jr. Postoperative ileus. Dig Dis Sci 1990;35:121ñ132.
3. Luckey A, Wang L, Jamieson PM, et al. Corticotropin-releasing factor receptor 1-deficient mice
do not develop postoperative gastric ileus. Gastroenterology 2003;125:654ñ659.
4. Barquist E, Bonaz B, Martinez V, et al. Neuronal pathways involved in abdominal surgeryinduced gastric ileus in rats. Am J Physiol 1996;270:R888ñR894.
5. de Jonge WJ, van den Wijngaard RM, The FO, et al. Postoperative ileus is maintained by
2003;125: 1137ñ1147.
6. The FO, de Jonge WJ, Bennink RJ, et al. The ICAM-1 antisense oligonucleotide ISIS-3082
prevents the development of postoperative ileus in mice. Br J Pharmacol 2005;146:252ñ258.
7. Kalff JC, Schraut WH, Simmons RL, et al. Surgical manipulation of the gut elicits an intestinal
muscularis inflammatory response resulting in postsurgical ileus. Ann Surg 1998;228:652ñ663.
8. Ulloa L. The vagus nerve and the nicotinic anti-inflammatory pathway. Nat Rev Drug Discov
2005;4:673ñ684.
9. Tracey KJ. The inflammatory reflex. Nature 2002;420:853ñ859.
10. de Jonge WJ, van der Zanden EP, The FO, et al. Stimulation of the vagus nerve attenuates macrophage activation by activating the Jak2-STAT3 signaling pathway. Nat Immunol
2005;6:844ñ851.
11. van Westerloo DJ, Giebelen IA, Florquin S, et al. The vagus nerve and nicotinic receptors
modulate experimental pancreatitis severity in mice. Gastroenterology 2006;130:1822ñ1830.
12. Bernik TR, Friedman SG, Ochani M, et al. Cholinergic anti-inflammatory pathway inhibition of
tumor necrosis factor during ischemia reperfusion. J Vasc Surg 2002;36:1231ñ1236.
13. Borovikova LV, Ivanova S, Zhang M, et al. Vagus nerve stimulation attenuates the systemic
inflammatory response to endotoxin. Nature 2000;405:458ñ462.
14. Wang H, Yu M, Ochani M, et al. Nicotinic acetylcholine receptor alpha7 subunit is an essential
regulator of inflammation. Nature 2003;421:384ñ388.
15. Wang H, Liao H, Ochani M, et al. Cholinergic agonists inhibit HMGB1 release and improve
survival in experimental sepsis. Nat Med 2004;10:1216ñ1221.
16. Ghia JE, Blennerhassett P, Kumar-Ondiveeran H, et al. The vagus nerve:a tonic inhibitory
influence associated with inflammatory bowel disease in a murine model. Gastroenterology
2006;131:1122ñ1130.
17. Thomas GA, Rhodes J, Mani V, et al. Transdermal nicotine as maintenance therapy for ulcerative colitis. N Engl J Med 1995;332:988ñ992.
18. Levin ED, Rezvani AH. Development of nicotinic drug therapy for cognitive disorders. Eur J
Pharmacol 2000;393:141ñ146.
19. Mullen G, Napier J, Balestra M, et al. (-)-Spiro[1-azabicyclo[2.2.2]octane-3,5í-oxazolidin-2íone], a conformationally restricted analogue of acetylcholine, is a highly selective full agonist at
the alpha 7 nicotinic acetylcholine receptor. J Med Chem 2000;43:4045ñ4050.
20. Grottick AJ, Trube G, Corrigall WA, et al. Evidence that nicotinic alpha(7) receptors are not
involved in the hyperlocomotor and rewarding effects of nicotine. J Pharmacol Exp Ther
2000;294:1112ñ1119.
21. Papke RL, Porter Papke JK, Rose GM. Activity of alpha7-selective agonists at nicotinic
and serotonin 5HT3 receptors expressed in Xenopus oocytes. Bioorg Med Chem Lett
2004;14:1849ñ1853.
22. Van Kampen M, Selbach K, Schneider R, et al. AR-R 17779 improves social recognition in rats
by activation of nicotinic alpha7 receptors. Psychopharmacology (Berl) 2004;172:375ñ383.
23. Takeda K, Clausen BE, Kaisho T, et al. Enhanced Th1 activity and development of chronic
enterocolitis in mice devoid of Stat3 in macrophages and neutrophils. Immunity 1999;10:39ñ49.
Chapter 5
Reference List
24. Bennink RJ, de Jonge WJ, Symonds EL, et al. Validation of gastric-emptying scintigraphy of
solids and liquids in mice using dedicated animal pinhole scintigraphy. J Nucl Med 2003;44:
1099ñ1104.
25. Desmedt M, Rottiers P, Dooms H, et al. Macrophages induce cellular immunity by activating
Th1 cell responses and suppressing Th2 cell responses. J Immunol 1998;160:5300ñ5308.
26. Matta SG, Balfour DJ, Benowitz NL, et al. Guidelines on nicotine dose selection for in vivo
research. Psychopharmacology (Berl) 2006.
27. Pullan RD, Rhodes J, Ganesh S, et al. Transdermal nicotine for active ulcerative colitis. N Engl
J Med 1994;330:811ñ815.
28. Sandborn WJ. Nicotine therapy for ulcerative colitis: a review of rationale, mechanisms, pharmacology, and clinical results. Am J Gastroenterol 1999;94:1161ñ1171.
29. Suzuki T, Hide I, Matsubara A, et al. Microglial alpha7 nicotinic acetylcholine receptors drive a
phospholipase C/IP3 pathway and modulate the cell activation toward a neuroprotective role.J
Neurosci Res 2006;83:1461ñ1470.
30. Borovikova LV, Ivanova S, Nardi D, et al. Role of vagus nerve signaling in CNI-1493-mediated
suppression of acute inflammation. Auton Neurosci 2000;85:141ñ147.
31. Bernik TR, Friedman SG, Ochani M, et al. Pharmacological stimulation of the cholinergic antiinflammatory pathway. J Exp Med 2002;195:781ñ788.
32. Luyer MD, Greve JW, Hadfoune M, et al. Nutritional stimulation of cholecystokinin receptors
inhibits inflammation via the vagus nerve. J Exp Med 2005;202:1023ñ1029.
33. Yu Z, Zhang W, Kone BC. Signal transducers and activators of transcription 3 (STAT3) inhibits
transcription of the inducible nitric oxide synthase gene by interacting with nuclear factor kappaB. Biochem J 2002;367:97ñ105.
34. Saeed RW, Varma S, Peng-Nemeroff T, et al. Cholinergic stimulation blocks endothelial cell
activation and leukocyte recruitment during inflammation. J Exp Med 2005;201:1113ñ1123.
35. Kawashima K, Yoshikawa K, Fujii YX, et al. Expression and function of genes encoding cholinergic components in murine immune cells. Life Sci 2007;80:2314ñ2319.
36. Giebelen IA, van Westerloo DJ, LaRosa GJ, et al. Stimulation of alpha 7 cholinergic receptors
inhibits lipopolysaccharide-induced neutrophil recruitment by a tumor necrosis factor alphaindependent mechanism. Shock 2007;27:443ñ447.
116
117
Chapter 5
6
6
Chapter
6
Central activation of the cholinergic anti-inflammatory pathway shortens postoperative
ileus in mice
submitted for publication
Frans O. The,
Jan van der Vliet,
Wouter J. de Jonge,
Roelof J. Bennink,
Ruud M. Buijs,
Guy E. Boeckxstaens
Abstract
Background & Aims: Electrical stimulation of the vagus nerve reduces the intestinal
inflammation following mechanical handling thereby shortening postoperative ileus in
mice. Previous studies in a sepsis model showed that this cholinergic anti-inflammatory
pathway can be activated pharmacologically by central administration of semapimod,
a p38 MAPKinase inhibitor. Aim: To evaluate the effect of semapimod icv on intestinal
inflammation and postoperative ileus in mice. Methods: Mice underwent a laparotomy (L)
or intestinal manipulation (IM) 1h after pre-treatment with 1µg/kg semapimod or saline icv.
Drugs were administered through a cannula placed in the right lateral ventricle one week
prior to experiments. 24h after surgery, gastric emptying of a semi-liquid meal was measured
using scintigraphy and the degree of intestinal inflammation was assessed. Finally, brain
region activation was assessed using quantitative c-Fos immunohistochemistry. Values
are depicted as mean ± s.e.m. P<0.05 was considered statistically significant. Results:
IM significantly delayed gastric emptying 24h after surgery in saline treated animals
(gastric retention at 80min (RT80) L=3±1%, vs. IM 19±4%, p<0.05, n=8) and induced
inflammation of the manipulated intestine (MPO-pos. cells/mm2: L= 48±7 vs. IM= 381±27,
p<0.05, n=8). Icv semapimod significantly reduced this inflammation and improved gastric
emptying (MPO-pos. cell/mm2: 227±28, p<0.05; RT80: 5±1%, p<0.05, n=8). Vagotomy
(VGX) enhanced IM induced inflammation and abolished the anti-inflammatory effect of
semapimod icv (MPO-pos.cell/mm2: VGX-saline 540±78 vs. saline, n=8, p<0.05 and vs.
VGX-semapimod 440±34, n=8, p=0.2). Semapimod but not saline induced a significant
increase in c-fos expression in the paraventricular nucleus, the nucleus of the solitary tract
and the dorsal motor nucleus of the vagus nerve. Conclusion: Our findings show that
icv semapimod reduces manipulation-induced intestinal inflammation and prevents POI
by central activation of the vagus nerve. In addition, we provide evidence suggesting that
semapimod activates the DMNV, possibly via activation of the paraventricular nucleus.
120
T
Introduction
The vagus nerve plays a crucial role in the control of gastrointestinal function, including
secretion, visceral perception and motility. Recently, Tracey et al. provided strong evidence
indicating that the vagus nerve modulates the innate immune system1. They showed that
electrical stimulation of the vagus reduced TNF levels and prevented arterial hypotension
after endotoxin injection1. Similarly, we demonstrated that vagus nerve stimulation reduced
the inflammatory response to mechanical manipulation of the intestine during surgery and
thereby prevented surgery-induced delayed gastric emptying2. This anti-inflammatory effect
is mediated by acetylcholine interacting with nicotinic receptors located on macrophages,
system controlling the inflammatory response to a wide range of threats to the organism.
Inflammation is sensed by afferent nerve fibers and is subsequently sent to the brain stem
for integration4. After integration, the motor neurons of the vagus nerve are activated and
an integrated anti-inflammatory signal is sent back to the inflamed area4. The presence of
such a feedback loop/reflex and its anatomical connections are still hypothetical and still
need to be demonstrated. Nevertheless, this system may represent an interesting tool to
control inflammation in a number of disorders. In contrast to anti-inflammatory cytokines
and the hormonal control by corticosteroids (HPA axis), this neural system provides an
integrated response that is lightning fast and target specific. Obviously, it may provide new
therapeutic means to control or dampen inflammation, not only in case of sepsis or ileus, but
most likely also in other inflammatory diseases like rheumatoid arthritis and inflammatory
bowel diseases.
Semapimod, a tetravalent guanyl-hydrazone also known as CNI-1493, prevents
macrophage activation via inhibition of mitogen activated protein kinase signaling5. While
studying the effect of semapimod in cerebral ischemia, Meistrell et al. found that central
application of this drug could reduce systemic inflammation6. Further studies revealed
that semapimod, when infused intracerebroventricular (icv), is up to 100.000 times more
effective compared to intravenous administration (iv)7. In addition, electrophysiological
studies have shown enhanced activity of the vagus nerve after infusion of semapimod8.
121
Central Activation of the Cholinergic Anti-inflammatory Pathway in Postoperative Ileus
cholinergic anti-inflammatory pathway is suggested to represent an additional regulatory
Chapter 6
leading to a reduction in macrophage activation and cytokine production3. This so-called
These findings strongly suggest that semapimod represents a pharmacological and central
activator of the cholinergic anti-inflammatory pathway. Animal studies on the pathogenesis of postoperative ileus have shown that gentle small
bowel manipulation during abdominal surgery results in a distinct inflammation of the
muscularis propria9, 10. This local innate inflammatory response activates an adrenergic
inhibitory neural reflex leading to generalized hypomotility or ileus10. Reduction of the
inflammatory response by pre-treatment with intercellular adhesion molecule (ICAM)-1
inhibitory antibodies or antisense oligonucleotides, normalizes gastric emptying10,
11
further illustrating its crucial role in the pathogenesis of postoperative ileus. Previously, we
showed that both electrical stimulation of the vagus nerve2 and systemic administration of
selective nicotinic agonists12 had an anti-inflammatory effect on surgery-induced intestinal
inflammation, suggesting that activation of the cholinergic anti-inflammatory pathway
indeed may represent an interesting approach to treat intestinal inflammation. In the
present study, we evaluated whether pharmacological activation of the vagus nerve by
central application of semapimod also leads to reduced inflammation and prevention of
ileus. In addition, we performed c-fos immunohistochemical analysis of the brain stem to
illustrate the involvement of the motor nucleus of the vagus nerve.
122
Methods
Animals
Female Balb/C mice (Harlan Nederland, Horst, The Netherlands), age 12 to 15 weeks,
were kept under environmentally controlled conditions (light on from 8:00 AM till 8:00 PM;
water and rodent nonpurified diet ad libitum; temperature 20°C-22°C; 55% humidity). All
experiments were performed with the approval of the Ethical Animal Research Committee
of the University of Amsterdam and according to their guidelines.
Study protocols
prior to surgery, as described below. Sixty min before the surgical procedure, animals were
treated with semapimod 1ug/kg icv or its vehicle (saline) in a volume of 5µl administered
in 10min, using an infusion (pump 22 multiple syringe pump, Harvard Apparatus, Holliston,
MA, USA). Twenty-four hrs after surgery, gastric emptying of a semi-liquid non-caloric test
meal was determined using a scintigraphic imaging technique
13
. After completion, mice
were sacrificed by cervical dislocation and ileal segments (4-6 cm proximal of cecum) were
quickly excised for the assessment of intestinal inflammation.
In a different set of experiments, a subdiaphragmal bilateral vagotomy was performed
30min prior to infusion of semapimod or vehicle to determine vagus nerve involvement. To
identify the brain nuclei involved in the central activation of the cholinergic anti-inflammatory
pathway, c-fos expression was studied after icv treatment with semapimod vs. saline. A
swivel equipped infusion pump was used to administer the drugs, allowing the animals to
move freely in their usual environment. Swivel pumps were connected at 8 am in all animals
and infusion was started only after 4 hrs to minimize stress-induced brain activity. Three
hrs after icv administration of saline or semapimod, mice were transcardially perfused (1.6
mL/min) with 8 mL of a 0.9% NaCl solution, followed by 50 mL of 4% paraformaldehyde
in phosphate buffer (0.1 mol/L; pH 7.4). After perfusion, the brain, brainstem and proximal
spinal cord were carefully removed, postfixed overnight in the same fixative at 4°C, and
cryoprotected until further analysis in 30% sucrose solution containing 0.05% sodium azide
at 4°C.
123
postoperative ileus10. An icv cannula was placed in the left lateral ventricle of the brain 7 days
Chapter 6
First, the efficacy of icv administered semapimod was evaluated in our mouse model of
ICV cannula placement
In anesthetized animals, a cannula (23 G needle) was stereotaxically implanted into the left
lateral cerebral ventricle using the following coordinates from Bregma: 0.46mm posterior,
1.0 mm lateral and 2.2 mm ventral. Dental cement was used to secure the cannula to three
screws inserted into the skull.
Surgical procedure
Anesthetized mice underwent a laparotomy (L) or a laparotomy followed by small intestinal
manipulation (IM) as described previously10. In short, a midline incision was made and
the peritoneal cavity was opened along the linea alba under sterile conditions. The small
intestine was carefully exteriorized from the distal duodenum until the cecum and gently
manipulated for 5 minutes using sterile moist cotton applicators. Contact or stretch
of stomach or colon was strictly avoided. After repositioning of the intestinal loops, the
abdomen was closed using a two-layer continuous suture (Mercilene Softsilk 6-0). Mice
recovered from surgery in a temperature controlled cage set at 32° C with free access
to water but not to food. Twenty-four hrs after surgery, gastric emptying was measured.
Thereafter, mice were anaesthetized and killed by cervical dislocation. The small intestine
was removed, flushed in ice-cold phosphate buffered saline (PBS), and snap frozen in
liquid nitrogen or fixed in ethanol for further analysis.
Subdiaphragmal vagotomy
A midline incision was made and a retractor was placed. Under microscopic view, both
the left and right vagal nerve trunks were cut, distal from the diaphragm but proximal
to the division of the hepatic branch. During this procedure, the intraperitoneal organs
were protected and kept moist using sterile gausses drenched in NaCl. Any palpation or
manipulation of the small intestine was carefully avoided. The abdomen was closed using
a two-layer continuous suture (Mercilene Softsilk 6-0). Animals were kept in a temperaturecontrolled cage at 32° C until drug infusion and surgery. Microscopic inspection and
postmortem evaluation of the stomach distention were utilized to determine a successful
vagotomy procedure.
124
Measurement of gastric emptying
As previously described, gastric emptying rate was determined after gavage of a semiliquid, non-caloric test meal (0.1ml of 3% methylcellulose solution containing 10 MBq of
99m
Tc-Albures11, 13. Mice were scanned using a gamma camera set at 140 keV13. The entire
abdominal region was scanned for 30 seconds, immediately and 80 minutes after gavage.
During the scanning period mice were conscious and manually restrained. The static images
obtained were analyzed using Hermes computer software (Hermes, Stockholm, Sweden).
gastric region of interest compared to the total abdominal region of interest11.
Quantification of intestinal muscle inflammation
ethanol for 10 minutes. Fixed preparations were kept in 70% ethanol at 4°C until further
analysis. Before final analysis segments were stretched 1.5 times to their original size and
pinned down on a glass-dish filled with 70% ethanol after which the mucosa was carefully
removed.
Myeloperoxidase (MPO) was stained using the method described in the specified section.
For quantification, the number of MPO-positive cells in five randomly chosen 1mm2 fields
was counted.
Quantification of brain regional C-fos expression4
The number of C-fos positive nuclei were counted in the nucleus of the solitary tract (NTS),
the dorsal motor nucleus of the vagus nerve (DMNV) and the paraventricular nucleus
(PVN) in 4 to 8 section of each individual animal and divided by the number of sections.
These mean numbers of C-fos positive nuclei per animal were used for further statistical
analysis.
125
its mesenteric border. Fecal content was washed out in ice-cold PBS and fixed in 100%
Chapter 6
After sacrifice, the mesentery was removed from the intestine, which was cut open along
Myeloperoxidase staining
Fixed preparations were rehydrated by incubation in 50% ETOH and phosphate buffered
saline pH 7.4 for 5 minutes. To visualize myeloperoxidase(MPO)-positive cells preparations
were incubated for 10 minutes with 3-amino-9-ethyl carbazole (Sigma, St. Louis, MO) as a
substrate, dissolved in sodium acetate buffer (pH 5.0) to which 0.01% H2O2 was added10.
Immunohystochemistry
C-fos immunohistochemistry was performed according to Bonaz et al.14, with modifications.
After fixation, the brain was embedded in Tissue-Tek (Sakura Finetek Inc., Torrance, CA)
and 40mm transversal sections were cryostat-cut. Free-floating sections were washed with
Tris-buffered saline (TBS; pH 7.4) 3 times and incubated overnight at 4°C with the primary
polyclonal sheep antibody (0.3 µg/mL; Sigma Genosys, St. Louis, MO) in 0.25% gelatin
and 0.5% TritonX-100 in TBS. Next, sections were washed in TBS (3x) and incubated
with biotinylated anti-sheep antiserum (Vector Laboratories, Burlingame, CA) for 1.5 hrs at
room temperature. After washing in TBS (3x), sections were processed for avidin– biotin–
peroxidase (Vectorstain; Vector Laboratories), and peroxidase was visualized by using
diaminobenzidine in 0.02% nickel sulphate in TBS as the chromogen.
Statistics
Statistical analysis was performed using SPSS 12.02 software for Windows. The data were
non-parametrically distributed and therefore analyzed using the non-parametric MannWhitney test. P<0.05 was considered statistically significant and results were depicted as
mean ± SEM.
126
Results
Semapimod administered icv ameliorates POI and diminishes
manipulation-induced intestinal muscle inflammation.
Manipulation of the small intestine during abdominal surgery (IM) initiated a significant
increase in gastric retention 24 hrs after the procedure when compared to mice undergoing
laparotomy (L) alone (gastric retention 80 min after gavage of test meal (GR80) IMsaline
19 ± 4% vs. Lsaline 3 ± 1%, n=8, p<0.001) (fig.1). The gastric stasis marking the extent
of postoperative ileus was accompanied by a marked myeloperoxidase(MPO)-positive
inflammatory cell influx in the manipulated segment. This local leukocyte recruitment was
IM-induced delay in gastric emptying (GR80 IMsemapomod 5 ± 1%, n=8, p= 0.02) (fig. 1). In
line with this observation, the number of MPO-positive cells in the intestinal muscle layer
also diminished significantly in the semapimod treated group compared to saline treated
control animals (IMsemapimod 227 ± 28 cells/mm2, n=8, p=0.003 vs. saline) (fig.2). In contrast,
semapimod did not alter gastric retention nor intestinal muscle inflammation compared to
saline in mice that underwent a L (GR80 Lsemapimod 5 ± 2%, n=8, p=0.4; MPO Lsemapimod 36 ±
relative gastric contents (%)
9 cells/mm2, n=8, p=0.3) (fig.1 and 2).
p<0.001
30
p=0.02
20
10
0
saline
semapimod
laparotomy
saline
semapimod
intestinal
manipulation
127
Figure 1 Effect of saline
(vehicle) or semapimod
icv pre-treatment on
gastric retention 24 hrs
after laparotomy (L) or
laparotomy followed by
gentle intestinal manipulation (IM). Gastric retention was determined
80 min. after gavage of
a semi-liquid test meal.
Data are expressed as
mean ± SEM (MannWhitney U test).
± 7 cells/mm2, n=8, p=0.001) (fig.2). Treatment with semapimod 1µg/kg icv ameliorated the
Chapter 6
not observed in L animals (MPO-positive cells/mm2 IMsaline 381 ± 27 cells/mm2 vs. Lsaline 48
The anti-inflammatory effect of semapimod is mediated through
the vagus nerve.
To assess the involvement of the vagus nerve, experiments were repeated in subdiaphragmal
vagotomized (VGX) animals. As gastric motility is strongly effected by VGX, gastric
emptying was not assessed. There was a significant inflammatory response 24hrs after
IM not seen 24hrs after L in mice subjected to VGX prior to abdominal surgey (IMvgx 540
± 78 cells/mm2 vs. Lvgx 52 ± 9 cells/mm2, n=8, p=0.004). This inflammatory response was
even severe when compared to the response observed in non-VGX animals undergoing
IM (IMVGX vs. IM, n=8, p=0.04) . In contrast, the anti-inflammatory effect in semapimod pretreated animals was absent after subdiaphragmal VGX (semapimod IMvgx 440 ± 34.3cells/
mm2 vs. saline IMvgx 404 ± 34, n=8, p=0.4) (fig. 2) illustrating that the anti-inflammatory
effect achieved with semapimod is mediated through vagus nerve signaling.
Semapimod induced c-fos expression in brain stem nuclei.
To obtain more insight in the central mechanism through which semapimod activates the
cholinergic anti-inflammatory pathway, c-fos immunohistochemical analysis of the brain
was performed. Based on a previous report demonstrating enhanced activity of vagal
efferent nerve fibers upon semapimod administration8 we first assessed c-fos expression
in the dorsal motor nucleus of the vagus nerve (DMNV) and the nucleus of the solitary tract
(NTS). Quantitative c-fos analysis 3 hrs after infusion of semapimod showed a significant
increased number of c-fos positive neurons in both the NTS (saline 23 ± 3 vs. semapimod
38 ± 5, n=6, p=0.02) and the DMNV (saline 4 ± 0 vs. 8 ± 0, n=6, p=0.002) when compared
to saline (fig. 3b-c).
Vagal efferent control of pancreatic protein secretion is mediated through M1-receptor
stimulation in the paraventricular nucleus (PVN).15 Moreover, Pavlov et al. recently found
that semapimod competitively binds to M1-muscarinic receptors.16 These observations led
to our hypothesis that semapimod triggers neurons in the PVN projecting to the dorsal
motor complex of the vagus nerve (DVC) hereby activating the cholinergic anti-inflammatory
pathway. Therefore we assessed c-fos activation in the PVN 3hrs after icv injection of
semapimod and found that c-fos expression was significantly increased in the PVN after
semapimod compared to saline infusion (19 ± 5 vs. 70 ± 23, n=6, p=0.03)(fig. 3a).
128
p=0.004
700
p=0.04
600
p=0.001
500
400
300
200
0
saline
VGX semapimod VGX
saline
saline
semapimod
laparotomy
VGX semapimod VGX
saline
semapimod
in t e s t i n a l
manipulation
Figure 2 Effect of saline (vehicle) or semapimod icv pre-treatment on manipulation-induced inflammatory cell recruitment to the muscularis propria. Quantative analysis of the number of MPO-positive
cells 24 hrs after laparotomy (L) or laparotomy followed by gentle intestinal manipulation (IM). Mice
that underwent a subdiaphragmal vagotomy prior to treatment are marked by VGX. Data are expressed as mean ± SEM (Mann-Whitney U test).
129
Chapter 6
100
MPO-pos. cells/mm
2
p=0.003
A
PVN
p=0.03
number of c-FOS
pos. neurons
100
80
60
40
20
0
B
saline PVN
semapimod PVN
DMNV
p=0.002
10
number of c-FOS
pos. neurons
8
6
4
2
0
semapimod DMNV
NTS
100
p=0.04
number of c-FOS
pos. neurons
C
saline DMNV
80
60
Figure 3 C-fos expression in in the
A) PVN, B) DMNV and C) NTS 3
hrs after icv saline or semapimod
treatment. Data are expressed as
mean ± SEM (Mann-Whitney U
test).
40
20
0
saline NTS
semapimod NTS
130
Discussion
Manipulation-induced inflammation of the intestine, a process orchestrated by innate
immune cells like mast cells and macrophages10, 17-23, is now generally believed to play
an imperative role in the pathophysiology of prolonged postoperative ileus. Recently, we
showed that vagus nerve stimulation reduces this inflammation and thereby shortens
postoperative ileus. Semapimod, a p38 MAPKinase inhibitor, has been shown to be a
central pharmacological activator of this so-called cholinergic anti-inflammatory pathway7, 8.
In the present study we showed that 1. icv semapimod reduces the inflammatory response
to intestinal manipulation and restores gastric emptying, 2. the anti-inflammatory effect
semapimod activates the cholinergic anti-inflammatory pathway leading to a reduction of
the manipulation-induced intestinal inflammation and restoration of gastric emptying.
Macrophages, present as a network between the circular and longitudinal muscle layers of
the intestine19, have been shown to play an important role in the pathogenesis of sustained
postoperative ileus in rodents20, 21. These phagocytes lay in close proximity of the myenteric
plexus and carry nicotinic acetylcholine receptors enabling neuro-immune interaction2.
Recently we demonstrated that the manipulation-induced inflammation can be diminished by
electrical stimulation of the vagus nerve in our experimental mouse model for postoperative
ileus2 A similar anti-inflammatory effect of vagus nerve stimulation has been demonstrated
in sepsis, ischemia-reperfusion, IBD and pancreatitis1, 24-26, and is currently referred to as
the cholinergic anti-inflammatory pathway. Stimulation of the efferent vagus nerve results
in reduction of pro-inflammatory cytokine release, i.e. TNF, IL1β and IL6, hereby improving
outcome in experimental septic-shock1. The peripheral mechanism of this cholinergic
anti-inflammatory response is mediated through alpha-7 nicotinergic receptors expressed
on macrophages3. Activation of these receptors results in JAK2/STAT3 signaling2 and
inhibition of NF-κB signal transduction27. Indeed, systemic application of selective alpha-7
nicotinergic agonists mimics the effect of vagus nerve stimulation in experimental models
for inflammatory bowel disease, pancreatitis and postoperative ileus.12, 24, 25
131
expression in the PVN and DMNV. These findings confirm that central administration of
Chapter 6
of semapimod is abolished by vagotomy, 3. icv semapimod leads to increased c-fos
Semapimod, a tetravalent guanylhydrazone also known as CNI-1493, has been suggested
a central, pharmacological activator of the cholinergic anti-inflammatory pathway.7,
8
Here we found that semapimod indeed suppresses inflammation in our mouse model
for postoperative ileus. Application of 1ug/kg semapimod administered icv diminished
manipulation induced inflammation and normalized gastric emptying. This effect was
abolished by subdiaphragmal vagotomy indicating central activation of the vagus nerve.
Pavlov et al. have demonstrated high affinity of semapimod for M1 muscarinic receptors16.
Moreover, they showed activation of the cholinergic anti-inflammatory pathway by central
administration of muscarinic agonists such as mucarine and the M1 selective agonist MCNA-343 with reduction of TNF release in an endotoxemia model16. However, the exact brain
areas involved remain to be clarified. In line with our functional data and the transient
increase of activity on vagal efferent recordings upon icv semapimod8, we observed an
increase in the number of c-fos positive neurons in the DMNV after semapimod but not
after saline. Accepting that semapimod interacts with central M1 receptors (Pavlov et al.)16,
direct activation of the DMNV is rather unlikely as this brain nucleus lacks M1 receptors28.
Previously, M1-receptor mediated activation of vagal output controlling pancreatic protein
secretion was shown to be mediated through stimulation in the paraventricular nucleus
(PVN).15 In line with this, we demonstrated c-fos activation in the PVN after icv administration
of semapimod. This finding and the knowledge that the PVN is interconnected with the
DMNV29, suggest that activation of the cholinergic anti-inflammatory pathway by semapimod
is indirect via interaction with M1 receptors in the PVN. Further studies however are required
to further confirm this hypothesis. In addition to increased c-fos expression in the DMNV,
we also observed enhanced c-Fos expression in the NTS following i.c.v. administration
of semapimod. This is in line with electrophysiological findings by Zhang et al.30 These
authors indeed demonstrated that although NTS neurons are predominantly inhibited, a
minority of NTS neurons was activated by electrical stimulation of PVN neurons.
Interestingly, we also demonstrated an increase in the inflammatory response to intestinal
manipulation in vagotomized animals. Similar findings have been reported in other models
of inflammation8 and of sepsis1. For example, vagotomy increased the mortality rate in
animals subjected to hemorrhagic shock, associated with an increase in TNF levels31.
Similarly, the degree of DSS colitis25 and pancreatitis24 was significantly augmented after
132
vagotomy. Together with our findings, these data would suggest endogenous activation
of the cholinergic anti-inflammatory pathway by the ongoing peripheral inflammatory
response, and would fit with the hypothesis that the vagus nerve exerts an important role
in modulating the innate immune system.
In conclusion, we showed that icv administration of semapimod reduces intestinal
inflammation and postoperative ileus induced by abdominal surgery via activation of the
cholinergic anti-inflammatory pathway. In addition, we provided indirect evidence that
semapimod activates the DMNV via activation of the PVN. These findings further demonstrate
Chapter 6
the anti-inflammatory properties of the cholinergic anti-inflammatory pathway.
133
Reference List
2. de Jonge WJ, van der Zanden EP, The FO, Bijlsma MF, van Westerloo DJ, Bennink RJ,
Berthoud HR, Uematsu S, Akira S, van den Wijngaard RM, Boeckxstaens GE. Stimulation of
the vagus nerve attenuates macrophage activation by activating the Jak2-STAT3 signaling
pathway. Nat.Immunol. 2005;6:844-851.
4. Tracey KJ. Physiology and immunology of the cholinergic antiinflammatory pathway. J Clin
Invest 2007;117:289-96.
5. Lowenberg M, Verhaar A, van den BB, ten Kate F, van Deventer S, Peppelenbosch M, Hommes D. Specific inhibition of c-Raf activity by semapimod induces clinical remission in severe
Crohn’s disease. J.Immunol. 2005;175:2293-2300.
6. Meistrell ME, III, Botchkina GI, Wang H, Di Santo E, Cockroft KM, Bloom O, Vishnubhakat JM,
Ghezzi P, Tracey KJ. Tumor necrosis factor is a brain damaging cytokine in cerebral ischemia.
Shock 1997;8:341-348.
7. Bernik TR, Friedman SG, Ochani M, DiRaimo R, Ulloa L, Yang H, Sudan S, Czura CJ, Ivanova
SM, Tracey KJ. Pharmacological stimulation of the cholinergic antiinflammatory pathway.
J.Exp.Med. 2002;195:781-788.
8. Borovikova LV, Ivanova S, Nardi D, Zhang M, Yang H, Ombrellino M, Tracey KJ. Role of vagus
nerve signaling in CNI-1493-mediated suppression of acute inflammation. Auton.Neurosci.
2000;85:141-147.
10. de Jonge WJ, van den Wijngaard RM, The FO, ter Beek ML, Bennink RJ, Tytgat GN, Buijs
2003;125:1137-1147.
11. The FO, de Jonge WJ, Bennink RJ, van den Wijngaard RM, Boeckxstaens GE. The ICAM-1
antisense oligonucleotide ISIS-3082 prevents the development of postoperative ileus in mice.
Br.J.Pharmacol. 2005.
12. The FO, Boeckxstaens GE, Snoek SA, Cash JL, Bennink R, Larosa GJ, van den Wijngaard
RM, Greaves DR, de Jonge WJ. Activation of the cholinergic anti-inflammatory pathway ameliorates postoperative ileus in mice. Gastroenterology 2007;133:1219-28.
13. Bennink RJ, de Jonge WJ, Symonds EL, van den Wijngaard RM, Spijkerboer AL, Benninga
using dedicated animal pinhole scintigraphy. J.Nucl.Med. 2003;44:1099-1104.
J.Comp Neurol. 1994;349:212-222.
15. Li Y, Wu X, Zhu J, Yan J, Owyang C. Hypothalamic regulation of pancreatic secretion is mediated by central cholinergic pathways in the rat. J.Physiol 2003;552:571-587.
16. Pavlov VA, Ochani M, Gallowitsch-Puerta M, Ochani K, Huston JM, Czura CJ, Al Abed Y,
Tracey KJ. Central muscarinic cholinergic regulation of the systemic inflammatory response
during endotoxemia. Proc.Natl.Acad.Sci.U.S.A 2006;103:5219-5223.
17. Wehner S, Schwarz NT, Hundsdoerfer R, Hierholzer C, Tweardy DJ, Billiar TR, Bauer AJ, Kalff
JC. Induction of IL-6 within the rodent intestinal muscularis after intestinal surgical stress. Surgery 2005;137:436-446.
134
135
Chapter 6
18. Kalff JC, Schraut WH, Billiar TR, Simmons RL, Bauer AJ. Role of inducible nitric oxide synthase in postoperative intestinal smooth muscle dysfunction in rodents. Gastroenterology
2000;118:316-327.
19. Mikkelsen HB. Macrophages in the external muscle layers of mammalian intestines. Histol.
Histopathol. 1995;10:719-736.
20. Wehner S, Behrendt FF, Lyutenski BN, Lysson M, Bauer AJ, Hirner A, Kalff JC. Inhibition of
macrophage function prevents intestinal inflammation and postoperative ileus in rodents. Gut
2006.
23. de Jonge WJ, The FO, van der CD, Bennink RJ, Reitsma PH, van Deventer SJ, van den
Wijngaard RM, Boeckxstaens GE. Mast cell degranulation during abdominal surgery initiates
postoperative ileus in mice. Gastroenterology 2004;127:535-545.
24. van Westerloo DJ, Giebelen IA, Florquin S, Bruno MJ, Larosa GJ, Ulloa L, Tracey KJ, van der
PT. The vagus nerve and nicotinic receptors modulate experimental pancreatitis severity in
mice. Gastroenterology 2006;130:1822-1830.
25. Ghia JE, Blennerhassett P, Kumar-Ondiveeran H, Verdu EF, Collins SM. The vagus nerve: a
tonic inhibitory influence associated with inflammatory bowel disease in a murine model. Gastroenterology 2006;131:1122-30.
26. Bernik TR, Friedman SG, Ochani M, DiRaimo R, Susarla S, Czura CJ, Tracey KJ. Cholinergic
antiinflammatory pathway inhibition of tumor necrosis factor during ischemia reperfusion.
J.Vasc.Surg. 2002;36:1231-1236.
27. Wang H, Liao H, Ochani M, Justiniani M, Lin X, Yang L, Al Abed Y, Wang H, Metz C, Miller EJ,
Tracey KJ, Ulloa L. Cholinergic agonists inhibit HMGB1 release and improve survival in experimental sepsis. Nat.Med. 2004;10:1216-1221.
28. Hoover DB, Hancock JC, DePorter TE. Effect of vagotomy on cholinergic parameters in nuclei
of rat medulla oblongata. Brain Res Bull 1985;15:5-11.
29. Rogers RC, Kita H, Butcher LL, Novin D. Afferent projections to the dorsal motor nucleus of the
vagus. Brain Res Bull 1980;5:365-73.
30. Zhang X, Fogel R, Renehan WE. Stimulation of the paraventricular nucleus modulates the
activity of gut-sensitive neurons in the vagal complex. Am.J.Physiol 1999;277:G79-G90.
31. Guarini S, Cainazzo MM, Giuliani D, Mioni C, Altavilla D, Marini H, Bigiani A, Ghiaroni V, Passaniti M, Leone S, Bazzani C, Caputi AP, Squadrito F, Bertolini A. Adrenocorticotropin reverses
hemorrhagic shock in anesthetized rats through the rapid activation of a vagal anti-inflammatory pathway. Cardiovasc Res 2004;63:357-65.
7
7
Chapter
7
Mast Cell Degranulation During
Abdominal
Surgery
Initiates Postoperative Ileus in
Mice
Wouter J. De Jonge,
Frans O. The,
Dennis van der Coelen,
Roelof J. Bennink,
Pieter H. Reitsma,
Sander J. van Deventer,
René M. van den Wijngaard
Guy E. Boeckxstaens
Abstract
Background & Aims: Inflammation of the intestinal muscularis following manipulation
during surgery plays a crucial role in the pathogenesis of postoperative ileus. Here, we
evaluate the role of mast cell activation in the recruitment of infiltrates in a murine model.
Methods: Twenty-four hours after control laparotomy or intestinal manipulation, gastric
emptying was determined. Mast cell degranulation was determined by measurement of mast
cell protease-I in peritoneal fluid. Intestinal inflammation was assessed by determination of
tissue myeloperoxidase activity and histochemical staining. Results: Intestinal manipulation
elicited a significant increase in mast cell protease-I levels in peritoneal fluid and resulted
in recruitment of inflammatory infiltrates to the intestinal muscularis. This infiltrate was
associated with a delay in gastric emptying 24 hours after surgery. Pretreatment with
mast cell stabilizers ketotifen (1 mg/kg, PO) or doxantrazole (5 mg/kg, IP) prevented both
manipulation-induced inflammation and gastroparesis. Reciprocally, in vivo exposure of
an ileal loop to the mast cell secretagogue compound 48/80 (0.2 mg/mL for 1 minute)
induced muscular inflammation and delayed gastric emptying. The manipulation-induced
inflammation was dependent on the presence of mast cells because intestinal manipulation in
mast cell-deficient Kit/Kitv mice did not elicit significant leukocyte recruitment. Reconstitution
of Kit/Kitv mice with cultured bone marrow-derived mast cells from congenic wild types
restored the manipulation-induced inflammation. Conclusions: Our results show that
degranulation of connective tissue mast cells is a key event for the establishment of the
intestinal infiltrate that mediates postoperative ileus following abdominal surgery.
138
P
Background
Postoperative ileus (POI) is characterized by dysmotility of the gastrointestinal tract that
occurs after essentially every abdominal procedure.1,2 Recent evidence indicates that
postoperative ileus following bowel manipulation is a biphasic process. An acute phase of
generalized enteric hypomotility is due to activation of inhibitory neural reflexes,3,4 which
is dependent on the release of α-calcitonin gene-related peptide (CGRP)5–7 and central
corticotropin-releasing factor.8 A subsequent prolonged phase is mediated by inflammation
of the intestinal muscularis externa that is induced by mechanical manipulation of the
gut.9–12 The muscular inflammation following bowel manipulation results in postoperative
motility changes of the manipulated small intestinal segment, i.e., impaired contractility and
delayed transit.10–12 However, the duration of POI is not determined by hampered peristalsis
of the small intestine only but rather by hypomotility of the entire gastrointestinal tract. In
only impair the neuromuscular function of the manipulated small intestine but also lead to
impaired gastric emptying.9 This gastroparesis resulted from the activation of an inhibitory
adrenergic neural pathway triggered by the intestinal infiltrates, explaining the generalized
nature of POI. Inflammatory infiltrates recruited to the bowel wall after manipulation are thus
Chapter 7
this light, we recently showed that the inflammatory infiltrates in the small intestine not
crucial in the pathogenesis of POI. The mechanism as to how mechanical manipulation of
the intestine induces inflammation has been shown to involve the activation of a resident
activation of these macrophages, however, remains unknown.
In this respect, intestinal manipulation has previously been described to initiate extensive
mast cell activation and degranulation.13 The murine intestine contains mast cells of the
mucosal (MMC) and connective tissue subtype (CTMC) that have distinct expression
of proteases.14 Intestinal mast cells contain numerous substances released upon
degranulation that are potent proinflammatory mediators such as tumor necrosis factor
(TNF) α,15 macrophage inflammatory protein-2 (MIP-2),15 and interleukin (IL)-8.16 Mast cells
have been shown to mediate granulocyte infiltration in a number of inflammatory conditions
involving delayed-type hypersensitivity reactions15 or in other diseases, such as bullous
pemphigoid,17 intestinal ischemia,18 and asthma.19 We therefore hypothesized that mast
cell degranulation may also initiate manipulation-induced intestinal inflammation in POI.
139
Mast cell Degranulation Initiates in Postoperative Ileus
macrophage network in the intestinal muscularis.10–12 The exact nature of the trigger behind
In the current study, we provide evidence that mast cells degranulate upon intestinal
manipulation and that mast cell degranulation initiates the muscularic inflammation that
mediates POI. By employing 2 mast cell stabilizers, doxantrazole, as a nonselective
stabilizer that acts on MMC as well as CTMC, and ketotifen, which stabilizes only CTMC,20
we demonstrate that CTMC are primarily involved in this process. Hence, these findings set
the stage for the clinical use of mast cell stabilizers to shorten POI.
140
Laboratory Animals
Mice (female BalB/C, Harlan Nederland, Horst, The Netherlands) were kept under
environmentally controlled conditions (light on from 8:00 AM to 8:00 PM; water and rodent
nonpurified diet ad libitum; 20°C–22°C, 55% humidity). Mast cell-deficient WBB6F1-W/WV
(Kit/Kitv) and the congenic mast cell-sufficient W/+ (Kit/WT) littermates were purchased
from The Jackson Laboratory (Bar Harbor, ME). Mice were maintained at the animal facility
of the Academic Medical Center in Amsterdam and were used at 15–20 weeks of age.
Animal experiments were performed in accordance with the guidelines of the Ethical Animal
Research Committee of the University of Amsterdam.
Surgical Procedures: Abdominal Surgery
mixture of fentanyl citrate/fluanisone (Hypnorm; Janssen, Beerse, Belgium) and midazolam
(Dormicum; Roche, Mijdrecht, The Netherlands). Surgery was performed under sterile
conditions. Mice (8–12 per treatment group) underwent control surgery of only laparotomy
Chapter 7
With Intestinal Manipulation Mice were anesthetized by an intraperitoneal (IP) injection of a
or laparotomy followed by intestinal manipulation. The surgery was performed as follows:
A midline abdominal incision was made, and the peritoneum was opened over the linea
alba. The small bowel was carefully externalized, layered on a sterile moist gauze pad,
cotton applicators. Contact or stretch on stomach or colon was strictly avoided. After the
surgical procedure, the abdomen was closed by a continuous 2-layer suture (Mersilene, 6-0
silk). After closure, mice were allowed to recover for 4 hours in a heated (32°C) recovery
cage with free access to drinking water but not food. At 4 hours postoperatively, mice were
completely recovered from anesthesia. At 24 hours after surgery, the gastric emptying rate
was measured using gastric scintigraphy. Thereafter, mice were anesthetized and killed by
cervical dislocation, and the small intestine was removed, flushed in ice-cold saline, and
snap frozen in liquid nitrogen or fixed in ice-cold ethanol for further analysis.
141
and manipulated from the distal duodenum to the cecum for 5 minutes, using sterile moist
Study Protocols
The effect of doxantrazole or ketotifen treatment on postoperative intestinal inflammation
and gastric emptying. One group of mice received the mast cell stabilizer doxantrazole (5
mg/kg in 5% NaHCO3 , pH 7.4; a kind gift of Agne`s Francois, Institut Gustave Roussy,
Villejuif, France) or its vehicle, via IP injection once daily for 3 days.21 Alternatively,
ketotifen (1 mg per kg; Sigma Chemical Co., St. Louis, MO) in 0.5% methylcellulose
solution in water was administered by oral gavage once daily for 5 consecutive days.22
Ketotifen controls received only 0.5% methylcellulose in water. Mice underwent abdominal
surgery with intestinal manipulation 1 hour after the final treatment. Twenty-four hours after
surgery, gastric emptying was measured. Thereafter, the mice were killed, and the intestine
was isolated. Intestinal tissue was cut open along the mesenterial border, washed in icecold saline, blotted dry, and frozen in liquid nitrogen for determination of MPO activity.
Alternatively, tissue was fixed in ice-cold 4% paraformaldehyde or ethanol and processed
for histologic analyses.
The effect of in vivo mast cell degranulation on intestinal inflammation and gastric emptying.
Local mast cell degranulation in the ileum was evoked as follows: A midline laparotomy
was performed, and 6 cm of ileum proximal to the cecum was carefully externalized and
placed in a sterile aluminum cup without touching or stretching other parts of the GI tract.
Of these 6 cm of ileum, the proximal 5-cm segment was incubated in either an 0.2 mg/mL
solution of compound 48/80 (C48/80, Sigma Co.) or vehicle (saline) for 1 minute at 37°C.
Leakage of C48/80 solution into the peritoneal cavity was strictly avoided. Segments were
incubated for a short period of only 1 minute to avoid potential systemic uptake of C48/80.
After incubation, the C48/80 solution was removed, and the ileal segment was washed 3
times with 0.9% NaCl, kept prewarmed at 37°C. After closure, the animals were allowed
to recover for 4 hours in a heated recovery cage. No mortality was observed following this
treatment.
The effect of intestinal manipulation in Kit/Kitv mice and mast cell reconstituted Kit/Kitv
mice. Kit/Kitv mice are completely devoid of mast cells in their gastrointestinal tract or
other anatomical sites.23 Kit/Kitv mice were reconstituted by the injection of bone marrowderived cultured mast cells into the peritoneal cavity, as described.24 In brief, femoral bone
marrow cells from Kit/WT control mice were maintained in vitro for 4 weeks in RPMI 1640
142
complete medium (Life Technologies Inc., Grand Island, NY) supplemented with 10% fetal
calf serum, in the presence of stem cell factor (50 ng/mL; Pepro Tech, Rocky Hill, NJ) and
interleukin-3 (1 ng/mL; Pepro Tech, Rocky Hill, NJ). During culture, medium was refreshed
once weekly. After this culture period, mast cells represented more than 95% of the total
cells as determined by Toluidine blue staining on cytospin preparations. Subsequently, mast
cells were harvested, and 2 x 106 cells in 100 µL PBS were injected in Kit/Kitv mice. PBS
alone (100 µL IP) was injected as a negative control. This procedure reconstitutes the mast
cell population without systemic effects.
24
To confirm mast cell reconstitution, we stained
peritoneal cells obtained by lavage as well as intestinal, mesenteric, and gastric tissue
sections with Toluidine blue and Giemsa. Mice were used 10 weeks after adoptive transfer
of mast cells. Twenty-four hours after intestinal manipulation, intestinal and gastric tissue
was obtained and analyzed by Toluidine blue staining. Tissue was snap frozen in liquid
nitrogen and stored at -80°C for determination of MPO enzyme activity. Peritoneal cells
400g for 5 minutes at 4°C. MMCP-1 levels were measured in blood plasma and peritoneal
lavage fluid by sandwich ELISA according to the manufacturer’s instructions (Moredun
Scientific, Edinburgh, Scotland).
Chapter 7
were harvested by lavage with 5 mL 0.9% NaCl, and the cell suspension was centrifuged at
Measurement of Gastric Emptying and Transit
Gastric emptying was determined as described previously.25 Gastric emptying rate
was determined after offering a caloric solid test meal (100 mg egg yolk containing 10
99m
Tc-Albures) that as consumed within 1 minute.26 Immediately after
complete consumption, mice were scanned using a gamma camera set at 140 keV. The
entire abdominal region was scanned for 30 seconds at 16-minute intervals for 112 minutes.
During the 30-second scanning period, mice were conscious and manually restrained.
The static images obtained were analyzed using Hermes computer software (Hermes,
Stockholm, Sweden). Gastric emptying was measured by determining the percentage of
activity present in the gastric region of interest, compared with the total abdominal region of
interest, for each image. Subsequently, the gastric half-emptying time (t1/2) was determined
for each individual mouse using DataFit software (version 6.1, Oakdale Engineering,
Oakdale, PA). A modified power exponential function y(t) = 1 - (1 - ekt)b was used, where
y(t) is the fractional meal retention at time t, k is the gastric emptying rate per minute, and
b is the extrapolated y-intercept from the terminal portion of the curve.
143
MBequerel(MBq) of
Quantification of Leukocyte Accumulation at the Intestinal
Muscularis
Myeloperoxidase (MPO) activity in full-thickness ileal segments was assayed as a measure
of leukocyte infiltration as described elsewhere.9 Tissue was blotted dry, weighed, and
homogenized in a 20 times volume of a 20 mmol/L potassium phosphate buffer, pH
7.4. The suspension was centrifuged (8000g for 20 minutes at 4°C), and the pellet was
taken up in 1 mL of a 50 mmol/L potassium phosphate buffer, pH 6.0, containing 0.5% of
hexadecyltrimethylammoniumbromide (HETAB) and 10 mmol/L ethylenediaminetetraacetic
acid (EDTA). Fifty microliters of the appropriate dilutions of the tissue homogenate was
added to 445 µL of assay mixture, containing 0.2 mg/mL tetramethylbenzidine in 50 mg
potassium phosphate buffer, pH 6.0, 0.5% HETAB, and 10 mmol/L EDTA. The reaction
was started by adding 5 µL of a 30 mmol/L H2O2 to the assay mixture, and the mixture was
incubated for 3 minutes at 37°C. After 3 minutes, 30 µL of a 300 µg/mL catalase solution
was added to each tube, and tubes were placed on ice for 3 minutes. The reaction was
ended by adding 2 mL of 0.2 mol/L glacial acetic acid. Absorbance was read at 655 nm.
A standard reference curve was established using purified MPO (Sigma Co.). One unit of
MPO activity was defined as the quantity of MPO activity required to convert 1 µmol of H2O2
to H2O per minute at 25°C, using purified MPO activity as a standard (Sigma, St Louis,
MO). MPO content was expressed as units of MPO activity per milligram of tissue.
Whole Mount Preparation
Whole mounts of ileal muscularis were prepared as previously described.9,11 In short,
ileal segments (2–6 cm distal from the cecum) were quickly excised, and mesentery was
removed. Ileal segments were cut open along the mesentery border, fecal content was
washed out in ice-cold PBS, and segments were pinned flat in a glass-dish filled with
preoxygenated Krebs-Ringer solution, pH 7.4. Mucosa was removed, and the remaining
full-thickness sheet of muscularis externa was fixed for 10 minutes in 100% ethanol.
Muscularis preparations were kept on 70% ethanol at 4°C until analysis.
144
Statistical Analysis
The data are expressed as mean ± SEM and were analyzed using the nonparametric
Mann–Whitney U test or 1-way ANOVA where indicated. A P value less than 0.05 was
Chapter 7
considered significant.
145
Results
Intestinal Manipulation Triggers Intestinal Mast Cell Degranulation
We determined whether the gentle intestinal manipulation induces mast cell degranulation
in our experimental model of POI. To this end, the level of the mast cell-specific soluble
chymase murine mast cell proteinase-1 (mMCP-1) in the serum and peritoneal lavage fluid
was measured 20 minutes after intestinal manipulation. Intestinal manipulation led to a
significant increase in peritoneal mMCP-1, compared with a control laparotomy (Table 1).
Serum mMCP-1 levels were not significantly altered (not shown). Pretreatment with mast
cell stabilizing agents ketotifen or doxantrazole effectively prevented mast cell degranulation
during intestinal manipulation in that the increase in peritoneal mMCP-1 levels after
intestinal manipulation was not observed following intestinal manipulation with ketotifen or
doxantrazole pretreatment (Table 1).
Table 1. mMCP-1 levels in peritoneal lavage fluid (n=4-5),
3 h PO det limit 1.25 ng/mL
surgery
L
IM
L
L
IM
IM
treatment
sham
VS 5V
sham
VS 5V
mMCP1 (ng/mL)
<1.3
4.1±0.5
<1.3
<1.3
3.7±0.7
3.5±0.9
Mast Cell Stabilization Prevents Intestinal Inflammation and the
Development of POI Following Abdominal Surgery
We aimed to investigate whether pretreatment with a mast cell stabilizer prevented
postoperative gastroparesis induced by intestinal manipulation. We measured gastric
emptying 24 hours after surgery in mice pretreated with either doxantrazole (5 mg/kg,
IP for 3 days once daily) or ketotifen (1 mg/kg, PO for 5 days once daily). As shown in
Figure 1, neither of the vehicles used for ketotifen or doxantrazole administration affected
basal gastric emptying (Figure 1A and B). However, intestinal manipulation significantly
increased halfemptying time and gastric retention of the test meal compared with control
mice treated with vehicle saline (Figure 1A) or 5% NaHCO3 (Figure 1B).
146
In contrast, pretreatment of mice with either ketotifen (Figure 1A) or doxantrazole (Figure 1B)
prevented manipulation-induced gastroparesis and resulted in a significant decreased halfemptying time (t1/2) back to laparotomy control values. In conjunction, ketotifen pretreatment
prevented the increase in gastric retention (Figure 1A and B). Although doxantrazole
pretreatment led to a normalized t1/2 , its effect was less effective compared with ketotifen
because gastric retention at 60 minutes was still significantly increased (Figure 1B).
60
60
IM + v
etotife
20
n
L+ ve
hicle
60
time (min)
doxantrazole
IM +
20
100
ehicle
doxa
ntraz
ole
L+ ve
hicle
20
60
time (min)
surgery
t
(min)
Ret60min (%)
+
+
L
L
IM
IM
30.5 ± 4.2
31.6 ± 2.4
47.5 ± 6.9 *
30.3 ± 8.3
25.3 ± 4.7
26.5 ± 2.6
41.1 ± 5.2 *
27.4 ± 7.1
+
+
L
L
IM
IM
28.8 ± 4.9
33.5 ± 10.7
59.1 ± 7.5 *
43.3 ± 7.5
24.6 ± 5.6
33.3 ± 4.0
43.1 ± 5.2*
39.0 ± 7.1*
pretreatment
ketotifen
IM + v
ehicle
IM + k
20
B
100
100
Figure 1. Delayed gastric emptying after intestinal manipulation is prevented by pretreatment with
mast cell stabilizers ketotifen or doxantrazole. (A) Gastric emptying curves determined by scintigraphic imaging of the abdomen after oral administration of solid caloric meal at 24 hours after
intestinal manipulation (IM + vehicle; solid squares), or IM after pretreatment with ketotifen (open
squares), compared with laparotomy (L + vehicle; gray circles). Values shown are percentages of
gastric content compared with the total abdominal region. (B) Treatment with doxantrazole prevents
gastroparesis 24 hours following bowel surgery (IM + doxantrazole), compared with vehicle treatment
(IM + vehicle). Lower panel: corresponding deduced half-emptying time (t1/2) as well as the retention
after 60 minutes. Ret60min is significantly increased after IM, irrespective of vehicle used, compared
with L. Pretreatment with either ketotifen or doxantrazole restores both t1/2 to normal. Note that
Ret60min is restored to normal only following ketotifen treatment. Treatment with ketotifen or doxantrazole did not alter basal emptying after L. Values are averages ± SEM of 5–10 mice per treatment
group. Asterisks indicate significant differences with respective L + vehicle group at P < 0.05.
147
Chapter 7
A
100
We next investigated whether the effect of mast cell stabilizers on normalizing postoperative
gastric emptying could be ascribed to an attenuation of the intestinal inflammatory response
to bowel surgery. Therefore, the inflammatory infiltrate in the intestine was quantified by
measuring MPO activity in intestinal homogenates 24 hours after intestinal manipulation.
Figure 2 shows that intestinal manipulation significantly increased intestinal MPO activity
in mice treated with vehicle. This increase in postoperative intestinal MPO activity was
prevented by pretreatment with ketotifen or doxantrazole, compared with their respective
vehicle-treated controls, indicating that mast cell degranulation is an important step in the
establishment of the leukocyte infiltrate. We subsequently analyzed the effect of mast cell
*
60
MPO activity (U/mg tissue)
*
40
20
0
treatment
surgery
L
keto
L
IM
keto
IM
L
dox
L
IM
dox
IM
Figure 2. The increase in ileal myeloperoxidase (MPO) activity after surgery with intestinal manipulation is prevented by ketotifen or doxantrazole pretreatment. MPO activity was determined in whole
homogenates of ileum, isolated 24 hours after surgery. The MPO activity after IM in mice pretreated
with ketotifen vehicle is increased, but no increase is seen after pretreatment with mast cell stabilizer
ketotifen (solid bars). Similarly, doxantrazole pretreatment prevented postoperative increase in MPO
activity, compared with its respective vehicle treated control (grey bars). Treatment with ketotifen or
doxantrazole had no effect on basal MPO activity in control animals. Asterisks indicate significant differences in ketotifen or doxantrazole treatment group using a 1-way ANOVA (P < 0.05), followed by
Dunnett’s multiple comparison test. Data represent means ± SEM of 5–8 mice.
148
stabilization on the inflammation of the intestinal muscularis by staining for MPO containing
leukocytes in muscularic whole-mount preparations. Figure 3 shows that, at 24 hours after
surgery, extensive leukocyte infiltration into the intestinal muscularis was detected in mice
that were treated with vehicle, corroborating earlier results9,11; however, pretreatment with
either ketotifen or doxantrazole significantly reduced the number of leukocytes infiltrating
the intestinal muscularis tissue.
In Vivo Mast Cell Degranulation in the Ileum Results in Leukocyte
Infiltration and the Development of POI
To evaluate further the importance of mast cell degranulation in initiating muscularic
inflammation and the development of POI, an isolated bowel segment was exposed to
the mast cell secretagogue C48/80.27 C48/80 has been shown to activate and degranulate
CTMC effectively.28 We studied whether the mast cell degranulation and the subsequent
manipulation. To this end, we measured gastric emptying 24 hours after selective exposure
of the ileum to C48/80 (Figure 4). C48/80 incubation led to a significant delay in gastric
emptying, compared with a similar treatment with vehicle (0.9% NaCl). Half-emptying
times, as well as gastric retention at 60 minutes after consumption of the test meal, were
Chapter 7
muscularic inflammation elicited gastroparesis, similar to that seen after intestinal
significantly increased. Ketotifen pretreatment prevented the delay in gastric emptying
after C48/80 treatment Figure 4), implying that the gastroparesis developed as a result
of degranulation of CTMC. C48/80-induced mast cell degranulation resulted in a marked
5). The degree of muscular inflammation observed after this treatment was equal to that
24 hours after intestinal manipulation. In conjunction, the MPO activity in intestinal loops
exposed to C48/80 (Figure 6) was significantly increased, compared with loops exposed to
vehicle (0.9% NaCl).
Pretreatment of mice with ketotifen ablated the increase in MPO activity after C48/80
exposure but did not affect the MPO activity measured after treatment with saline (Figure
6), demonstrating that the increase in MPO activity seen after C48/80 treatment selectively
resulted from CTMC degranulation.
149
leukocyte infiltration into the intestinal muscularis 24 hours after C48/80 exposure (Figure
Figure 3. The appearance of leukocyte infiltrates in ileal muscularis after intestinal manipulation is
prevented by ketotifen or doxantrazole pretreatment. (A–D) Whole mount preparations of ileal intestinal muscularis tissue 24 hours after L (A), IM with ketotifen vehicle (B), IM with ketotifen pretreatment
(C), and IM with doxantrazole pretreatment (D) are stained for MPO positive leukocytes. IM with
either ketotifen or doxantrazole (not shown) vehicle pretreatment induced a massive influx of MPOpositive leukocytes to the ileal muscularis, compared with L (A and B). Pretreatment with ketotifen (C)
or doxantrazole (D) prevented this influx of inflammatory cells. Preparations shown are representative for 5–8 mice per treatment group. Bar is0.6 mm. (E) Shows that the significant increase in the
number of MPO positive leukocytes per mm2 of muscularis tissue after IM with (ketotifen) vehicle
pretreatment was prevented by ketotifen or doxantrazole pretreatment. Asterisk indicates significant
difference using a 1-way ANOVA (P < 0.05), followed by Dunnett’s multiple comparison test. Values
shown are the mean cell counts ± SEM of muscularis prepared from 5–8 mice.
150
60
C48/8
0
C48/8
0 + ke
20
totifen
saline
60
time (min)
ketotifen
pretreated
intestinal
exposure
t
+
saline
C48/80
C48/80
32.6 ± 2.2
59.1 ± 10.4*
37.6 ± 7.9
(min)
100
Ret60min (%)
30.4 ± 4.4
52.3 ± 7.4*
34.5 ± 5.3
Figure 4. Intestinal exposure to C48/80 delays gastric emptying. A Gastric emptying curves, determined by scintigraphic imaging of the abdomen after oral administration of solid caloric meal at 24
hours after exposure to C48/80 (solid circles), exposure to C48/80 after pretreatment with ketotifen
(open circles), and L alone (squares). Exposure to C48/80 results in a delay in gastric emptying,
which can be prevented by ketotifen pretreatment. Values are given as percentage of gastric content
compared with the total abdominal region. Corresponding half-emptying time (t1/2) as well as the
retention after 60 minutes. Ret60min is significantly increased after C48/80 exposure, compared
with vehicle (saline) (B). Pretreatment with ketotifen restores both t1/2 as well as Ret60min back to
normal. Values are means ± SEM of 8–12 mice per treatment group. Asterisks indicate significant
differences at P < 0.05.
151
Chapter 7
20
100
Figure 5. (see fullcolor chapter 11) Mast cell degranulation results in infiltration of leukocytes in ileal
muscularis. (A and B) Whole mount preparations of ileal intestinal muscularis tissue 24 hours after
exposure to vehicle (0.9% NaCl) (A) or C48/80 (B) are stained for MPO-positive leukocytes. Extensive inflammatory infiltrates were observed after exposure to C48/80 but not saline. Preparations
shown are representative for 6–8 mice per treatment group. Bar is 0.6 mm. Panel C shows that the
number of MPO-positive leukocytes was significantly increased after incubation with C48/80, compared with incubation with vehicle. Asterisk indicates significant difference (P < 0.05). Values shown
are the means ± SEM of 6–8 mice.
152
*
Chapter 7
30
10
ketotifen pretreatment
i ntesti nal exposure
-
NaCl
+
NaCl
C48/80
+
C48/80
Figure 6. Mast cell degranulation elicits an increase in ileal MPO activity that can be prevented by
ketotifen pretreatment. MPO activity was determined in whole homogenates of ileum isolated 24
hours after exposure to C48/80, or vehicle only (saline). The MPO activity after exposure to C48/80
is significantly increased, whereas exposure to saline did not affect MPO activity, irrespective of ketotifen pretreatment. Note that the increase in MPO activity elicited by C48/80 exposure was prevented
by ketotifen pretreatment. Asterisk indicates significant difference using a 1-way ANOVA (P < 0.05),
followed by Dunnett’s multiple comparison test. Data represent means ± SEM of 6–8 mice.
153
50
Mast Cell-Deficient Mice Are Resistant to Manipulation-Induced
Muscularic Inflammation
To confirm further that mast cells participate in the generation of the intestinal inflammation
that mediates POI, we performed abdominal surgery on Kit/Kitv mutant mice. These mice
have been shown to lack mast cells in all anatomical sites investigated.23 Indeed, no mast
cells could be identified in tissue sections of intestine, mesentery, or cytospins of peritoneal
fluid after staining with Toluidine blue (Figure 7, left panels). Intestinal manipulation
performed on Kit/WT congenic wild-type mice elicited mast cell degranulation indicated by
the increased level of MMCP-1 in their peritoneal lavage fluid measured 20 minutes after
surgery, compared with control laparotomy (13.0 ± 2.0 vs. 0.3 ± 0.2 ng/mL, respectively).
As expected, levels of peritoneal MMCP-1 levels were hardly detectable in Kit/Kitv mutant
mice and did not increase upon intestinal manipulation (0.2 ± 0.1 and 0.3 ± 0.2 ng/mL after
laparotomy and intestinal manipulation, respectively).
154
The absence of mast cells led to a significant reduction in the manipulation-induced
inflammation of the intestine. Intestinal manipulation performed on Kit/WT congenic wildtype mice resulted in an increased MPO activity into the intestinal tissue, compared with
laparotomy alone (Figure 8, solid bars). In Kit/Kitv mutant mice, however, MPO activity
was not significantly increased after intestinal manipulation. In concert, the number of
leukocytes infiltrating the intestinal muscularis after abdominal surgery performed on Kit/
Kitv mutant mice was significantly reduced compared with the number seen in Kit/WT wild-
Figure 7. (see fullcolor chapter 11) Reconstituted mast cells in Kit/Kitv mutant mice have a normal
phenotypic appearance. (A and C) Mast cells were absent in Kit/Kitv small intestinal muscularis and
Peyer’s patch (LM, longitudinal muscle layer; CM, circular muscle layer; PP, Peyer’s patch), as well
as in peritoneal fluid (E). Sections of small intestinal muscularis (B and D) and peritoneal fluid (F) of
Kit/Kitv mice reconstituted with cultured bone marrow-derived Kit/WT wild-type mast cells. The number of mast cells recovered in reconstituted mice is similar to that in wild-type mice, and they have
a normal histology and granule content (arrows). Giemsa staining. Sections are representative of 5
mice examined in each group. Bar is 75 μm.
155
Chapter 7
type muscularis (Figure 9A and B).
50
*
*
30
10
L
Kit/Kit v
IM
Kit/Kit v
IM
Kit/WT
IM
Kit/Kit v
IM
Kit/Kit v
PBS
Kit/WT MC
Figure 8. Intestinal inflammation after intestinal manipulation depends on the presence of mast cells.
MPO activity was determined in whole homogenates of ileum isolated 24 hours after L or IM. A significant increase in MPO activity and inflammation was observed after IM in wild-type mice but not in
mast cell deficient Kit/Kitv mutants (solid bars). Reconstitution of Kit/Kitv mutant mice with cultured
Kit/WT mast cells restored the granulocyte infiltration after intestinal manipulation to wild-type levels.
Asterisks indicate significant differences (P < 0.05). Data represent means ± SEM of 5 mice.
156
Chapter 7
157
Figure 9. (see fullcolor chapter 11) Granulocyte infiltration into the intestinal muscularis after
intestinal manipulation in mast cell deficient- and mast cell–reconstituted mice. (A–D) Whole
mount preparations of ileal intestinal muscularis tissue 24 hours after IM stained for MPO-positive
leukocytes. Extensive inflammatory infiltrates were observed after IM in Kit/WT mice (A), but the
number was drastically reduced in Kit/Kitv mutant mice (B). Reconstitution of Kit/Kitv mutant mice
with Kit/WT mast cells restored the inflammatory response to IM (D), whereas reconstitution with
PBS did not (C). Preparations shown are the representative for 5 mice per treatment group. Bar
is 0.6 mm. The number of MPO-positive leukocytes was significantly decreased after IM in Kit/
Kitv mutant mice, compared with Kit/WT wild-type, whereas mast cell reconstitution restored the
inflammatory response to IM. Asterisk indicates significant difference (P < 0.05). Values shown
are the means ± SEM of 5 mice.
Reconstitution of Mast Cells Restores Manipulation-Induced
Intestinal Inflammation in Mast Cell-Deficient Mice
To demonstrate directly the role of mast cells in the manipulation-induced inflammatory
response in the intestinal muscularis, mast cell populations were restored in Kit/Kitv mice
by adoptive transfer of cultured mast cells derived from congenic Kit/WT wild-type mice. If
mast cell deficiency alone would account for the lack of muscularic inflammation observed
in Kit/Kitv mice, reconstitution of the mast cell population in these animals should restore
the inflammation to the level of Kit/WT wild-type mice. To this end, we performed mast cell
reconstitution in Kit/Kitv recipients by IP injection of cultured bone marrow-derived mast
cells to repair the mast cell deficit. Reconstitution of Kit/Kitv mice gave rise to phenotypically
(Figure 7, right panels) and quantitatively normal mast cell populations in peritoneal lavage
fluid, mesentery, and intestine 10–12 weeks after transplantation, confirming earlier
reports.23 We performed abdominal surgery on mast cell reconstituted Kit/Kitv mice and
investigated the intestinal MPO activity and leukocyte infiltration 24 hours after surgery. In
Figure 8, it is shown that MPO activity was significantly increased after bowel manipulation
in reconstituted animals, compared with non-reconstituted, age-matched, mast cell
deficient mice (Figure 8, gray bars). Analysis of the MPO containing leukocytes in the
intestinal muscularis stained by muscularic whole-mount staining (Figure 9, panels C and
D) showed that intestinal manipulation gave rise to a significant increase in the number of
infiltrating leukocytes in mast cell-reconstituted mice, whereas surgery performed on nonreconstituted Kit/Kitv mice injected with PBS did not.
158
Discussion
In previous studies, it has been established that inflammation of the small intestinal muscularis
resulting from bowel manipulation is the main contributor to the prolonged phase of POI.9,11
The mechanism leading to the inflammatory response to bowel manipulation however, is
not known. Intense activation of visceral afferents, i.e., because of mechanical stretch, can
result in the local release of sensory neurotransmitters, especially substance P and CGRP.
These neuromediators have pro-inflammatory effects in that their release has been shown
to elicit a neurogenic inflammation at the activated tissue site.29,30 In our current model,
visceral afferents most likely are triggered to release these neuropeptides during intestinal
manipulation. Mast cells have been shown to be in close contact with visceral afferent
nerve terminals,31 and secreted neuropeptides can directly activate mast cells.32–34 We
therefore investigated the role of mast cell activation in the recruitment of the manipulation-
of peritoneal mast cell protease mMCP-1. Second, stabilization of mast cells using either
doxantrazole or ketotifen prevented this increase and prevented the intestinal inflammation
and delayed gastric emptying following bowel manipulation as well. The involvement of
Chapter 7
induced inflammation. First, we showed that bowel manipulation indeed increased the level
mast cells in this process was further demonstrated by exposure of a segment of small
intestine to the mast cell-degranulating compound C48/80. Similar to bowel manipulation,
local mast cell degranulation induced by C48/80 resulted in inflammation of the exposed
mast cells in the generation of the inflammation following intestinal manipulation.
To confirm further the requirement of mast cells in the induction of the manipulation-induced
intestinal inflammation, we conducted experiments in mast cell-deficient Kit/Kitv animals.
These mice displayed a significantly reduced inflammation after bowel manipulation. The
mast cell deficiency of these mice is due to mutations in the c-kit receptor gene, which
impairs the development of functional mast cells derived from the bone marrow. Because
of the fact that this mutation also affects other cell lineages, such as red blood cells and
melanocytes,35 these mice are mildly anemic, although immune responses have been
described to be generally similar to wild-type mice.36 In addition, the lack of a functional
c-kit receptor affects proper development of the network of the interstitial cells of Cajal,37
159
bowel segment and delayed gastric emptying. These data clearly illustrate a crucial role of
resulting in a disturbed gastrointestinal motility in these mice. Loss of these cells has been
associated with aberrations in gastric emptying.38 Therefore, data of gastric emptying were
not obtained from these mice.
To rule out the possibility that resistance of mast cell-deficient mice to manipulation-induced
inflammation is due to anemia or other defects resulting from the W mutation in Kit/Kitv
mice apart from mast cell deficiency, adoptive transfer of immature mast cells derived from
bone marrow cells of congenic normal Kit/WT mice was performed to repair selectively
the mast cell deficiency of the Kit/Kitv recipients. The final maturation and phenotype of
bone marrow-derived cultured mast cells transferred to mast cell-deficient Kit/Kitv mice
has been shown to be determined by the tissue in which mast cells are located.39 Hence,
mast cell-reconstituted Kit/Kitv mice differ only from Kit/Kitv mice in their presence of mast
cells. We found the mast cell populations in reconstituted mice histologically normal, and
normal numbers of mast cells were recovered from intestine, stomach, mesentery, and
peritoneum, which confirms earlier reports.23 We observed that mast cell reconstitution
of Kit/Kitv mice with mast cells derived from wild-type animals restored the manipulationinduced inflammatory response in the
intestinal muscularis back to wild-type levels,
showing that neutrophil infiltration and subsequent muscularic inflammation triggered by
intestinal manipulation are mast cell dependent.
MMC in the intestinal mucosa or CTMC in mesentery, serosa, and lamina propria have distinct
phenotypes and functions.14 Treatment with doxantrazole, which stabilizes both MMC and
CTMC,20,40 and ketotifen, a stabilizer of mainly CTMC,20 were both effective in preventing
the occurrence of muscular inflammation that follows bowel surgery. Unexpectedly, our
data also indicate that doxantrazole tended to be less effective in reducing postoperative
gastroparesis, although both stabilizers were equally potent in attenuating mMCP-1
release in the peritoneal cavity. Nevertheless, we must conclude that CTMC are involved
early in the process of the recruitment of inflammatory cells following bowel manipulation.
In particular, CTMC, and not MMC, respond to C48/80.41 C48/80 exposure mimicked the
manipulation-induced inflammation and gastroparesis, again suggesting that only CTMC
are involved in the initiation of the manipulation-induced inflammation. Paradoxically
however, intestinal manipulation elicited an increase in the level of mMCP-1, a soluble
chymase derived from MMC and not CTMC.14 These 2 observations indicate that both
160
MMC and CTMC degranulate as a result of intestinal manipulation. Our observation that
CTMC degranulation using C48/80 results in an inflammation in the muscularis externa, i.e.,
a site anatomically distinct from the mesentery, implies that CTMC (for instance adhering
to the intestinal serosa) can easily exert proinflammatory effects on surrounding intestinal
tissue.
Kalff et al. have previously suggested that the intestinal inflammation observed after
manipulation results from activation of resident macrophages, in rodents11 as well as in
humans,42 possibly through the enhanced expression of LFA-1.11 The mechanism by which
these macrophages may be activated, however, has not been studied. One likely possibility
is that macrophages are activated through proinflammatory mast cell mediators released,
such as TNF-α.43 In addition, other mast cell-derived mediators, such as histamine,44
prostaglandines, and tryptase,45 can orchestrate inflammation, possibly via activation of
cannot be concluded from our data but is currently the subject of ongoing studies.
In conclusion, our findings demonstrate that mast cells play an essential role in the genesis
of the muscularic inflammation mediating POI. Furthermore, we showed that mast cell
Chapter 7
macrophages. To what extent mast cells activate resident macrophages or vice versa
stabilization resulted in shortening the period of POI following bowel surgery. Mast cell
stabilizing agents are commonly used in the treatment of asthma and allergic disorders,46
and mast cell stabilizers have been shown in animal models22,47 and humans48,49 to attenuate
cell stabilization as a possible treatment for POI are certainly warranted.
161
the severity of active colitis. Based on our data, clinical studies evaluating the effect of mast
Reference List
1. Prasad M, Matthews JB. Deflating postoperative ileus. Gastroenterology 1999;117:489–492.
2. Livingston EH, Passaro EP Jr. Postoperative ileus. Dig Dis Sci 1990;35:121–132.
3. De Winter BY, Boeckxstaens GE, De Man JG, Moreels TG, Herman AG, Pelckmans PA. Effect
of adrenergic and nitrergic blockade on experimental ileus in rats. Br J Pharmacol 1997;120:
464–468.
abdominal surgery-induced gastric ileus in rats. Am J Physiol 1996;270:R888–R894.
celiac ganglia partly prevent postoperative gastric ileus. Peptides 1993;14:1225–1229.
6. Raybould HE, Kolve E, Tache Y. Central nervous system action of calcitonin gene-related peptide to inhibit gastric emptying in the conscious rat. Peptides 1988;9:735–737.
7. Zittel TT, Reddy SN, Plourde V, Raybould HE. Role of spinal afferents and calcitonin gene-related peptide in the postoperative gastric ileus in anesthetized rats. Ann Surg 1994;219:79–87.
8. Bonaz B, Tache Y. Corticotropin-releasing factor and systemic capsaicin-sensitive afferents
are involved in abdominal surgeryinduced Fos expression in the paraventricular nucleus of the
hypothalamus. Brain Res 1997;748:12–20.
9. de Jonge WJ, van den Wijngaard RM, The FO, ter Beek ML, Bennink RJ, Tytgat GN, Buijs
2003;125:1137–1147.
dysfunction in rat. Surgery 1999;126:498–509.
11. Kalff JC, Carlos TM, Schraut WH, Billiar TR, Simmons RL, Bauer AJ. Surgically induced leukocytic infiltrates within the rat intestinal muscularis mediate postoperative ileus. Gastroenterology 1999;117:378–387.
12. Kalff JC, Cicalese L, Exner B, Schraut WH, Bauer AJ. Role of phagocytes in causing dysmotility after each stage of small bowel transplantation. Transplant Proc 1998;30:2568.
13. Moriwaki K, Fujii K, Yuge O. Protein exudation induced by manipulation of the intestines and
mesentery during laparotomy in rat. A study of the mechanism of “third space” loss. In Vivo
1997;11:325–327.
14. Miller HR, Pemberton AD. Tissue-specific expression of mast cell granule serine proteinases
and their role in inflammation in the lung and gut. Immunology 2002;105:375–390
15. Biedermann T, Kneilling M, Mailhammer R, Maier K, Sander CA, Kollias G, Kunkel SL, Hultner
L, Rocken M. Mast cells control neutrophil recruitment during T-cell-mediated delayed-type
hypersensitivity reactions through tumor necrosis factor and macrophage inflammatory protein
2. J Exp Med 2000;192:1441–1152.
16. Galli SJ, Gordon JR, Wershil BK. Cytokine production by mast cells and basophils. Curr Opin
Immunol 1991;3:865–872.
17. Chen R, Ning G, Zhao ML, Fleming MG, Diaz LA, Werb Z, Liu Z. Mast cells play a key role in
neutrophil recruitment in experimental bullous pemphigoid. J Clin Invest 2001;108:1151–1158.
18. Kanwar S, Kubes P. Mast cells contribute to ischemia-reperfusion-induced granulocyte infiltration and intestinal dysfunction. Am J Physiol 1994;267:G316–G321.
19. Williams CM, Galli SJ. Mast cells can amplify airway reactivity and features of chronic inflammation in an asthma model in mice. J Exp Med 2000;192:455–462.
20. Pearce FL, Befus AD, Gauldie J, Bienenstock J. Mucosal mast cells. II. Effects of anti-allergic
compounds on histamine secretion by isolated intestinal mast cells. J Immunol 1982;128:
2481–2486.
21. Brown JF, Chafee KA, Tepperman BL. Role of mast cells, neutrophils and nitric oxide in
endotoxin-induced damage to the neonatal rat colon. Br J Pharmacol 1998;123:31–38.
162
163
Chapter 7
22. Eliakim R, Karmeli F, Okon E, Rachmilewitz D. Ketotifen ameliorates capsaicin-augmented
acetic acid-induced colitis. Dig Dis Sci 1995;40:503–509.
23. Kitamura Y, Go S, Hatanaka K. Decrease of mast cells in W/Wv mice and their increase by
bone marrow transplantation. Blood 1978;52:447–452.
24. Wershil BK, Mekori YA, Murakami T, Galli SJ. 125I-fibrin deposition in IgE-dependent immediate hypersensitivity reactions in mouse skin. Demonstration of the role of mast cells using
genetically mast cell-deficient mice locally reconstituted with cultured mast cells. J Immunol
1987;139:2605–2614.
25. Bennink RJ, De Jonge WJ, Symonds EL, Van Den Wijngaard RM, Spijkerboer AL, Benninga
using dedicated animal pinhole scintigraphy. J Nucl Med 2003;44: 1099–1104.
26. Symonds EL, Butler RN, Omari TI. Assessment of gastric emptying in the mouse using the
[13C]-octanoic acid breath test. Clin Exp Pharmacol Physiol 2000;27:671–675.
27. Koibuchi Y, Ichikawa A, Nakagawa M, Tomita K. Binding of active components of compound
48/80 to rat peritoneal mast cells. Eur J Pharmacol 1985;115:171–177.
28. Jaffery G, Coleman JW, Huntley J, Bell EB. Mast cell recovery following chronic treatment with
compound 48/80. Int Arch Allergy Immunol 1994;105:274–280.
29. Quinlan KL, Song IS, Naik SM, Letran EL, Olerud JE, Bunnett NW, Armstrong CA, Caughman
SW, Ansel JC. VCAM-1 expression on human dermal microvascular endothelial cells is directly
and specifically up-regulated by substance P. J Immunol 1999;162: 1656–1661.
30. Smith CH, Barker JN, Morris RW, MacDonald DM, Lee TH. Neuropeptides induce rapid expression of endothelial cell adhesion molecules and elicit granulocytic infiltration in human skin.
J Immunol 1993;151:3274–3282.
31. Williams RM, Berthoud HR, Stead RH. Vagal afferent nerve fibres contact mast cells in rat
small intestinal mucosa. Neuroimmunomodulation 1997;4:266–270.
32. Ansel JC, Brown JR, Payan DG, Brown MA. Substance P selectively activates TNF-alpha gene
expression in murine mast cells. J Immunol 1993;150:4478–4485.
neurite-mast cell communication in vitro occurs via the neuropeptide substance P. J Immunol
1999;163: 2410–2415.
34. Theoharides TC, Singh LK, Boucher W, Pang X, Letourneau R, Webster E, Chrousos G.
Corticotropin-releasing hormone induces skin mast cell degranulation and increased vascular permeability, a possible explanation for its proinflammatory effects. Endocrinology
1998;139:403–413.
35. Wershil BK, Tsai M, Geissler EN, Zsebo KM, Galli SJ. The rat c-kit ligand, stem cell factor,
induces c-kit receptor-dependent mouse mast cell activation in vivo. Evidence that signaling
through the c-kit receptor can induce expression of cellular function. J Exp Med 1992;175:245–
255.
36. Wershil BK, Furuta GT, Wang ZS, Galli SJ. Mast cell-dependent neutrophil and mononuclear
cell recruitment in immunoglobulin E-induced gastric reactions in mice. Gastroenterology
1996;110:1482–1490.
37. Huizinga JD, Thuneberg L, Kluppel M, Malysz J, Mikkelsen HB, Bernstein A. W/kit gene required for interstitial cells of Cajal and for intestinal pacemaker activity. Nature 1995;373:347–
349.
38. Ordog T, Takayama I, Cheung WK, Ward SM, Sanders KM. Remodeling of networks of interstitial cells of Cajal in a murine model of diabetic gastroparesis. Diabetes 2000;49:1731–1739.
39. Kitamura Y, Morii E, Ogihara H, Jippo T, Ito A. Mutant mice: a useful tool for studying the development of mast cells. Int Arch Allergy Immunol 2001;124:16–19.
40. Francois A, Ksas B, Aigueperse J, Griffiths NM. The recovery of the neurally evoked secretory response of rat colonic mucosa after irradiation is independent of mast cells. Radiat Res
2002;157:266–274.
41. Befus AD, Pearce FL, Gauldie J, Horsewood P, Bienenstock J. Mucosal mast cells. I. Isolation
and functional characteristics of rat intestinal mast cells. J Immunol 1982;128:2475–2480.
Bauer AJ. Intra-abdominal activation of a local inflammatory response within the human muscularis externa during laparotomy. Ann Surg 2003;237:301–315.
43. Bissonnette EY, Enciso JA, Befus AD. Inhibitory effects of sulfasalazine and its metabolites on
histamine release and TNF-α production by mast cells. J Immunol 1996;156:218–223.
44. Torres R, de Castellarnau C, Ferrer LL, Puigdemont A, Santamaria LF, de Mora F. Mast cells
induce upregulation of P-selectin and intercellular adhesion molecule 1 on carotid endothelial
cells in a new in vitro model of mast cell to endothelial cell communication. Immunol Cell Biol
2002;80:170–177.
45. Vergnolle N. Proteinase-activated receptor-2-activating peptides induce leukocyte rolling, adhesion, and extravasation in vivo. J Immunol 1999;163:5064–5069.
46. Slater JW, Zechnich AD, Haxby DG. Second-generation antihistamines: a comparative review.
Drugs 1999;57:31–47.
47. Eliakim R, Karmeli F, Okon E, Rachmilewitz D. Ketotifen effectively prevents mucosal damage
in experimental colitis. Gut 1992;33:1498–1503.
48. Marshall JK, Irvine EJ. Ketotifen treatment of active colitis in patients with 5-aminosalicylate
intolerance. Can J Gastroenterol 1998;12:273–275.
49. Jones NL, Roifman CM, Griffiths AM, Sherman P. Ketotifen therapy for acute ulcerative colitis
in children: a pilot study. Dig Dis Sci 1998;43:609–615.
164
165
Chapter 7
8
8
Chapter
8
Intestinal Handling Induced
Mast
Cell
Inflammation
Activation
in
Postoperative Ileus
and
Human
Gut 2008; 57: 33-40
Frans O. The,
Roelof J. Bennink,
Willem M Ankum,
Marrije R. Buist,
Olivier R.C. Busch,
Dirk J. Gouma,
Sicco van der Heide,
Wouter J. de Jonge,
Guy E. Boeckxstaens
Abstract
Background & Aims: Murine postoperative ileus results from intestinal inflammation triggered by manipulation-induced mast cell activation. As its extent depends on the degree
of handling and subsequent inflammation, we hypothesise that the faster recovery after
minimal invasive surgery results from decreased mast cell activation and impaired intestinal inflammation. Objective: to quantify mast cell activation and inflammation in patients
undergoing conventional and minimal invasive surgery. Methods: 1)Mast cell activation
(i.e. tryptase release) and pro-inflammatory mediator release were determined in peritoneal lavage fluid obtained on consecutive time-points during open, laparoscopic and transvaginal gynaecological surgery. 2)LFA-1, ICAM-1 and iNOS mRNA as well as leukocyte
influx were quantified in non-handled and handled jejunal muscle specimens collected during biliary reconstructive surgery. 3)Intestinal leukocyte influx was assessed by 99mTc labelled leukocyte SPECT-CT scanning before and after abdominal or vaginal hysterectomy.
Results: 1)Intestinal handling during abdominal hysterectomy resulted in an immediate
release of tryptase followed by enhanced IL-6 and IL-8 levels. None of the mediators rose
during minimal invasive surgery except for a slight increase in IL-8 during laparoscopic
surgery. 2)Jejunal mRNA transcription for ICAM-1 and iNOS as well as leukocyte recruitment were increased after intestinal handling. 3)Leukocyte scanning 24hrs after surgery
revealed increased intestinal activity after abdominal but not after vaginal hysterectomy.
Conclusions: This study demonstrates that intestinal handling triggers mast cells activation and inflammation associated with prolonged postoperative ileus. These results may
partly explain the faster recovery after minimal invasive surgery and encourage future clinical trials targeting mast cells to shorten postoperative ileus.
168
P
Background
Postoperative ileus, characterized by a lack of coordinated motility of the entire
gastrointestinal tract leads to increased morbidity and prolonged hospitalisation1,2 and
represents a substantial socio-economical burden. In the US alone, the additional annual
healthcare expenses related to this condition have been estimated to surpass 1 billion
dollars3. At present, treatment is rather disappointing and limited to predominantly supportive
measures4.
The introduction of minimal invasive surgical techniques (e.g. laparoscopy) has fastened
postoperative recovery significantly5. This major improvement is believed to result from
minimal wound trauma and decreased release of stress hormones6,7. In addition, it is
becoming increasingly clear that intestinal inflammation is a key event in the pathogenesis
of postoperative ileus. In rats, the degree of gut paralysis is directly proportional to the
degree of intestinal handling and inflammation8. This inflammation leads to local impaired
muscle contractility9 and the activation of an adrenergic inhibitory neural pathway10.
induced inflammation. As minimal invasive surgery implies limited handling of the intestine,
faster recovery of motility may result from an impaired influx of inflammatory cells.
Although the exact mechanism remains unclear, we previously showed that mast cells
play a pivotal role in triggering the inflammatory process. In mice, intestinal handling led
to degranulation of mast cells with increased levels of mouse mast cell protease-1 in
peritoneal lavage fluid. In contrast, W/Wv mice, deficient of mast cells, failed to develop an
intestinal muscle inflammation in response to manipulation of a bowel loop. Reconstitution
of W/Wv mice with mast cells from wild type animals restored the handling induced
inflammatory response, clearly demonstrating the importance of mast cells. Activation of
resident macrophages has also been demonstrated, possibly secondary to influx of luminal
bacteria during a brief episode of increased mucosal permeability12. To what extent mast
cell activation is the trigger leading to increased mucosal permeability and macrophage
activation remains to be determined.
169
Mast Cell induced Inflammation in Human Postoperative Ileus
shortens postoperative ileus10,11 and further underscores the importance of this handling
Chapter 8
Reduction of inflammatory cell influx accomplished by blockade of adhesion molecules
At present the evidence supporting the importance of inflammation in human is rather
scarce13, and data on the relationship between the degree of inflammation and clinical
outcome are lacking. In addition, although we provided convincing evidence for a crucial
role of mast cell degranulation in mice, no data are available in man. In the present study,
therefore, we studied whether intestinal manipulation leads to mast cell degranulation
and inflammation in patients undergoing conventional or minimal invasive surgery and
hypothesised that clinical recovery is determined by the degree of manipulation induced
mast cell activation and inflammation.
170
Patients and Methods
Participants
Between December 2003 and July 2005 a total of 44 patients were enrolled in 3 clinical
research protocols. The physical condition and co-morbidity of potential participants
was assessed during pre-assessment at the outpatient clinic of the department of
anaesthesiology, which is part of the standard pre-operative work-up. The American Society
of Anesthesiologists-Physical Status classification (ASA-PS)14,15 was used and comprises
a scale from 1 to 6 in which 1 equals a normal healthy patient; 2 equals a patient with
mild systemic disease; 3 equals a patient with severe systemic disease; 4 is a patient
with severe systemic disease that is a constant threat to life; 5 equals a moribund patient
who is not expected to survive without the operation and 6 equals a declared brain-dead
patient whose organs are being removed for donor purposes14. In the present study, only
patients categorized as ASA-PS 1 to 3 were asked to participate. In addition, patients were
screened for the following exclusion criteria: intra-abdominal inflammation, pre-operative
Study design
The activation of mast cells and the inflammatory mediator response to intestinal handling
were evaluated in protocol 1. Pro-inflammatory gene transcription and leukocyte recruitment
within the handled intestinal muscle layer were studied in protocol 2. The occurrence of
manipulation induced leukocyte recruitment in relation to clinical recovery of bowel function
and duration of hospital admission was evaluated in study protocol 3. All protocols were
evaluated and approved by the Medical Ethical Review Board of the Academic Medical
Center, Amsterdam, the Netherlands.
Anaesthesia
To correct for the influence of used anesthetic technique and medication, patients were
subjected to standardized peri-operative care according to our anaesthesiologist’s protocol.
In short, patients were pre-medicated with paracetamol 1000mg and lorazepam 1mg on
the evening before surgery and approximately 2hrs before surgery. Induction of general
171
were included after informed consent was obtained.
Chapter 8
radiation therapy and the use of anti-inflammatory or mast cell stabilizing drugs. Patients
anaesthesia was attained with propovol 2-2.5mg/kg; fentanyl 1.5-3μg/kg; rocuronium
0.6mg/kg. Anesthesia was maintained using et. 0.8% isoflurane. Postoperative pain
medication was introduced after last sampling in protocol 1 and 2 and administered similar
for all patients according to our anesthesiologist’s pain protocol.
Protocol 1: Mast cell activation and inflammatory mediator release during abdominal
surgery
Peritoneal lavage fluid samples were collected from 18 patients, either undergoing an
abdominal hysterectomy (n=6), a laparoscopic resection of an adnexum (n=6) or a transvaginal hysterectomy (n=6). Three consecutive lavages were performed in each individual
patient. The first lavage sample was collected immediately after opening of the peritoneum
(basal). A second lavage sample was collected immediately after abdominal inspection
and first gentle small intestinal handling (early). The final lavage sample was collected at
the end of the procedure (late). As the intestine is not handled in the trans-vaginal group,
only two lavages were performed in the trans-vaginal hysterectomy group (i.e. basal and
late sampling). The systemic release of mediators in response to abdominal surgery was
assessed in 2 blood samples (abdominal hysterectomy only), the first sample taken before
induction of general anaesthesia (1day prior to surgery) and the second sample taken
at the end of the surgical procedure, i.e. just before closure of the abdominal cavity. The
harvested fluid and serum were used to measure the release of tryptase, Tumor Necrosis
Factor (TNF)-α Interleukin (IL)-1β, IL-6 and IL-8 in relation to surgical handling. The
abdominal lavage was performed using 100ml of warm (42°C) sterile 0.9% NaCl solution,
which was sprinkled gently onto the small intestine and its mesentery. After approximately
30 seconds, peritoneal fluid (between 20 and 40 ml) was collected using a 22 French Foley
catheter (Bard Limited, West Sussex, England) connected to a 50ml catheter tip syringe.
Protocol 2: Regulatory gene transcription and leukocyte influx upon intestinal
handling
Jejunal muscle specimens were used to quantify regulatory gene transcription and
assess the degree of inflammation. Full thickness biopsies were obtained from patients
undergoing biliary reconstructive surgery. This specific procedure was chosen because of
its considerable length, providing at least sufficient time for gene transcription to occur16.
Two consecutive jejunal tissue samples were collected from 10 patients. The first specimen
172
was collected at the beginning of the procedure and had not been touched by the surgeon
until resection. The second tissue specimen, exposed to the usual handling during surgery,
was collected approximately 3 hrs thereafter. Following mucosa removal, both specimens
were partitioned (5mm2 segments) and snap frozen in liquid nitrogen in the operating
theatre and stored at -80°C.
Protocol 3: Abdominal leukocyte recruitment and clinical recovery
Abdominal leukocyte Single Photon Emission Computed Tomography (SPECT) CT scans
were performed in 16 gynaecological patients to quantify the leukocyte recruitment in
response to surgical handling. Eight patients undergoing an abdominal hysterectomy were
compared with 8 patients undergoing a vaginal hysterectomy. In each patient a reference
(basal) leukocyte scintigraphy was performed on the day of admission, 24 hrs prior to
surgery. A second leukocyte scintigraphy was performed on the first postoperative day, i.e.
approximately 24 hrs after surgery. Clinical recovery was assessed until hospital discharge
Chapter 8
(see methods for detailed description).
173
Methods
Tryptase release
Tryptase concentrations were assessed at the routine clinical laboratory of the Department
of Allergy, University Medical Center, Groningen, the Netherlands. Total tryptase
(α-protryptase and β-tryptase) concentration was measured in peripheral blood and lavage
fluid samples using a commercial fluoro-immunoenzyme assay (FIA) (Pharmacia Uppsala,
Sweden)17.
Cytokine and chemokine release
Cytokine levels were determined by cytometric bead array (BD PharMingen, San Diego,
CA, USA). In short, 5μl of each test sample was mixed with 5μl of mixed capture beads
and 5μl of human phycoerythrin (PE) detection reagents consisting of PE-conjugated antihuman IL-1β, TNF-α, IL-6 and IL8. These mixtures were incubated at room temperature in
dark for 3 hrs, washed and resuspended in 300ul wash buffer. Acquisition was performed
on a FACSCalibur using a high throughput-sampling interface (BD Biosciences, Sunnyvale,
CA, USA). Generated data were analyzed using CBA software (BD PharMingen, San
Diego, CA, U SA) and interpolated from corresponding standard curves generated using
the mixed cytokine standard provided by the supplier18.19.
Real Time Reverse Transcription-Polymerase Chain Reaction
Tissue specimens were homogenized and total RNA was extracted using Trizol
(Invitrogen, Carlsbad, CA, USA). The total RNA fractions were treated with DNAse
and reverse-transcribed using Superscript II (Invitrogen, Carlsbad, CA, USA).
cDNA (150ng) was subjected to 45 cycles of lightcycler PCR (FastStartDNA
Masterplus
SYBR
Green;
Roche,
Basel
Switzerland).
The
following
primers
were used: LFA-1 antisense 5’-GACCCAAGTGCTCTCAGGAA-3’ and sense 5’AGGAGCACTCCACTTCATGC-3’; ICAM-1 antisense 5’-CATAGAGACCCCGTTGCCTA-3’
and sense 5’-GGGTAAGGTTCTTGCCCACT- 3’; iNOS antisense 5’-TGGAAGCGGTA
ACAAAGGAGA- 3’ and sense 5’ CGATGCACAGCTGAGTGAAT- 3’; GAPDH antisense 5’CGACCACTTTGTCAAGCTCA-3’ and sense 5’-AGGGGAGATTCAGTGTGGTG-3’. PCR
quantification was performed by a linear regression method using the Log(fluorescence)
174
per cycle number20 and normalized for GAPDH housekeeping gene expression. In each
individual patient the late sample value was expressed as fold increase of the early control
sample value.
Immunocytochemistry
Immunocytochemical staining was performed on peritoneal cell cytospins obtained from
the harvested abdominal lavage fluid. In short, spins containing 1x105 cells were fixed in
carnoy’s fixation fluid (60% ethanol, 30% chloroform and 10% glacial acidic acid) for 30
min at room temperature and washed with TBST (0.1%). Non-specific binding of antibody
was blocked by incubation with TBS containing 10% normal goat serum for 20 min. Spins
were incubated with anti-tryptase antibodies (mouse anti-human, 1:250) (Chemicon,
Temecula, CA, USA) for 2 hrs at room temperature. Goat-anti mouse alexa-488 was used
as secondary antibody (Molecular probes, Invitrogen, Carlsbad, CA, USA). After final
washing, the spins were mounted using Vectashield mounting medium containing 5μg/ml
DAPI (Vector Laboratories, Burlingame, CA, USA).
inflammation
Handled and non-handled jejunal muscle sections were used to assess the extent of
inflammation. Leukocyte infiltration was visualized by myeloperoxidase (MPO) staining as
Chapter 8
Semi-quantitative evaluation of the degree of intestinal muscle
(8 μ) were incubated for 10 minutes with 3-amino-9-ethyl carbazole (Sigma, St. Louis,
MO) as a substrate, dissolved in sodium acetate buffer (pH 5.0) to which 0.01% H2O2 was
added10. To evaluate the degree of inflammation, unmarked myeloperoxidase stained early
and late collected section from 10 patients were scored independently by 3 observers
(TK, OW and RVDW). A semi-quantitative scoring scale from 0 to 4 was utilized; 0 being
non-inflamed, 1=very mildly inflamed, 2=mildly inflamed, 3= inflamed and 4 being clearly
inflamed. The mean of 3 scores, calculated for each segment, was used for statistical
analysis (Wilcoxon signed rank test).
In-vivo quantification of leukocyte recruitment
White blood cells (WBC) were labeled using technetium-99m hexamethylpropyleneamine
oxime (99mTc-HMPAO) (Ceretec, GE Health, Eindhoven, The Netherlands) according to
175
described previously(11). After 10min fixation in ice-cold acetone, transverse frozen section
the consensus protocol for leukocyte labelling21. The harvested WBC fraction of 100ml of
blood labeled with an average of 450 ± 10 MBq of
99m
Tc-HMPAO was reinjected into the
patient. Sixty min later, a SPECT scan of the abdomen was performed (GE Millennium
Hawkeye, GE Healthcare, Den Bosch, the Netherlands) followed by a low dose CT-scan
without contrast on the same gantry. CT data were used for attenuation correction and as
an anatomical reference for region of interest (ROI) analysis. After data acquisition, images
were processed on an Entegra workstation (GE Healthcare, Den Bosch, The Netherlands)
using attenuation corrected iterative reconstruction and analyzed on a Hermes workstation
(Nuclear Diagnostics, Stockholm, Sweden). Five consecutive abdominal SPECT slices
were summed and ROI’s were drawn around small intestine and lumbar spine at the level
of the ileac crest. Small bowel uptake of leukocytes was calculated as an uptake ratio
expressed as a fraction of bone marrow activity, similar to analysis of leukocyte uptake
assessment in inflammatory bowel disease22. The small bowel uptake ratio determined
prior to surgery was considered as basal leukocyte activity. The relative percentage of
difference in leukocyte activity 24h after surgery was calculated using the following formula:
(postoperative small bowel ratio/preoperative small bowel ratio)*100%.
Clinical evaluation
All patients received standard postoperative medical care according to the ward accustomed
care protocol. Patients were visited by the research physician once daily until discharge
to assess postoperative clinical recovery of bowel function (time of first flatus and time of
first defecation). Patients were discharged when the following criteria were met: normal
urinary-tract function, spontaneous defecation, tolerance of oral fluid and solid food intake,
adequate pain relief with oral analgesics and adequate mobilization and self-support.
Statistical analysis was performed using SPSS 12.02 software for Windows. Data were
non-parametrically distributed and expressed as median values and inter quartile range
or median increase compared to basal values. In protocol 1, all serum but only vaginal
hysterectomy lavage samples were analysed using a Wilcoxon signed rank test for 2 paired
samples. For all other lavage sample-series (consisting of 3 samples) a Friedman’s two
way analysis of variance was applied. When a statistical difference was observed, a MannWhitney test used to identify the specific sample(s) of significant difference. In protocol
176
2, the quantitative PCR data and the semi-quantitative inflammation data were analyzed
using a Wilcoxon signed rank test. In protocol 3, leukocyte recruitment was analyzed using
the Wilcoxon signed rank test. Clinical data were analyzed with a Mann-Whitney test for
Chapter 8
independent samples. P-values <0.05 was considered statistically significant.
177
Results
Patient demographics
Eighteen patients participated in study protocol 1, 6 in each surgical intervention group.
Overall mean age was 47 years, range 21 to 70 (trans–vaginal: 52 years, range 43 to
70; laparoscopy: 36 years, range 21 to 49; laparotomy: 49 years, range 44 to 53). The
indications for surgery in this patient population were leyomyomata (n=8), prolaps (n=4) or
a benign ovarian tumour (n=6). In study protocol 2, jejunal tissue samples were collected
from 10 patients (6 male, mean age 42 years, range 32 to 53) who underwent biliary
reconstructive surgery because of iatrogenic biliary tract injury. Study protocol 3 involved
16 patients; 8 patients underwent an abdominal hysterectomy (mean age 50 years, range
42 to 70) and 8 patients underwent a vaginal hysterectomy (mean age 55 years, range 42
to 66). The indications for surgery were uterine leiomyomata in the abdominal hysterectomy
patient group and uterine prolapse (n=4), leyomyomata (n=3) and primary dysmenorrhoea
(n=1) in the vaginal hysterectomy group.
Study protocol 1:
Mast cell activation and inflammatory response during abdominal surgery
To assess the activation of mast cells in response to intestinal handling, the expression and
release of tryptase, a pre-stored mast cell specific protease23, was analyzed. Peritoneal
lavage fluid harvested during abdominal surgery (laparotomy) contained a distinct mast
cell population, as illustrated by the number of tryptase positive cells in fig. 1a. In the
basal lavage sample collected immediately after opening of the peritoneal cavity, the
basal median tryptase concentration was 5.2 (InterQuartile Range (IQR) 2.7-11.3) μg/l.
Tryptase release was significantly increased to a median concentration of 23.1 (IQR 15.146.9) μg/l, (p=0.02) in early samples taken after gentle palpation of the small intestines,
necessary to allow inspection of the pelvic organs. In the late sample taken at the end
of surgery, tryptase levels had increased even further (late: median concentration 51.7
(IQR 25.8-90.2) μg/l, n=6, p=0.002) (fig.1b). In contrast, neither laparoscopic nor transvaginal intra-peritoneal surgery (n=6 in both types of surgery) elicited a significant mast cell
response (fig. 1c-d). To evaluate possible release of mast cell mediators in the systemic
circulation, we also determined pre- and postoperative serum tryptase concentrations in
178
B
laparotomy
tryptase conc. ( g/L)
150
*†
125
100
75
25
0
basal
CC
early
time during surgery
late
laparoscop
laparoscopy
y
( /
150
12
125
10
100
757
0
bas
basal
DD
ear
early
timeduring
duri gsurgery
surger
time
la
late
trans-vagina
trans-vaginal
15
150
12
125
10
100
757
505
252
0
179
bas
basal
ear
early
timeduring
dur gsurgery
surger
time
a
late
252
Chapter 8
505
y
. ( g/
Figure 1 (see fullcolor chapter 11)
a) Peritoneal cells collected in late lavage fluid stained for tryptase
(in green). Cell nuclei
2 0
were counterstained with DAPI (bleu). Individual patient tryptase concentrations during
b) open surgery, c) laparoscopic surgery and
d) trans-vaginal surgery; measured in lavage
fluid collected immediately after opening of
the peritoneal cavity (basal), after first palpation
of the small intestine during inspection
apa
oscopy
ar scop
of pelvic organs (early) and at the end of the
procedure (late). A Wilcoxon signed rank
test (vaginal samples) and Friedman’s two
way analysis of variance (laparoscopic and
laparotomy samples) were used to determine statistical significance. Tryptase levels
increased significant in patient undergoing a
laparotomy (n=6, p=0.002) in contrast to the
laparoscopic (n=6, p=0.5) or vaginal (n=6,
p=0.06) approach. Note that no “early” lavage was performed in patients undergoing
trans-vaginal surgery. Dotted line represents
late
median change in tryptase
concentration of
du
ng
su
g
du
ing
surge
all 6 patients
*
50
the laparotomy group (n=6 patients). Serum tryptase levels did not increase and remained
within the normal range of 1 to 11.4 μg/l24 (pre-operative median concentration 4.1 (IQR
2.8-7.5) μg/l and postoperative 1.6 (IQR 1.3-3.5) μg/l respectively). The release of the
pro-inflammatory cytokines TNF-α IL-1β, IL-6 and the chemokine IL-8 was analyzed in
the same peritoneal lavage samples. Gentle handling of the intestine during laparotomy
did not lead to an immediate increase in any of these mediators. However, at the end of
the surgical procedure, IL-6 and IL-8 were increased significantly (table 1). In laparoscopic
treated patients, intra-peritoneal IL-8, but not IL-6, was increased, but not as profound as
in the laparotomy group (table 1). On the other hand, transvaginal surgery did not affect
any of the measured cytokines and chemokines. TNF-α and IL-1β levels did not change
upon first handling or at the end of any of the types of surgery evaluated. The serum levels
of the studied inflammatory mediators remained unaltered (median increase compared to
pre-operative for: TNF-α: 0.0 (IQR 0.0-0.0) pg/ml; IL-1β: 0.0 (IQR 0.0-35.3) pg/ml; IL-6: 4.2
(IQR 0.0-11.7) pg/ml; IL-8: 0.9 (IQR 0.0-310.4) pg/ml).
Table 1: Inflammatory mediator release during surgery
treatment group
laparotomy (n=6)
laparoscopy (n=6)
trans-vaginal (n=6)
TNF-
0.0 (0.0-3.4)
0.0 (-6.6-0.0)
0.0 (0.0-0.0)
IL-1
0.0 (-1.8-18.1)
-2.5 (-3.3-0.0)
0.2 (0.0-0.8)
IL-6
135.6 (4.2-5130.0) 1
6.1 (1.3-15.2)
1.5 (-0.4-4.0)
IL-8
114.2 (32.9-208.7) 2
28.9 (1.3-166.5) 1
0.8 (-2.8-6.0)
median increase of mediator concentration in late vs. basal lavage sample collected during indicated
type of surgery (pg/ml) with (IQR)
1
2
p=0.02; p=0.006
Table 1
Pro-inflammatory mediator release in lavage fluids collected immediately after opening of the peritoneal cavity (basal), after first palpation of the small intestine during inspection of pelvic organs (early)
and at the end of the surgical procedure (late). A Wilcoxon signed rank test (vaginal samples) and
Friedman’s two way analysis of variance (laparoscopic and laparotomy samples) were used to determine statistical significance. TNF-α and IL-1β did not change during surgery. IL-8 increased significant at the end of laparoscopic as well as open (laparotomy) surgery. IL-6 only increased at the end
of a laparotomy. None of the pro-inflammatory proteins increased in the trans-vaginal surgery group.
2: Qualavage
tita ive
e tra sc
iption undergoing
nalys s. trans-vaginal surgery.
Note thatable
no “early”
wasge
performed
in patients
180
0
Study protocol 2
Regulatory gene transcription upon intestinal handling
Recruitment of leukocytes to the muscularis propria strongly depends on the upregulation
of adhesion molecules and the synthesis of pro-inflammatory proteins. Therefore, ICAM-1,
LFA-1 and iNOS gene expression was determined in muscle specimens collected during
abdominal surgery. As the synthesis of functional proteins requires several hours16, mRNA
quantification was used to evaluate the kinetics of these inflammatory proteins in jejunal
muscle tissue. As shown in Table 2, iNOS and ICAM-1 levels were significantly increased
after
(table r2)r In acontrast
remained unchanged.
e 1: intestinal
Inflamm handling
tory mediat
durinLFA-1
s gery
Degree of intestinal muscle inflammation upon intestinal handling.
Histological evaluation of leukocyte recruitment was performed before and after
surgical handling on the same tissue specimens used for gene-transcription analysis.
.6 stained
.2
0.
.
. - recruitment in response 0to intestinal
Myeloperoxidase was
to visualize leukocyte
1 4 2 sections
32.9 208of7the jejunal
8 muscularis
9 1.3-166.5
0.8 (-2.8 6.0)
handling in transverse
propria. Non-handled
early samples
contained
only
a ator
small
number
of nleukocytes
layer
2 ing
upper
an increase of
m d
concen
rati n
late vs.
s in the
va emuscle
sample col
ec (fig
ed du
in cleft
tedpanel). In
f surg
(pg ml) wi h
(IQR)
p=0 specimens
2; p=0 00 showed a marked extravasation of inflammatory
contrast,
routinely
handled
late
cells (fig. 2 upper right panel), confirmed by semi-quantitative evaluation (fig.2 lower panel).
These recruited leukocytes predominantly reside in and around the vasculature of the
handled intestinal muscle layers as is illustrated in figure 2b. This extravasation marks the
Chapter 8
1
Table 2: Quantitative gene transcription analysis.
Gene
median fold increase of gene expression
LFA-1
0.9 (0.3-20.0)
ICAM-1
3.3 (1.3-139.9) 1
iNOS
3.3 (0.7-20.0) 2
median fold increase of gene transcription in late (handled) vs. early (non-handled)
jejunal muscle layer with (IQR) (n= 10 patients)
1
2
p=0.017; p=0.022
Table 2
Relative increase of mRNA expression in manipulated small intestinal muscle tissue compared to
non-handled control specimens. A Wilcoxon signed rank test was used to determine statistical differences. The relative increase was significant for iNOS (median fold increase 3.3 (IQR 0.7-20.0),
p=0.022) and ICAM1 (median fold increase 3.3 (IQR 1.3-139.9), p=0.017).
181
ongoing inflammatory process.
early
late
100x
100x
p=0.005
degree of inflammation
(scale 0-4)
d
ee
4
3
2
1
0
early
late
Figure 2
Handling induced leukocyte infiltration of the jejunal muscularis propria, visualized by myeloperoxidase staining in early non-handled (upper panel, left) and late handled tissue segments (upper panel
right). Note the ongoing extravasation, illustrated by the predominant peri-vascular localization of
leukocytes in the handled late tissue sample. Semi-quantitative evaluation of handling induced leukocyte recruitment is depicted in the lower panel (scale 0= non-inflamed through 1=very mildly inflamed,
2=mildly inflamed, 3=inflamed to 4= clearly inflamed). Early non-handled (median score: 1 (IQR 1-2))
vs. Late handled (median score: 3 (IQR 2-4)), n= paired samples from10 patients, p=0.005 tested
with a Wilcoxon signed rank test. The dotted line represents the median increase in intestinal muscle
inflammation of all 10 patients.
182
Study protocol 3
Abdominal leukocyte recruitment 24hrs after open and minimal invasive hysterectomy
In vivo leukocyte recruitment in response to intestinal handling was investigated by
99m
Tc
labeled leukocyte imaging. Abdominal leukocyte influx was assessed on 5 consecutive
leukocyte-SPECT images at the level of the ileac crest and compared with that of the bone
marrow(22). The change in leukocyte activity before compared to after surgery showed
no increase in the vaginal hysterectomy group (median % of activity before compared
to after surgery 91% (IQR 84-102), n=8). In the abdominal hysterectomy group however
leukocyte recruitment was significantly increased to a median of 127% of the pre-operative
abdominal activity ((IQR 113-148), n=8, p=0.01) (fig 3). To determine the exact anatomical
location, plain CT-images were made immediately after SPECT imaging. The region in
which the enhanced leukocyte activity was observed coincided with small intestinal loops
*
175
150
125
100
75
abdominal
hysterectomy
vaginal
hysterectomy
Figure 3
Quantification of postoperative leukocyte recruitment to the small intestinal region expressed as percentage (%) of the preoperative scan. A significant increase (Wilcoxon signed rank test) in intestinal
leukocyte activity was observed after an abdominal hysterectomy (median % of preoperative scan
127% (IQR 113-148), n=8, p=0.01), but not after a vaginal hysterectomy (median % of preoperative
scan 91% (IQR 84-102), ns, n=8).
183
post-operative leukocyte
activity (% of basal)
200
Chapter 8
and its mesentery, as shown in figure 4.
Figure 4 (see fullcolor chapter 11)
Representative example of leukocyte SPECT-CT imaging 24 hrs before (left column) and after (right
column) an abdominal hysterectomy was performed. a) Coronary SPECT overview slide with anatomical references: (1) liver, (2) bladder, (3) ileac spine and (4) lumbar vertebral-range between which
quantification was performed. b) transverse CT-slide and c) corresponding SPECT image at same
position in quantification range, visualizing increased leukocyte activity in the abdominal region (arrows). Finally, d) transfers CT- and SPECT overlay showing the specific leukocyte activity in the small
intestine (arrow heads).
184
Clinical recovery after open and minimal invasive hysterectomy
In conjunction with the assessed leukocyte recruitment, clinical recovery was also evaluated.
Time until first flatus did not differ significantly between the two patient groups. First bowel
movement and duration of hospital admission however were significantly prolonged after
abdominal hysterectomies compared to the vaginal procedure (table 3).
mean age (years)
mean ASA-PS*
mean time of surgery (min)
abdominal
vaginal
hysterectomy
hysterectomy
50 (range 42-70)
55 (range 42-66)
2 (range 1-3)
1 (range 1-2)
182 (range 130-298) 150 (range 113-179)
mean time until first flatulence (days)
2 (range 2-3)
1 (range 1-2)
mean time until first defecation (days)
4 (range 4-5) 1
2 (range 2-3)
mean time until discharge (days)
8 (range 7-8) 2
4 (range 4-5)
*ASA-PS: American Society of Anaesthesiologists-Physical Status (scale 1 (being a normal
healthy patients) to 6 (being a patient declared brain-dead) see methods section for detailed
description)
1
2
p=0.02, p=0.001
Table 3
Patient demographics and clinical recovery data from patients undergoing a vaginal hysterectomy or
an abdominal hysterectomy assessed in protocol 3. To identify potential confounders in clinical parameters and to test significant difference a Mann-Whitney test was performed. Age and ASA-score
did not differ between the two patient populations. Time till first defecation (1 p=0.02) and time until
discharge 2p=0.001) were both significantly prolonged in patients undergoing an abdominal hysterectomy when compared to those undergoing a vaginal hysterectomy. All data are depicted as mean
and (range).
185
treatment group
Chapter 8
Table 3: Clinical evaluation of post-operative recovery
Discussion
Inflammation of the muscularis propria following surgical manipulation of the intestine is
increasingly recognized to postpone the recovery of gastrointestinal motility. Animal studies
indeed have revealed that prevention of this inflammatory process, either by antibodies
or antisense oligonucleotides to the adhesion molecule ICAM-1, macrophage inactivation
or COX-2 inhibition enhances gastrointestinal transit and shortens postoperative ileus11,
25-27
. Recently, we demonstrated that mast cell activation plays an important role in this
process and may be one of the first steps triggering the inflammatory response. Intestinal
manipulation indeed induces the immediate activation of mast cells leading to increased
levels of the murine mast cell proteinase-1 in the abdominal cavity25. Three hours later,
inflammatory mediators such as MIP-2, MIP-1α TNFα and IL-6 can be detected27, 28 which
on their turn enhance the expression of adhesion molecules such as ICAM-111, recruitment
of leukocytes and inflammation of the intestine.
In the present study, we investigated whether this cascade of events also plays a role in the
pathogenesis of human postoperative ileus. We found that gentle palpation of the intestines
during first inspection of the pelvic organs resulted in the instantaneous intra-peritoneal
release of tryptase, a mast cell specific protease23, in patient undergoing a laparotomy.
This increase in tryptase increased even further towards the end of the procedure and
was accompanied by an increase in IL-6 and IL-8. As the latter is known to be released in
response to mast cell activation and initiates leukocyte recruitment via ICAM-129, we also
determined ICAM-1 and iNOS mRNA in intestinal tissue that was handled at the start of a
surgical procedure, but was only removed approximately three hours later. In addition to
an upregulation of ICAM-1 and iNOS, the number of inflammatory cells was significantly
increased in these late tissue samples compared to untouched specimen harvested at the
beginning of the procedure. Interestingly, leukocytes were localized predominantly around
blood vessels in both the serosa and the muscularis propria, partly adhering to the endothelial
lining marking the ongoing recruitment and extravasation in this early stage of inflammation.
Kalff et al. also reported an intestinal inflammatory response during abdominal surgery in
patients13. To assess the degree of inflammation in a later stage, we also performed
99m
Tc
labelled leukocyte SPECT scanning 24 hrs after surgery. Using this technique, we showed
186
increased intra-abdominal activity compared to the pre-operative baseline scan in patients
subjected to an abdominal hysterectomy. As the actual resection, performed in the pelvic
region, did not comprise any gastrointestinal organs, this observed increase in leukocyte
activity can not be explained by the primary surgical trauma. Clearly, leukocytes could
reside anywhere in the abdominal cavity and may not be restricted to the intestinal wall.
The additional CT scanning however showed that the increased leukocyte activity observed
with the SPECT scans coincided with intestinal loops. When the uterus was resected
transvaginally, a surgical approach that leaves the intestines largely untouched, no such
increase was observed, indirectly suggesting that the intestinal inflammation is triggered by
intestinal manipulation. From these data we conclude that also in man, manipulation of the
intestine during surgery leads to mast cell degranulation and a local inflammatory process,
which, like in our animal model, plays an important role in postoperative hypomotility.
In rodents, the extent of gastrointestinal hypomotility or ileus is proportionally related to the
degree of intestinal handling and subsequent inflammation8. As intestinal manipulation is
and the subsequent inflammatory response in the intestine will be less and thus may
contribute to the faster clinical recovery observed after minimal invasive surgery. To test
this hypothesis, tryptase and inflammatory mediators were quantified during 2 minimal
invasive surgical procedures, i.e. laparoscopic and trans-vaginal hysterectomy. In contrast
Chapter 8
minimal in laparoscopic surgery, one might argue that the degree of mast cell degranulation
was observed in the peritoneal lavage fluid. Only IL-8 levels were increased, although
less profound compared to laparotomy. Moreover, during trans-vaginal surgery, leaving the
intestines largely untouched, none of the evaluated parameters increased. These findings
underscore that the degree of intestinal handling to a large extent determines the degree
of mast cell activation and the subsequent inflammatory response. The latter was further
confirmed by the
99m
Tc labelled leukocyte SPECT scanning 24 hrs after surgery showing
increased intra-abdominal activity in patients subjected to an abdominal hysterectomy but
not in patients who underwent a trans-vaginal hysterectomy. Finally, clinical recovery in our
study was significantly delayed after abdominal compared to vaginal hysterectomy, a finding
in line with previous clinical studies showing faster postoperative gastrointestinal recovery
after minimal invasive surgery6, 30-35. It should be emphasized though that differences in
postoperative pain medication, especially opioids, may have contributed to the delay in
187
to gentle handling during open surgery, no mast cell degranulation or increase of IL-6
normalisation of gastrointestinal motility36,37. In the current study, however, postoperative
analgesia in both patient groups was provided according to a standardized postoperative
pain protocol, making this explanation less likely. Therefore, the observation that delayed
clinical recovery is associated with increased influx of radio-labeled leukocytes indirectly
adds to the hypothesis that the degree of intestinal handling, mast cell degranulation and
subsequent inflammation determine the duration of postoperative ileus.
A drawback of this study is that the study-protocols were conducted in different groups of
patients. Ideally, the same patient cohort should have been studied to better understand
the causative association between mast cell degranulation and the subsequently observed
inflammatory responses upon intestinal handling. Especially as mRNA levels of ICAM-1
and iNOS peak only 2 to 24hrs after stimulation16, 38 a long-lasting surgical procedure had to
be chosen in order to allow the detection of the upregulation of these inflammatory markers
in response to intestinal handling. Therefore patients undergoing biliary reconstructive
surgery were selected instead.
Our current findings may have important clinical implications. First, they clearly illustrate
that manipulation of the intestine should be limited whenever possible in order to reduce
the release of mast cell mediators and limit postoperative intestinal inflammation. This
knowledge should urge further development of minimal invasive surgical or even endoscopic
techniques to minimise intestinal handling. Second, if mast cell degranulation is indeed
an important initial step in the pathophysiology of postoperative ileus in man, mast cells
may represent an important therapeutic target. As we previously showed reduction of
postoperative ileus by mast cell stabilisation in our mouse model, our current findings in
humans warrant further studies evaluating the effect of a mast cell stabilising agent in
patients.
188
189
1. Collins TC, Daley J, Henderson WH et al. Risk factors for prolonged length of stay after major
elective surgery. Ann Surg 1999;230(2):251-9.
2. Longo WE, Virgo KS, Johnson FE et al. Risk factors for morbidity and mortality after colectomy
for colon cancer. Dis Colon Rectum 2000;43(1):83-91.
3. Prasad M, Matthews JB. Deflating postoperative ileus. Gastroenterology1999;117(2):489-92.
4. Kehlet H, Holte K. Review of postoperative ileus. Am J Surg 2001;182(5A Suppl):3S-10S.
5. Schwenk W, Haase O, Neudecker J et al. Short term benefits for laparoscopic colorectal resection. Cochrane Database Syst Rev 2005;(3):CD003145.
6. Chen HH, Wexner SD, Iroatulam AJ et al. Laparoscopic colectomy compares favorably with colectomy by laparotomy for reduction of postoperative ileus. Dis Colon Rectum
2000;43(1):61-5.
7. Glaser F, Sannwald GA, Buhr HJ et al. General stress response to conventional and laparoscopic cholecystectomy. Ann Surg 1995;221(4):372-80.
8. Kalff JC, Schraut WH, Simmons RL et al. Surgical manipulation of the gut elicits an intestinal
muscularis inflammatory response resulting in postsurgical ileus. Ann Surg 1998;228(5):65263.
9. Kalff JC, Carlos TM, Schraut WH et al. Surgically induced leukocytic infiltrates within the rat
intestinal muscularis mediate postoperative ileus. Gastroenterology 1999;117(2):378-87.
10. de Jonge WJ, van den Wijngaard RM, The FO et al. Postoperative ileus is maintained by
2003;125(4):1137-47.
11. The FO, de Jonge WJ, Bennink RJ et al. The ICAM-1 antisense oligonucleotide ISIS- 3082
prevents the development of postoperative ileus in mice. Br J Pharmacol 2005.
12. Schwarz NT, Beer-Stolz D, Simmons RL et al. Pathogenesis of paralytic ileus: intestinal manipulation opens a transient pathway between the intestinal lumen and the leukocytic infiltrate
of the jejunal muscularis. Ann Surg 2002;235(1):31-40.
13. Kalff JC, Turler A, Schwarz NT et al. Intra-abdominal activation of a local inflammatory response within the human muscularis externa during laparotomy. Ann Surg 2003;237(3):301-15.
14. Saklad M. Grading of patients for surgical procedures. Anesthesiology 1941;2:281-4.
15. New Classification of Physical Status. American Society of Aesthesiologists,Inc. Anesthesiology 1963;24:111.
16. Yan HC, Juhasz I, Pilewski J et al. Human/severe combined immunodeficient mouse chimeras.
An experimental in vivo model system to study the regulation of human endothelial cell-leukocyte adhesion molecules. J Clin Invest 1993;91(3):986-96.
17. Schwartz LB, Kepley C. Development of markers for human basophils and mast cells. J Allergy
Clin Immunol 1994;94(6 Pt 2):1231-40.
18. Tarnok A, Hambsch J, Chen R et al. Cytometric bead array to measure six cytokines in twentyfive microliters of serum. Clin Chem 2003;49(6 Pt 1):1000-2.
19. Chen R, Lowe L, Wilson JD et al. Simultaneous Quantification of Six Human Cytokines
in a Single Sample Using Microparticle-based Flow Cytometric Technology. Clin Chem
1999;45(9):1693-4.
20. Ramakers C, Ruijter JM, Deprez RH et al. Assumption-free analysis of quantitative realtime
polymerase chain reaction (PCR) data. Neurosci Lett 2003;339(1):62-6.
21. Roca M, Martin-Comin J, Becker W et al. A consensus protocol for white blood cells labelling
with technetium-99m hexamethylpropylene amine oxime. International Society of Radiolabeled
Blood Elements (ISORBE). Eur J Nucl Med 1998;25(7):797-9.
22. Weldon MJ, Masoomi AM, Britten AJ et al. Quantification of inflammatory bowel disease activity
using technetium-99m HMPAO labelled leucocyte single photon emission computerised tomography (SPECT). Gut 1995;36(2):243-50.
23. Hogan AD, Schwartz LB. Markers of mast cell degranulation. Methods 1997;13(1):43-52.
Chapter 8
Reference List
24. Schwartz LB, Bradford TR, Rouse C et al. Development of a new, more sensitive immunoassay
for human tryptase: use in systemic anaphylaxis. J Clin Immunol 1994;14(3):190-204.
25. de Jonge WJ, The FO, van der CD et al. Mast cell degranulation during abdominal surgery initiates postoperative ileus in mice. Gastroenterology 2004;127(2):535-45.
26. Schwarz NT, Kalff JC, Turler A et al. Prostanoid production via COX-2 as a causative mechanism of rodent postoperative ileus. Gastroenterology 2001;121(6):1354-71.
27. de Jonge WJ, van der Zanden EP, The FO et al. Stimulation of the vagus nerve attenuates macrophage activation by activating the Jak2-STAT3 signaling pathway. Nat Immunol
2005;6(8):844-51.
28. Wehner S, Behrendt FF, Lyutenski BN et al. Inhibition of macrophage function prevents intestinal inflammation and postoperative ileus in rodents. Gut 2006.
29. Compton SJ, Cairns JA, Holgate ST et al. The role of mast cell tryptase in regulating endothelial cell proliferation, cytokine release, and adhesion molecule expression: tryptase induces
expression of mRNA for IL-1 beta and IL-8 and stimulates the selective release of IL-8 from
human umbilical vein endothelial cells. J Immunol 1998;161(4):1939-46.
30. Isik-Akbay EF, Harmanli OH, Panganamamula UR et al. Hysterectomy in obese women: a
comparison of abdominal and vaginal routes. Obstet Gynecol 2004;104(4):710-4.
31. Veldkamp R, Kuhry E, Hop WC et al. Laparoscopic surgery versus open surgery for colon cancer: short-term outcomes of a randomised trial. Lancet Oncol 2005;6(7):477-84.
32. Graber JN, Schulte WJ, Condon RE et al. Relationship of duration of postoperative ileus to
extent and site of operative dissection. Surgery 1982;92(1):87-92.
33. Huilgol RL, Wright CM, Solomon MJ. Laparoscopic versus open ileocolic resection for Crohn’s
disease. J Laparoendosc Adv Surg Tech A 2004;14(2):61-5.
34. Bohm B, Milsom JW, Fazio VW. Postoperative intestinal motility following conventional and
laparoscopic intestinal surgery. Arch Surg 1995;130(4):415-9.
35. Milsom JW, Hammerhofer KA, Bohm B et al. Prospective, randomized trial comparing laparoscopic vs. conventional surgery for refractory ileocolic Crohn’s disease. Dis Colon Rectum
2001;44(1):1-8.
36. Miedema BW, Johnson JO. Methods for decreasing postoperative gut dysmotility. Lancet Oncol
2003;4(6):365-72.
37. Bauer AJ, Boeckxstaens GE. Mechanisms of postoperative ileus. Neurogastroenterol Motil
2004;16 Suppl 2:54-60.
38. Yoo HS, Rutherford MS, Maheswaran SK et al. Induction of nitric oxide production by bovine alveolar macrophages in response to Pasteurella haemolytica A1. Microb Pathog
1996;20(6):361-75.
190
191
Chapter 8
9
Chapter
9
Mast
9
Cell
Treatment
Stabilization
of
as
Postoperative
Ileus: a Pilot Study
submitted for publication
Frans O. The,
Marrije R. Buist,
Aaltje Lei,
Roelof J. Bennink,
Jan Hofland,
Wouter J. de Jonge,
Guy E. Boeckxstaens
Abstract
Background & Aim: Postoperative ileus is mediated by intestinal inflammation resulting
from manipulation-induced mast cell activation. Therefore, mast cell stabilization may
represent a new therapeutic approach to shorten postoperative ileus. Aim: To study the
effect of ketotifen, a mast cell stabilizer, on postoperative gastrointestinal transit in patients
who underwent abdominal surgery. Methods: In this pilot study, 60 patients undergoing
major abdominal surgery for gynecological malignancy with standardized anesthesia were
randomized to treatment with ketotifen (4 or 12mg) or placebo. Patients were treated for
6 days starting 3 days prior to surgery. Gastric emptying of liquids, selected as primary
outcome parameter, was measured 24 hrs after surgery using scintigraphy. Secondary
endpoints were, scintigraphically assessed, colonic transit represented as geometrical
center of activity (segment 1=cecum to 7=stool) and clinical parameters. Results: Gastric
retention 1 hr after liquid intake was significantly reduced by 12mg (median 3% (1-7),
p=0.01), but not by 4mg ketotifen (18 % (3-45), p=0.6) compared to placebo (16 %
(5-75)). Twenty-four hr colonic transit in placebo was 0.8 (0.0-1.1) vs. 1.2 (0.2-1.4) colon
segments in 12 mg ketotifen group (p=0.07). Abdominal cramps improved significantly in
patients treated with12mg ketotifen, whereas other clinical parameters were not affected.
Conclusion: Ketotifen significantly improves gastric emptying and showed a tendency
to improvement of colonic transit after abdominal surgery. These results warrant further
exploration of mast cell stabilizers as putative therapy for postoperative ileus.
194
P
Background
Postoperative ileus, characterized by generalized gastrointestinal hypomotility is a major
determinant of prolonged hospitalization after extensive abdominal surgery1. The annual
costs related to ileus have been estimated to exceed $1,000,000,000 in the US, illustrating
its socio-economical impact2. Until recently, neurogenic inhibition of gastrointestinal motility
was considered as main pathophysiological mechanism underlying postoperative ileus.
Animal experiments indeed revealed activation of adrenergic and non-adrenergic noncholinergic inhibitory pathways during and shortly after abdominal surgery3-5. To overcome
this inhibitory input, treatment so far has mainly focused on prokinetic drugs, such as
metoclopramide6-8, cisapride9-11 or erythromycin12, 13 with however disappointing results14, 15.
Therefore, there is a large need for other more efficient therapeutic strategies.
Recently, we and others have shown that local inflammation of the intestine triggered
by handling of bowel loops during surgery plays a crucial role in the pathogenesis of
postoperative ileus. In rodents, abdominal surgery indeed leads to influx of inflammatory
local inflammatory response not only leads to impaired neuromuscular function of the
affected intestinal segment, but also activates an adrenergic neural pathway inhibiting
the motility of the entire gastrointestinal tract. Most importantly, prevention of the influx
Chapter 9
cells, approximately 4 to 6 hours after the intestinal segment has been manipulated. This
of inflammatory cells by for example blocking adhesion molecules such as ICAM-1
inflammatory response in the pathophysiology of postoperative ileus. Also in man, we and
others provided evidence that abdominal surgery triggers an inflammatory response in
intestinal tissue resected at the end of the procedure18, 19. Moreover, using SPECT imaging,
we were able to provide in vivo evidence for influx of radiolabeled leukocytes into the
intestine after open hysterectomy but not after laparoscopic hysterectomy18.
One of the initial steps attracting inflammatory cells to the site of manipulation is mast
cell degranulation. Intestinal handling triggers the release of mast cell mediators both in
rodents20 and man18, whereas W/Wv mice, deficient of mast cells, fail to develop intestinal
inflammation in response to bowel manipulation. Reconstitution of these animals with
195
Mast Cell Stabilization in Ileus: a Randomized Trial
restores gastric emptying and intestinal transit16, 17, indicating the eminent role of this local
mast cells from their wild type littermates restores the occurrence of manipulation-induced
inflammation. Finally, pretreatment of mice with the mast cell stabilizing agents ketotifen
and doxantrazole prevents the occurrence of inflammation and normalizes postoperative
gastric emptying, suggesting that mast cell stabilization may be an attractive alternative
approach to treat postoperative ileus.
To investigate this hypothesis and to prove the concept that interference with the mast
cell – inflammation sequence indeed improves postoperative gastrointestinal motility, we
designed a double blind placebo controlled randomized pilot study evaluating the effect of
ketotifen on postoperative gastrointestinal transit.
196
Methods
Study subjects
The present study is a randomized, double blind, placebo controlled, single center proof
of principle study conducted in the Academic Medical Center (AMC), Amsterdam, the
Netherlands. This study was approved by the Medical Ethical Committee of the AMC.
Patients (18 – 80 years of age), scheduled to undergo a radical hysterectomy, debulking
of ovarian malignancy, or an oncological explorative laparotomy were invited to participate.
The exclusion criteria were: 1) evident intra-abdominal inflammation (diagnosed by imaging
and/or laboratory test results), 2) use of anti-allergic drugs, 3) use of anti-inflammatory
pharmaca during the first 3 days after surgery, 4) use of laxatives and/or prokinetic agents
during the first 3 post-operative days, 5) colostomy or ileostomy, 6) intestinal resection as
part of the surgical procedure, 7) American Society of Anesthesiologists physical-health
status(ASA-PS)21 > III.
with placebo, ketotifen 4mg or ketotifen 12mg in 2 daily oral doses. In order to avoid side
effects such as sedation, the drug was gradually introduced (1/4 of the full dose on day 1,
1/2 of the full dose on day 2 and the full dose from day 3 until day 6). Randomization was
Chapter 9
After written informed consent was obtained, patients were randomized to receive treatment
Study protocol
Treatment was started 3 days prior to surgery and was continued until the second
postoperative day. Patients were admitted to the hospital 1 day prior to surgery. On the
evening before the operation the patients were pre-medicated with lorazepam 1mg orally,
followed by 1mg on the day of operation, to which then paracetamol 1000mg was added
(table 1). Anesthesia, analgesia, peri-operative intravenous (iv) fluids and respiratory
support were standardized according to a pre-defined protocol (table 1). The nasogastric
decompression tube was removed in the recovery room or on the ward the morning of the
first postoperative day. Postoperative analgesia was attained with paracetamol 500mg 6
times a day, orally. Non-steroidal anti-inflammatory drugs (NSAID’s) and/or tramadol were
197
performed according to a 2:2:2 block ratio.
added on demand when iv. or epidural analgesia was ceased, however, NSAID’s were not
allowed before post-operative day 4.
On post-operative day 1 (24hrs after surgery), patients were asked to drink 100mL of
diethylenetriaminepentaacetate (111In-DTPA) labeled tap water. One, 24 and 48 hrs after
intake scintigraphical scans of the abdomen were performed (see section gastrointestinal
transit studies for details). Clinical recovery was monitored and symptoms were noted until
hospital discharge (see section data collection for details).
198
199
level
lorazepam 1mg
morning of surgery
choice
maintenance
top-up
test dose
table 1
Chapter 9
500ml BL 500ml Voluven; 1000ml BL 500ml
Voluven; 1500ml BL transfusion according to
hemoglobin
Start 500ml Voluve
substitution blood loss (BL)
thereafter 8ml/kg/hr
VT6-8ml/kg endtitle CO2 4-4.5kPa
1st hr 20ml/kg
PEEP 5cm H
2O
test-stop 6am day 3
post-operative
rocuronium 0.6mg/kg
lacteted ringer’s solution
FiO2 40% (air-O2)
continues i.v. morphine
patient controlled morphine i.v.
bupivacaïne 0.125% + fentanyl μg/mL
25
25mg bupivacaïne 0.25% + 100
μg fentanyl
4mL bupivacaine 0.25% + epinephrin 1:200,000
Th12-L1 ± 1
fentanyl 1.5-3μg/kg
paracetamol 1000mg
maintenance
Peri-operative i.v. fluids
Respiration
3rd choice
2nd
1st choice
epidural
isoflurane.0.8% end tidal
Maintenance
analgesia
propofol 2-2.5mg/kg
Induction
anesthesia
lorazepam 1mg
evening prior to surgery
pre-medication
peri-operative care and post-operative analgesia protocol
Gastrointestinal transit studies
Gastric emptying was assessed 24hrs after surgery. Patients were asked to drink 100mL
of tap water labeled with 4 MBq
111
In-DTPA (Tyco Healthcare, Petten, The Netherlands).
Sixty min after ingestion, a 5-min acquisition was performed in a 128 matrix with the
patient in supine position using a single head gamma camera (Siemens Diacam, Siemens,
Hoffman Estates, Il, USA) fitted with a medium energy collimator. The following formula
was used to calculate the relative gastric content (counts stomach-(pixels stomach*(counts
background/pixels background))* 100 and was depicted as percentage of activity present
in the stomach compared to the total activity in the abdominal region of interest, corrected
for background.
Colon transit was assessed 48 and 72 hrs after surgery. For this, 2 additional 5minacquisitions were performed 24 and 48 hrs after ingestion of the radio-labeled water
using the same single head gamma camera and settings also used for gastric emptying.
To enable calculation of colon transit, the colon was subdivided in 7 segments (i.e. 1 =
ascending colon, 2 = right colonic flexure, 3 = transverse colon, 4 = left colonic flexure, 5
= descending colon, 6 = sigmoid / rectum and 7 = stool). The centre of mass model22 was
applied expressing colonic transit as 24 and 48hrs postprandial geometrical centre (GC)
of activity. To correct for the influence of oro-cecal transit the 24hr shift (i.e. delta) in colon
GC was also calculated (GC 48hrs – GC 24hrs). Interpretation and calculation of gastric
retention and colon transit was done by one staff physician (RJB) of the nuclear medicine
department on a Hermes (Nuclear Diagnostics, Sweden) workstation.
Data collection
During hospital admission, patients were visited at least once daily, by a trial nurse and/
or research physician, for clinical evaluation (i.e. diet, first passage of flatus, first bowel
movement, vomiting, pain and discomfort). Prior to surgery, most frequently reported
adverse-events for ketotifen known from literature (i.e. drowsiness, dizziness, nausea and
headache)23 were scored daily, using a 100mm visual analog scale (VAS). After surgery,
patients were asked to rate the severity of pain, nausea and abdominal cramping on a VAS
scale every day until discharge. As department policy dictates a minimal hospitalization of
10 days for patients undergoing a radical hysterectomy, duration of hospitalization could
not be used as parameter to evaluate clinical recovery. Instead, patients were deemed
200
ready for discharge after tolerance of solid food, occurrence of first bowel movement and
adequate post-surgical pain control with oral analgesics in absence of complications.
The pre-defined primary endpoint of efficacy was formulated as the percentage of
111
In-
DTPA labeled liquid present in the stomach 1 hr after ingestion, measured 24hrs after
surgery. The secondary endpoints of this study were defined as follows: 1) GC of intracolonic mass 24 and 48hrs postprandially and the 24hr colon transit, i.e. delta GC between
24hrs and 48hrs after ingestion of 111In-DTPA labeled water; 2) time until ready for hospital
discharge; 3) time until first flatus in hrs after surgery; 4) time until first bowel movement in
hrs after surgery and 5) degree of post-operative pain, nausea and abdominal cramping
during first 5 days post-operative (mean time until ready for discharge) calculated as area
under the curve. As this study was designed as a proof of principle study, per-protocol
analysis was applied on all data.
Previous studies on gastrointestinal transit in healthy females24 indicated that 16 patients
ingestion of a non-caloric liquid test meal between placebo and ketotifen treated patients,
providing a 90% power.
Data were non-parametrically distributed and therefore expressed as median values and
Chapter 9
would suffice to identify a >10% significant (p<0.05) difference in gastric retention 1hr after
inter-quartile range. For paired and unpaired data the Wilcoxon signed rank or the Mann-
sets a Chi-square test was applied. For analyses of clinical symptom VASscores, the area
under the curve (AUC) was calculated for each individual patient. These AUC values were
statistically tested using an independent sample test. P<0.05 was considered statistically
significant. Statistical analysis was performed using SPSS 12.02 software for Windows
(SPSS Inc. Chicago, Ill, USA).
201
Whitney U test was used respectively. To identify potential confounders in ordinal data
Results
Study subjects
Between June 2004 and March 2006 a total of 60 female patients were enrolled in this
study, 20 patients in each treatment group (i.e. placebo, 4mg ketotifen and 12mg ketotifen
per day). One patient, randomized for placebo, withdrew from the trial because of non-drug
related personal reasons. A protocol violation was reported in 15 cases, 3 in the placebo,
7 in the 4 mg and 5 in the 12mg of ketotifen group. Therefore data from 44 patients was
available for full analysis. Six cases were excluded on post-operative day 2 or 3 because
of laxative/prokinetic drug administration (placebo n=1, ketotifen 4mg n=3 and 12mg n=2).
Hence, only gastric emptying studies of these patients were included in the analysis (n=50;
fig.1).
The majority of patients (36) underwent a radical hysterectomy, 14 patients underwent
tumor debulking and 10 patients an explorative laparotomy because of suspicion of
malignant disease. The distribution of the three types of surgery was statistically equal in
all three treatment groups. From the remaining baseline characteristics only ASA-health
classification was not equally distributed (table 2). Pain medication (i.e. paracetamol,
NSAIDs and tramadol) consumption, calculated as median AUC for daily consumption until
ready for discharge, did not differ between the 3 treatment groups (table 2).
Evaluation of adverse events potentially related to ketotifen use revealed no serious
adverse events. None of the patients indicated they were considering withdrawal from the
trial because of side effects. In accordance with previous reports on ketotifen drowsiness
was the most frequently noted adverse event (median VAS for placebo: 0.0 (InterQuartile
Range (IQR) 0.0-0.0 vs. ketotifen 4mg: 0.6 (IQR 0.0-5.5), p=0.02 and 12mg: 2.5 (IQR 0.07.1), p=0.002).
202
203
completed study
n=15
protocol violation
laxatives/prokinetics (n=2)
completed gastric emptying
n=17
Chapter 9
completed study
n=13
protocol violation
placebo
n=20
protocol violation
anti-histamines (n=2)
withdrew informed consent (n=1)
completed study
n=16
protocol violation
n=17
protocol violation
missed medication (n=2)
colostomy (n=1)
anti-inflammatory agents (n=1)
ketotifen 4mg
n=20
Enrolled
n=60
n=16
protocol violation
missed medication (n=1)
surgery canceled (n=1)
anti-inflammatory agents (n=1)
ketotifen 12mg
n=20
Figure 1
Flowchart of patient enrolment, protocol violation, exclusion and number of patients completing the
study.
204
n=20
47 ± 12
enrollment
age (yrs)
1
morphine continues i.v.
75 (0-200)mg
tramadol
125 (0-213)mg
0 (0-0)mg
2792 (2725-2919)mg
2
3
15
3.3 ± 1.6 SD
2 (2-2)
5
8
7
55 ± 11
n=20
ketotifen 4mg
100 (0-238)mg
0 (0-0)mg
2808 (2646-2951)mg
1
2
16
3.8 ± 1.7
2 (1-2)
4
3
13
48 ± 10
n=20
ketotifen 12mg
a
b
table 2 exact value, mean ± SD or median (inter-quartile range) =excluded prior to gastric emptying studies and = after gastric emptying studies
0 (0-125)mg
NSAIDs
2875 (2716-2893)mg
4
morphine patient controlled i.v.
paracetamol (median AUC)
15
epdural
analgesia
4.7 ± 1.8
duration of surgery (hrs)
1
explorative laparotomy
1 (1-2)
3
debulking
ASA-PS score (scale 1-5)
16
radical hysterectomy
type of surgery
placebo
parameter
patient characteristics
ns
ns
ns
ns
ns
p=0.02
ns
ns
p-value
Gastrointestinal transit
1. Gastric emptying
Gastric emptying was determined 24 hrs after surgery. One hour after ingestion of 100 mL
of radio-labeled tap water, the residual gastric radioactivity was calculated as measure of
emptying. As shown in figure 2, gastric emptying of patients treated with placebo varied
considerable, ranging from complete emptying to gastric stasis with more than 90% of
radiolabeled material still present in the stomach. The median gastric retention was 16 %
(IQR 5-75). Treatment with 4 mg ketotifen did not significantly change gastric emptying
(gastric retention: 18 % (IQR 3-45), p=0.6) compared to placebo. In contrast, gastric
emptying of patients treated with 12mg ketotifen was significantly improved with a median
gastric retention of 3% (IQR 1-7), p=0.01) (fig. 2).
relative
gastric contents (%)
100
90
80
70
60
50
40
30
20
10
0
placebo
ketotifen
4mg
ketotifen
12mg
Figure 2
Scintigraphical evaluation of gastric emptying of an 111In labeled non-caloric liquid test-meal 24hrs
after surgery. Gastric retention determined 1hr postprandial, depicted as median percentage in stomach compared to total abdominal region corrected for background.
205
Chapter 9
p=0.01
Fifteen of the 17 patients had almost completely emptied their stomach one hour after
ingestion of the radiolabeled water. To put things in perspective, 47% of patients in the
placebo group and 44% in 4mg ketotifen showed >20% residual gastric content one hour
after ingestion of radiolabeled tab water in comparison to only 12% of patients in the 12mg
A
ketotifen.
placebo
cebo
B
26
l
6
seg
colon colon
segment
colon segmentc
n
5
g
4
04
n
3
c
15
5
3
2
2
5
4
3
02
1
0
C
1
GC
24h
postprandial
postprandial
0
GC
48h
postprandial
po tprandial
ketotifen 12mg
GC
24h postprandial
D
24 hour colon transit
colon segment
5
m nt
4
3
2
o
1
GC
GC
24h postprandial
GC
48h postprandial
p=0.07
3
6
0
ketotifen 4mg
e
4
A
2
1
0
-1
GC
GC
48h postprandial
placebo
ketotifen
4mg
ketotifen
12mg
Figure 3
Scintigraphical evaluation of intestinal transit of an 111In labeled non-caloric liquid testmeal 24hrs
after surgery. a) colonic transit time, depicted as median shift of geometrical center (GC) of colonic
contents in number of (predefined) segments per 24hrs; b) individual GC’s of colon transit 24 and c)
48hrs postprandial for each consecutive treatment group. Note the smaller distribution in the ketotifen
treated group 24hrs postprandial, not present at 48hrs postprandial. d) 24 to 48hr GC-shift calculated
to correct for potential study drug related influences on (small) intestinal motility. Dottedline indicates
median.
206
2. Colonic transit
Colonic transit was determined 48 and 72 hrs after surgery, or 24 and 48 hrs after ingestion
of the radio-labeled tap water. As depicted in figure 3, colonic transit varied significantly
in patients treated with placebo; 24 hrs after intake the GC was still located in the small
intestine in one patient but had already moved towards the left colonic flexure in others. The
median GC was 1.9 (1.0-2.8). A similar distribution was observed in the group of patients
treated with 4 mg ketotifen. The median GC at t=24hrs after intake was 1.2 (1.0-2.4) (NS
compared to placebo). In contrast to placebo and 4 mg, the GC of patients treated with 12
mg ketotifen was less dispersed and varied mainly (except in one patient) between 1 and
2.5, with a median of 1.5 (1.3-2.4) (NS compared to placebo). These data indicate that most
of the radiolabeled material was located in the right colon. In this respect, it is important to
notice that patients were still on medication at this point, which was only discontinued at
the end of the day.
The next day, patients were off medication when the colonic transit was assessed at t=48hrs
after intake. In patients treated placebo, the GC had shifted more distally with 0.8 (0.0-1.1)
between the placebo and the two doses of ketotifen (fig.3a-c).
Based on the observation that the GC at t=24hrs in patients treated with 12 mg ketotifen
tended to be located more proximally, despite the improvement of gastric emptying, we
Chapter 9
unit towards 2.5(1.9- 4.0). The calculated GC of activity did not show a statistical difference
hypothesized that the highest dose of ketotifen might delay intestinal/colonic transit. To
and t=48hrs. In the placebo treated group this GC shift over 24hrs was 0.8 (0.0-1.1)
segments compared to 0.6 (0.0-1.2) and 1.2 (0.2-1.4) segments in the ketotifen 4 and
12mg treated groups respectively, showing a trend towards significance for the 12 mg
group (p=0.07 placebo vs. ketotifen 12mg) (fig.3d).
207
eliminate this possible confounding effect, we calculated the shift in GC between t=24hrs
Clinical evaluation
Table 3 depicts the outcome of clinical endpoints and marks the time interval between
the end of the surgical procedure and the occurrence of the indicated event. None of the
clinical endpoints of gastrointestinal recovery were significantly improved after surgery.
VAS scores for pain, nausea or abdominal cramping are plotted in fig. 4. The area under the
curves (AUC) were calculated showing a significant improvement for abdominal cramping
in the 12mg dose (median AUC for placebo 10.4 (3.2-18-9); ketotifen 4mg 6.4 (0.0-13.0),
p= ns; ketotifen 12mg 4.6 (0.5-8.4), p= 0.03) but not for pain (AUC for placebo 6.9 (3.310.6) vs. ketotifen 4mg 8.0 (7.2-12.5), p= ns and 12mg (2.4 (0.0-18.8), p= ns) and nausea
(AUC for placebo 3.0 (0.6-6.9) vs. ketotifen 4mg 7.0 (0.2-15.0), p= ns and 12mg 1.7 (0.013.0), p= ns).
10
9
8
7
6
5
4
3
2
1
0
nausea
placebo
ketotifen 4mg
ketotifen 12mg
VAS-score
VAS-score
pain
ns
0
1
2
3
day post-surgery
4
5
10
9
8
7
6
5
4
3
2
1
0
placebo
ketotifen 4mg
ketotifen 12mg
0
1
2
3
day post-surgery
4
5
ns
VAS-score
abdominal cramping
10
9
8
7
6
5
4
3
2
1
0
placebo
ketotifen 4mg
ketotifen 12mg
p=0.03
0
1
2
3
day post-surgery
4
5
208
Figure 4
Clinical evaluation of pain, nausea and abdominal cramping over the first 5 postoperative days
(i.e. median time till ready for discharge). Note
that median time of epidural/i.v. analgesia is 3
days.
209
119 (94-125)
time until ready for hospital discharge
119 (105-147)
96 (72-132)
118 (108-120)
43 (20-74)
ketotifen 4mg
p-value
ns
ns
ns
ns
vs. placebo
119 (96-130)
72 (72-96)
114 (98-125)
30 (19-51)
ketotifen 12mg
Chapter 9
median time till event occurred in hrs after surgery (inter-quartile range)
96 (72-108)
time until solid food intake
table 3
114 (91-125)
23 (20-36)
time until first flatus
time until first bowel movement
placebo
endpoint
clinical (secondary) endpoints
ns
ns
ns
ns
p-value
vs. placebo
Discussion
Postoperative ileus is an iatrogenic disorder characterized by impaired and disturbed
motility of the entire gastrointestinal tract. Spontaneous recovery of intestinal transit or
coordinated motility is initiated first in the small intestine, approximately 24hrs after surgery,
but it may last up to 3 to 5 days before gastric and colonic function have returned to normal25.
One approach to enhance this process is stimulation of gastrointestinal motility with
potent prokinetics, of which cisapride is the most studied drug. Cisapride administered i.v.
induced a significantly faster propagation of radiopaque markers in the colon accompanied
by a significantly earlier first bowel movement in patients undergoing cholecystectomy26.
Most studies however fail to demonstrate clinical improvement15. Treatment with 3 times
30 mg rectal cisapride induced some changes in motor activity but did not enhance the
recovery to normal motility or clinical outcome in patients who underwent major intraabdominal surgery27. Similarly, Hallerback et al. failed to demonstrate changes in time to
first bowel movement after upper gastrointestinal or colonic surgery by rectal administration
of cisapride10. One might argue that the absence or moderate effect of prokinetics could
result from the fact that the underlying cause of postoperative ileus, i.e. increased inhibitory
neural input to the gastrointestinal tract, has not been targeted. Especially as recent findings
indicate that inflammation induced by handling of the intestine continuously drives this
inhibitory input16, 28, prevention of this inflammatory response could embody an alternative
therapeutic approach.
Previously, we demonstrated that mast cells play an important role in the development
of the inflammatory response to intestinal handling20. Animals lacking mast cells do not
develop intestinal inflammation after surgery, whereas treatment with mast cell stabilizers
block the occurrence of handling-induced inflammation in wild type animals. Conversely,
mast cell degranulation with compound 48/80 induces a local inflammatory response in the
exposed intestinal loop inducing delayed gastric emptying20. In line with these animal data,
we recently showed in man that a conventional open hysterectomy, but not a laparoscopic
adnexectomy or transvaginal hysterectomy results in the release of the mast cell mediator
tryptase in peritoneal lavage fluid and triggers the influx of leukocytes in the intestinal
muscularis18. Although there is abundant evidence in animals that prevention of surgery-
210
induced intestinal inflammation shortens postoperative ileus and is an effective alternative
treatment, human studies supporting this principle are still lacking. Therefore, we designed
the current pilot study investigating the effect of ketotifen, a mast cell stabilizer used in
allergic disorders such as hay fever, on postoperative gastric emptying of liquids. This
parameter was chosen as primary outcome parameter as it parallels our animal model, and
is a reliable and reproducible read-out of gastric motility with accepted clinical relevance.
Patients ingested 100 mL of radiolabeled tap water 24 hrs after the surgical procedure and
gastric retention was determined by scintigraphic imaging one hr later. In patients treated
with placebo or 4 mg ketotifen, gastric emptying varied considerable: some patients had
emptied their stomach almost completely whereas others had a severe gastric stasis with
more than 90% of the radiolabel still present in the stomach. In contrast, gastric emptying
of patients treated with 12 mg ketotifen was significantly faster with almost complete
emptying in 15 of the 17 patients. These findings show that mast cell stabilization restores
gastric emptying after abdominal surgery and provide indirect support for the concept that
intervention with the mast cell – inflammation cascade may represent a new therapeutic
approach for postoperative ileus. It should be emphasized though that a direct prokinetic
ketotifen or mast cell stabilizers on gastric emptying are available. In rats however, the
mast cell stabilizers disodium cromoglycate and FPL-52694 significantly inhibited gastric
motor activity indirectly arguing against this possibility29.
Chapter 9
effect of ketotifen can not be excluded, especially as no human data on the effect of
In addition to gastric emptying, we also monitored clinical parameters and colonic transit,
as secondary outcome parameter of this pilot study. No significant effect of ketotifen was
detected on colonic transit. As shown in figure 2, the variation in colonic transit in the placebo
group was very large, implying that this negative finding might represent a type II error. In
fact, the same applies for the effect on symptoms and clinical recovery. Only abdominal
cramping was reduced by ketotifen 12 mg, whereas nausea, vomiting, pain and clinical
recovery parameters remained unaltered. Alternatively, animal data indicate that ketotifen
inhibits colonic motility possibly obscuring the beneficial effect of interference with the mast
cell-inflammation cascade. Ketotifen indeed has mild anti-cholinergic properties23, but also
relaxes the mouse colon and inhibits small intestinal contractions evoked by carbachol and
nerve stimulation30. In this respect, it should be emphasized that patients were still treated
211
measured by means of the GC 24 and 48 hrs after ingestion of radiolabeled tab water,
with ketotifen when colonic transit at t=24 hrs was determined. The observation that the GC
of all but one patient treated with 12 mg ketotifen was situated in the right colon at t=24 hrs,
whereas it had moved up to the left colonic flexure in some patients treated with placebo,
indirectly supports this possibility. Moreover, when ketotifen was stopped at the end of
postoperative day 2, the transit of the GC between t=24 and t=48 hrs indeed tended to be
faster in the ketotifen group compared to the placebo group. These considerations would
imply that ketotifen treatment in future studies must be stopped immediately after surgery,
similar to the treatment regimen used in our animal study20.
Although this study clearly opens perspective for future treatment of postoperative ileus,
there are some drawbacks that need to be considered. First, there were a relatively large
number of dropouts, mainly due to protocol violation in the initial phase of the study. This
was caused by the administration of drugs on the ward that were not allowed according
to the study protocol, defined as exclusion criteria. However, the number of dropouts is
comparable in all three patient groups making it rather unlikely that this will affect the
outcome of the study. Second, a large variation was observed in gastric emptying and
colon transit. Especially gastric emptying for liquids was almost completed in a significant
proportion of patients, even after placebo treatment. Emptying of a solid caloric testmeal would have been more appropriate and might have yielded more consistent results,
perhaps showing an even greater difference between the treatment arms. However as
data on early postoperative gastric emptying in patients are lacking, manly for safety
reasons, emptying of liquids was evaluated in stead. A third point is the lack of baseline
gastrointestinal transit measurements. This design would have allowed comparison before
and after surgery in each individual patient, reducing variability and increasing the power of
the study. Nevertheless, even with the current design our study showed a dose-dependent
effect of ketotifen on gastric emptying.
Our study may be of great clinical relevance, as it partly confirms our animal data
demonstrating that mast cell stabilization restores surgery-induced delayed gastric
emptying. Based on our animal research data and our previous findings in patients, this
effect most likely result’s from blockade of the inhibitory neural input to the stomach driven
by intestinal inflammation. If this concept indeed proves to be important in the pathogenesis
of postoperative ileus in human, the treatment of this iatrogenic disorder will change
212
dramatically in the near future. There are however important issues that still need to be
addressed or require improvement. It remains to be studied if the observed effect on gastric
emptying indeed results from blockade of intestinal inflammation. Secondly, future studies
are required to demonstrate whether results can be further improved, i.e. improvement of
colonic transit and clinical recovery. This may be achieved by changing the concentration
or route of administration for ketotifen. Higher dosages injected intravenously before and
during surgery or even lavage of the abdominal cavity with ketotifen could be alternative
treatment protocols, avoiding the possible inhibitory effect of ketotifen on gastrointestinal
motility. Nevertheless, we feel that our observation is an important step forward encouraging
Chapter 9
larger clinical studies with ketotifen or other more potent mast cell stabilizers.
213
Reference List
major elective surgery. AnnSurg 1999;230(2):251-9.
2. Prasad M, Matthews JB. Deflating postoperative ileus. Gastroenterology 1999;117(2):489-92.
3. Boeckxstaens GE, Hirsch DP, Kodde A, et al. Activation of an adrenergic and vagally-mediated NANC pathway in surgery-induced fundic relaxation in the rat. NeurogastroenterolMotil
1999;11(6):467-74.
Effect of adrenergic and nitrergic blockade on experimental ileus in rats. BrJPharmacol
1997;120(3):464-8.
5. De Winter BY, Boeckxstaens GE, De Man JG, et al. Effect of different prokinetic agents and a
novel enterokinetic agent on postoperative ileus in rats. Gut 1999;45(5):713-8.
6. Seta ML, Kale-Pradhan PB. Efficacy of metoclopramide in postoperative ileus after exploratory
laparotomy. Pharmacotherapy 2001;21(10):1181-6.
7. Cheape JD, Wexner SD, James K, Jagelman DG. Does metoclopramide reduce the
length of ileus after colorectal surgery? A prospective randomized trial. DisColon Rectum
1991;34(6):437-41.
8. Jepsen S, Klaerke A, Nielsen PH, Simonsen O. Negative effect of Metoclopramide in
postoperative adynamic ileus. A prospective, randomized, double blind study. BrJSurg
1986;73(4):290-1.
9. Brown TA, McDonald J, Williard W. A prospective, randomized, double-blinded, placebo-controlled trial of cisapride after colorectal surgery. AmJSurg 1999;177(5):399-401.
10. Hallerback B, Bergman B, Bong H, et al. Cisapride in the treatment of postoperative ileus.
AlimentPharmacolTher 1991;5(5):503-11.
11. Boghaert A, Haesaert G, Mourisse P, Verlinden M. Placebo-controlled trial of cisapride in postoperative ileus. Acta AnaesthesiolBelg 1987;38(3):195-9.
12. Smith AJ, Nissan A, Lanouette NM, et al. Prokinetic effect of erythromycin after colorectal surgery: randomized, placebo-controlled, double-blind study. DisColon Rectum 2000;43(3):333-7.
13. Bonacini M, Quiason S, Reynolds M, Gaddis M, Pemberton B, Smith O. Effect of intravenous
erythromycin on postoperative ileus. AmJGastroenterol 1993;88(2):208-11.
14. Bungard TJ, Kale-Pradhan PB. Prokinetic agents for the treatment of postoperative ileus in
adults: a review of the literature. Pharmacotherapy 1999;19(4):416-23.
15. Holte K, Kehlet H. Postoperative ileus: a preventable event. BrJSurg 2000;87(11):1480-93.
16. de Jonge WJ, van den Wijngaard RM, The FO, et al. Postoperative ileus is maintained by
2003;125(4):1137-47.
17. The FO, de Jonge WJ, Bennink RJ, van den Wijngaard RM, Boeckxstaens GE. The ICAM-1
antisense oligonucleotide ISIS-3082 prevents the development of postoperative ileus in mice.
BrJPharmacol 2005.
18. The FO, Bennink RJ, Ankum WM, et al. Intestinal handling induced mast cell activation and
inflammation in human post-operative ileus. Gut 2007.
19. Kalff JC, Turler A, Schwarz NT, et al. Intra-abdominal activation of a local inflammatory response within the human muscularis externa during laparotomy. AnnSurg 2003;237(3):301-15.
20. de Jonge WJ, The FO, van der CD, et al. Mast cell degranulation during abdominal surgery
initiates postoperative ileus in mice. Gastroenterology 2004;127(2):535-45.
21. Saklad M. Grading of patients for surgical procedures. Anesthesiology 1941;2:4.
22. Kamm MA. The small intestine and colon: scintigraphic quantitation of motility in health and
disease. EurJNuclMed 1992;19(10):902-12.
23. Grant SM, Goa KL, Fitton A, Sorkin EM. Ketotifen. A review of its pharmacodynamic and
pharmacokinetic properties, and therapeutic use in asthma and allergic disorders. Drugs
1990;40(3):412-48.
214
Chapter 9
24. Bennink R, Peeters M, Van dMV, et al. Evaluation of small-bowel transit for solid and
liquid test meal in healthy men and women. EurJNuclMed 1999;26(12):1560-6.
25. Miedema BW, Johnson JO. Methods for decreasing postoperative gut dysmotility. Lancet Oncol 2003;4(6):365-72.
26. Tollesson PO, Cassuto J, Rimback G, Faxen A, Bergman L, Mattsson E. Treatment of
postoperative paralytic ileus with cisapride. Scandinavian journal of gastroenterology
1991;26(5):477-82.
27. Benson MJ, Roberts JP, Wingate DL, et al. Small bowel motility following major intraabdominal surgery: the effects of opiates and rectal cisapride. Gastroenterology
1994;106(4):924-36.
28. Kalff JC, Carlos TM, Schraut WH, Billiar TR, Simmons RL, Bauer AJ. Surgically induced
leukocytic infiltrates within the rat intestinal muscularis mediate postoperative ileus.
Gastroenterology 1999;117(2):378-87.
29. Takeuchi K, Nishiwaki H, Okabe S. Cytoprotective action of mast cell stabilizers
against ethanol-induced gastric lesions in rats. Japanese journal of pharmacology
1986;42(2):297-307.
30. Abu-Dalu R, Zhang JM, Hanani M. The actions of ketotifen on intestinal smooth
muscles. EurJPharmacol 1996;309(2):189-93.
215
01
10
Chapter
10
Summary and conclusions
M
Summary and conclusions
More than a millennium after its first written documentation, postoperative ileus still is a
prevalent clinical condition with significant morbidity and socio-economical impact1, 2. Even
to date, every patient undergoing abdominal surgery remains hospitalized for several
days as he or she will experience a period of nausea, lack of appetite and inability to eat
or defecate3. Both studies in animal models4-7 and in man1, 8 reveal that these symptoms
result from absence or disturbed gastrointestinal motility and failed propulsion of intestinal
contents, known as postoperative ileus. Mainly pain stimuli during surgery, induced by
skin incision, opening of the peritoneum, but above all handling of the intestine, have
been identified as a major cause of this instantaneous “paralysis” of the intestines4, 9, 10.
Until recently, it was generally accepted that activation of inhibitory neural pathways by
nociceptive / mechanical stimuli explained the generalized impairment of gastrointestinal
motility. Pharmacological neural blockade, section of nerves or the spinal cord, afferent
nerve ablation with capsaicin and identification of activated nerve pathways and brain
nuclei all confirmed this hypothesis4, 5, 7, 11, 12. Most experiments however were performed
during surgery or evaluated gastrointestinal function within a time frame of up to 3 hours
after surgery. By now, we know that this period reflects the first early phase of postoperative
ileus and only represents the tip of the iceberg. In this thesis, we indeed describe that this
early phase is followed by a second prolonged phase triggered by inflammation of the
handled intestine.
During each surgical procedure in the peritoneal cavity, the intestines will have to be replaced
from their original location, either to reach the organ of interest, to inspect the intestine for
abnormalities or to isolate the diseased segment for resection. Although the exact initial
trigger remained unclear, it became clear that the extent of intestinal handling might be
one of the major determinants of the severity of postoperative ileus. This observation
undoubtedly has been an important stimulus for the development and the introduction of
minimal invasive surgery, associated with a significant reduction in postoperative ileus and
duration of hospitalization13-15. It was not until a few years ago that we began to understand
how intestinal handling could lead to prolonged inhibition of intestinal neuromuscular
function. Kalff and coworkers demonstrated that 3 to 4 hours after mechanical manipulation
of the intestine, the muscularis became infiltrated by inflammatory cells16. In rodents, this
218
phenomenon led to impaired transit of intestinal content and reduced in vitro contractile
activity of inflamed muscle strips for more than 24 hours17, this thesis. This finding has led
to the “inflammatory” hypothesis suggesting that the prolonged duration of postoperative
ileus rather results from inflammation-induced impairment of gastrointestinal motility,
and not from activation of inhibitory neural pathways. Based on these observations, the
pathophysiology of postoperative ileus is now subdivided in 2 phases; an instantaneous
short lasting neurogenic phase resulting from activation of nociceptive neural pathways
during surgery, and a second late-onset (after 3-4 hours) inflammatory phase. Given its
duration, the second phase is obviously the most important one, at least from a clinical
point of view. In the current thesis, we have been focusing on this second phase and
have tried to unravel the cells and mechanisms involved in its pathophysiology in order to
develop more efficient therapeutic strategies to shorten postoperative ileus.
Although the inflammatory theory certainly explains the prolonged nature of postoperative
ileus or in other words, explains how gastrointestinal motility remains disrupted even
though the initial surgical stimulus has ceased, it fails to explain the generalized nature
of postoperative ileus. Indeed, if we accept that intestinal handling leads to inflammation
of the handled segment, then how does this explain dysfunction of those areas of the
intestine that have not been handled during surgery? One explanation could be that surgery
triggers a systemic inflammatory response via for instance the release of pro-inflammatory
the small intestine, we observed inflammation of the manipulated segments, but not in other,
non-handled, areas of the gastrointestinal tract. To investigate the mechanisms leading to
the generalized impairment of gastrointestinal motility, we developed a mouse model in
which gastric emptying was used a read-out to determine the degree of postoperative ileus
Chapter 10
mediators / cytokines in the systemic circulation. However, 24 hours after manipulation of
(chapter 2). The small intestine was gently manipulated during 5 minutes after which the
abdomen was closed and the animals were allowed to recover. Twenty-four hours later,
that underwent intestinal manipulation during a laparotomy revealed delayed gastric
emptying, whereas those that underwent a laparotomy only had normal transit (chapter
2). Similarly, inflammation of the muscularis was only observed in animals that underwent
abdominal surgery (intestinal manipulation) (chapter 2). Most importantly, the inflammation
was limited to the handled region, i.e. the small intestine, but was absent in the stomach.
219
a radio-labeled meal was gavaged and gastric emptying was determined. Only animals
Nevertheless, gastric emptying was significantly delayed in our model up to 48 hours after
surgery (chapter 2). Moreover, we showed that prevention of influx of inflammatory cells by
several interventions, like antibodies or antisense olignucleotides to the adhesion molecule
ICAM-1, restored gastric emptying, indicating that this infiltrate was indeed responsible for
the observed ileus (chapter 3).
How then can local inflammation of the small intestine lead to delayed gastric emptying?
We hypothesized that activation of neural pathways by the infiltrate must be involved. To
confirm this concept, animals were pretreated with the ganglion blockers guanethidine
and hexamethonium to inhibit neurotransmission (chapter 2). These experiments
indeed showed that gastric emptying was normalized by these agents, even though the
inflammation in the small intestine was still present. To further prove that the local infiltrate
activated inhibitory neural pathways, we stained the spinal cord for c-fos expression, a
marker of neural activation (chapter 2). Animals that underwent abdominal surgery, but not
those subjected to a laparotomy only, showed c-fos expression in the spinal cord. When
manipulation-induced inflammation was blocked by pretreatment with adhesion molecule
neutralizing antibodies, the increase in c-fos expression was prevented, clearly confirming
our hypothesis (chapter 2). From these experiments, we concluded that the prolonged late
phase of postoperative ileus results from neuro-immune interaction between inflammatory
cells in the manipulated segment and its afferent innervation. This interaction activates
an inhibitory adrenergic neural pathway synapsing in the spinal cord, affecting the entire
gastrointestinal tract.
The next crucial question that arose was how handling of the intestine triggers the influx of
inflammatory cells. Obviously, tissue damage will attract immune cells and will contribute to
the local inflammatory process. However, we reasoned that intense activation of nociceptive
nerve fibers would play a more important role, especially as earlier studies showed that
intense activation of afferent nerve fibers is an important trigger for local inflammation, also
referred to as neurogenic inflammation18, 19. Afferent nerve fibers, when intensely activated,
release neuropeptides like Calcitonine Gene Related Peptide (CGRP) and substance P
at the site of stimulation19. These peptides are potent pro-inflammatory mediators, mainly
by their capacity to stimulate mast cells18, 20. Mediators released by mast cells will not only
directly attract inflammatory cells, but will also lead to the transient increase in mucosal
220
permeability21 and bacterial translocation previously described after intestinal handling22.
Schwarz et al. indeed elegantly demonstrated that intestinal manipulation leads to influx
of intraluminal micro-spheres during a time window of 4 hours after manipulation22. This
transient disruption of the intestinal barrier allows intraluminal bacteria to enter the intestinal
wall, an important trigger to activate the immune system. Activation of resident macrophages,
located in between the longitudinal and circular muscle layer23, has been described to
occur a few hours after intestinal manipulation, as shown by upregulation of IL-6, iNOS
and LFA-117, 24, 25. In chapter 7, we confirmed the important role of mast cells and showed
that manipulation-induced inflammation and ileus were reduced in animals pre-treated with
mast cell stabilizers and in animals lacking mast cells (W/Wv mice). Reconstitution of mast
cells in W/Wv mice restored the capacity to develop an inflammatory response following
intestinal manipulation. Also in man, we demonstrated the release of mast cell mediators
in the peritoneal cavity, even after gentle inspection of the intestinal at the beginning of
the surgical procedure (chapter 8). In line with our animal findings, mast cell activation
was followed by the upregulation and release of inflammatory mediators such as IL-6,
IL-8, iNOS and ICAM-1. This process ultimately led to the influx of inflammatory cells in
to the muscularis propria of the resected intestinal tissue specimen at the end of surgery.
Interestingly, this cascade of events occurred almost exclusively in patients who underwent
conventional (open) surgery, but not in patients who underwent minimal invasive surgery.
Using radio-labelled leukocyte SPECT scanning, an increase in influx of leukocytes 24
after open, but not after laparoscopic abdominal surgery in patients (chapter 8). These
findings demonstrate both in mice and man that mast cell activation triggered via surgical
bowel manipulation represents an important initial step in the cascade of events leading to
intestinal inflammation and postoperative ileus.
Chapter 10
hr after surgery (compared to baseline pre-operative scanning) was only demonstrated
As shown in the Summarizing Figure, increased permeability induced by mast cell activation
macrophages after intestinal handling, described earlier by Kalff et al.24. Interestingly,
Borovikova et al. reported dampening of macrophage activation by the vagus nerve26. In
a model of sepsis, these investigators demonstrated increased survival and improvement
of blood pressure after LPS infusion when the vagus nerve was electrically stimulated.
Acetylcholine, released by the vagus nerve, was shown to interact with alpha7 nicotinic
221
leads to bacterial translocation, most likely contributing to the activation of resident
222
receptors on macrophages, resulting in a reduction in the release of the pro-inflammatory
cytokines TNF-alpha27. Especially as the gastrointestinal tract is largely under control of
the vagus nerve, we investigated whether electrical nerve stimulation could also intervene
with the activation of the resident macrophages thereby reducing the inflammatory
response and ileus following abdominal surgery (chapter 4). Indeed, electrical vagus
nerve stimulation diminished intra-peritoneal release of TNF, MIP-2 and IL-6 three hrs after
surgery in our ileus model, indicating a reduction of the activation of macrophages during
surgery. Accordingly, the number of leukocytes recruited to the intestinal muscle layer was
significantly reduced 24 hrs later, associated with a normal gastric emptying rate (chapter
4). To further confirm the anti-inflammatory properties of the vagus nerve, experiments
were performed with CNI 1493, a p38 MAPKinase inhibitor shown to reduce inflammatory
responses in a vagus nerve dependent manner when injected i.c.v.28, 29. Like electrical
nerve stimulation, this intervention reduced inflammation and restored gastric emptying,
an effect abolished by vagotomy (chapter 6). This set of experiments indicates that also
in the gastrointestinal tract, the vagus nerve exerts an important anti-inflammatory input
contributing to the control of the innate immune response. To elucidate how acetylcholine
exerts its anti-inflammatory effect on macrophages, peritoneal macrophages were isolated
and activated in vitro with LPS in the presents of nicotine. This agonist indeed reduced the
release of TNF, IL-6 and MIP-2 via its alpha7 acetylcholine receptor subtype (chapter 4).
We identified the signal transduction pathway mediating the inhibitory effect of nicotine and
mediators (chapter 4). The importance of this signaling cascade is illustrated by the fact
that manipulation induced inflammation cannot be reduced through vagus nerve stimulation
in STAT3 conditional knock-out mice (chapter 4).
Summarizing figure: (see fullcolor chapter 11)
Summary of the pathophysiology of postoperative ileus. The inevitable handling of the intestines during abdominal surgery (A) results in the instant activation and degranulation of mast cells (B). The
latter leads to transient intestinal barrier dysfunction enabling luminal bacteria to enter the intestinal
wall (C). A network of macrophages residing between the circular and longitudinal muscle layer probable phagocytize these bacteria and become activated (D). These events result in upregulation of the
adhesion molecules ICAM-1 and LFA-1 and recruitment of leukocytes from the circulation in to the
intestinal muscle layer (E). This local inflammation then activates inhibitory neural pathways explaining the sustained general inhibition of gastrointestinal motility during postoperative ileus (F). This
cascade identifies several new targets for therapeutic strategies to shorten or prevent postoperative
ileus (indicated in rectangles on right).
223
subsequent phosphorylation of Jak2 and STAT3 decreasing the release of inflammatory
Chapter 10
demonstrated that activation of the alpha7 receptor subtype on macrophages results in the
This so-called cholinergic anti-inflammatory pathway is suggested to represent an
additional system controlling the inflammatory response to a wide range of threats to the
organism30. Inflammation is sensed by afferent nerve fibers and is subsequently relayed to
the brain. After integration of afferent information, the motor neurons of the vagus nerve are
activated and an integrated anti-inflammatory signal is sent back to the inflamed area. Still,
the presence of such a feedback loop (i.e. reflex) and its anatomical connections clearly
need to be demonstrated, and is currently being investigated. Nevertheless, this system
may represent an interesting new tool to contain undesired inflammatory processes. In
contrast to anti-inflammatory cytokines and the hormonal control by corticosteroids (HPA
axis), this neural system provides an integrated response that is lightning fast and localized.
Obviously, it may provide new therapeutic targets to control or dampen inflammation, not
only in case of sepsis or ileus, but most likely also in other inflammatory disorders like
rheumatoid arthritis and inflammatory bowel diseases.
Therapeutically these finding might have great impact. As we repeatedly have demonstrated
the importance of the local inflammatory response in the pathogenesis of postoperative
ileus, any therapeutic intervention preventing its occurrence could be an interesting
approach to treat this disorder. In the first chapters, we showed that interference with
adhesion molecules, necessary for leukocytes to leave the circulation and enter the area
of manipulation, either with antibodies or antisense oligonucleotides, indeed shortened the
ileus. Alternatively, we showed that interference with the release of mast cell mediators,
one of the first events in the pathophysiological cascade, is effective in our mouse model
(chapter 7). Ketotifen and doxantrazole, agents known to stabilize mast cells, prevented
handling induced inflammation and indeed shortened postoperative ileus in mice (chaper
7). Based on these findings, we designed a pilot proof-of-principle clinical study evaluating
the effect of ketotifen versus placebo treatment on postoperative gastric emptying in a
series of gynecological patients (chapter 9). Interestingly, we demonstrated that similar to
our animal experiments, ketotifen reduced the delay in gastric emptying evoked by surgery.
Although a larger study with a different dosing scheme is certainly required, this study
confirms our hypothesis in man and suggests that more specific mast cell stabilizers may
represent an interesting new approach to shorten postoperative ileus. Finally, interventions
that activate the cholinergic anti-inflammatory pathway might embody an attractive therapy.
Vagus nerve stimulation can be obtained either by electrical stimulation or administration
224
of central application of drugs like CNI 1493, as shown in chapters 4 and 6. A much more
interesting approach would be to activate the vagus nerve by more physiological stimuli,
such as for example feeding. Recently, an interesting study was reported showing that a
meal containing high concentration of long-chain fatty acids activates vagal afferents via
endogenous cholecystokinin release31. In a model of hemorrhagic shock, feeding reduced
the production of TNF and the degree of inflammation and prevented the increase in
mucosal permeability. Based in these observations, we will study the potential beneficial
effect of early feeding of a high fat meal in the peri-operative period as potential treatment
of postoperative ileus. Finally, we showed that acetylcholine released by the vagus nerve
dampens the cytokine production of macrophages via binding to the alpha7 nicotinic
receptor (chapter 5). Drugs interacting with this receptor will mimic the effect of vagus
nerve stimulation and are theoretically interesting agents with potential anti-inflammatory
properties. Treatment with AR-R17779, a specific agonist to the alpha7 nicotinic receptor,
indeed prevented inflammation and shortened postoperative ileus in our mouse model
(chapter 5). Surprisingly though, the production of cytokines by macrophages in vitro was
only slightly reduced, in contrast to nicotine itself, indicating that other nicotinic receptors
and/or other cells may be involved explaining the in vivo effect. Nevertheless, clinical
studies evaluating the efficacy of alpha7 nicotinic receptor agonists are certainly warranted
and will be studied in the near future.
into the pathogenesis of prolonged postoperative ileus and have identified new therapeutic
targets. Our work is indirectly also a plea for minimal invasive surgery as our data clearly
indicate that intestinal handling during surgery should be avoided as much as possible.
Chapter 10
In summary, the data presented in the current thesis have provided substantial new insight
225
Reference List
WG, Daley J, Khuri SF. Risk factors for morbidity and mortality after colectomy for colon cancer. Dis.Colon Rectum 2000;43:83-91.
1997;120:464-468.
6. De Winter BY, Boeckxstaens GE, De Man JG, Moreels TG, Schuurkes JA, Peeters TL, Herman
AG, Pelckmans PA. Effect of different prokinetic agents and a novel enterokinetic agent on
postoperative ileus in rats. Gut 1999;45:713-718.
8. Clevers GJ, Smout AJ. The natural course of postoperative ileus following abdominal surgery.
Neth J Surg 1989;41:97-9.
10. Holzer P, Lippe IT, Amann R. Participation of capsaicin-sensitive afferent neurons in gastric motor inhibition caused by laparotomy and intraperitoneal acid. Neuroscience 1992;48:715-22.
J.Comp Neurol. 1994;349:212-222.
13. Schwenk W, Haase O, Neudecker J, Muller JM. Short term benefits for laparoscopic colorectal
resection. Cochrane.Database.Syst.Rev. 2005:CD003145.
14. Chen HH, Wexner SD, Iroatulam AJ, Pikarsky AJ, Alabaz O, Nogueras JJ, Nessim A, Weiss
EG. Laparoscopic colectomy compares favorably with colectomy by laparotomy for reduction of
postoperative ileus. Dis.Colon Rectum 2000;43:61-65.
15. Veldkamp R, Kuhry E, Hop WC, Jeekel J, Kazemier G, Bonjer HJ, Haglind E, Pahlman L,
Cuesta MA, Msika S, Morino M, Lacy AM. Laparoscopic surgery versus open surgery for colon
cancer: short-term outcomes of a randomised trial. Lancet Oncol. 2005;6:477-484.
1999;163:2410-2415.
226
Chapter 10
23. Mikkelsen HB, Mirsky R, Jessen KR, Thuneberg L. Macrophage-like cells in muscularis externa
of mouse small intestine: immunohistochemical localization of F4/80, M1/70, and Ia-antigen.
Cell Tissue Res. 1988;252:301-306.
2000;85:141-147.
J.Exp.Med. 2002;195:781-788.
30. Tracey KJ. The inflammatory reflex. Nature 2002;420:853-859.
31. Luyer MD, Greve JW, Hadfoune M, Jacobs JA, Dejong CH, Buurman WA. Nutritional stimulation of cholecystokinin receptors inhibits inflammation via the vagus nerve. J.Exp.Med.
2005;202:1023-1029.
227
11
11
Chapter
11
Samevatting en Conclusies
Dankwoord
Colour Figures
M
Samenvatting en conclusies
Meer dan een eeuw na de eerste beschrijvingen in de wetenschappelijke literatuur is
postoperatieve ileus nog steeds een frequent voorkomend medisch probleem met niet
onaanzienlijke morbiditeit en sociaal-economische consequenties1, 2. Tot op de dag van
vandaag blijft iedere patiënt die een buikoperatie ondergaat tot enkele dagen na de ingreep
in het ziekenhuis opgenomen in verband met klachten van misselijkheid, gebrek aan
eetlust, niet kunnen eten en het uitblijven van ontlasting3. Zowel dierexperimenteel-4-7 als
patiëntgebonden-onderzoek1, 8 hebben aangetoond dat deze symptomen, beter bekend als
postoperatieve ileus, het gevolg zijn van afwezige of gestoorde maag-, darm motoriek en
het onvermogen van de darmen om hun inhoud voort te stuwen. Voornamelijk pijnprikkels
gedurende de operatieve ingreep, opgewekt door de huid incisie, het openen van buikholte
maar voornamelijk door het aanraken van de darmen gedurende de procedure, blijken
belangrijke veroorzakers te zijn van deze instantane paralyse van het spijsverteringskanaal4,
9, 10
. Tot voor kort werd de activatie van inhiberende zenuwbanen als gevolg van activatie
van pijn- en mechano-sensoren, beschouwd als belangrijkste onderliggende oorzaak.
Farmacologische neuronale blokkade, klieving van spinale zenuwbanen, depletie van
afferente zenuwbanen met capsaicine en identificatie van de betrokken zenuwbanen
en hersenencentra, bevestigen allen deze hypothese4,
5, 7, 11, 12
. De meeste van deze
experimenten zijn echter tijdens de operatie uitgevoerd of hebben alleen het effect op de
maag- darm motoriek bestudeerd gedurende de 1e 3uur na de ingreep. Inmiddels weten
we echter dat deze “vroege fase” slechts een fractie van het klinische probleem is, vooral
omdat postoperatieve ileus veel langer aanhoudt en enkele dagen duurt. In dit proefschrift
beschrijven we inderdaad dat postoperative ileus vooral bepaald wordt door een latere en
langdurige fase die veroorzaakt wordt ontsteking van de darm.
Gedurende operatieve ingrepen in de buikholte is het hanteren van darmlissen onvermijdelijk,
of het nu is om het doelorgaan te bereiken of om de darm te inspecteren voor afwijkingen.
Alhoewel de exacte initiële prikkel onopgehelderd blijft, is het inmiddels duidelijk geworden
dat de mate van darmmanipulatie gedurende een operatie een belangrijke voorspeller kan
zijn voor de ernst van de postoperatieve ileus. Deze observatie heeft dan ook ongetwijfeld
bijgedragen aan de ontwikkeling van minimaal invasieve chirurgische technieken. Onderzoek
230
heeft aangetoond dat deze relatief nieuwe wijze van opereren inderdaad resulteert in
verkorting van de duur van postoperatieve ileus en ziekenhuis opname13-15. Het is echter
pas sinds enkele jaren dat we zijn gaan beginnen te begrijpen hoe darmmanipulatie kan
leiden tot aanhoudende neuronmusculaire dysfunctie. Kalff en collega’s hebben aangetoond
dat 3 tot 4 uur na het hanteren van dunne darm lissen (darmmanipulatie) de spierlaag in
de darmwand ontstoken raakt16. In knaagdieren leidt deze locale ontsteking tot vertraging
van de darmtransit (voortstuwen van darminhoud) en verminderde spierfunctie17. Deze
bevindingen hebben geresulteerd in de ontstekingshypothese die stelt dat de persisterende
fase van postoperatieve ileus veeleer het gevolg is van manipulatie geïnduceerde ontsteking
en als gevolg hiervan gestoorde maag-, darmmotoriek door activatie van inhibitoire
zenuwbanen. Op basis van deze observaties kunnen we de pathofysiologie nu indelen in
2 fasen; een acute, kortdurende, neurogene fase en een 2e late inflammatoire fase (vanaf
3 uur na chirurgie). Gezien de duur is deze 2e fase vanuit klinisch oogpunt veruit de meest
belangrijke van de twee. In dit proefschrift hebben we ons dan ook gericht op het verder
ontrafelen van het onderliggende cellulaire mechanisme met als doel de (preventieve)
behandeling van postoperatieve ileus te verbeteren.
Hoewel de inflammatoire theorie verklaart hoe de maag-, en darmmotoriek gestoord blijft
na beëindiging van de chirurgische ingreep, blijft het onduidelijk waarom de peristaltiek van
gans de gastrointestinale tractus verstoord is. Aannemende dat darmmanipulatie resulteert
in een locale ontsteking van de darmspierlaag blijft het onduidelijk hoe dit leidt tot gestoorde
propulsieve functie van die delen van het maagdarmstelsel die niet gemanipuleerd zijn.
dat 24 uur na abdominale chirurgie alleen die segmenten ontstoken zijn die gemanipuleerd
zijn geweest gedurende de procedure. Om het mechanisme verder te onderzoeken dat
leidt tot gegeneraliseerde verstoring van maag-, en darmmotoriek hebben we daarom een
muismodel ontwikkeld waarin 24uur na abdominale chirurgie de maagontledigingssnelheid
bepaald wordt als maat van gegeneraliseerde ileus (hoofdstuk 2). In deze experimenten
wordt in de ene groep de dunne darm gedurende 5 minuten voorzichtig gemanipuleerd
terwijl in de controle groep alleen de buikholte wordt geopend (laparotomie). Vervolgens
worden de darmlissen weer voorzichtig in de buikholte teruggeplaatst waarna de buikholte
wordt gesloten en 24 uur later wordt de maagontledigingssnelheid bepaald. De dieren die
231
Samenvatting en conclusies,
dankwoord en colour figures
gegeneraliseerde ontsteking. Echter experimenten in ons laboratorium hebben uitgewezen
Chapter 11
Eén verklaring zou kunnen zijn dat manipulatie gedurende de chirurgie resulteert in een
darmmanipulatie ondergingen tijdens de operatie toonden een significante vertraging van de
maagontlediging ten opzichte van controle dieren (hoofdstuk 2). Ook de ontstekingsreactie
in spierlaag van de darmwand was alleen aantoonbaar in muizen die darmmanipulatie
hadden ondergaan (hoofdstuk 2). Belangrijk hierbij is te vermelden dat deze ontsteking
alleen aanwezig was in die segmenten die waren gemanipuleerd tijdens de operatie.
Desalniettemin was tot 48 uur na de operatie de maaglediging significant vertraagd in
deze dieren (hoofdstuk 2). Daarnaast hebben we ook ontdekt dat het voorkomen van
de manipulatie gemedieerde ontsteking door middel behandeling met oa. antilichamen of
antisense oligonucleotiden gericht tegen het adhesie molecuul ICAM-1 (belangrijk bij de
rekrutering van ontstekingscellen vanuit de bloedsomloop) resulteert in de normalisatie
van maagontlediging (hoofdstuk 3).
Hoe kan locale ontsteking van de dunne darm leiden tot vertraging van de maagontlediging?
We veronderstelden dat neuronale reflexbanen geactiveerd raken door het
locale
ontstekingsinfiltraat. Om deze hypothese te onderzoeken werden dieren voorbehandeld
met de ganglionerge blokkers guanethidine en hexamethonium(hoofdstuk 2). De
resultaten van dit experiment toonde inderdaad een normalisering van de maagfunctie
terwijl de locale ontsteking in de dunne darm nog wel aanwezig was. Om de aanwezigheid
van een inhiberende zenuwreflex in de pathofysiologie van postoperatieve ileus verder
te onderzoeken hebben we vervolgens het ruggemerg gekleurd voor c-fos (een zenuw
activatie marker) (hoofdstuk 2). Dieren die darmmanipulatie hadden ondergaan maar
niet de controle muizen toonden c-fos expressie in het ruggenmerg, wat onze hypothese
andermaal bevestigde (hoofdstuk 2). Op basis van deze resultaten concluderen we
dat de aanhoudende (late) fase in postoperatieve ileus het gevolg is van neuro-immuun
interactie tussen de ontstekingscellen in de gemanipuleerde darmsegment en de afferente
(sensorische) innnervatie van het maag-, en darmstelsel.
Dit samenspel activeert
vervolgens adrenerge zenuwbanen die via het ruggemerg het gehele maag-, darmstelsel
negatief beïnvloeden.
Het volgende vraagstuk was op welke wijze darmmanipulatie leidt tot een locale
ontstekingsreactie met rekrutering van ontstekingscellen. Natuurlijk kan weefselschade
die het gevolg is van manipulatie zorgen voor de attractie van ontstekingscellen en dus
bijdragen tot de totstandkoming van een locale ontstekingsreactie. Onze gedachte was
232
echter dat sensorische zenuwvezels een belangrijkere rol zouden kunnen spelen, mede
gezien het feit dat eerdere studies hebben laten zien dat intensieve stimulatie van afferente
zenuwvezels een belangrijke prikkel vormen voor de ontwikkeling van locale inflammatie,
ook wel neurogene inflammatie genoemd18, 19. Afferente zenuwvezels stellen neuropeptiden
zoals Calcitonine Gene Related Peptide (CGRP) en substance P vrij wanneer ze intens
worden geactiveerd19. Deze eiwitten zijn potente pro-inflammatoire mediatoren, vooral
door hun vermogen om mestcellen te activeren18, 20. Het vrijstellen van mediatoren door
mestcellen heeft niet alleen een direct pro-inflammatoir effect maar resulteert ook in een
kortstondig verhoogde permeabiliteit (doorlaatbaarheid) van het darmslijmvlies (mucosa)21.
Dit laatste maakt het voor bacteriën mogelijk de darmwand te penetreren, een fenomeen
dat al eerder beschreven is ten gevolge van darmmanipulatie22. Schwarz et al. hebben
laten zien dat darmmanipulatie leidt tot de influx van luminale micropartikels naar de
darmwand ongeveer 4 uur na manipulatie22. Zoals gezegd kunnen bacteriën ten gevolge
van deze tijdelijke opening van de mucosale barrière, vanuit het darmlumen de darmwand
penetreren en vormen daar een belangrijke stimulus voor het immuunsysteem. Macrofagen
die als een soort netwerk van poortwachters tussen de longitudinale en circulaire spierlaag
van de darm liggen23, worden enkele uren na darmmanipulatie geactiveerd zoals onder
andere blijkt uit de toename van IL-6, iNOS en LFA-1concentraties17, 24, 25.
In hoofdstuk 7 hebben we de belangrijke rol die mestcellen vervullen in de pathofysiologie
van postoperatieve ileus aangetoond door aan te tonen dat manipulatie geïnduceerde
ontsteking en ileus gereduceerd zijn in muizen die voorbehandeld zijn met mestcel
van mestcel mediatoren in de buikholte aansluitend op subtiele darmmanipulatie kunnen
aantonen (hoofdstuk 8). Vergelijkbaar met onze observaties in muizen resulteert mestcel
activatie ook bij de mens in het vrijkomen of opreguleren van ontstekingsmediatoren zoals
IL-6, IL-8, iNOS en ICAM-1. Dit proces leidt uiteindelijk ook hier tot de rekrutering van
ontstekingscellen naar de spierlaag van de darm. Opvallend hierbij is overigens dat dit
proces nagenoeg alleen waarneembaar is in patiënten die een conventionele open buik
operatie (laparotomie) ondergaan en niet in patiënten die een minimaal invasieve ingreep
ondergingen. Ook visualisatie van ontstekingscelrekrutering 24 uur voor en na chirurgie,
middels het markeren van witte bloedcellen met een radioactieve merkstof (leukocyten
233
W/Wv muizen herstelt de ontstekingsrespons. Ook in patiënten hebben wij het vrijkomen
Chapter 11
stabilisatoren of die mestcel deficiënt zijn (W/Wv muizen). Mestcel reconstitutie in deze
SPECT scintigrafie), toonde vergelijkbare resultaten (hoofdstuk 8). Deze resultaten tonen
duidelijk aan dat zowel in proefdieren als in mensen, mestcel activatie ten gevolge van
chirurgische darmmanipulatie een belangrijke eerste stap vormt in de cascade die leidt tot
locale darmontsteking en postoperatieve ileus.
Zoals weergegeven in Summarizing figure (zie kleuren katern) gaat mestcel gemedieerde
toename van darm permeabiliteit gepaard met microbiële translocatie. Dit laatste fenomeen
is waarschijnlijk verantwoordelijk voor de eerder door Kalff et al. beschreven activatie
van het netwerk van macrofagen gelegen tussen de darmspierlagen24.
Interessant
hierbij is dat Borovikova et al. hebben aangetoond dat stimulatie van de nervus vagus
de activatie van macrofagen kan doen verminderen26. In een experimenteel sepsis model
hebben deze onderzoekers aangetoond dat elektrische stimulatie van de nervus vagus
een betere overleving en bloeddruk controle tonen na infusie van LPS. Acetylcholine, de
neurotransmitter vrijgesteld door de nervus vagus, bindt aan alfa7 nicotinerge receptor
op macrofagen27 met verminderde vrijstelling van pro-inflammatoire mediatoren zoals
TNF-alfa27. Aangezien het maag-, darmstelsel overwegend onder de controle staat van
de nervus vagus, hebben wij onderzocht of elektrische stimulatie van de nervus vagus
ook de activatie van macrofagen in de darmwand kan beïnvloeden om op deze wijze de
ontstekingreactie en ileus na darmmanipulatie te verminderen (hoofdstuk 4). Hieruit is
gebleken dat elektrische stimulatie van de vagus intra-peritoneale vrijstelling van TNF,
MIP-2 en IL-6 3 uur postoperatief inderdaad kan verminderen in ons model, een maat
voor verminderde macrofaag activatie. Bovendien worden er minder ontstekingscellen
gerekruteerd wat resulteerde in een normalisatie van de maagontlediging (hoofdstuk4). Om
de rol van dit anti-inflammatoire mechanisme verder te exploreren hebben we vervolgens
experimenten uitgevoerd met CNI-1493, een MAPKinase remmer die vagus afhankelijke
anti-inflammatoire eigenschappen heeft28, 29. Net als elektrische stimulatie van de nervus
vagus vermindert intraventriculaire toediening van CNI-1493 de ontsteking en verbetert het
de maagontledigingsfunctie, een effect dat te niet wordt gedaan door vagotomie (hoofdstuk
6). Deze serie van proeven heeft aangetoond dat de nervus vagus ook in het maag-,
darmstelsel een belangrijke regulatoire invloed heeft op het immuunsysteem. Vervolgens
hebben we in geïsoleerde macrofagen aangetoond dat acetycholine macrofaagactivatie
remt via de alfa7 nicotine receptor (hoofdstuk 4) en het Jak2/STAT3 signaleringspad. Het
belang van dit signaleringsmechanisme wordt benadrukt door het feit dat vagus stimulatie
234
de ontstekingsrespons niet kan onderdrukken in STAT3 geconditioneerde knock-out
muizen (Hoofdstuk 4).
Dit zogenaamde cholinerge anti-inflammatoire pad wordt beschouwd als een additioneel
regulatoir systeem van het immuunsysteem30. Hierin wordt de ontsteking gedetecteerd door
sensibele zenuwbanen en doorgegeven aan het brein. Na de verwerking van deze afferente
informatie worden de motorneuronen van de vagus geactiveerd en wordt er een geïntegreerd
anti-inflammatoir signaal teruggestuurd naar het ontstoken gebied. Echter het bestaan van
een dergelijk controle systeem (reflex) en betrokken anatomische verbindingen moeten
nog daadwerkelijk worden aangetoond. In tegenstelling tot anti-inflammatoire cytokinen en
hormonale regulatie middels corticosteroïden (via de HPA-as) zorgt dit neuronale syteem
voor een geïntegreerde respons die extreem snel en locatie specifiek is. Dit concept zal
ongetwijfeld resulteren in de ontwikkeling van nieuwe behandelstrategieën die niet alleen
toepasbaar zijn in sepsis of ileus maar ook in een scala aan andere ontsteking gerelateerde
aandoeningen.
De hier gepresenteerde resultaten hebben mogelijk belangrijke therapeutische gevolgen.
Gezien het feit dat we bij herhaling hebben laten zien dat de locale inflammatoire respons
belangrijk is in de pathogenese van postoperatieve ileus is iedere interventie die deze
respons kan voorkomen in beginsel een interessante therapeutische optie. In de eerste
hoofdstukken hebben we laten zien dat interventie op het niveau adhesie moleculen, nodig
bij de rekrutering van ontstekingscellen vanuit de bloedsomloop naar het ontstoken weefsel,
van mestcel degranulatie (vrijkomen van mestcel specifieke pro-inflammatoire eiwitten), een
van de eerste processen in de pathofysiologische cascade, een gunstig effect heeft op het
beloop van postoperatieve ileus (hoofdstuk 7). Ketotifen en doxantrazole, twee farmaca
die bekend staan als mestcel stabiliserende agens, voorkomen manipulatie gemedieerde
ontsteking en verkorten het beloop van postoperatieve ileus in muizen (hoofdstuk 7). Op
basis van deze resultaten hebben we een pilot-studie ontworpen waarin we het concept
van mestcel stabilisatie als behandeling voor postoperatieve ileus hebben onderzocht. In
deze studie hebben we op een dubbelblind gerandomiseerde wijze gekeken naar het effect
van ketotifen behandeling ten opzichte van placebo op de postoperatieve maagontlediging
235
ileus gunstig kunnen beïnvloeden. Daarnaast hebben we ook laten zien dat het voorkomen
Chapter 11
doormiddel van antilichamen of antisense oligonucleotiden het beloop van postoperatieve
in een gynaecologische patiëntenpopulatie (hoofdstuk 9). Naar analogie met ons
dierexperimenteel werk was de maagontlediging sneller na behandeling met ketotifen.
Alhoewel een grotere studie mogelijk met een ander doseringsschema nodig is, bevestigt
deze studie onze hypothese en suggereert dat meer specifieke mestcel stabilisatoren een
attractieve behandel optie zouden kunnen vormen om postoperatieve ileus te voorkomen.
Tot slot kunnen farmaca die het cholinerge anti-inflammatoire mechanisme activeren een
interessante benadering zijn om postoperatieve ileus te verkorten. Dit zou kunnen worden
bewerkstelligd door middel van elektrische stimulatie van de nervus vagus of toediening
van farmaca zoals CNI-1493 (beschreven in hoofdstuk 4 en 6). Een veel elegantere en
meer fysiologische methode van vagus activatie is voeding. Een interessante recente
publicatie heeft aangetoond dat voeding die hoge concentratie lange-keten vetzuren
bevat afferente vagale zenuwvezels activeert door middel van endogene cholecystokinine
vrijstelling31. In een model voor hemorhagische shock hebben deze auteurs aangetoond dat
voeding de productie van TNF vermindert, de ontstekingsreactie dempt en toename van de
darmpermeabiliteit voorkomt. Gebaseerd op deze gegevens willen wij het effect van vroege
voeding met vetrijke maaltijden in de peri-operatieve fase op het beloop van postoperatieve
ileus gaan bestuderen. In hoofdstuk 5 hebben we al laten zien dat acetylcholine, vrijgesteld
door de nervus vagus, de cytokine vrijstelling door macrofagen vermindert via alfa7 nicotine
receptor binding. Agonisten voor deze receptor bootsen het effect van vagusstimulatie na
en zijn in theorie potente anti-inflammatoire medicijnen. Behandeling met AR-R17779, een
specifieke alfa7 nicotine receptor agonist, vermindert inderdaad de inflammatoire respons
en verbetert de maagledigingsfunctie in ons postoperatieve ileus model (hoofdstuk 5).
Vreemd genoeg is het cytokine productie reducerende vermogen van dit middel in stimulatie
proeven slechts minimaal. Dit in tegenstelling tot het effect van nicotine wat suggereert dat
ander nicotine receptoren en/of celtypen betrokken zijn in dit proces. Desalniettmin zijn
klinische studies naar het effect van alfa7 nicotinerge agonisten gerechtvaardigd en zullen
zeker in de nabije toekomst worden uitgevoerd.
236
Samenvattend kunnen we stellen dat de gegevens gepresenteerd in dit proefschrift een
schat aan nieuwe inzichten heeft gegenereerd met betrekking tot de pathofysiologie van
postoperatieve ileus en hierbij meerdere nieuwe therapeutische targets heeft geidentificeerd.
Vooral ook omdat we duidelijk hebben aangetoond dat darmmanipulatie gedurende
heelkundige ingrepen zoveel mogelijk dient te worden vermeden, is dit proefschrift tevens
een indirect pleidooi voor de verdere ontwikkeling van minimaal invasieve chirurgische
Chapter 11
technieken.
237
Referentie Lijst
WG, Daley J, Khuri SF. Risk
��
factors for morbidity and mortality after colectomy for colon cancer. Dis.Colon Rectum 2000;43:83-91.
1997;120:464-468.
6. De Winter BY, Boeckxstaens GE, De Man JG, Moreels TG, Schuurkes JA, Peeters TL, Herman
AG, Pelckmans PA. Effect of different prokinetic agents and a novel enterokinetic agent on
postoperative ileus in rats. Gut 1999;45:713-718.
8. Clevers GJ, Smout AJ. The natural course of postoperative ileus following abdominal surgery.
Neth J Surg 1989;41:97-9.
10. Holzer P, Lippe IT, Amann R. Participation of capsaicin-sensitive afferent neurons in gastric motor inhibition caused by laparotomy and intraperitoneal acid. Neuroscience 1992;48:715-22.
J.Comp Neurol. 1994;349:212-222.
13. Schwenk W, Haase O, Neudecker J, Muller JM. Short term benefits for laparoscopic colorectal
resection. Cochrane.Database.Syst.Rev. 2005:CD003145.
14. Chen HH, Wexner SD, Iroatulam AJ, Pikarsky AJ, Alabaz O, Nogueras JJ, Nessim A, Weiss
EG. Laparoscopic colectomy compares favorably with colectomy by laparotomy for reduction of
postoperative ileus. Dis.Colon Rectum 2000;43:61-65.
15. Veldkamp R, Kuhry E, Hop WC, Jeekel J, Kazemier G, Bonjer HJ, Haglind E, Pahlman L,
Cuesta MA, Msika S, Morino M, Lacy AM. Laparoscopic surgery versus open surgery for colon
cancer: short-term outcomes of a randomised trial. Lancet Oncol. 2005;6:477-484.
1999;163:2410-2415.
238
Chapter 11
23. Mikkelsen HB, Mirsky R, Jessen KR, Thuneberg L. Macrophage-like cells in muscularis externa
of mouse small intestine: immunohistochemical localization of F4/80, M1/70, and Ia-antigen.
Cell Tissue Res. 1988;252:301-306.
2000;85:141-147.
J.Exp.Med. 2002;195:781-788.
30. Tracey KJ. The inflammatory reflex. Nature 2002;420:853-859.
31. Luyer MD, Greve JW, Hadfoune M, Jacobs JA, Dejong CH, Buurman WA. Nutritional stimulation of cholecystokinin receptors inhibits inflammation via the vagus nerve. J.Exp.Med.
2005;202:1023-1029.
239
H
Dankwoord
Het einde nadert. Een sentimenteel moment van overpeinzing maakt zich meester van
de auteur. Onvermijdelijk denkt hij op zo’n moment even terug aan wat hij nu eigelijk de
afgelopen jaren allemaal heeft uitgespookt.
Ik beschouw mijn onderzoeksperiode als één groot speelkwartier waarbij het AMC de
speeltuin was waar ik me als een kind in luilekkerland heb kunnen uitleven. Het is dan ook
met gemengde gevoelens dat ik een slotwoord op papier zet.
Dit wetenschappelijke avontuur zou ik nooit hebben kunnen volbrengen zonder de steun
van velen. Zij die al jaren deel uit maakten van mijn leven en mij hebben bijgestaan tijdens
de high’s en low’s die deze onderzoeksjaren met zich mee hebben gebracht. Velen van jullie
heb ik mogelijk tekort gedaan in mijn misschien wel bijna echocentrische preoccupatie met
dit boeiende maar soms ook zeer frustrerende werk. Ik ben jullie eeuwig dankbaar voor de
onvoorwaardelijke vriendschap die jullie mij hebben gegeven. Door het multidisciplinaire
karakter van het onderzoeksproject heb ik ook vele nieuwe mensen leren kennen. Ik
beschouw het als een groot voorrecht dat ik in de keuken van verscheidene disciplines en
instituten heb mogen snuffelen en ben dankbaar voor de gastvrijheid die men mij daarbij
geboden heeft. De inspiratie die het geeft om met mensen met verschillende expertise van
gedachte te wisselen over het onderzoek en meer..., heeft zeker bijgedragen aan het grote
plezier dat ik aan het doen van onderzoek heb beleeft.
Alhoewel ik het hier misschien wel het liefste bij zou willen laten, uit vrees in mijn
dankbetuiging te kort te schieten, ontkom ook ik er niet aan om een aantal mensen in
het bijzonder te noemen. Toch wil ik vanuit de grond van mijn hart hier alvast iedereen
bedanken die op welke wijze dan ook aan het tot stand komen van dit proefschrift heeft
bijgedragen!
Professor Dr. Boeckxstaens, beste Guy. Mijn wetenschappelijke mentor en spellingscontrole.
Jij hebt me wegwijs gemaakt in de wereld van neurogastroenterologie en wetenschap.
Ik beschouw het als een eer om onder jou te hebben mogen promoveren. Onder jouw
gedreven leiderschap heb ik een kleine eigenzinnige onderzoeksgroep zien uitgroeien
240
tot een autoriteit en heb getuigen mogen zijn van meerdere grootse wetenschappelijke
momenten waarvan velen alleen kunnen dromen. Ik dank je voor de mogelijkheden die je
me in de afgelopen jaren hebt geboden.
Dr. de Jonge, beste Wouter. Je bent op vele wijzen een onnavolgbaar voorbeeld voor me
geweest. Je ambitie en gedrevenheid zijn fenomenaal. Met veel bewondering heb ik vaak
gedwee aanschouwd hoe jij gedreven door je oprechte wetenschappelijke nieuwsgierigheid
met alles en iedereen een gesprek aanknoopte om iets te realiseren of om het naatje
van de kous, ten aanzien van een onderwerp, te weten te komen. Ondanks deze enorme
drive was er altijd tijd voor wat slap geouwehoer, een goed gesprek of flauwe “de Jonge”
grappen. Je enthousiasme is aanstekelijk en heeft me meer dan eens gemotiveerd. Waar
velen van je collega’s het lab verruilen voor een werkkamer ben jij niet uit het lab te slaan.
Tussendoor schrijf je daarnaast dan nog even de ene succesvolle subsidieaanvraag na
de andere, iets wat ik in je bewonder. Inmiddels heb je al je eigen onderzoeksgroep en is
je benoeming tot hoogleraar mijns inziens slechts een kwestie van tijd. Ik ben blij dat je
mijn co-promotor bent en hoop dat de toekomst ons weer samenbrengt om “belangrijke
enigmata” te ontrafelen.
Het motiliteitscentrum op C2, het kloppend hart. Lieve Aaltje, zonder jou was de ketotifentrial nooit wat geworden. Met veel plezier denk ik terug aan onze samenwerking en leuke
gesprekken. Je bent meer dan een fijne collega en ik hoop dat we snel weer eens tijd
kunnen vrijmaken om onder het genot van een hapje en drankje de wereldproblematiek
Andreas Ziekenhuis. Dank voor het bijbrengen van de fijne kneepjes van het ano-rectaal
functie onderzoek. Gelukkig zijn onze wegen na je vertrek al meer dan eens gekruist en
komen we elkaar zeker nog tegen. Bram, met jou heb ik lange tijd lief en leed gedeeld. Je
onuitputtelijke geduld en sociale instelling bewonder ik enorm. Dank voor je vriendschap
en de leuke tijd samen. Laten we snel het al lang geleden aan elkaar beloofde biertje
gaan drinken! Cynthia, helaas, maar niet onverwacht, heb je de motiliteit verruild voor een
nieuwe werkgever. Dank voor je ondersteuning en oplossend vermogen. Tamira, dank voor
de leuke gesprekken, het meedenken en het geven van je oprechte mening. Hanneke,
altijd in voor iets leuks. Dank voor de gezellige tijd samen, ik zal nooit vergeten hoe we
241
mijn promotie avontuur begonnen. Sjoerd, jammer genoeg ging jij al snel naar het Lucas
Chapter 11
door te nemen. Sjoerd en Bram, toen C2-310 nog een mannenkamer was... Met jullie is
samen op het Rembrandtplein lagen! Rene, amice! Rots in de branding, kamergenoot om
vijf voor twaalf en kritisch oor. De (ten onrechte) soms te stille (lees bescheiden) kracht van
de motiliteit. De leuke en inzichtelijke gesprekken met jou waren onbetaalbaar! Dennis,
dank voor je labsupport en soms bijna niet te volgen gevoel voor humor. Ik hoop dat je het
naar je zin hebt op je nieuwe werk. De vagus-girls, Esmerij en Susan. Jammer genoeg was
onze samenwerking relatief kort maar wel gezellig. Heel veel succes met jullie onderzoek,
dank! Ramona, Olaf (je hebt een moedige en goede beslissing genomen), Sjoerd B (the
next generation), Breg, Cathy en de poep-poli boys and girls (Mark, Fleur, Wieger, Maartje,
Michiel, Marloes, Noor en Olivia): allen dank voor de leuke tijd samen!
Dr. Bennink, beste Roel. Het was fantastisch om met je te mogen werken. Ik dank je voor
je onuitputtelijke vindingrijkheid, meedenkend vermogen en behulpzaamheid. Niets was
onmogelijk! Natuurlijk wil ik ook Formijn, Cynara, Jan, Ilse, Marsha en het hele team van de
Nucleaire Geneeskunde bedanken. Zonder jullie inzet (soms zelfs in het weekend!) waren
mijn experimenten en de klinische studies nooit het succes geworden wat het nu is!
Mijn steun en toeverlaat in het lab, Angelique! Bijna altijd goed gehumeurd en geïnteresseerd
in hoe het met je medemens gaat. We hebben elkaar leren kennen toen je, met veel
tegenzin, van G1 naar G2 moest verhuizen. Je hebt je ontpopt tot een top analist die
iedereen een helpende hand biedt. Inmiddels heb je een cardioloog aan de haak geslagen
en ben je moeder geworden van een lieve dochter: alle ingrediënten voor het geluk. Helaas
zien wij elkaar te weinig sinds ik het AMC verlaten heb. Dank voor je hulp en vriendschap.
Hoop snel weer eens bij Arko, Meike en jou te kunnen komen buurten.
Professor Buijs, beste Ruud, Jan en Caroline. Dank voor de goede en vooral ook gezellige
samenwerking op het NIH (tegenwoordig NIN). Jullie expertise was onmisbaar in de
“neuro-immuun interactie”. Inmiddels zijn jullie allen elders gaan werken. Ik wens jullie heel
veel succes en geluk in de nieuwe omgeving. Jan, het spijt me dat ik de minimale 1.5 x
Balkenende niet hebben kunnen realiseren.
Dr. te Velden, beste Anje. Onder jouw toeziend oog heb ik de eerste voorzichtige schreden
de wetenschap gezet. Zonder jou was ik waarschijnlijk nooit in contact gekomen met
Wouter et al., dank!
242
Dr. Buist, beste Marrije. Dank voor je inzet bij de klinische studies. Je enthousiasme voor het
onderzoek, je visie op het leven en je gevoel voor humor zal ik niet snel vergeten. Zonder jouw
inzet hadden we nooit de inclusie voor de ketotifen-trial, nooit binnen de deadline gehaald!
Dr Ankum, beste Pim. Je input bij de klinische studies was van onschatbare waarde. Je
toegankelijkheid en belangeloze inzet waardeer ik enorm. Professor Matthe Burger, Dr. Ko
van der Velden, Dr. Mark van Beurden (je komt toch wel je beloofde biertje innen?). Ik dank
jullie voor de medewerking en het enthousiasme waarmee jullie me geholpen hebben.
Ook de verpleging van H5zuid (ondanks alle bisacodylletjes) en de dames van de poli
gynaecologie wil ik heel erg bedanken voor hun behulpzaamheid en gastvrijheid.
Professor Hollmann, beste Markus en Dr. Hofland, beste Jan. Jullie expertise op het
gebied van de anesthesie was van onschatbare waarde voor het tot stand komen van de
klinische studies. Jullie deur stond altijd open, fantastisch. Ik hoop dat we nog eens wat
leuke projecten samen kunnen gaan opzetten. Mijn dank aan jullie is groot.
Professor Gouma, Dr. Olivier Busch en Professor Willem Bemelman, beste heren. Ondanks
het grote aantal studies dat op de afdeling chirurgie loopt was de behulpzaamheid en
gastvrijheid vanaf het begin groot. In een constructieve sfeer en met de motivatie om
gezamenlijk mooi onderzoek te doen was het altijd mogelijk om een oplossing te zoeken.
Ik dank jullie voor de goede samenwerking en de behulpzaamheid.
werken. Professor Bartelsman, beste Joep. Het enthousiasme waarmee jij het vak weet over
te brengen is aanstekelijk en heeft er voor gezorgd dat ik dit vak graag wil uitoefenen.
Beste Robert, mister Apple! De beste “sidekick” die een arts-assistent zich kan wensen!
Leuke en boeiende gesprekken over medische, ethische, maatschappelijke en elektronische
onderwerpen in het OLVG en op de fiets naar huis blijven me bij. Solidair tot in de late
uurtjes. Zonder jou was dit boekwerk letterlijk nooit geworden wat het nu is (een esthetische
aanwinst voor iedere boekenkast)! Ik bewonder je veelzijdigheid en extreem sociale inborst.
Ik hoop dat we elkaar niet uit het oog verliezen, dank!
243
inspirerende werkplek waar ik van af het eerste moment het gevoel had graag te willen
Chapter 11
De afdeling Maag-, Darm- en Leverziekten wil ik bedanken voor het bieden van een
Gabor en Jesse. Bij het typen van jullie namen laat ik een traan. Jullie steun, zowel
vakinhoudelijk, wetenschappelijk en als vrienden is met geen pen te beschrijven. Jullie
onvoorwaardelijke vriendschap heb ik meer dan eens op de proef gesteld. Jullie stonden
altijd klaar met raad en daad. Zonder jullie hulp had ik dit nooit kunnen volbrengen. De drie
musketiers ride once more!
Zonder vrienden is het leven zinloos! Sanne, Alain, Fanny, Bart, Barbara, Michiel, Petra,
Laurens, Marjolein, Taco, Annet, Martine maar ook al die anderen. Jullie steun, vertrouwen
en relativeringsvermogen waren onontbeerlijk. Ik hoop nog lang en vooral ook vaker van
jullie vriendschap te mogen genieten.
Anne-Mei, Onno, Mey Mey en Ying. Lieve zus en familie, dank voor jullie peptalk en de
heerlijke momenten samen. Ze waren nodig om zo nu en dan weer even inspiratie op te
doen. Bert en Joke, dank voor het in mij gestelde vertrouwen.
Willemijn, de liefde van mijn leven! Ik ben je eeuwig dankbaar voor de onvoorwaardelijke
steun die me gegeven hebt. Je hebt gezorgd voor de basis en de ideale conditie waaronder
ik kon werken. Nooit heb je geklaagd als ik weer in het weekend naar het AMC moest of
wanneer er tijdens de vakantie aan een stuk gewerkt moest worden. Je weet niet half wat
dit voor mij betekend heeft. En dan te bedenken dat je zelf met een prestigieus AGIKO
project bezig bent! Ik hoop dat ik je de komende tijd iets kan teruggeven van alles wat je
mij gegeven hebt
Hauw en Mariet, mijn lieve ouders. Wie had ooit gedacht dat het dromertje dat zich door
zijn school carrière heen moest worstelen ooit nog eens zou promoveren! Zonder jullie
blindelings vertrouwen en steun had ik het in ieder geval nooit gered! Ik ben jullie innig
dankbaar voor alles wat jullie me hebben meegegeven.
244
245
Chapter 11
Colour
figures
g 3
A
B
C
D
E
F
G
H
Chapter 2 - figure 4
246
E
Chapter 11
C
A
B
200 m
D
247
IM sham
IM VNS1V
IM VNS5V
IM VNS5V + vehicle
IM VNS5V + Hexa
C
L VNS5V + Hexa
248
A
macrophages
B
SM
CM
LM
MP
L sham
IM sham
IM VNS
PYStat-3 dextran DaPi
Chapter 11
C
249
B
F4/80
PY-STA T3
Vehicle
AR-R17779
Nicotine
250
m er g e F4/80 & PY-STA T3
Chapter 11
A
B
C
D
251
A
B
A
B
Kit/WT
Kit/Kit v
C
D
Kit/Kit v
Kit/Kit v
PBS
Kit/WTMC
252
C
CM
Chapter 11
LM
A
B
CM
D
LM
PP
E
F
253
A
20.00 m
254
Chapter 11
preoperative scan
postoperative scan
4
255
A
B
mast cell stabilizers
C
bacterial
translocation
vagus nerve
electrical stimulation
D
macrophage activation
E
rolling activation
adhesion
diapedesis
ICAM-1 antibody/antisense
LFA-1
ICAM-1
F
G
Inhibitory neural
pathway
activation
generalized hypomotility
i.e. POSTOPERATIVE ILEUS
Chapter 10 - Summarizing figure
256
257
Chapter 11

Pathophysiology of Postoperative Ileus: from Bench to Bedside

Transcription

Similar documents

November - December 2009

Dit PDF bestand downloaden

Abstract - Maastricht University Research Portal

Crash - Bekegem

Abstract

Woord vooraf…………………………..

Sparender bestraling van de axilla?

In het Kielzog van Maurice Griffiths door Thijs Vliegers deel 1

Internodium - Semper Florens

Thoughts on rain - Gerrit Rietveld Academie