Contractile effects and intracellular Ca2+ ... - Wiley Online Library

2 downloads 0 Views 341KB Size Report
§Chugai Pharmaceutical Company, Gotemba, Japan. Abstract Motilin has excitatory effects on the colon of the rabbit and the dog, but little is known of its.
Neurogastroenterol. Mot. (2001) 13, 27±35

Contractile effects and intracellular Ca2+ signalling induced by motilin and erythromycin in the circular smooth muscle of human colon G . VAN ASSCHE ,* I . DEPOORTERE ,* T . THIJS ,* L . MISSIAEN ,  F . PENNINCKX ,à H . TAKANASHI ,§ K. GEBOES ,* J . JANSSENS * & T . L . PEETERS *

*Center for Gastroenterological Research,  Laboratory of Physiology, and àDepartment of Surgery, University of Leuven, B-3000 Leuven, Belgium §Chugai Pharmaceutical Company, Gotemba, Japan

Abstract Motilin has excitatory effects on the colon of the rabbit and the dog, but little is known of its effect on the human colon. The aim of this study was to investigate the effects induced by motilin and erythromycin A (EMA) on muscle strips and on single cells from primary cultures from human colon. Isotonic contraction was recorded in circular muscle strips from macroscopically normal resection specimens of patients operated on for colonic neoplasm. Agonist-induced intracellular Ca2+ ([Ca2+]i) signalling was studied in primary cultures of colonic smoothmuscle cells using the ratiometric Ca2+ indicator 1 Indo 1, on a laser-scanning confocal epi¯uorescence microscope. In circular muscle strips, norleucine13porcine motilin ([Nle13]-pm)and EMA induced tonic contractions with an EC50 of 92 ‹ 21 nmol L)1 and 31 ‹ 16 lmol L)1, respectively. The maximal contraction was 21 ‹ 4% (motilin) and 33 ‹ 12% (EMA) of the 2 response to 10)4 mol L)1 acetylcholine (ACh). The motilin antagonist OHM-11526 (10)5.5 mol L)1) abolished the effects of both [Nle13]-pm and EMA. Neither tetrodotoxin (10)5.5 mol L)1), L-nitro-D-arginine methyl 3 ester (L-NAME) (10)3.5 mol L)1) nor guanethidine (10)5 mol L)1) interfered with the effects of [Nle13]-pm or EMA. [Nle13]-pm (10)11)10)6 mol L)1) induced rises of [Ca2+]i in cultured colonic myocytes. At 10)6 mol L)1, 94% of the cells responded, and half of the cells responded at 1.4 nmol L)1 [Nle13]-pm. 81%

(35/43) and 95% (75/79) responded to EMA (10)6 mol L)1) and acetylcholine (ACh, 10)4 mol L)1), respectively. The motilin antagonist GM-109 inhibited motilin- and EMA-induced [Ca2+]i rises. In the absence of extracellular Ca2+, only 13% (7/52) of the cells responded to [Nle13]-pm (10)6 mol L)1) vs. 90% (47/ 52) to ACh (10)4 mol L)1). Motilin and EMA have direct excitatory effects on circular smooth muscle from the human colon and these effects are mediated via a smooth-muscle motilin receptor. These ®ndings suggest that motilin may regulate colonic motility and that motilides may have therapeutic potential for the treatment of colonic hypomotility. Keywords human colon, motility, myocytes, motilin, erythromycin, transduction.

INTRODUCTION Contrary to the multitude of studies investigating the effects of motilin and macrolides on gastric and small intestinal motility, limited and somewhat contradictory information is available on the effects in the mammalian colon. In vivo stimulatory actions of motilin and erythromycin have been described in the human, rabbit and canine colon.1±8 Other authors failed to observe an effect in the human colon.9±11 In the rabbit, the contractile effects of motilin are only partially inhibited by the muscarinic antagonist atropine, which suggests that, similar to the gastric antrum, motilin excites the rabbit colon, at least in part, by mechanisms independent of cholinergic motornerves.7 In vitro, the response of the rabbit colon to motilin and erythromycin has been well characterized. It was observed that circular smooth muscle is more

Address for correspondence Dr T. L. Peeters, Gut Hormone Laboratory, Gasthuisberg O & N, Herestraat 49, B-3000 Leuven, Belgium. Tel.: + 32 1634 5757; fax: + 32 1634 5939; e-mail: [email protected] Received: 18 February 2000 Accepted for publication: 5 September 2000 Ó 2001 Blackwell Science Ltd

27

G. van Assche et al.

Neurogastroenterology and Motility

sensitive than longitudinal muscle to motilin12,13 and to erythromycin.14 These effects are mediated via a smooth-muscle motilin receptor, which has also been characterized in binding studies.14±16 Accordingly, isolated rabbit colonic myocytes respond to both motilin and erythromycin.15 Only one study has paid attention to the effect of motilin on strips from the human colon.12 In this study, motilin was shown to have excitatory effects on taenia coli preparations, but not on circular smooth muscle. Because of the potential therapeutic implications of motilides and motilin antagonists in colonic motility disorders, the aim of the present study was to further characterize the actions of motilin and erythromycin on the human colonic smooth muscle in vitro.

or EMA (10)7 ± 10)3.5 mol L)1) in some experiments. The effect of neuronal blockers and of OHM-11526 was studied by incubating the preparations for 10 min prior to a challenge with [Nle13]-pm or EMA. ACh (10)4 mol L)1) was given at the end of every experiment as a reference contraction. The maximal response to ACh before and after [Nle13]-pm or EMA administration was similar.

Smooth-muscle cultures Human colonic circular muscle explants were obtained from ganglionated colonic segments from two patients (aged 2 and 6 months) operated on for congenital megacolon. Colon segments were cleared of mucosa and submucosa and bands of circular smooth muscle were prepared by sharp dissection. Samples of the dissected muscle layers were rapidly frozen and haematoxylin±eosin-stained sections were examined using an inverted microscope (Olympus IMT-2; Olympus, Tokyo, Japan) to verify the absence of submucosal remnants. Fragments of the muscle layers were explanted in 60 mm culture dishes (Corning, NY, USA) and grown in Dulbecco's Modi®ed Eagles Medium (DMEM; Gibco BRL, Grand Island, NY, USA) with 100 U mL)1 penicillin G, 100 lg mL)1 streptomycin and 0.25 lg mL)1 amphotericin B (Gibco) and supplemented with 10% fetal bovine serum (FBS; Gibco). The primary cultures were incubated in a 5% CO2 incubator at 37 °C and the medium was changed twice per week. Upon reaching con¯uence, cells were dissociated with 0.125% trypsin±EDTA (Gibco), concentrated 4 by centrifugation at 55 g for 5 min and replated at the desired density. Cells were used between passages 2 and 5. For cell counts, trypsinized cells in suspension were diluted 1:4 with trypan blue (0.4%) and loaded into a Burker chamber. Viability, as judged by trypan blue exclusion, was always preserved in more than 90% of trypsinized cells. For immunohistochemistry, cells were seeded in 24-well plates on Thermanox inserts (Nunc, Naperville, IL, USA). Inserts with con¯uent cells attached were removed from the media and shortly ®xed in acetone. Immunohistochemistry was performed using a three-step indirect immunoperoxidase method with monoclonal antibodies against a-smooth-muscle actin (dilution 1:400; Sigma). The expression of a and c enteric smooth-muscle actin was determined with the reverse transcriptase± polymerase chain reaction (RT±PCR). Glyceraldehydephosphate dehydrogenase (GAPDH) mRNA expression was used as an internal control.17 Total RNA was prepared from con¯uent cells and from circular muscle tissue using TRIzol reagent (Gibco). One

MATERIALS AND METHODS Chemicals Norleucine13-porcine motilin ([Nle13 ]-pm) was purchased from Eurogentec, Namur, Belgium. Acetylcholine, L-NAME, tetrodotoxin, guanethidine, and atropine were all from Sigma (St Louis, MO, USA). Erythromycin A lactobionate (EMA) was from Abbott Laboratories (Chicago, IL, USA), and the motilin antagonist GM-109 was obtained from Chugai Pharmaceutical Company (Gotemba, Japan). The motilin antagonist OHM-11526 was a gift from Dr A. Galdes (Ohmeda, Murray Hill, NJ, USA).

Contraction studies Macroscopically intact segments were obtained from resection specimens of left or right colon from 16 patients (mean age 65 ‹ 7 years, n ˆ 16) undergoing surgery for colonic neoplasm. Mucosa and submucosa were removed under a dissection microscope and muscle strips (20 ´ 2 mm) were cut along the circular axis. Strips were suspended in water-jacketed organ baths at 37 °C in 10 mL of Krebs solution containing (in mmol L)1) NaCl 120.9;, NaH2PO4 2.0; NaHCO3 15.5; KCl 5.9; CaCl2 2.5; MgCl2 1.2; D-glucose 11.5, oxygenated by 95% O2/5% CO2 with a counter-weight of 1 g. Contractions were measured isotonically. Tissues were left to equilibrate for at least 60 min while the bath solution was changed regularly. During equilibration ,repeated challenges of acetylcholine (ACh; 10)4 mol L)1) were given until a stable response was obtained. Pharmacological agents were administered as 100 lL aliquots to the organ baths either as single doses or in cumulative doses for [Nle13]-pm (bath concentrations: 10)9 ± 10)6 mol L)1) 28

Ó 2001 Blackwell Science Ltd

Volume 13, Number 1, February 2001

Motilin on human colon

lg of RNA was used to synthesize single-stranded cDNA using 200 units of SuperscriptTM II RNase-H reverse transcriptase and oligodeoxynucleotide (25 lg mL)1; Gibco) as primer. The obtained cDNA served as a template for the PCR, consisting of a denaturation step at 94 °C for 4 min followed by 32 cycles of ampli®cation (94 °C for 1 min, 57 °C for 1 min, 72 °C for 1 min with a ®nal extension duration of 10 min at 72 °C), using 2.5 U of TaqTM DNA polymerase (Pharmacia Biotech, Uppsala, Sweden) and speci®c primers for GAPDH (forward, GTTCCAATATGATTCCACC; reverse, TGAGTCCTTCCACGATAC) and actin (c-actin forward: ACCACAGCTGAGAGAGAA, reverse: TGTCTCATGAATTCCAGC; a-actin forward: TTCCTGCTCCTCTCTGTCTC, reverse: TGTGATTCTACCCTTTGCC), selected from the respective cDNA sequences reported in the literature.17,18 This resulted in cDNA fragments of 228 bp for c-enteric smooth-muscle actin, 291 bp for a-smooth-muscle actin and 386 bp for GAPDH. PCR fragments were loaded on a 1.5% agarose gel and after electrophoresis DNA bands were visualized with ethidium bromide under UV light.

above the cell layer to allow for rapid ¯uid exchange without mechanical disturbance by the superfusate. In experiments where Ca2+ was removed from the extracellular medium, 2 mmol L±1 EGTA was added to the modi®ed Ca2+-free Krebs solution. Pharmacological compounds were added to the superfusate. Indo-1 was excited with an argon laser at 351 nm and emission was measured at two wavelengths (405 and 470 nm). The ¯uorescence emitted by Indo-1 decreases at 470 nm and increases at 405 nm with elevations of [Ca2+]i. The ratio of the ¯uorescence registered at both wavelengths therefore re¯ects changes of [Ca2+]i and allows the elimination of artifacts. Because it is generally accepted that calibration curves are dif®cult to obtain using this method,20 we have always expressed the results as ratios of emitted ¯uorescence.

Data analysis Results of the contraction studies were expressed relative to the isotonic contraction obtained with 10)4 mol L)1 acetylcholine (ACh). Statistical signi®cance was evaluated using ANOVA or Student's t-test when appropriate. The traces of ¯uorescence ratios (405/470 nm), re¯ecting [Ca2+]i shown in the ®gures, are representative for experiments performed in at least four chamber slides. Results of [Ca2+]i measurements were expressed as the percentage of responding cells and evaluated with chi-squared statistics. Statistical signi®cance was accepted at the 95% level.

Single-cell agonist-induced intracellular Ca2+ measurements Agonist-induced intracellular Ca2+ ([Ca2+]i) measurements were performed as described previously,19 using a laser scanning confocal ¯uorescence microscope (MRC-1024; Bio-Rad, Hertfordshire, UK) linked to an inverted epi¯uorescence microscope (NikonDiaphot-300; Nikon, Tokyo, Japan) with a CF Fluor 40x (NA 1.3) oil-immersion objective. The scanning box was set at a resolution of 256 ´ 256, allowing for the selection of 8)17 individual cells, depending on the cell type and density. Scanning was performed every 0.5±0.75 s and the images were not averaged. For loading the cells with the ratiometric Ca2+ indicator Indo-1, the culture medium was removed and cells were washed with a modi®ed Krebs solution containing (in mol L)1): NaCl 135 , KCl 5.9, CaCl2 1.5, MgCl2 1.2, HEPES 11.6 and glucose 11.5, at pH 7.3. Cells were subsequently incubated for 1 h at 22 °C with 10 lmol L)1 Indo-1-AM and Pluronic F-127 (0.04%; Molecular Probes, Leiden, the Netherlands) dissolved in modi®ed Krebs solution. Cells were washed twice with modi®ed Krebs solution and further incubated for another 1±2 h in the absence of Indo1-AM. Coverglass chambers were mounted on the microscope and continuously superfused with modi®ed Krebs solution at 22 °C from a pipette placed 1 mm Ó 2001 Blackwell Science Ltd

RESULTS Contractile effects of norleucine13 porcine motilin and erythromicin A on muscle strips from human colon In seven out of ten specimens from the left colon and four out of six from the right colon, addition of both [Nle13]-pm (10)9 ± 10)6 mol L)1) or EMA (10)7± 10)3.5 mol L)1) resulted in a tonic contraction of the preparation, superimposed on the spontaneous phasic activity. The amplitude of the phasic oscillations was reduced at high concentrations of both [Nle13]-pm (³10)6.5 mol L)1) and EMA (³10)4 mol L)1) (Fig. 1), but their frequency was unaffected (control 5.5 ‹ 0.4 contractions min)1, n ˆ 10; [Nle13]-pm (10)6 mol L)1) 5.6 ‹ 0.6 contractions min)1, n ˆ 6; EMA (10)4 mol L)1) 5.3 ‹ 0.4 contractions min)1, n ˆ 6). The half maximal response to [Nle13]-pm and EMA was observed at 92 ‹ 21 nmol L)1 (pEC50 7.0 ‹ 0.3) and 31 ‹ 16 lmol L)1 (pEC50 4.5 ‹ 0.2), respectively, and 29

G. van Assche et al.

Neurogastroenterology and Motility

the maximal contraction was 21 ‹ 4% (Nle13 ]po-motilin) and 33 ‹ 12% (EMA) of the maximal response to 10)4 mol L)1 ACh. In single dose experi-

ments the maximal response induced by [Nle13]-pm or EMA was not different from the response in cumulative concentration±response experiments (Table 1).

Figure 1 Representative tracing of the concentration-dependent response of human colon circular muscle to EMA (top) and to norleucine13 porcine motilin (bottom). The motilin experiment was performed in the presence of tetrotodoxin (3 ´ 10)6 mol L)1). The isotonic contraction is expressed as mm displacement.

Table 1 Potency and maximal effect of norleucine13 porcine motilin and erythromycin A on in vitro human colon contractile activity

[Nle13] porcine motilin pEC50 0.70 ‹ 0.3 EC50 92 ‹ 21 nM Maximal contraction Cumulative 21 ‹ 4% Single dose 22 ‹ 3% (10)6 mol L)1) )5.5 )1 + TTX (10 mol L ) 21 ‹ 3% 24 ‹ 4% + L-NAME (10)3.5 mol L)1) + guanethidine (10)5 mol L)1) 26 ‹ 10%

EMA 4.5 ‹ 0.2 31 ‹ 16 lmol L)1 33 22 27 18 22

‹ ‹ ‹ ‹ ‹

12% 5% (10)4 mol L)1) 8% 2% 7%

pEC50 is the negative logarithm of the concentration that induced a half maximal response. Maximal contraction values are expressed as a percentage relative to the response obtained with acetylcholine (10)4 mol L)1). Values shown are means ‹ SE (n = 3). TTX, tetrododoxin; L-NAME.

30

Ó 2001 Blackwell Science Ltd

Volume 13, Number 1, February 2001

Motilin on human colon

Figure 2 Inhibition of the norleucine13 porcine motilin- and erythromycin A-induced contractions by the motilin antagonist OHM-11526. Tracing of one, representative, experiment. The isotonic contraction is expressed as mm displacement.

Figure 3 Preservation of the contractile response to norleucine13 porcine motilin (10)6 mol L)1) after the inhibition of nitrergic (A) and adrenergic (B) neurotransmission. Note that L-NAME induced a tonic increase in baseline activity.

Pharmacological characterization of the response to norleucine13 porcine motilin

actin staining was orientated in clear ®brils along the longitudinal axis. Moreover, electrophoresis of RT±PCR products obtained with speci®c primers demonstrated a similar pro®le of both a and c enteric smooth-muscle actin mRNA expression in the subcultured cells and in the parent smooth-muscle layers (Fig. 4). Alpha smooth-muscle actin is considered to be a speci®c marker for smooth-muscle cells and has been shown to be present in cultured smooth muscle, whereas c enteric smooth-muscle actin is exclusively found in myocytes from the gut wall21,22.

The motilin antagonist OHM-11526 (10)5.5 mol L)1) did not induce contractions, but prevented the contractile response to [Nle13]-pm (10)6± )5.5 )1 )3.5 )1 10 mol L ) and EMA (10 mol L ). OHM-11526 did not interfere with the ACh-induced contraction (Fig. 2). To differentiate a direct effect of motilin on colonic smooth muscle from an excitatory action on enteric nerves, preparations were incubated with 10)5.5 mol L)1 tetrodotoxin (TTX) prior to the addition of motilin. TTX induced a rise in the baseline tonus (10.9 ‹ 2.8% of the response to ACh), but did not affect the [Nle13]-pm response (Fig. 1, Table 1). Similarly, inhibition of nitric oxide synthesis by L-NAME (10)3.5 mol L)1) resulted in a tonic increase of baseline activity (12.2 ‹ 1.7% of response to ACh), but [Nle13]pm (10)6 mol L)1) still induced a contractile response (Fig. 3, Table 1). Guanethidine, a blocker of neurotransmitter release from adrenergic nerves (10)5 mol L)1) did not affect baseline contractility and did not interfere with motilin-induced responses (Fig. 3, Table 1).

Norleucine13 porcine motilin and acetylcholineinduced intracellular Ca2+ transients in human colon myocytes

Figure 4 Ethidium bromide stain of a DNA-agarose gel loaded with PCR-products obtained after ampli®cation of cDNA from human colon using primers for c-enteric smooth-muscle actin (228 bp), a-smooth-muscle actin (291 bp) and GAPDH (386 bp). Lane 1: circular muscle layer of human colon; Lane 2: cultured myocytes from the same preparation; Lane 3: negative control (sample without cDNA), Lane 4: molecular weight marker giving bands of 100 bp spacing.

The cells subcultured from colonic smooth-muscle explants preserved a smooth-muscle phenotype, as they displayed a distinct `hill and valley' appearance when con¯uent, and more than 80% of the cells stained intensively for a smooth-muscle actin. The Ó 2001 Blackwell Science Ltd

31

G. van Assche et al.

Neurogastroenterology and Motility

Baseline [Ca2+]i oscillations were observed in 41% (12/29) of the myocytes, with a mean frequency of 0.3 oscillations min)1 (0.2±0.8). 93% (28/30) of the myocytes responded when depolarized by superfusion with 75 m mol L)1 KCl, in modi®ed Krebs solution. Both ACh and [Nle13]-pm induced [Ca2+]i transients in human colonic myocytes. 95% (75/79)of the cells responded to ACh (10)4 mol L)1) and 94% (92/98) to [Nle13]-pm (10)6 mol L)1). The threshold concentration for the [Nle13]-pm effect was 10)10 mol L)1, and with increasing doses, the percentage of cells responding increased (Fig. 5). It can be estimated from this ®gure that 50% of the cells responded at a [Nle13]pm concentration of 1.4 ´ 10)9 mol L)1. EMA induced elevations of [Ca2+]i in 81% (35/43) of myocytes at 10)6 mol L)1. To investigate the speci®city of the motilin- and EMA-induced [Ca2+]i transients, the motilin receptor antagonists OHM-11526 and GM-109 were superfused for 90 s before and during agonist administration. OHM-11526 (10)6 mol L)1) induced a [Ca2+]i transient in 69% (36/52) of myocytes, but abolished the subsequent response to [Nle13]-pm (10)7 mol L)1) (0/52 cells responding; Table 2). GM-109 (10)5 mol L)1) was effective in 24% (52/216) of the cells, abolishing the response to [Nle13]-pm (10)7 mol L)1) and signi®cantly reducing the response to EMA (10)6 mol L)1) (Table 2, Fig. 6A). The response to ACh (10)4 mol L)1) and KCl (75 mmol L)1), however, was preserved in 100% (30/ 30) and 90% (36/40) of the cells, respectively, in the presence of GM-109 (10)5 mol L)1) (Fig. 6A,B). EMA was also found to block the subsequent response to

Table 2 Effect of motilin antagonists on intracellular Ca2+ transients induced by the subsequent administration of motilin, erythromycin A and acetylcholine Motilin EMA ACh (10±7 mol L±1) (10±6 mol L±1) (10±4 mol L±1) 100% (21/21) 81% (35/43) + GM-109 0% (0/52) * (10±5 mol L±1) + OHM-11526 0% (0/52) * (10±6 mol L±1)

96% (26/27) 28% (19/69) * 100% (30/30) ND

94% (49/52)

Data are expressed as percentage responding cells with the total number of cells between brackets (*: P < 0.001 vs. response to agonist alone, chi-square). ND, not determined; EMA, erythromycin A; ACh, acetylcholine.

motilin but not to ACh. An example of such an experiment is shown in Figure 6(C). To investigate the role of Ca2+ in¯ux in the generation of the [Ca2+]i oscillations, agonists were superfused in Ca2+-free conditions (0 mmol L)1 extracellular Ca2+,; 2 mmol L)1 EGTA). In the absence of extracellular Ca2+ the [Nle13]-pm (10)6 mol L)1) response was preserved in 13% (7/52) and the ACh (10)4 mol L)1) response in 90% (47/52) of the myocytes (Fig. 7). Interestingly, [Ca2+]i increased transiently in 77% of the myocytes when Ca2+-free modi®ed Krebs solution was superfused (Fig. 7), probably due to release of Ca2+ from intracellular stores. This suggests that background Ca2+ in¯ux inhibits Ca2+ release from these stores.

DISCUSSION The results of the present study provide the ®rst in vitro evidence of a direct excitatory action of both motilin and EMA on circular smooth muscle from left and right human colon in vitro. Both agonists induced contractions in muscle strips and [Ca2+]i transients in cultured myocytes. However, motilin effects on Ca2+ levels occurred at signi®cantly lower concentrations compared to the contractile effects in colonic muscle strips. Only one study has investigated the effect of motilin on the human colon in vitro. Strunz et al.12 reported that the taeniae from the sigmoid colon, but not circular muscle strips from left and right colon, responded to motilin, although the amplitude of the response was low compared to the one induced by ACh. We provide evidence for motilin-induced contractile effects and [Ca2+]i transients in human circular smooth muscle. However, similar to the results obtained by Strunz et al.12 in the taeniae, the amplitudes of the motilin-

Figure 5 Effect of [Nle13] porcine motilin on human colon myocytes. (A) Fluorescence ratio of one single cell in response to 10)10 and 10)9 mol L)1 motilin. (B) Effect of increasing norleucine13 porcine motilin concentrations on the percentage of cells responding to the agonist. [total number of cells tested: 98 (10)6 mol L)1), 21 (10)7 mol L)1), 43 (10)8 mol L)1), 64 (10)9 mol L)1), 58 (10)10 mol L)1), 41 (10)11 mol L)1)].

32

Ó 2001 Blackwell Science Ltd

Volume 13, Number 1, February 2001

Motilin on human colon

Figure 6 (A) Inhibition of the effect of norleucine13 porcine motilin ([Nle13]-pm; 10)7 mol L)1) by the motilin antagonist GM-109 (10)5 mol L)1). (B) Absence of an effect of GM-109 on the response to KCl (75 mM). Note that in this cell GM-109 induced a rise in [Ca2+]i. (C) Evidence for cross-tachyphylaxis between erythromycin A (10)6 mol L)1) and [Nle13]-pm (10)7 mol L)1). Following the response to erythromycin-A, the cell responds to subsequent administration of acetylcholine (10)4 mol L)1) but not to [Nle13]-pm (10)7 mol L)1).

amplitude, may explain why they escaped the attention of Strunz et al.12 Because inhibition of the neuronal conductance by TTX did not interfere with motilin- or EMA-induced contractions in the circular muscle strips, the observed effects appear to be mediated by a motilin receptor on the smooth-muscle membrane. Inhibition of nitrergic or sympathetic neurotransmission was also unable to block the motilin-induced contractions of human colonic circular smooth muscle, indicating that motilin does not act by blocking these inhibitory nerves. Indeed, both TTX and L-NAME induced a tonic contraction, indicating that nitrergic inhibitory nerves are preserved and persistently active in isolated muscle strips derived from human colon. A similar phenomenon was observed in the rabbit colon although in contrast to our study, adrenergic nerves account for the tonic inhibition in this species.14,23 The motilin-induced contractions were abolished by OHM-11526, an antagonist of the direct smooth-muscle effects of motilin in the rabbit duodenum and human antrum.24 Surprisingly, OHM-11526 increased [Ca2+]i in a substantial proportion of our isolated cells,

Figure 7 Effect of [Nle13] porcine motilin (10)6 mol L)1) and acetylcholine (10)4 mol L)1) on intracellular Ca2+ ([Ca2+]i) in Ca2+-free medium. Note that the removal of Ca2+ resulted in a transient rise of [Ca2+]i. Results are expressed as the ratio of emitted ¯uorescence intensity (405/470 nm).

and EMA-induced contractions in the circular muscle were low compared to the effect of ACh. The responses of colonic smooth-muscle strips to motilin and EMA in our study developed rather slowly. This, and their low Ó 2001 Blackwell Science Ltd

33

G. van Assche et al.

Neurogastroenterology and Motility

yet inhibited the subsequent response to motilin. As OHM-11526 has agonist properties in the canine and avian intestine,25,26 and is equipotent to motilin in inducing Ca2+ currents in isolated human and canine jejunal smooth-muscle cells,27 it cannot be excluded that the compound has weak agonist activity not revealed in the contractility measurements. Therefore, experiments were also performed with the antagonist GM-109, a compound with a lower af®nity for the motilin receptor than OHM-11526, but for which, as yet, no agonist activity has been reported. However, GM-109 was also found to induce increases in [Ca2+]i, although in far fewer cells than OHM-11526. GM-109 abolished the subsequent responses to motilin or to EMA, but not to ACh or KCl. One may speculate that the changes in [Ca2+]i induced by OHM-11526 and by GM-109 are too small to result in a contraction, but the phenomenon deserves further study. However, our data do indicate that the motilin- and EMA-induced effects are mediated via a motilin receptor on smooth-muscle cells. Our results also indicate that motilin-induced Ca2+ signalling is dependent on Ca2+ in¯ux, and this in contrast to ACh, which seems to trigger the release from intracellular Ca2+ stores. This is in accordance with earlier studies on the role of extra- and intracellular Ca2+ in motilin- and ACh-induced contractions in the rabbit and human small intestine.27±29 It may be noted that the preservation of KCl effects in the presence of GM-109 indicates that the compound does not block Ca2+ in¯ux nonspeci®cally, for instance, by depolarizing the membrane. On cultured myocytes, motilin was effective at 100-fold lower concentrations than in the tissue bath. A very high sensitivity of isolated smooth muscle to agonists including motilin has been observed before.30±33 In freshly isolated rabbit colonic myocytes, Hasler et al.15 have reported contractile effects of motilin with an EC50 of 2 pmol L)1, much lower than the EC50 in the tissue bath (10 nmol L)1) and than the af®nity of the motilin receptor (1 nmol L)1) reported in the rabbit colon.14 The reason for the discrepancy is unclear. Taken together, the data from contraction studies and Ca2+-signalling experiments provide evidence for an excitatory role of motilin on human colonic circular muscle, which suggests that motilin may have a role in the regulation of colonic motor activity and that motilin receptor agonists, such as motilides, may have prokinetic effects in vivo. The therapeutical implications are important, since colonic dysmotility underlies frequent clinical entities such as chronic constipation and irritable bowel syndrome and until now few drugs have proven to be ef®cient at amelior-

ating colonic hypomotility. As yet, in vivo studies investigating the effects of erythromycin on colonic motility and transit have yielded con¯icting data. In constipated patients oral erythromycin (1 g day)1) has been shown to enhance colonic transit and to increase stool frequency,8 and a similar acceleration in transit time was found in healthy volunteers.5 However, in other studies erythromycin had no effect9±11 or a very limited effect.6 As new and more potent derivatives of erythromycin are now available, an exploration of their effect on the colon seems to be warranted.

ACKNOWLEDGMENTS Supported by grants from the Fund for Scienti®c Research, Flanders (Belgium) (FWO grant number G 0109.00) and the Belgian Ministry of Science (GOA 98/ 011 and IUAP P4/16). I. Depoortere is a postdoctoral research fellow of the Fund for Scienti®c Research, Flanders (Belgium).

REFERENCES 1 Rennie JA, Christo®des ND, Mitchener P, Johnson AG, Bloom SR. Motilin and human colonic activity. Gastroenterology 1980; 78: A1243. 2 Zara GP, Qin XY, Pilot MA, Thomson HH, Maskell JP. Erythromycin and gastrointestinal motility. Lancet 1987; 2: 1036. 3 Bickel M, Belz U. Motilin and a synthetic enkephalin induce colonic motor complexes (CMC) in the conscious dog. Peptides 1988; 9: 501±7. 4 Hasler W, Heldsinger A, Soudah H, Owyang C. Erythromycin promotes colonic transit in humans: mediation via motilin receptors. Gastroenterology 1990; 98: A358. 5 Lehtola J, Jauhonen P, KesaÈniemi P, Wikberg R, Gordin A. Effect of erythromycin on the oro-caecal transit time in man. Eur J Clin Pharmacol 1990; 39: 555±8. 6 Bradette M, Poitras P, Boivin M. Effect of motilin and erythromycin on the motor activity of the human colon. J Gastroint Mot 1993; 5: 247±51. 7 Kitazawa T, Ichikawa S, Yokoyama T, Ishii A, Shuto K. Stimulating action of KW-5139 (Leu13-motilin) on gastrointestinal motility in the rabbit. Br J Pharmacol 1994; 111: 288±94. 8 Sharma SS, Bhargava N, Mathur SC. Effect of oral erythromycin on colonic transit in patients with idiopathic constipation. A pilot study. Dig Dis Sci 1995; 40: 2446±9. 9 Delvaux M, Louvel D, Fioramonti J, Staumont G, Bueno L, Frexinos J. Effect of various doses of erythromycin on colonic myoelectrical activity in IBS patients. Neurogastroenterol Motil 1994; 6: 205±12. 10 Yeaton P, Weber F, Chen J, McMallum R. Effect of erythromycin on segmental colonic motility in humans. Gastroenterology 1992; 102: A536. 11 Jameson JS, Rogers J, Misiewicz JJ, Raimundo AH, Henry MM. Oral or intravenous erythromycin has no effect on

34

Ó 2001 Blackwell Science Ltd

Volume 13, Number 1, February 2001

12

13 14 15 16

17 18

19

20 21 22

Motilin on human colon

human distal colonic motility. Aliment Pharmacol Ther 1992; 6: 589±95. Strunz U, Domschke W, Mitznegg P et al. Analysis of the motor effects of 13-norleucine motilin on the rabbit, guinea pig, rat, and human alimentary tract in vitro. Gastroenterology 1975; 68: 1485±91. Adachi H, Toda N, Hayashi S et al. Mechanism of the excitatory action of motilin on isolated rabbit intestine. Gastroenterology 1981; 80: 783±8. Depoortere I, Peeters TL, Vantrappen G. Motilin receptors of the rabbit colon. Peptides 1991; 12: 89±94. Hasler WL, Heldsinger A, Chung OY. Erythromycin contracts rabbit colon myocytes via occupation of motilin receptors. Am J Physiol 1992; 262: G50±G55. Sakai T, Satoh M, Sonobe K, Nakajima M, Shiba Y, Itoh Z. Autoradiographic study of motilin binding sites in the rabbit gastrointestinal tract. Regul Pept 1994; 53: 249±57. Tokunaga K, Nakamura Y, Sakata K et al. Enhanced expression of a glyceraldehyde-3-phosphate dehydrogenase gene in human lung cancers. Cancer Res 1987; 47: 5616±9. Miwa T, Manabe Y, Kurokawa K. Structure, chromosome location, and expression of the human smooth muscle (enteric type) gamma-actin gene: evolution of six human actin genes. Mol Cell Biol 1991; 11: 3296±306. Missiaen L, Parys JB, Bootman MD, De Smedt H, Droogmans G, Casteels R. Normal Ca2+ signalling in glutathione-depleted and dithiothreitol-treated HeLa cells. P¯ugers Arch 1993; 423: 480±4. Grynkiewicz G, Poenie M, Tsien RY. A new generation of Ca2+ indicators with greatly improved ¯uorescence properties. J Biol Chem 1985; 260: 3440±50. Chamley-Campbell J, Campbell GR, Ross R. The smooth muscle cell in culture. Physiol Rev 1979; 59: 1±61. McHugh KM, Lessard JL. The development expression of the rat alpha-vascular and alpha-enteric smooth muscle isoactins: isolation and characterization of a rat gammaenteric actin cDNA. Mol Cell Biol 1988; 8: 5224±31.

Ó 2001 Blackwell Science Ltd

23 Snape WJ, Yoo S. Effect of amino acids on isolated colonic smooth muscle from the rabbit. J Pharmacol Exp Ther 1985; 235: 690±5. 24 Depoortere I, Macielag MJ, Galdes A, Peeters TL. Antagonistic properties of [Phe3,Leu13] porcine motilin. Eur J Pharmacol 1995; 286: 241±7. 25 Matsuo H, Peeters TL, Janssens J, Depoortere I. Effect of the putative motilin antagonist ANQ-11168 on the migrating motor complex (MMC) in conscious dogs. Neurogastroenterol Motil 1994; 6: 148. 26 De Clercq P, Depoortere I, Macielag M, Vandermeers A, Vandermeers-Piret MC, Peeters TL. Isolation, sequence, and bioactivity of chicken motilin. Peptides 1996; 17: 203±8. 27 Farrugia G, Macielag MJ, Peeters TL. Motilin and OHM11526 activate a calcium current in human and canine jejunal circular smooth muscle. Am J Physiol 1997; 273: G404±12. 28 Matthijs G, Peeters TL, Vantrappen G. Effect of different calcium modulators on motilin-induced contractions of the rabbit duodenum. Comparison with acetylcholine. Regul Pept 1988; 21: 321±30. 29 Peeters TL, Matthijs G, Vantrappen G. Ca2+ dependence of motilide-induced contractions in rabbit duodenal muscle strips in vitro. Naunyn-Schmiedeberg's Arch Pharmacol 1991; 343: 202±8. 30 Bitar KN, Makhlouf GM. Receptors on smooth muscle cells: characterization by contraction and speci®c antagonists. Am J Physiol 1982; 242: G400±7. 31 Collins SM, Crankshaw DJ. Dissociation of contraction and muscarinic receptor binding to isolated smooth muscle cells. Am J Physiol 1986; 251: G546±52. 32 Louie DS, Owyang C. Motilin receptors on isolated gastric smooth muscle cells. Am J Physiol 1988; 254: G210±G216. 33 Moummi C, Magous R, Bali JP. Gastrointestinal hormone receptors on isolated smooth muscle cells from gastric antrum of the rabbit. Biochem Pharmacol 1989; 38: 2895±901.

35