Introduction

Top Abstract Introduction Methods Results Discussion Conclusions References

NO is an established messenger in the gastrointestinal tract (for reviews see Konturek and Konturek, 1995, Makhlouf and Grider, 1993, Stark and Szurszewski, 1992, Whittle 1994). NO either alone or with the putative cotransmitters VIP, PACAP and ATP is thought to mediate the nonadrenergic noncholinergic inhibitory component of the peristaltic wave. However, NO has also been shown to induce contractions by a direct effect in rat small intestine and by release of acetylcholine in guinea pig small intestine (Bartho and Lefebvre, 1995).

NO-generating myenteric neurons are numerous in the mammalian gut as demonstrated immunocytochemically by visualization of NOS and by NADPH diaphorase histochemistry (Bredt et al., 1991, Costa et al., 1992, Ekblad et al., 1994a, Stark and Szurszewski, 1992) or by in situ hybridization for NOS mRNA (Ekblad et al., 1994b). NOS-containing myentric neurons issue anal projections to the smooth muscle layers and other myenteric ganglia (Costa et al., 1992, Ekblad et al., 1994a). These data are compatible with a transmitter role of NO in inhibitory enteric neurons. In rat small intestine NOS-containing nerve fibers running to the smooth muscle contain in addition VIP and NPY although those running within myenteric ganglia lack these neuropeptides (Ekblad et al., 1994a, b). Some of them also contain PACAP (Hannibal et al., in press). An interplay between VIP and NO has been suggested. VIP may cause NO release from nerve fibers (Huizinga et al., 1992) and from smooth muscle cells (Murthy and Makhlouf, 1994). However, the ability of VIP to induce NO release from smooth muscle cells has been questioned (Keef et al., 1994, Chakder and Rattan 1996). Furthermore, NO has been found to stimulate VIP release from isolated myenteric ganglia (Grider and Jin, 1993). Information on which role NPY might play as a cotransmitter is scarce; it has been reported to inhibit the release of both VIP and NO in canine ileum (Fox-Threlkeld et al., 1993).

PACAP, which belongs to the VIP family of peptides, has been shown to stimulate NOS activity in isolated smooth muscle cells, with a potency equal to that of VIP (Murthy and Makhlouf 1994); in another study inhibition of NOS was found to inhibit electrically induced PACAP release (Grider et al., 1994). However, nothing is known about interactions of NO and PACAP in the regulation of gut motility.

The aims of our study were to examine NO effects, both contractile and relaxatory, on strips of longitudinal muscle with adherent myenteric ganglia from rat ileum using NO donors and electrical field stimulation. NO effects were also examined after addition of the neuropeptides VIP, NPY, PACAP-27 or PACAP-38, after stimulation of adenylate cyclase by forskolin, or after inhibition of soluble guanylate cyclase by MB or ODQ in order to find out if these substances alter the NO evoked responses. The effects of atropine, L-NAME, TTX or the substance P antagonist Spantide on NO evoked responses were also examined, as were possible alterations in electrically induced acetylcholine release due to NO addition.

	Methods

Top Abstract Introduction Methods Results Discussion Conclusions References

Fifty female Spraque-Dawley rats (160-180 g) were used. The animals were anaesthetized in vaporized diethylether and killed by bleeding. The ileum was dissected out and placed in ice-cold Krebs solution with the following composition (in mM): NaCl 133.0, KCl 4.7, CaCl₂ × 2 H₂O 2.5, MgCl₂ 1.0, NaHCO₃ 16.3, NaH₂PO₄ 1.4 and glucose 7.8. Strips (10 mm long) from longitudinal smooth muscle with adherent myenteric ganglia were dissected out using a dissection microscope. The strips were mounted in organ baths containing 5 ml of Krebs solution aerated with a mixture of 5% CO₂ and 95% O₂ and continuously recorded for isometric tension with a Grass FTO3C force-displacement transducer and registered on a Grass model 79D polygraph. Bath temperature was maintained thermostatically at 37°C and the pH was kept around 7.40 (range 7.35-7.45). The strips were given an initial passive load of 20 to 25 mN and allowed to equilibrate for 1 hr before experimentation started. During this period the preparations gradually relaxed and the basal tension stabilized at 3 to 5 mN. Rythmic spontaneous contractions with varying amplitude developed in all preparations.

The following drugs were used: Atropine sulphate, TTX, acetylcholine, SNP, SNAP, L-arginine, L-NAME, forskolin, MB, VIP, NPY, PACAP-38 and PACAP-27 were purchased from Sigma, St. Louis, MO. ODQ was from Tocris Cookson, Bristol, UK. Prostaglandin F₂ (PGF₂; Prostin) was from Upjohn Co., Kalamazoo, MI. Spantide ([D-Arg¹, D-Trp^7,9, Leu¹¹]SP) was from Ferring, Malmö, Sweden. SNAP and ODQ were dissolved in 0.1 M DMSO, the other substances were dissolved in 0.9% saline.

EFS. Platinum electrodes (0.5 mm in diameter and 10 mm apart) were mounted on both sides of the muscle strip. The electrodes were connected to a Grass S4C stimulator for field stimulation with square wave pulses (4 V over the electrodes, 400 mA, 1 msec duration, frequency 4-20 Hz); pulse trains lasting for 5 sec. Pretreatment with forskolin, TTX, atropine or L-arginine was by addition of the substances 5 min before EFS; L-NAME, ODQ or MB were added 30 min before EFS.

5

4

NO donors. Experimentation started with the addition of acetylcholine (10⁵ M) to the bath followed by wash out and recovery. The strips were then precontracted with PGF₂; when stable, acetylcholine (10⁵ M) was added and, after wash out the strips were again allowed to recover. The precontraction of the muscle strips was by a submaximal concentration of PGF₂. The PGF₂ concentration (10⁸-10⁷ M) was chosen individually for each muscle strip to ensure that the contraction level was the same, relative to acetylcholine (10⁵ M) in all experiments. Concentration response curves for the NO donors SNP or SNAP were constructed on precontracted muscle strips and on muscle strips at basal tension. Single doses of the NO donors were tested one at a time and the strips were allowed to recover for 30 min between each dose. Each strip was exposed to six to eight different concentrations of SNP or SNAP, always starting with the lowest dose. Spantide, TTX, atropine, forskolin, VIP, NPY, PACAP-27 or PACAP-38 were added 5 min before precontraction. L-NAME, ODQ and MB were added 30 min before precontraction. Neither of the substances used for pretreatment affected the baseline or the PGF₂-induced contraction.

5

	Results

Top Abstract Introduction Methods Results Discussion Conclusions References

Responses evoked by EFS. EFS (4-10 Hz, 5 sec) resulted in contractions that could be almost completely abolished by atropine (10⁶ M) (n = 12) or TTX (10⁶ M) (n = 12). L-NAME (10⁴ M) (n = 14) had no effect on the amplitude of the contractions (fig. 1). EFS (20 Hz, 5 sec) of atropine-pretreated and PGF₂-precontracted strips caused a biphasic response; a relaxation followed by a contraction. Addition of L-NAME (10⁴ M) reduced the amplitude of the relaxation to 15.1 ± 2.7% (n = 16) of control but had no effect on the contraction (fig. 2a). L-arginine (10⁴-10³ M) (n = 8) reversed the effect of L-NAME by restoring the relaxation to 82.5 ± 7.1% of control (fig. 2a). MB (3 × 10⁵ M) (n = 20), ODQ (3 × 10⁶ M) (n = 16) or forskolin (3 × 10⁷ M) (n = 8) affected neither the EFS-induced relaxations nor the contractions (figs. 2b and c) while TTX (10⁶ M) (n = 16) abolished both responses.

View larger version (8K): [in this window] [in a new window]   Fig. 1.   Tracings of electrically induced contractions of rat ileal longitudinal muscle. Stimulation (4 Hz, 4V, 400 mA) was maintained for 5 sec (vertical bars). The contractile response was unaffected by addition of L-NAME but was blocked by atropine.

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Tracings of electrically induced contractions of precontracted rat ileal longitudinal muscle. Stimulation (20 Hz, 4V, 400 mA) was maintained for 5 sec (indicated by vertical bars). In strips at basal tension EFS evoked a biphasic contractile response. After precontraction by PGF2 and in the presence of atropine a relaxation followed by a contraction was evoked by EFS. The EFS-induced relaxation was abolished by L-NAME but could be reversed by addition of L-arginine (a). The EFS-induced relaxation was unaffected after pretreatment with ODQ (b) or forskolin (c).

NO modulation of cholinergic contractions. Continuous stimulation (4-10 Hz, 4V, 400 mA, 4 min) evoked a stable contraction (fig. 3a) which could be blocked by atropine (n = 10). Addition of SNAP (3 × 10⁵ M) (n = 12) resulted in a quick and transient relaxation after which the contraction was restored to the same magnitude as before (fig. 3a). When ACh (10⁶ M) was used to achieve the same level of contraction as that elicited by EFS, addition of SNAP (3 × 10⁵ M) (n = 11) evoked the same quick transient relaxation followed by restoration of the contraction to the same magnitude (fig. 3b).

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Tracings showing NO modulation of cholinergic responses of rat ileal longitudinal muscle. Effect of SNAP on contractions evoked by (a) continuous EFS (10 Hz, 4V, 400 mA) or (b) acetylcholine (Ach). Addition of SNAP caused a transient relaxation in both experiments.

Responses evoked by NO donors. Both NO donors (SNP and SNAP) induced a biphasic response, a contraction followed by a relaxation, on precontracted muscle strips. At low concentrations of SNAP the contractile effect dominated although at higher concentrations (> 10⁶ M) the relaxation predominated (figs. 4 and 5a). However, with SNP (10⁷-10³ M) both the contraction and the relaxation increased in a concentration dependent fashion (fig. 5b). Strips at basal tension (3-5 mN) did neither contract nor relax when exposed to SNP (n = 12) or SNAP (n = 13). Of the two NO donors, SNAP was the most potent and gave the most reproducible results and it was therefore chosen for the following experiments.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Tracings of rat ileal longitudinal muscle showing responses to different concentrations of SNAP on strips precontracted with PGF2. SNAP (10[sup]6 M) evoked a contraction (a), SNAP (5 × 10[sup]6 M) evoked both a contraction and a relaxation (b) although SNAP (10[sup]4 M) evoked relaxation only (c).

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Concentration-response curves showing the contractile and the relaxatory effects of SNAP (a) and SNP (b) on rat ileal longitudinal muscle. a, SNAP evoked contractile effects are shown in the upper panel and relaxatory effects in the lower panel. Presence of PACAP-27 or forskolin almost totally abolished the contractions but caused a marked potentiation of the relaxations. VIP and NPY did not affect the contractile or the relaxatory responses. b, SNP evoked contractile effects are shown in the upper panel and relaxatory effects in the lower panel. Each value is the mean of 12 to 24 experiments. Vertical bars give S.E.M. *P < .05, **P < .01, ***P < .001.

View larger version (44K): [in this window] [in a new window]   Fig. 6.   Contractile effects of SNAP (10[sup]6 M) (a) and relaxatory effects of SNAP (3 × 10[sup]5 M) (b) on rat ileal longitudinal muscle. a, Presence of atropine (10[sup]6 M), Spantide (10[sup]6 M), TTX (10[sup]6 M), NPY (10[sup]7 M) or VIP (10[sup]7 M) did not affect the SNAP evoked contraction although MB (3 × 10[sup]5 M), ODQ (3 × 10[sup]6 M), PACAP-27 (10[sup]7 M), PACAP-38 (10[sup]7 M) or forskolin (3 × 10[sup]7 M) effectively inhibited it. b, Presence of atropine (10[sup]6 M), Spantide (10[sup]6 M), TTX (10[sup]6 M), NPY (10[sup]7 M) or VIP (10[sup]7 M) did not affect the SNAP evoked relaxation, MB (3 × 10[sup]5 M) and ODQ (3 × 10[sup]6 M) caused a marked inhibition although PACAP-27 (10[sup]7 M), PACAP-38 (10[sup]7 M) and forskolin (3 × 10[sup]7 M) caused potentiation. Each value is the mean of 18 to 24 experiments. Vertical bars give S.E.M. **P < .01, ***P < .001.

6

6

6

6

5

5

6

6

5

7

7

7

7

7

7

7

	Discussion

Top Abstract Introduction Methods Results Discussion Conclusions References

Our results together with those of Bartho and Lefebvre (1995) suggest that NO is involved in both relaxatory and contractile responses in rat ileal longitudinal muscle. Because L-NAME almost totally abolished the EFS-induced relaxation of atropine -pretreated and precontracted muscle strips, and because L-arginine restored the relaxatory response it is probably mediated by neuronal release of NO. The relaxatory response induced by the NO donor SNAP was unaffected by atropine, Spantide or TTX suggesting a direct effect of NO (or a NO-related substance) on the smooth muscle. This is well in line with results from a number of investigations on various gastrointestinal regions from different species (see reviews Konturek and Konturek, 1995, Makhlouf and Grider, 1993, Sanders and Ward, 1992, Stark and Szurszewski, 1992, Whittle 1994). NO-evoked relaxations are generally thought to be mediated by an elevation of intracellular cGMP. Our study shows, however, that the inhibitors of soluble guanylate cyclase MB and ODQ, had no effect on the NO-mediated EFS-induced relaxation although significantly reducing the relaxatory response to the NO donor SNAP. Similar results have been presented on the mouse cecum, longitudinal muscle. In this preparation ODQ totally blocked the SNP induced relaxation but only a partial block of the EFS-induced NO-mediated relaxation was found (Young et al., 1996). Both MB and, in particular, ODQ were proven to be efficient tools when assessing the mechanisms behind NO evoked responses. A clear inhibition by both MB (Furchgott and Jothianandan 1991) and ODQ (Brunner et al., 1996, Moro et al., 1996) of cGMP accumulation induced by NO donors has been shown. The explanation for the failure of MB and ODQ to block the neuronally evoked NO-mediated relaxation may be that it is independent of a rise in intracellular cGMP. One possible mechanism behind the cGMP-independent NO-mediated relaxation is through hyperpolarization of smooth muscle cells via a direct activation of K⁺ channels. Koh et al. (1995) provided evidence for this by demonstrating that NO can directly, and independently of cGMP, activate at least two types of K⁺ channels (K_NO1 and K_NO2) in colonic smooth muscle cells. The authors further suggested that NO-induced hyperpolarization is initiated by the direct activation of K⁺ channels and maintained by a cGMP-dependent mechanism. The ability of NO to act directly on K⁺ channels has been suggested to occur also in rat pulmonary artery (on voltage-gated K⁺ channels) (Yuan et al., 1996) and in porcine trachea (on Ca⁺⁺-activated K⁺ channels) (Yamakage et al., 1996).

The relaxatory response induced by neuronal NO release does not exclude the participation of other inhibitory transmitters. Relaxations induced by low frequency stimulation have been found to be more dependent on NO than those evoked by high frequency stimulation which have been found to involve also VIP (Li and Rand, 1990).

The contractions induced by NO in rat ileum have been suggested to be mediated via Ca⁺⁺ influx through L-type calcium channels since nifedipine, but not MB, abolished them (Bartho and Lefebvre, 1995). In our study both MB and ODQ abolished the SNAP-evoked contraction suggesting that it is mediated via guanylate cyclase activation. This has previously been found also in opposum esophageal longitudinal muscle, where NO donors caused contractions that were abolished by MB (Saha et al., 1993). It was enigmatic that contractions were revealed only in precontracted muscle strips and not in those at a basal tension. This is in contrast to the findings of Bartho and Lefvebre (1995) who registered a contraction of strips both precontracted and at a basal tension of 3 mN using NO gassed solutions. A possible explanation for this discrepancy may be that NO donors liberate NO at a different redox state than solutions of NO gas or NO released from neurons and therefore activate different second messenger systems.

The EFS-induced response in precontracted rat ileal strips has been suggested to be triphasic: a slow contraction followed by a relaxation again followed by a contraction (Bartho and Lefebvre, 1995). The first contraction was claimed to be neuronally mediated and due to NO release since it could be blocked by L-NAME, while in the relaxatory response NO was thought to play a minor role (poorly blocked by L-NAME). In our experiments EFS did not reveal any neuronally mediated contractions that could be attributed to NO since L-NAME was without effect. In guinea pig intestine acetylcholine has been found to be released in response to exogenously applied NO (Bartho and Lefebvre, 1995). The contractions evoked by NO donors in the present study were not blocked by atropine, Spantide or TTX suggesting that they are not neuronally mediated. In our experiments the NO donor SNAP did not stimulate the release of acetylcholine induced by EFS, neither was NOS blockade by L-NAME found to elicit any excitatory responses.

In our study pretreatment of the rat ileal strips with VIP or NPY affected neither the contractions nor the relaxations induced by SNAP. However, PACAP as well as forskolin effectively potentiated the SNAP-induced relaxations and almost totally inhibited the SNAP-induced contractions.

PACAP-containing enteric neurons have been demonstrated in several mammals, including the rat (Sundler et al., 1991). The fibers are distributed in association with smooth muscle as well as within the enteric ganglia. PACAP is a potent relaxant (50 times as potent as VIP) of rat ileal strips through smooth muscle receptors that are distinct from those mediating VIP responses (Katsoulis et al., 1993a), although in the guinea pig ileum PACAP causes contractions via release of acetylcholine and substance P (Katsoulis et al., 1993b). Excitatory PACAP receptors have been shown to be linked to adenylate cyclase in guinea pig small intestinal AH/type 2 myenteric neurons (Christofi and Wood, 1993). In guinea pig taenia coli PACAP-27 causes relaxation with a concomitant increase in cAMP but without any increase in cGMP (McConalogue et al., 1995). In contrast to PACAP (and forskolin) NO can increase intracellular cGMP levels via activation of soluble guanylate cyclase (Furchgott and Jothianandan, 1991). Thus, a rise in both cAMP and cGMP leads to a relaxation of gastrointestinal smooth muscle. The mechanisms behind the potentiation by PACAP and forskolin of the SNAP-induced relaxations are enigmatic but indicate that a rise in cAMP facilitates the relaxations evoked by NO donors. This has been suggested to occur also in human bronchi where a rise in cAMP by pretreatment with rolipram (a cAMP-specific phosphodiesterase inhibitor) potentiated the relaxations induced by NO donors (Fernandes et al., 1994). The potentiation of NO evoked responses by substances known to increase the levels of cAMP is probably due to an increased accumulation of cGMP after NO stimulation. Forskolin has been shown to potentiate SNP stimulated cGMP accumulation in slices of cerebellum from guinea pig (Hernandez et al., 1994) as well as in the palmar lateral vein of pig (Wright et al., 1994). Further, a rise in cAMP has been shown to enhance NO formation by a Ca⁺⁺-dependent NOS in a neuronal cell line (Reiser, 1992) and by inducible NOS in vascular smooth muscle (Hirokawa et al., 1994, Imai et al., 1994). Induction of NOS by increased levels of cAMP may also occur in the gut (Murthy and Makhlouf, 1994). However, this cannot entirely explain the facilitating effect of PACAP and forskolin on NO-mediated relaxations because this was registered only after application of NO donors. It is of interest in this context that forskolin did not potentiate the EFS-induced NO-mediated relaxation. This relaxation was found to be independent of a rise in intracellular cGMP (not blocked by MB or ODQ). Therefore, it seems that a rise in cAMP facilitates only the cGMP-dependent part of the NO-evoked relaxation.

Our study revealed that the SNAP-evoked contraction could be inhibited by both PACAP-27 and -38 as well as by forskolin. The mechanisms behind this inhibition are unclear. Because the SNAP-induced NO-mediated relaxation is markedly facilitated by cAMP-stimulating agents the relaxatory response may predominate to the extent that contraction is prevented. Alternatively, the rise in cAMP abolishes the release of excitatory substances such as eicosanoids, which in some preparations have been found to mediate the NO-evoked contraction (Irie et al., 1994, Saha et al., 1993).

	Conclusions

Top Abstract Introduction Methods Results Discussion Conclusions References

NO is involved in both relaxatory and contractile responses in innervated rat ileal longitudinal muscle strips. Relaxations due to neuronal release of NO are mediated via a direct effect on the smooth muscle that is independent of a rise in intracellular cGMP (not blocked by ODQ or MB). The relaxations evoked by NO liberated from NO donors (SNAP) involve a cGMP-dependent mechanism (partially blocked by MB or ODQ). The SNAP-induced relaxation is markedly potentiated by forskolin, PACAP-27 and PACAP-38, but not by VIP or NPY, although the EFS-induced (mainly NO-mediated) relaxation is unaffected by forskolin pretreatment. The mechanisms behind the NO-induced contraction are poorly understood; it is cGMP dependent (blocked by MB and ODQ), but it is evoked only by NO donors. It is abolished by PACAP-27, PACAP-38 and forskolin, but unaffected by VIP or NPY.