Introduction

Abstract Introduction Methods Results Discussion References

It is well accepted that I derivatives such as clonidine and idazoxan bind to alpha-2 adrenoceptors. In recent years, it has been recognized that in addition, they interact with distinct nonadrenergic binding sites or I-preferring receptors. Radioligand binding studies have demonstrated that I receptors exist in two main subtypes: the clonidine-preferring receptor (I₁) and the idazoxan-preferring receptor (I₂) (for reviews, see Michel and Insel, 1989; Regunathan and Reis, 1996).

Recent studies have indicated that a number of drug effects previously attributed to activity at alpha-2 adrenoceptors may in fact be mediated through activation of I receptors. In particular, there is strong evidence that the antihypertensive actions of clonidine and related drugs, moxonidine and rilmenidine are primarily mediated by stimulation of central I₁ receptors in the rostral ventrolateral medulla rather than by alpha-2 adrenoceptors (Bousquet et al., 1984; Ernsberger et al., 1990, 1992; Haxhiu et al., 1994; Head, 1995; Tibirica et al., 1991). There also is evidence that activation of I₁ receptors increases renal excretion of sodium and water (Allan et al., 1993; Penner and Smyth, 1994; Smyth et al., 1992). In addition, moxonidine has been suggested to regulate gastric secretion and protect against gastric mucosal injury through activation of I₁ receptors (Carlisle et al., 1995; Glavin and Smyth, 1995).

We demonstrated that alpha-2D adrenoceptors are involved in the control of intestinal motility and fluid transport in rats (Liu and Coupar, 1996, 1997). Whether I receptors are involved in the regulation of these intestinal activities has not yet been determined. Therefore, moxonidine, which has been shown to display a much greater selectivity and affinity for I₁ receptors over alpha-2 adrenoceptors (70-600 fold, Ernsberger et al., 1992, 1993), was used to investigate how it affects intestinal motility and fluid transport. UK 14,304, which is a selective alpha-2 adrenoceptor agonist (Cambridge, 1981) with low affinity for I₁ receptors (I₁/alpha-2 affinity ratio <0.01, Bricca et al., 1993; Ernsberger et al., 1992), was used for comparison. The interactions with yohimbine, a selective alpha-2 adrenoceptor antagonist (Goldberg and Robertsson, 1983) with very low affinity for I binding sites (I₁/alpha-2 affinity ratio <0.01, Ernsberger et al., 1987, 1992; Hamilton et al., 1988, Senard et al., 1990), efaroxan, an I₁/alpha-2 receptor antagonist with 30-fold selectivity for I₁ receptors compared to alpha-2 adrenoceptors (Ernsberger et al., 1992; Haxhiu et al., 1994), and EEDQ, an irreversible alpha-2 adrenoceptor antagonist with very low affinity for I receptors (Pineda et al., 1993; Ruiz-Ortega et al., 1995), were also studied.

It has been suggested that agmatine (decarboxylated arginine) is an endogenous neurotransmitter at alpha-2 adrenoceptors and I receptors (Li et al., 1994; Gonzalez et al., 1996). Like clonidine, it binds to alpha-2 adrenoceptors of all subclasses (Bylund, 1995) and I receptors of both subclasses (Li et al., 1994; Piletz et al., 1995). Agmatine and its biosynthetic enzyme, arginine decarboxylase, have been identified in a number of mammalian tissues, including brain, stomach, intestine and aorta (Li et al., 1994; Raasch et al., 1995). However, possible functions of agmatine mediated by alpha-2 adrenoceptors and I receptors are not yet clearly apparent. We have, therefore, attempted to examine whether agmatine modulates rat intestinal activities in the in vivo and in vitro models described in this study.

	Methods

Abstract Introduction Methods Results Discussion References

Peristalsis

The method used in this study has been described in detail by Coupar (1995). Briefly, hooded Wistar male rats (250-350 g) were stunned by a blow to the head and killed by exsanguination. Segments of ileum (15 cm proximal to the caecum) were excised and tied at the aboral end onto a glass J tube. Each tissue was placed in a 30 ml jacketed organ bath containing Krebs-Henseleit solution of the following composition (in mM): NaCl, 118; KCl, 4.7; NaHCO₃, 25; KH₂PO₄, 1.2; CaCl₂, 2.5; MgSO₄, 1.2;and D-(+)-glucose, 11, maintained at 37°C and continuously gassed with 95% oxygen/5% carbon dioxide. Segments were tied off at the oral end to give lengths of between 7 and 9 cm, and the cotton thread was connected to an isotonic transducer (Ugo Basil, Varese, Italy) to monitor length changes of the longitudinal smooth muscle under a resting tension of 1 g. The free end of the J tube was connected by flexible tubing to a 5-ml float chamber half-filled with Krebs-Henseleit solution.

The float was connected to a second isotonic transducer to record volume changes related to circular muscle contraction. Both the reservoir and transducer (volume meter) were attached to a wormscrew stand, enabling them to be raised gradually above the fluid level of the organ bath to increase intraluminal pressure in the ileum and initiate peristaltic contractions. The frequency and intensities of the longitudinal and circular muscle contractions were recorded on a Grass model 79D polygraph.

After a 45-min equilibration period, the intraluminal pressure was increased gradually at 3 mm/sec until it initiated a peristaltic wave. This was recorded as a contraction of the longitudinal muscle (preparatory phase) closely followed by expulsion of fluid (circular muscle contraction). The pressure resulting in a peristaltic contraction was maintained for two periods of 15 min separated by a rest period of 10 min at zero pressure. In each period, three to four control peristaltic contractions were recorded before the agonists, moxonidine, rilmenidine or UK 14,304, were added cumulatively to the bath fluid until obvious decreases in the rate of peristaltic contractions occurred. The antagonists, yohimbine or efaroxan, added 10 min before the control contractions were present in the bathing fluid for 15 min before addition of agonist. Some tissues were pretreated with the irreversible alpha-2 adrenergic antagonist EEDQ at a concentration of 3 µM for 20 min, followed by a 20-min washout period before application of agonists. In some experiments, agmatine was tested either as an agonist or as an antagonist against UK 14,304 in the presence of the DAO inhibitor aminoguanidine 1 µM.

Intestinal Fluid Transport

Surgical and analytical procedures. The method follows that described in detail by Coupar (1985). In brief, hooded Wistar male rats (220-300 g weight) were deprived of food over night but had free access to water. Rats were anesthetized with pentobarbital sodium (60 mg/kg s.c., supplemented when necessary) and placed on an electrically heated pad at 35°C. The trachea was cannulated and a second cannula was introduced into the left jugular vein for administration of the agonist and/or antagonist, or in the case of controls, saline or vehicle at volumes of 0.1 ml/100 g. Another cannula was introduced into the left common carotid artery for constant infusions of saline as control, agmatine or VIP (.8 µg/min) in saline into the aortic arch at a rate of 40 µl/min by means of a microinjection pump (CMA/microdialysis, Stockholm, Sweden). MAP was recorded to monitor the condition of the animals from a side-arm off the carotid cannula by means of a Gould pressure transducer (Statham, Costa Mesa, CA) connected to a Neotrace (Neomedix Systems, Sydney, Australia) recorder.

Systemic drug administration. In studies to determine the effect of drug treatments on basal absorption and VIP-induced secretion, the administration sequence commenced with the i.v. injection of moxonidine, UK 14,304, vehicle or saline as the control. At 5 min later, the i.a. infusion of saline (for the basal absorption series of experiments) or VIP (for the secretion series) was commenced with associated measurement of MAP. The loops of jejunum and ileum were washed before intestinal perfusion was started at 10 min. At 30 min, the loops were removed and the perfusion fluid was collected for analysis. The segments were weighed after the removal of fluid content. The testing of the antagonists followed a similar protocol to above, except that the antagonists were injected i.v. 5 min before injecting moxonidine, UK 14,304 or saline. In some experiments, agmatine was infused i.a. at rates of 0.1, 0.3 and 1 µmol/min and continued for the length of the 20 min luminal perfusion which commenced 5 min later. Agmatine was administered by i.a. infusion to reduce the influence of metabolism by endogenous enzymes because it has been reported that agmatine is a substrate for DAO.

Drugs

Agmatine sulfate, efaroxan hydrochloride and UK 14,304 [5-bromo-6-(2-imidazolin-2-yl-amino)-quinoxaline] were obtained from Research Biochemicals (Natick, MA); moxonidine hydrochloride from Beiersdorf-Lilly (Hamburg, Germany); rilmenidine dihydrogenphosphate from Servier Laboratories (Hawthorn, Victoria, Australia); and bonyl-2-ethoxy-1,2-dihydroquinoline) from ICN Biochemicals (Aurora, OH).

EEDQ was dissolved in ethanol and then diluted in saline so that the final concentration of ethanol in the bath was <0.05%. UK 14,304 was dissolved in DMSO and further diluted with saline. Preliminary control experiments established that the bath concentration of DMSO did not affect tissue responses. Also, the dilution of DMSO (8%) did not influence basal values of absorption or VIP-induced secretion (Liu and Coupar, 1997; Hancock and Coupar 1997). All other drugs were dissolved in 0.9% saline.

Statistical Analysis

Results are presented as arithmetic mean ± S.E.M. The peristaltic inhibitory effect of the agonists was expressed as a percentage of the contraction rate present before addition of the agonists. Semilogarithmic linear regression analysis (Tallarida and Murray, 1987) was used to determine the potencies of the agonists, which are expressed as IC₅₀ values defined as the concentration causing a 50% inhibition in the rate of peristalsis. The negative logarithm of receptor dissociation constants of yohimbine and efaroxan (pK_B) were calculated using the equation: pK_B = log [B] + log (DR  1) (Furchgott, 1972), where [B] is the one point molar concentration of antagonist and DR (dose ratio) is the ratio of the agonist IC₅₀ in the presence of antagonist, divided by the IC₅₀ in the absence of antagonist. Student's unpaired t test was used to compare individual means for differences, and one-way analysis of variance followed by Dunnett's t test was used to compare treatment means with a common control and Bonferroni's test for multiple comparisons (Prism, GraphPAD Software, San Diego, USA). The criterion for statistical significance was set at P < .05.

	Results

Abstract Introduction Methods Results Discussion References

Peristalsis

In control preparations of rat ileum, a mean pressure of 4.9 ± 0.2 cm H₂O (n = 11) was required to initiate rhythmic contractions of the longitudinal muscle and associated volume expulsions. The frequency of peristalsis remained the same over 15 min at 0.66 ± 0.05 peristaltic waves/min (n = 11).

Inhibitory effect of agonists on peristalsis. The alpha-2 adrenoceptor agonist UK 14,304 (1-10 nM) and I₁/alpha-2 receptor agonists moxonidine (10-100 nM) and rilmenidine (50-300 nM) all induced concentration-dependent decreases in the rate of peristalsis with the after order of potency: UK 14,304 > moxonidine > rilmenidine. IC₅₀ values were (95% CI in parentheses): UK 14,304, 4.1 nM (2.3-7.3, n = 5); moxonidine, 35.0 nM (19.6-62.2, n = 5) and rilmenidine, 145.0 nM (91.9-228.8, n = 5). At concentrations 10 times higher than the IC₅₀ values, each agonist completely blocked the peristaltic reflex, but the inhibitory effects, however, were reversible on wash out.

Influence of yohimbine and efaroxan on the inhibitory effect of agonists. The addition of yohimbine (3 µM, 10- min incubation) or efaroxan (0.01 or 1 µM, 10-min incubation) to the bathing fluid did not influence the frequency of peristaltic contractions compared to the control rate. The influence of yohimbine and efaroxan on the inhibitory effects of moxonidine and UK 14304 is shown in figure 1. Yohimbine (3 µM) and efaroxan (0.01 and 1 µM) caused parallel rightward shifts to the concentration-response curves of moxonidine and UK 14,304. The apparent pK_B values for yohimbine against moxonidine, rilmenidine or UK 14,304 and for efaroxan against moxonidine or UK 14,304 were summarized in table 1.

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Effects of moxonidine (A) and UK 14,304 (B) at inhibiting the peristaltic reflex of the rat ileum. The peristaltic inhibition by the agonists was expressed as a percentage of the contraction rate present before addition of the agonists. Tissues were pretreated without antagonist () or with yohimbine 3 µM (), efaroxan 0.01 µM (), efaroxan 1 µM () and EEDQ 3 µM (). Mean values with S.E.M. are shown from 4 to 5 rats.

Influence of EEDQ on the inhibitory effect of moxonidine and UK 14,304. Pretreatment of rat ileum preparations with the irreversible alpha-2 adrenoceptor antagonist EEDQ (3 µM) did not influence the frequency of peristaltic contractions compared to the control. EEDQ, however, almost entirely prevented the inhibitory activities of moxonidine and UK 14,304. The maximal inhibitions were <20% at the high concentrations up to 30 µM for moxonidine and 10 µM for UK 14,304 (fig. 1).

The effect of the putative endogenous I receptor agonist agmatine. Agmatine (0.01-100 µM) did not significantly affect the rate of peristalsis. Pretreatment of tissues with the DAO inhibitor aminoguanidine (1 µM, 20 min) did not reveal any effect of agmatine on peristalsis. The frequencies of peristalsis in the presence of 100 µM agmatine were 0.55 ± 0.03 (n = 4) and 0.57 ± 0.05 (n = 5) waves/min, respectively, without and with aminoguanidine.

Fluid Transport

Absorption. Basal net fluid transport was an absorptive state. The values from the jejunum and ileum of rats infused i.a. and injected i.v. with saline as a control were 192 ± 24 (n = 16) and 362 ± 22 (n = 16) µl/g wet weight in 20 min, respectively.

Dose-dependent effect of moxonidine. Intravenous injection of moxonidine resulted in a dose-dependent increase in the rate of net fluid absorption from both jejunum and ileum (fig. 2). Moxonidine had no obvious effect on the basal rates of absorption from the jejunum or ileum at the two lower doses, 0.36 µmol/kg (0.1 mg/kg) and 1.08 µmol/kg (0.3 mg/kg), but significantly enhanced the fluid transport rate by 124 ± 21% (n = 6) in the jejunum and 45 ± 13% (n = 6) in the ileum at 3.6 µmol/kg (1.0 mg/kg). Moxonidine also produced a biphasic effect on MAP consisting of a transient hypertension followed by prolonged hypotension. The depressor effect of moxonidine remained stable during the 20-min period of intestinal perfusion. The MAP measured at the midpoint of perfusions were 110 ± 2 (.36 µmol/kg, P < .001, n = 5), 99 ± 5 (1.08 µmol/kg, P < .001, n = 6) and 85 ± 4 (3.6 µmol/kg, P < .001, n = 5) compared with the control of 130 ± 3 mm Hg (n = 12).

View larger version (34K): [in this window] [in a new window]   Fig. 2.   Effects of i.v. injection of moxonidine on the rate of net absorption from the jejunum and ileum. Rats were injected i.v. with saline (open columns), moxonidine 0.36 µmol/kg (hatched columns), 1.08 µmol/kg (cross-hatched columns) and 3.6 µmol/kg (filled columns). **, P < 0.01; ***, P < 0.001 vs. the individual control (Dunnett's t test). Values are mean ± S.E.M. of 5 to 16 rats.

Effects of moxonidine and UK 14,304 on the responses to yohimbine and efaroxan. Intravenous injection of the selective alpha-2 adrenoceptor antagonist yohimbine and I₁/alpha-2 receptor antagonist efaroxan caused a decrease in the rate of fluid absorbed from both the jejunum and ileum (table 2). Yohimbine at 10 µmol/kg (3.9 mg/kg) and efaroxan at 10 µmol/kg (2.5 mg/kg) reduced absorption to 17 ± 41 (n = 8) and 13 ± 63 (n = 5) µl/g in 20 min, respectively, in the jejunum, and 173 ± 31 and 131 ± 50 (n = 5) µl/g in 20 min, respectively, in the ileum.

The effect of the putative endogenous I receptor agonist agmatine. Infusion of agmatine into the aortic arch via the left common carotid artery resulted in an infusion-dependent decrease in the rate of net fluid absorption from both jejunum and ileum (fig. 3). Agmatine at 0.1 µmol/min (0.02 mg/min, total dose of 2.1 mg/kg) had no effect on the fluid transport rates in both jejunum and ileum (n = 5) compared to the controls, but at 0.3 µmol/min (0.07 mg/min, total dose of 6.3 mg/kg) and 1.0 µmol/min (0.23 mg/min, total dose of 21.1 mg/kg), there were significant decreases in fluid transport rates to 79 ± 20 and 95 ± 20 (n = 5) µl/g in 20 min, respectively, in the jejunum, and 232 ± 23 and 176 ± 54 (n = 5) µl/g in 20 min, respectively, in the ileum. No further reductions were observed from both jejunum and ileum at higher infusion rates of agmatine. Agmatine had no effect on MAP at all infusion rates used.

View larger version (32K): [in this window] [in a new window]   Fig. 3.   Effects of i.a. infusions of agmatine on the rate of net absorption from the jejunum and ileum. Rats were administrated with saline (i.a., open columns), agmatine 0.1 µmol/min (shaded columns), 0.3 µmol/min (hatched columns), and 1 µmol/min (filled columns). *, P < 0.05; **, P < 0.01 vs. the individual control (Dunnett's t test). Values are mean ± S.E.M. of 5 to 16 rats.

Secretion Induced by VIP

Infusion of VIP (0.8 µg/min i.a.) starting 5 min before and continuing during the 20-min perfusion totally reversed water transport from net absorption to secretion. The values were 265 ± 17 µl/g in 20 min secreted into the lumen of the jejunum and 368 ± 33 µl/g in 20 min secreted by the ileum (n = 10 for both regions, table 3).

Dose-dependent antisecretory effect of moxonidine. Intravenous injection of moxonidine caused a dose-related inhibition of the VIP-induced fluid secretion in both jejunum and ileum (fig. 4). The effect of the highest dose of moxonidine extended to complete reversal of the VIP-induced effect in the jejunum, but in the ileum, it only blocked the secretory component of the response to VIP.

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Inhibition by moxonidine (i.v.) of VIP-induced secretion in the rat jejunum and ileum. The open columns indicate the control values of fluid secretion in response to i.a. infusion of VIP at 0.8 µg/min. Hatched columns are responses of rats administrated with moxonidine at 0.036 µmol/kg, cross-hatched columns at 0.36 µmol/kg, and filled columns at 3.6 µmol/kg. ***, P < 0.01 vs. the individual control (Dunnett's t test). Values are mean ± S.E.M. of 4 to 10 rats.

Comparison of the antisecretory with the proabsorptive effect of moxonidine. Moxonidine (3.6 µmol/kg i.v.) enhanced basal absorption by 239 ± 41 µl/g in 20 min in the jejunum and by 164 ± 47 µl/g in 20 min in the ileum (n = 6, for actual values see table 2). In animals infused with VIP, the same dose of moxonidine reversed secretion by 426 ± 26 µl/g in 20 min in the jejunum and by 277 ± 11 (n = 4) in the ileum (for actual values see table 3). The antisecretory effect of moxonidine was significantly greater than the proabsorptive effect in the jejunum (P < .01, Student's unpaired t test). Although not statistically significant, the antisecretory effect of moxonidine also appeared larger than the proabsorptive effect in the ileum (P = .09).

Effects of moxonidine and UK 14,304 on the responses to yohimbine and efaroxan. Yohimbine and efaroxan did not affect the values of VIP-induced secretion in either jejunum or ileum, when tested separately at doses of 10 µmol/kg (P > .05, n = 3). The results of agonist-antagonist interaction experiments are shown in table 3, where, yohimbine and efaroxan significantly antagonized the inhibitory effect exerted by moxonidine (3.6 µmol/kg) and UK 14,304 (1 µmol/kg) on VIP-induced secretion in the jejunum and ileum (P < .001, n = 5 in both regions). The degree of antagonism of yohimbine and efaroxan was similar against moxonidine and UK 14304.

Effect of the putative endogenous I receptor agonist agmatine. The antisecretory activity of agmatine was tested at the low dose (0.1 µmol/min) because higher infusion rates decreased the rate of absorption in the rat intestine, an action contrary to that of moxonidine and UK 14,304. Infusion of agmatine, together with VIP, starting 5 min before and continuing during the 20-min perfusion did not influence the rate of secretion induced by VIP. The values were 286 ± 53 µl/g in 20 min in the jejunum (n = 5) and 345 ± 58 µl/g in 20 min in the ileum (n = 5) vs. corresponding values of 265 ± 17 and 368 ± 33 µl/g in 20 min in animals infused with VIP alone (P > 0.05, student's unpaired t test).

	Discussion

Abstract Introduction Methods Results Discussion References

The I receptors are now generally recognized as a unique set of nonadrenergic, high-affinity binding sites for a number of agents that also specifically bind to alpha-2 adrenoceptors (Regunathan and Reis, 1996). It has been suggested that I₁ receptors are important in mediating the hypotensive actions of clonidine, moxonidine and rilmenidine and may regulate renal sodium and water excretion (see the introduction). However, the involvement of I₁ receptors in cardiovascular and renal functions are not fully accepted. For example, the results of studies in conscious rats and rabbits argued in favour of alpha-2 adrenoceptors as sites of mediating hypotensive actions of clonidine, rilmenidine and moxonidine (Hieble and Kolpak, 1993; Szabo et al., 1993; Urban et al., 1994, 1995). Also, experiments conducted in conscious dogs provided no support for a role of I receptors in regulation of renal function (Evans and Anderson, 1995).

The starting point for the present study was that there are reports showing moxonidine inhibits gastric acid secretion through activation of I₁ receptors (Carlisle et al., 1995; Glavin and Smyth, 1995). Also, agmatine is an endogenous ligand for alpha-2 adrenoceptors and I receptors (Li et al., 1994) and is widely distributed in rat organs, particularly in the stomach and small intestine (Raasch et al., 1995). A series of experiments was designed to investigate whether agmatine and moxonidine could regulate functions of the intestine and, if so, determine the receptors by which they might exert their effects. The effects of moxonidine on intestinal motility and fluid transport were compared with those of the pure alpha-2 adrenoceptor agonist UK 14,304. It is reasonable to presume that if moxonidine and UK 14,304 acted through different receptors, then the profiles of their effects might be expected to differ. Moreover, it would also be expected that if the effects of moxonidine were mediated by I receptors, they would be relatively resistant to antagonism by the alpha-2 adrenoceptor antagonists yohimbine and EEDQ.

Peristalsis. It is well established that peripheral alpha-2 adrenoceptors mediate inhibition of small intestinal contraction by inhibiting acetylcholine release from enteric cholinergic nerves (Andrejak et al., 1980; Coupar and Liu, 1996; Doherty and Hancock, 1983; Wikberg, 1977). It has been shown recently that the alpha-2 adrenoceptors in the rat ileum are the alpha-2D subtype, which when activated by agonists results in inhibition of peristalsis (Liu and Coupar, 1996). The present study showed that alpha-2 adrenoceptors, but not I receptors, were responsible for moxonidine mediating inhibition of intestinal motility. First, this was substantiated by comparing IC₅₀ and relative potency values of agonists used in the present experiments with those obtained in previous reports. The IC₅₀ value of clonidine for inhibition of peristalsis was 2.8 nM (Liu and Coupar, 1996). This shows that the relative potencies for clonidine, moxonidine and rilmenidine at inhibiting the peristaltic reflex are in accordance with their relative affinities at alpha-2 adrenoceptor binding sites of the bovine ventrolateral medulla (1:12:52 vs. 1:19:47) but are quite different to their relative affinity at I₁- binding sites (vs. 1:2:6) (Ernsberger et al., 1993). Second, the affinity values (apparent pK_B values) of the selective alpha-2 adrenoceptor antagonist yohimbine against moxonidine, rilmenidine and UK 14,304 were the same, suggesting that these three compounds acted at the same receptors. Third, the pK_B values of the I₁/alpha-2 antagonist efaroxan at the low (0.01 µM) and high (1 µM) concentration against moxonidine (7.66 and 7.86) and UK 14,304 (8.03 and 7.96) were consistent. These values are correspondent to the affinity of efaroxan at the alpha-2 binding sites in bovine rostral ventrolateral medulla (pK_i = 7.72-8.22) but not at the I binding sites (pK_i = 9.82-9.96) (Haxhiu, et al., 1994). Fourth, the alkylating agent EEDQ almost abolished the inhibitory effect of both moxonidine and UK 14,304 on peristalsis. EEDQ is a useful tool because it inactivates brain alpha-2 adrenoceptors in vivo and in vitro without influencing I receptors (Pineda et al., 1993; Ruiz-Ortega et al., 1995; Szabo et al., 1996). The final evidence discounting the presence of functional I receptors in the control of peristalsis was obtained from the series of experiments using the putative endogenous ligand agmatine. In these experiments, agmatine failed to mimic the effect of moxonidine and UK 14,304, as in the case of rat vas deferens and guinea pig ileum, which are tissues containing alpha-2 adrenoceptors (Pinthong et al., 1995). It was reported by Holt and Baker (1995) that agmatine was a substrate for DAO but not for monoamine oxidase and the metabolism of agmatine by DAO was inhibited by aminoguanidine. Hence, administration of a DAO inhibitor may increase endogenous agmatine levels and lead to a detectable biological response. The potential effect of agmatine on peristalsis, however, was not revealed in the presence of aminoguanidine (1 µM). The possibility of agmatine being an alpha-2 adrenoceptor antagonist has been ruled out because the potency of UK 14,304 in suppressing the peristaltic reflex was not affected by the presence of agmatine (100 µM).

Fluid transport. The influence of the I₁/alpha-2 agonist moxonidine and putative endogenous I receptor agonist agmatine on intestinal fluid transport by the small intestine was investigated under basal conditions and during active secretion induced by intra-arterial infusion of VIP. This part of the study also supplied no functional role for I receptors in mediating fluid transport.