Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

It has been shown that three distinct enzymes, AsA, PsE and CsD, play important roles in the inactivation of exogenously added met-enk in three types of isolated preparations: guinea pig ileum (Aoki et al., 1984), mouse vas deferens (Aoki et al., 1986) and rat vas deferens (Cui et al., 1986). The close proximities of these enzymes to the opioid receptors in guinea pig ileum (Aoki et al., 1984) and mouse vas deferens (Aoki et al., 1986) suggest that they play important roles in terminating the physiological action of endogenous met-enk as well. Additionally, when ileal and striatal membrane fractions are incubated with met-enk for 60 min at 37°C in the presence of three PIs (amastatin, captopril and phosphoramidon), approximately 98% and 94%, respectively, of met-enk remains intact (Hiranuma and Oka, 1986). This shows that met-enk is almost exclusively hydrolyzed by the three enzymes AsA, PsE and CsD, at least in these membrane preparations. Moreover, the s.c. administration of met-enk was shown to produce two naloxone-reversible effects: inhibition of the tail-flick response and loss of the righting reflex in 10-day-old rats pretreated with all three of these PIs, but not in those pretreated with any combination of two PIs (Oka et al., 1992). This indicates that all three enzymes also play important roles in the inactivation of met-enk after its systemic administration.

However, the enzymes other than met-enk that are involved in the inactivation of endogenous opioid peptides have not been thoroughly investigated. In the case of met-enk-RGL, an endogenous opioid peptide derived from proenkephalin A (Noda et al., 1982), its potency in guinea pig ileum, mouse vas deferens and rat vas deferens has been shown to be significantly increased by pretreatment of the preparations with the combination of four PIs (McKnight et al., 1983). This result suggests the involvement of several peptidases in its inactivation. Additionally, a biochemical study (Norman and Chang, 1985) has indicated that the carboxypeptidase activities, especially that of angiotensin-converting enzyme, participate in the degradation of met-enk-RGL when it is incubated with synaptic membranes from rat striatum. Although these reports have shown the important aspects of met-enk-RGL degradation, it was still unclear how to prevent completely the hydrolysis of met-enk-RGL. This made it impossible to determine the potency of met-enk-RGL accurately, inasmuch as the determinations were made on partially degraded peptide. Therefore, in the present investigation, we explored methods to protect met-enk-RGL completely from degradative hydrolysis. Upon establishing a method that affords the peptide complete protection, we extended our study to determine the potency of met-enk-RGL accurately and compared it with the potencies of representative opioid agonists.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Chemicals. Captopril was kindly provided by Sankyo Company (Tokyo, Japan). YGGFMRG was a gift from Meiji Seika Kaisha Ltd. (Yokohama, Japan). Other reagents were purchased from the following sources: YGGFMR, met-enk-RGL, DAMGO and CTOP (Peninsula Laboratories, Inc., Belmont, CA); Y, YGGFM, amastatin, phosphoramidon and dynorphin-(1-8) (Peptide Institute, Inc., Minoh, Japan); YG, YGG and YGGF (BACHEM Feinchemikalien AG, Bubendorf, Switzerland) and morphine HCl (Sankyo Company, Tokyo, Japan).

Preparation of membrane fractions. Male Hartley guinea pigs weighing 400 to 600 g were used for this study. The myenteric plexus-longitudinal muscle strip of guinea pig ileum was prepared as previously described (Oka et al., 1982). The striatum of the guinea pig brain was dissected by the method of Glowinski and Iversen (1966). The strips or striata were cut with scissors into small pieces in 50 mM Tris-HCl buffer, pH 7.4, and then homogenized in 15 volumes of this buffer with a Teflon-glass homogenizer. The homogenate was diluted to a 3% suspension and centrifuged at 800 × g for 15 min. The supernatant was collected. The pellet was resuspended with 5 ml of the buffer, and the suspension was centrifuged again at 800 × g for 15 min. The supernatant was combined with the initial supernatant. The combined supernatants were centrifuged at 49,000 × g for 10 min. The resulting pellet was washed three times by suspension in the buffer and centrifugation. The washed pellet was suspended and adjusted with the buffer to a concentration of 2 mg/ml of protein, which yielded the sample used in the following experiment as a membrane fraction. Protein concentrations were measured by the method of Lowry et al. (1951), with bovine serum albumin as the standard.

Hydrolysis of met-enk-RGL. The sample of a membrane fraction (0.5 mg of protein) in 50 mM Tris-HCl buffer, pH 7.4, was incubated with met-enk-RGL (5 nmol) in either the absence or the presence of the PIs for various times at 37°C in a total volume of 0.5 ml. The reaction was stopped by the addition of 0.5 ml of 10% trichloroacetic acid, followed by centrifugation at 3000 rpm for 15 min. A sample (10-20 µl) of the supernatant was then analyzed by HPLC-ECD to assess the product formation.

Separation and detection of met-enk-RGL and its hydrolysis products by HPLC-ECD. The apparatus consisted of a PM-60 pump (Bioanalytical Systems, Inc., West Lafayette, IN), a µBondasphere 5-µm C18-1000A column (3.9 × 150 mm) (Waters/Nihon Millipore, Ltd., Tokyo, Japan) and an LC-4B electrochemical analyzer (Bioanalytical Systems). The chromatographic mobile phase was 0.1 M phosphate buffer at pH 3.0 containing 2.5% acetonitrile for the Y, YG and YGG assays (fig. 1A); that containing 10% acetonitrile for the YGGF, YGGFM, YGGFMR and YGGFMRG assays (fig. 1B) and that containing 15% acetonitrole for the met-enk-RGL assay (fig. 1C), at a flow rate of 0.8 ml/min. The ECD worked at an applied voltage of 1 V vs. the Ag/AgCl reference electrode. Under these conditions, Y, YG, YGG, YGGF, YGGFM, YGGFMR, YGGFMRG and met-enk-RGL exhibited retention times of approximately 3.58, 5.38, 4.77, 5.50, 12.1, 6.62, 14.9 and 7.82 min, respectively (fig. 1, A, B and C), and the detection limit of each compound was approximately 2 pmol. The recovery of met-enk-RGL and those of its hydrolysis products added to the heat-treated membrane fraction (100°C for 5 min) were all above 95%. Therefore, we estimated the content of each compound without correcting for percent of its recovery.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Typical chromatograms of met-enk-RGL and its seven hydrolysis products. The chromatograms were produced by applying 10 µl of the standard solution containing 10 pmol of each compound. The chromatographic mobile phase was 0.1 M phosphate buffer (pH 3.0) containing 2.5% acetonitrile for Y (1), YG (3) and YGG (2) assays (panel A); that containing 10% acetonitrile for YGGF (4), met-enk (6), YGGFMR (5) and YGGFMRG (7) assays (panel B); and that containing 15% acetonitrile for met-enk-RGL (8) assays (panel C), at the flow rate of 0.8 ml/min.

Calculations. The content of each compound was determined by comparing the peak heights in the sample with the standard. Working standards were run at regular intervals to check the sensitivity of the detector. Y was always detected, even when the sample of a membrane fraction was incubated in the absence of met-enk-RGL for 60 min (Hiranuma and Oka, 1986). Therefore, the amount of Y in the absence of met-enk-RGL was always subtracted from that in the presence of met-enk-RGL. In contrast to Y, met-enk-RGL and its hydrolysis products other than Y were undetectable in any of the samples incubated in the absence of met-enk-RGL.

In vitro isolated preparation. The myenteric plexus-longitudinal muscle strip of guinea pig ileum was set up for electrical stimulation as described previously (Oka et al., 1982). The percent inhibition of the stimulated muscle twitch produced by an opioid was plotted against the log concentration of the opioid to estimate the IC₅₀ (opioid concentration producing 50% inhibition of the twitch). When the effect of PIs on the IC₅₀ value of an opioid peptide was studied, they were given 10 min before the administration of the opioid peptide. The ratio of the potency and the percent difference, shown in the tables, were calculated from the following formulas: ratio of potency = IC₅₀ before additional treatment/IC₅₀ after additional treatment, and % difference = [(IC₅₀ before additional treatment  IC₅₀ after additional treatment)/IC₅₀ before additional treatment] × 100. The statistical significance of percent differences between IC₅₀ values of two adjacent groups shown in the tables was determined by the paired Student's t test. The K_e (equilibrium dissociation constant) value of CTOP against the opioid agonists was determined by the "single dose" method of Kosterlitz and Watt (1968). Because the potency of an opioid in one preparation was sometimes significantly different from its potencies in the others, the same preparation was employed to carry out the experiment on the relative potency of opioids.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

The time course of the met-enk-RGL hydrolysis and the generation of hydrolysis products during the incubation with membrane fractions. Met-enk-RGL (5 nmol) was incubated at 37°C in 50 mM Tris-HCl buffer (pH 7.4) with a membrane fraction (0.5 mg of protein) prepared from either the myenteric plexus-longitudinal muscle strip of guinea pig ileum or the striatum of guinea pig brain. The contents of met-enk-RGL and its seven hydrolysis products: Y, YG, YGG, YGGF, YGGFM, YGGFMR and YGGFMRG in the reaction mixture were estimated at 0, 5, 15, 30, 45 and 60 min after the incubation began. During the initial 5-min incubation, met-enk-RGL was hydrolyzed by approximately 85% and 82% in the ileal (fig. 2A) and striatal (fig. 2B) membrane fractions, respectively, and was hydrolyzed almost completely during the subsequent 10-min incubation in both membrane preparations (figs. 2, A and B). The major hydrolysis products during the initial 15-min incubation were YGGFMR, YGGF and Y in both preparations (fig. 2, A and B), which indicates that the dipeptidyl carboxypeptidase and aminopeptidase activities are mainly involved in the met-enk-RGL hydrolysis.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   The time course of the change in the contents of met-enk-RGL () and its seven hydrolysis products Y (), YG (), YGG (), YGGF (), YGGFM (), YGGFMR () and YGGFMRG () in the reaction mixture during the incubation of met-enk-RGL (5 nmol) with either the ileal (panel A) or the striatal (panel B) membrane fraction (0.5 mg of protein each). Each point represents the mean of three determinations. Each standard error was within 10% of the mean for both panels A and B.

Effects of the PIs on the hydrolysis of met-enk-RGL by the ileal membrane fraction. When the met-enk-RGL (5 nmol) was incubated with the ileal membrane in the absence of PI for 60 min at 37°C, Y, YG, YGG and YGGF were detected in the sample, whereas YGGFM, YGGFMR, YGGFMRG and met-enk-RGL were not. The percent production of Y, YG, YGG and YGGF was 76.3, 7.2, 1.3 and 13.3, respectively (table 1). Additionally, the percent generation of the met-enk-RGL hydrolysis products in the presence of phosphoramidon (1 µM) was quite similar to that in the absence of the PIs, except that the YGG fragment was undetectable in the presence of phosphoramidon (table 1). In contrast to the results with phosphoramidon, the presence of 1 µM amastatin decreased the percent production of free Y to 1.5%, increased both the YG and the YGGF fragments to 54.3% and 28.2%, respectively, and left 3% of the met-enk-RGL intact (table 1). In the presence of 1 µM captopril, the percent formation of free Y was decreased to 59.1%, and 33.4% of met-enk-RGL remained intact (table 1). In the presence of both amastatin and captopril, the main hydrolysis product was the YGG fragment, and 56.3% of the met-enk-RGL remained intact (table 1). When amastatin, captopril and phosphoramidon were all present, 95% of the met-enk-RGL remained intact (table 1), which indicates that the hydrolysis of met-enk-RGL by the ileal membrane fraction was almost completely prevented by the presence of the three PIs.

Effects of the PIs on the hydrolysis of met-enk-RGL by the striatal membrane fraction. The hydrolysis products of met-enk-RGL formed during a 60-min incubation with the striatal membrane fraction were qualitatively similar to, but quantitatively different from, those yielded by the ileal membrane fraction in either the absence or the presence of the PIs (tables 1 and 2). The formations of both the YG and the YGGF fragments in the absence of the three PIs, and that of the YG fragment in the presence of either amastatin or phosphoramidon, after incubation with the striatal membrane fraction were higher than those observed with the ileal membrane fraction (tables 1 and 2). In contrast to the YG fragment, the free Y generation in the absence of the PIs and in the presence of either captopril or phosphoramidon after incubation of met-enk-RGL with the striatal membrane fraction was lower than that with the ileal membrane fraction (tables 1 and 2). On the other hand, the amount of met-enk-RGL remaining intact 60 min after incubation with the striatal membrane fraction was higher than that with the ileal membrane fraction in the presence of either captopril alone or both amastatin and captopril (tables 1 and 2). In the presence of amastatin, captopril and phosphoramidon, however, 95% of the met-enk-RGL remained intact after a 60-min incubation with either the ileal or the striatal membrane fraction (tables 1 and 2).

Effects of PI on the potency of met-enk-RGL in guinea pig ileum. The inhibitory effect of met-enk-RGL on the electrically evoked contractions of the myenteric plexus-longitudinal muscle preparation of guinea-pig ileum was significantly enhanced by pretreatment of the preparation with either amastatin or phosphoramidon, but not captopril (table 3). The magnitude of the enhancement by amastatin was higher than that by phosphoramidon (table 3).

Potencies of YGGFMR and YGGF relative to that of met-enk-RGL in guinea pig ileum. We examined the potencies of the YGGFMR and YGGF fragments, the two main hydrolysis products, at 5 min after the incubation of met-enk-RGL with the ileal membrane fraction (fig. 2A), relative to that of met-enk-RGL in guinea pig ileum pretreated with three PIs (amastatin, captopril and phosphoramidon) at the final concentration of 1 µM each. Pretreatment of the preparation with the three PIs significantly decreased the respective IC₅₀ value of met-enk-RGL, YGGFMR and YGGF in guinea pig ileum (data are not shown). In the presence of the three PIs, the YGGFMR fragment was approximately 3-fold more potent than met-enk-RGL, and the potency of the YGGF fragment was about 3% that of met-enk-RGL (table 4). Furthermore, the K_e values (mean ± S.E. of four experiments) of CTOP, a selective µ-opioid receptor antagonist (Pelton et al., 1986; Gulya et al., 1986), against the selective µ-agonist DAMGO (Kosterlitz and Paterson, 1981; Handa et al., 1981), the endogenous -agonist dynorphin(1-8) (Corbett et al., 1982; Oka et al., 1983) and the three test compounds met-enk-RGL, YGGFMR and YGGF, in guinea-pig ileum pretreated with the three PIs, were 16.0 ± 1.1, 589 ± 230, 35.0 ± 6.4, 23.9 ± 4.7 and 23.9 ± 2.2, respectively, indicating that the three test compounds acted on µ-receptors.

Potencies of DAMGO, morphine and met-enk relative to that of met-enk-RGL in guinea pig ileum. Before the present study, the potency of met-enk-RGL had not yet been estimated because its degradation was not completely prevented. For this reason, and because met-enk-RGL was shown to act on µ-receptors in guinea pig ileum, we compared the potency of met-enk-RGL in guinea pig ileum pretreated with the three PIs with the potencies of representative µ-opioid agonists. Pretreatment of the guinea pig ileum with the three PIs significantly decreased the IC₅₀ value of met-enk-RGL and that of met-enk but did not significantly change the IC₅₀ value of either DAMGO or morphine (data not shown). In the presence of the three PIs, the potencies (mean ± S.E. of four experiments) of DAMGO, a selective peptide µ-agonist; morphine, a representative alkaloid µ-agonist; and met-enk, a representative endogenous opioid peptide acting on µ-receptors in guinea pig ileum (Lord et al., 1977), were found to be 292%, 23.2% and 162% that of met-enk-RGL, respectively (table 5).

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

We examined the hydrolysis of met-enk-RGL by incubating with the membrane fraction prepared from the myenteric plexus-longitudinal muscle strip of guinea pig ileum or guinea pig striatum in either the absence or the presence of the PIs. Our results show that met-enk-RGL is completely hydrolyzed during the initial 15-min incubation in the absence of the PIs but that in the presence of three PIs, its hydrolysis is almost completely prevented during the 60-min incubation. Until now, the potency of met-enk-RGL has not been accurately estimated because of its susceptibility to degradation. Employing the three PIs in the present study to eliminate hydrolytic inactivation of met-enk-RGL, we have been able to determine its potency and compare it with the potencies of representative opioids. The present results demonstrate that met-enk-RGL, which acts on µ-receptors in guinea pig ileum, is several-fold more potent than morphine.

The fact that more than 50% of the met-enk-RGL molecules are hydrolyzed to the YGGFMR fragment during the first 5 min of incubation in both the ileal and the striatal membrane fractions indicates that the dipeptidyl carboxypeptidase activity is mainly involved in the hydrolysis of met-enk-RGL, a result consistent with the finding reported by Norman and Chang (1985). Additionally, the observation that the amount of the YGGFMR fragment decreases, while that of the YGGF fragment concomitantly increases, from 5 to 15 min after initiation of the incubation indicates that the YGGFMR and the YGGF fragments are sequentially produced in this order by the dipeptidyl carboxypeptidase.

The dipeptidyl carboxypeptidase activity in both membrane fractions can be identified as CsD by the fact that YGGF fragment generation during the 60-min incubation was almost completely prevented by the presence of captopril, a specific inhibitor of dipeptidyl carboxypeptidase I (Rubin et al., 1978). Similarly, the enzyme activity producing the free Y and the YGG fragment can be identified as AsA and PsE, respectively, because the production of the free Y and the YGG fragment was almost completely prevented by amastatin, a potent inhibitor of aminopeptidase (Matsas et al., 1985), and phosphoramidon, a specific inhibitor of endopeptidase-24.11 (Matsas et al., 1983), respectively.

The time course of the met-enk-RGL hydrolysis, illustrated in figure 2, shows that although the involvement of CsD in the met-enk-RGL hydrolysis is the highest among the three peptidases, CsD is less involved in the YGGFMR hydrolysis than AsA, because the main substrate for peptidases from 15 to 30 min after initiation of the incubation is the YGGFMR fragment, and the amount of the YGGF fragment production is lower than that of the free Y generation during this period. Interestingly, a previous study (Hiranuma and Oka, 1986) has demonstrated that CsD is significantly less involved than AsA in met-enk (YGGFM) hydrolysis in both membrane preparations.

In contrast to CsD and AsA, the magnitude of the involvement of PsE in the met-enk-RGL hydrolysis is quite low when both CsD and AsA are not inhibited. When met-enk-RGL is incubated with the ileal or the striatal membrane fraction for 60 min in the presence of both amastatin and captopril, however, 39% or 20%, respectively, of met-enk-RGL is hydrolyzed to the YGG fragment. This indicates that when both CsD and AsA are inhibited, PsE is able to hydrolyze a significant amount of met-enk-RGL. The involvement of PsE in the hydrolysis of met-enk-RGL is also supported by the observation that when enkephalin octapeptide is incubated with the ileal and with the striatal membrane fractions for 60 min in the presence of the three PIs, (amastatin, captopril and phosphoramidon), approximately 95% of the met-enk-RGL remains intact in both cases.

Because the products of the hydrolysis of met-enk-RGL by either AsA or PsE, such as free Y and the YGG, [des-Y]-met-enk-RGL and [des-YGG]-met-enk-RGL fragments, are suggested to have very low, if any, agonist activity against opioid receptors (Morley, 1980), the potency of met-enk-RGL is expected to be decreased by its hydrolysis by these two peptidases. In fact, the potency of met-enk-RGL in guinea pig ileum has been shown to be significantly increased by either amastatin or phosphoramidon in the present investigation. Additionally, the pharmacological evidence obtained in the present study that the magnitude of the enhancement of the met-enk-RGL potency by amastatin is significantly higher than that by phosphoramidon is consistent with the present biochemical data indicating that AsA is involved in the met-enk-RGL hydrolysis to a significantly higher degree than PsE.

In contrast to the hydrolysis of met-enk-RGL by AsA and PsE, the change in the potency of met-enk-RGL after its hydrolysis by CsD, which was shown to have the highest involvement in the met-enk-RGL hydrolysis among the three peptidases, depends on the balance between the rate of the metabolic change from met-enk-RGL to the YGGFMR fragment and that from the YGGFMR to the YGGF fragment. This is because the YGGFMR fragment, the initial hydrolysis product of met-enk-RGL by CsD, is approximately 3-fold more potent than met-enk-RGL, whereas the YGGF fragment, the subsequent product of hydrolysis by CsD, is approximately 30-fold less potent than met-enk-RGL in guinea pig ileum. The potency of met-enk-RGL in the presence of captopril is not significantly different from that in the absence of the PI, which indicates that two-thirds of the YGGFMR fragment produced by CsD must be further hydrolyzed either to the YGGF fragment by CsD or to the free Y by AsA, as shown in figure 2.

The electrically evoked contractions of guinea pig ileum were shown to be inhibited by µ- and -agonists but not -agonists (Lord et al., 1977). In the present study, we showed that met-enk-RGL and its hydrolysis products, the YGGFMR and YGGF fragments, act on the µ-receptors in guinea pig ileum pretreated with the three PIs. Because the potency of met-enk-RGL has yet to be estimated under conditions where its degradation is completely prevented, here we determined its potency and compared it with those of representative µ-agonists in guinea pig ileum pretreated with three PIs, a condition that totally prevents hydrolysis. The present results show that the rank order of potencies is as follows: YGGFMR > DAMGO > met-enk > met-enk-RGL > morphine  YGGF. Indeed, our preliminary experiment on the antinociceptive effect measured by the tail-immersion assay with 55°C as the nociceptive stimulus indicated that met-enk-RGL is more potent than morphine in inhibiting the tail-flick response when both opioids are given i.c.v. to rats pretreated with the three PIs.

The data obtained in the previous study (Hiranuma and Oka, 1986), in which the effects of three PIs (amastatin, captopril and phosphoramidon) on the hydrolysis of met-enk were investigated, together with those obtained in the present investigation, show that both met-enk and met-enk-RGL are almost exclusively hydrolyzed by three distinct enzymes, AsA, CsD and PsE, in both ileal and striatal membrane fractions. Additionally, our preliminary experiment showed that the hydrolysis of both [Met⁵]-enkephalin-Arg⁶-Phe⁷ and dynorphin-(1-8) in the membrane fractions is also almost completely prevented by the mixture of these three PIs. Because previous studies have shown all three enzymes to be located very close to opioid receptors (Aoki et al., 1984; Aoki et al., 1986; Cui et al., 1986; Numata et al., 1988), it would seem that they must play critical roles in the physiological inactivation of several endogenous opioid peptides.