Glucagon-like peptide isolated from the eel intestine: effects on atrial beating

Many bioactive peptides have been found in mammalian gut, and termed brain-gut peptides. Relatively few, however, have been isolated from teleost gut (Uesaka et al., 1994a; Uesaka et al., 1994b; Uesaka et al., 1995; Uesaka et al., 1996), although the existence of various peptides has been demonstrated immunohistochemically using antisera raised against mammalian peptides (Elbal and Agulleiro, 1986; Rombout et al., 1986; Abad et al., 1987). To clarify the role of gut peptides in teleosts, it is necessary to isolate the gut peptides and to examine their effects on fish organs.

We have determined that the eel atrium provides a suitable bioassay system, because it is both easy to prepare and is sensitive to peptides. The eel atrium can be easily isolated from the heart and beats spontaneously for more than 10h in artificial saline. In addition, it has been demonstrated that the isolated eel atrium is sensitive not only to adrenaline or acetylcholine (ACh) (Yasuda et al., 1996), but also to eel neuropeptide Y (eNPY) (Uesaka, 1996). As a candidate intestinal hormone capable of enhancing atrial beating, we have isolated a glucagon-like peptide (HSQGTFTNDY¹⁰SKYLETRRAQ²⁰DFVQWLMNSK³⁰RSGGPT) from the eel intestine. This peptide is shown to exhibit positive inotropic and chronotropic actions in the atrium. The inotropism is due to an increase in intracellular free Ca²⁺ concentration ([Ca²⁺]_i).

Japanese eels (Anguilla japonica Temminck & Schlegel) taken from a commercial source were decapitated and the gut excised immediately and stored at –40°C. A boiled-water extract of eel gut (361gwetmass) was prepared following a method previously described (Uesaka et al., 1994a; Uesaka et al., 1994b). The extract was applied to Sep-Pac C18 cartridges (Millipore, Milford, MA, USA). The retained material was eluted with 50% (v/v) acetonitrile containing 10% (v/v) 2-propanol and 0.1% (v/v) trifluoroacetic acid (TFA), and the eluate was evaporated. The concentrated material was applied to a column of Toyopearl HW-40F (2.6cm×100cm; Tosoh, Tokyo, Japan) and eluted with 1moll⁻¹ acetic acid and 10% (v/v) 2-propanol at rate of 1.5mlmin⁻¹. Each fraction was assayed for its ability to enhance atrial contraction and beating rate.

Bioactive fractions were pooled and subjected to high-performance liquid chromatography (HPLC) separation (LC-6AD, Shimadzu, Kyoto, Japan). Retained material was eluted with a 50min linear gradient of 0%–90% acetonitrile containing 10% 2-propanol and 0.1% TFA, and each fraction was bioassayed. The active fractions were applied to a C18 reverse-phase column (TSK ODS-80T_M, Tosoh) and eluted with a 100min linear gradient of 15%–35% acetonitrile containing 5% 2-propanol and 0.1% TFA. The active fractions were further applied to a cation-exchange column (TSK CM-5PW, Tosoh) and eluted with a 35min linear gradient of 0–0.35moll⁻¹ NaCl in 10% 2-propanol and 20mmoll⁻¹ phosphate buffer (pH 6.8). The bioactive peak was rechromatographed on the C18 reverse-phase column (TSK ODS-80T_M) with a 50min linear gradient of 24%–34% acetonitrile containing 5% 2-propanol and 0.1% TFA. Final purification was performed using the same column under isocratic conditions, as shown in Fig.1A, to give a single peak, EI-14.

The amino acid composition of the isolated peptide was determined by precolumn derivation with phenylisothiocyanate using a PICO-TAG amino acid analysis system (Millipore). The sequence of the N-terminal 32 residues was determined directly by automated Edman degradation using a gas-phase sequencer (PPSQ-10, Shimadzu). The sequence of residues 32–36 was determined after cleavage with lysyl endopeptidase; a small sample (approximately 30pmol) was treated with 0.1moll⁻¹ Tris-HCl (pH 9) and 2moll⁻¹ urea for 6h at 37°C. The molecular mass of the entire molecule was obtained by fast atom bombardment mass spectrometry (JMX-HX110A, Jeol, Tokyo, Japan). A search for homologous sequences was performed with the Protein Information Resource database (National Biomedical Research Foundation, Washington, DC, USA).

Japanese eels, weighing approximately 220g, were kept in sea water (20°C) for more than 1 week. After decapitation, the heart was rapidly excised and the atrium was isolated on ice. It was then tied with two cotton threads and connected to a force transducer (type 451996, Sanei, Tokyo, Japan). The details of this procedure have been described previously (Uesaka, 1996; Yasuda et al., 1996). The isolated eel atrium was bathed in artificial saline consisting of (in mmoll⁻¹): 118.5 NaCl, 4.7 KCl, 3.0 CaCl₂, 1.2 MgCl₂, 1.2 KH₂PO₄, 24.9 NaHCO₃ and 10 sodium lactate, bubbled with a 95% O₂/5% CO₂ gas mixture (pH 7.4) at room temperature (24–28°C). Although the pH of fish blood is approximately pH 7.9, equilibrated with 0.5% CO₂ (Farrell, 1984), experiments in the present study were usually performed in the presence of 5% CO₂ for convenience, since we have established that the effects of adrenaline, ACh, eNPY and eel oxyntomodulin (eOXM) were similar in the presence of 5% CO₂ (pH 7.4) and 0.5% CO₂ (pH 7.9) (T. Uesaka and M. Ando, unpublished observation). In Ca²⁺-free Ringer’s solution, 3.0mmoll⁻¹ CaCl₂ was replaced with 4.5mmoll⁻¹ NaCl. After preloading by 5.9mN, spontaneous isometric contractions were converted into electrical signals by a transducer connected to a strain amplifier (6M82, Sanei) and these were recorded using an electric polyrecorder (EPR-10B, Toa, Tokyo, Japan). The rate of contraction was measured simultaneously using a tachometer (type 1321, Sanei). When examining atrial contractions in the absence of Ca²⁺, the preparation was flushed twice with a Ca²⁺-free Ringer’s solution.

The effects of adrenaline (Sigma Chemical, St Louis, MO, USA), betaxolol (Mitsubishi Kasei, Tokyo, Japan) and verapamil HCl (Wako Pure Chemical, Osaka, Japan) were also examined.

[Ca²⁺]_i in the myocardium was measured as described previously (Uesaka, 1996). Briefly, the isolated atrium was cut into pieces in Ringer’s solution and treated with collagenase (1mgml⁻¹; Wako Pure Chemical) in the same solution for 1h at room temperature. After rinsing with Ringer’s solution, the partially dispersed tissue was incubated with Calcium Green-1 acetoxymethyl ester (5μmoll⁻¹; Molecular Probes, Eugene, OR, USA) and with Pluronic F-127 (0.05%, Molecular Probes) as a surfactant for 40min. The atrial cells were stuck onto a glass slide using coverslips, and mounted in a small chamber on the stage of an inverted microscope (Axiovert 135 MTV, Zeiss, Tokyo, Japan) and superfused with normal Ringer’s solution. The fluorescent signal from Calcium Green-1 was recorded with a laser scanning confocal imaging system (MRC-600, Bio-Rad, Tokyo, Japan; dichroic reflecter 510LP; emission filter 515LP) equipped with a Kr–Ar laser (5470K, Ion Laser Technology, Salt Lake City, UT, USA; 488nm for excitation). The intensity of the fluorescence was expressed in arbitrary units ranging from 0 to 255.

Data are reported as means ± s.e.m. N represents the number of preparations. The statistical significance of difference between means was examined using a Mann–Whitney U-test. The null hypothesis was rejected for P<0.05.

Fig.1A shows the final purification profile of the peptide, temporarily given the name EI-14, and Fig.1B shows the biological activity of this peptide. It is clear that EI-14 gave a single peak (Fig.1A), indicating the homogeneity of the peptide. When EI-14 was applied to eel atrium, both the contractile force and the beating rate increased gradually after a latent period (Fig.1B). The sequence of EI-14 was analysed twice. The first trial gave the N-terminal 32 residues, although the detection of Arg¹⁸ and Trp²⁵ was scanty (Table1). The second trial confirmed the existence of Arg¹⁸ and Trp²⁵, while Arg³¹ and Ser³² were obscure. Combining these results, the N-terminal 32 residues of EI-14 was determined as HSQGTFTNDY¹⁰SKYLETRRAQ²⁰DFVQWLMNSK³⁰RS (Table1).

After digestion of EI-14 with lysyl endopeptidase, three peptides were obtained. Of these three peptides, two contained fragments of the above 32 residues; HSQGTFTNDYSKYLET and AQDFVQWLMNS. However, another peptide had the sequence (R)(S)GGPT (Table1). Comparing this sequence with the result of the second trial, this peptide was considered to be the C-terminal fragment of EI-14. We concluded, therefore, that the entire sequence of EI-14 is HSQGTFTNDY¹⁰SKYLETRRAQ²⁰DFVQWLMNSK³⁰RSGGPT. This conclusion was supported by quantitative amino acid analysis and molecular mass spectrometry.

Table2 shows the amino acid composition of EI-14, which was almost identical to that expected from the sequence analysis described above. The relative molecular mass (M_r) of EI-14 (4207±1) was almost identical to that predicted from the sequence (M_r=4207). To identify the structure, a peptide consisting of the 36 amino acid residues was synthesized using the sequence result. Fig.2 compares the synthesized peptide (S) with native EI-14 (N) on the same HPLC. The retention time of the synthesized peptide was identical to that of the native EI-14 in both reverse-phase and cation-exchange HPLC. When a mixture of native and synthesized peptide was applied, only a single peak was observed with both types of HPLC.

The isolated eel atrium continues to beat spontaneously at a constant rate for more than 10h in artificial saline. The basal atrial contractile force was 1.9±0.2mN and the beating rate was 60.5±1.9beatsmin⁻¹ (N=36). When synthesized EI-14 (10⁻⁷moll⁻¹) was added to the bathing fluid, both the contractile force and the beating rate increased gradually as in the case of the native EI-14 (Fig.1B). This increase in both the force and the contraction rate were concentration-dependent, with a threshold of 10⁻⁹moll⁻¹ and a maximal effect at 3×10⁻⁷moll⁻¹ (Fig.3). Similar enhancements in both the force and the rate were observed after blocking β₁-adrenoceptors with betaxolol. In the presence of betaxolol (10⁻⁵moll⁻¹), the effects of adrenaline were completely abolished, confirming previous results (Uesaka, 1996). Under the same conditions, however, the synthetic EI-14 (10⁻⁷moll⁻¹) increased the contractile force by 2.9±0.4mN and the beating rate by 24.0±3.3beatsmin⁻¹ (N=6, P<0.01).

[Ca²⁺]_i showed spontaneous oscillations in the eel atrium (Fig.4), with an oscillatory frequency of 20–30cyclesmin⁻¹. After application of synthetic EI-14, the peak level of [Ca²⁺]_i doubled, but the frequency tended to decrease (to 18–25cyclesmin⁻¹).

To determine whether the increase in [Ca²⁺]_i is due to an influx of Ca²⁺ from the extracellular medium or to a release of Ca²⁺ from intracellular stores, Ca²⁺ in the bathing medium was omitted. In the absence of external Ca²⁺, both the contractile force and the beating rate gradually decreased and the inotropic effect of EI-14 was markedly reduced (Fig.5B). A similar reduction in the inotropic effect of EI-14 was also observed in the presence of verapamil (10⁻⁵moll⁻¹), a Ca²⁺ channel blocker (Fig.5C). However, the positive chronotropic effect of EI-14 was retained after the inhibition of Ca²⁺ entry (Fig.5B,C).

The present study determined the primary structure of a new glucagon-like peptide (HSQGTFTNDY¹⁰SKYLETRRAQ²⁰DFVQWLMNSK³⁰RSGGPT) isolated from the Japanese eel intestine. The N-terminal 29 residues were identical to the eel glucagon isolated from the European eel pancreas (Conlon et al., 1988). Since the structure of the new peptide is similar to oxyntomodulins (OXMs) isolated from a variety of vertebrates (Table3) and since glucagon is processed from the OXM in the mammalian intestine (Plisetskaya and Mommsen, 1996), it is plausible to designate this peptide (EI-14) eel OXM (eOXM). Although Conlon et al. (1988) isolated another OXM-like peptide from European eel pancreas (Conlon et al., 1988), its sequence is slightly different from that reported in our study (see Table3). It is probable that processing of the preproglucagon protein differs between the intestine and the pancreas, as described by Plisetskaya and Mommsen (Plisetskaya and Mommsen, 1996). Eel oxyntomodulin is similar to that of alligator gar (Lepisosteus spatula) (Pollock et al., 1988b) or bullfrog (Rana catesbeiana) (Pollock et al., 1988a), but differs from that of elasmobranchs, with 73% identity with dogfish (Scyliorhinus canicula) (Conlon et al., 1994) and only 58% identity with ratfish (Hydrolagus colliei) (Conlon et al., 1987). Although the structures are not presented in Table3, Raufman et al. (Raufman et al., 1992) also isolated two truncated glucagon-like peptides from paddlefish (Polyodon spathula) consisting of 31 amino acid residues, one with 71% identity and the other 68% identity to eOXM. With the exception of the elasmobranch data, the information given in Table3 suggests that the N-terminal sequence is relatively conserved and the C-terminal sequence variable among vertebrates. The N-terminal sequence (glucagon sequence) may, therefore, be an essential structure for the action of OXMs in vertebrates.

There is general agreement that glucagon increases cardiac contractile force and heart rate. Farah and Tuttle (Farah and Tuttle, 1960) first established the positive inotropic effect of glucagon in a dog heart–lung preparation. Thereafter, the effect of glucagon on cardiac contractility was confirmed and reproduced in anaesthetised animals, isolated heart preparations and cultured heart cells (Parmley et al., 1968; Smitherman et al., 1978; Chernow et al., 1986; Iwanij and Hur, 1987; Mery et al., 1990). There have, however, been reports of species differences in the response of the heart to glucagon. Dogs, rats and cats are responsive to glucagon, whereas in guinea pig, rabbit and pig the effect of the peptide has been questioned (Prasad, 1975; MacLeod et al., 1981; Farah, 1983). The difference in response to glucagon is explained by the species difference in the endopeptidase that degrades glucagon into a miniglucagon [glucagon-(19-29), an active form of glucagon] (Pavoine et al., 1991). It has, however, been reported that the effects of glucagon on arterial blood pressure and heart rate in rats are smaller than those of glucagon-like peptide-1-(7-36) amide (Barragan et al., 1994). These results suggest that the C-terminal structure of glucagon-like peptides is also important for positive inotropic and chronotropic effects on the heart.

The eel glucagon-like peptide (eOXM) has positive inotropic and chronotropic effects on the eel atrium (Fig.1B, Fig.5A). Although these effects of eOXM are similar to those of adrenaline (Yasuda et al., 1996), eOXM appears to stimulate atrial beating independently of the β₁-adrenoceptors. In the presence of betaxolol, a β₁-adrenoceptor antagonist, the effects of eOXM remain (present study), while the effects of adrenaline are completely blocked, as previously described (Uesaka, 1996).

Eel oxyntomodulin increases [Ca²⁺]_i in the eel myocardium (Fig.4). Similar increases in [Ca²⁺]_i are observed after treatment of the eel atrium with adrenaline or neuropeptide Y (Uesaka, 1996). It is thus plausible that eOXM enhances contractility via an increase in [Ca²⁺]_i. In the eel atrium, contraction appears to be induced by entry of Ca²⁺ from the extracellular fluid, since atrial contractility is markedly reduced in Ca²⁺-free Ringer’s solution or after treatment with verapamil, a Ca²⁺ channel blocker (Fig.5). In general, heart contractions in poikilotherms can be induced by an influx of Ca²⁺ from the extracellular fluid. Electron microscopy of a number of poikilotherm hearts demonstrates both a sparsity of sarcoplasmic reticulum and an absence of transverse tubules (Santer, 1985). Furthermore, contraction of the amphibian (Bers, 1985) and teleost (Driedzic and Gesser, 1988) ventricle appears to be relatively insensitive to ryanodine, a blocker of the sarcoplasmic reticulum Ca²⁺ release. More directly, Mery et al. (Mery et al., 1990) demonstrated, using a patch-clamp technique, that the Ca²⁺ current is increased by glucagon in frog and rat ventricular myocytes. It is likely therefore, that eOXM enhances Ca²⁺ influx, enhancing [Ca²⁺]_i, which enhances the contractile force of the eel atrium. However, the chronotropic effect of eOXM is still present after Ca²⁺ entry has been inhibited with Ca²⁺-free solutions or verapamil. This indicates that the chronotropic effect of eOXM is independent of extracellular Ca²⁺. Eel oxyntomodulin might stimulate beating rate by increasing levels of other intracellular mediators, such as cyclic AMP. The fact that the oscillations in [Ca²⁺]_i are much slower than the beating rate in our preparations may be because of damage to the pacemaker cells by collagenase and/or Pluronic F-127.

Eel oxyntomodulin may act as a hormone, being produced in the intestine and targeted to the heart. It is not known what triggers release of eOXM from the intestine, but a recent finding in the rat small intestine is of value for reference. When the lumen of the small intestine is perfused with 200mmoll⁻¹ NaCl Ringer’s solution (hyperosmotic), guanylin (an intestinal peptide composed of 15 amino acid residues) secretion into the lumen increased threefold, accompanied by a slight increase in uroguanylin secretion (Kita et al., 1999). Uroguanylin is another intestinal peptide of 15 amino acid residues and is considered to be a hormone targeting the kidney and pancreas (Nakazato et al., 1998), while guanylin is a luminocrine substance (Forte and Currie, 1995). If similar factors, such as concentrated NaCl, stimulate eOXM secretion from the intestine into the circulation, eOXM may enhance the contractile force and beating rate of the heart.

This research was supported in part by Grants-in-Aid for Scientific Research (C) no. 08640864 and JSPS fellowship no. 3524 from the Ministry of Education, Science, Sports and Culture, Japan, and also by the Fisheries Agency of Japan.