Effects of Substance P and Vasoactive Intestinal Polypeptide on Gastrointestinal Blood Flow in the Atlantic Cod Gadus Morhua

The innervation of the cardiovascular system was for long considered to involve only adrenergic and cholinergic nerves: all nervous influence on cardiovascular events was explained by more or less complex interactions between adrenergic and cholinergic autonomic nerves. However, following more recent discoveries, primarily from studies of the enteric nervous system, it is evident that the control of the cardiovascular system also depends on non-adrenergic, non-cholinergic factors, notably peptidergic nerves (see Burnstock and Griffith, 1988). Immunohistochemical studies, performed mainly in mammals, have identified a number of neuropeptides in perivascular nerves. Among these are vasoactive intestinal polypeptide (VIP), with a widespread distribution in the cardiovascular system (Uddman et al. 1981; Della et al. 1983), and substance P (SP), with an especially dense distribution in vessels of the abdominal viscera in mammals (Furness et al. 1982; Barja et al. 1983) and in the toad Bufo marinus (Morris et al. 1986). Both neuropeptides produce vasodilation in several vascular beds in mammals (Said and Mutt, 1970; Hallberg and Pernow, 1975; Blitz and Charbon, 1983; Rozsa and Varro, 1987), SP probably being involved in the vasodilator response to antidromic stimulation of sensory nerves (Lembeck and Holzer, 1979).

Not much is known about the peptidergic innervation of the cardiovascular system in teleost fishes. Vessels surrounded by nerve fibres containing VTP- or SP-like immunoreactivity have been observed in the gastrointestinal tract of some fishes (Lundin and Holmgren, 1984; Bjenning and Holmgren, 1988; Burkhardt-Holm and Holmgren, 1989). In the cod (Gadus morhua), nerve fibres showing VTP-like immunoreactivity have been demonstrated surrounding both the coeliac and mesenteric arteries, which supply the gut with blood (Lundin and Holmgren, 1984; C. Bjenning and S. Holmgren, unpublished observations). SP-like immune-reactivity has, however, not been found in nerves associated with these major vessels, and only a few branches of the stomach vasculature appear to be innervated (J. Jensen, A-C. Jonsson and S. Holmgren, unpublished observations). Porcine VIP and VIP extracted from the gut of rainbow trout (Salmo gairdneri) or catfish (Ictalurus melas) produce vasodilation in the vascularly perfused intestine of the catfish (Holder et al. 1983). Furthermore, VIP induces an increased flow through the perfused gas gland of the cod (Gadus morhua) swimbladder and through isolated gill arches from the brown trout Salmo trutta (Bolis et al. 1984; Lundin and Holmgren, 1984). The cardiovascular effects of SP in fish are, however, largely unknown.

The aim of this investigation was to study the influence of VIP and SP in vivo on the cardiovascular system in an unanaesthetized teleost fish, the cod. Special interest was focused on the control of blood flow to the abdominal viscera.

Atlantic cod, Gadus morhua, of either sex and with a body mass of 900–1700 g, were used in this study. They were kept in recirculating, well-aerated sea water at 10–11 °C, and were used within 10 days of capture.

The animals were anaesthetized in MS222 (3-aminobenzoic acid ethylester methanesulphonate; Sigma; 100 mg l⁻¹) until all breathing movements ceased. They were then transferred to the operating table, and aerated sea water containing anaesthetic (50 mg 1⁻¹) was continuously passed over the gills during surgery.

A cannula (PE 50) was inserted into the afferent branchial artery of the third gill arch on the left side for measurement of ventral aortic (prebranchial) blood pressure (P_VA) and for recording the heart rate (fH). A second cannula was implanted in the efferent branchial artery of the same gill arch to measure the dorsal aortic (postbranchial) blood pressure (P_DA). Both cannulae were filled with heparinized (100 i.u. ml⁻¹) 0.9% NaCl and secured with skin sutures. During the experiment, the cannulae were attached to Statham P23 pressure transducers connected to a Grass polygraph (model 7D) for recording blood pressure.

To measure cardiac output (ventral aortic flow) an incision was made at the base of the pectoral fins and the ventral aorta was dissected free from the surrounding tissues without disrupting the pericardium. A cuff-type Doppler flow probe (2.5–3.0 mm i.d., single crystal, P. Pohl International Inc.) was placed around the vessel.

The fish was then placed on its left side and an incision approximately 4 cm long was made between the pectoral and pelvic fins starting 1–1.5 cm posterior to the edge of the operculum. A cannula was inserted into the gonadal vein for the injection of drugs. The cannula was passed through the body wall and sutured to the skin.

To measure blood flow in the coeliac and mesenteric arteries, the vessels were exposed and freed from connective tissue. During this procedure, care was taken not to damage the nerve running along the mesenteric artery. Doppler flow probes (1.3–1.6mm i.d., single crystal, P. Pohl International Inc.) were placed around the vessels. The leads from the flow probes were taken out via a small incision posterior to the pelvic fin. The incision was closed and the leads were sutured to the skin.

For recordings of gastric motility, a balloon attached to a cannula was inserted into the stomach through an incision in the posterior part of the stomach. The incision was closed and the cannula tunnelled through the body wall. The balloon was filled with water and connected to a Statham P23 pressure transducer for recording changes in intragastric pressure.

In one group of fish, small incisions were made dorsal to the opercula, and the vagi were cut bilaterally immediately outside the skull.

After surgery, the animals were transferred to a tank with circulating sea water, where they recovered rapidly from anaesthesia. They were allowed to recover in this tank for 24-48 h before any experiments were started. During this time the effects of MS222 and handling wore off and the flow and pressure recordings stabilized. The flow signal was electrically dampened to obtain a mean blood flow instead of a beat-to-beat flow.

The peptides were injected as bolus doses (0.2 ml) and in the case of SP also infused during a longer period (8–18 min). The doses were calculated per kilogram body mass of the fish.

After the experiments were finished, some of the killed fish (N=6) were injected with yellow and red microfil (Canton Bio-Medical Products Inc.) to reveal the distribution of the vascular beds supplied by the coeliac and the mesenteric arteries, respectively. Cannulae were inserted into the proximal part of the coeliac and mesenteric arteries, and connected to syringes containing the microfil. The microfil was infused simultaneously into the two vessels, and the distribution of microfil was drawn (see Fig. 1).

The following drugs were used: atropine sulphate (Sigma), synthetic substance P (Sigma), natural porcine vasoactive intestinal polypeptide (VIP; a kind gift from Professor V. Mutt, Stockholm).

In addition to Grass polygraph recordings, data acquisition software and an IBM PPC computer were used for calculation of mean values during sampling periods. For the statistical tests the control period (2 min) was compared to a 1 min (VIP) or 20 s (SP) period at the maximal flow through the coeliac artery.

A Wilcoxon matched-pairs, signed ranks test was used to evaluate the statistical significance of the recorded effects of SP and VIP. Differences where P<0.05 were regarded as statistically significant. Values are presented as means±s.E.M.

The directional pulsed Doppler flowmeter used in the present work accurately measures blood flow velocity in the vessels studied and displays this velocity in kHz Doppler shift. However, there is a direct relationship between the blood velocity and instantaneous volume flow; previous studies on cod visceral arteries (M. Axelsson and R. Fritsche, unpublished results) and ventral aortic blood flow in the hagfish (Axelsson et al. 1990), in which the mean blood flow has been calibrated in absolute terms, have demonstrated a high degree of linear correlation between the Doppler signal and mean volume flow. We are therefore confident that the percentage changes in Doppler shift recorded in the present experiments are directly correlated to changes in blood flow in the arteries.

Heart rate was derived from the pulsatile blood pressure signal via a Grass 7P44 tachograph unit, and is expressed in beats per minute. Changes in the vascular resistance were calculated as the pressure drop across the vascular bed divided by the percentage change in blood flow through the same vascular circuit, assuming zero pressure in the central venous system.

The vascular beds supplied by the coeliac artery and the mesenteric artery, as indicated by microfil injections, are shown schematically in Fig. 1.

Resting values for heart rate (/H; 42.9±1.8 beatsmin⁻¹; N= 10), ventral aortic pressure (P_VA; 4.9±0.3kPa; N=10) and dorsal aortic pressure (P_DA; 3.5±0.2kPa; N= 10) were similar to the values previously reported in the cod (Axelsson and Nilsson, 1986; Fritsche and Nilsson, 1989).

Injection of VIP (100 pmol kg⁻¹; Fig. 2; Table 1) caused an increase in cardiac output (Q̇) and increased flows in the coeliac and mesenteric arteries (F_COA and F_MEA, respectively). Blood pressures increased in both the ventral and dorsal aorta. The increase in cardiac output was produced solely by an increase in stroke volume (Vs) since the heart rate (fH) was unaffected by VIP. The coeliac vascular resistance (VR_COA) was reduced by VIP, while no significant change could be found in the mesenteric vascular resistance (VR_MEA). The vascular effects of VIP were slow in onset (4–6min) and longlasting (15–25min).

Recordings of the intragastric pressure in most cases showed spontaneous contractions of the stomach wall. This activity was abolished when VIP was injected into the fish (Fig. 2; N=5).

10pmolkg⁻¹ VIP usually had no effect on the recorded variables (N=5).

SP (10 or 100pmolkg⁻¹; Fig. 3) produced a triphasic response in coeliac artery blood flow. An initial increase was followed by a rapid decrease, down to or below the initial level, which subsequently reversed to a secondary increase in flow. During these three changes in coeliac artery flow, corresponding changes in the coeliac vascular resistance occurred (Table 2). At the same time as the increase in coeliac artery flow, there was a smaller increase in mesenteric artery flow, an increase in cardiac output and a decrease in mesenteric and systemic vascular resistance. No significant change could, however, be found in P_VA, P_DA, FH or Vs during this period (Table 1). During the transient decrease in flow in the coeliac artery, there was often a corresponding small further increase in flow in the mesenteric artery.

Similar, but weaker effects were obtained with 1pmolkg⁻¹ SP in 50% of the animals tested, while the others appeared to be unaffected by this dose, indicating that this dose is within the range of the threshold value of the effects of SP.

Injections of SP had weak excitatory effects on the stomach motility in five out of seven fish. The effect occurred within the first minutes after injection and was recorded as a phasic contraction, if the stomach was previously inactive, or as an enhanced amplitude of spontaneously occurring contractions. Infusion of SP (65–100 pmol kg⁻¹; 8–18 min) slightly increased the frequency of the rhythmic contractions of the stomach in five out of six fish during the infusion period. The second change in F_COA in response to SP, the decrease in flow, could not be correlated to the contractions of the stomach, whether induced by SP or spontaneous. The response in F_COA was not due to recirculation of the bolus dose of SP in the cod, since the triphasic response also occurred when SP was infused.

One group of animals was subjected to vagotomy to elucidate whether a vagal reflex was involved in the triphasic response to SP. There was, however, no significant difference between vagotomized and non-vagotomized animals in the response of F_COA (Fig. 4; N = 5). However, atropine treatment (1.2mgkg⁻¹) blocked the second phase, the decrease, in F_COA in response to SP, while the initial increase in flow remained unchanged (Figs 3 and 5; N=6).

Several studies have been performed on fish to investigate the importance of adrenergic control of the cardiovascular system (Nilsson and Axelsson, 1987). The present study is, however, to our knowledge the first report on the influence of regulatory peptides on the cardiovascular system in fish in vivo.

The 28 amino acid peptide VIP, originally isolated from pig intestine, has been found to produce a decrease in systemic blood pressure, an increase in cardiac output (both inotropic and chronotropic effects) and to have a vasodilatory effect in gastrointestinal vascular beds in several mammalian species (Said and Mutt, 1970; Eklund et al. 1979; Blitz and Charbon, 1983; Unverferth et al. 1985). In vitro studies in fish have also indicated a vasodilatory role for VIP in different vascular beds (Holder et al. 1983; Bolis et al. 1984), including the mesenteric-artery-derived swimbladder vasculature of the cod (Lundin and Holmgren, 1984). In the present in vivo study of the cod, porcine VIP increased blood flow in the coeliac artery and the mesenteric artery by 48.6% and 16.7%, respectively. This was accomplished by the positive inotropic effect that increases the cardiac output by 16.2% and, in the case of the coeliac artery, partly by a decrease in vascular resistance. VIP had no significant effect on the mesenteric vascular resistance, which may seem to contrast with the results obtained by Lundin and Holmgren (1984). However, the vasodilatory effect in the isolated swimbladder preparation was only revealed after longlasting infusions of VIP (10⁻⁷moll⁻¹) into preparations where the tonus of the vessels was increased by pretreatment with adrenaline, while single injections of VIP (1ml at 10⁻⁷mol l⁻⁻¹ = 100pmol), equivalent to those administered to the whole fish in the present study, had no recordable effect. In any case, the vasodilatory effect of VIP in vascular beds of the gut observed in mammals also seems to be present in the cod, although different vascular beds may be affected by different concentrations and times of exposure, at least under experimental conditions.

The amino acid sequence of the VIP present in the cod differs in five positions from the porcine VIP used in the present study. Cod VIP and porcine VIP have, however, been found to be approximately equally potent in stimulating amylase secretion from pancreatic acini of the guinea pig, suggesting that the substitutions may be of minor importance for the biological activity (Thwaites et al. 1989). This is supported by the observation that the dose range of porcine VIP needed to produce an effect in the cod in the present study agrees well with those used in mammalian (canine) studies in vivo (Said and Mutt, 1970; Blitz and Charbon, 1983; Unverferth et al. 1985). Furthermore, the measured plasma concentrations of VIP-like immunoreactive material in the cod (125 pmol 1⁻¹) are similar to the levels found, for example, in the cat (50pmoll⁻¹) (Fahrenkrug et al. 1978; Holstein and Humphrey, 1980).

In mammals, VIP is suggested to be the neurotransmitter in the reflex mediating gastric receptive relaxation, which follows distension of the oesophagus (see Dockray, 1987). An inhibitory effect of VIP on the spontaneous motility of the gut was observed in the cod stomach in the present study. VIP is, however, probably not involved in a gastric receptive relaxation reflex in the cod (although this may be the case in other fish species), since the excitatory response to gastric distension is unaffected by VTP (D. J. Grove and S. Holmgren, unpublished observations).

SP has previously been shown to have a vasodilatory effect in the mesenteric vascular bed of mammals (Hallberg and Pernow, 1975; Burcher et al. 1977; Schrauwen and Houvenaghel, 1980; Rozsa and Varro, 1987), and in the present study SP produced an increase in flow through both the coeliac and mesenteric arteries. In the coeliac artery the increase in flow was followed by a rapid decrease and a subsequent increase. A similar triphasic response to SP has been described in the mesenteric vascular bed of the pig in vivo, when high doses of SP were infused (1–10μg animal⁻¹ min⁻¹; Schrauwen and Houvenaghel, 1980). The mechanism behind this response in the pig was, however, not further examined. In the cod, the decrease in flow could not be attributed to stomach contractions induced by SP. Nor was the response altered by vagotomy. Atropine, in contrast, blocked the rapid decrease in coeliac blood flow without affecting the initial increase, which suggests that a local cholinergic mechanism, rather than a vagal reflex, is responsible for the decrease in flow. The significance of this response needs further investigation.

Nerve fibres and endocrine cells containing SP-like immunoreactivity have been demonstrated in both the stomach and the intestine of the cod. Only very few vessels are, however, innervated by immunoreactive fibres in the stomach and no immunoreactive nerve fibres have been observed innervating the coeliac and mesenteric arteries (Jensen and Holmgren, 1985; Jensen et al. 1987; J. Jensen, A.-C. Jonsson and S. Holmgren, unpublished results). This may indicate that, if an endogenous SP-like peptide is involved in vascular control, it is not released from perivascular nerves but acts mainly as a hormone released from endocrine cells or non-vascular nerves in the gut wall.

Studies of the cod stomach, using a preparation vascularly perfused via the mesenteric artery, have shown that the threshold dose of SP needed to elicit a contraction is approximately 10pmol (J. Jensen, A.-C. Jonsson and S. Holmgren, unpublished results). In the present experiments, SP was injected into the gonadal vein in vivo in the dose range 1–100 pmol kg⁻¹, but the effects were weak even with the higher doses. However, considering that part of the injected SP is distributed to other vascular beds and some of it is likely to be degraded, the amount of SP actually reaching the gut smooth muscle is probably considerably reduced towards or below the threshold dose for an effect on stomach motility. The concentration reaching the vascular receptors may be higher, or the vascular receptors may be more sensitive to substance P than the gut wall smooth muscle receptors.

It has been estimated that as much as 40% of the cardiac output in the cod is directed to the coeliac and mesenteric arteries during resting conditions, and after feeding there is a marked increase in coeliac and mesenteric flow equivalent to the ‘increase in cardiac output (M. Axelsson and R. Fritsche, unpublished results). In the present study, both VIP and SP increased the cardiac output and the visceral blood flow. Assuming a Q+̇ value of 18mlmin⁻¹ kg⁻¹, an F_COA value of 3.5 ml min⁻¹ kg⁻¹ and an F_MCA value of 4.1 ml min⁻¹ kg⁻¹ (as calculated by M. Axelsson and R. Fritsche, unpublished results), VIP (100pmolkg⁻¹) increases cardiac output by 2.9 ml min⁻¹ kg⁻¹. The main part of this increase is distributed to the viscera (2.4 ml min⁻¹ kg⁻¹); this may be partly achieved by the decreased resistance in the coeliac vascular bed. Making the same assumptions for 10 and 100pmolkg⁻¹ SP gives increases in cardiac output of 1.4 and 2.2mlmin⁻¹kg⁻¹, respectively, while the corresponding increases in visceral blood flow are 2.3 and 3.4 ml min⁻¹ kg⁻¹. This indicates that the increase in visceral blood flow produced by SP is accomplished both by an increase in cardiac output and by a redistribution of blood from other vascular beds. The present results suggest that both VTP and SP may be mediators of postprandial hyperaemia in the gut of the cod.

This study was supported by the Swedish Natural Science Research Council, the A. F. Regnell Zoological Foundation and the Erna and Victor Hasselblad Foundation. We thank Professor Stefan Nilsson for critical reading and valuable comments on the manuscript.