DIVERGENCE IN VASCULAR ACTIONS OF PROSTACYCLIN DURING VERTEBRATE EVOLUTION

Prostacyclin (PGIz) generation by teleost and elasmobranch fish vascular wall was detected, as 6-keto-PGFlm-like immunoreactivity, in the incubation medium of ventral aorta rings from Anguilla anguilla, Conger conger and Scyliorhinus stellaris. The average yields after 15 min incubation were the following (ng of 6-keto-PGFle-1ike/g of wet tissue): A.anguilla = 116.8 k30; C. conger = 7.50; S.stellaris = 58.0+ 10. Accordingly, the vascular responses to synthetic PGIz were examined on the isolated and saline-perfused head of four teleost (A. anguilla, C. conger, Scor-paena porcus and Solea solea) and two elasmobranch (S.stellaris and Torpedo marmorata) species. In teleosts, PGIz gave a dose-related increase in overall gill vascular resistance, unaffected by indomethacin, phenoxybenzamine and phentolamine pretreatment or by previous decerebration of the animal. Con-versely, in elasmobranchs PGIz elicited a diphasic response, characterized by a transient, not dose-related constriction, followed by a longer lasting and dose-dependent dilation. The effects of PGIz on isolated branchial arch (in A.anguilla, C.conger and S. stellaris) and ventral aorta strips (in C. conger and S. stellaris) were also examined. Both preparations shared the same responsivity of the head in toto, confirming the divergence observed in this respect between teleosts and elasmobranchs.

In mammals, PGIZ, the major metabolite of arachidonate (AA) in large vessel wall, is a potent vasodilator and the strongest endogenous inhibitor of platelet aggregation known. A physiological antagonism to thromboxane A2 (another component of the prostanoid system, mainly generated by platelets) with respect to effects on vascular tone and platelet aggregability has been proposed (Moncada and Vane,'79). Similar vasoactive properties have been described in the amphibian Rana catesbeiana (Leffler et al. '80;Herman et al. '82). The aim of our experiments was to investigate both the occurrence of PGIz biosynthesis from endogenous AA stores and the vascular actions of this autacoid on different segments of teleost and elasmobranch fish branchial circulation.

Fish.
S. stellaris (1,000-3,000 g), T marmorata (200-400 g) and C. conger (500-1,500 g) were obtained during winter and spring from the Aquarium of the Zoological Station of Naples where they were kept in holding tanks with circulating seawater for at least ten days at 16-18'. A. anguilla (90-180 g), S.porcus (400-500 g) and S. solea (200-250 g) were obtained from local dealers. The fish were kept at 16-18" either in circulating tap water or seawater until used.

Incubation and prostanoid extraction
The animals were sacrificed by decapitation and the ventral aorta was excised. The artery was carefully freed from surrounding tissue and cut into rings. The tissues were incubated for 15 min in a metabolic shaker at 18", 0.5 ml of saline were added to ventral aorta rings. After incubation, the medium was acidified to pH 3.0 with 0.5 M citric acid and 0.25 ml of saturated NaCl solution was added, twice extracted with an equal volume of ethylacetate and evaporated. The residue was dissolved in 1 ml of 50 mM buffer (pH 7.4) and used for radioimmunoassay determination of 6-ket0-PGF1,~.

Isolated and perfused head
Perfused head was prepared according to Payan and Matty ('75) with minor modifications. The heart was exposed, heparin (500 I.U./lOO g) injected into the ventricle and allowed to flow through the gills. After having tied the sinus venosus, the fish was decapitated just behind the pectoral fins and a polyethylene catheter inserted into the ventral aorta. The head was then put in a plastic box filled with circulating water at constant temperature. During preparation and experiment gills were ventilated either with filtered tap water or seawater (T: 18"; flow: 600 mumin) continously bubbled with a stream of air. Perfusion was carried out by means of a peristaltic pump with a steady flow of 2.5 mYlOO g (Satchel1 '71). After this procedure, 10-15 min were allowed to reach a steady pressure before beginning the experiment. Cannulation of dorsal aorta was also performed in some cases, but, apart from the expected rise in inflow pressure, no variation in the response was observed.

Isolated branchial arch
Branchial arch was prepared as follows (Bergman et al. '74). The gills were cleared of the blood by passing heparinized saline through the isolated head for 15 min; the gill arch was dissected out of the branchial basket and a cannula inserted into the afferent branchial vessel. Position of ligature was such that approximately 80% of the filaments were perfused. Each gill arch was suspended in a bath with conditions similar to those described for the isolated head. The pump output to each arch was defined assuming an equally divided flow to all gill arches and considering that only 80% of the arch was perfused.

Pressure measurements and calculations
The inflow pressure both in isolated head and branchial arch was measured with a Statham P23 pressure transducer connected to a Beckmann polygraph. The overall gill vascular resistance (Nilsson '84) was calculated (Burton '72) from Rg = 6P/Q. Where Rg = gill vascular resistance; 6P = drop in pressure through the gills; Q = perfusion flow (ml/min/100 g).

Ventral aorta strip
The ventral aorta was dissected out and freed of surrounding tissue. The artery was then cut spirally and suspended in an organ bath containing an air-bubbled saline at room temperature (18"). All preparations were left suspended in saline for at least 1 hr before experiments were begun. Preparations were exposed to substances for three min. Responses were measured by means of an isotonic transducer connected to a recorder (Basile).
PGI2 administered in bolus injections ranging from 10 ng to lo4 ngI100 g produced a dose-related vasoconstriction in the isolated and perfused head of all the teleost species examined (Fig. 1). This effect was not inhibited by indomethacin (1.0" 10-5M), phenoxybenzamine (1.0*10-6M) or phentolamine (l.O* 10-5M) pretreatment or by previous meng*lOOg-'b.Y 1 chanical decerebration of the animal (data not shown). Vehicle alone (phosphate buffer, pH 8.4) did not affect inflow pressure. Conversely, in the elasmobranch, s. stellaris and Tmarmorata, a dilation preceded by a transient, not dose-related constriction was observed ( Fig. 1). As shown in Figure 2, the action of PGI2 on gill arch preparation of A. anguilla and C. conger was consistent with the one observed in the head in toto. A dosedependent increase in vascular resistance followed PGIZ administration at doses as low as 10 ng/100 g. Similarly, in isolated arches from S. stellaris the typical pattern constriction-long-lasting dilation was reproduced (Fig. 2). Changes in tension of spiral ventral aorta strips from C. conger and S. stellaris induced by PG12 were also examined. The effects of PGIZ on arterial smooth muscle preparations of each species paralleled those observed in branchial vasculature (Fig. 3). A concentration-related contraction was elicited in C. conger (threshold = 100 ng/ml of bath), whereas a relaxation (preceded by a brief contraction) occurred in S. stellaris (thresh-old= 10 ng/ml of bath). Furthermore, neither effect was inhibited by a mixture of antagonists (Gilmore et a1 '68) of the following composition: methysergide maleate (2.7" gl ml), propopranolol HC1 (2.6*10-6 g/ml), atropine sulfate g/ml) and indomethacin (2*10-~ g/ml).

DISCUSSION
From the findings here reported it appears that teleost and elasmobranch vessel walls are both able of producing PGIz using endcgenous AA as a substrate and of responding in a dose-related fashion to the administration of this autacoid. In our laboratory, generation and vascular activities of other prostanoids and prostanoid precursors in fish circulation are in study (Piomelli et al., in preparation). We were able to show that teleost and elasmobranch branchial vasculature can convert exogenous AA to vasoactive products of cyclo-oxygenation, and that synthetic prostaglandins (PGFz, and PGE2) can elicit an array of responses in the perfused head, branchial arch and spiral aortic strip of several fish species. Furthermore, we observed that the vasopressor response that follows the administration of angiotensin I1 to the isolated head of the freshwater adapted eel, A.anguilla, can be blocked by the cyclooxygenase inhibitor indomethacin (Piomelli and Pinto, submitted). These results, taken together, may lend support to the hypothesis of a role of eicosanoids in teleost and elasmobranch gill circulation.
In this frame, the observation of a divergence in the vascular activity of PGIZ between teleost fish, on one hand, and elasmobranchs, amphibians and mammals, on the other, may be of interest. We have attempted to rule out reflex responses and mediation via some other transmitters by the use of decerebrated animals, isolated tissues (as opposed to in toto perfused head) and various pharmacological antagonists.
Even though these experiments do not clarify the mechanism of PGI2-induced vasoconstriction in teleost fish, they clearly suggest a local site of action, which is consistent with the finding of a local production of this humor by the vessel. Two characteristics of teleost responsivity to PGIZ, as they appear from our results, are worthy of being mentioned. First, its apparent group-specificity: within teleosts, species phylogenetically distant showed a qualitatively similar response pattern. Second, the lack of a direct relationship with any particular environmental conditions. In fact, in spite of the diversity of habitat and behaviour, the four teleost species studied here shared a common responsivity to PGIZ. Consistently, teleost and elasmobranch species dwelling in the same areas and showing related behavioural patterns (such as T marmorata and S. oorcus) were dif- ferent in this respect. It seems reasonable therefore to suggest that the divergence Observed here is not immediately explainable in terms of a specific environmental need of homeostatic control of the milieu interieur modern teleosts. More subtle changes in the