Adrenaline and Branchial Nerve Stimulation Inhibit ⁴⁵Ca Influx into the Gills of Rainbow Trout, Salmo Gairdneri

Adrenaline and acetylcholine affect the movement of water and electrolytes across the branchial epithelium in teleosts (see Isaia, 1984; Rankin & Bolis, 1984) and chloride cell function in opercular membranes (see Zadunaisky, 1984). Autonomic nerves have been shown to be involved in the regulation of water and electrolyte transport in tissues of other vertebrates, e.g. mammalian intestine (Sjövall, 1984; Tapper, 1983). The possibility of neural involvement in epithelial function in fish is suggested by the finding (Mayer-Gostan & Hirano, 1976) that transection of the glossopharyngeal and vagus nerves disturbed water and electrolyte balance in the eel, Anguilla anguilla. Although these authors attributed the effect to changes in ion and water fluxes in the stomach, it is possible that concomitant changes in fluxes through the gills may have contributed to the overall result.

In this paper the possibility of autonomic mechanisms affecting an aspect of epithelial function has been examined using the uptake of ⁴⁵Ca into the gills as a marker of epithelial activity. The inhibitory effects of adrenaline and stimulation of the branchial nerves on the influx of ⁴⁵Ca into the gills of S. gairdneri are reported.

Rainbow trout, Salmo gairdneri Richardson (100–300 g), were obtained from the Silverstream Trout Farm, Buxton, Victoria. The fish were kept in a recirculating freshwater tank at 17–18 °C with a sand/charcoal filter, and a Ca²⁺ concentration of between 0·9 and 1·1 mmol l⁻¹. The fish were equilibrated for 1 week before experimentation, and were held for no longer than 4 weeks.

The experiments were performed using ⁴⁵Ca, as CaCl₂ (0·17 μCi μmol⁻¹, Amersham) and [³H]mannitol (New England Nuclear). Mannitol was used as a marker of external space to enable the calculation of the amount of calcium taken up by the gills. Preliminary experiments showed that sufficient ⁴⁵Ca influx occurred in 5 min and that the influx of mannitol was negligible over this time.

Trout were anaesthetized with MS222 (tricaine methane sulphonate, Rural Chemicals) dissolved in the freshwater medium, and the heart and bulbus arteriosus were exposed by a ventral midline incision. A loose cotton ligature was placed around the bulboventricular junction. The bulbus arteriosus was cannulated with polyethylene tubing (i.d. 0·86 mm, o.d. 1·27 mm, Dural Plastics) via an incision in the ventricle, and was perfused at constant pressure with heparinized (5 i.u. ml⁻¹) Hepes-buffered physiological saline [composition in mmol l⁻¹: NaCl, 138·0; KC1,5·0; MgCl₂.6H₂O, 0·5; CaCl₂, 2·5; glucose, 5·6; Hepes (Sigma), 10·0; NaOH, 7·0; bubbled with O₂] that had been passed through a 0·2 μm filter (Sartorius). Immediately after commencement of perfusion, irrigation of the gills with recycled fresh water was begun. The perfusion rate was measured using a photoelectric drop-counting apparatus and was recorded on a Grass 79D polygraph. The perfusion rate was initially adjusted by altering the height of the reservoir, until it approximated the in vivo cardiac output of rainbow trout, approximately 37 ml kg⁻¹ min⁻¹ (Wood & Shelton, 1980).

After 10 min of perfusion, the first and second gill arches from both sides were removed and placed in a dish of physiological saline. The branchial nerve supply of the first and second gill arches from both sides was exposed and harnessed with a fine cotton ligature. Each arch was then attached to a jig and the nerve was passed through platinum ring electrodes in a small reservoir of physiological saline attached to the jig. The gill on its jig was immersed for 5 min in 100 ml of magnetically stirred distilled water containing 20 μCi of ⁴⁵Ca and 5 μCi of [³H]mannitol. The nerves to the gills from one side only were stimulated with 1 ms, 20 V pulses at 10 Hz delivered from a Grass S9 stimulator throughout the period of immersion. The contralateral gills acted as an unstimulated control. Prior to addition of the gills, isotope activities in the bath were measured. The gill arches were placed in 4 ml of 0·1 mol l⁻¹ nitric acid for 24 h to elute the radioactivity. A 0·5 ml sample was added to 10 ml of scintillation cocktail (Insta-gel, Packard) for determination of ⁴⁵Ca and ³H activities (L’Annunziata, 1979). The effect of nerve stimulation was also tested in trout which had received an intraperitoneal injection of the adrenergic neurone-blocking drug, bretylium (1 mgkg⁻¹), 2h before the experiment.

The influx of ⁴⁵Ca was expressed as μmolCa²⁺ h⁻¹ g gill wet mass⁻¹.

The ⁴⁵Ca influxes are shown in Table 1. Stimulation of the branchial nerves throughout the 5 min exposure to ⁴⁵Ca significantly reduced the influx of ⁴⁵Ca into the gills, compared with the unstimulated contralateral gills. In preparations taken from fish injected 2 h previously with bretylium, there was no significant difference in ⁴⁵Ca influx between stimulated and unstimulated gills (Table 1).

Addition of adrenaline (10⁻⁶ mol l⁻¹) to the perfusion fluid of the test gill arches significantly reduced the influx of ⁴⁵Ca compared with the influx observed in the contralateral control gills perfused with normal physiological saline (Table 1).

In the present experiments the inhibition of calcium influx by adrenaline was in contrast to the results of Payan et al. (1981), who found that adrenaline increased the uptake of calcium into the perfused head of rainbow trout. The conflicting results may be due to differences in experimental design. In the perfusion experiments of Payan et al. (1981), adrenaline infusion would have altered the haemodynamic state of the gills, which in turn could have altered the ‘functional surface area’ of the gills available for molecular exchanges. In the nonperfused gill preparations used in this study, the inhibition of calcium influx by adrenaline and nerve stimulation was considered to be independent of their effect on gill haemodynamics.

Perry & Wood (1985) provided evidence that Ca²⁺ uptake into the gills of freshwater fish is mediated by active transport – it showed saturable Michaelis-Menten kinetics – and they concluded that either a Ca²⁺ carrier and/or a selective ion channel was involved. However, there is no clear evidence to show which epithelial cells are involved in the Ca²⁺ movement. Ultrastructural studies have shown that the chloride cells are the only cells in the gills that appear to be morphologically specialized for active transport, since they are the only epithelial cells to have the numerous mitochondria, surface microvilli and cytoplasmic tubulovesicular extensions that characterize ion-transporting cells (see Laurent et al. 1985).

Substantive evidence that neurotransmitters can affect chloride cell function comes from the studies on isolated opercular membranes of teleost fish, in which adrenaline has been shown to inhibit the transport of chloride by the chloride cells (see Zadunaisky, 1984). It is possible that in this study nerve stimulation and adrenaline were affecting the ability of chloride cells in the filament epithelium to inhibit the uptake of ⁴⁵Ca into the gills. Thus, in addition to vasomotor regulation, autonomic mechanisms could be involved in the regulation of molecular exchanges across the filament epithelium of the gills.

I would like to thank Professor Graeme Campbell for many useful discussions during the course of this work. For part of this work the author was supported by a Commonwealth Postgraduate Research Award. Steve Petrou is thanked for writing a computer program for analysis of the data. Pat Williams at the Silverstream Trout Farm is thanked for providing the trout.