A study of the unidirectional fluxes of Na and Cl across the gills of the dogfish Scyliorhinus canicula (Chondrichthyes)

The marine chondrichthyean (cartilaginous) fishes, which include the sharks and rays, show some remarkable differences in the process of their osmoregulation as compared with most osteichthyean (bony) fishes, to which they are distantly related. Most bony fishes are hypo-osmotic to sea water but the Chondrichthyes that live in this situation are nearly iso-osmotic (Smith, 1936). This latter condition reflects a retention of additional salts and urea in the body fluids and results in some profound differences in their physiology. The gills of chondrichthyeans are, for instance, nearly impermeable to urea and, compared with marine teleost fish, relatively impermeable to salts. Sodium and chloride move across the gills in chondrichthyean fish less rapidly than in teleosts, chloride moving 20 times slower and sodium 100 times slower. Chondrichthyean gills are 10 times more permeable to Cl than to Na (Maetz & Lahlou, 1966; Burger & Tosteson, 1966; Horowicz & Burger, 1968; Payan & Maetz, 1970, 1973; Carrier & Evans, 1972; Haywood, 1975).

In the present paper we have described measurements of the unidirectional fluxes of Na and Cl across the gills of the European dogfish, Scyliorhinus canicula, and have attempted to relate these ion movements to the electrical potential across the integument, as indicated by the potential difference (PD) between the inside and the outside of the fish. We have also studied the effects of changes in the pH and NaCl concentration of the external bathing solution on these ion fluxes.

Dogfish, Scyliorhinus canicula, were a gift from the aquarium at Monaco. They had been caught by trawling off the coast at Monaco and were kept in aquaria with circulating sea water. The fish were fasted for about 2 weeks before the experiments, to give more uniform results and minimize vomiting. The fish weighed 100−175 g and were paired by weight for the measurements of influx and efflux.

Polyethylene catheters, through which isotopes were injected, and a 3 M-KCl-agar bridge, for measuring electrical potential across the integument, were inserted into the peritoneal cavity through the ventral body wall. These tubes, as well as a plastic cloacal cannula in which the urine and rectal salt gland secretion were collected, were fixed into place with a purse-string suture. These procedures were carried out after the fish had been rapidly anaesthetized by packing them in ice. The fish recovered almost immediately after they were placed in sea water in the tanks in which the experiments were to be performed.

The experiments were carried out at 15−17 °C. Although the periods of handling and surgical procedures were kept as short as possible, the fish were in what is often referred to as a ‘handled’ or mildly ‘shocked’ state. This is known to influence ions movements in fishes (see later).

Influxes of Na and Cl were measured simultaneously, using the isotopes ²⁴Na and ³⁶Cl. The method has been described in detail by Payan & Maetz (1973). Each fish was rapidly weighed and placed in a tank containing 1500 ml of sea water, or physiological solution, which was mixed and oxygenated by bubbling with air. The external solution contained about 50μCi/l of ³⁶Cl and 100 μCi/1 of ²⁴Na. After 30 min exposure the fish was removed, briefly rinsed and a 1 ml blood sample was collected from the caudal artery. The ²⁴Na in the fish was then counted while it was in a chamber of circulating sea water which was placed in a whole-body counter (Packard, model 446). Standardization was made by collecting 0·5 ml samples from the external medium and measuring activity with the same counter immediately before and after a count was made upon the fish. The influx was then calculated by the usual methods (Payan & Maetz, 1973).

The ³⁶Cl influx was calculated after 7 days, allowing for the ²⁴Na in the external standard solution and the plasma to decay. The ³⁶C1 was measured in a scintillation counter (Intertechnique SL 40). The Cl space in these dogfish is known to be identical to the Na space (Payan & Maetz, 1970). This was calculated in each fish from the ratio of the total ^MNa in the fish and the concentration in the sample of plasma, the activity of which was counted with the external standard in a gamma-well counter (Mecaserto model MO 13).

The effluxes of ³⁶Cl and ²⁴Na were measured in dogfish in which a peritoneal catheter and a cloacal cannula had been inserted on the previous day (see Surgical Procedures). The efflux thus represents that from the integument, principally the gills according to Payan & Maetz, 1970. The isotope, 50 μCi ³⁶Cl or 100 μCi ²⁴Na, was injected through the peritoneal catheter, and after allowing 3 h for equilibration with the body fluids, the rate of loss of the isotopes into the bathing media was determined. This was measured with a gamma-detector, for ²⁴Na, or a beta-detector, for ³⁶Cl, through which the external fluid was circulated by a pump. For comparing the ion effluxes in media with different NaCl concentrations, the bath was flushed out with 51 of the solution. The experiment on each fish was commenced in sea water, which was replaced with solution of successively lower NaCl concentration. The fish was finally returned to sea water. The period of exposure to each solution was about 40 min. The influx of ions into these fish is slow, so that a terminal blood sample from the caudal artery was used to calculate the isotopes’ specific activity in the fish. The efflux was then calculated using this value and the slope of the appearance of the radioactivity in the external solution.

The electrical PD between the fish and the bathing solution was measured using standard procedures (see Kirschner, 1970). On the day preceding the measurements, the KCl-agar bridge was inserted through the ventral body wall into the peritoneal cavity, into which it projected by about 4 cm. The PD was measured with the aid of this bridge and a similar one placed in contact with the external bathing solution, employing a voltmeter (Keithley model 610 C) and a pair of calomel cells. The PD measured in this way is thought principally to reflect the potential across the branchial surfaces. This is suggested by studies made in the flounder where the value of the potential measured in vivo by a technique similar to that used in the present study (Potts & Eddy, 1973) was found identical with that of the perfused in vitro gill (Shuttleworth, Potts & Harris, 1974).

The fish were kept in sea water (from the bay of Villefranche-sur-Mer) with a pH of 7·8 and an osmotic concentration of 1140 m-osmole/kg water. The average concentration of chloride was 630 mm while that of Na was 530 mm. An artificial seawater solution was prepared with the following composition (mm): Na, 500; Ca, 10; Mg, 50; Cl, 500; SO₄, 30; NO₃, 60; urea, 80. The pH of this solution when aerated was 6·9 and in order to ensure that precipitation of the divalent cations did not occur it was not elevated to seawater level. The divalent cations are known to play an important role in maintaining normal branchial permeability in teleost fishes (Kerstetter, Kirschner & Rafuse, 1970; Maetz, 1974; Pic & Maetz, 1975). The urea was added to keep the osmotic concentration at that of sea water. Alexander et al. (1968) observed that Scyliorhinus canicula gained weight and soon died when kept in a mixture of sea water and tap water (60 % v/v). but not if the osmotic pressure of this mixture was restored to that of sea water by adding urea. The fish then survived and maintained weight. To determine the effects of Na and Cl in the external medium upon ion exchanges, the concentrations of these ions were lowered in the artificial sea water by replacing them with urea. This artificial solution had a pH nearly one unit lower than that of sea water, and hence, to determine whether the exchange fluxes were influenced by such a pH change, they were measured in sea water of lowered pH, the pH being decreased to 6·9 by the addition of 0·2 N nitric acid.

The concentration of sodium was determined with a flame photometer (Eppendorf) and that of Cl with a chloridometer (Buchler). The osmotic concentrations of the external solutions were measured with an osmometer (Knauer).

In sea water, Na and Cl were taken up across the integument (Tables 1, 2). The skin has been shown to contribute less than 10 % of such ion exchanges (Horowicz & Burger, 1968; Payan & Maetz, 1970) so they principally reflect movements across the gills. The unidirectional fluxes of Cl were several times greater than those of Na confirming earlier observations (Maetz & Lahlou, 1966). The flux ratio Na_in/Na_out, determined in separate experiments, was 4·6, which is more than double the value of the concentration gradient of Na between the inside and outside of the fish. The potential across the integument was 2 mV, the inside being negative. The Ussing passive flux ratio equation criterion (Ussing, 1960) indicates that the observed fluxes were not occurring passively, since this would require a PD of 20·3 mV, inside negative. The discrepancy suggests that an electrically neutral active transport of Na is occurring into the fish.

The value of the flux ratio for Cl was similar to that of the concentration gradient across the gills, and the observed PD was close to the predicted one, indicating passive movements of Cl.

The influxes of both Na and Cl were significantly (P < 0·05 and P < 0·01, respectively) decreased when the pH of the sea water was lowered from 7·8 to 6·9 (Tables 1, 2). The efflux of both these ions was unaffected by this change in pH. In five fish this change in the pH of the sea water resulted in a slight change in the PD, being a mean change from — 2·1 to —1·2 mV; a mean difference of 0·9 mV; S.E., 0·18. This contrasts with the small increase in electronegativity seen in the artificial sea water solution. It is interesting that although the change in pH resulted in a substantial decline in the influx of both Na and Cl, no significant change in the PD, other than could be expected from the increased H⁺ concentration, could be detected. The permeation process appears to be electrically neutral. The ratio of the influxes of Na and Cl were also similar, 3·18 and 3·14 respectively, at each pH value.

The artificial sea water solution did not contain K and it seemed possible that this might be the reason that the PD observed when the fish were bathed in this solution was different from that observed in sea water. When 12 mm-KCl was added to the artificial sea water, no significant change in the PD was observed. The initial PD averaged —6·7 mV in 4 fish, and 10 min after adding the KC1 it was —6·9 mV; a mean difference of 0·2 mV; S.E. 0·12.

The influx and efflux of Na and Cl were compared over a 50-fold range in NaCl concentration (see Methods). The basal influx of Na and Cl in the 500 mm-Na solution was lower than in sea water, a difference which can be accounted for by the difference in pH rather than the difference in sodium concentration. As the external NaCl concentration was decreased, the influxes of Na and Cl declined, but that of Na, in contrast to Cl, reached a saturation level at about 200 mm, This is consistent with there being an active transport process for Na as proposed above. Chloride influx appears to be passive.

The efflux of Na increased 3-4-fold at the lower NaCl concentrations but that of Cl remained relatively constant.

The potential acrosss the integument became more negative as the external NaCl concentration was decreased. The observed PD’s rarely corresponded with the PD’s predicted by the Ussing flux ratio equation. The results were variable and no firm conclusions can be drawn from them, but it seems likely that at low external Na concentrations there is an increased passive leak of Na to the exterior so that the fish experience a negative Na balance. The efflux of Cl always exceeded its influx in the physiological solutions and the observed PD was always less than the theoretical one, suggesting that an active Cl extrusion may have been occurring. The net Cl loss that becomes apparent in artificial sea water is due, not to a change in Cl efflux, but to a decline in the influx due to the lower pH of the artificial solutions and the reduced external Cl levels.

The present results suggest that there is an electrically neutral active transport of Na into the body fluids of the dogfish Scyliorhinus canicula, across the gills, from bathing sea water. It is also possible that a net active extrusion of Cl is occurring, under our experimental conditions, which in sea water is masked by a high passive Cl influx. When considering these results it should be recalled that the fish were in a ‘handled’ or even ‘shocked’ condition, which is known to alter the fluxes of ions across the gills of such fish. In Scyliorhinus canicula, handling has been shown to increase the Na influx and probably decreases the efflux (Payan & Maetz, 1973). It is also unlikely that the fish are in strict ionic balance under such conditions.

Previous studies on chondrichthyean fishes have suggested that the branchial influx of Na exceeds the efflux (Maetz & Lahlou, 1966; Burger & Tosteson, 1966; Horowicz & Burger, 1968; Payan & Maetz, 1973), but since the accompanying potential across the integument was unknown it remained uncertain whether or not this reflected an active inward transport of Na. It now seems likely that such a process does occur, since, apart from a deviation from the PD predicted if diffusion alone was involved, the influx of Na has been shown to involve saturation kinetics. In addition, the gills of chondrichthyeans have been shown to possess enzymic processes that could be involved in Na transport; a Na-K activated ATPase (Jampol & Epstein, 1970) and carbonic anhydrase (Hodler et al. 1955). The latter enzyme appears to mediate an internal H⁺/external Na⁺ exchange across the gills (Payan & Maetz, 1973). The present results show that an increase in the external H⁺ concentration depresses the Na influx which can be explained by this process. Such an effect of pH has also been shown to occur in the goldfish Carassius auratus where such a H⁺/Na⁺ exchange occurs (Maetz, 1973).

The evidence that there may also be an active Cl extrusion mechanism across the gills of the dogfish is more tenuous. Influx appears to be passive because it does not show any saturation at the concentrations tested. The reduction of the Cl influx as pH was lowered is reminiscent of the decline in Cl permeability which has been observed when the external H⁺ concentration is increased in solutions bathing frog skeletal muscle (Hutter & Warner, 1967) and cornea (Candia, 1973). The reason for such changes is uncertain but they do not appear to involve fixed charges on the membrane. The latter are expected to become more positive under such conditions and thus would facilitate anion permeability. It is possible that the change in Cl permeability is linked to the decline in the Na influx which is also observed under these conditions. The efflux of Cl was maintained at all of the external Cl concentrations, an observation which is not inconsistent with either a passive permeation or an active transport. The observed potential across the integument could, however, not be reconciled, under any set of experimental conditions, with a passive process. Balance studies on the influx of Cl in dogfish and its rate of extrusion from the kidneys and rectal salt gland have suggested that an active Cl extrusion process is likely (Maetz & Lahlou, 1966; Payan & Maetz, 1970). Chloride-secreting cells which could perform this function have been identified in the gills of Scyliorhinus canicula (Garcia Romeu & Masoni, 1970).

The ionic permeability of the gills of chondrichthyean fishes has been compared with that of freshwater teleosts, the gills of both types of fishes being less permeable than those of marine teleosts (Payan & Maetz, 1970). To emphasize further such a similarity, it now appears that chondrichthyeans, like freshwater teleosts, possess an active Na uptake mechanism that also involves a Na⁺/H⁺ exchange (Kerstetter et al. 1970). The physiological use of such a process in chondrichthyeans is uncertain. It could represent a phylogenetic remnant from fresh water ancestors (Payan & Maetz, 1973). Such a process may be necessary to maintain the general metabolic integrity of the gills in sea water or, as the gills are important sites for H⁺ excretion in such fish (Piiper, Meyer & Drees, 1972), it may be important for acid-base balance (Payan & Maetz, 1973). Some uptake of Na across the gills may also be desirable in order to maintain the nearly electrically neutral conditions observed, since this could influence the fluxes of other ions. As indicated by the efflux of Na, the gills of the dogfish have a much lower permeability to this ion than have the gills of marine teleosts. An active type of Na-channel may provide the best way of keeping the passive permeability low while retaining a limited access of Na. It is interesting that the gills become more leaky when the external Na concentrations are low and this could be due to a failure of the Na pump allowing a backleak through the pump sites. Ouabain is known not only to block influx of Na across various epithelia but also to facilitate the efflux in the cornea (Candia, Bentley & Cook, 1974) and in gills of the perfused head of trout in fresh water (Payan, Matty & Maetz, 1975). A number of chondrichthyeans live in fresh water and it has often been suggested that such species may, like freshwater teleosts, be able to accumulate salts actively across their gills from dilute external solutions. This possibility now seems quite likely. Such species would, however, need to attain saturation levels of Na transport at lower external Na levels and be able to control Na efflux more effectively than can the dogfish when placed in such solutions.

If an active Cl extrusion mechanism exists in the dogfish this occurs in a direction which is opposite to that of the active Na transport. This is an unusual, but not unique, situation as it has also been shown to occur in the amphibian cornea (Candia & Askew, 1968; Candia et al. 1974). A physiological role for such a Cl extrusion is easy to envisage in marine chondrichthyeans which can accumulate this ion in large amounts from sea water. In this respect dogfish may be similar to marine teleosts which extrude Cl actively across their gills (Maetz, 1971). Such a process could vary in its quantitative importance depending on the physiological condition of the fish. If it became salt-loaded, due for instance to damage to the integument or drinking, Cl extrusion could be stimulated. Drinking is rare in chondrichthyeans but seems to occur in starved pyjama sharks (Poroderma africanum) which adopt a hypo-osmotic form of osmoregulation due to inadequate formation of urea (Haywood, 1973).

The potential across the integument of Scyliorhinus canicula in sea water is about 2 mV, the interior being electronegative. This compares with about 10-30 mV, inside positive, in marine teleosts (Maetz, 1971; Potts & Eddy, 1973; House & Maetz, 1974; Greenwald, Kirschner & Sanders, 1974; Pic & Maetz, 1975). In these teleosts, the potential results from the diffusion of sodium and potassium ions along their chemical gradients, the gill being far more permeable to monovalent cations than to anions. A small contribution of an electrogenic Cl pump has also been demonstrated (Shuttleworth et al. 1974; Pic & Maetz, 1975). The low potential in the dogfish suggests, as already stated, that Na⁺ transport is an electrically neutral process. Since the net influx of Cl⁻ exceeds that of Na⁺, the slight internal negativity could indicate a small diffusion potential. In any case this low potential appears to offer a minimal electrical barrier to active ion movements.

The data obtained with low external Na and Cl concentrations do not provide an explanation for the observed direction and magnitude of the potential observed across the fish integument. If the potential was the result of diffusion, the gill potential should be positive inside because the chloride permeability appears greater than the Na permeability. As described earlier, however, the gills are important sites for acidbase balance in these fish; H⁺ and NH⁺₄ are known to be extruded and it is also considered likely that may play a role (Payan & Maetz, 1973). The latter ions could presumably also contribute to the observed PD.

The present results may also have some practical implications with respect to the life of chondrichthyeans in polluted environments and the ability of these fish to move into estuaries and rivers where the sea water becomes diluted. Urea and water permeability are maintained in 50% sea water (Payan, Goldstein & Forster, 1973) and our results indicate that this is also the case for Na⁺ and Cl⁻ permeabilities at such concentration. However, very low external NaCl concentrations (< 100 mm) can be seen to result in a substantial loss of body salts from the gills of such marine fish.

Sea water is a poorly buffered alkaline solution; thus the pH can be readily altered by small amounts of acidic effluents. Such changes would also be expected to have a considerable influence on the salt balance of chondrichthyean fishes living in such environments.