Evidence That K-Dependent Transport Components in Addition to Na-K ATPase Are Involved in Na and Cl Excretion in Marine Teleost Gills

In marine teleosts, the branchial ionocytes are probably responsible for Na+ and Cl⁻ exchange fluxes. These cells are very rich in Na-K ATPase activity suggesting that this enzyme may be involved in Na+ excretion (e.g. Maetz & Bornancin, 1975). Using autoradiography with [₃H]ouabain, this enzyme has been localized in the basolateral membrane of the cell and its numerous invaginations, the tubular reticulum (Karnaky et al. 1976). Ouabain inhibits the Na-K ATPase more effectively from the serosal side, supporting the idea that the enzyme is located on the serosal face of the cell (Karnaky et al. 1976; Silva et al. 1977). This situation, similar to that in fresh water fish, favours the transport of Na⁺ into the blood instead of excretion to the external medium. It therefore raises the problem of the exact role of the Na-K ATPase in marine teleosts, particularly regarding the extrusion of NaCl into the external medium.

Since the inhibition of Na-K ATPase by ouabain parallels the inhibition of Na+ and Cl⁻ efflux, the enzyme seems to be involved in ionic excretion (Silva et al. 1977,). Evidence for coupling between Na⁺ and Cl⁻ fluxes had previously been provided by the fact that both fluxes are simultaneously activated by external K+ ions or inhibited by SCN injection, when a marine fish is rapidly transferred from salt to fresh water (Epstein, Maetz & de Renzis, 1973; Maetz & Pic, 1975). On this basis it was suggested that the Cl⁻ pump is associated with Na⁺/K+ exchange.

As the Na-K ATPase of the basolateral membrane faces the internal medium, plasma K+ would be high enough to activate the enzyme irrespective of the K+ level in the external medium. Therefore, the activation of ionic excretion by external K suggests the involvement of other mechanisms, complementary to the sodium pump in ionic extrusion across the gill epithelium in sea water.

In the present work, we try to confirm the proposal that other systems are involved by comparing the effects of external T1+ with K+. Thallium (T1+) can substitute for K+ with a 10− to 100− fold greater affinity when activating a rabbit kidney Na-K ATPase preparation (Britten & Blank, 1968) or human red cell sodium pump activity (Cavieres & Ellory, 1974). In a recent study (Bell, Tondeur & Sargent, 1977), the seawater-adapted eel gill Na-K ATPase has been shown to be very similar in its ionic affinities to the kidney and brain Na-K ATPase preparations. It is therefore interesting to see whether the T1+ ion can mimic the K⁺ effect on Na+ and Cl⁻ fluxes in marine teleosts, to determine if activation of Na-K ATPase is directly involved in this stimulation.

All experiments were carried out on juvenile mullet, Mugil capito, weighing 5– 20 g. The fish were maintained in sea water for at least 3 weeks before use.

Blood was taken by cardiac puncture into a heparinized saline medium containing 5 mm Na EDTA. The saline was NaCl 140 mM, KC1 5 mM, Tris 20 mM, pH 7·5 at 20°C. The cells were washed three times by centrifugation (2500 g× 5 min) and loaded with ²⁴Na by incubation for 2 h in a medium containing NaCl 150 mm (+ ²⁴NaCl, 100 μCi/ml), glucose 10 mm, Tris 10 mm, pH 7·5 at 20°C. The cells were washed five times in ice cold MgCl₂ 110 mm, Tris 10 mm and dispensed into a tube to measure ²⁴Na efflux. Cells were suspended at 3% hematocrit in a solution similar in composition to the loading medium, but containing either KC1 10 mM, or TINO3 ^{0·2 or} 1 mM-Samples were run in triplicate, together with a series containing 1 mM-ouabain. At time intervals of 5, 25 and 45 min the samples were centrifuged and aliquots of the supernatant counted for ^MNa by Cerenkov radiation in a β-counter. Total ²⁴Na counts in a deproteinized cell preparation were also determined, and the efflux calculated as a rate constant.

A microsomal ATPase preparation was made from Mugil gill scrapings by homogenization and differential centrifugation following the methods previously used for intestine (Ellory & Smith, 1969) and fish gill (Bornancin & de Renzis, 1972). The final pellet was treated with sodium deoxycholate to a concentration of 1·5 mM and ouabain-sensitive ATPase activity assayed at 25°C for 15 min with incubation in a medium containing 15omM-NaCl, 2·5 mM-Mg ATPase, 15 mm-Tris, pH 7·5, with various concentrations of KC1 or T1NO₃ added. The reaction was stopped by the addition of TCA to a final concentration of 5% and the inorganic phosphate liberated assayed as previously described (Ellory & Maher, 1977).

Measurements of the transgill potential difference, and continuously monitored ²⁴Na and ³⁶C1 efflux were made following ‘rapid transfer’ of whole conscious fish from seawater to freshwater. The experimental techniques were as previously described (Maetz & Pic, 1975, 1979; Pic, 1978).

The data presented in Table 1 indicate that T1+ can substitute equally effectively and at lower concentrations for K⁺ in supporting ouabain-sensitive Na+ efflux. These cells showed a ouabain-sensitive Na+/Na+ exchange in the absence of K+ (e.g. Garrahan & Glynn, 1967) but K⁺ 10 mm or T1+ 1 mm gave a significant activation. Similar results for gill ATPase activity (Table 2) indicate that, in line with other tissues so far studied, the sodium pump in mullet cells will accept Tl+ as an effective substitute for potassium.

Table 3 presents the results for Na⁺ and Cl⁻ effluxes and P.D. measurements. The absolute values are very similar to previous observations (Maetz & Pic, 1975, 1979). Addition of 10 mm-KCl in freshwater causes an increase in the Na+ (200%) and Cl⁻ (100%) effluxes and a depolarization of the gill. In contrast the addition of 0·75 mm-TlN0₃ does not significantly alter the transgill P.D. or Na efflux. A small increase in Cl⁻ efflux is on the limit of significance (P ˜ 0·05). Thus there is no action in vivo of a T1 concentration sufficient to activate the sodium pump in erythrocytes and microsomal Na-K ATPase of the gill in vitro.

The difference between T1+ and K+ in affecting efflux across the gills could be mediated in at least three possible ways. First, neither K+ nor Tl⁺ could be reaching the Na-K ATPase, but rather the external K+ only acts on apical mechanisms independent of the Na-K ATPase; alternatively, only K⁺ reaches the Na-K ATPase, Tl+ being prevented from access to the basolateral membranes by diffusion through the cell. Finally, it is possible that both cations are activating the Na-K ATPas and increasing Na+ (and perhaps indirectly, Cl⁻) concentration in the tubular reticulum, however there is a K+-dependent step subsequent to this, at which T1+ cannot substitute effectively. Both the first and second hypotheses seem unlikely, since lanthanum added to the external medium penetrates quickly through short leaky junctions between the ionocytes (Sardet, Pisam & Maetz, 1978). This indicates a possible route, communicating with the tubular reticulum, for the entry of T1+ and K⁺ ions. Moreover, the first hypothesis is inconsistent with the effect of ouabain on ionic fluxes (Silva et al. 1977)-In the second hypothesis it is difficult to reconcile the location of Na-K ATPase on the basolateral membrane with sodium extrusion since the pumps are wrongly directed. In the third hypothesis, activation of Na-K ATPase is necessary for Na+ and Cl⁻ excretion, which is compatible with the inhibition of effluxes by ouabain, but this activation is not enough for ionic extrusion.

For Na+ extrusion, several authors have suggested that the transgill electrical potential effectively compensates the chemical gradient in seawater and that the increased Na⁺ efflux after K⁺ addition results from a further gill depolarization (Potts & Eddy, 1973; Greenwald, Kirschner & Sanders, 1974; Kirschner, Greenwald & Sanders, 1974). However, activation of Na+ efflux by K⁺ is respectively two and three times greater than one would expect from depolarization in Dormitator maculatus (Evans, Carrier & Bogan, 1974) and Mugil (Maetz & Pic, 1975, 1979). Further the Cl⁻ secretion in seawater occurs against an electrochemical gradient, and the Cl⁻ efflux is in an opposite direction from that predicted from the potential change. Thus the alteration in transgill electrical potential induced by K⁺ will only partly explain an increased Na⁻+ efflux and cannot be responsible for the Cl⁻ efflux component. It is therefore impossible that the difference between K+ and T1+ effects can be only attributed to their different effects on transgill P.D.

The magnitude of Na± and Cl⁻ gill exchanges in seawater, and the K+-activation mechanism depends on the presence of Ca²+ (and Mg²⁺) in the external solution (Bornancin, Cuthbert & Maetz, 1972; Isaia & Masoni, 1976). Since Ca²* is an inhibitor of Na-K ATPase (Fahn, Koval & Albers, 1966; Tobin et al. 1973) this involvement of Ca^21- is a further pointer to the involvement of other mechanisms in addition to the Na+ pump in this effect. Finally, the inhibition of ionic excretion and K+ activated mechanisms by colchicine (Maetz & Pic, 1977) suggests that microtubules, cell polarity and motility may play a role in Na⁺ and Cl⁻ excretion in seawater.

The authors wish to thank, for his judicious advice, Dr J. Maetz who tragically died in August 1977. J.C.E. received an EMBO short-term fellowship, during this work.

Dr C. Lucu would like to express his thanks to FAO/UNEP Joint Coordinated Project on Pollution in the Mediterranean for financial support.