1. In all marine fish, both stenohaline and euryhaline, the net sodium excretion by the gill, compensating for intestinal absorption, is the result of a Naint-Kext exchange.

  2. In a euryhaline fish, such as the eel, in which a sodium exchange diffusion (linkage of sodium influx and outflux) also occurs, the two kinds of exchange (Na-Na and Na-K) can be regarded as alternative modes of behaviour of the same transport system. In normal sea water both exchanges take place at the same time; in K-free sea water the system exchanges Na for Na exclusively and in Na-free solution containing K the system exchanges Na for K exclusively.

  3. The transport system is located on the external membrane since it is inhibited by ouabain added in the external medium.

  4. The Na-K process is ouabain-sensitive but K can antagonize the inhibitory action of the glycoside. The Na-Na process on the other hand is not ouabain-sensitive.

The ability to maintain body fluids hypotonic to the environment is a characteristic of marine teleosts. Homer Smith (1930) showed that to replace body water lost along the osmotic gradient the teleost swallows the external medium continuously, absorbing ions and water from the gut. To abolish the salt load incurred by drinking sea water the fish excretes Na and Cl through the gills. By using radioactive tracers, however, the sodium outflux through the gill has in fact been found to be more than ten times greater than the sodium extrusion expected from the drinking rate (Motais & Maetz, 1965). On experimental grounds it has been suggested (Motais, 1967) that the branchial sodium efflux in flounder and eel may be divided into at least three components, as shown in Fig. 1.

Fig. 1.

Components of the sodium outflux in seawater-adapted eel. Outflux fraction I = net flux; outflux fraction II = exchange-diffusion component; outflux fraction III = residual outflux observed after rapid transfer to fresh water. Fractions I and II are dependent on external cations, fraction III is independent.

Fig. 1.

Components of the sodium outflux in seawater-adapted eel. Outflux fraction I = net flux; outflux fraction II = exchange-diffusion component; outflux fraction III = residual outflux observed after rapid transfer to fresh water. Fractions I and II are dependent on external cations, fraction III is independent.

  1. A small net extrusion component balancing the intestinal absorption (amounting to about 100 μ-equiv/h/100 g body weight in the eel).

  2. A very large fraction, dependent of the external sodium concentration in a manner suggesting Michaelis-Menten saturation kinetics, linked with the branchial sodium influx and considered as an exchange diffusion component (amounting to about 700 μ-equiv/h/100 g body weight in the eel).

  3. A residual outflux observed after rapid transfer to fresh water, i.e. independent of external cations (amounting to about 200 μ-equiv/h/100 g body weight in the eel).

Indirect evidence suggests a close correlation between the first two components (Maetz, Motais & Mayer, 1969). Until recently the nature of the sodium-excreting pump was completely unknown. But Maetz (1969) showed in the flounder that the net Na extrusion rate is identical to the K influx and is blocked in K-free sea water. So he considered that the sodium-excreting pump is a sodium-potassium exchange pump. Furthermore, since external Na competes with K on the K site of the pump and since sea water contains 50 times more Na than K, he suggested that at least a part of the exchange-diffusion fluxes are the result of this competitive process.

This model is significant for two reasons. First, it adequately explains the correlation between the Na-Na and Na-K exchanges mentioned above, and secondly, it agrees well with the hypothesis that Na-K-activated ATPase is involved in the excretion of sodium ions by the gills of marine teleosts - indeed parallel variations of the enzyme activity and gill ionic transfers either during adaptation to salt water (see Kamiya & Utida, 1968, 1969; Epstein, Katz & Pickford, 1967; Motais, 1970a; Zaugg( & McLain, 1969), or in seawater-adapted fish after hypophysectomy (Epstein et al. 1967) or after treatment with actinomycin D (Motais, 1970b) are suggestive of an important role of this enzyme in sodium extrusion by the gill.

But to be acceptable this model must fulfil a specific requirement with regard to the action of cardiotonic steroids. As suggested by the facts that the sodium extrusion is sensitive to environmental K+ and that external Na competes on the K site, it seems reasonable to assume that the sodium-excreting pump is located on the external membrane with the K site facing outward (Maetz, 1971; Kirschner, 1969). Ouabain and related cardiotonic steroids are known to be specific inhibitors of active cation transport in many types of cell. Their action is known to be asymmetric, acting only on the side of the membrane from which K is actively transported - that is, for marine teleosts, the outer border of the branchial epithelium. Thus it is possible to subject Maetz’s hypothesis to a very simple test: Na extrusion and at least part of the Na-Na exchanges should be sensitive to ouabain added to sea water. It has been claimed, however, that addition of ouabain to the medium had no effect either on the Na turnover rate in eels (J. Maetz, personal communication) and in flounders (Kirschner, 1969) adapted to sea water or on the net flux (Kamiya, 1967). This lack of agreement between the model and these particular data requires elucidation. In this study we attempt to clarify this point.

(A) Animals

The euryhaline eels (Anguilla anguilla) averaged 70 g in weight and came from the estuary of the Rhône, and euryhaline flounders (Platichthys flesus) (200–300 g) came from the estuary of the Loire. The stenohaline fish used were sea perch (Serranus scriba) caught in the bay of Monaco and of an average weight of 50 g.

(B) Outflux measurements by rapid transfer experiments

Since the net extrusion component of the sodium efflux only represents a very small fraction of the total sodium outflux through the gill, the Na turnover rate should not be significantly modified if this component alone is inhibited by ouabain. To test the effect of ouabain, all the measurements were made with a more accurate technique permitting the comparison of the sodium outfluxes in media with or without ouabain during rapid transfer experiments. Techniques and calculations had previously been described by Motais (1967).

Plasma concentrations of sodium and potassium were measured by flame photometry (Eppendorf).

As shown in Fig. 2, there is evidence that in the seawater-adapted eel ouabain (10−1 M) added in the external medium produces a small but consistent reduction of the sodium outflux ( –16·1 ±2·3%, n = 17, P < 0·001). This reduction amounts to a value at least equal to the net extrusion rate of sodium. Furthermore, a 24 h exposure to 10−4 M ouabain in sea water causes an increase of the internal sodium (control fish, 140·5 ±4·7 m-eqmv/1, n = 8; treated fish, 167·0+ 1·3, n = 8, P <0·001), indicating an impairment of the sodium-excreting pump. From this first set of data it can be assumed that in the eel a ouabain-sensitive pump, located on the external membrane, is responsible for the net excretion of sodium.

Fig. 2.

Comparison of the sodium outflux in different media with or without ouabain. Experiments carried out on the seawater-adapted eel. C = control; O = ouabain-treated fish (ouabain added in the external medium). The outflux is expressed in % outflux in sea water control fish. The cross-hatched column represents the basic sodium outflux measured in fresh water, which it can be assumed does not pass through the pumping system since it is independent of external cations.

Fig. 2.

Comparison of the sodium outflux in different media with or without ouabain. Experiments carried out on the seawater-adapted eel. C = control; O = ouabain-treated fish (ouabain added in the external medium). The outflux is expressed in % outflux in sea water control fish. The cross-hatched column represents the basic sodium outflux measured in fresh water, which it can be assumed does not pass through the pumping system since it is independent of external cations.

Ouabain is generally thought to inhibit Na-K pumps. The existence of a Na-K exchange in eels can be demonstrated directly by a comparison of the sodium outflux in sea water (SW), in fresh water containing 10 mM-K (FWK) which is the concentration of K in sea water and in K-free fresh water (FW) (Fig. 2). After transfer from SW to FWK the sodium outflux remains unchanged ( – 0·4 ±9·1 %, n = 10), while the transfer from SW to FW results in a reduction of the flux by 77·6 ± 3·5 % (n = 20). These results indicate first that under these particular conditions an important Na-K exchange exists, and thus under normal conditions in sea water net sodium extrusion is presumably brought about by a Na-K pump as suggested by the results with ouabain. Secondly, they show that in the absence of external sodium, i.e. in FWK, the Na-Na exchange (the component II of the sodium outflux) is replaced by a K-Na exchange. In other words, K can act as a substitute for Na in the exchange diffusion mechanism. This obviously must result in a rapid drop of the plasma sodium concentration and an increase of the potassium concentration. Indeed it has been shown that after 30 min in FWK there was a 19 ± 1 m-equiv/1 (n = 3) drop in plasma sodium. The potassium concentration of the plasma increased threefold. But it must be noticed that after 24 h in this solution the animals were in good condition and the plasma concentration of Na and K were similar to those measured 30 min after transfer. Presumably there is a rapid reduction of activity in the pump to enable the animals to survive in KC1. According to Motais (1967), adaptation of a fish to any change of external salinity would be accompanied by a variation in the quantity of available carrier in the gill epithelium. These results suggest that it is the Na rather than the K which induces these variations. It was also found that when ouabain was added to the 10 mM-KCl solution there was a reduction of the sodium outflux ( –27·0±6·0%, n = 10), this reduction being greater than that observed when the ouabain was added to sea water. From this one may deduce that the exchange-diffusion mechanism, which is insensitive to ouabain when it catalyses the Na-Na exchange, becomes sensitive to the inhibitor when the exchange is Na-K. This would indicate, as Maetz suggested, that both Na-Na and Na-K exchanges are catalysed by a common mechanism.

Nevertheless, under these experimental conditions (10 mM external K) ouabain does not completely inhibit the external cation-dependent fraction of the outflux (i.e. fractions I+ 11) as one would expect if the totality of the exchanges passed by a Na-K pump. It is well known that in the human red corpuscle K directly antagonizes the fixation of ouabain on the membrane (Hoffman, 1969), and that the inhibition by cardiac glycoside can be completely reversed by increasing the external potassium concentration (Glynn, 1957). It is thus reasonable to suppose that in the eel experiments ouabain inhibition in the 10 mM-KCl solution was not complete as there would have been antagonism between the K and the ouabain. It can be seen from Fig. 2 that the inhibitory effect of ouabain is clearly enhanced when the K concentration is reduced from 10 to 2 mM. The sodium outflux in 2 mM-KCl is lower than in 10 mM-KCl - a result which is similar to that obtained in the flounder by Maetz. It amounts to 51·7 ± 7·4% (n = 8) of the flux measured in sea water (which is equivalent to 10 mM-kCl). About half of this flux corresponds to the basic sodium outflux in fresh water, which it can be assumed does not go through the pumping system since it is independent of external cations. When ouabain is added to this medium the flux falls ( – 46·5 ±4·8%, n = 8, P < 0·001) to a value more or less that of the basic flux. Thus when only 2 mM of K are present in the external medium ouabain inhibition of the exchange system would appear to be virtually complete.

We have pointed out that when no Na is present in the external medium the Na-Na exchanges (which are insensitive to ouabain) are replaced by ouabain-sensitive Na-K exchanges. The question arises as to whether the Na-K exchanges which bring about the net sodium excretion in sea water are replaced by Na-Na exchanges in K-free sea water. No significant reduction of the Na efflux was observed when eels were transferred from normal SW to K-free SW (reduction of – 5·7±7·2%, n = 8), indicating that the net excreting pump exchanging internal Na for external K in normal sea water switches to exchanging Na for Na after removal of the external K. Furthermore, whereas in sea water the sodium efflux is ouabain-sensitive, in K-free sea water it is no longer so (inhibition –4·1 ±4·5%, n = 8). Thus although these Na-Na exchanges are going through the excreting pump, they are ouabain-insensitive.

In conclusion, our observations on the branchial sodium pump in the eel suggest that the same system is responsible for the two components of the Na efflux which are dependent on the external cation (fraction I or net excretion resulting from a Na/K exchange and fraction II or exchange diffusion of Na). The two kinds of exchange should be regarded as alternative modes of behaviour of the cation transport system: in normal sea water both exchanges proceed at the same time, in K-free sea water the system exchanges Na for Na exclusively, and in a Na-free solution containing 10 mM-K only the Na-K exchange occurs. The Na-K process is ouabain-sensitive, but K, as in the red blood cell, can antagonize the inhibitory action of the glycoside. The fraction of the system which operates as a Na-Na exchange is not ouabain-sensitive, unlike that of the human blood red cell (Garrahan & Glynn, 1967). This lack of sensitivity of the pump when it catalyses Na-Na exchanges may be explained in at least two ways: interactions between ouabain and pump is prevented either by an antagonism resulting from steric interference between cation and inhibitor, or by a change in membrane conformation when the Na-K pump turns into a Na-Na pump. The hypothesis of a steric interference seems relatively improbable since at one and the same concentration of external sodium (500 mM) ouabain can have either an inhibitory effect (in sea water) or none (in K-free sea water).

Similar results have been obtained in the flounder to those in the eel, so it may be assumed that the central mechanism of the excreting system is also a Na-K pump. There are, however, two differences in the reactions of the two euryhaline fish which should be noted:

  1. In the flounder no inhibition by ouabain occurs in normal sea water although evidence of a Na-K exchange has been clearly demonstrated (Maetz et al. 1969). This discrepancy with the data obtained from the eel could be explained by assuming that the relative affinities of ouabain and K for the reactive centres are different in the two fishes.

  2. In the flounder it has been shown (Maetz, 1969) that a large reduction of the Na outflux occurs ( – 58·6 ± 7·5 %, n = 10) when sea water is replaced by a 10 mM-KCl solution, and further increase of the potassium concentration of the external solution does not alter this value. It would seem that in the flounder, in contrast to the eel, about 40% of the sodium efflux in sea water does not go through the Na-K pump. Whether this external sodium-sensitive component of the efflux is diffusive in nature (Maetz, 1971) or represents a ouabain-insensitive exchange-diffusion mechanism distinct from the pump (Motais & Garcia-Romeu, 1972) remains to be elucidated.

    In a stenohaline fish, the sea perch Serranas scriba, in which no Na exchange diffusion occurs (Motais, 1967), no inhibition of the Na efflux by ouabain is observed in sea water ( –0·1±2·9, n – 10), but transfer from normal SW to K-free SW induces a reduction of Na outflux ( – 20·0 ± 5·6%, n = 6, P < 0 ·02), suggesting that, in these fish also, a Na-K exchange pump exists.

The present data indicate that in all the marine fish studied, whether euryhaline or stenohaline, sodium excretion by the gill is an active process, a fact which has in the past been questioned (Kirschner, 1970), mediated by a mucosal pump. They also permit to consider that, at least in the eel, the so-called sodium exchange-diffusion component (Motais et al. 1966) is a real exchange diffusion and not a true leak as suggested by Smith (1969) and Maetz (1971).

  1. In all marine fish, both stenohaline and euryhaline, the net sodium excretion by the gill, compensating for intestinal absorption, is the result of a Naint-Kext exchange.

  2. In a euryhaline fish, such as the eel, in which a sodium exchange diffusion (linkage of sodium influx and outflux) also occurs, the two kinds of exchange (Na-Na and Na-K) can be regarded as alternative modes of behaviour of the same transport system. In normal sea water both exchanges take place at the same time; in K-free sea water the system exchanges Na for Na exclusively and in Na-free solution containing K the system exchanges Na for K exclusively.

  3. The transport system is located on the external membrane since it is inhibited by ouabain added in the external medium.

  4. The Na-K process is ouabain-sensitive but K can antagonize the inhibitory action of the glycoside. The Na-Na process on the other hand is not ouabain-sensitive.

We wish to thank Drs F. Garcia-Romeu and J. Maetz for their critical comments.

Epstein
,
F. H.
,
Katz
,
A. I.
&
Pickford
,
G. E.
(
1967
).
Sodium and potassium activated adenosine triphosphatase of gills: role in adaptation of teleosts to salt water
.
Science, N. Y
.
156
,
1245
7
.
Garrahan
,
P. J.
&
Glynn
,
I. M.
(
1967
).
The behaviour of the sodium pump in red cells in the absence of external potassium
.
J. Physiol., Land
.
192
,
159
74
.
Glynn
,
I. M.
(
1957
).
The action of cardiac glycosides on sodium and potassium movements in human red cells
.
J. Physiol., Lond
.
136
,
148
73
.
Hoffman
,
J. F.
(
1969
).
The interaction between tritiated ouabain and the sodium-potassium pump in red blood cells
.
J. gen. Physiol
.
54
,
343
53
.
Kamiya
,
M.
(
1967
).
Changes in ion and water transport in isolated gills of the cultured eel during the course of salt adaptation
.
Annot. Zool. Jap
.
40
,
123
9
.
Kamiya
,
M.
&
Utida
,
S.
(
1968
).
Changes in activity of Na-K activated adenosine triphosphatase in gills during adaptation of the Japanese eel to sea water
.
Comp. Biochem. Physiol
.
26
,
675
85
.
Kamiya
,
M.
&
Utida
,
S.
(
1969
).
Sodium-potassium activated ATPase activity in gills of fresh water, marine and euryhaline teleosts
.
Comp. Biochem. Physiol
.
31
,
671
4
.
Kirschner
,
L. B.
(
1969
).
ATPase activity in gills of euryhaline fish
.
Comp. Biochem. Physiol
.
29
,
871
4
.
Kirschner
,
L. B.
(
1970
).
The study of NaCl transport in aquatic animals
.
Am. Zool
.
10
,
365
76
.
Maetz
,
J.
(
1969
).
Sea-water teleosts: evidence for a sodium-potassium exchange in the branchial sodium-excreting pump
.
Science, N.Y
.
166
,
613
15
.
Maetz
,
J.
(
1971
).
Fish gills: mechanisms of salt transfer in fresh water and sea-water
.
Phil. Trans. Roy. Soc. B
262
,
209
49
.
Maetz
,
J.
,
Motáis
,
R.
&
Mayer
,
N.
(
1969
).
Isotopic kinetic studies on the endocrine control of teleostean regulation
.
Proc. 3rd Int. Cong. Endocr., Mexico
(ed.
C.
Gual
), pp.
225
32
.
Amsterdam
:
Excerpta Med
.
Motáis
,
R.
(
1967
).
Les mécanismes d’échanges ioniques branchiaux chez les Téléostéens
.
Am. Inst. Ocean
.
45
,
1
84
.
Motáis
,
R.
(
1970a
).
Les mecanismes branchiaux des échanges ioniques chez les Téléostéens en rapport avec la salinité: aspect biochimique
.
Bull. Inform. Sci. Tech. Commis. Energ. At. Fr
.
146
,
3
19
.
Motáis
,
R.
(
1970b
).
Effect of actinomycin D on the branchial Na-K dependent ATPase activity in relation to sodium balance of the eel
.
Comp. Biochem. Physiol
.
34
,
497
501
.
Motáis
,
R.
&
Maetz
,
J.
(
1965
).
Comparaison des échanges de sodium chez un Téléostéen euryhalin (le Fiet) et un Téléostéen sténohalin (le Serran) en eau de mer. Importance relative du tube digestif et de la branchie dans ces échanges
.
C. r. hebd. Séanc. Acad. Sci
.,
Paris
261
,
532
5
.
Motáis
,
R.
&
Garcia-Rombu
,
F.
(
1972
).
Transport mechanisms in the Téleostean gill and Amphibian skin
.
A. Rev. Physiol
.
34
,
141
76
.
Motáis
,
R.
,
Garcia-Romeu
,
F.
&
Maetz
,
J.
(
1966
).
Exchange diffusion effect and euryhalinity in Teleosts
.
J. gen. Physiol
.
50
,
391
422
.
Smith
,
H. W.
(
1930
).
The absorption and excretion of water and salts by marine teleosts
.
Am. J. Physiol
.
93
,
480
505
.
Smith
,
P. G.
(
1969
).
The ionic relations of Artemia salina L. II. Fluxes of sodium, chloride and water
.
J. exp. Biol
.
51
,
739
57
.
Zaugc
,
W.
&
Mclain
,
L.
(
1969
).
Inorganic salt effects on growth, salt adaptation, and gill ATPase of Pacific salmon
.
In Fish in Research
, ed.
J. E.
Halver
and
O.
Neuhaus
.
New York
:
Academic Press
.