Branchial Sodium Transport Mechanisms in Scyliorhinus Canicula: Evidence for Na⁺/NH₄⁺ And Na⁺ H⁺ Exchanges and for A Role of Carbonic Anhydrase

Re-invasion of the sea occurred at the end of the palaeozoic era and re-adaptation of the body fluids to the hyperosmotic environment followed two different patterns according to whether the fishes emerging from the protovertebrate phylum were bony or cartilaginous fishes. While the former became hypo-osmotic regulators, the latter remained more or less iso-osmotic to their environment. To keep the osmolarity of the body fluids at the same level as that evolved during freshwater adaptation, the teleosts developed an extrarenal salt excretion mechanism in their gills. The elasmobranchs, however, retained the nitrogenous waste urea as an ‘osmotic filler’ making their internal medium iso-osmotic to sea water.

Various studies suggest that the branchial ionic transport mechanisms of marine elasmobranchs are, on the one hand, widely different from those of marine teleosts and, on the other hand, rather similar to those of freshwater fishes despite the huge salt concentration difference of their respective external media.

Maetz & Lahlou (1966) observed that in Scyliorhinus, a marine elasmobranch, the turnover rate of the exchange between external and internal sodium involves only about 0·5% of the internal sodium per hour, a value similar to that calculated in freshwater fishes and 30−100 times lower than that characterizing sea-water teleost (Motais, 1967). When branchial sodium influx and efflux are compared by means of isotopic techniques it appears that in both marine elasmobranchs and freshwater teleosts a net uptake of sodium occurs, since the influx is higher than the efflux (Payan & Maetz, 1970; Maetz, 1956b). In marine teleosts the reverse situation prevails and a net excretion of sodium is observed (Motais, 1967).

Jampol & Epstein (1970) observed that the branchial Na-K-dependent ATPase, an enzyme which is presumably linked with sodium transport, has a similar activity level in freshwater teleosts and in marine elasmobranchs, while the level is much higher in sea-water teleosts.

The diffusionsl water permeability of the elasmobranch gill is similar or greater than that of the freshwater teleost gill, while in sea-water teleosts the branchial water permeability is generally much lower than that of their freshwater relatives (Payan, Goldstein & Forster, 1973; Payan & Maetz, 1971; Motais et al. 1969; Evans, 1969). The present study further emphasizes the similarities between the ionic transport mechanisms in the marine elasmobranch Scyliorhinus and in freshwater teleosts such as Carassius auratus and Salmo gairdneri. In both these freshwater fish recent evidence demonstrated that the branchial sodium uptake mechanism is associated with the maintenance of the acid-base balance of the body fluids by means of a linkage between uptake of sodium and excretion of NH₄⁺ or H⁺ ion (Maetz & Garcia Romeu, 1964; Maetz, 1971, 1972a, b,c; Kerstetter, Kirschner & Rafuse, 1970). Moreover, in both these fish branchial carbonic anhydrase activity plays an essential role in the sodium uptake mechanism. Injection of acetazolamide, a powerful inhibitor of carbonic anhydrase, is followed by a considerable depression of the sodium influx (Maetz, 1956b; Kerstetter et al. 1970).

Carbonic anhydrase, has also been found in the elasmobranch gill (Hodler et al. 1953). Iⁿ elasmobranchs, as in teleosts, the gill is the main side of ammonia excretion (Smith, 1929, 1931)- Furthermore, after acute perturbations of the acid-base balance, Mudaugh & Robin (1967) observed that the gill of Squalus acantinas is the main site of H⁺ ion excretion.

This work furnishes evidence that in Scyliorhinus excretion of ammonia and H⁺ by the gill are also linked to sodium uptake and that carbonic anhydrase activity is involved in the process.

Individuals of Scyliorhinus canicula of both sexes and mean weight of 300 g were used. The fish were caught on the northern continental shelf off Corsica and kept in running sea water at 15−18 °C and fed up to a week previous to the flux measurements.

Two techniques were used. The first allowed for the determination of both sodium influx and sodium space.

Technique A necessitated a total-body counter PACKARD (model 446) connected to a multiscale analyser INTERTECHNIQUE (Model SA 43) giving a visual display and a digital print-out of the successive 2 min radioactive counts.

The fish was kept for a time t (usually 30−60 min) in a radioactive sea-water solution containing 50−100μCi.l₋₁, at a constant temperature (17° ± 1 °C). It was then placed in a cylindrical aquarium, inside the body counter and rinsed continuously with running sea water during the counting. After about 10 min, the radioactivity of the fish stabilized at a constant level (Q_int). The fish was then removed from the counter and a blood sample was taken (from the caudal artery) by means of a heparinized syringe.

This first influx sequence was followed by several others lasting for 1−20 h, the fish being kept overnight in the radioactive bath.

Plasma radioactivity was measured in a gamma-well counter (MECASERTO) and compared with that of an aliquot of the radioactive bath. This aliquot was also counted in the total-body counter to allow for the assessment of the relative yields of both counters. The decay of the isotope was taken into account when calculating the values. As the sodium turnover rate is very low (0°5%.h⁻¹) the radioactive loss by the fish may be neglected in the influx calculations.

Fig. 1 illustrates the evolution of the sodium space as a function of time calculated for 1, 2, 3, 4 and 20 h of content with ²⁴Na. These values were used for the calculation of the influxes according to the second technique.

Technique A was used to study the effects of handling on the sodium influx. The animals were placed in a closed-circuit aquarium at 17 °C the day previous to the influx measurement. ²⁴Na was added to the bath without disturbing the fish. Thus, the first influx period (30−60 min) represents the control period, while the following periods correspond to fluxes obtained after handling the fish.

For the experiments of long duration (17 h) the fish were not catheterized. A blood sample was taken by mean of a syringe at the end of the influx periods.

The day before the flux measurements, the fish receives an intra-arterial injection of 200 μCi ²⁴Na through an in-dwelling catheter. This procedure allows for the sodium space to be tagged homogeneously, yet because of the slow sodium turnover very little tracer is lost (see Payan & Maetz, 1970). At the beginning of the experiment the fish is placed in a closed circuit (about 2 1/fish) of aerated sea water at a constant temperature (17°± 1 °C). The experimental protocol is similar to that of the influx measurements of type B (short duration). The experiments are divided into two or three periods, a control period followed by one or two experimental periods. The fish receives either an ammonium salt or acid solution at the beginning of the first experimental period and an acetazolamide injection at the beginning of the second period. In one experiment acetazolamide was given immediately after the control period.

Samples of the external medium were taken every half hour in order to follow appearance rates of the ²⁴Na and ammonia. Blood samples were taken at the beginning of the control period, at the end of the experiment and between each period in order to allow for the measurement of the plasma ²⁴Na specific activity (SRA_int) and internal ammonia level. The sample radioactivity was measured in a well counter MECASERTO. SRA_int was found to be constant during the experiments.

In the present investigation the total sodium efflux was thus calculated, since the urinary papilla and the rectal gland were not catheterized.

Ablation of the hypophysis was performed according to the technique described previously (Payan & Maetz, 1970, 1971). The animals were operated about 3 months before the flux experiments.

Ammonia was given at the dose of 300 μ-equiv. (100 g) ⁻¹ in the form of equimolar solutions of ammonium sulphate (3 experiments), ammonium acetate (2 experiments), ammonium bicarbonate (2 experiments). While the former solutions were acid (pH < 6), the latter were alkaline (pH > 8).

H⁺ was given at a dose of 100 μ-equiv. (100 g) ⁻¹ in the form of a molar solution of HC1. Acetazolamide (DIAMOX) was given at a dose of 10 mg (100 g) ⁻¹ in the form of a 50 g l⁻¹ solution (pH ∼ 8).

The volume injected, 900 μl/fish for NH⁺₄ and 300 μl for H⁺, was delivered in approximately 15 min.

pH was measured in both external and internal media (whole blood) with a TACUSSEL macroelectrode and an IL microelectrode respectively.

Sodium concentration was measured in a flame photometer EPPENDORF. Ammonia was measured with the help of the TECHNICON autoanalyser (see Maetz, 1972a). To avoid precipitation of the calcium during treatment of the sea-water samples it was found necessary to saturate the samples with sodium oxalate.

Table 1 shows that handling produced a significant (230% increase of the influx f_in. This effect was observed in intact as well as in hypophysectomized fish. In order to evaluate the duration of the stress effect, the fluxes were measured in one fish during 5 successive hours immediately after each handling. The fluxes were 112, 86, 80, 65 and 67 μ-equiv. h⁻¹ (100 g) ⁻¹. During the 9 h overnight period after the last handling the mean flux was found to be 54 μ-equiv. Thus, the effect of handling had almost worn off during the 5 h following the initial effect, although the handling procedure was repeated every hour.

Table 2 shows the results obtained with technique B (short duration). It may be noted that in this table all the control periods, including those preceding HC1 injection (see below), have been pooled. From the high value of the control flux it is apparent that the technique B causes stress in the fish. Nevertheless, injection of an ammonium salt solution produced a 120% increase of the sodium influx, an increase which lasted for at least 4 h, but owing to the small number of animals followed during the second experimental period the values obtained are not significantly different from those obtained during the control period. It is important to note that an increase of the influx was observed irrespective of the anion accompanying ammonia.

Table 3 shows that ammonium salt injection produced metabolic acidosis, but that the degree of acidosis varied with the accompanying anion. With sulphate, 1 h after injection, blood pH reached 7·04 (3 experiments) with acetate, it reached 7·24 (n = 2) and with bicarbonate, 7·51 (n = 2). In Table 3 all pH values have been pooled, 4 h after injection of ammonium sulphate blood pH was still significantly lower than during the control period.

Blood ammonia levels are given in Table 4. It may be seen that the injected ammonia load (300 μ-equiv.) was eliminated within 2−3 h after injection, and ammonia excretion which increased 14-fold during the hours following injection returned to its original level within 2 h (see Table 2). Fig. 2 represents a typical experiment showing the effect of an ammonia load on sodium influx and ammonia excretion.

Table 5 shows that HCl injection produced a 110% increase of the sodium influx lasting for at least 3 h. Fig. 3 is a typical experiment showing a case in which the increase lasted 4 h.

Table 3 shows that HCl injection was followed by a severe acidosis, more severe than after ammonia loading.

It may be noted that ammonia excretion remained unchanged after HCl injection. This ammonia excretion was followed in four fish in which the rate was found to be 9·2 ± 2·9μ moles h₋₁ (roo g) ₋₁, a value not significantly different from the control value given in Table 2.

The effect of this carbonic anhydrase inhibitor was ascertained in long-term and short-term experiments using the technique B for influx measurements.

In the short-term experiments acetazolamide was injected at the beginning of the second experimental period in three fish treated with ammonium acetate. In an additional experiment acetazolamide was given after the control period. The results were similar. Table 6 shows that acetazolamide produced a severe depression of about 75 % of the sodium influx observed during the control period. In the three fish treated with ammonium acetate before the acetazolamide injection the reduction of the influx was even more important attaining about 85−90% of the flux observed during the experimental period. Fig. 4 illustrates a typical experiment showing the effects of ammonia-loading and of acetazolamide.

In the long-term experiments lasting 17 h, two groups of fish were compared, a control group of fish receiving a 15% NaCl solution and an acetazolamide-treated group of fish. The controls showed a relatively low sodium influx as expected in view of the transient effects of handling (see above). As shown in Table 6, the inhibitory effects of acetazolamide on sodium influx were nevertheless quite clear-cut. Thus, at the dose utilized in the present experiments, acetazolamide produces a long-term inhibition.

In Table 3 it can be seen that in Scyliorhinus, acetazolamide produces an acidosis, a decrease of blood pH being observed for at least 4 h after the injection.

The results obtained during short-term experiments on stressed fish are summarized in Table 7. It may be seen that both types of treatment produced an increase of the total sodium efflux by about 100%. Fig. 5 illustrates a typical experiment concerning the effects of HCl.

Fig. 5 also shows that acetazolamide injection does not produce any effect of sodium efflux. Table 6 summarizes the result of sodium efflux measurements in four fish receiving acetazolamide injection. Three were studied 1 or 2 h after HC1 injection, while the fourth received acetazolamide immediately after the control period. No significant effect of acetazolamide was observed.

Previous observations from our laboratory concerned sodium exchanges in Scyliorhinus studied for 20−24 h (Maetz & Lahlou, 1966). Such fishes may be considered as unstressed since the effects of initial handling wear off within 4 or 5 h. The fluxes observed (f_in = 56 ±5 μ-equiv. h⁻¹ (100 g)⁻¹ (n = 5), and f_out = 69 ± 13μ-equiv. h⁻¹ (100 g)⁻¹ (n = 9)) were found to be in balance.

The values of f_in and f_out reported in the present study for fish immediately after handling (compare Tables 2 and 6) are strikingly different, f_in being about 6 times greater than f_out. Thus handling results in an important pertubation of the sodium balance.

The stimulating effect of handling on the influx has been amply documented above. It may be noted that the mean influx value given in Table 1 for unstressed fish is not significantly different from that reported by Maetz & Lahlou (1966).

The mean efflux value given in Table 6 for handled fish is significantly smaller than that previously reported for unstressed fish (Maetz & Lahlou, 1966; Payan & Maetz, 1970). This difference suggests that handling results in a reduction of the sodium efflux. In unpublished experiments we observed that handling causes antidiuresis in Scyliorhinus. Urine flow was found to be 18·8 ± 1·4 μl h⁻¹ (100 g)⁻¹ in 5 fish immediately after handling and 30·5 ± 4·0 μl 24-48 h later. At the same time urinary sodium excretion rose from 5·3 ± 0·7μ-equiv. h⁻¹ (100 g)⁻¹ to 10·5 ± 1·0μ-equiv. Thus the reduction of the sodium efflux after handling results at least in part from a 50% diminution of the urinary sodium loss. No information is available concerning the effects of handling on the sodium secretion by the rectal gland, but in undisturbed Scyliorhinus this excretion has been shown to play a minor role (about ⁠) in the total sodium efflux (Payan & Maetz, 1970). The branchial sodium efflux on the contrary contributes to about two-thirds of the total sodium efflux. Thus, handling almost certainly results in a decrease of this parameter of the sodium balance.

It may be noted that in freshwater Carassius (Maetz, 1956b) as in Scyliorhinus, handling produces a perturbation of the sodium exchanges in the direction of the chemical gradient of this ion across the gill resulting in a positive sodium net flux in the marine fish and a negative sodium net flux in the freshwater fish.

In both fish the effects of handling wear off within a few hours. This may result from the endogenous release of an endocrine factor with a short half-life. The fact that handling still produces a perturbation of the sodium exchanges in hypophysectomized Scyliorhinus seems to exclude the hypophysial hormones from playing a role in this process. It may be noted that 3 months after ablation the neurohypophysis may have regenerated in Scyliorhinus as in other fish (see review by Maetz, 1968). Further investigations are planned to study the effects of catecholamines and of neurohypophysial hormones on the sodium exchanges of Scyliorhinus.

Injection of ammonia salt solutions (especially sulphate and acetate) and of HCl solution result in long-lasting metabolic acidosis (see Table 3). There are three possible ways to restore the acid-base balance: the passive intervention of the blood buffers, or an increased H⁺ ion excretion by the active intervention of either kidney or gill.

According to Murdaugh & Robin (1967) the buffering capacity of the elasmobranch is extremely poor. For instance, in Squalus acanthias plasma bicarbonate concentration is as low as 6 m-equiv. 1⁻¹ (Burger, 1967).

The role of the kidney in the maintenance of the acid-base balance is reputed to be negligible in elasmobranchs. Urine pH remains constant despite important variations in blood pH produced by injection of acids, alkali or acetazolamide (W. W. Smith, 1939 ; Hodler et al. 1955) or by bubbling a mixture of oxygen and CO₂ in the external medium (Cross, Packer, Linta, Murdaugh & Robin, 1969). Cross and his colleagues also showed that the contribution of the kidney in ammonia elimination represents barely 1 % of the total excretion rate. This refutes the possibility that renal NH₄⁺ excretion plays a role in H⁺ elimination. Thus the gill is obviously the main route of the acid excretion maintaining the internal pH. Murdaugh & Robin (1967) suggested that acid is excreted both in the form of CO₂ resulting from the buffering activity of the blood, and in the form of H⁺ or NH₄⁺. Branchial clearance studies of the ‘non-carbonic acid’ component of blood in Squalus acanthias following experimental metabolic or respiratory acidosis suggest that H⁺ is excreted as such and that NH₄⁺ plays a minor role in proton elimination (Peirce, 1967; Murdaugh & Robin, 1967; Peirce & Kent, 1968; Cross et al. 1969).

The results reported above show that in the case of metabolic acidosis elimination of H⁺ is linked to Na⁺ absorption, the gill function being directed towards the replacement of a ‘fixed acid’ by a ‘fixed base’. It may be noted that the net sodium uptake is increased by about 100 μ-equiv. h⁻¹ (100 g)⁻¹ after injection of HCl. This value corresponds to that found by Murdaugh & Robin (1967) in their branchial clearance studies of ‘non-carbonic acid’ in Squalus injected with HCl at the same dose as that given to Scyliorhinus. Our observation showing that ammonia excretion remains unchanged at a level of about 10 μ-equiv. h⁻¹ (100 g)⁻¹ confirms that in this experimental condition ammonia does not play any role in the process of acid excretion. Ammonia excretion also remains unchanged during respiratory acidosis (Cross et al. 109).

The stimulating effects of handling on the sodium uptake may at least in part be explained by our observations on blood pH which we found to be significantly lower in disturbed than in undisturbed fish (7·60 ± 0·021 for n = 18 against 7·72 ± 0·013 for n = 4; P < 0 ·001). Peirce (1967) noted that handling is followed in Squalus acanthias by metabolic acidosis resulting from an endogenous release of lactic acid. Thus the increased sodium influx observed after handling may reflect increased H⁺ ion elimination by the gill. Whether this metabolic perturbation is related to the release of hormones - neurohypophysial principles or catecholamine - remains to be studied.

The effects of ammonia-loading on the sodium influx are very similar to those produced by HCl injection. Table 2 clearly shows that the increased ammonia excretion rate lasts only for 2 h and that only about two-thirds of the injected dose is eliminated at that time. It is probable that the remaining 100 μ-equiv. are metabolized in the liver by way of the Krebs-urea cycle. Urea production from ammonia entails metabolic acidosis. Thus the decreased blood pH which is reported in Table 3 can be explained. It is important to note, however, that the increase in sodium influx produced by ammonia-loading lasts much longer than the increase in ammonia excretion. It is probable that the initial NH₄⁺/Na⁺ exchange is progressively replaced by a H⁺/Na⁺ exchange during the second hour following the injection.

The fact that in Scyliorhinus Na⁺ uptake is linked to H⁺ excretion is illustrated in Fig. 6. In this figure the sodium influx has been plotted against the blood pH observed simultaneously irrespective of the origin of the metabolic acidosis (ammonia loading or HCl injection). An excellent correlation is obtained showing that sodium influx is inversely proportional to blood pH.

The intervention of carbonic anhydrase in the mechanism of sodium uptake by the gill is indicated by the inhibitory effect of acetazolamide. At the same time acetazolamide produces a decrease of the blood pH (see Table 3) which results from the inhibition of the enzyme located in the red cells and from a decreased efficiency of the CO₂ elimination by the gill (Hodler et al. 1955). As a decreased blood pH would normally induce an increase H⁺ ion elimination by the gill linked with an increased Na⁺ uptake, the inhibitory effect of acetazolamide is even more apparent. It may be noted that acetazolamide is also effective in inhibiting the enhanced Na⁺ uptake observed after ammonia or HCl injection.

The effects of ammonia-loading are similar in Scyliorhinus canicula and in the two freshwater teleosts Carassius auratus (Maetz & Garcia Romeu, 1964; Maetz, 1972 a, b) and Salmo gairdneri (Kerstetter et al. 1970). An increased sodium influx and sodium net uptake is observed in both types of fish. In the teleosts there is, however, no consistent increase of the sodium efflux.

The effects of acetozolamide are identical in both groups of fish. A depression of the sodium influx is observed, the sodium efflux remaining unchanged.

Thus it seems probable that the model proposed by Maetz & Garcia Romeu (1964) for the mechanism of sodium uptake by the gill of freshwater teleosts also applies to Scyliorhinus. This model has recently been modified (Maetz, 1971) to take into account the possibility of a Na⁺/H⁺ exchange in addition to the Na⁺/NH₄⁺ exchange, and NH₈ excretion in the unionized form (Kerstetter et al. 1970; Maetz, 1972 c). The occurrence of an Na⁺/H⁺ exchange is even better demonstrated in Scyliorhinus than in Carassius. In the latter, injection of HCl is only followed by an increased Na⁺ net uptake in 50% of the fish investigated (Maetz, 19720). In Scyliorhinus the effect of HCl injection is very clear-cut. The high buffering capacity of the blood in Carassius and the fact that in freshwater teleosts the kidney definitely plays a role in the maintenance of the acid-base balance (Maetz, 1956 a) may account for the 50% lack of response. A direct demonstration of H⁺ excretion by the gill of Carassius has, however, been obtained recently by following the titrable adidity of the external medium. This is only possible in the absence of parallel C1⁻/HCO₃⁻ exchanges by the gill, the fish being kept in buffered sodium sulphate solutions (Maetz, 1972 c ; see also Kerstetter et al. 1970). Such a demonstration cannot be obtained in Scyliorhinus kept in sea water because of the possibility of C1⁻/HCO₃⁻ exchanges in the gill, which will be investigated later.

A final consideration of importance is whether the branchial sodium exchange in Scyliorhinus is effected by an active transport mechanism as in Carassius. Such a hypothesis is unwarranted in the absence of any information on the electrical potential across the gill epithelium. The presence of mitochondria-rich ‘chloride cells’ in the gill of Scyliorhinus (A. Masoni, personal communication) strongly suggests the possibility of an active transport mechanism. Such a transport would, however, occur along the chemical gradient. The participation of an enzyme, carbonic anhydrase, also suggests active transport in relation to Na⁺ uptake but the function of the enzyme may alternatively be related to a mechanism of ‘facilited diffusion’ involving H⁺ and HC0₃⁻ ions.

The present results demonstrate a similarity between the mechanisms of sodium transport in the freshwater teleosts and the marine elasmobranch so far studied. If these observations are confirmed in other species including freshwater elasmobranchs an evolutionary significance for this functional similarity must be sought. We suggest that living elasmobranchs have retained branchial mechanisms inherited from their freshwater ancestors. These mechanisms are of importance in the maintenance of the acid-base balance of the body fluids.

1. Branchial sodium exchanges were measured with the help of ²⁴Na in the marine elasmobranch Scytiorhinus canicula. Handling causes a transient increase of the sodium influx and decrease of the sodium efllux in both intact and hypophysectomized fish.

2. Ammonia-loading (300 μ-equiv./100 g) is followed by an increase of both influx and efflux of sodium resulting in an augmented net sodium uptake lasting for at least 4h. Ammonia excretion is also increased but only for 2 h. Ammonia-loading results in a metabolic acidosis lasting for at least 4 h.

3. HC1 injection (100μ-equiv./100 g) produces an increase of both influx and efflux of sodium resulting in an augmented net sodium uptake lasting for at least 4 h. Ammonia excretion is not affected.

4. Acetazolamide injection (10 mg/100g) results in a depression of the sodium influx, while the sodium efflux remains unchanged. This inhibitory effect is observed in control fish as well as in fish treated with HC1 or ammonium salt injections.

5. These observations confirm that the gill plays a major role in the maintenance of the pH of the body fluids. The similarities between the sodium transport mechanisms of Scyliorhinus and of the freshwater teleosts are emphasized. These results suggest that living elasmobranchs may have retained branchial mechanisms inherited from their freshwater ancestors.

The authors wish to thank Mr R. Tanguy for his technical help in the experiments involving radioactive measurements, Mr P. Lahitette for his help in the ammonia concentration measurements and Dr B. M. Walshe-Maetz for correcting the manuscript.