Salinity Acclimation and Intestinal Salt Transport in the Flounder: the Role of the Basolateral Cell Membrane

Most euryhaline teleosts closely regulate the osmotic pressure of their body fluids to about 350 mosmol l⁻¹, whilst that of their environment can fluctuate between near Omosmoll⁻¹ in fresh water (FW) and 1200 mosmol l⁻¹ in sea water (SW) (e.g. Evans, 1979). Our understanding of the overall strategy enabling fish to survive this complete reversal of osmotic challenge has developed from the classical studies of Smith (1930), Keys (1933) and Krogh (1939): thus in SW, euryhaline teleosts replenish branchial water loss by absorbing NaCl and water through their intestines and excreting excess salt across their gills; in FW, the gills absorb NaCl whilst the kidney and bladder serve to produce copious dilute urine thereby conserving salt and eliminating excess water (reviewed in Ellory & Gibson, 1984).

The central role proposed for the intestine in teleostean osmoregulation has made it the subject of several recent studies (e.g. Olde & Utida, 1967; Lau, 1985), and there is general agreement from isolated sacs, voltage-clamp and in vivo experiments that intestinal Na⁺ and CE absorption is higher in tissues from SW-acclimated fish than FW-acclimated ones (see Discussion). However, the magnitude and significance of these changes vary and, further, the cellular mechanisms underlying the acclimatory response have received less attention.

The work described in this paper represents an attempt to identify the site(s) for acclimatory change in intestinal transport. From conventional models of leaky epithelia, such as teleost intestine (Field et al. 1978), these would include the ionic conductances of apical, basolateral and junctional membranes and the basolateral Na⁺,K⁺-ATPase activity (shown schematically in Fig. 1). The European flounder, Platichthys flesus, was chosen as an experimental model and we have measured bidirectional Na⁺ and Cl⁻ fluxes under short-circuit conditions in the intestines of SW and FW fish. Application of amphotericin B to the mucosal surface was used to increase the univalent ion permeability of the apical cell membrane, thus removing its limiting function. We could then assess its original role in NaCl transport in this epithelium. This approach has been established for a variety of epithelia (Bentley, 1968; Nielsen, 1971 ; Frizzel & Turnheim, 1978; Reuss, 1978; Wills, Lewis & Eaton, 1979) and has been used by us to study the relative contributions of apical and basolateral cell membranes to transepithelial transport in other acclimatory responses (Gibson & Ellory, 1984; Gibson, 1985; Gibson, Ellory & Cossins, 1985).

Our results indicate that the apical cell membrane does not represent a regulatory site, but rather that the principal acclimatory change in the tissues of SW-acclimated fish is located in the basolateral cell membrane mediated by an increased Na⁺-pumping capacity and Cl⁻exit step.

European flounders, Platichthys flesus, weighing 200—500g were obtained from the Fisheries Research Laboratory, Lowestoft, Suffolk; The Marine Biological Association Laboratories, Plymouth; and L’Institut des Pêches (IFREMER), Nantes. After transit, the fish were left for 1 week in sea water (SW) to recover, and then divided into two groups. One group was left in recirculating aerated SW at 11–15 °C, while the other was transferred in two stages (50 % SW, 20 % SW) to fresh water (FW), at the same temperature. Fish were kept unfed in FW or SW for at least 2 weeks before use.

The entire intestine was removed from fish immediately after they had been killed, and sections of anterior intestine were washed with saline (see below), opened longitudinally and pinned mucosal surface downwards in a soft plastic dish to allow stripping of the serosal muscle layers. Pieces of stripped intestine were then mounted in conventional Ussing chambers (Ramos & Ellory, 1981) 0·7cm² in area, under voltage-clamp conditions at room temperature (21–24°C).

The saline had the following composition (in mmol l⁻¹): NaCl, 130; CH₃COOK, 5; MgSO₄, 1·1; CaCl₂, 2·5; NaHCO₃, 25; D-glucose, 10; L-alanine, 2; gassed with 95 % O₂/5 % CO₂ to pH 7·2 at 22°C.

To measure bidirectional Na⁺ and Cl⁻ fluxes, the following radioisotopes were used: ²⁴Na⁺ was obtained from Amersham International, Amersham, Bucks, or The Northern Universities Research Reactor, Warrington (see Gibson et al. 1985, for details), or from Commissariat à l’Energie Atomique, Saclay, France; ²²Na^{+, 36}C1⁻ came from Amersham International; ⁷⁷Br⁻ was obtained as a carrier-free solution from The Medical Research Council Cyclotron Unit, Hammersmith Hospital, London.

For simultaneous bidirectional Na⁺ flux.measurements, ²⁴Na⁺ at 40–80 KBq ml⁻¹ was added to the mucosal solution to measure J_ms, whilst ²²Na⁺ at 4–10 KBqml⁻¹ was added to the serosal solution to measure J_sm. In similar parallel experiments, ⁷⁷Br⁻ at 20 KBq ml⁻¹ was used to measure ⁠, and ³⁶C1⁻ at 4–8 KBq ml⁻¹ to measure ⁠. Preliminary experiments (Gibson, 1985) established the validity of ⁷⁷Br⁻ as a tracer for Cl⁻ fluxes in flounder intestine.

Tissues were stable electrically for the 2-h experimental period, which followed a 1-h tracer preloading period. Typically, four 15-min control periods were followed by four similar periods in the presence of amphotericin B (nominally 40μgml⁻¹: Fungizone, E. R. Squibb & Sons Inc., Princeton, NJ). Samples were processed for Cerenkov (²⁴Na⁺), y-scintillation (⁷⁷Br⁻) and liquid scintillation (Picofluor-30) counting (²²Na^{4−, 36}C1⁻), using quench correction routines for the effects of amphotericin B, and separation by isotope decay for ²²Na⁺/²⁴Na⁺; ⁷⁷Br⁻/³⁶Cl⁻ resolution as previously described (Ramos & Ellory, 1981; Gibson et al. 1985). Because of the time course of the effect of amphotericin on short-circuit current (Fig. 3), data in Tables 3 and 4 and Fig. 2 are calculated for the final 30min after application of the drug.

PD, transepithelial potential difference, expressed in mV, with the polarity of the serosal side relative to that of the mucosa.

SCC, short-circuit current, passing from mucosa to serosa taken as positive, expressed in μA cm⁻².

J_ms, mucosa-to-serosa flux.

J_sm, serosa-to-mucosa flux.

J_net, net flux (absorption positive) = J_ms – J_sm.

Superscripts denote chloride or sodium ions. All fluxes expressed in μmol cm⁻²h⁻¹.

Results are expressed as mean ± S.E.M. (N), whereN is the number of tissues; two pieces of intestine were typically taken from each fish. Comparisons were made using unpaired Student’s t-test.

Tissues from both SW- and FW-acclimated fish gave a stable negative transepithelial PD and SCC (Table 1) when measured 30 min after mounting. As noted previously by Smith, Ellory & Lahlou (1975), intestines from FW-acclimated fish, initially showed positive values on mounting before stabilizing at the control values after 10–15 min. The magnitude of SCC and PD of SW-acclimated tissues was about twice that of FW-acclimated ones, whilst transepithelial resistance was similar in the two groups. The absolute values of PD and SCC for SW-and FW-acclimated fish compared well with previous values of – 1·9±0·14mV and – 45 ±5·1 μA for SW-acclimated fish and –1·24 ± 0·14mV and –18·2 ± 3·6μA for FW-acclimated fish (Smith et al. 1975).

The results for bidirectional chloride fluxes are given in Fig. 2 and Table 2 and show that net Cl⁻ absorption occurs in both FW-and SW-acclimated tissues, although both the net and unidirectional Cl⁻ fluxes are about 40 % greater in the SW-acclimated tissues.

Similarly, for the Na⁺ fluxes (Fig. 2; Table 2), of the SW-acclimated fish is greater than that of FW-acclimated ones. However, unidirectional Na⁺ fluxes are about the same in both groups, underlining the necessity of measuring bidirectional fluxes.

A comparison of the Na⁺ and Cl⁻ data reveals that the net Cl⁻ fluxes are higher than the net Na⁺ fluxes, and that the sum of and is not significantly different from the SCC expressed in μequivcm⁻²h⁻¹ (P<0·1) (Table 1). Further, unidirectional Na⁺ fluxes, (particularly > ) and hence the overall Na⁺ permeability, exceed those of Cl⁻.

These values are similar to previous published data for Na⁺ fluxes on Platichthys flesus (Smith et al. 1975) and Cl⁻ and Na⁺ fluxes in other seawater teleosts (e.g. plaice, Ramos & Ellory, 1981).

Application of amphotericin B (40μgml⁻¹) to the mucosal solution reduced the negative PD and SCC, with a half-time of about 7·5 and 15 min in SW-and FW-acclimated fish, respectively (Fig. 3). In SW-acclimated tissues these parameters became positive whilst in the FW-acclimated case they declined to near zero (Table 3). Transepithelial resistance was not affected by application of amphotericin B.

For every tissue tested in both groups of fish, the net Cl” fluxes were significantly reduced (by about 30%; P<0·01) following amphotericin B treatment, whilst the unidirectional Cl⁻ fluxes increased in all tissues. The values of and remained greater in the tissues from SW-compared with FW-acclimated fish (Table 4; Fig. 2), whilst the difference in the values of between FW-and SW-acclimated groups became insignificant, after amphotericin B treatment.

Amphotericin B had a large stimulatory effect on the net Na⁺ flux measured in SW-acclimated tissues, almost doubling it (Fig. 2) and thus revealing a large reserve Na⁺ pumping capacity.

For FW-acclimated tissues, the net Na⁺ flux was again stimulated by amphotericin B but to a lesser extent, so that the difference between SW-and FW-acclimated fish was enhanced to about three times the control difference.

After amphotericin B, SCC (in μequivcm⁻²h⁻¹) is +0·28 ± 0·08 and +0·09 ± 0·06 for SW-and FW-acclimated fish, respectively, much less than the sum of and which are + 1·53 ± 0·50 and +0·54 ± 0·23 μequiv cm⁻² h⁻¹, respectively, (P for difference <0·001) in SW-and FW-acclimated tissues. This indicates that net secretion of unknown cations or absorption of anions must have occurred.

In summary, the flux data show that removal of the selective permeability barrier of the apical cell membrane with amphotericin B enhanced the difference in Na⁺ absorption, but reduced that of Cl⁻ absorption, between the two fish populations.

The classical paradigm for salt adaptation in teleost fishes involves enhanced intestinal Na⁺ and Cl⁻ transport in SW-adapted animals. Although widely accepted, this theory is based on limited data which often reflect rather small changes in transport which do not appear to be significant statistically. In the present work we confirm that a modest (50%) increase in Na⁺ and Cl⁻ transport in SW-adapted flounders occurs, which is nevertheless highly significant. Because of the relatively large unidirectional vs net fluxes, it is essential to use bidirectional tracer fluxes to determine net transport in these experiments. Additionally, the unidirectional Na⁺ fluxes are always greater than those of Cl⁻, in accordance with the higher Na⁺ conductance of the paracellular pathway observed in many leaky epithelia (e.g. Barry, Diamond & Wright, 1971 ; Field et al. 1978).

The flounder, in common with other SW teleosts (e.g. Evans, 1979; Ramos & Ellory, 1981, in plaice), has a negative PD and SCC and the values for SW-acclimated fish are consistently higher than those for FW-acclimated ones in this and other studies (Smith et al. 1975; Mackay & Lahlou, 1980). However, it is not possible to relate SCC directly to net ion fluxes in leaky epithelia or when coupled ion transport is significant (Schultz, 1974). Nevertheless, the net transepithelial movement of charged ions must still be equal in magnitude, though opposite in polarity, to the injected current, expressed in μequiv cm⁻² h⁻¹ (Ussing & Zerahn, 1951). From Tables 1 and 2, is equal to Jscc within the limits of error (P>0·l) in control SW-and FW-acclimated tissues, and thus it is unnecessary to postulate further net transepithelial ion movements, although coupled neutral fluxes may occur. One possible consequence of this equivalence is that most K⁺ pumped into the cell to account for the Na⁺ absorption recycles through the basolateral cell membrane, as postulated in Fig. 1 (Ellory, Gibson & Lau, 1984).

Finally, as noted above, the transepithelial resistance is the same in both SW-and FW-acclimated tissues and, together with the similarity of in the two cases, this suggests that the junctional permeability is not a site of acclimatory change.

Amphotericin B applied to the mucosal surface of epithelia renders the apical cell membrane highly permeable to many univalent ions. The electrochemical potentials of Na⁺ (Graf & Giebisch, 1979; Reuss, 1981) and Cl⁻ (Russell, Eaton & Brodwick, 1977; Ellory et al. 1984) become equal to those in the bathing solution within about 10 min and any net transepithelial movement of these ions under such circumstances must be mediated by the basolateral cell membrane.

After application of amphotericin B, the residual Cl⁻ flux is no longer significantly higher in SW-than FW-acclimated tissues (Fig. 2). Compared to control conditions, is reduced in both groups of fish by a similar percentage. The transport mechanism responsible for the residual net Cl⁻ flux is still obscure but may well involve K⁺, pumped into the cell by the Na⁺,K⁺-ATPase, recycling across the basolateral cell membrane coupled to Cl⁻ ions (Fig. 1). The evidence for this includes the fact that it is abolished by serosal addition of ouabain or bumetanide (Ellory et al. 1984). Recently, however, Hoffman has suggested a more direct role for the sodium pump in anion-coupled Na⁺ fluxes by the red blood cell (Hoffman et al. 1985).

The unidirectional Cl⁻ fluxes all increase significantly, as expected if the intact apical cell membrane imposes some limitation on transepithelial Cl⁻ movement. However, ⁠, of SW-acclimated tissues still exceeds that of FW-acclimated ones, even following treatment with amphotericin B. This may indicate that the basolateral cell membrane Cl⁻ permeability is elevated in salt water, a factor which would account for part of the observed acclimatory response under control conditions.

In contrast to net Cl⁻ fluxes, net Na⁺ fluxes are considerably enhanced in both SW-and FW-acclimated fish by the mucosal addition of amphotericin B. Both the magnitude of the increase and the final absolute value of the net Na⁺ flux are greater in SW-than FW-acclimated tissues. Na⁺ transport across the basolateral cell membrane is mediated by the ouabain-sensitive Na⁺,K⁺-ATPase, and hence if similar junctional permeabilities apply in FW-and SW-acclimated tissues the activity of this enzyme must be increased in the latter.

The action of amphotericin B demonstrates a reserve pumping capacity in SW-and FW-acclimated tissues, which is normally limited by the Na⁺ permeability of the apical cell membrane. The amphotericin-B-induced stimulation will be via an increased cell Na⁺ concentration (as seen in other systems such as Necturus gallbladder, Graf & Giebisch, 1979; and plaice intestine, Ellory, Lahlou & Ramos, 1981).

The minimal effect of amphotericin on unidirectional Na⁺ fluxes might be thought inconsistent with the action of this ionophore. However, junctional Na⁺ permeability is very high, especially relative to Cl⁻, and Cl⁻ fluxes increased by only about 30% with amphotericin B. A much higher junctional Na⁺ permeability, observed in many leaky epithelia (Barry et al. 1971; Field et al. 1978), may well reduce the expected transepithelial Na⁺ flux increment to within the error of the present measurements. The unchanged transepithelial resistance after amphotericin B has made permeable one cell membrane indicates the high conductance of the intercellular pathway.

The conclusion drawn from these experiments with amphotericin B is that the principal site of acclimation accounting for the increase in intestinal Na⁺ and Cl⁻ absorption resides in the basolateral cell membrane and changes in the apical cell membrane and the junctions are unnecessary.

A number of other observations in the literature militate further against the apical cell membrane as a site of regulation, and point towards a primary role for changes at the basolateral cell membrane. Thus the electrochemical driving force for Cl⁻ across the basolateral cell membrane is the same in SW-and FW-acclimated tissues (Lau, 1982), whilst net Cl⁻ absorption is greater in the former; and second, Cl⁻ uptake into the intestinal cells across the apical cell membrane is similar in SW-and FW-acclimated fish (Smith et al. 1975; N. A. A. Macfarlane & B. Lahlou, cited in Lahlou, 1975). [The observation that mucosal application of piretanide results in the loss of active Cl⁻ accumulation in SW-acclimated tissues only (Lau, 1982, 1985) could follow from a reduction in the rate of Cl⁻ exit across the basolateral cell membrane in FW-acclimated tissues, rather than from a difference in apical cell membrane properties.]

The most economical scheme which could account for these observations and the present results is that the apical cell membrane is not limiting to Cl⁻ transport. Thus, intracellular Cl⁻ activity reflects the Na⁺ gradient into the cell and the membrane potential (equal in SW-and FW-acclimated fish), whilst transepithelial Cl⁻ transport can be regulated at the level of the basolateral cell membrane without affecting the level of Cl⁻ accumulation within the cell.

We have argued so far that acclimation is accompanied mainly by changes in the basolateral cell membrane. There are, however, at least two possible sites which could be involved in acclimatory changes at this membrane, the Na⁺,K⁺-ATPase and the Cl⁻ exit step (see Fig. 1).

An increase in Na⁺,K⁺-ATPase activity in SW-acclimated tissues is suggested by the present data for net Na⁺ transport after amphotericin B (Fig. 2) and from measurements of the enzyme activity in tissue homogenates (Olde, 1967; Jampol & Epstein, 1970; Pickford et al. 1970), although the latter may not reflect the in vivo Na⁺ transporting capacity because of methodological artefacts, e.g. homogenization may expose pumps not in the original transporting pool, and the expression of activity in terms of turnover per unit mass of protein may obscure changes if the total amount of the latter per cell does not remain constant.

Also, the data of Table 4 and the experiment of Lau (1985) described above suggest a change in the Cl⁻ exit step across the basolateral cell membrane. After amphotericin B, remains greater in SW-than in FW-acclimated tissues and, as discussed above, this difference is unlikely to be across the junctions.

In conclusion, the present simultaneous bidirectional flux studies for Na⁺ and Cl⁻ transport confirm the classical acclimatory response of teleostean intestine to altered environmental salinities. The use of amphotericin B further helps to localize the acclimatory processes to the basolateral cell membrane. Our data indicate that changes in the activity of the Cl⁻ exit step at the basolateral cell membrane may occur on salinity acclimation, and this property deserves further study. An obvious next step is to relate the cellular effector mechanisms to the endocrine changes associated with salinity acclimation.

JSG thanks the MRC, Magdalene College, Cambridge and the Gordon Wigan Fund for financial assistance. BL was supported by CNRS grant UA 651. During this work JCE held the title of Professor Associé in the University of Nice.