ABSTRACT
Mechanisms of active NaCl uptake across the posterior gills of the shore crab Carcinus maenas were examined using radiochemical and electrophysiological techniques. In order to measure short-circuit current (Isc), transepithelial conductance (Gte) and area-related unidirectional fluxes of Na+ and Cl−, single split gill lamellae (epithelium plus cuticle) of hyperregulating shore crabs were mounted in a modified Ussing chamber. The negative short-circuit current measured with haemolymph-like NaCl saline on both sides of the epithelium could be inhibited by application of basolateral ouabain (ouabain inhibitor constant KOua=56±10 μmol l−1), 5-nitro-2-(3-phenylpropylamino)-benzoic acid (NPPB; KNPPB=7.5±2.5 μmol l−1) or Cs+ (10 mmol l−1). From the apical side, Isc was nearly completely blocked by Cs+ (10 mmol l−1) or Ba2+ (15 mmol l−1), whereas apical addition of furosemide (1 mmol l−1) resulted in only a small current decrease. Cl− influxes were linearly related to negative Isc.
The ratio between net influxes of Cl− and Na+ was found to be approximately 2:1. With a single membrane preparation, achieved by permeabilizing the basolateral membrane with amphotericin B, Cl− influxes which were driven by a concentration gradient were shown to depend on the presence of apical Na+ and K+. On the basis of these observations, we propose that active and electrogenic absorption of NaCl across the gill epithelium of hyperregulating shore crabs proceeds as in the thick ascending limb of Henle’s loop in the mammalian nephron. Accordingly, branchial NaCl transport is mediated by apical K+ channels in cooperation with apical Na+/K+?2Cl− cotransporters and by the basolateral Na+/K+-ATPase and basolateral Cl− channels.
Introduction
The shore crab Carcinus maenas is a euryhaline species inhabiting coastal environments of the North Sea, western parts of the Baltic Sea and the east coast of North America. The animals tolerate a wide range of environmental salinities between approximately 10 ‰ and full-strength sea water. The body fluids of shore crabs adapted to sea water are iso-osmotic to the ambient medium. When acclimated to brackish water, the crabs hyperregulate their haemolymph osmolarity and counterbalance the ensuing salt losses by active uptake of NaCl across the posterior gills (Péqueux et al. 1988; Lucu, 1990; Towle, 1993).
Studies on isolated perfused gills (Lucu and Siebers, 1987) and on split gill lamellae in an Ussing chamber (Onken and Siebers, 1992) demonstrated an active and electrogenic (inward movement of negative charge) NaCl absorption across the low-resistance gill epithelium of hyperregulating shore crabs which proceeds in a coupled mode. The Na+/K+-ATPase energizes NaCl absorption and represents the basolateral pathway for Na+ uptake (Lucu and Siebers, 1987; Onken and Siebers, 1992). Basolateral Cl− channels allow the translocation of Cl− from the cytosol to the haemolymph (Siebers et al. 1990). Thus, the basolateral transporters involved in transbranchial NaCl absorption seem to be identified.
With respect to transapical NaCl uptake, two different mechanisms have been proposed: parallel Na+/H+ and Cl−/ HCO3− antiports (Lucu and Siebers, 1986; Siebers et al. 1987; Lucu, 1989) or Na+/K+/2Cl− symport in parallel with K+ channels (Onken and Siebers, 1992; Onken et al. 1995). An electrogenic 2Na+/1H+ exchanger has been identified in gill membrane vesicles (Shetlar and Towle, 1989). However, its localization and its involvement in NaCl absorption are still uncertain. Moreover, taking into consideration the polarity of the electrophysiology of the epithelium with identical salines on both sides (outside-positive transepithelial potential difference, PDte, and inward-negative short-circuit current, Isc) as well as the independence of NaCl absorption and Isc from a functioning carbonic anhydrase (Böttcher et al. 1991; Onken and Siebers, 1992), a major contribution of apical antiports to NaCl uptake seems unlikely. The alternative proposal of transapical NaCl absorption via Na+/K+/2Cl− symport, operating in parallel with K+ channels, seems to be more consistent with the electrophysiological observations (Onken and Siebers, 1992; Onken et al. 1995). However, this hypothesis suffers from the lack of any direct experimental support.
Another feature of the Carcinus maenas gill epithelium is its interaction with the well-known diuretic amiloride, which reduces unidirectional Na+ (but not Cl−) fluxes (Lucu and Siebers, 1986), hyperpolarizes the outside-positive PDte (Lucu and Siebers, 1986; Siebers et al. 1987) and increases inward-negative Isc (Onken and Siebers, 1992) when applied to the apical bath. These effects were interpreted as evidence for the presence of either Na+/H+ antiports (Lucu and Siebers, 1986; Siebers et al. 1987) or Na+ channels (Onken and Siebers, 1992) in the apical membrane. Recently, however, a cuticular, amiloride-sensitive Na+ conductance has been reported (Riestenpatt and Siebers, 1995), suggesting that the effects of the diuretic amiloride do not reflect an interaction with the epithelium itself.
In summary, the apical mechanisms involved in electrogenic NaCl absorption across the gills of the shore crab are still unclear. Although it appears that active NaCl absorption is, at least in part, an electrogenic process, it is not known whether electroneutral, transcellular NaCl uptake participates in the overall absorption of NaCl.
In the present investigation, simultaneous measurements of area-related radioactive tracer fluxes and electrophysiological parameters in an Ussing chamber were conducted for the first time for crab gills. Our experiments focused on the mechanisms of active and coupled NaCl absorption across the gill epithelium of hyperregulating shore crabs, using several specific inhibitors of transport proteins. A single membrane preparation, achieved by permeabilization of the basolateral membrane with amphotericin B, was used to determine the K+- and Na+-dependence of the apical Cl− transporter. Our findings support the view of completely electrogenic NaCl absorption via apical Na+/K+/2Cl− symport in parallel with apical K+ exit through K+ channels.
These results were reported in part at the annual meeting of the Deutsche Zoologische Gesellschaft (Riestenpatt and Siebers, 1994).
Material and methods
Crabs
Shore crabs (Carcinus maenas L.) were caught by commercial fishermen in Kiel Bay (Baltic Sea). Before experimental use, the crabs were kept at 16 °C for at least 1 month in dilute sea water (10 ‰ salinity) which was continuously aerated and filtered. The animals were fed three times a week with pieces of bovine heart.
Preparation
After killing the crabs by destroying the ventral ganglion (a needle was pressed through the ventral side of the body wall), the carapace was lifted and the three posterior gills were removed. Single gill lamellae were isolated and split according to the method described by Schwarz and Graszynski (1989). In this way, a single epithelial layer covered by an apical cuticle was obtained. The preparations were mounted in an Ussing chamber modified after De Wolf and Van Driessche (1986), allowing experimentation on epithelial areas of 0.02 or 0.01 cm2. In order to minimize edge damage, silicone grease was used to seal the edges of the preparation. The chamber compartments were continuously perfused with salines at a rate of 0.5 ml min−1 by means of a peristaltic pump.
Salines and chemicals
The haemolymph-like saline was composed of (mmol l−1): 248 NaCl, 5 KCl, 2 NaHCO3, 4 MgCl2, 5 CaCl2, 5 Hepes, 2 glucose. Immediately before use, the pH was adjusted to 7.7 with Tris.
Amphotericin B solution (0.27 mmol l−1 amphotericin B in deionized water) was obtained from Sigma; ouabain was purchased from Fluka; CsCl, BaCl2 and dimethylsulphoxide (DMSO) were obtained from Merck. Furosemide and 5-nitro-2-(3-phenylpropylamino)-benzoic acid (NPPB) were gifts from Farbwerke Hoechst (Frankfurt am Main, Germany). All reagents, with the exception of NPPB (stock solution of 10 mmol l−1 NPPB in DMSO), were dissolved directly in the salines.
Electrical measurements
where Rs,apical is the resistance of the saline in the apical bath and Rs,basolateral is the resistance of the saline in the basolateral bath. The correction of followed Ohm’s law. The transepithelial conductance Gte was calculated as Gte=1/Rte. In the Results section, only the corrected values are given.
Measurement of unidirectional Na+ and Cl− fluxes
where A and B represent the mean values and a and b represent the standard errors of unidirectional influxes and effluxes, respectively.
Statistics
All results represent means ± S.E.M. Differences between groups were tested using the paired Student’s t-test. Significance was assumed for P<0.05.
Results
Using haemolymph-like salines on both sides of the flat sheets, consisting of the cuticle and the underlying single layer of epithelial cells, a short-circuit current (Isc) of −375±14 ΔA cm−2 (N=77) was measured, representing negative charge flow from the apical to the basolateral side of the preparation. The corresponding transepithelial conductance (Gte) amounted to 45±1 mS cm−2. Determinations of unidirectional Cl− fluxes showed a mean Cl− influx (ClJa→b) of 23.7±2.0 Δmol cm−2 h−1 (N=11) and a mean Cl− efflux (ClJb→a) of 5.5±0.6 Δmol cm−2 h−1 (N=10). From these values, a net Cl− influx (ClJnet,a→b) of 18.8±2.1 Δmol cm−2 h−1 was calculated. Furthermore, ClJa→b and the corresponding Isc were linearly related (Fig. 1A), indicating the Cl−-dependence of Isc and the electrogenicity of Cl− uptake mechanisms. For Na+, we measured a mean unidirectional influx (NaJa→b) of 22.0±2.2 Δmol cm−2 h−1 (N=10) and a mean efflux (NaJb→a) of 13.7±1.4 Δmol cm−2 h−1 (N=10). Thus, the mean net Na+ influx (NaJnet,a→b) was calculated to be 8.3±2.6 Δmol cm−2 h−1. The net influx of Na+is markedly smaller than the net influx of Cl−. A comparison of these flux values gives an approximately 1:2 relationship for the net transport of Na+ and Cl− across the gills.
Effect of ouabain
The involvement of the Na+/K+-ATPase in NaCl absorption by Carcinus maenas gill has been demonstrated previously (Lucu and Siebers, 1987; Onken and Siebers, 1992). However, dose-dependent blockage of NaCl uptake by ouabain, a specific inhibitor of the Na+/K+-ATPase (Skou, 1965), has not been reported so far. In five experiments, Isc and Gte were measured when ouabain was added in stepwise increasing concentrations (0.001–10 mmol l−1) to the basolateral superfusion saline. The effect of different concentrations of ouabain on the control current (−398±37 ΔA cm−2) is shown in Fig. 2. A more detailed analysis of blocker saturation kinetics may be obtained by the transformation of the data in a Hanes–Woolf plot (inset to Fig. 2). In this plot, the ratio of the ouabain concentration to the respective Isc decrease ([ouabain]/ΔIsc) is plotted against the ouabain concentration. From the reciprocal slope of the lines obtained for the five individual preparations, we calculated the mean maximal ouabain-induced current decrease ΔImax to be
−371±38 ΔA cm−2, indicating a strict dependence of Isc on the activity of the Na+/K+-ATPase. The average half-maximal effect of the drug (KOua=55.5±10.2 Δmol l−1) was determined from the intercept of the lines with the abscissa. Ouabain also affected the conductance of the gill epithelium. Using high doses of the inhibitor (10 mmol l−1), Gte was significantly reduced by 34±9 % to 32±7 mS cm−2 (N=5) (P<0.05).
Effect of NPPB
As shown previously for several epithelia, NPPB is a potent and specific inhibitor of Cl− channels (Wangemann et al. 1986; Gögelein, 1988). In isolated perfused gills, basolateral Cl− channels have been identified by the use of NPPB and other specific inhibitors of Cl− channels in potentiometric and ion-flux studies (Siebers et al. 1990). Following the addition of NPPB (0.1 mmol l−1) to the basolateral NaCl saline of the short-circuited split lamella preparation, the negative Isc of −213±40 ΔA cm−2 decreased to −32±13 ΔA cm−2 (N=5). Gte remained unchanged. DMSO, which was used as primary solvent for the drug, was without any effect on the electrical parameters when added to the basolateral perfusion medium at a final concentration of 0.5 % (data not shown). The dose-dependence of the inhibition of Isc by NPPB was studied in five experiments (Fig. 3). After transformation of the data into a Hanes–Woolf plot, a straight line was obtained, indicating simple Michaelis–Menten kinetics for the current blockage by NPPB. The half-maximal inhibitor concentration (KNPPB) was 7.5±2.5 Δmol l−1 NPPB in the basolateral bath. The maximum NPPB-induced current decrease ΔImax) was estimated to be −201±43 ΔA cm−2.
Effect of K+ channel inhibitors
In order to identify the electrogenic pathway of the apical membrane of the branchial epithelium, the specific K+ channel inhibitor Ba2+ (Zeiske, 1990) was added to the apical bath. Following application of 15 mmol l−1 BaCl2, the control Isc (−519±90 ΔA cm−2) was reduced by 73±9 % to −122±29 ΔA cm−2 and Gte (48±5 mS cm−2) decreased by 56±3 % to 26±1 mS cm−2 (N=4) (P<0.05).
Using CsCl, another effective K+ channel inhibitor (Zeiske, 1990; Draber and Hansen, 1994), we found an even more effective block of Isc (Fig. 4; Table 1). Addition of 10 mmol l−1 CsCl to the apical bath decreased Isc by 93±1 % and Gte by 62±2 % (P<0.05). The measurement of unidirectional Na+ and Cl− fluxes under these conditions revealed substantially decreased unidirectional influxes of Na+ (NaJa→b) from 23.2±2.4 Δmol cm−2 h−1 to 14.2±1.6 Δmol cm−2 h−1 (P<0.05) and of Cl− (ClJa→b) from 27.1±2.5 Δmol cm−2 h−1 to 5.3±0.5 Δmol cm−2 h−1 (P<0.05). Therefore, the NaCl net influxes (ClJnet,a→b=1.5 Δmol cm−2 h−1 and NaJnet,a→b=−0.7 Δmol cm−2 h−1) were nearly abolished since the unidirectional NaCl effluxes (NaJb→a and ClJb→-a) were not significantly affected by CsCl (Table 1).
When CsCl (10 mmol l−1) was added to the basolateral bath, the negative Isc (−567±132 ΔA cm−2) decreased by 85±2 % to 79±12 ΔA cm−2 (N=3) (P<0.05). The simultaneously measured Gte (48±9 mS cm−2) was reduced by 28±3 % to 34±5 mS cm−2 (P<0.05) (an example is shown in Fig. 4).
Effect of furosemide
The identification of apical K+ channels and their involvement in transepithelial NaCl uptake (see above) suggested a mode of apical NaCl entry via a Na+/K+/2Cl− cotransport as described in the thick ascending limb of Henle’s loop of the mammalian nephron. The diuretic furosemide has been shown to inhibit this Na+/K+/2Cl− cotransporter in various vertebrate tissues (Greger and Kunzelmann, 1990). In order to test the involvement of this enzyme in branchial NaCl absorption, we examined the effect of furosemide on Isc and Gte in split lamellae preparations. Following apical addition of 1 mmol l−1 furosemide, negative Isc (−343±63 ΔA cm−2) was decreased by only 13±1 % to −296±51 ΔA cm−2 (N=5) (P<0.05). The transepithelial conductance (Gte) remained unchanged.
Permeabilization of the basolateral membrane
Since the above experiment did not clearly identify the pathway of apical NaCl entry, we investigated whether the apical Cl− entry depends on the presence of other ions (e.g. K+ and Na+) in the apical bath. With this aim in mind, the basolateral membrane of the gill epithelium was electrically eliminated by the use of a relatively unspecific ionophore. The channel-forming polyene amphotericin B permeabilizes membranes for monovalent ions, water and small polar molecules (Holz and Finkelstein, 1970; Dawson, 1987). Single membrane preparations of epithelia were obtained by treating one side of the epithelium with amphotericin B (Dawson et al. 1990; Schirmanns and Zeiske, 1994). Basolateral addition of 14 Δmol l−1 amphotericin B reduced the Isc by 90±5 % to −39±23 ΔA cm−2 (N=10), indicating the permeabilization of the basolateral membrane for single-charged ions by the incorporation of polyene molecules (Fig. 5). Active NaCl absorption is assumed to have collapsed under these conditions since the depolarization of the cellular potential and the breakdown of the inwardly directed Na+ gradient eliminate the driving force of secondary active transport processes across the apical membrane. Gte was not significantly affected by treatment with amphotericin B. As shown for other epithelia (Onken et al. 1991; Dijkstra et al. 1994), the transcellular conductance is mainly determined by the conductance of the apical membrane. Thus, with respect to the unchanged Gte, it could be anticipated that the polyenic membrane perforation is restricted to the treated membrane. Using a NaCl gradient across this preparation by applying 500 mmol l−1 NaCl + 5 mmol l−1 KCl in the apical bath and 250 mmol l−1 NaCl + 5 mmol l−1 KCl in the basolateral bath, an inward-positive Isc of 1354±226 ΔA cm−2 and a Gte of 141±21 mS cm−2 (N=9) were measured. Following substitution of Na+ by choline on both sides of the preparation, Isc (−16±5 ΔA cm−2; N=5) and Gte (10±1 mS cm−2) decreased drastically. Moreover, the polarity of Isc was changed. These results indicate that the single membrane preparation has a much higher conductance for Na+ than for Cl−. In the presence and in the absence of 5 mmol l−1 KCl and following substitution of Na+ by choline, the unidirectional Cl− influxes were measured in the presence of the Cl− concentration gradient directed from the apical to the basolateral side. The Cl− influx was significantly reduced when Na+ or K+ was absent from the superfusion solutions. The data from these experiments are summarized in Table 2.
Discussion
Methodological aspects
In the past, fluxes of radioactive tracers across crab gill epithelia have only been measured using whole perfused gills (Péqueux et al. 1988; Lucu and Siebers, 1986). Because of the complex morphology of the organs, the fluxes could not be related to the epithelial area and electrophysiological measurements were restricted to the determination of the transepithelial potential difference (PDte), which does not measure transport quantities. Moreover, even with identical salines as perfusing and bathing media, the calculated net flux is not necessarily equal to the rate of active transport because the measurements were conducted under open-circuit conditions. Of course, a PDte influences the unidirectional fluxes used to determine the net flux. Strictly, the most comprehensive conclusions that can be drawn from the results of such experiments are whether or not active transport and/or electrogenic mechanisms are involved. The effects of transport inhibitors on fluxes and PDte may allow conclusions to be drawn about the mechanisms involved in transcellular ion translocation. However, the interpretation of these results is often ambiguous.
For some years, experiments have been performed with split lamellae of crab gills mounted into micro versions of an Ussing chamber (Schwarz and Graszynski, 1989). The validity of this preparation has been demonstrated and discussed extensively (Onken et al. 1991; Onken and Siebers, 1992). So far, split gill lamellae have been used to determine area-related transepithelial short-circuit currents (Isc), reflecting the quantity of active charge transfer across the epithelium, and conductances (Gte). More advanced electrophysiological techniques, such as measurements of membrane potentials using microelectrodes (Onken et al. 1991) or current fluctuation analysis (Zeiske et al. 1992), have also been successfully employed on this preparation. These techniques allowed new insights into crab gill physiology (Onken et al. 1995). However, until the present study, the ionic nature of the currents has not been directly identified using flux measurements but was determined by substitution of the ion in question. Moreover, the restriction to electrophysiological techniques did not allow putative electroneutral ion movements across the epithelium to be investigated.
All these problems can be solved when both influxes and effluxes of radioactive tracers and short-circuit currents across a single split gill lamella, mounted in an Ussing chamber, are measured simultaneously. In the present investigation, simultaneous measurements of influxes or effluxes and Isc have been conducted for the first time on crab gills. Because of the small epithelial area in the Ussing chamber, it was necessary to use relatively high activities of 22Na and 36Cl. Unfortunately, a complete wash-out of the radioactive isotopes took too long for unidirectional influx and efflux, and thus the net flux, to be determined on the same preparation. Therefore, net fluxes had to be calculated from the mean influxes and effluxes determined on different preparations. Nevertheless, the validity of the respective results is shown by a more detailed analysis of the data (see Results and below). For example, when we compute the mean net inward movement of negative charge from the difference between the calculated net influxes of Na+ and Cl−, we obtain a value (10.5±3.3 Δmol cm−2 h−1) which is not significantly different from the mean measured current (13.5±0.7 Δmol cm−2 h−1). As in similar investigations on other epithelial tissues (Ussing and Zerahn, 1951; Diaz and Lorenzo, 1991), the simultaneous measurement of fluxes and currents across crab gill epithelia represents significant progress in methodology which may advance our understanding of their transport characteristics.
NaCl absorption across shore crab gills
Recent ion-flux measurements on isolated, perfused gills of hyperregulating shore crabs have demonstrated their ability to absorb Na+ and Cl− actively (Lucu and Siebers, 1986, 1987; Siebers et al. 1987, 1990). In the present study, the mean unidirectional influxes of Na+ and Cl− with identical salines on both sides of the short-circuited gill epithelium clearly exceeded the mean unidirectional effluxes (see Results). Thus, net influxes were obtained for both ions. Because of the different methodological approaches (Ussing chamber, short-circuit), these findings not only verify the presence of active NaCl absorption across shore crab gills but may also be used to measure active transport of Na+ and Cl− and to relate them to the epithelial area.
Electrogenic versus electroneutral ion movements
In recent investigations using perfused whole gills (Lucu and Siebers, 1986, 1987; Siebers et al. 1987, 1990) and split gill lamellae (Onken and Siebers, 1992), outside-positive transepithelial potential differences (PDte) and inward-negative short-circuit currents (Isc), respectively, have been measured using identical NaCl salines on both sides of the epithelium, indicating that electrogenic mechanisms are involved in NaCl absorption. The results of the present investigation allowed the relationship between the unidirectional NaCl fluxes and the simultaneously measured Isc to be investigated. The effluxes of Na+ and Cl− showed no dependence on Isc (Fig. 1A,B). In contrast, the Cl− influx increased linearly with rising Isc values (see Fig. 1A). It is interesting that the mean Cl− influx at Isc=0 (5.3 Δmol cm−2 h−1), which is obtained from the intercept of the line with the ordinate, was found to be almost identical to the mean Cl− efflux (5.5±0.6 Δmol cm−2 h−1) determined on different preparations. Thus, it seems that net influx of Cl− only occurs concomitantly with an inward-negative Isc, indicating the absence of any active, cellular, electroneutral uptake of Cl−. This strict Cl−-dependence of Isc also excludes any active, electrogenic and Cl−-independent absorption of other ions (e.g. Na+). For this reason, active Na+ absorption across the gill tissue, which is evident from the observation of net influxes of Na+ (Table 1), seems either to be directly coupled to electrogenic Cl− uptake or to proceed via Cl−-independent, electroneutral mechanisms (Na+/H+ antiports). The inhibition of Isc following substitution of Na+ or Cl− in the perfusion salines (Onken and Siebers, 1992) and after blockage of the Na+/K+-ATPase with ouabain (Fig. 2) suggests that Na+ uptake is directly coupled to Cl− absorption. In this case, a correlation between net Na+ influx and negative Isc would be expected. A correlation between unidirectional influxes of Na+ and negative Isc was not found (Fig. 1B). However, effluxes of Na+ across the gill epithelium were found to be much higher than those of Cl−. Moreover, effluxes and influxes varied greatly from preparation to preparation (see Fig. 1B). Therefore, it remains doubtful whether the unidirectional Na+ influxes which represent transepithelial active and passive Na+ movements are an appropriate measure of net Na+ fluxes related to active transport. To determine whether transcellular, electroneutral Na+ absorption occurs in the gill epithelium of Carcinus maenas, different isotopes (22Na, 24Na) could be used in the future to measure influxes and effluxes on the same preparation.
Transporters involved in electrogenic NaCl absorption Basolateral Na+/K+-ATPase
In recent studies with whole gills (Lucu and Siebers, 1987) and with split gill lamellae (Onken and Siebers, 1992), the dependence of NaCl absorption across the gill epithelium on a functioning Na+/K+-ATPase in the basolateral membrane has been demonstrated. The results of the present investigation (see Fig. 2) confirm this finding. Ouabain, a specific blocker of the Na+/K+ pump (Skou, 1965), completely inhibited the negative Isc which is, as discussed above, associated with the entire NaCl absorption. The half-maximal Isc inhibition by basolateral ouabain (KOua, see inset to Fig. 2) agrees with ouabain inhibitor constants (KOua) obtained for gill tissues of other crustaceans such as Eriocheir sinensis (Graszynski and Bigalke, 1986; Schwarz and Graszynski, 1990), several species of fiddler crab (Wanson et al. 1984; D’Orazio and Holliday, 1985; Graszynski and Bigalke, 1986) and Callinectes sapidus (Neufeld et al. 1980). The conductance decrease, observed following ouabain addition, probably reflects a secondary influence on the apical membrane, as in the thick ascending limb of Henle’s loop in the mammalian nephron (Greger et al. 1984).
Basolateral K+ channels
Cs+, which is known to block K+ channels (Zeiske, 1990; Draber and Hansen, 1994), inhibited Isc when applied to the basolateral bath (see Fig. 4). This finding indicates that, as in many other epithelia (Wills and Zweifach, 1987; Dawson et al. 1990) including crab gills (Onken et al. 1991), the basolateral membrane contains K+ channels.
Basolateral Cl− channels
In a variety of NaCl-absorbing epithelia (Gögelein, 1988), including crab gills (Drews and Graszynski, 1987; Onken et al. 1991), Cl− channels have been shown to be the pathway for Cl− to cross the basolateral membrane. Siebers et al. (1990) demonstrated that some blockers of Cl− channels reduced PDte as well as NaCl fluxes across isolated, perfused gills of hyperregulating shore crabs when applied to the basolateral saline, indicating the presence of a Cl− conductance in the basolateral membrane. This finding is confirmed in the present study by the effect of NPPB on Isc (see Fig. 3). As in experiments on whole gills, a KNPPB of about 7 Δmol l−1 was obtained when the dose–response relationship was analysed (see inset to Fig. 3). In the thick ascending limb of Henle’s loop, NPPB half-maximally inhibited the Cl− conductance at a concentration of 8/10−8 mol l−1 (Wangemann et al. 1986). However, as pointed out by Greger (1990), the blocking effect of NPPB and related substances varied considerably from tissue to tissue. The maximal NPPB-induced Isc decrease (ΔImax) is not significantly different from the control current, indicating that the whole transepithelial current flow proceeds via basolateral Cl− channels.
Apical K+ channels
Because of the limited knowledge of the transport mechanisms of the apical membrane, the question concerning which apical transport protein mediates the current flow occurring simultaneously with Cl−absorption was unresolved. Since basolateral Cl− channels have been demonstrated in Carcinus maenas gill epithelium (see above), the presence of these channels in the apical membrane can be excluded for thermodynamic reasons. Another candidate for the role of the apical electrogenic pathway was tested in this study by using two blockers of K+ channels. The effects of apical addition of Cs+ or Ba2+, which are known as potent inhibitors of these channels (see above), on Gte, negative Isc and NaCl influxes (see Results and Table 1) strongly indicate the presence of K+ channels in the apical membrane of shore crab gills and their involvement in active NaCl absorption.
Apical Na+/Cl− symport
Influxes of Na+ and Cl− as well as the outside-positive PDte and inward-negative Isc across the gill epithelium of the shore crab depend on a functioning Na+/K+-ATPase (see above). Moreover, in a previous study, the current was almost completely abolished following substitution of Na+ and Cl− in the perfusion salines (Onken and Siebers, 1992). Finally, Cl− influx across single membrane preparations, achieved by basolateral addition of the ionophore amphotericin B, depended on the presence of apical Na+ (see Table 2). All these observations strongly indicate a coupled mode of transapical NaCl entry. In NaCl-absorbing epithelia, three different mechanisms of coupled NaCl absorption have been described so far: double ion exchange via Na+/H+ and Cl−/HCO3− antiports (Murer and Burckhardt, 1983; Aronson and Seifter, 1983), Na+/Cl− symport (Velasquez, 1987; Cremaschi et al. 1992) and Na+/K+/2Cl− symport (Greger, 1985). Until now, attempts at pharmacological distinction between these mechanisms had not been successful in the gills of Carcinus maenas: SITS (Lucu, 1989), bumetanide (Lucu, 1989) and furosemide (see Results) showed only small effects on ion fluxes, PDte and Isc across shore crab gills. However, it is not yet clear whether apically applied inhibitors can cross the cuticle of crustacean gills and affect apically located transport proteins. In the isolated gill cuticle of several crustaceans, it has been shown that the permeability of the gill cuticle for large ions such as Tris, tetraethylammonium or thiocyanate is rather low. Moreover, blockers of epithelial ion transport such as amiloride or SCN− also modified the transport properties of the cuticle (Lignon and Péqueux, 1990). When the gill cuticle of the shore crab was isolated and mounted in an Ussing-type chamber, it was shown that the conductance of the isolated cuticle was reduced by approximately 95 % following apical addition of amiloride. A half-maximal inhibitor concentration KAmi of 1.2 μmol l−1 was estimated for the amiloride-induced conductance decrease (Riestenpatt and Siebers, 1995). Thus, it seems that this diuretic at least cannot pass the gill cuticle of Carcinus maenas.
So far, the electrogenicity of NaCl absorption and its dependence on apical K+ channels (see Fig. 4 and Table 1) suggests that transapical NaCl uptake proceeds via Na+/K+/2Cl− symport. This interpretation is supported by the finding of an approximately 1:2 relationship (1:2.27) between net Na+ and Cl− influxes under short-circuit conditions and by the K+-dependence of Cl− influxes across single membrane preparations (see Table 2).
Modelling active and electrogenic NaCl absorption across shore crab gills
Summarizing the results of the present study, the following model of NaCl uptake across the gill epithelium of Carcinus maenas is proposed (Fig. 6). Active NaCl absorption across the posterior gills of the shore crab is electrogenic and proceeds in a coupled mode. The Na+/K+-ATPase energizes NaCl uptake by establishing and maintaining Na+ and K+ concentration gradients across the plasma membranes. The Na+ gradient (directed into the cells) drives transapical entry of Na+, K+ and Cl−via Na+/K+/2Cl− symport. The K+ gradient (directed out of the cells), in cooperation with the apical and basolateral K+ conductances (K+ channels), generates a negative cellular potential. This cellular negativity is responsible for driving the basolateral exit of Cl− against the concentration gradient via Cl− channels. Thus, the negative Isc is carried via the apical membrane by K+ (which is recycled via the apical symport) through K+ channels and via the basolateral membrane by Cl− through Cl− channels.
The Na+ efflux under control conditions and the influx of Na+ following inhibition of active NaCl absorption with Cs+ exceed the respective fluxes of Cl− by a factor of about 3 (see Table 1). Furthermore, ion substitution experiments showed that the conductance decrease after Na+ substitution was much higher than that after Cl− substitution (Onken and Siebers, 1992). Thus, the leak pathway of shore crab gills seems to be predominantly permeable to Na+, allowing 1Na+:2Cl− absorption under open-circuit conditions to be balanced. Also, the gill cuticle of Carcinus maenas alone has been shown to be cation-selective, suggesting ion-selective channel-like pores in this cuticle (Lignon and Péqueux, 1990). However, owing to its manifold higher conductance (Lignon, 1987) when compared with that of the split gill lamella (see Results), the contribution of the cuticle to rate-limiting transport phenomena seems to be rather small.
In this transport model (Fig. 6), K+ channels are the sole electrogenic pathway in the apical membrane. Thus, the Cs+-induced conductance decrease (14±1 mS cm−2) reflects the cellular conductance, whereas the remaining conductance of 26±1 mS cm−2 reflects the paracellular pathway (Gp). With a Gp/Gte ratio of 0.65, the gill epithelium of the shore crab belongs to the group of ‘leaky’ epithelia (see Schultz, 1979) which maintain only moderate osmotic gradients.
Interestingly, not only the transport mechanisms but also the characteristics of the paracellular pathway show significant similarities to the epithelium of the thick ascending limb of Henle’s loop in the mammalian nephron (Greger, 1985; Molony et al. 1989).
ACKNOWLEDGEMENTS
This work was supported by the Deutsche Forschungsgemeinschaft.