Hypotonicity induced K⁺ and anion conductive pathways activation in eel intestinal epithelium

Epithelia are physiologically exposed to osmotic stress, resulting in alteration of cell volume in several aspects of their functioning. Epithelial transport, accomplished by the entry or extrusion of osmotically active substances at the two cellular membranes, represents a continuous challenge to cell volume constancy, since slight changes in the large apical or basolateral fluxes could lead to rapid changes in cell volume. For example, in intestine,gallbladder and renal proximal tubules the luminal uptake of substrates by Na⁺-coupled transport tends to swell the cells, leading to volume regulatory activation of K⁺ channels in the basolateral cell membrane (Beck et al., 1991, 1994; Breton et al., 1996; Cemerikic and Sackin, 1993; Furlong and Spring, 1990; Harvey, 1994; Lang et al., 1986; Lau et al., 1984; Reuss and Cotton, 1994; Schultz, 1994; Turnheim, 1994). Moreover,epithelia, being interfaces between the internal and the external environment of the organism, can experience changes of extracellular osmolarity. For example, kidney medulla epithelial cells face alterations of extracellular osmolarity during transition between antidiuresis, when the osmolarity can reach very high levels, and diuresis (Beck et al., 1988), or intestinal epithelial cells experience hypotonic stress following excess intake of water.

For epithelial cells, like all the other cell types, alteration in cell volume ultimately results in membrane damage and loss of the cellular structural integrity; therefore, activation of `emergency' systems of rapid cell volume regulation is fundamental in their physiology. The mechanisms for the control of cell volume after osmotic stress are highly conserved and in principle similar in cells from various tissues as well as between evolutionary distant species (Gilles,1988; Chamberlin and Strange,1989; Lang et al.,1998). Following a hypotonic cell swelling, they comprise the release of osmolytes followed by loss of osmotically obliged water, termed Regulatory Volume Decrease (RVD; Hoffmann and Dunham, 1995). A widely established strategy of electrolyte transport regulation in RVD response in epithelial and non-epithelial cells is the activation of K⁺ and Cl^- efflux through independent K⁺ and anion channels (Hoffmann 1978; Busch and Maylie,1993; Deutsch and Chen,1993; Felipe et al.,1993; Hoffmann and Dunham,1995; Fürst et al.,2000; Hoffmann,2000; Niemeyer et al.,2000; Nilius et al.,2000; Strange et al.,1996; Okada, 1997; Valverde et al., 2000). Electroneutral K⁺-Cl^- cotransporter is an alternative system contributing to RVD in some cell types(Lauf and Adragna, 2000).

In the present study the physiological response to hypotonic stress was investigated in a salt-transporting epithelium, the intestine of the euryhaline teleost Anguilla anguilla. This tissue plays a key role in the osmoregulation of the eel, especially in seawater, where the intestinal epithelium carries out active salt absorption necessary in turn to replenish passively lost water from the body, while excess salt is secreted by the gills(Smith, 1930). The intestinal epithelium is a useful model system for functional studies of epithelia that perform near-isosmotic fluid absorption. The mechanisms accounting for ion and water transport in the eel middle intestine have been characterized and considerable information is also available on regulation of the transport (for a review, see Schettino and Lionetto,2003). Briefly, it develops a transepithelial Cl^-absorption, measurable as transepithelial potential and short circuit current,sustained by the operation of the luminal Na⁺-K⁺-2Cl^- cotransporter, in series with a basolateral Cl^- conductance and in parallel with an apical K⁺ conductance. The Na-K-ATPase on the basolateral membrane, by generating an inwardly directed electrochemical gradient for Na⁺,provides the driving force for the Na⁺-K⁺-2Cl^- cotransporter and thus for the active intracellular accumulation of Cl^- (Trischitta et al., 1992a,b). This tissue is physiologically exposed to anisosmotic conditions, particularly during the transfer of the animal from freshwater to seawater and vice versa. As previously demonstrated (Lionetto et al., 2001, 2002) it is sensitive to the osmolarity of the extracellular medium, making this tissue a good physiological model for epithelial cell volume regulation research.

To our knowledge this is one of the few works in which the ionic transport mechanisms that sustain the physiological response to hypotonic stress in the intestinal epithelium are examined not on cultured or isolated cells, but on a native tissue, and studied through an integrated analysis of the ion transport involved on the two plasma membranes.

Seawater-acclimated yellow eels Anguilla anguilla L. (weighing about 200 g) were purchased from a commercial source (Agroittica,Lesina-Foggia, Italy). They were kept unfed in seawater aquaria under conditions of controlled photoperiod cycles and water temperature(18-20°C) for at least a week before killing. The animals were anaesthetized with 2-phemoxyetanol (0.3 ml dm^-3) prior to sampling. All experiments were carried out in accordance with the European Committee Council Directive (86/609/EEC).

All chemicals were reagent grade. Cacodylate buffer, glutaraldehyde,OsO₄ and Epon 812 were purchased from Flukachemie GmbH (Buchs,Switzerland); all the other chemicals were purchased from Sigma (St Louis, MI,USA). Stock solutions of BAPTA-AM (50 mmol l^-1 in DMSO),thapsigargin (1 mmol l^-1 in DMSO), trifluoroperazione (10 mmol l^-1 in distilled water), bumetanide (10 mmol l^-1 in DMSO), apamin (100 μmol l^-1 in distilled water) and iberiotoxin(100 μmol l^-1 in distilled water) were prepared and kept at-20°C until use. Unless otherwise noted, solutions were freshly prepared.

The middle intestine of seawater-acclimated eels was removed, stripped of longitudinal and circular layers using two pairs of fine forceps and mounted vertically in a modified Ussing chamber (CHM6, World Precision Instruments,Berlin, Germany) (membrane area: 0.6 cm²), where it was perfused on both sides by isotonic teleost Ringer solution (NaCl 133 mmol l^-1,KCl 3.2 mmol l^-1, NaHCO₃ 20 mmol l^-1,MgCl₂ 1.4 mmol l^-1, CaCl₂ 2.5 mmol l^-1, KH₂PO₄ 0.8 mmol l^-1, glucose 20 mmol l^-1; osmolarity: 315 mOsm kg^-1).

The preparations were kept `open circuited' through the time course of the experiments, except for a few seconds every 5 min for recording the short-circuit current (I_sc). Tissues were connected to an automatic short-circuit current device (DVC-1000, World Precision Instruments)by four Ag/AgCl electrodes (two voltage electrodes and two current electrodes)that made contact with the bathing solutions via agar-Ringer filled cartridges. Transepithelial voltage (V_te) was measured with respect to the mucosal bath (grounded); I_sc was measured by passage of sufficient current through Ag/AgCl electrodes to reduce the spontaneous V_te to zero automatically (resistance of the chamber fluid was subtracted automatically). The I_scis referred as negative when current flows across the tissue from the apical membrane to the basolateral membrane. Transepithelial resistance(R_te) was measured by pulsed current injection (33 μA cm^-2, 500 ms) through the tissue. This injected current produces a voltage deflection (ΔV_te) from which R_te was calculated.

In the hypotonic stress experiments the osmolarity of the teleost Ringer solution was bilaterally decreased to 175 mOsm in two steps: first the concentration of NaCl was reduced to 66.5 mmol l^-1 at unaltered osmolarity by mannitol replacement, then the hypotonic stress was applied by removing mannitol. This experimental maneuver was necessary to avoid the superimposing of diffusion potential in the unstirred layers, due to the NaCl reduction, on the hypotonic stress effect.

Drug treatments were performed by 1 h incubation of the tissue with the drug before removing mannitol. The same time protocol was respected in the corresponding control sample. In the drug test the drug was still present during the hypotonic stress application.

In hypotonic stress experiments Ringer osmolarity was decreased only once in each experiment. The hypotonic stress response was quantified asΔ V_te and ΔI_sc. Δvalues were calculated as the differences between the V_teand I_sc values after hypotonic stress application and the value before every 5 min during the time course of the exposure.

In all Ussing chamber experiments the temperature of the perfusing Ringer solution was kept constant at 18°C.

Small segments of eel middle intestine, stripped from their outer layers,mounted in the Ussing chamber for electrophysiological measurements and perfused with Ringer solution (315 mOsm), were fixed before (control), after 5 min and after 45 min exposure to hyposmotic Ringer solution (160 mOsm),respectively. Solutions utilized for fixation and postfixation were respectively: (a) cacodylate buffer 0.1 mol l^-1, pH 7.9, 315 mOsm with mannitol; (b) cacodylate buffer 0.1 mol l^-1, pH 7.9; (c) 2.5%glutaraldehyde in cacodylate buffer a; (d) 2.5% glutaraldehyde in cacodylate buffer b; (e) 1% OsO₄ in cacodylate buffer a; (f) 1%OsO₄ in cacodylate buffer b. Fixation was performed as follows: the epithelium segments were incubated for 36 h at 4°C in buffer c (control)or buffer d (hypotonicity exposed segments). Thereafter, three washings (15 min each) in cacodylate buffer a (control) or b (hypotonicity exposed tissues)were performed. Intestinal segments were then postfixed in 1% OsO₄in cacodylate buffer e (control) or f (hypotonicity exposed segments) and embedded in Epon 812. Semithin (0.5 μm) sections were cut along planes perpendicular to the luminal epithelium surface and stained with 1% Toluidine Blue. Sections were placed on an optic microscope (Axiolab, Zeiss, Oberkochen,Germany; objective utilized: 100× oil immersion), and the images obtained from a video camera (Polaroid Digital Microscope Camera DMC 1, CCD;1600×1200 pixels) were digitalized using NIH Image 1.62; each image was calibrated with respect to overall magnification and in each specimen epithelium thickness was measured in 80 points, where the plane of section appeared to be exactly perpendicular to the basal membrane.

European eels fully adapted to seawater were transferred to freshwater. Plasma samples were obtained immediately before the transfer and at regular intervals thereafter up to 163 h. Plasma was obtained from blood collected from the caudal vein of the eel by a heparinized syringe and immediately centrifuged at 9000 g for 10 min (Beckman Microfuge® Lite Centrifuge, Fullerton, CA, USA). The osmolarity of plasma samples was measured by vapor pressure osmometer (5520, Delcon, Milano, Italy), also utilized for measuring osmolarity of experimental solutions.

Values are given as means ± s.e.m. Statistical tests utilized to evaluate statistical significance of differences were paired Student' t-test, one-way ANOVA test and Keuls-Newman multiple comparison test, as indicated in legends to figures. ^*P<0.05; ^**P<0.01.

Eels transferred directly from seawater to freshwater showed a significant reduction in plasma osmolarity by 20 h(Fig. 1), reaching its maximum value (18.6%, corresponding to a decrease of approximately 60 mOsmol kg^-1) in 90 h, then followed by a partial recovery.

This finding clearly indicates that in the eel, cells are physiologically exposed to hypotonic stress when the animal faces decreases in environmental osmolarity.

The effect of hypotonic stress on the volume of enterocytes in the native tissue was studied by morphometrical analysis of the epithelium height as described in Materials and methods. When the eel intestinal epithelium was exposed to hypotonic stress (performed by decreasing Ringer osmolarity from 315 to 175 mOsm) the thickness of the epithelium showed a significant(P<0.01) 18% increase after 5 min exposure (Figs 2A,B, 3), but decreased significantly(P<0.01) and returned back to the initial value during the next 30 min(Figs 2A,C, 3). This finding suggests that the eel intestinal epithelium exposed to hypotonic stress initially swells because of the osmotic intake of water, but is able to regulate its volume in a RVD response, even though the osmotic stress still persists.

In all the experiments the hypotonic stress applied was a 45% percentage decrease in osmolarity of the perfusion solution; this represents a commonly used hypotonic stress in volume regulation protocols(Hoffmann, 1978; Hazama and Okada, 1990; Diener et al., 1992).

In order to investigate the nature of the observed physiological response to hypotonicity, functional studies were performed on the Ussing chamber-mounted epithelium by transepithelial electrophysiological measurements (transepithelial potential, short circuit current and transepithelial resistance).

The exposure of seawater eel intestine to an hypotonic teleost Ringer produced a biphasic decrease in V_te and I_sc, shown in Fig. 4: a first rapid transient phase, which lasted for about 5-10 min,followed by a more gradual and sustained one. There were also significant changes in R_te (Fig. 5), which rapidly increased soon after the hypotonic stress application, but tended to slowly decrease towards the initial value during the RVD phase.

With the aim of investigating the ionic nature of the V_te and I_sc responses to hypotonic stress, we first tested if the observed biphasic decreases in V_te and I_sc could be due to a reduction in the rate of transepithelial Cl^- absorption via the luminal Na⁺-K⁺-2Cl⁺cotransporter. To test this, 10 μmol l^-1 bumetanide (specific inhibitor of the Na⁺-K⁺-2Cl^-) was added to the luminal Ringer solution before decreasing external osmolarity(Fig. 6A). Addition of bumetanide in itself decreased both V_te and I_sc but did not have any significant effect on the response to hypotonicity (neither V_te nor I_sc, not shown), suggesting that other membrane ion transport mechanisms must be involved in the observed phenomenon.

By contrast, addition of the Na⁺-K⁺-ATPase inhibitor ouabain (1 mmol l^-1) to the serosal bathing solution(Fig. 6B), by dissipating K⁺ and Na⁺ gradients maintained by the pump operation(Lew et al., 1979),significantly decreased the hypotonicity induced V_te and I_sc response in both the first(Fig. 6B) and the second phase(data not shown). Since the decreases in V_te and I_sc are not secondary to a decrease in cotransport but are dependent on an intact ionic gradient, we investigated whether the observed changes are a result of a swelling activated K⁺ conductance,K⁺ being the most abundant intracellular cation in eel enterocyte(Marvão, 1994).

In basal conditions the main cation electrodiffusive pathway is a Ba²⁺-sensitive K⁺ conductance on the apical membrane that permits recycling of K⁺ into the lumen and contributes to the operation of the luminal Na⁺-K⁺-2Cl^-transporter system (Trischitta et al.,1992a; Marvão et al.,1994). Marvão et al.(1994), using conventional and selective microelectrodes, also found a Ba²⁺-inhibitable K⁺ conductance on the basolateral membrane.

In our experiments application of 2 mmol l^-1 Ba²⁺either on the luminal or the basolateral side of the epithelium had no effect on the V_te and I_sc (not shown)responses to hypotonicity (Fig. 7A,B), suggesting that activation of other swelling-dependent ion conductances, different from the basal ones, could be responsible for the hypotonicity induced V_te and I_scchanges. Thus, the eventual involvement of other K⁺ channels, such as Ca²⁺-activated K⁺ channels, important in cell volume regulation in other cell types(Fernandez-Fernandez et al.,2002; Roman et al.,2002) were tested by utilizing iberiotoxin, a specific blocker of the high conductance Ca²⁺-activated K⁺ channels (BK channels; Galvez et al.,1990), and apamin, specific blocker of small conductance Ca²⁺-activated K⁺ channels (SK2 and SK3; Blatz and Magleby, 1986; Bond et al., 1998).

In isotonic conditions, 0.1 μmol l^-1 iberiotoxin (Ibtx) had no significant effect on the basal electrophysiological parameters V_te and I_sc either on the serosal or mucosal membranes (Fig. 8A,B),but in hypotonic conditions, when applied on the serosal side, it decreased the hypotonicity induced depolarization of V_te in both the first and the second phases (Fig. 8C); when it was applied on the mucosal side it increased the hypotonicity induced depolarization of V_te in the second phase (Fig. 8D). These results might be explained by considering the `well type' potential profile of the enterocyte, where the K⁺ efflux through the basolateral membrane produces a depolarization of V_te (therefore its inhibition shows a decrease of the hypotonicity induced depolarization of V_te), while the K⁺ efflux through the apical membrane produces a hyperpolarization of V_te (therefore its inhibition shows an increase of the hypotonicity induced depolarization of V_te).

In isotonic conditions 1 μmol l^-1 apamin had no effect on the basal electrophysiological parameters of either the basolateral or the mucosal membrane (Fig. 9A,B), but in hypotonic conditions, when applied on the serosal side, it decreased the hypotonicity induced depolarization of V_te only during a limited temporal window in the second phase(Fig. 9C). When apamin was applied on the mucosal side it slightly increased this depolarization in the second phase (Fig. 9D).

Effective volume regulation on each membrane requires a parallel efflux of anions. Therefore, the involvement of a volume-activated anion channel was tested using 0.5 mmol l^-1 DIDS as an inhibitor. In isotonic conditions mucosal DIDS had no effect on the basal electrophysiological parameters V_te and I_sc(Fig. 10B), but in hypotonic conditions, when applied on the mucosal side(Fig. 10C), it significantly decreased the hypotonicity induced depolarization either in the first or the second phase. This result is consistent with the activation of anion channels on the apical membrane during the hypotonic stress response. On the basolateral (serosal) side DIDS significantly inhibited basal electrophysiological parameters (Fig. 10A), because of its inhibition of basal Cl^-conductance (Bicho et al.,1999). In hypotonicity conditions, when applied on the basolateral side, DIDS also induced a significant decrease in the second phase of the hypotonicity response (Fig. 10C). Similar results (not shown) were obtained with 0.5 mmol l^-1 NPPB, another inhibitor of basolateral Cl^-conductance.

As reported in Fig. 11, 1 h incubation of the tissue with BAPTA-AM (50 μmol l^-1), a chelator of intracellular Ca²⁺, significantly reduced the V_te hypotonicity induced depolarization. Similar results were obtained when the tissue was preincubated with 10 μmol l^-1trifluoroperazine, a calmodulin inhibitor(Fig. 12).

Ca²⁺ removal from the Ringer solution nullified the electrophysiological response both in the first and in the second phase(Fig. 13B), on the contrary 1μmol l^-1 thapsigargin [a selective inhibitor of the sarco(endo)plasmic reticulum type of Ca²⁺-ATPase leading to depletion of the intracellular Ca²⁺ stores(Tharstrup et al., 1990)] did not elicit any significant effect (Fig. 13A).

Euryhaline teleosts are able to adapt to varying osmolarity of the external environment, so are good physiological models for studying cell volume regulation in anisosmotic conditions in vivo(Schmidt-Nielsen, 1977) and in tissues in vitro (Hoffmann et al., 2002; Zadunaisky et al.,1995; Lionetto et al., 2001, 2002).

In the present study we investigated the integrated ion transport response activated by external hypotonicity in a native epithelium, the intestine of the euryhaline teleost Anguilla anguilla, physiologically exposed to hypotonic stress. When the eel moves from seawater to freshwater the intestine is exposed to hypotonicity either on the basolateral side, because of the dramatic decrease in the eel plasma osmolarity observed during the first 90 h after transfer of the animal (Fig. 1), or the apical side, which can be exposed to a very diluted medium because of the drinking behavior known in freshwater eels(Maetz and Skadhauge,1968).

Morphometrical analysis of the height of the epithelium revealed that the eel intestine initially swelled following an hypotonic stress, but soon after,in the following 45 min, it performed a regulatory volume decrease (RVD), as do many other epithelia (for reviews, see Hoffmann and Kolb, 1991; Larsen and Spring, 1987; Macknight, 1991; Hoffmann and Ussing, 1992). The parallel monitoring of the transepithelial electrophysiological parameters revealed that during the swelling phase and the RVD response there was a biphasic decrease in V_te as well as in I_sc. There were also significant changes in transepithelial resistance induced by hypotonicity exposure. In a leaky epithelium such as the eel intestine the transepithelial resistance is dominated by the resistance of the paracellular pathway, which in turn is due to the resistance of the junctional complex in series with the lateral intercellular space. Following a hypotonic stress the transepithelial resistance showed an increase during the swelling phase, followed by a slow recovery towards the initial value after 60 min exposure. These R_te changes, similar to those observed for other intestinal epithelia such as rat colonic epithelium(Diener et al., 1992) and rat small intestine (Diener et al.,1996), are in the direction expected for narrowing of lateral intercellular spaces during the swelling phase, followed by their slow widening in the RVD phase. However, hypotonicity induced changes in junctional resistance cannot be ruled out.

The use of specific ion transport inhibitors allowed us to clarify the ionic nature of the hypotonicity induced decrease in V_teand I_sc.

Previous studies have shown that under basal conditions 80-90% of the eel intestinal V_te and I_sc is the result of net transepithelial Cl^- absorption, which is accomplished by the Na⁺-K⁺-2Cl^- cotransporter(Trischitta et al., 1992a),and that external hypertonicity increases V_te and I_sc by stimulation of the triporter(Lionetto et al., 2001). The possibility that the observed decrease of V_te and I_sc induced by hypotonicity might be the result of an inhibition of the luminal Na⁺-K⁺-2Cl^-cotransporter was ruled out by the observation that the decrease of V_te and I_sc induced by external hypotonicity remained unaltered after luminal blockage by bumetanide of Na⁺-K⁺-2Cl^- triporter, suggesting that other ion transport mechanisms could be involved.

The decrease of V_te and I_sc induced by the hypotonic media reflects an electrogenic response of the tissue arising from the ion movements across the enterocyte cell membranes following their conductance changes. The Ba²⁺-sensitive basal K⁺conductance is not involved, since the electrogenic response was unaffected by serosal or mucosal application of Ba²⁺. By contrast, the response was reduced by serosal application of iberiotoxin and apamin, specific inhibitors of high conductance (BK) and small conductance (SK2 and SK3)Ca²⁺-activated K⁺ channels, respectively, but was enhanced by the luminal application of iberiotoxin and slightly increased by luminal apamin. These results are consistent (as explained in the Results)with the hypotonicity induced activation of BK and SK Ca²⁺-activated K⁺ channels on the apical and the basolateral membranes. Furthermore, the iberiotoxin and apamin-sensitive components showed different time courses of activation: the iberiotoxin-sensitive pathway was rapidly activated in the first minutes of the response on the basolateral membrane, but on the mucosal membrane its main contribution was in the second phase of the response; the apamin-sensitive pathway contributed to the second phase of hypotonicity response on the serosal membrane, but across the apical membrane appeared to be minor, if present at all. The importance of K⁺ efflux in the hypotonic stress response is also suggested by ouabain experiments, where dissipation of the K⁺ gradient resulting from the pump inhibition significantly decreased the response.

The hypotonicity induced activation of high conductance Ca²⁺-activated (BK) K⁺ channels is reported to occur in other epithelial cell types such as human bronchial epithelial cell line(Fernandez-Fernandez et al.,2002), cultured proximal tubule cells(Kawahara et al., 1991) and rat collecting tubules (Stoner and Morley,1995). In a diverse range of epithelial cells, from kidney(Bolivar and Cereijido, 1987; Hirsch et al., 1993), male reproductive tract (Sohma et al.,1994), oviduct (James and Okada, 1994), and vestibulum (Takeuchi, 1992), high conductance Ca²⁺-activated K⁺ channels have been detected on the apical membrane. However, they have also been reported in the basolateral domain of enterocytes (Burckhardt and Gogelein, 1992) and epithelial cells from acinar glands (Maruyama,1983). In the present investigation, functional evidence of the hypotonicity induced activation of these channels on both membranes is presented. Moreover,the channels do not seem to contribute to ion transport processes under basal conditions.

In our experimental model hypotonicity also seems to activate apamin-sensitive K⁺ conductance, which is described in the literature as small conductance Ca²⁺-sensitive K⁺channels of the type SK2 and SK3 (Bond et al., 1998), reported also to be involved in the RVD response after swelling in human liver cell line (Roman et al., 2002).

To our knowledge this is the first evidence of functional involvement of both high conductance and small conductance K⁺ channels in the hypotonicity induced ionic response, although in human liver cell, Roman et al. (2002) also suggested that K⁺ channels additional to the SK2 channel were involved in RVD.

In eel intestinal epithelium the hypotonicity induced K⁺ channel activation is paralleled by an anion channel activation on the apical membrane, where DIDS, inhibitor of volume-activated anion channels(Strange et al., 1996),significantly reduced the V_te response. Activation of the DIDS-sensitive conductance appears rapidly in the first minutes of the V_te depolarization and is also maintained in the second phase. DIDS applied on the mucosal side was effective only during the hypotonic challenge and had no significant effect on basal electrophysiological parameters, suggesting a specific role of this conductance on the apical membrane in the hypotonic stress. On the serosal side in isotonic conditions DIDS inhibited the basal electrophysiological parameters, via inhibition of basal Cl^- channels(Bicho et al., 1999) present on the basolateral membrane. In hypotonic conditions basolateral DIDS also induced a significant decrease in the second phase of the hypotonicity response, similar to results obtained with basolateral K⁺ channel blockers. Similar results were obtained using NPPB, another inhibitor of basolateral Cl^- conductance (data not shown), suggesting that the hypotonicity induced activation of basolateral K⁺ conductance requires the functional integrity of basal Cl^- channels.

On the basis of these results we conclude that in the eel intestinal epithelium a hypotonic stress activates separate K⁺ and anion conductances, both on the basolateral and the apical membranes(Fig. 14). The electrogenic nature of the response can be explained by a K⁺ and anion conductance increase in a ratio different from 1:1 on the two membranes. Since the electrogenic response is a depolarization of V_te, we suggest that on the basolateral membrane the increase in the K⁺conductance, exceeds the increase in the Cl^- conductance; by contrast on the apical membrane, the increase in the anion conductance exceeds the increase in the K⁺ conductance. It seems that the asymmetric increase of cation (predominantly in the basolateral membrane) and anion(predominantly in the luminal membrane) conductances in response to hypotonicity are needed in order to balance cation and anion fluxes across the respective membrane of the enterocyte. Before adding the hypotonic solution the apical and basolateral membrane conductances are very different(Trischitta et al., 1992b). Together the various conductance changes on the two membranes result in the measured potential difference depolarization across the epithelium (normal polarity: serosal side negative) and thus also in the measured decrease in I_sc. The typical time course of the electrogenic response could result from the sum of the different time courses of activation of these separate conductive pathways on the two membranes.

Previously studies on Japanese eel have demonstrated the presence of aquaporin 1 in the apical membrane of the enterocytes, and another epithelial aquaporin isoform is supposed to be present on the basolateral membrane(Aoki et al., 2003). Therefore,it is possible to argue that in our experimental model KCl efflux on the apical as well as on the basolateral membrane is followed by the osmotically required loss of water through water-permeant aquaporins present on the two membranes, thus accounting for the RVD response observed on morphometrical analysis of the epithelium.

An increase in intracellular Ca²⁺ concentration is important for the hypotonicity induced decrease in V_te and I_sc, since chelation of intracellular Ca²⁺ by BAPTA-AM produced a significant although not complete inhibition of the response; comparison of the time courses of the effects of BAPTA-AM and trifluoroperazine allows us to hypothesize that the role of intracellular calcium in the hypotonicity response is mediated by a Ca²⁺-calmodulin pathway. The lack of any effect of thapsigargin ruled out the hypothesis that the release of intracellular Ca²⁺stores is a Ca²⁺ signal in the eel intestinal hypotonicity response. On the other hand, when Ca²⁺ was removed from bathing solution the decrease in V_te and I_scwas completely abolished, clearly indicating a dependence of extracellular Ca²⁺ in the hypotonicity response. We therefore suggest that a swelling-activated influx of Ca²⁺ increases local cytosolic Ca²⁺ levels, thereby activating BK and SK channels, both being strongly dependent on intracellular Ca²⁺(Vergara et al., 1998). With respect to the increase in Cl^- conductance, we cannot yet distinguish whether we are dealing with Ca²⁺-activated Cl^- channels or swelling-activated Cl^- channels, since both can be inhibited by DIDS. Considering that some Cl^- channels are activated by extracellular Ca²⁺(Uchida et al., 1995;Waldegger and Jentsch, 2000), it is possible to argue that in the eel intestine extracellular calcium could also be essential for the activation of hypotonicity activated anion channels.

This work was supported by M.I.U.R. grants.