Chloride Conductance Across Toad Skin: Effects of Ionic Acclimations and Cyclic Amp and Relationship to Mitochondria-Rich Cell Density

The Cl⁻ conductance (G_Cl) across amphibian skin epithelium has been studied extensively in the past two decades (for reviews, see Larsen, 1991; Katz and Nagel, 1994). The transport of Cl⁻ is passive, but a selective transepithelial conductance is activated by an inwardly directed (inside-positive) electrical potential. The voltage-activated conductance shows a characteristic, non-linear pattern of time-dependent increase. The pathway is assumed to be localized to the mitochondria-rich (MR) cells of the epithelium, but this is still under debate because recent studies could not localize more than 20 % of the total Cl⁻-dependent current over MR cells (Nagel et al., 1998). Cyclic AMP and IBMX (isobutylmethyl xanthine) exert different effects on transepithelial anionic conductance (Katz and Nagel, 1995), but the molecular mechanisms involved have not been resolved. The Cl⁻ conductance is greatly decreased in the skin of toads acclimated to high external [NaCl] (Katz and Larsen, 1984), and this was paralleled by a decrease in MR cell density. However, later studies found that, in both frog and toad skin after acclimation to KCl, the density of mitochondria-rich cells (D_mrc) is maintained or increased (Ehrenfeld et al., 1989; Katz and Gabbay, 1995). Direct methods have failed to solve the problem of localization of G_Cl in amphibian skin. Therefore, an indirect approach using the conventional voltage-clamp technique was undertaken, taking advantage of the large variability in G_Cl and D_mrc in Bufo viridis (Katz and Gabbay, 1995) acclimated to different conditions. In the present study, the effects of long-term acclimation to NaCl and KCl, the effects of cyclic AMP and IBMX on the anionic conductance and the density of MR cells in the skin of the toad under these conditions were investigated.

Toads (Bufo viridis) were of local origin. They were maintained in the laboratory (19–22 °C) with free access to tapwater (40–50 mosmol kg⁻¹; approximately 15 mmol l⁻¹ Cl⁻) and fed mealworms once a week. For acclimation, the toads were kept in a solution 2–3 cm deep of appropriate ionic composition; acclimation to high [NaCl] (200 mmol l⁻¹) was achieved in steps over 6–8 days. No mortality occurred over the course of the experiments. Plasma (collected from the vena angularis in the mouth) and urine (collected through the cloaca in plastic tubes) composition were determined using a Wescor 5520 vapour pressure osmometer (osmolality), a Corning 480 flamephotometer ([Na⁺] and [K⁺]) and a Radiometer (Copenhagen) chloridometer ([Cl⁻]).

Electrophysiological measurements were performed on pieces of abdominal skin, as described previously (Katz and Nagel, 1995). Briefly, the skin was mounted in a modified Ussing chamber (area 0.5 cm²) continuously perfused at more than 4 ml min⁻¹ with Ringer’s solution on both sides. Normal Ringer had the following composition (in mmol l⁻¹): Na⁺, 115; K⁺, 2.5; Ca²⁺, 1; Mg²⁺, 1; Cl⁻, 117; Hepes, 3.5; pH 7.6. Ionic replacements were made on an equimolar basis. After elimination of Na⁺ transport by the addition of 10 μmol l⁻¹ amiloride to the mucosal fluid, tissues were depolarized to −30 mV to deactivate the Cl⁻ conductance (G_Cl). To activate G_Cl, transepithelial potential was clamped intermittently to +80 mV (serosa-positive). Reported values of G_Cl reflect the difference between the total conductance (G_t) at −30 and +80 mV.

Dose–response relationships for the effects of cyclic AMP and IBMX on the deactivated conductance and on voltage-activated G_Cl and G_t were measured after normalization of the values to that under the respective control conditions. A sigmoidal regression function was obtained by data fitting using a commercial software package (Origin, Microcal Inc.). CPTcAMP (8-(4-chlorophenylthio) cyclic AMP), the permeable, non-hydrolysable derivative of cyclic AMP, and IBMX were purchased from Sigma.

The density of MR cells (D_mrc) was estimated by counting silver-stained cells in the pieces of skin used in the electrophysiological analyses. Briefly, the skin was rinsed with distilled water, incubated for 5 min with 0.25 % AgNO₃ and then washed and exposed to strong illumination for approximately 30 min. The diameter of the apical aperture of these cells was measured using a scaled eyepiece micrometer.

Statistical analyses were performed using analysis of variance (ANOVA). Results are presented as means ± S.E.M.

The toads used in the present study were from two series of experiments (40 toads in winter and 12 toads in summer). The animals were acclimated to tapwater and distilled water, to 50 and 200 mmol l⁻¹ NaCl and to 50 mmol l⁻¹ KCl for more than 12 days. Table 1 summarizes the plasma and urine ionic composition in toads (winter series) acclimated to tapwater, 50 mmol l⁻¹ KCl and 200 mmol l⁻¹ NaCl. The plasma osmolality remained relatively high in control, tapwater conditions in Bufo viridis, which is a characteristic of this species, compared with the value in most other amphibians (Table 1). Urea accounts for the difference between the measured osmolality and that calculated from measurements of the ionic composition. In all conditions, plasma ion levels remained fairly stable, but the urine was always hypotonic to the plasma. Plasma [K⁺] nearly doubled in toads acclimated to KCl and 200 mmol l⁻¹ NaCl (although the difference is not significant; P=0.053), and plasma [K⁺] in the urine was significantly higher than the control value only in the KCl-acclimated toads (P<0.001).

The physiological state of the skin can be judged from the short-circuit current (I_sc; active Na⁺ transport) recorded at the start of each experiment. I_sc was 21±4 μA cm⁻² in skins from distilled-water-acclimated toads, 21±3 μA cm⁻² in skins from tapwater-acclimated toads and 36±7 μA cm⁻² in skins from KCl-acclimated toads. The I_sc of skins from NaCl-acclimated toads tended to reduce (P=0.5) for those acclimated in 50 mmol l⁻¹ NaCl (18±5 μA cm⁻²) and diminished for those acclimated to 200 mmol l⁻¹ (8±5 μA cm⁻²) compared with tapwater-acclimated toads. Only in the last group was the value significantly lower than in the other groups (P<0.05).

Movements of Cl⁻ across the skin were analyzed in non-transporting conditions after apical application of 10 μmol l⁻¹ amiloride. Similar results were obtained in the two series, and the data were pooled. Numerical values of current and conductance were obtained from recordings such as that shown in Fig. 1. The effects of IBMX and CPTcAMP on the total conductance (G_t) are shown for a representative skin in Fig. 2. Voltage activation (+80 mV) was sensitized reversibly by IBMX, whereas high concentrations of cyclic AMP led to an elevation of the baseline conductance. Moreover, following the addition of cyclic AMP, transepithelial conductance was almost insensitive to voltage activation, and became time-independent within the resolution of the system (Katz and Nagel, 1995). External Cl⁻ was not required for activation in this condition, and the effects were reversible.

The concentrations of cyclic AMP and IBMX used were supramaximal, as is evident from dose–response curves (Fig. 3) with either CPTcAMP (Fig. 3A,B) or IBMX (Fig. 3C). The EC₅₀ values calculated from these experiments were 9–13 μmol l⁻¹ for CPTcAMP and 69 μmol l⁻¹ for IBMX. The baseline transepithelial conductance (−30 mV depolarization) and the voltage-activated (+80 mV) Cl⁻-specific conductance (G_Cl) are shown in Fig. 4 for control (tapwater) and other conditions of acclimation. The non-specific conductance (hatched columns) was similar under all acclimation conditions. The voltage-activated (+80 mV) Cl⁻-dependent conductance G_Cl (open columns) varied among the various acclimation groups. The higher the [NaCl] in the acclimation solution, the lower the absolute value of the activated conductance. Note that this conductance was lost from the skins of toads acclimated to 200 mmol l⁻¹ NaCl. The high G_Cl values of skins from KCl- and distilled-water-acclimated toads did not differ significantly from each other. The effects of IBMX and CPTcAMP are summarized separately for the deactivated and activated conductances in Fig. 5. The effects in the deactivated (−30 mV) condition in skins from all acclimation groups are shown in Fig. 5A. IBMX (100 μmol l⁻¹) had a very small and usually only transient effect (not significantly different from the control value in all groups), while cyclic AMP (200 μmol l⁻¹) greatly increased the transepithelial conductance in this condition. Similar results were obtained under hyperpolarized (+80 mV) conditions (hatched columns), showing that the conductance is insensitive to the transepithelial potential. Only in the skins from distilled- and tapwater-acclimated toads was the difference between the deactivated and activated conductances in the presence of cyclic AMP significant (P<0.05). Inhibition of the Na⁺/K⁺/2Cl⁻ cotransporter by bumetanide (100 μmol l⁻¹; serosal side) did not affect the response to cyclic AMP (not shown). Fig. 5B shows the voltage-activated time-dependent conductance G_Cl in control conditions and following treatment with IBMX (stippled columns) or cyclic AMP (hatched columns). The stimulation of G_Cl by IBMX was most pronounced in skins from NaCl- and tapwater-acclimated toads compared with their control. The lower the control G_Cl, the greater was the relative effect of IBMX on the activated conductance. Thus, G_Cl in skins of distilled-water- and KCl-acclimated toads, which was already at its maximal level, was not affected significantly by IBMX. A completely different response was observed upon application of CPTcAMP. The time-dependent voltage-activated conductance was greatly reduced; in skins from NaCl-acclimated toads, it was abolished. The response became virtually time-independent within the time frame of activation (see Fig. 2).

Fig. 6 relates all the values for the voltage-activated (+80 mV) transepithelial conductance (G_Cl) under control condition to the corresponding MR cell density (D_mrc). The calculated regression line had a very shallow slope (y=0.0011x+0.93) that was not significantly different from zero. Separate calculations of regression lines are given in Table 2 for skins of all acclimation groups in the two series (winter and summer). Similar values were obtained in all calculations, including those for the tapwater- and distilled-water-acclimated groups calculated separately.

Silver-stained MR cells from the five acclimation groups are shown in Fig. 7. It has already been demonstrated in histological cross sections that the silver deposit is confined to the apical membrane of the cells (Whitear, 1972; Katz et al., 2000). The diameter of the apical silver-stained aperture of the MR cells was largest in the skins of distilled-water- and KCl-acclimated toads (mean diameter 13.8±0.37 μm; N=8). Note that MR cells with two distinctly different appearances are detected in the skin of KCl-acclimated toads. A smaller cell aperture was measured in the skin of tapwater- and 50 mmol l⁻¹ NaCl-acclimated toads (mean diameter 9.4±0.45 μm; N=6). In the 200 mmol l⁻¹ NaCl acclimation conditions, skin MR cell density changed little in the first 6–7 days (not typical) and was similar to the value for the 50 mmol l⁻¹ acclimation condition. The number and appearance of these cells was subsequently decreased, as shown in Fig. 7E,F. The anion conductance of these skins was reduced similarly.

Our data extend previous studies (Katz and Larsen, 1984; Katz and Gabbay, 1995) showing that Cl⁻ conductance is influenced in the long term by the salinity of the ambient solution. It was also established that the short-term effects of cyclic AMP and IBMX on the anionic conductance across the skin are influenced by the ionic acclimation of the animals.

The Cl⁻ conductance across amphibian skin is controlled by β-adrenergic receptors (Willumsen et al., 1992; Nagel and Katz, 1998) that activate adenylate cyclase, leading to increased cellular cyclic AMP levels and activation of protein kinase. Cyclic AMP and IBMX affect the transepithelial anion conductance in fundamentally different ways (Katz and Nagel, 1995), although cellular cyclic AMP levels must be increased by IBMX because of inhibition of phosphodiesterase (PDE; Ukena et al., 1993). Two steps can be distinguished in the movement of anions across toad skin, activation and conductance (Larsen, 1982; Nagel and Katz, 1997). In this scheme, Cl⁻-dependent voltage activation is followed by anion conductance (Larsen, 1991; Lacaz-Vieira and Procopio, 1988; Nagel and Katz, 1997). IBMX may have a structural effect, different and separate from its inhibition of PDE that sensitizes voltage activation. Indeed, it was reported recently that xanthine derivatives lacking an inhibitory effect on PDE stimulated the cystic fibrosis transmembrane conductance regulator (CFTR), whereas no such effect was elicited by specific inhibitors of PDE (Chappe et al., 1998). At high concentrations, cyclic AMP exerts a more profound effect, proximal to the voltage sensor, and the pathway becomes insensitive to voltage, keeping the conductive path open to the passage of anions.

The effects of cyclic AMP and IBMX on the anion conductance of the skin were influenced by the conditions of acclimation of the toads. IBMX was effective only on the activated conductance, particularly when the initial value had not attained its maximal level. It had only a small and transient effect on the deactivated conductance, which could be the result of a temporal increase in cellular cyclic AMP levels. The increase in transepithelial anionic conductance in response to a high concentration of cyclic AMP was insensitive to transepithelial potential and became independent of external [Cl⁻]. Because the total anionic conductances, after voltage activation or in response to cyclic AMP, were similar in magnitude, they may represent the same conductance pathway. Thus, an apparently ‘dormant’ conductance is uncovered by cyclic AMP in skins from NaCl-acclimated toads, in which the activated conductance was greatly reduced. This observation suggests that acclimation to high [NaCl] desensitizes the voltage sensor and that cyclic AMP, at relatively high concentrations, relieves this block, opening the path to the conductive passage of anions.

In earlier studies, it was reported that D_mrc is greatly decreased following adaptation to NaCl, in parallel with a reduction in skin Cl⁻ conductance. This was assumed to result from the long-term exposure of the animals to high ambient [Cl⁻] (Katz and Larsen, 1984). However, it was discovered subsequently that D_mrc is not reduced in response to acclimation to KCl (Ehrenfeld et al., 1989; Katz and Gabbay, 1995), which is shown in the present study to be accompanied by an increase in G_Cl that reaches a maximal level similar to that of the conductance of skins from distilled-water-acclimated toads. It has been suggested previously that acclimation to KCl induces plasma acidosis (Ehrenfeld et al., 1989), which necessitates an increased secretion of protons. Indeed, it was shown that MR cells in frog skin react to anti-H⁺-ATPase (Klein et al., 1997), and this has been confirmed recently for Bufo viridis (U. Katz and S. Gabbay, unpublished experiments). This supports the assumption that MR cells are involved in proton secretion, a role similar to that of α-type intercalated cells in other tight epithelia (Steinmetz, 1986). The elevated G_Cl in KCl-acclimated toads could therefore be explained as a consequence of the maintained D_mrc, if MR cells are also endowed with an anion conductance pathway. This remains to be verified more directly.

Our measurements covered a wide range of values of transepithelial conductance and D_mrc, but there was no correlation between these two parameters in any of the acclimation conditions. A closer relationship was observed in previous studies (Katz and Gabbay, 1995), and some correlation has been found between Cl⁻ conductance and D_mrc in some studies in other species. Voute and Meier (1978) reported some correlation in frog (Rana esculenta) skin, as did Willumsen and Larsen (1985) in toad (Bufo bufo) skin. Although no activated conductance can be elicited in skins with no MR cells or only a few MR cells (Katz et al., 2000), Nagel and Dorge (1990) found no correlation between these parameters in frog (Rana esculenta) skin. This relationship is therefore unpredictable, and it has not yet been possible to determine the reason for these variable results. It could perhaps be explained if MR cells were functionally heterogeneous (Katz et al., 1997) and if they were not the exclusive route for passive anion conductance. Proton secretion, for example, if carried out by MR cells, could also influence cell density in certain conditions, and Cl⁻ conductance is not necessarily, therefore, the sole determinant influencing MR cell density. Maximal skin G_Cl differed considerably among the various acclimation conditions, and these two functions, i.e. G_Cl and H⁺ secretion, which are assumed to be localized in MR cells, may be expressed differently or in different cells according to the physiological demands at a given condition.

The activated conductance was always very low in toads acclimated to high [NaCl], regardless of the density of MR cells, indicating that the conditions of acclimation, rather than D_mrc, determine the anion conductance that can exist in different states. We have limited evidence for the existence of different MR cell types. This evidence, at the cellular level, is based on silver staining and immunological methods (Katz et al., 1997).

The primary signal(s) responsible for the acceleration of cell division and/or differentiation is still unknown. The actual functional properties of the cells are yet to be resolved and certainly depend on more than one determinant, external [Cl⁻] being only one. This could explain why no correlation between D_mrc and G_Cl was found, although current results suggest that they must be associated.

In conclusion, we have shown that the transepithelial anionic conductance across Bufo viridis skin is influenced differently by long-term acclimation to external NaCl and KCl solutions. These changes could not be corrected by IBMX or by cyclic AMP. No correlation was found between MR cell density and Cl⁻ conductance across the isolated skin. It is suggested that, as well as being sites of Cl⁻ conductance across the skin, the MR cells must also be involved in other transport functions that would mask any possible relationship between the density of these cells and Cl⁻ conductance.

This research was supported by Technion V.P.R. Fund for the promotion of sponsored research.