INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

THERE HAVE BEEN FEW electrophysiological studies of the mouse gallbladder epithelium (18, 22). The organ in the mouse is a small, ovoid sac, with a long axis of 2-3 mm, making the conventional short-circuit current (I_sc) approach technically demanding. However, the advent of transgenic mice, in which genes of interest are "knocked out" or "knocked in," has led to important understandings of function. Consequently, the physiology of this species has become an important area of study. Here we have made observations of the electrogenic transport of ions across the gallbladder epithelium to gain insight into the mechanisms of bile secretion in this organ.

The literature on bile secretion is derived from species other than the mouse. It is known that bile is concentrated in the gallbladder by fluid absorption driven by active salt transport. Bile formation occurs in the bile canaliculi by bile salt-dependent and -independent secretion of salts and water. Bile salt-independent secretion extends into the bile duct and the gallbladder itself, both the secretory and absorptive functions being the responsibility of a simple columnar epithelium (7). However, the apparently identical columnar epithelial cells have heterogeneous functions (1). Human gallbladder epithelial cells have both Ca²⁺- and cAMP-dependent Cl efflux pathways, and natural secretion may be under the control of vasoactive intestinal polypeptide (VIP) (7). A variety of agents can modify the transporting activities of gallbladder epithelial cells in a species-dependent manner. In the dog an increase in intracellular Ca²⁺ converts absorption to secretion by a prostaglandin-dependent mechanism (16); in humans somatostatin analogs decrease the secretion of bile acids, HCO₃, and lipase (17). The effects in both tissues are due to interactions with the adenylate cyclase system. Here we show that agents that increase intracellular cAMP in mouse gallbladder epithelium cause electrogenic anion secretion, with HCO₃ carrying the major fraction of the current.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Animals. All experiments were carried out in 2- to 3-mo-old mice killed by exposure to an increasing concentration of CO₂. The care and use of the animals conformed to the requirements set by the Home Office. The gallbladder and as much of the common bile duct as possible were removed by fine dissection under a microscope. Tissues were placed in cold Krebs-Henseleit solution (KHS) immediately and mounted, after opening, as soon as possible.

I_sc recording. Gallbladders were mounted in a specially constructed chamber, cushioned with silicone washers and with a window area of 4 mm², and bathed on both sides with KHS (20 ml warmed to 37°C), continually circulated by bubbling with 95% O₂-5% CO₂. The chambers were initially assembled without tissues for the purpose of balancing the electrodes. In experiments of long duration, periodic checks were made to ensure the electrodes remained in balance. However, if asymmetry was introduced by having solutions of different composition on either side of the tissue, the asymmetric electrode potentials introduced an apparent change in basal I_sc. The tissues were short circuited with a World Precision Instruments dual-voltage clamp with series resistance compensation, and the I_sc was recorded continuously with a MacLab and associated AppleMac computer. KHS had the following composition (in mM): 117 NaCl, 4.7 KCl, 2.5 CaCl₂, 1.2 MgSO₄, 1.2 KH₂PO₄, 25 NaHCO₃, and 11.1 glucose (pH 7.4 at 37°C). In Cl-free solution, sodium isethionate, potassium sulfate, and calcium sulfate replaced the corresponding Cl salts. To make HCO₃-free solutions, NaHCO₃ and KH₂PO₄ were replaced either with sucrose or NaCl and KCl and the solutions were buffered with either 5 mM Tris or 10 mM HEPES. In general, HCO₃-free solutions were bubbled with 100% O₂ and had a pH of 7.3 at 37°C. On occasion HCO₃-free solution was bubbled with 95% O₂-5% CO₂ when the pH dropped to 6. Several agents were used frequently throughout this study as follows: 10 µM amiloride applied apically, 10 µM forskolin applied to both sides, 1 mM furosemide applied basolaterally, and 100 µM acetazolamide applied to both sides.

Base content of mouse bile. A Radiometer VIT90 videotitrator with an ABU91 autoburette fitted with a 1-ml syringe used in the pH-stat mode was employed to construct a linear calibration curve for the volume of 0.01 N HCl added to maintain pH at 3.5 when aliquots of NaHCO₃ solution (25 mM) of 5-25 µl were added to a starting volume of 2 ml. Small volumes of bile were withdrawn directly from the gallbladders of freshly killed mice, using a Hamilton syringe. Volumes of bile from 2 to 8.5 µl were obtained. The concentration of base in the bile was calculated as NaHCO₃, using the calibration curve.

Statistical analysis. Student's t-test (paired or unpaired) was used as appropriate to test for significance. P < 0.05 was considered significant.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Electrogenic ion transport in wild-type mouse gallbladder epithelium. In an initial series of experiments, forskolin, an adenylate cyclase activator, was applied to short-circuited mouse gallbladders. The basal I_sc was 7.2 ± 4.3 µA/cm² (n = 21), and addition of 10 µM forskolin to the solution bathing both sides of the tissue caused a rapid increase in current of 48.2 ± 6.1 µA/cm² (n = 21). The large standard errors underlie the considerable variation in both basal and stimulated I_sc, with some tissues having virtually no I_sc before addition of forskolin. The transepithelial potential (TEP) in unstimulated bladders was close to zero but increased after forskolin. In a series of four measurements, the TEP was 0 mV with an I_sc of 1.25 ± 5.75 µA/cm². After forskolin, the mean TEP increased to 1.02 ± 0.37 mV and the mean I_sc to 39.0 ± 5.5 µA/cm², indicating a low transepithelial resistance (TER) of 26  · cm². However, the low surface area-to-circumference ratio (i.e., 0.5× radius) in these small preparations adversely affects edge damage (6) and lowers TEP. As the TEP was close to zero in untreated bladders, the TER was measured by recording the current required to impose a potential of 1.5 mV across the epithelium. For seven untreated gallbladders with a basal I_sc of 3.2 ± 4.4 µA/cm², the resistance obtained by imposing a potential was 16.6 ± 2.6  · cm².

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Effects of cAMP on short-circuit current (Isc) in mouse gallbladder. A: all records are from the same gallbladder epithelium (area 4 mm2). Agents were applied as indicated to bathing solutions on both sides of the membrane, with extensive washing between top, middle, and bottom traces. Forskolin and dibutyryl-cAMP (DBcAMP) were used at concentrations of 10 µM and 1 mM, respectively, unless indicated otherwise. B: record from a second gallbladder epithelium with DBcAMP alone. Basal Isc is shown at the beginning of each trace.

Effects of transport inhibitors. Individual gallbladder preparations were able to withstand multiple solution changes, as shown in Fig. 2. Furosemide and then acetazolamide were added after I_sc was increased with forskolin. The whole sequence was repeated after extensive washing at 90-min intervals. The low value of the basal I_sc, shown at the beginning of each trace, indicates the protocol of washing and reequilibration is adequate to reverse the effects of previous drug exposure. Initially, furosemide had no effect on current, whereas acetazolamide caused inhibition. Later, furosemide did cause a small inhibition of I_sc, and the responses to acetazolamide remained constant. In other instances (see Figs. 4 and 6), furosemide caused a minor stimulation in I_sc when first added. Results from 12 experiments were pooled in which the sequence of agents (forskolin, furosemide, and acetazolamide) was added more than once to gallbladders. The data are given in Fig. 3, which shows that acetazolamide caused a significantly greater effect than furosemide (P < 0.0002 in applications 1 and 2, n = 12) and that furosemide caused no changes that were significantly different from zero. The mean inhibition by acetazolamide of the forskolin response on first application was 38% (18.2 ± 2.1 µA/cm² for forskolin responses of 48.2 ± 6.1 µA/cm², n = 21).

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Effects of forskolin, furosemide, and acetazolamide on Isc in mouse gallbladder. The sequence of forskolin (10 µM, applied to both sides), furosemide (1 mM, applied basolaterally), and acetazolamide (100 µM, applied to both sides), with 10 min between each addition, was repeated four times with extensive washing between each addition over a period of 8 h. Basal Isc at the beginning of each trace is indicated at left.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Changes in the responses to furosemide and acetazolamide with time. Using the protocol depicted in Fig. 2, we measured the inhibition of the Isc responses to forskolin by furosemide (1 mM) and acetazolamide (100 µM). In each experiment, the whole protocol was repeated a second, and in some instances a third, time at 90-min intervals after thorough washing. Means ± SE are shown. The responses to acetazolamide remain constant over time, unlike those to furosemide. The x-axis refers to the 3 repetitions of the protocol.

Effects of ion substitutions on responses to forskolin in mouse gallbladder epithelium. The responses to acetazolamide reported above strongly suggest that HCO₃ is involved in the anion transport processes of the gallbladder. When HCO₃ and CO₂ were removed and replaced by sucrose and O₂, respectively, the responses to forskolin were severely attenuated by 84% (Table 1), as depicted in Fig. 4. Removal of HCO₃ alone, leaving CO₂ present, caused 78% inhibition (Table 1), but in this situation the pH of the bathing solution dropped to 6. In both sets of experiments, tissues were used to measure the effect of forskolin both in KHS and in the modified solution, with each tissue acting as its own control. Care was taken to ensure that the first exposure was alternately to KHS or the modified solution in a series of experiments or that the responses in the modified solution were bracketed between responses in KHS. This precaution was to avoid bias from any adverse effects of the modified solutions. Further evidence for the role of HCO₃ was sought by examining the effects of unilateral removal of HCO₃ and CO₂, replacing HCO₃ with NaCl and substituting O₂ for 95% O₂-5% CO₂ on the appropriate side. Again, care was taken to alternate the solution to which the tissue was first exposed. No effect was seen when HCO₃ was removed from the apical side (Fig. 5), i.e., the compartment into which anions were being transported. However, when HCO₃ was removed from the basolateral side, presumably the source of the transported ions, the response was reduced in eight of the nine experiments by a mean of 58% compared with controls. Not only was there a significant reduction in the responses in the absence of basolateral HCO₃ (P < 0.003), but the responses were less well maintained after the initial peak (Fig. 5).

View this table: [in this window] [in a new window]   Table 1.   Effect of HCO3 and CO2 or HCO3 removal on responses to forskolin in murine gallbladder epithelium

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Effects of HCO3 and CO2 removal on the responses to forskolin. Again using the protocol depicted in Fig. 2, we obtained sequences of Isc responses to forskolin (F; 10 µM), furosemide (Fr; 1 mM), and acetazolamide (A; 100 µM) at 90-min intervals. The protocol was repeated 5 times, but the solutions bathing the tissue were alternated between Krebs-Henseleit solution (KHS) and HCO3 and CO2-free solution. The extent of the forskolin responses is shown in µA/cm2 at bottom. The x-axis refers to the 5 repetitions of this protocol.

View larger version (13K): [in this window] [in a new window]   Fig. 5.   Effects of unilateral removal of HCO3 and CO2 on responses to forskolin. Pairs of traces shown in A and B are Isc records from single bladders that were exposed to forskolin, both in KHS and in the modified solution in which HCO3 and CO2 were absent. Modified solution was applied basolaterally in A and apically in B. C, left: data from 9 experiments identical to that shown in A. C, right: data from 4 experiments identical to that shown in B. NS, not significant.

View larger version (12K): [in this window] [in a new window]   Fig. 6.   Effects of Cl removal from both sides of the epithelium and from the basolateral side only on the responses to forskolin. Two bladders were subjected to solution changes as indicated, and the effects of forskolin (F; 10 µM), furosemide (Fr; 1 mM), acetazolamide (A; 100 µM), and ionomycin (I; 5 µM) were recorded. A: Cl was absent from the basolateral side only on the first exposure to the drugs and subsequently responses were obtained in KHS. B: responses in the complete absence of Cl are shown between responses obtained in KHS. F, forskolin; Fr, furosemide; A, acetazolamide; I, ionomycin.

View larger version (15K): [in this window] [in a new window]   Fig. 7.   Effects of Cl removal from the apical bathing solution only on the responses to forskolin. Shown are Isc records of a gallbladder subjected to the sequential drug regimen of forskolin (F; 10 µM), furosemide (Fr; 1 mM), and acetazolamide (A: 100 µM), followed by extensive washing and repeat application of the same drugs 1 h later. Tissue was exposed to the protocol 5 times with Cl-free apical solution replacing KHS in alternate sequences. The extent of the forskolin responses in µA/cm2 is shown at bottom.

View this table: [in this window] [in a new window]   Table 2.   Effect of Cl removal on responses to forskolin in murine gallbladder epithelium

Base content of mouse bile. A simple pH-stat method was used to estimate the total base content of mouse bile, expressing the result as HCO₃. In 10 measurements, the base equivalent corresponded to 38.9 ± 3.8 meq/l (range 21.4-58.0 meq/l).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

It is shown that the murine gallbladder epithelium is capable of electrogenic ion transport. The basal current, which has not been investigated, is small and often close to zero, but after treatment with forskolin to activate adenylate cyclase, an increase in both TEP and I_sc occurs. The epithelium is of the leaky type, with a low TER of 15-30  · cm², comparable to those of other leaky epithelia, such as Necturus gallbladder (13), goldfish intestine, frog choroid plexus, and the rectal gland of Squalus acanthias (see Ref. 9). However, we were not able to allow for edge damage, which is accentuated in tissues of small dimensions (6), although we took possible precautions by using soft silicone washers to cushion the tissue.

The direction of the current flow and the lack of effect of low concentrations of amiloride indicate the current is due to the electrogenic secretion of anions. This behavior is reminiscent of studies with the guinea pig gallbladder, in which cAMP converts electroneutral transport to electrogenic secretion (19, 21), which in the mouse can be activated by VIP. No response to secretin was seen, although this peptide is effective in the human gallbladder (7). The effects of other agents that interact with the cAMP cascade, such as an inhibitor of phosphodiesterase (IBMX) and the lipid-soluble cAMP analog DBcAMP, justify the assumption that the secretory effect is cAMP dependent. Although the Ca²⁺ ionophore ionomycin did stimulate I_sc in the gallbladder, the responses were relatively small compared with those with forskolin. They did not appear to be due to prostaglandin formation, as in the dog (16). We did not further investigate how increased intracellular Ca²⁺ causes secretion.

To discover the nature of the major transported species, changes in the composition of the bathing fluid plus the use of inhibitors were informative. Acetazolamide inhibited the forskolin-sensitive I_sc by 40%. Although not complete, this degree of inhibition is greater than that found in other HCO₃-secreting epithelia (15) and similar to that found in the human duodenum (12). Inhibition at this level may mean that although carbonic anhydrase-dependent hydration of CO₂ is important in generating intracellular HCO₃, it is not the only or even the predominant way that HCO₃ enters the cell. Uncatalyzed formation of HCO₃ may play a part, but retention of CO₂ in the absence of HCO₃ was not more efficient in maintaining secretion than was removing both HCO₃ and CO₂ (Table 1). It seems likely that HCO₃ enters the basolateral face of the epithelial cells directly, for example by using an Na⁺-HCO₃ cotransporter (23) or Cl/HCO₃ exchanger.

Complete removal of HCO₃ and CO₂ from the bathing solution reduced the forskolin response to 16% of normal (Table 1). Because the concentration of the major permanent anion, Cl, was 120 mM, it is likely that conventional Cl secretion involving a basolateral Na⁺-K⁺-2Cl cotransporter and apical Cl channels would have been maintained in this situation. This suggests that only a small fraction of the secretory response could be due to Cl secretion. When HCO₃ and CO₂ were removed only from the basolateral side, the response to forskolin was significantly reduced and not well maintained (Fig. 5), a situation that would have attenuated the activity of any basolateral HCO₃ cotransporters. However, with unilateral solution changes it is possible that HCO₃ crosses the epithelium from the apical side by a nonelectrogenic process, especially as the epithelium is of the low-resistance type, preventing a true HCO₃-free situation remaining in close proximity to the basolateral face and accounting for the residual response. In contrast, HCO₃ and CO₂ removal from the apical solution made no difference in the forskolin responses, which were as well maintained as in the normal situation.

Complete removal of Cl from the bathing solutions reduced the forskolin response to a low level, 10% of normal. However, the total amount of permeant anion was reduced to only 25 mM and was accounted for entirely by HCO₃. Two explanations can be offered for the failure of secretion in Cl-free conditions. First, HCO₃ might cross the basolateral face of the epithelial cells by a Cl/HCO₃ exchange mechanism, which would fail in the complete absence of Cl. However, removal of Cl from the basolateral side alone fails to cause a significant reduction in secretion, suggesting that basolateral Cl is not essential for the forskolin response, although leakage of Cl from the apical side cannot be assumed to have no effect. Second, the absence of sufficient permeant anion, when Cl is totally removed, may lead to cell shrinkage such that the secretory processes are disrupted. A more significant result is the effect of Cl removal from the apical side alone. Anion secretion in response to forskolin is significantly reduced, but only by 33%. It has been suggested, for pancreatic duct cells, that HCO₃ secretion occurs via a Cl channel in parallel with a Cl/HCO₃ exchanger (10). When Cl is absent from the apical solution, any efflux of Cl via Cl channels would be infinitely diluted, given the large volume of the rapidly stirred bathing solution, so that Cl/HCO₃ exchange could not occur. Yet, in this situation, 70% of the response to forskolin remains. It is concluded that the major fraction of the secretory response is due to the direct secretion of HCO₃ via anion channels. The minor fraction of the secretory response could be due to the operation of the parallel channel-exchanger mechanism or simply to Cl secretion. Although no significant effects of furosemide were measured, it is difficult not to accept that in some instances furosemide had a real, but minor, inhibitory effect (Fig. 2), perhaps representing a minor contribution of conventional Cl secretion.

Anion channels with a finite permeability to both Cl and HCO₃ have been described previously (11, 14), so direct efflux of HCO₃ through the apical face is possible if appropriate electrochemical gradients are present after forskolin stimulation. In summary, the major fraction of the electrogenic anion secretory response in the murine gallbladder is sensitive to acetazolamide, insensitive to furosemide, prevented by HCO₃ removal, does not depend on basolateral or apical Cl, and is attenuated by HCO₃ removal from the basolateral side. Many investigations have been made with the guinea pig gallbladder with results closely similar to those found here for the mouse. For example, the cAMP-dependent secretion was insensitive to piretanide but sensitive to acetazolamide (2) and dependent on basolateral HCO₃ (21). It was concluded, for the guinea pig, that HCO₃ exits the apical face of the epithelium mainly by a conductance pathway and partly by a parallel channel-exchanger mechanism. Furthermore, similar conclusions were reached for transport in the mammalian epididymis (4), frog gastric mucosa (3), and turtle bladder (20). In the only other studies of electrogenic transport in the mouse gallbladder (18, 22), it was assumed that forskolin caused a Cl secretory response powered by the basolateral Na⁺-K⁺-2Cl cotransporter in series with apical Cl channels. The simple expedient of using appropriate inhibitors would have shown this was incorrect. Here, the retention of a major part of the secretory response in the absence of apical Cl and with retained sensitivity to acetazolamide (Fig. 7) indicates that electrogenic secretion of HCO₃ proceeds directly through apical anion channels. Although we have not measured HCO₃ flux, because of the small window area, we could find no value in the literature for the HCO₃ content of mouse bile. The values for the HCO₃ content in human, dog, rabbit, guinea pig, and turkey bile range from 20 to 60 µeq/l, and values range from 10 to 30 µeq/l in rats and sheep (8). The mean value and range found here for the mouse, expressed as base, are comparable to those found in the majority of other species. The low value for rat bile could be related to the absence of a gallbladder in this species, so that concentration by electroneutral absorption from the gallbladder cannot occur.

As mentioned earlier, the main reason for investigating anion secretion in the mouse gallbladder epithelium was the advent of transgenic animals. As shown by Curtis et al. (5), using cystic fibrosis animals, the apically located anion channels responsible for HCO₃ secretion are the cystic fibrosis transmembrane conductance regulator channels.