ABSTRACT
Utilizing a pH-stat method, the rates of mucosal and serosal alkalinization were measured separately in the seawater eel intestine. These two rates were dependent on contralateral HCO3− concentration and were inhibited by contralateral application of DIDS, an inhibitor of HCO3− transport, indicating that the mucosal and serosal alkalinization are due to HCO3− secretion and absorption, respectively. The mucosal alkalinization was enhanced after inhibiting Na+/K+/Cl− cotransport by treatment with bumetanide, furosemide or Ba2+, with a latent period of more than 10 min, suggesting that HCO3− absorption from mucosa to serosa depends on Na+/K+/Cl− cotransport. The serosal alkalinization caused by HCO3− absorption was completely abolished after mucosal application of bumetanide. After pretreatment with bumetanide, mucosal omission of Cl− halved the enhanced rate of mucosal alkalinization, and Na+ omission had no effect on it; this indicates that the exit of HCO3− into the lumen depends on luminal Cl−, i.e. on the existence of the usual C1−/HCO3− exchange on the brushborder membrane. When serosal Na+ was removed under the same conditions, mucosal alkalinization was reduced, indicating that HCO3− entry from the serosal fluid depends on Na+. Serosal omission of Cl− did not reduce mucosal alkalinization. In addition, serosal alkalinization was enhanced by serosal removal of Na+ but not of Cl−. These results suggest that there is a Na+/HCO3− cotransport on the basolateral membrane. A possible model for HCO3− transport systems in the seawater eel intestine is proposed, and a possible role for these transport systems is discussed in relation to Na+, Cl− and water transport.
INTRODUCTION
In the preceding paper (Ando, 1990), it was proposed that HCO3− transport systems may contribute to a homeostasis in the intracellular H+ concentration (pHi), which will control Na+/K+/Cl− cotransport via pHi-sensitive K+ channels on the brush-border membrane of the epithelium in the intestine of the seawater eel. The present study aimed to elucidate how HCO3− is transported across the intestinal epithelium. However, the HCO3− flux cannot be detected directly by using radioisotopes, because labels on HCO3− are promptly dispersed into CO2 and H2O. Therefore, in the present study, the HCO3− transport rate was estimated from the rate of alkalinization of the bathing fluid.
HCO3− transport in the fish intestine has been little studied. So far as we know, the only study is that of Dixon and Loretz (1986), who observed HCO3− secretion in the goby intestine using a pH-stat method. However, they clamped the pH manually, and therefore they were not able to analyse precisely the time course of HCO3− secretion. Using an automatic pH-stat, we analysed more precisely the time course of HCO3− secretion as well as HCO3− absorption, and examined the effects of Na+, Cl−, 4,4’-diisothiocyanostilbene-2,2’-disulphonic acid (DIDS) and bumetanide on HCO3− transport. The results obtained indicate that some HCO3− absorption is linked with the Na+/K+/Cl− cotransport system, and that at least two kinds of HCO3− transport system exist in the seawater eel intestine.
MATERIALS AND METHODS
Japanese cultured eels Anguilla japonica, weighing 200-240 g, were kept in seawater aquaria (20°C) for more than 1 week before use. After decapitation, the intestine was removed and stripped of its serosal muscle layers. The stripped intestine was opened by cutting longitudinally and mounted as a flat sheet in an Ussing-Rehm chamber with an exposed area of 0.785 cm2. One side of the intestine was bathed with normal HCO3− Ringer’s solution (6.5 ml), and the other side was bathed with an unbuffered Ringer’s solution (5.0 ml). Both solutions were kept at 20°C and circulated continuously; they were gassed with a 95 % O2/5 % CO2 gas mixture or 100% O2.
Table 1 shows the composition of the Ringer’s solutions used in this experiment. Solution A is the normal HCO3− Ringer’s solution. In Na+-free Ringer’s solution (solution B), all Na+ was replaced with choline4”. Cl−-free Ringer’s solution (solution C) was made by replacing NaCl, KC1 and CaCl2 with sodium gluconate, KNO3 and Ca(NO3)2, respectively. These HCO3−-buffered Ringer’s solutions were bubbled with a 95% O2/5% CO2 gas mixture (pH7.4). Solution D is phosphate-buffered Ringer’s solution, gassed with 100% O2 (pH 7.4). Solution E is the standard unbuffered Ringer’s solution, in which HCO3− is replaced with gluconate and acetate, and MgCl2 is substituted for MgSO4. In low-Na+ unbuffered Ringer’s solution (solution F), Na+ was replaced with choline4”, and this solution was used within 1 week. Cl−-free unbuffered Ringer’s solution (solution G) was made by replacing NaCl, KC1, CaCl2 and MgCl2 with sodium gluconate, KNO3, Ca(NO3)2 and magnesium acetate, respectively. These unbuffered solutions were gassed with 100 % O2 and the pH was clamped at 7.4 using a pH-stat (TOA, HSM-10A).
The rate of alkalinization (JOH) was calculated from the amount of 20 mmol l−1 HC1 titrated automatically to clamp the unbuffered fluid pH at 7.4 using the pH stat. The amount of HC1 titrated was recorded automatically (TOA, EPR-121A) and the pH in the unbuffered medium was monitored throughout the experiment with a polyrecorder (TOA, EPR-10A). A similar technique has been used for measuring H+ secretion rate in the eel stomach (Ando et al. 1986). The transepithelial potential difference (PD) was recorded with the polyrecorder (TOA, EPR-121A) as the serosal potential with respect to the mucosa through a pair of calomel electrodes (A. H. Thomas Co.). The PD was short-circuited every 10 min for less than 10 s and the tissue resistance (Rt) was calculated from the ratio of the PD to the short-circuit current (Isc). Under short-circuit conditions, current flow from mucosa to serosa is reported as a positive Isc. The fluid resistance was 18.8 Qcm2 and this factor was also used to correct each Isc and Rt value as usual.
After these four variables had reached steady levels under the standard condition, 4-4′ -diisothiocyanostilbene-2-2′ -disulphonic acid (DIDS, Sigma), acetazolamide (Sigma) or bumetanide (a gift from Sankyo Co., Tokyo) was added to either the serosal or the mucosal fluid.
RESULTS
Mucosal and serosal alkalinization are due to HCO3− transport
When the mucosa was bathed with standard unbuffered Ringer’s solution (solution E), while the serosa was bathed with normal HCO3− Ringer’s solution (solution A), the mucosal fluid was alkalinized at a constant rate (Fig. 1). The ferosa-negative PD and Isc were maintained under these conditions. After replacement of the HCO3−-buffered solution with phosphate-buffered solution, the rate of mucosal alkalinization was reduced to zero, accompanied by a decrease in PD and Isc. The tissue resistance (Rt) tended to increase.
When DIDS, an inhibitor of HCO3− transport, was added to the serosal fluid under the same conditions, decreased gradually, accompanied by a decrease in PD and and by an increase in Rt (Fig. IB). Addition of acetazolamide, an inhibitor of carbonic anhydrase, enhanced the inhibitory effects of DIDS. When DIDS was applied to the mucosal fluid under the same conditions, decreased slightly, accompanied by a slight decrease in PD and Isc, whereas Rt did not change significantly (data not shown).
Similar experiments were performed after bathing the mucosa and the serosa with normal HCO3− Ringer and standard unbuffered Ringer, respectively (Fig. 2). After removal of HCO3− from the mucosal fluid, the rate of serosal alkalinization was reduced to zero, accompanied by a decrease in PD and Isc. and by an increase in Rt (Fig. 2A). When DIDS was added to the mucosal fluid, decreased gradually (Fig. 2B). PD and Isc also decreased after treatment with DIDS, accompanied by an increase in Rt. Acetazolamide also enhanced the inhibitory effects of mucosal DIDS on these four parameters. Serosal addition of DIDS inhibited slightly, accompanied by a slight decrease in PD and Isc (data not shown).
Effects of inhibition of Na+/K+/Cl− cotransport
To clarify the relationship between HCO3− transport and Na+/K+/Cl− cotransport, the following experiments were performed. Whilst bathing the mucosa and the serosa with standard unbuffered Ringer’s solution and normal HCO3− Ringer’s solution, respectively, 1μmoll−1 bumetanide, an inhibitor of Na+/K+/Cl− cotransport, was added to the mucosal fluid (Fig. 3A). After addition of bumetanide, PD and Isc decreased immediately and Rt increased more slowly, indicating that Na+/K+/Cl− cotransport is blocked by this drug and that The luminal K+ channels are blocked secondarily. The mucosal alkalinization increased gradually after a latent period of 10.0±1.0min (N=14). This enhancement in was completely blocked by DIDS and acetazolamide added to the serosal fluid. A similar increase in DIDS-sensitive was also observed after application of furosemide (10 μmoll−1) to the mucosal fluid. When Ba2+, a well-known blocker of K+ channels, was added to the mucosal fluid, the DIDS-sensitive was also enhanced with a latent period of 18.8±1.9min (N=5). However, PD and decreased immediately, accompanied by an immediate increase in Rt (Fig. 3B). Since bumetanide, furosemide and Ba2+ are known to inhibit Na+/K+/Cl− cotransport, these results suggest that inhibition of the cotransport either stimulates HCO3− secretion or inhibits HCO3− absorption. The following result supports the latter explanation.
Fig. 4 shows the ‘sidedness’ of the effects of bumetanide. In this experiment, the serosal HCO3− was omitted and bumetanide was added either to the serosal side or to the mucosal side. Although serosal addition of bumetanide had no effects or any of the four parameters (PD, Isc,Rt and ), mucosal application abolished , reduced PD and Isc, and caused an increase in Rt. These changes in the electrical parameters were similar to those shown in Fig. 3A.
Effects of Na+ and Cl− on HCO3− transport systems
Since HCO3− reabsorption was blocked by mucosal bumetanide, as shown in Figs 3 and 4, the following experiments were designed to clarify the mechanisms of HCO3-secretion in the presence of bumetanide. Fig. 5A shows the effects of removal of mucosal CP on mucosal alkalinization after pretreatment with bumetanide. When Cl− was omitted from the mucosal solution, was reduced by 50 % ; it recovered after the réintroduction of CD into the mucosal fluid. In the absence of Cl− in the mucosal fluid, PD and shifted their polarity to become serosa-positive and Rt increased significantly. These three electrical parameters recovered to their original levels after réintroduction of Cl− into the mucosal fluid.
The effects of mucosal Na+ on mucosal alkalinization were also examined (Fig. 5B). was not affected by lowering the mucosal Na+ concentration. When the mucosal Na+ concentration was lowered, the serosa-negative PD and Isc increased dramatically and Rt also increased significantly. These three electrical parameters returned to their original levels after réintroduction of the standard solution into the mucosal fluid.
Under the same conditions, when serosal Na+ was removed, however, was gradually reduced by 40 % (Fig. 6A). PD and Isc became more serosa-positive and Rt increased significantly. After réintroduction of Na+ into the serosal fluid, all these four parameters returned to their original levels.
In contrast, serosal omission of Cl− did not affect mucosal alkalinization (Fig. 6B). PD and Isc increased gradually and Rt increased dramatically after removal of Cl− from the serosal fluid. When normal Ringer’s solution was reintroduced, these electrical parameters recovered to their original levels.
After bathing the mucosa and the serosa with normal HCO3− Ringer’s solution and with standard unbuffered Ringer’s solution, respectively, the effects of serosal Na+ or Cl− on serosal alkalinization were examined (Fig. 7). When the serosal Na+ concentration was lowered from 142.8 to 24.3 mmol l−1 increased significantly (Fig. 7A). PD and Isc become more serosa-positive and Rt also increased significantly. When the standard solution was reintroduced into the serosal fluid, all these four parameters returned to their initial levels.
In contrast, when serosal Cl− was omitted, was not affected (Fig. 7B). PD and Rt increased significantly but Isc increased only slightly. After reintroduction of the standard solution into the serosal fluid, Rt returned to its original level, but PD and Isc were slightly lower than their original values.
DISCUSSION
The present study demonstrates that mucosal and serosal alkalinization in the seawater eel intestine are due to HCO3− secretion and absorption, respectively, since these two rates of alkalinization depend on contralateral HCO3− concentration and are inhibited by contralateral DIDS, an inhibitor of HCO3− transport (Cabantchik and Rothstein, 1972; Marsh and Spring, 1985; Jentsch et al. 1988). Acetazolamide, an inhibitor of carbonic anhydrase, enhanced these inhibitory effects of DIDS. When HCO3− transport was inhibited in both directions, the serosa-negative PD and Isc decreased and Rt increased simultaneously. These phenomena may be explained by an inhibition of luminal K+ channels, since the serosa-negative PD is mostly due to K+ leakage from the cell to the lumen in the seawater eel intestine (Ando and Utida, 1986).
Mucosal alkalinization was enhanced by the addition of bumetanide to the mucosal fluid. Since mucosal bumetanide blocks HCO3− absorption from mucosa to serosa (Fig. 4), this enhanced seems to be due to the inhibition of HCO3− reuptake from the luminal fluid. Similar enhancement in was also observed after the addition of furosemide or Ba2+ to the mucosal fluid. Since these three drugs are known inhibitors of the Na+/K+/Cl− cotransport system, these results suggest that the HCO3” reuptake processes are closely linked with Na+/K+/Cl− cotransport. However, it is unlikely that the cotransport itself carries HCO3− because the inhibition of HCO3 reuptake (enhancement of ) is delayed by more than 10 min after the initiation of changes in PD, Isc and Rt.
After blocking the HCO3− reuptake processes with bumetanide, omission of Cl− from the mucosal side halved the enhanced but Na+ omission had no effect on it, indicating that the movement of HCO3− into the lumen depends on luminal Cl−. In other words, this suggests that there is CD/HCO3− exchange on the brush-border membrane: this idea is also supported by the inhibitory effect of mucosal DIDS on , since DIDS is known to inhibit C1−/HCO3− exchange.
Mucosal alkalinization was reduced by removing Na+ from the serosal fluid but not by removing Cl− (Fig. 6), and blocked by serosal DIDS (Fig. 1). In addition, serosal alkalinization was enhanced by lowering serosal Na+ concentration, but not by removing serosal Cl− (Fig. 7). These results indicate that HCO3− entry from the serosal fluid depends on Na+ but not on Cl−, and suggest that there is a DIDS-sensitive Na+/HCO3− cotransporter which may be driven by the Na+ gradient across the basolateral membrane. Similar DIDS-sensitive Na+/ (HCO3−)n cotransport has been reported in the renal tubules of amphibians (Boron and Boulpaep, 1983; Guggino et al. 1983; Wang et al. 1987) and mammals (Good et al. 1984; Alpem, 1985; Good, 1985; Yoshitomi et al. 1985; Akiba et al. 1986; Biagi and Sohtell, 1986; Grassl and Aronson, 1986; Jentsch et al. 1986a,b;,Grassl et al. 1987; Kondo and Fromter, 1987; Sasaki et al. 1987; Ullrich and Papavassiliou, 1987), in the frog gastric fundus (Curci et al. 1987) and in bovine corneal endothelial cells (Jentsch et al. 1984, 1985; Wiederholt et al. 1985).
Although the relationship between Na+/K+/Cl− cotransport and HCO3− reuptake across the brush-border membrane is not clear yet, a plausible explanation is a coupling between HCO3− reuptake and Cl− movement out of the cell, such as C1−/HCO3− exchange, since HCO3− absorption is blocked by mucosal DIDS (Fig. 2B). Considering driving forces for such C1−/HCO3− exchange, however, the exchanger must be driven by other force(s), such as the Na+ gradient. Such a DIDS-sensitive Na+/(HCO3−)n/Cl− transport has been reported in Necturus proximal tubule (Guggino et al. 1983; Matsumura et al. 1984) and in invertebrate cells (Thomas, 1977; Boron et al. 1981). We have no direct information about how HCO3− moves from the cell into the serosal fluid, except that this process is independent of serosal Cl− and inhibited by serosal DIDS.
All the responses of the electrical parameters (PD, Isc and Rt) observed after replacement of Na+ or Cl− indicate that this tissue is substantially permeable not only to Na+ but also to Cl−, although the permeation pathways are not clear from this study.
Fig. 8 shows a possible model for HCO3− transport systems in the seawater eel intestine: the HCO3− absorption process (Na+/HCO3−/Cl− exchange and HCO3− conductance) is based on speculation from circumstantial evidence. Since NaCl and water absorption depend on HCO3− transport (Ando, 1990) and HCO3− transport also depends on Na+/K+/Cl− cotransport (present study), all these transport systems appear to be mutually interrelated. The HCO3− transport systems discussed in this paper will control the pHi homeostasis in the intestinal epithelium. Although other intracellular organic osmolytes may also control the pHi homeostasis, their contribution may be smaller than that of the HCO3−/CO2 buffer system, since amino acid metabolism is very active in this tissue (Ando, 1988). The amino acid metabolism may continuously acidify the cytoplasm. Intracellular pH may control K+ channels on the brush-border membrane, and secondarily regulate Na+/K+/CT cotransport, as discussed in the preceding paper (Ando, 1990). Among these HCO3− transport systems, the Na+/HCO3− cotransport system on the basolateral membrane might be the most important in controlling pHi, since serosal deficiency of HCO3− and serosal addition of DIDS effectively inhibit the serosa-negative PD and water absorption (Ando, 1990).
ACKNOWLEDGEMENTS
We wish to thank Professors Makoto Kobayashi and Yojiro Muneoka, Faculty of Integrated Arts and Sciences, Hiroshima University, for their helpful advice. This research was supported in part by a Grant-in-Aid no. 01304027 from the Ministry of Education, Science and Culture, Japan.