The gulf toadfish (Opsanus beta) intestine secretes base mainly in the form of HCO3− via apical anion exchange to serve Cl− and water absorption for osmoregulatory purposes. Luminal HCO3− secretion rates measured by pH-stat techniques in Ussing chambers rely on oxidative energy metabolism and are highly temperature sensitive. At 25°C under in vivo-like conditions, secretion rates averaged 0.45 μmol·cm−2·h−1, of which 0.25 μmol·cm−2·h−1 can be accounted for by hydration of endogenous CO2 partly catalyzed by carbonic anhydrase. Complete polarity of secretion of HCO3− and H+ arising from the CO2 hydration reaction is evident from equal rates of luminal HCO3− secretion via anion exchange and basolateral H+ extrusion. When basolateral H+ extrusion is partly inhibited by reduction of serosal pH, luminal HCO3− secretion is reduced. Basolateral H+ secretion occurs in exchange for Na+ via an ethylisopropylamiloride-insensitive mechanism and is ultimately fueled by the activity of the basolateral Na+-K+-ATPase. Fluid absorption by the toadfish intestine to oppose diffusive water loss to the concentrated marine environment is accompanied by a substantial basolateral H+ extrusion, intimately linking osmoregulation and acid-base balance.
- bicarbonate transport
- chloride absorption
- epithelial water transport
- seawater ingestion
- pH-stat titrations
osmoregulating marine fish maintain extracellular fluid osmolality at 300–350 mosmol/kg and, as a consequence, experience continuous diffusive water loss to the hypertonic surrounding seawater (∼1,000 mosmol/kg). To withstand this osmoregulatory challenge, fish drink seawater in amounts equivalent to ∼5% of their body mass per day (25). The ingested fluid is modified along the gastrointestinal tract, and the intestine plays a key role in marine fish osmoregulation by absorbing NaCl and water (25). After desalinization of the imbibed seawater in the water-impermeable esophagus (18, 30), fluid just slightly hyperosmotic to the extracellular compartment enters the intestine (26). Fluid absorption across the intestinal epithelium is driven by the active transport of Na+ and Cl− and occurs without or even against net osmotic gradients (15, 24, 26, 32, 42). Uptake of Na+ and Cl− from the intestinal lumen across the apical membrane of the intestinal epithelium is mediated in part by two parallel cotransport systems: Na+-Cl− and Na+-K+-2Cl− (9, 10, 16, 28), both of which rely on the electrochemical Na+ gradient established by basolateral Na+-K+-ATPase (42).
However, in addition to these cotransport systems, apical anion exchangers have recently been associated with Cl− and water absorption across the marine teleost intestine (15) and appear to contribute significantly to overall Cl− absorption. Evidence for intestinal anion-exchange activity, indicated by highly alkaline intestinal fluids with high concentrations of total CO2 or large cation-anion gaps when total CO2 was not measured, is abundant in the literature. The earliest report of high total CO2 in intestinal fluids dates back three-quarters of a century (33), and observations now include a large number of species, including elasmobranchs and sturgeon (14, 31, 39, 43, 46), suggesting that alkaline luminal fluids are common to fish inhabiting marine environments. Experiments on isolated intestinal preparations demonstrate that the source of luminal total CO2 is HCO3− secretion, which may occur along the entire length of the intestine (13–15, 46). A recent study revealed that the intestinal epithelium performs secondary active secretion of HCO3−, which accounts for the high luminal total CO2 concentrations while providing for the active absorption of Cl− via apical anion exchange in seawater-acclimated European flounder (15). A prediction of strong temperature dependence of this secondary active transport system in the gulf toadfish (Opsanus beta) was tested in the present study by direct measurement of HCO3− secretion using a pH-stat approach at 15, 25, and 35°C. Furthermore, whether aerobic energy metabolism and, thus, cellular ATP are required for the putative active HCO3− transport was examined.
Studies on the source of HCO3− secreted by the intestinal epithelium range from demonstrating that extracellular HCO3− transported across the intestinal epithelium provides the majority of HCO3− secreted into the intestinal lumen (1) to showing that endogenous epithelial CO2 constitutes the main source of luminal HCO3− (15, 45). Therefore, one goal of the present study was to determine the contribution of endogenous metabolic CO2 and extracellular HCO3− to the overall apical HCO3− secretion by the intestine of a marine teleost, the gulf toadfish. To the extent that endogenous CO2 provides for apical HCO3− secretion, cytosolic CO2 hydration must occur. This prompted examination of the potential involvement of carbonic anhydrase (CA) in intestinal HCO3− secretion. Our findings demonstrate that hydration of endogenous CO2 provides for apical secretion of HCO3−, which led to investigations of the fate of H+ from CO2 hydration. The prediction was that H+ from CO2 hydration would be eliminated from the epithelial cells to maintain cellular pH and to prevent reversal of the hydration reaction. Furthermore, because the epithelium exhibits net base secretion, it was predicted that this H+ extrusion would occur across the basolateral membrane. The possibility that basolateral H+ extrusion by the intestinal epithelium of the gulf toadfish is mediated by Na+/H+ exchange (NHE), as in the rat pancreatic ducts (38), was examined by characterization of the dependence of apical HCO3− secretion on serosal Na+ and by pharmacological manipulations of Na+ gradients. Finally, the high apical HCO3− secretion rates and the resulting need to eliminate H+ produced from the CO2 hydration reaction prompted direct measurements of basolateral H+ extrusion. The osmoregulatory role of the intestinal epithelium of marine teleost fish is salt and water absorption to maintain water balance, whereas the gill extrudes salt from the intestine (25). Thus the water absorption across the intestinal epithelium is obligately associated with the need for branchial salt secretion. Our observations of high basolateral H+ secretion rates in the water-absorbing intestinal epithelium prompted measurement of net water transport by isolated intestinal epithelia, which ultimately allowed for calculation of the pH in the absorbed fluids. These measurements revealed that intestinal fluid absorption in marine teleosts is associated with a significant acid load in addition to the unavoidable salt gain.
MATERIALS AND METHODS
Gulf toadfish (Opsanus beta) from Biscayne Bay, FL, were obtained from shrimp fishermen during spring and early summer 2005 and transferred to holding facilities at the University of Miami Rosenstiel School of Marine and Atmospheric Sciences. Procedures were approved by our Institutional Animal Care and Use Committee. On arrival, the fish received a prophylactic treatment for ectoparasites, as previously described (27). The fish were provided short lengths of polyvinylchloride tubing as shelters in 80-liter glass aquaria that received a continuous flow of filtered seawater (salinity 27–30 ppt, 22–26°C, from Bear Cut, FL). Fish were fed frozen squid twice weekly, but food was withheld 48 h before experimentation.
pH-stat experiments: general experimental protocol.
Segments of the anterior intestine of toadfish (25–40 g body wt) were obtained by dissection, cut open, and placed in tissue holders exposing 0.7-cm2 surface area with the serosal muscle layer intact. The true exposed surface area of the intestinal epithelium clearly exceed 0.7 cm2 because of the presence of villi and microvilli, but all transport rates are reported as a function of this known gross surface area. The intact intestinal epithelium was subsequently mounted in Ussing chambers (model P2300; Physiological Instruments) containing 1.6 ml of appropriate, pregassed saline in each half-chamber (Table 1). Mixing was achieved by gassing with O2 through airlifts unless otherwise stated. Pregassing of the luminal saline and continued gassing with CO2-free gas in the luminal half-chamber ensured stable titration curves throughout the experiments. The Ussing chambers were mounted in chamber holders connected via a pump to a thermostatic water bath maintained at 25°C unless otherwise stated. Current and voltage electrodes connected to amplifiers (model VCC600; Physiological Instruments) recorded the transepithelial potential (TEP) differences under current-clamp conditions at 0 μA. At 60-s intervals, 3-s 50-μA pulses were passed from the mucosal to the serosal side. The TEP measurements were logged on a personal computer using BIOPAC systems interface hardware and Acqknowledge software (version 3.8.1). TEP values are reported with a luminal reference of 0 mV. The luminal half-chamber (unless otherwise stated) was fitted with a combination pH electrode (model PHC4000.8; Radiometer) and a microburette tip, both of which were connected to a pH-stat titration system (model TIM 854 or 856; Radiometer). The pH-stat titration system was grounded to the amplifier to allow for pH readings during current pulsing. The pH-stat titrations were performed on luminal saline solutions at pH 7.800 throughout all experiments, with pH values and rate of acid addition logged on personal computers using Titramaster software (versions 1.3 and 2.1). Luminal pH was generally maintained within ±0.003 pH unit around the set point throughout the experiments. Base secretion rates were calculated from the rate of addition and concentration of titrant (0.0005 N HCl). Common to all experiments was an initial control period of ∼60 min with stable TEP and base secretion rates before the manipulations outlined below.
Nature of titrated base secretion.
To test the assumption that measured base secretion consisted of HCO3−, a separate set of experiments were performed. The luminal chamber was mixed using a peristaltic pump circulating a total of 3 ml of mucosal saline through the mucosal half-chamber with no gassing while the serosal chamber, containing HCO3−-free saline, was gassed with O2. In this preparation, O2 supply on the serosal side alone is sufficient to maintain normal base secretion rates and electrophysiological parameters (7). Secreted base was allowed to accumulate in the mucosal saline for 3 h without pH-stat titration, then total HCO3− and CO concentrations in the luminal saline were determined using double end-point titrations, which determine the titratable alkalinity of the saline. This method is described in detail elsewhere (12). Briefly, the sample was gassed with CO2-free gas for 30 min to remove gaseous CO2 from the solution; pH was titrated to 3.800 with HCl using Gilson microburettes, and the volume of HCl was recorded. At this low pH, all HCO3− and CO will convert to gaseous CO2, which was removed from the sample by continuous gassing for ≥15 min. Thereafter, pH was titrated back to the initial value during continuous gassing and careful recording of the amount of NaOH added by Gilson microburettes. From the volumes of HCl and NaOH used and the normality of these titrants, the titratable alkalinity (HCO3− equivalents) was determined from the difference in amounts of H+ needed to bring the solution to pH 3.800 and the amount of OH− needed to return the solution to the starting pH after HCO3− and CO were removed as gaseous CO2. The rate of excretion of HCO3− equivalents by the epithelium in these experiments was calculated from the total HCO3− equivalents in the mucosal saline at the end of the 3-h experimental period, the epithelial surface area, and the elapsed time. Because these experiments revealed that the majority of base secreted was HCO3−, base secretion will be referred to as HCO3− secretion.
Longevity of toadfish intestinal epithelium.
Two series of control experiments were performed to assess the temporal performance of the preparation, with gut saline on the luminal side and HCO3−/CO2 saline on the serosal side of the epithelium. In the first series of control experiments, three subsequent 60-min flux periods were separated by a change of the serosal saline. In the second series of control experiments, an initial 60-min control period was followed by two 120-min flux periods. The initial 60-min flux period and the two 120-min flux periods were again separated by serosal saline changes. The serosal saline in the first of the two 120-min flux periods contained 0.1% DMSO, which served as a vehicle control for some of the pharmacology experiments (see below).
Is aerobic energy metabolism required for HCO3− secretion?
The intestinal HCO3− secretion in marine teleosts appears to be of a secondary active nature (15) and can, therefore, be expected to rely on aerobic energy metabolism and, thus, cellular ATP. Cellular ATP levels were suppressed by N2 gassing of mucosal and serosal saline solutions. Also, for these experiments, the serosal HCO3−-free saline was replaced after 60 and 120 min with saline preequilibrated for >90 min with N2 or O2, respectively. From 60 to 120 min, gassing of mucosal saline solutions was changed from O2 to N2.
Temperature dependence of HCO3− secretion.
A strong temperature dependence of luminal HCO3− secretion was hypothesized to be due to the secondary active transport characteristics of luminal alkalinization. To test this hypothesis, measurements were performed after an initial 60-min control period at 25°C and for 120 min at 15 or 35°C, then the temperature was returned to 25°C for an additional 120-min recovery period. Exposure temperatures were verified by measurements of saline temperature in the Ussing chambers, which revealed that up to 30 min were required for a complete change in temperature from 25°C to 15 or 35°C.
Determining the source of luminal HCO3− secretion.
To assess whether secreted HCO3− is derived from endogenous epithelial CO2 or is dependent on serosal CO2 and/or HCO3−, experiments were performed to compare secretion rates obtained with serosal HEPES-saline gassed with O2 with rates obtained with serosal saline containing 5 mM HCO3− and gassed with 0.3% CO2 in O2. In these experiments, an initial 60-min control period with a HEPES-O2 serosal saline solution was followed by a 60-min period with the above-mentioned serosal HCO3−/CO2-containing saline solution before a final 60-min control period with HEPES-O2 serosal saline solution. The HCO3− concentrations and CO2 levels resulted in pH and CO2 partial pressure typical of the extracellular fluids of toadfish.
Does CA mediate CO2 hydration?
Having established that hydration of endogenous CO2 is the main source of luminal HCO3− secretion, the potential involvement of CA was examined by application of the lipophilic CA inhibitor etoxzolamide (10−3 M in 0.1% DMSO) to the luminal saline. The CA inhibitor was added after a 60-min control period, and measurements were continued for 120 min thereafter. These experiments were performed with HEPES-O2-containing serosal saline.
Involvement of serosal H+ secretion.
To test the hypothesis that serosal H+ export is important for luminal HCO3− secretion, luminal HCO3− secretion was measured during manipulation of serosal pH and, thereby, H+ gradients across the basolateral membrane, with the prediction that reduced serosal pH would inhibit luminal HCO3− secretion. In these experiments, which were performed with HEPES-O2-containing serosal saline, an initial 60-min control period at serosal pH 7.800 was followed by a 60-min period with reduced serosal pH and, finally, a 60-min recovery period at serosal pH 7.800. A total of three series were performed on three sets of preparations with serosal pH reduced to 7.4, 7.0, and 6.6.
Rate of basolateral H+ extrusion.
Considering that CO2 hydration within the intestinal epithelium provides the main source for luminal HCO3− secretion and that basolateral proton extrusion is critical for normal rates of apical HCO3− secretion, we tested the prediction that basolateral H+ secretion rates across the basolateral membrane would be high. In these experiments, pH-stat titrations were performed with the electrode and burette tip in the serosal, rather than the mucosal, half-chamber, and 0.0005 N NaOH, rather than HCl, was employed as the titrant. The serosal saline for these experiments contained neither HEPES nor HCO3− (but was osmotically compensated with mannitol; Table 1) and was gassed with O2. These titrations were also, in this case, performed with a titration set point of pH 7.800 (normal teleost fish blood plasma pH). Experiments were continued until stable titration and electrophysiological values were obtained for ≥30 min.
Na+ dependence of serosal H+ secretion: involvement of NHE.
The possibility that secretion of H+ across the basolateral membrane occurred via an Na+-H+ antiport was examined by measuring luminal HCO3− secretion under conditions with no serosal HCO3− and reduced serosal Na+. In these experiments, an initial 60-min control period was followed by a 60-min period in which the serosal saline Na+ concentration was reduced to 11 mM and a final 60-min recovery period under control conditions. In the low-Na+ saline, the majority of NaCl was replaced with choline chloride. The pH and osmolality were adjusted to match those of the serosal control saline.
Subsequently, experiments with the Na+-H+ antiport inhibitor ethylisopropylamiloride (EIPA) were performed. In these experiments, the initial 60-min control period was followed 120 min during which serosal saline contained 10−3 M EIPA in a final concentration of 0.1% DMSO.
Na+-K+-ATPase inhibitor studies.
To further examine whether serosal H+ extrusion and, thereby, luminal HCO3− secretion rely on Na+ gradients, experiments with the Na+-K+-ATPase inhibitor ouabain were performed. A 60-min control period was followed by a 120-min experimental period in which the serosal saline contained 10−3 M ouabain. Ouabain was dissolved by sonication in a low volume of serosal saline.
Rates of fluid absorption.
For measurement of rates of fluid absorption in the anterior intestine, the fish were killed by an overdose of tricaine methanesulfonate (MS-222; 0.25 g/l), and the anterior segment of the intestine was obtained by dissection. A short length of heat-flared polyethylene tubing was tied in the anterior end of the intestinal segment with double silk ligatures, and the segment was flushed with 20 ml of mucosal saline (Table 1). The proximal end of the ∼30-mm-long segments was subsequently closed with double silk ligatures. These intestinal sacs were then filled with gut saline containing 14C-labeled polyethylene glycol (PEG; 0.01 μCi/ml), and a sample of the mucosal saline was obtained as follows: the sac preparations were moderately overfilled, and the saline from the preparations was drawn from the sacs through the polyethylene catheter into a syringe and flushed back and forth three times before a sample was obtained. This procedure ensured that the sample was completely mixed and that this initial sample of gut saline was truly representative of the content of the sac preparation. The catheter was sealed, and the preparations were rinsed in serosal saline before they were submerged individually in 20 ml of serosal saline in glass vials. The serosal saline for these studies contained HCO3− and was gassed with the 0.3% CO2-in-O2 gas mixture for 90 min before and during experimentation. At the beginning and end of a 3-h flux period, samples were taken from the serosal saline. Finally, at the end of the flux, a sample of the [14C]PEG-labeled mucosal saline was obtained, the intestinal sac was cut open, and the gross surface area of the exposed epithelial surface was determined from a trace of its outline on graph paper. This type of preparation has been employed in the past and exhibits stable transport characteristics for ≥6 h (15). Water absorption was determined from the increase in the 14C radioactivity of the luminal saline during the 3 h of experimentation. Measurements of 14C radioactivity in the serosal saline solutions confirmed that all [14C]PEG was retained in the mucosal saline, despite considerable rates of fluid absorption.
Data presentation and statistical analysis.
Values are means ± SE of five to eight observations. Statistical evaluation revealed that the data were normally distributed and was performed by paired t-tests with Bonferroni's multisample comparison correction to evaluate the difference between individual time points after treatment and an average control value (34). This average control value was based on the last 30 min of the corresponding initial 60-min control flux. Samples were considered statistically significantly different at P < 0.05.
On the basis of relatively stable HCO3− secretion rates, TEP, and conductance, the anterior toadfish intestine appears viable and stable for >5 h under the conditions employed in the present study and is not influenced by 0.1% DMSO (Fig. 1). Control experiments of 3-h duration with saline changes every 60 min also revealed stable transport rates and electrophysiological parameters (data not shown). In these initial experiments in the presence of serosal HCO3−, HCO3− secretion rates were 0.3–0.5 μmol·cm−2·h−1 (Fig. 1), whereas TEP and conductance consistently were approximately −20 mV and 10 mS/cm2 for all control experiments. Compared with an overall mean control base secretion rate of ∼0.25 μmol·cm−2·h−1 (n = 42) in the absence of serosal HCO3−, double end-point titration determination of HCO3− equivalents building up in the mucosal saline in the absence of titration revealed an HCO3− secretion rate of 0.15 μmol·cm−2·h−1 (Table 2), accounting for 60% of overall base secretion. For these experiments, perfusion of the luminal chamber, rather than gas-lift mixing, did not influence TEP or conductance (Table 2, Fig. 1).
As predicted, intestinal HCO3− secretion was greatly reduced by O2 deprivation (Fig. 2). A trend toward recovery in secretion rates on return to aerobic conditions suggested reversibility of the anaerobically induced inhibition, although full return to control values was not observed in the 60-min recovery period. The 60-min anaerobic period did not significantly influence TEP or conductance.
Intestinal HCO3− secretion rates displayed marked temperature dependence, especially when temperature was reduced from 25 to 15°C (Fig. 3). Elevation of bath temperature from 25 to 35°C appeared less potent and caused a transient and modest elevation of HCO3− secretion rates. In the experiments with reduced temperature, HCO3− secretion exhibited Q10 values of 1.8–3.0. Similar responses to temperature changes were observed for conductance, which was significantly reduced at 15°C and elevated at 35°C (Fig. 3). HCO3− secretion and conductance demonstrated the reversibility of the temperature-induced changes, but this was not the case for TEP. The temperature increase from 25 to 35°C caused a gradual decline in TEP and an increased variation among preparations. In contrast, a trend toward increased TEP at the end of 120 min at 15°C was observed also with a clear increase in variation.
Compared with HCO3−- and CO2-free conditions, the presence of serosal HCO3− and CO2 caused an approximately twofold reversible increase in luminal HCO3− secretion without a change in conductance or TEP (Fig. 4). However, control HCO3− secretion rates in the absence of serosal HCO3− and CO2 were substantial (Fig. 4).
Luminal HCO3− secretion rates in the absence of serosal HCO3− and CO2 were greatly reduced on addition of the lipophilic CA inhibitor to the luminal saline (Fig. 5). This effect was accompanied by a significant reduction of TEP. Interestingly, HCO3− secretion rates, but not TEP, gradually recovered in the presence of etoxzolamide. Addition of etoxzolamide caused a slight acidification of the luminal saline, but luminal pH returned to 7.800 in all experiments within 10–15 min after addition, at which time pH-stat titrations resumed. Etoxzolamide treatment elevated conductance for some individual time points toward the end of the 120 min of exposure (Fig. 5).
As hypothesized, reducing serosal pH in the absence of HCO3− and CO2 caused a reversible and H+ concentration-dependent inhibition of luminal HCO3− secretion (Fig. 6). Serosal pH of 7.40 resulted in maximal inhibition of 33% after 60 min, whereas serosal pH of 7.0 and 6.6 resulted in maximal inhibitions of 44 and 47%, respectively (Fig. 6). The manipulations of serosal pH did not significantly alter TEP or conductance.
In a separate set of experiments, H+ secretion across the basolateral membrane was measured directly by pH-stat titration. The H+ secretion rates measured in six individual preparations were in good agreement with the overall mean control HCO3− secretion rates across the apical membrane (Fig. 7). Electrophysiological parameters for these experiments were similar to values obtained for luminal pH-stat titration experiments with TEP and conductance of −22.09 ± 2.63 mV and 8.41 ± 0.47 mS/cm2.
Having established that serosal H+ secretion is important for luminal HCO3− secretion and that the apical HCO3− and basolateral H+ fluxes occur at approximately equal rates, the Na+ dependence of the serosal H+ secretion was investigated. Reduction of serosal Na+ concentrations from 152 to 11 mM in the absence of serosal HCO3− caused a marked reduction in luminal HCO3− secretion (Fig. 8). Reduced serosal Na+ greatly reduced and even reversed TEP and, as could be expected, reduced conductance. Although HCO3− secretion showed only a slight trend toward recovery after return to control conditions, TEP and conductance almost full recovered (Fig. 8).
Surprisingly, if we consider the apparent Na+ dependence of basolateral H+ extrusion, 10−3 M EIPA did not influence luminal HCO3− secretion or conductance, whereas TEP was reduced, after 40 min of exposure (Fig. 9). These results are similar to those obtained in preliminary experiments using 10−3 M amiloride (results not shown).
Addition of the Na+-K+-ATPase inhibitor ouabain (10−3 M) caused a gradual reduction of luminal HCO3− secretion and a slight reduction of TEP with no change in conductance (Fig. 10).
Water absorption by isolated anterior intestinal sac preparations was 3.92 ± 0.92 μl·cm−2·h−1.
The isolated toadfish intestine displays stable HCO3− transport and electrophysiological characteristics over 5 h, demonstrating that this preparation, in agreement with previous work on European flounder (15), is suitable for the experimental protocols employed in the present study. Comparisons of base secretion rates obtained using pH-stat techniques with rates obtained by direct double-end-point titration measurements of HCO3− and CO equivalents revealed that the majority of intestinal base secretion occurs in the form of HCO3−. These measurements permit use of the term “HCO3− secretion,” rather than “base secretion.”
As predicted, luminal HCO3− secretion appears to rely on aerobic energy metabolism and, thus, ATP. This observation is in agreement with previous observations of reduced H+ secretion, which fuels Cl− uptake (via anion exchange) across anuran skin under anaerobic conditions (21–23), and with findings of HCO3− transport in cortical collecting ducts relying on oxidative metabolism (17). Strong temperature dependence is expected for biological reactions mediated by enzymatic reactions and transporters. Typically, increases in reaction velocity of >1.5-fold per 10°C increase are ascribed to biological, rather than physiochemical, processes (19). A 10°C reduction in temperature resulted in a 1.8- to 3.0-fold reduction in luminal HCO3− secretion, clearly indicating that intestinal HCO3− secretion in the gulf toadfish is biologically mediated. Experiments with increased temperature revealed a lower HCO3− secretion response and, perhaps, a general deterioration, as indicated by substantial nonrecoverable reduction in TEP. This latter response may reflect that 35°C is approaching the thermal maximal tolerance for the gulf toadfish, which experience temperatures ranging from 13 to 35°C in its natural environment (3). In any case, the apparent dependence of HCO3− secretion on aerobic energy metabolism and, thus, ATP and the strong reduction in secretion rates observed with reduced temperature support previous conclusions that intestinal HCO3− secretion in marine teleosts is of a secondary active nature (15).
The continued, although reduced, HCO3− secretion in the absence of serosal CO2 and HCO3− demonstrates that endogenous CO2 provides a significant (∼50%) source for luminal HCO3− secretion under resting conditions, with transepithelial HCO3− transport or serosal CO2 accounting for the remaining secretion. These observations are in agreement with reports of 30–60% of the luminal HCO3− secretion being sustained by endogenous epithelial CO2 in the European flounder and the goby (8, 15, 45) but are in contrast to findings from the Japanese eel, where serosal HCO3− seems to fully account for luminal HCO3− secretion (1). However, the studies on the Japanese eel employed hyperphysiological serosal HCO3− levels [25 mM vs. 5–8 mM normally present in teleost fish extracellular fluids (25)]. Therefore, it seems that endogenous epithelial CO2 generally contributes significantly to intestinal HCO3− secretion in marine teleosts, at least under resting, nonstimulated conditions. The remainder of the studies presented here were designed to investigate the transport processes associated with HCO3− secretion from hydration of endogenous CO2.
The HCO3− secretion rates measured in the present study agree well with previous results from marine teleost fish (1, 8, 15, 45, 46), which seem to be on the order of 0.5 μmol·cm−2·h−1 under resting conditions in the presence of serosal HCO3−. These intestinal HCO3− secretion rates from marine teleost fish are ∼10-fold higher than the only other resting HCO3− secretion rate reported from an exothermic vertebrate, the bullfrog (11), and are comparable to the range (0.5–1 μmol·cm−2·h−1) of resting duodenal HCO3− secretion rates reported from mammals at 37°C (6, 20, 40, 41). The high intestinal HCO3− secretion rates in the marine teleost, often at temperatures much below 37°C, and the significant contribution of endogenous CO2 to luminal alkalinization led to the hypothesis that CA might be involved in accelerating CO2 hydration to fuel the apical anion exchange. The lipophilic CA inhibitor etoxzolamide clearly inhibited luminal HCO3− secretion, although a gradual recovery controlled secretion rates, despite the continued presence of the inhibitor. We interpret this as an immediate depletion of cytosolic HCO3− caused by CA inhibition followed by a gradual recovery mediated by noncatalyzed cellular CO2 hydration, perhaps fueled by an increased cellular partial pressure of CO2. An increase in partial pressure of CO2 might be expected in a situation with sustained metabolic CO2 production and without CA activity. What appears to be a complete inhibition of HCO3− secretion in the first 10 min after etoxzolamide addition is caused in part by a slight acidification of the buffer-free mucosal saline in response to drug addition. However, luminal pH of 7.80 was fully recovered in all preparations after 10–15 min of drug addition, when a 50% reduction in luminal HCO3− secretion rates persisted. This observation is in agreement with previous reports of reduced, but not fully inhibited, intestinal HCO3− secretion by marine teleosts in the presence of CA inhibitors (8, 44) and with the strong physical and functional association between CA and anion exchangers in mammals (35–37).
The H+ arising from hydration of CO2 must be extruded from the epithelial cells to maintain intracellular pH, and furthermore, whereas HCO3− is excreted across the apical membrane, the H+ must be extruded preferentially across the basolateral membrane, because the intestinal epithelium exhibits strong net base secretion. This polarization of HCO3− and H+ secretion is demonstrated for the toadfish intestinal epithelium by the reduced HCO3− secretion when serosal H+ concentration is increased, presumably leading to reduced basolateral H+ extrusion, and by direct measurements of basolateral H+ secretion. A similar polarization of HCO3− and H+ secretion is characteristic of the pancreatic duct, which secretes HCO3− to reach concentrations of 70–140 mM (38), values similar to HCO3− concentrations in marine teleost intestinal fluids (14, 25, 26, 46).
Manipulation of the H+ gradient across the basolateral membrane by reduction of serosal pH revealed a concentration-dependent inhibition of luminal HCO3− secretion that appeared to be at least partly reversible within 60 min after return to control conditions. Serosal pH as low as 6.6 did not influence epithelial integrity, as indicated by unaltered TEP and conductance. A likely explanation for these observations is that basolateral H+ extrusion is reduced by the increased serosal H+ concentrations and that this in turn slows the cellular CO2 hydration and, thus, depletes the cytosolic HCO3− available for apical anion exchange. Direct evidence for a full polarization of HCO3− and H+ secretion come from measurements of H+ secretion to serosal fluids, which match those of the luminal HCO3− secretion. The basolateral H+ secretion is important for the secondary active apical HCO3− secretion and must be carrier mediated, because it occurs against an electrochemical gradient. Under serosal HCO3−-free conditions, the involvement of a basolateral Na+-dependent Cl−/HCO3− exchanger (5) in H+ extrusion can be excluded, leaving NHE and H+-ATPase as two possible mechanisms of basolateral H+ excretion. The marked reduction in apical HCO3− secretion in response to lowered serosal Na+ concentrations clearly indicates the involvement of an Na+-dependent H+ extrusion mechanism in HCO3− secretion by the marine teleost intestine and is consistent with transport mechanisms involved in rat pancreatic duct HCO3− secretion (29, 38). However, although the dependence of luminal HCO3− secretion on serosal Na+ strongly suggests the involvement of a basolateral NHE-like protein, this could not be confirmed pharmacologically, because neither serosal amiloride nor EIPA at 10−3 M influenced luminal HCO3− secretion. The inhibitor concentrations should be sufficient to inhibit H+ extrusion via EIPA-sensitive NHE isoforms (4), but the lack of amiloride and EIPA effects does not necessarily exclude the involvement of NHE, inasmuch as certain NHE isoforms or spliced variants of NHE isoforms are EIPA insensitive (2, 47).
Serosal addition of the Na+-K+-ATPase inhibitor ouabain (10−3 M) results in an ∼70% inhibition of luminal HCO3− secretion after 120 min. The reduced HCO3− secretion is likely related to reduced basolateral H+ extrusion via NHE resulting from the ouabain-induced reduction in basolateral electrochemical Na+ gradient.
The basolateral H+ extrusion is necessary for apical HCO3− secretion, as discussed above, and this basolateral H+ extrusion seems to occur via an EIPA-insensitive NHE mechanism. The basolateral NHE is driven by the electrochemical Na+ gradient established by the basolateral Na+-K+-ATPase, which thereby fuels the apical, secondary active HCO3− secretion (Fig. 11).
Although luminal HCO3− secretion is reduced by reduced serosal pH, it persists, although at lower rates, even at serosal pH 6.6, and the reduction observed at serosal pH 7.0 and 6.6 is not much different. These observations may indicate the recruitment of additional H+ extrusion mechanisms in response to reduced extracellular pH.
The substrate for luminal HCO3− secretion by the gulf toadfish intestine is concluded to be a combination of endogenous epithelial CO2 hydration and serosal CO2 and/or HCO3−, with endogenous CO2 hydration accounting for ∼50% of the total HCO3− secretion rates under resting, in vivo-like conditions. The present study demonstrates that the polarized extrusion of H+ and HCO3− (arising from cellular CO2 hydration) accounts for 50% of the net epithelial secretion of HCO3−. The nature of the remaining 50% of the overall HCO3− secretion, which relies on serosal HCO3− and, possibly, CO2, offers an interesting area for further studies. An Na+-HCO3− cotransporter is a likely candidate for basolateral HCO3− uptake by intestinal epithelial cells (Fig. 11) and could provide for apical anion exchange, as in the case of the rat, pig, and guinea pig pancreatic ducts (38), but this possibility remains to be investigated.
High rates of intestinal anion exchange represent a significant contribution to marine teleost osmoregulation and, thus, can be expected to be regulated, depending on salt and water balance. In addition, intestinal HCO3− secretion in fish likely serves the same functions as in mammals, where duodenal HCO3− secretion protects against acidic gastric effluent (6) and can be potently stimulated by multiple factors. The characteristics of the gulf toadfish intestinal HCO3− secretion described in the present study apply to unstimulated, resting epithelia, and HCO3− secretion by stimulated epithelia may display different characteristics. The involvement of an H+-ATPase in H+ extrusion and a more significant contribution of transepithelial HCO3− transport to overall luminal HCO3− secretion may apply to stimulated intestinal epithelia.
In a recent study, it was suggested that basolateral H+ extrusion from marine teleost intestine may account for a substantial fraction of intestinal H+ absorption and that the H+ secretion from the epithelium across the basolateral membrane could be as high as 0.3 μmol·cm−2·h−1 (15). In the present study, this suggestion was verified by direct measurement of basolateral H+ extrusion at ∼0.3 μmol·cm−2·h−1, which, combined with direct measurements of water absorption, yields a theoretical H+ concentration of 76.5 mM in the absorbed fluid. From this calculation and previous reports (15), it is clear that intestinal fluid absorption by marine teleosts is accompanied by acid absorption (equivalent to a theoretical pH of 1.1 in the absorbed fluid with the assumption of no buffer capacity) and that maintenance of systemic acid-base balance requires compensation for the acid load associated with vital intestinal water absorption.
Interestingly, elevated extraintestinal excretion of acidic equivalents when intestinal base secretion is stimulated has been reported recently (45), illustrating this intimate link between osmoregulation and acid-base balance.
The marine teleost intestine, which is capable of secondary active HCO3− secretion, resulting in luminal concentrations >100 mM in some cases, displays polarized apical HCO3− and basolateral H+ secretion arising from endogenous CO2 hydration. CA-catalyzed CO2 hydration provides significantly for the apical anion exchange, whereas the electrochemical Na+ gradient fuels basolateral H+ extrusion via an EIPA-insensitive NHE mechanism. Basolateral H+ extrusion is required for continued secondary active HCO3− secretion, and the energy for this transport process is ultimately supplied from the basolateral Na+-K+-ATPase, as illustrated by ouabain-sensitive HCO3− secretion. Under nonstimulated conditions, transepithelial HCO3− transport, possibly via Na+-HCO3− cotransporter-like proteins, appears to contribute ∼50% to overall secretion rates. The intestinal apical anion exchange contributes significantly to Cl− and, thereby, water absorption across marine teleost intestine, and a consequence of the CO2 hydration required for this transport pathway is a highly acidic absorbate. If we consider that the immense diversity among marine teleosts and possible diversity in epithelial transport mechanisms, it is likely that the macroscopic intestinal epithelium may provide useful models for mammalian HCO3− secreting and/or fluid-absorbing epithelia.
This work was supported by National Science Foundation Grant 0416440 to M. Grosell.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society