Maintenance and regulation of the pH microclimate at the luminal surface of the distal colon of guinea-pig
Author's present address
R. Busche: Department of Physiological Chemistry, School of Veterinary Medicine, Hannover, Germany.
W. v. Engelhardt: Department of Physiology, School of Veterinary Medicine, Hannover, Bischofsholer Damm 15/102, D-30175 Hannover, Germany. Email: firstname.lastname@example.org
The existence of a constant pH microclimate at the luminal surface of the large intestinal mucosa had been shown by various research groups in recent years. These measurements hitherto had been performed almost exclusively by means of pH-sensitive microelectrodes. In studies with the rat, guinea-pig and human colon (McNeil & Ling, 1984; Rechkemmer et al. 1986) it was demonstrated that this surface pH was relatively independent of changes in proton concentration of the bulk luminal fluid. Neither removal of sodium from the perfusion solution, nor the inhibition of Na+-H+ exchange by mucosal amiloride affected the pH microclimate in the colon of guinea-pig (Rechkemmer et al. 1986).
Such a region of relatively constant pH has a decisive influence on the absorption of weak electrolytes such as short-chain fatty acids (SCFAs) (Rechkemmer, 1991). A constant luminal pH microclimate is efficient, for instance, in preventing the rapid entry of protonated, lipid-soluble SCFA molecules into the enterocytes, when the pH in the luminal bulk phase is decreased. Lowering the pH of the luminal solution from 7.2 to 5.5 in the human colon (McNeil et al. 1978), from 7.4 to 5.4 in the rat caecum (Flemming et al. 1991), from 8.6 to 5.0 or from 7.4 to 5.6 in the guinea-pig colon (Rechkemmer & Engelhardt, 1988; Oltmer & Engelhardt, 1994) had little or no effect on SCFA absorption rates. This independence of SCFA absorption from pH can be attributed to a microclimate of constant pH at the luminal surface of the large intestine.
The reasons for the constancy of the pH microclimate at the surface of the large intestine are not known. A limiting factor for studies in this area has been the lack of an appropriate method for surface pH estimations. Measurement using microelectrodes provides several possible sources of error, including the lack of a defined position for the electrode tip and, over the course of time, the uncertainty of the location of the tip due to movement of the mucosa, especially when a change of the perfusion solution is required. Therefore, estimations of surface pH over a longer period of time are generally not possible with microelectrodes. Shiau et al. (1985) estimated the surface pH in rat jejunum with the pH indicators Kodak 5325 and Bromocresol Purple. For measuring the luminal pH in colonic crypts of mice Chu & Montrose (1995, 1996) used the pH-sensitive fluorescent dye SNARF-1.
It was the aim of our study to use a pH-sensitive dye that attaches to the luminal surface of enterocytes. With such a probe factors affecting and regulating the pH at the luminal surface of colonocytes could be studied. We used the pH-sensitive fluorescent dye 5-N-hexadecanoyl-aminofluorescein (HAF). Due to its lipophilic character the alkyl tail of this probe inserts into the membrane of the colonocytes. Hitherto, HAF had been used for investigations of lateral diffusion of lipids in biomembranes (Dragsten et al. 1981; Dupou et al. 1987; Tournier et al. 1989) and for measurements of resonance energy transfer in lipid membranes (Remmers & Neubig, 1993).
The localization of the dye molecule at the epithelial surface was demonstrated by means of a confocal laser-scanning microscope and by using IgG antibodies. With the pH-sensitive dye attached to the apical membrane the influence of bicarbonate, chloride and SCFAs on the pH microclimate were investigated. Our findings underline the importance of bicarbonate in establishing and maintaining the pH microclimate at the surface of the colonic mucosa of guinea-pig.
Animals and tissue preparation
Male guinea-pigs (body weight of 550-650 g) were fed on a pelleted standard diet (Altromin No. 3122, Altromin, Lage, Germany). Water and food were available ad libitum. The guinea-pigs were maintained on a 12 h light: 12 h dark photoperiod with two or three animals being held together in one cage. For experiments the guinea-pigs were killed by decapitation. The distal colon was cut in approximately 3 cm-long sections, and these sections were immediately put into ice-cold, carbogen (95% O2, 5% CO2)-gassed Krebs-Henseleit solution (KH; pH 7.4) after the luminal contents had been flushed off. The sections were stored in this solution until the beginning of the preparations. A section of the distal colon was cut open longitudinally at the mesenterial side and was fixed on a PVC plate by needles with the mucosal side uppermost. The mucosa was stripped off the submucosal tissue with glass slides, and was mounted, slightly stretched with its serosal side on a supporting silicone ring, with Histoacryl (Braun-Melsungen, Melsungen, Germany). The experiments were carried out according to the guidelines of our local Animal Care Committee.
Fluorescent probe for the estimation of the luminal surface pH
The pH-sensitive fluorescent dye 5-N-hexadecanoylaminofluorescein (HAF) was purchased from Molecular Probes (Eugene, OR, USA). The HAF molecule consists of two domains, the fluorescein molecule and a palmitoyl chain connected to this fluorescing moiety. Dragsten et al. (1981) have shown that this palmitoyl chain inserts into the apical membrane of cells due to its lipophilic character. The sensitivity of HAF to pH changes relies on the presence of dissociable groups in the fluorescing moiety.
A stock solution of HAF in a concentration of 30 mmol l−1 was prepared with ethanol (98%; Merck, Darmstadt, Germany) and stored at -20°C for no longer than 1 week. A final HAF concentration of 15 μmol l−1 was used to superfuse the mucosal side of the epithelium of the distal colon for 20 min. An excitation spectrum of HAF was determined with the help of a Kontron SFM 23 spectrofluorometer (Kontron, Zürich, Switzerland).
The intracellular pH (pHi) was measured using the fluorescent dye 2′,7′-biscarboxyethyl-5,6-carboxyfluorescein acetoxymethylester (BCECF AM) as described for an HT29 cell line (Busche et al. 1993).
Freeze-dried monoclonal IgG antibody (750 μg) against free and bound fluorescein (Boehringer Mannheim GmbH, Germany) was dissolved in 50 μl of Hepes buffer. Three drops of this stock solution were used to quench the fluorescence intensity of the epithelia labelled with HAF. Concentrations of 10 mmol l−1 of 1,4-dithio-L-threitol (DTT; Sigma, Deisenhofen, Germany) were used before and during dye loading to diminish the amount of mucus at the epithelial surface. During the following perfusion DTT was not present in the solutions. Nigericin was purchased from Sigma, carbonyl cyanide 4-trifluoromethoxyphenyl-hydrazone (FCCP) from Fluka (Buchs, Switzerland) and BCECF AM from Serva (Heidelberg, Germany). For the inhibition of the apical K+-H+-ATPase ouabain (4β,20-cardenolide-1β,3b,5α,14,19-hexol-3-[6′deoxy-α-L-mannopyranosyl]-G-strophanthin (Sigma)) was added to the mucosal superfusate.
For measuring the resting surface pH, the epithelium was superfused with Krebs-Henseleit (KH) solution containing (in mmol l−1): 140 Na+, 124 Cl−, 21 HCO3−, 5.4 K+, 2.4 HPO42−, 0.6 H2PO4−, 1.2 Mg2+, 1.2 Ca2+, 10 glucose; pH 7.4. In bicarbonate-free solutions bicarbonate was replaced by 21 mmol l−1 of N-[2-hydroxyethyl]piperazine-N′-[2-ethane sulfonic acid] (Hepes) buffer. In chloride-free solutions, chloride was replaced by gluconate. In butyrate-containing solutions 21 or 113.6 mmol l−1 butyrate was added as the sodium salt. The sodium concentration was kept constant by reducing the amount of sodium chloride as appropriate. Therefore, in the solution containing 113.6 mmol l−1 butyrate the concentration of chloride was only 10.4 mmol l−1. To adjust the pH of the bicarbonate-free Hepes solutions between 6.5 and 8 the solutions were titrated with NaOH; to obtain the required sodium concentration an appropriate amount of sodium gluconate was added. To adjust the lower pH in the bicarbonate-containing solution less bicarbonate was added (for pH 6.4, 1 mmol l−1; for pH 6.9, 4.5 mmol l−1; for 7.2, 10 mmol l−1), and the solutions were gassed with carbogen. The osmolarity of all solutions was adjusted to 300 mosmol l−1 with mannitol. All of the chemicals required for the solutions were purchased from Merck (Darmstadt, Germany). During the experimental procedure the solutions were kept in stock bottles in a water-bath at 37°C. All solutions containing bicarbonate were gassed with carbogen and bicarbonate-free solutions were gassed with O2. To ensure that the solutions had a constant pH they were heated to 37°C and gassed for at least 30 min before the experiments were started.
For calibrations nigericin (10 μg l−1) was added to the calibration buffers containing (in mmol l−1: 152 Cl−, 133.4 K+, 25 Hepes, 15 Na+, 1.8 Ca2+, 0.8 Mg2+, 0.8 SO42−). The pH of these solutions was adjusted to a defined pH in the range 6.0-8.0 by different amounts of 1 n NaOH. The osmolarity of the buffers was 300 mosmol l−1. The calibration buffers were added to both the mucosal and serosal sides of the epithelium.
The stripped epithelium fixed to the supporting ring was placed in a microperfusion chamber (Busche et al. 1993) with the basolateral side uppermost. Measurements were performed by means of an inverted microscope (Axiovert 35M; Zeiss, Oberkochen, Germany). The microperfusion chamber, consisting of two silicone rubber gaskets and two in- and outflow cannulas at each side, allowed separate perfusion of the mucosal and the serosal sides of the epithelium. A Zeiss LD40 objective (NA 0.60) with a working distance of 1.5 mm (no oil immersion) was used. The volume of each side of the chamber was 20 μl. The perfusion rate was 100 μl min−1. It could be altered by varying the height of the table on which the solutions were kept in stock bottles. For the measurements the chamber was placed onto the heatable stage of the microscope. The chamber was kept at 33°C during the experiments. Two to three surface cells were focused on for measuring the fluorescence intensity. A computer-directed, motor-driven stage allowed a parallel measurement of fluorescence intensity at three different locations within the epithelium. A minimal time interval of 15 s was necessary between two measurements. The correct position of the focus could be controlled visually using a video camera connected to the microscope.
For the fluorescence measurements a 75 W xenon lamp was used and two wavelengths (436 and 485 nm) were selected by bandpass filters for the excitation of the fluorescent dye. The excitation intensity at 485 nm was reduced to 25% by a neutral density filter to allow measurements using the same gain settings of the photomultiplier at both wavelengths. To minimize photobleaching, the intensity at both wavelengths was further reduced to 10% by a neutral density filter, and the excitation time was restricted to 200 ms for each time point using a high-speed shutter. The fluorescence emission was collected by a photomultiplier in the range 515-530 nm. The ratio of the fluorescence intensities (excitation at 436 and 485 nm) was calculated continuously. After calibration the actual surface pH (pHs) was calculated by taking the ratios of at least two defined pH values and assuming a linear relation between ratio and pH within this range (Busche et al. 1993).
Experimental procedure for measurements of surface pH (pHs)
The epithelium mounted into the microperfusion chamber was superfused with KH (carbogen gassed, 37°C), and the autofluorescence of the epithelium was determined at the excitation wavelengths of 436 and 485 nm. After an equilibrium period of 15 min the mucosal KH was replaced by the dye solution of 15 μmol l−1 HAF in KH for 20 min. Thereupon the remaining non-inserted dye was washed off with standard KH. At the end of the experiments calibration with at least two calibration buffers of different pH was performed.
Confocal laser-scanning microscopy and quenching of the fluorescence intensity with antibodies
The position of HAF after the epithelium was stained with dye was investigated using confocal laser-scanning microscopes (Bio-rad, MRC 500: Hemel Hempstead, UK; and Leica TCS NT: Benzheim, Germany) using a ×20 objective (NA 0.5) and a ×40 objective (NA 1.25), respectively. The HAF-labelled epithelium was focused through in the z-direction in micrometre steps.
For quenching the fluorescence intensity of an HAF-labelled epithelium monoclonal IgG antibodies against free and bound fluorescein (Boehringer Mannheim GmbH, Germany) were used. After mounting the stripped epithelium of guinea-pig distal colon into the microperfusion chamber and staining with HAF, the fluorescence intensity of this epithelium was measured near the isosbestic point. To diminish the influence of mucus 10 mmol l−1 of DTT was added to the mucosal perfusion solution prior to dye loading. Three drops of the antibody solution were added to the mucosal side of the epithelium. After 45 min of incubation with the antibody the fluorescence intensity was measured again.
Ussing chamber experiments
The viability of the dye-labelled epithelium was controlled with mucosal sheets of the distal colon of guinea-pig mounted in Ussing chambers as described by Engelhardt et al. (1993). Both sides of the epithelium were superfused with 10 ml of Krebs-Henseleit solution at pH 7.4 and 37°C. Transepithelial conductance (gt) and short-circuit current (Isc) were determined.
Results are presented as means and standard deviations. Differences were evaluated using Student's paired and unpaired two-sided t tests as appropriate. n corresponds to the number of epithelia that were used for the experiments.
Estimation of microclimate pH at the apical surface of colonocytes with HAF
The pH-sensitive fluorescent dye 5-N-hexadecanoylaminofluorescein (HAF) is a well-suited probe for measuring the microclimate pH at the intestinal surface of the distal colon of guinea-pig. As observed earlier by Dragsten et al. (1981) for epithelial cell lines derived from the kidney, and as confirmed in our experiments, the lipophilic palmitoyl chain of HAF inserts into the apical membrane of the colonocytes, and the pH-sensitive fluorescing moiety is thus located at the epithelial surface. By means of a confocal laser-scanning microscope it was shown that fluorescence intensity was not seen in the intracellular space or at the basolateral membrane. This implies that the dye molecule is exclusively bound to the apical membrane of the cells (Fig. 3). This corresponds to the observations by Dragsten et al. (1981). These authors reported that the length of the fatty acyl chain was responsible for the behaviour of such membrane-bound probes. With an acyl chain consisting of 16 carbon atoms (like HAF) the fluorescein moiety remains at the outer leaflet of the membrane, whereas a chain length of 12 carbon atoms causes the fluorescein moiety to dive into the lipid membrane (Dragsten et al. 1981). The localization of fluorescein at the luminal surface of the apical membrane of the colonocytes was confirmed by a 90% quenching of the fluorescence intensity after adding monoclonal IgG antibodies against fluorescein (Fig. 4). According to Ahlers et al. (1992) the antibody-fluorescein complex does not emit fluorescence radiation. In their experiments the fluorescence emission of the fluorescein molecules was subsequently quenched by addition of the antibody.
The calibration technique used in our experiments was adopted from measurements of the intracellular pH of different cell types (Thomas et al. 1979; Busche et al. 1993). To confirm that the high K+-nigericin technique is suitable for calibrating the microclimate pH measurements with HAF, another ionophore, FCCP, was used. Calibration curves obtained in our experiments using either nigericin or FCCP did not differ.
Resting microclimate pH at the luminal surface of the distal colon of guinea-pig
The microclimate pH measured with HAF at the luminal surface of the distal colonic epithelium of guinea-pig was 7.54 ± 0.14 when the pH of the luminal solution was 7.4. This pH is more alkaline than the mean microclimate pH obtained with a pH-sensitive microelectrode in the proximal colon (pH 7.1) and in the distal colon (pH 6.9) of guinea-pig (Rechkemmer et al. 1986). McNeil et al. (1987) had estimated even lower values for the rat and human distal colon. One explanation for the differing microclimate pH levels obtained with HAF and microelectrodes is probably the difference between the methods. The position of the tips of the glass electrodes cannot be defined exactly at any time of the measurement, and the position can easily change as a result of minimal movements of the epithelium. Furthermore, the surface of the epithelium of the colon is not smooth because of crescentic folds and crypts. It should also be noted that the tip might occasionally injure the apical membrane of the enterocytes. In contrast, the distance between the fluorescing moiety of HAF molecules and the epithelial surface is considerably smaller than the distance between electrode tip and epithelial surface, and this distance does not change with epithelial movements. Therefore, by using HAF, the pH at the luminal side of the apical membrane can be estimated directly at the surface.
Chu & Montrose (1995) used a soluble pH-sensitive fluorescent dye (carboxy-SNARF-1) for measuring the crypt luminal pH in the murine distal colon. They also obtained a slightly alkaline pH (about pH 7.3), and findings are similar to those we had measured with HAF. Thus using two different pH-sensitive fluorescent dyes, similar results have been observed.
Bulk luminal pH and microclimate pH
Rechkemmer et al. (1986) observed with microelectrodes a constant independent pH microclimate at the surface of the proximal and distal colon of guinea-pig that was independent of changes of pH in the bulk luminal solution. In in situ as well as in in vitro studies the surface pH remained constant when the luminal pH was changed between pH 5 and pH 8.6. A similar constant pH microclimate was reported by Shiau et al. (1985) for the rat small intestine. In our studies the pH at the surface of the colonocytes in the distal colon of guinea-pig changed by 0.56 pH units (from pH 7.54 to 6.98) when the pH in the bulk luminal solution was decreased by 1.0 pH unit (from pH 7.4 to 6.4; Fig. 9). Thus, in our studies a slight decrease in surface pH was seen when the pH of the bulk luminal solution was decreased step by step, whereas in in vitro experiments with microelectrodes (Shiau et al. 1985; Rechkemmer et al. 1986) no major changes of surface pH were observed. In in vivo experiments in rat proximal jejunum (Lucas, 1983) the surface pH changed, as we also found in the colon, when the pH in the bulk luminal solution was altered. With pH microelectrodes it seems likely that the mean pH value had been measured in the unstirred layer at the surface of the colonic epithelium. These mean values may explain the reported lower values and the missing changes when bulk luminal pH was altered.
Bicarbonate: the most important factor in maintaining a microclimate surface pH
A continuous secretion of bicarbonate at the luminal surface of the colonic epithelium (Feldman et al. 1990) might be the cause of the alkaline surface pH measured with HAF at the colonic surface (Fig. 9) and also with SNARF-1 in the colonic crypt lumen (Chu & Montrose, 1995). A similar explanation for the neutral surface pH was given by Rechkemmer (1991). Removal of bicarbonate and CO2 from the luminal side creates a bicarbonate and CO2 gradient across the cell membrane. The luminal CO2 pool would be ‘refilled’ by diffusion of CO2 from the serosal to the luminal side of the epithelium. This promotes the reaction of CO2 and H2O towards HCO3− and H+ at the luminal side, which is additionally accelerated by the high carbonic anhydrase activity in the mucin attached to the epithelial surface (Kleinke et al. 1998). As a consequence more protons would be available at the surface of the epithelium, and pHs would decrease. Similarly, Chu & Montrose (1996) have shown a decrease in pH in the lumen of mouse crypts by 0.12 units with CO2-HCO3− at the serosal side and HCO3−-free solution at the mucosal side. Moreover, due to a bicarbonate gradient HCO3− anions would be transported out of the cell by the apical Cl−-HCO3− exchanger. Bicarbonate anions would neutralize the free protons at the luminal surface of the epithelium, and thus pHs would increase. However, due to a rapid CO2 diffusion-mediated equilibrium reaction the effect of slower bicarbonate transport becomes less apparent (Fig. 10).
When bicarbonate-free solutions were added to the serosal side of the epithelium CO2 diffusion equilibrium reactions and bicarbonate secretion were directed towards the serosal side. The flux of CO2 from the mucosal to the serosal side would provoke an alkalization of the luminal surface as seen in our experiments and also in the colonic mouse crypt (Chu & Montrose, 1996). After removal of bicarbonate from the mucosal as well as the serosal side only the very small amounts of CO2 and H2O produced by intracellular metabolism are present, and thus very little bicarbonate would be available for the apical Cl−-HCO3− exchanger. This explains the lower pHs, which is similar to that when only the luminal solution was bicarbonate free. Under these bicarbonate-free conditions an alkaline microclimate pHs cannot be maintained, and when the bulk luminal pH is decreased the pHs decreases correspondingly (Fig. 9).
The findings in our experiments with chloride-free solutions support the existence of an apical Cl−-HCO3− exchange being involved in the regulation of pHs. This exchanger is obviously the main mechanism responsible for apical colonic bicarbonate secretion in the large intestine (Feldman et al. 1990; Feldman & Stephenson, 1990). The activity of the apical Cl−-HCO3− exchanger depends on the presence of luminal chloride. The transport direction of the exchanger is reversed when chloride is removed from the luminal superfusate as was shown by Chaillet et al. (1986) with the LLC-PK1 porcine renal cell line. The observed acidification of the luminal surface pH with luminal Cl−-free solution (Fig. 11A) can be explained by an inhibition of bicarbonate secretion. This results in a decrease of pHs. A chloride gradient in the opposite direction (no chloride on the serosal side) led to an increase of pHs probably due to an increased bicarbonate secretion across the apical membrane. Busche et al. (1997) have demonstrated that under these conditions (removal of chloride from the serosal side) the intracellular pH also increased. An interesting observation is that chloride has the same effect on pHs when bicarbonate is replaced by an equal amount of butyrate (Fig. 11B). This indicates that butyrate may substitute for bicarbonate in exchange with chloride in the apical membrane. This is in agreement with recent observations by Rajendran & Binder (1994), in which these authors postulate a Cl−-butyrate− exchange in apical membrane vesicles isolated from the distal colon of rats.
Proton secretion and microclimate pH
Rechkemmer et al. (1986) have speculated that the surface pH may depend on proton secretion in connection with the regulation of intracellular pH. In the distal colon of guinea-pig protons are secreted by the apical K+-H+-ATPase (Suzuki & Kaneko, 1989; Watanabe & Suzuki, 1990), not by Na+-H+ exchange. This proton secretion might affect the pH microclimate at the apical surface. However, in our experiments with epithelia from the distal colon of guinea-pig inhibition of the apical K+-H+-ATPase by addition of ouabain to the sufficiently buffered mucosal KH did not alter the basic pHs. Even after an experimental decrease of intracellular pH the pHs was not, or at least very little, affected, independently of whether or not K+-H+-ATPase was inhibited (authors’ unpublished observations). Although Abrahamse et al. (1992, 1994) have postulated a contribution of the K+-H+-ATPase to intracellular pH regulation, we conclude that the K+-H+-ATPase in the distal colon of guinea-pig may not be sufficiently active under physiological conditions to cause acidification at the luminal surface. In this respect it is important to consider that mechanisms involved in intracellular pH regulation of enterocytes in the distal colon of guinea-pig as well as in the polarized layer of colonic tumour cells (Busche et al. 1993, 1997) are located primarily in the basolateral membrane, not in the apical membrane.
In bicarbonate-free low buffer concentrations (5 mmol l−1 Hepes) of luminal solution pHs was lower than in 21 mmol l−1 of Hepes, and after the inhibition of the apical K+-H+-ATPase under these conditions pHs increased slightly (Fig. 13). Obviously under physiological conditions the luminal buffer system neutralizes the protons secreted by K+-H+-ATPase. An effect of the luminal buffering power has also been observed by Chu & Montrose (1995). They obtained an acidification of the crypt luminal pH when the buffering capacity of the solution had been reduced. The buffering capacity of normal colonic contents of guinea-pig is high ([HCO3−]≈ 21 mmol l−1, [SCFA]≈ 100 mmol l−1).
Effect of SCFA transport on pHs
In the large intestine short-chain fatty acids, mainly acetate, propionate and butyrate, are produced by anaerobic bacterial fermentation. These SCFAs are absorbed in the non-ionized and also in the ionized form (Engelhardt et al. 1993, 1994). Although at the physiological pH in colonic contents 99% of these SCFAs are present as anions, in a number of recent studies SCFA absorption is believed to occur mainly by diffusion in the protonated form (Chu & Montrose, 1996; Charney et al. 1998; Sellin & De Soignie, 1998). Other investigators postulate a predominantly ionized transport of SCFAs (Mascolo et al. 1991; Harig et al. 1996; Ritzhaupt et al. 1998a). Obviously major regional and species differences exist with regard to transport in the protonated and ionized forms (Engelhardt et al. 1993, 1994, 1995, 1997).
The alkalization of pHs obtained in our experiments after butyrate was added to the mucosal side of the epithelium (Fig. 12) confirms a net proton flux from the lumen into the colonocytes. From this information alone it cannot be distinguished whether the non-ionic diffusion or a carrier-mediated transport of SCFA in exchange with bicarbonate caused the observed changes in pHs. However, this alkalization was also observed when butyrate was added in the absence of bicarbonate, although only when the mucosal buffering capacity was reduced. pHs was acidified in bicarbonate-free solutions, on the other hand, when butyrate was added to the serosal side. Both these observations suggest non-ionic diffusion of SCFA across the apical membrane. With 21 mmol l−1 of Hepes buffer in the absence of bicarbonate butyrate applied to the luminal side did not affect pHs. That is further indirect evidence for the participation of an apical SCFA−-HCO3− exchange (Mascolo et al. 1991; Engelhardt et al. 1994, 1997; Ritzhaupt et al. 1998a). It has been shown in a number of studies that bicarbonate accumulates in the lumen of the large intestine proportionally to SCFA absorption (see Engelhardt, 1995). Such a bicarbonate gain could also explain the observed increase of pHs. The stronger alkalization of pHs in our studies in the presence of bicarbonate compared with that in bicarbonate-free media substantiates the role of bicarbonate secretion. A monocarboxylate (MCT1) transporter that may be involved in SCFA transport in the apical membrane has been recently identified (Ritzhaupt et al. 1998b).
We conclude that the fluorescent dye HAF can be used for measurements of the surface pH in the colon. The integration of the fatty acyl chain of the HAF molecule into the apical membrane of the epithelial cells allows a continuous measurement of pHs for several hours. A pH microclimate does exist at the surface of the distal colon of guinea-pig only in the presence of bicarbonate. Bicarbonate plays a major role in maintaining pHs. Our findings support the presence of a SCFA−-HCO3− exchange in the apical membrane of the colon.
This project was supported by the Deutsche Forschungsgemeinschaft, SFB 280, ‘Gastrointestinale Barriere’. The skilled technical assistance of K. Hansen and G. Becker is greatly appreciated. We would like to thank Professor Forssmann, Niedersächsiches Institut für Peptidforschung, Hannover and Professor Niemann, Department of Physiological Chemistry, Medical School, Hannover, for the use of their confocal microscopes, and J. Bartels for his support in measuring the intracellular pH.