We hypothesized that duodenal HCO3− secretion alkalinizes the microclimate surrounding intestinal alkaline phosphatase (IAP), increasing its activity. We measured AP activity in rat duodenum in situ in frozen sections with the fluorogenic substrate ELF-97 phosphate and measured duodenal HCO3− secretion with a pH-stat in perfused duodenal loops. We examined the effects of the IAP inhibitors l-cysteine or l-phenylalanine (0.1–10 mM) or the tissue nonspecific AP inhibitor levamisole (0.1–10 mM) on AP activity in vitro and on acid-induced duodenal HCO3− secretion in vivo. AP activity was the highest in the duodenal brush border, decreasing longitudinally to the large intestine with no activity in stomach. Villous surface AP activity measured in vivo was enhanced by PGE2 intravenously and inhibited by luminal l-cysteine. Furthermore, incubation with a pH 2.2 solution reduced AP activity in vivo, whereas pretreatment with the cystic fibrosis transmembrane regulator (CFTR) inhibitor CFTRinh-172 abolished AP activity at pH 2.2. l-Cysteine and l-phenylalanine enhanced acid-augmented duodenal HCO3− secretion. The nonselective P2 receptor antagonist suramin (1 mM) reduced acid-induced HCO3− secretion. Moreover, l-cysteine or the competitive AP inhibitor glycerol phosphate (10 mM) increased HCO3− secretion, inhibited by suramin. In conclusion, enhancement of the duodenal HCO3− secretory rate increased AP activity, whereas inhibition of AP activity increased the HCO3− secretory rate. These data support our hypothesis that HCO3− secretion increases AP activity by increasing local pH at its catalytic site and that AP hydrolyzes endogenous luminal phosphates, presumably ATP, which increases HCO3− secretion via activation of P2 receptors.
- brush-border membrane
- ELF-97 phosphate
the duodenal brush-border membrane (BBM) has multiple highly expressed ectoenzymes with extracellular catalytic sites, including membrane-bound carbonic anhydrase (CA) and intestinal alkaline phosphatase (IAP). These ectoenzymes are mostly zinc metalloenzymes tethered to the BBM with a transmembrane domain or glycosylphosphatidyl inositol anchor (44, 69). Ectoenzyme activities are highest in the proximal duodenum along the proximal-caudal axis (30, 58). Since HCO3− secretion is frequently invoked as a primary duodenal defense mechanism against concentrated gastric acid (8), the high expression of CA and IAP in duodenum might be involved in protective HCO3− secretion. Indeed, the ecto-CA and intracellular CA expressed in duodenal villous enterocytes play important roles in coordinating the protective response to luminal acid (3–5, 7, 26, 43).
Among three isoenzymes of AP, placental AP, IAP, and liver/bone/kidney tissue-nonspecific AP (TNAP), IAP has the highest specific catalytic activity, especially in duodenum and jejunum (13). IAP has long been used as a marker of the intestinal brush border, with activity expressed predominantly in villous tip cells (48, 53). Localization has been aided by in situ quantitative measurement of AP kinetic activity performed in cryostat sections of rat duodenum (10, 45) and in isolated rat duodenocytes (45) using a microdensitometer-based histochemical method.
IAP function remains somewhat speculative, particularly with regard to its apparently nonphysiological pH optimum. Furthermore, IAP knockout mice have no overt intestinal phenotype. Nevertheless, enhanced fat absorption is observed after fat loading in IAP knockout mice (46), which, combined with prior studies of the role of IAP in intestinal fat absorption, suggests that IAP activity helps regulate intestinal lipid absorption and surfactant-like protein (SLP) secretion (19, 20, 61), in addition to hydrolyzing ingested organic phosphates.
The unstirred layer overlying the duodenal brush border has an alkaline pH disequilibrium with the bulk solution, thought to result from active HCO3− secretion, with a microclimate pH stabilized by the mucus gel layer (9, 24, 39, 51). These studies, however, were performed using pH electrodes, which were likely unable to measure the predicted <5-μm-thick microclimate bathing the brush-border ectoenzymes (21). One of the purposes of the study was thus to test the hypothesis that HCO3− secretion, by alkalinizing the microclimate, enhances AP activity by exposing the catalytic site to an environment nearer to its pH optimum.
HCO3−-secreting organs such as the duodenum, bile duct, pancreatic duct, the airway, and vas deferens coexpress cystic fibrosis transmembrane conductance regulator (CFTR) and AP (17, 23, 31, 33, 36, 56). AP activity correlates well with the presence of CFTR-dependent electrogenic HCO3− secretion (57), with the presence of apical P2Y purinergic receptors, and with ATP secreted into the lumen, which then increases the HCO3− secretory rate (14, 38, 65). On the basis of the presence of this luminal ATP-based signaling system and the known HCO3−-ATPase activity of IAP, we hypothesized that IAP activity may help regulate duodenal HCO3− secretion by hydrolyzing luminal ATP.
Only a few studies have measured IAP activity in situ in an intact preparation (10, 45). Furthermore, because of methodological concerns, high-resolution localization and kinetic assays have not been studied in duodenal BBM, in vitro or in vivo. Enzyme-labeled fluorescence (ELF)-97 is a unique, fluorogenic phosphatase substrate in which water-soluble ELF phosphate is hydrolyzed to water-insoluble ELF alcohol, which generates strong, stable green fluorescence (11, 50), enabling localization and characterization of phosphatase activity with the use of quantitative fluorometry in several organs and species (18, 25, 47, 60). At alkaline pH, ELF fluorescence provides a high-resolution image of BBM IAP activity, as shown in the zebrafish intestine (18). We used this combination of high resolution and sensitivity to measure AP kinetics in situ in frozen sections and also in the duodenal brush border of living rats to test the hypothesis that duodenal brush border AP activity is increased during active HCO3− secretion and that this increased AP activity diminishes HCO3− secretion by hydrolyzing a luminal purine phosphate such as ATP. In this article, we report the first measurement of IAP activity measured in vivo, the first use of ELF to measure IAP activity in intact intestine, and the first observations testing the novel hypothesis that IAP activity is positively regulated by the HCO3− secretory rate and that IAP activity inversely regulates the HCO3− secretory rate through presumed hydrolysis of luminal purine phosphates.
Chemicals and animals.
ELF-97 phosphate, propidium iodide (PI), and 5(6)-chloromethyl SNARF-1 acetate were obtained from Molecular Probes (Invitrogen, Carlsbad, CA). l-Cysteine, l-phenylalanine, d-cysteine, d-phenylalanine, glycerol phosphate, levamisole, prostaglandin E2 (PGE2), suramin, HEPES, and other chemicals were obtained from Sigma Chemical (St. Louis, MO). CFTRinh-172 was synthesized by Dr. Samedy Ouk (Department of Chemistry, University of California, Los Angeles) and purified with high-performance liquid chromatography, with the structure verified using nuclear magnetic resonance (6). Tris-buffered saline solution (TBS) contained (in mM) 135 NaCl and 50 Tris·HCl at pH 7.0, 8.0, 8.5, 9.0, or 10.0. Krebs solution contained (in mM) 136 NaCl, 2.6 KCl, 1.8 CaCl2, and 10 HEPES at pH 7.0. All studies were performed with the approval of the Veterans Affairs Institutional Animal Care and Use Committee. Male Sprague-Dawley rats weighing 200–250 g (Harlan, San Diego, CA) were fasted overnight but had free access to water.
ELF-based AP assay on frozen sections.
After terminal exsanguination under pentobarbital sodium anesthesia, the stomach, duodenum, jejunum, ileum, and colon were immediately removed, cut into the longitudinal strips with a sharp blade, mounted in OCT compound, frozen at −20°C, and cut on a cryostat (Leica Microsystems, Wetzlar, Germany) at 8-μm thickness, based on prior studies documenting a linear relationship between AP activity and cryostat section thickness up to 8–9 μm (10, 45). The segment between the pylorus and papilla of Vater was defined as proximal duodenum (PD); from the papilla of Vater to the ligament of Treitz was defined as distal duodenum (DD); the 5-cm segment distal to the ligament of Treitz was defined as jejunum (J); the 5-cm segment proximal to the cecum was defined as ileum (I); and the 5-cm segment distal to cecum was defined as colon (C). Sections were mounted on silanized nonfluorescent glass slides (Dako North America, Carpinteria, CA) and surrounded by a delimiting pen. Sections were prestained with PI (1 μM) for 5 min to provide a fluorescent counterstain in the tissue plane. Sections were examined with a Zeiss microscope initially using 535-nm excitation and 590-nm emission with narrow-bandpass interference filters (Chroma Technology, Brattleboro, VT) to visualize PI. ELF alcohol, the fluorescent phosphatase product, was visualized with 365-nm excitation, 390-nm dichroic, and 515-nm emission filters (Chroma) with a ×10 objective. Images were recorded with a cooled charge-coupled device video camera (Hamamatsu Orca-EN; Hamamatsu USA, Bridgewater, NJ) and captured and digitized using an Apple G4 microcomputer and an image analyzer software (OpenLab; Improvision, Lexington, MA).
The ELF phosphate solution (originally 5 mM) was diluted to 0.1, 0.167, 0.5, 1, and 2.5 mM with TBS solution. After PI prestaining followed by a TBS rinse, villi were identified using PI nuclear fluorescence. The wavelength was then shifted to measure ELF fluorescence, which had a very low background due to minimal overlap with PI fluorescence. At time 0, the section was covered with 100 μl of ELF-TBS solution. Images were captured and recorded every 15 s for up to 5 min by using automated time-lapse recording and analyzed by selecting three areas of interest in the upper villous region (the upper one-third of villi), middle villous region (the middle one-third of villi), and mucus layer, which were followed throughout the experiment. The fluorescent intensity at time 0 was defined as background due to dark current signal, which was subtracted from each time-point intensity. Mean fluorescence intensity of the three selected areas was defined as the value of each section. Each data point was given from 2 sections from each of 6 rats, for a total of 36 areas in 6 rats (n = 6). All reactions were performed at room temperature. The initial velocity (Vint), calculated from the plotted time-fluorescence intensity curve, was for comparative calculations of AP activity. A plot of ELF concentration vs. Vint was analyzed using GraphPad Prism software (San Diego, CA) to calculate Km and Vmax. To examine the optimal pH of BBM AP, TBS solution containing ELF (167 μM) at pH 7.0, 8.0, 8.5, 9.0, or 10.0 was reacted on the duodenal sections and BBM AP activity was calculated as described above.
To examine the effect of known AP inhibitors on in situ AP activity measured with ELF fluorescence, duodenal frozen sections were incubated with l-cysteine, l-phenylalanine, or levamisole in TBS containing ELF phosphate (167 μM). In conventional assays of catalytic activity, l-cysteine and l-phenylalanine inhibit IAP at ∼10 mM (10, 35, 45, 54), whereas levamisole only inhibits TNAP at ∼1 mM and has little effect on IAP (10, 54). Each inhibitor at 0.1, 1, or 10 mM was reacted on the sections with ELF (167 μM). Furthermore, the effect of enantiomer d-cysteine or d-phenylalanine, or a competitive AP inhibitor, glycerol phosphate, on ELF-based AP activity was also examined.
Measurement of ELF-based AP activity in duodenum in vivo.
In vivo visualization of BBM AP activity in rat duodenum using ELF phosphate solution was performed using the method of in vivo fluorescent microscopy as previously described (3, 7). The exposed, chambered duodenal mucosa was first incubated with 5(6)-chloromethyl SNARF-1 acetate (20 μM) in pH 7.0 Krebs buffer for 15 min to load, visualize, and focus on the upper villous cells by excitation at 488 nm and emission at 640 nm through narrow-bandpass filters (Chroma). After the stabilization was perfused with pH 7.0 Krebs buffer for 30 min, the chambered mucosa was gently rinsed with normal saline and incubated with ELF phosphate (167 μM) in normal saline (since buffer solution would mask the microclimate pH regulation by the basal and stimulated HCO3− secretion) for 5 min. In some experiments, the mucus gel layer loosely present over the mucosa was gently removed by suctioning, using a PE-50 catheter with a syringe as previously reported (52), before ELF application for the direct contact of ELF to the villous surface.
After the villous cells were delineated by using SNARF red fluorescence, green fluorescent images of the microscopically observed chambered segment of duodenal mucosa at 365-nm excitation and 515-nm emission were recorded, captured, and digitized every 15 s as described above. The images were analyzed by selecting each of three areas of interest in the upper villous surface, which were followed throughout the experiment. Mean fluorescence intensity of the three selected areas was defined as the value of each time point. To examine the effect of the stimulated HCO3− secretion on ELF-based AP activity, PGE2 (0.3 mg/kg), a well-known HCO3− secretagogue in duodenum (1, 59), was intravenously injected 5 min before ELF application. Furthermore, to examine the effect of IAP inhibitors on AP activity measured in vivo, l-cysteine (10 mM in saline followed by pH adjusted at pH 7.0, since pH of l-cysteine is ∼5.5) was coincubated with the ELF solution. Since AP activity is pH dependent, we measured AP activity when the mucosa was incubated with pH 2.2 saline containing ELF. Furthermore, since activity of the CFTR is needed for stimulated HCO3− secretion (6, 15, 32), some animals were pretreated with the potent selective CFTR inhibitor CFTRinh-172 (1 mg/kg ip) 1 h before the experiments. Pretreatment with CFTRinh-172 eliminates acid-induced HCO3− secretion in rat duodenum (6).
Effects of AP inhibition on duodenal HCO3− secretion.
Duodenal loops were prepared and perfused as previously described (2, 6, 26, 43). In brief, after animal preparation under isoflurane anesthesia (1.5–2.0%) as described above, the abdomen was incised, both stomach and duodenum were exposed, and the forestomach wall was incised 0.5 cm using a thermal cautery. A polyethylene tube (diameter 5 mm) was inserted through the incision until it was 0.5 cm caudal from the pyloric ring, where it was secured with a nylon ligature. The distal duodenum was ligated proximal to the ligament of Treitz before the duodenal loop was filled with 1 ml of saline prewarmed at 37°C. The distal duodenum was then incised, and another polyethylene tube was inserted through the incision and sutured into place. To prevent contamination of the perfusate from bile or pancreatic juice, the pancreaticobiliary duct was ligated just proximal to its insertion into the duodenal wall and cannulated with a PE-10 tube to drain the juice. The resultant closed proximal duodenal loop (perfused length 2 cm) was perfused with prewarmed saline by using a peristaltic pump (Fisher Scientific, Pittsburgh, PA) at 1 ml/min. The saline perfusate and effluent were circulated through a reservoir in which the perfusate was bubbled with 100% O2 and stirred and warmed at 37°C with a heating stirrer (Barnstead International, Dubuque, IA). The pH of the perfusate was kept constant at pH 7.0 with a pH stat (models PHM290 and ABU901; Radiometer Analytical, Lyon, France). Furthermore, to eliminate the buffer action of inhibitors, which would over- or underestimate the titration volume using pH-stat, two sets of flow-through pH and CO2 electrodes were connected in the circulation, where pH and [CO2] were simultaneously and continuously measured to calculate the total CO2 output equivalent to the secreted HCO3− as previously described (43). After stabilization with continuous perfusion of pH 7.0 saline for ∼30 min, the time was set as time 0. The duodenal loop was perfused with pH 7.0 saline from time 0 (t = 0 min) until t = 10 min (basal period). The perfusate was then changed to pH 7.0 or pH 2.2 acid saline from t = 10 min until t = 20 min (challenge period), with or without inhibitors (described below). At t = 10 and 20 min, the system was gently flushed to rapidly change the composition of the perfusate. During the challenge period, the solution was perfused with a syringe pump, followed by the perfusion of pH 7.0 saline with a peristaltic pump through a reservoir from t = 20 min to t = 45 min (recovery period).
To examine the effect of the inhibition of AP on duodenal HCO3− secretion, we perfused the duodenal loop with l-cysteine, l-phenylalanine, or levamisole (10 mM) dissolved in pH 2.2 saline solution during the challenge period. To test the specificity of amino acid IAP inhibitors, we examined the effect of the enantiomers d-cysteine or d-phenylalanine. We also tested the effect of the organophosphate competitive AP inhibitor glycerol phosphate, which has been used clinically to inhibit IAP activity (42).
To test our hypothesis that IAP hydrolyzes luminally released endogenous phosphate compounds such as ATP, which then activates P2 receptors on the enterocyte apical membrane to augment duodenal HCO3− secretion, we measured the effect of perfusion of the nonselective P2 receptor antagonist suramin (1 mM) (40, 62) on acid-induced HCO3− secretion. In addition, to associate AP inhibition with P2 receptor-related HCO3− secretion, we perfused l-cysteine or glycerol phosphate (10 mM) with or without suramin (1 mM). Since the high concentrations of the inhibitors altered medium buffering, we used flow-through pH and CO2 electrodes to measure HCO3− secretory rates.
All data are means ± SE. Data for in vitro study were derived from two sections from each of three rats (total n = 6). Data for in vivo study were from six rats in each group. Comparisons between groups were made by one-way ANOVA followed by Fischer's least significant difference test. P values of 0.05 were taken as significant.
Kinetics of ELF-based AP activity measured in duodenal frozen sections.
The time course of ELF fluorescence at pH 8.5 on the duodenal frozen sections was examined to localize and analyze AP activity at room temperature. ELF-positive fluorescence was recognized on the BBM of the whole villous cells and in the mucus gel as a time-dependent increase of fluorescence intensity, which appeared mostly on the BBM, with minor fluorescence of the cytoplasm and basolateral membranes (Fig. 1, A–D). With the ×10 objective, ELF staining on the BBM was distinguishable from intracellular staining by their morphological localization. To minimize contributions from overlap from adjoining villi, we analyzed only the upper and middle thirds of the villous epithelial cells and overlying mucus. In the upper and middle villous cells and in the mucus gel, fluorescent intensity increased time and substrate concentration dependently (Fig. 1, E–G). The calculated initial velocities of fluorescence intensity (FI) increase (Vint, FI/s), defined as relative AP activity, of the upper and middle villous BBM and the mucus gel are shown in Fig. 1H. AP activities of the upper and the middle villous BBM were similar, and the mucus AP activity was less than BBM AP activity. Calculated Km and Vmax values were 0.99 ± 0.18 mM and 61.4 ± 4.9 FI/s in the upper villous BBM, 0.81 ± 0.17 mM and 58.9 ± 5.0 FI/s in the middle villous BBM, and 0.45 ± 0.18 mM and 13.5 ± 1.7 FI/s in the mucus gel, respectively.
Since there was no significant difference among pH 7.0, 8.0, and 8.5, the pH optimum for BBM AP activity was 7.0–8.5 and the pH maximum was ∼8.5 in the upper and middle villous segments (Fig. 2). Activity steeply declined at pH > 9.0, similar to published observations of AP activity measured in cryostat sections (10, 45).
Longitudinal axis of ELF fluorescence in the rat gastrointestinal tract.
ELF-based AP activity at pH 8.5 was measured in frozen sections of the crypt-villous axis and in the longitudinal axis from stomach to colon. In stomach, no ELF fluorescence in the mucosal cells of the fundus (Fig. 3A) or antrum (Fig. 3B), whereas the antral lumen was occasionally positive in the prepyloric area, probably due to AP trapped in mucus gel migrating from the duodenum. In the proximal duodenum (Fig. 3, C and D), marked ELF fluorescence was observed on the BBM of the villous enterocytes, whereas the crypt cells and Brunner's glands were negative, consistent with previous studies (10, 45, 48). The overlying mucus gel was weakly positive. Distal duodenum (Fig. 3, E and F) and jejunum (Fig. 3, G and H) displayed a similar fluorescence pattern. In contrast, the interstitium of the villi and of the crypts was diffusely stained in the ileum, presumably reflecting SLP migration to the lymphatic ducts (19, 20, 61), with faint fluorescence observed in the BBM and mucus gel (Fig. 3, I and J). Colonic mucosa had no fluorescence (Fig. 3K). Figure 3L depicts the longitudinal axis of ELF-based BBM AP activity (or apical membrane AP activity in the stomach), which was highest in the proximal and distal duodenum and declined in the jejunum, but low in the fundus, antrum, ileum, and colon, consistent with previous studies (13, 29, 55, 68).
Effect of AP inhibitors on ELF-based AP activity in vitro.
l-Cysteine concentration-dependently inhibited AP activity measured at pH 8.5 on the BBM and in the mucus gel (Fig. 4A), consistent with previous studies (10, 45). The enantiomer d-cysteine also abolished AP activity at 10 mM, indicating nonstereospecificity of this inhibition. l-Phenylalanine inhibited BBM AP activity maximally at 1 mM (Fig. 4B). In contrast, its enantiomer, d-phenylalanine, had no effect (Fig. 4C), confirming prior reports of its specificity (10). Levamisole, a TNAP inhibitor, inhibited BBM AP activity in the upper villus but was less effective in the middle villus and had no effect on AP activity in the mucus gel (Fig. 4D), suggesting that AP activity present in the mucus gel may be IAP cleaved or secreted from the villous cell BBM (19, 20, 61). Glycerol phosphate, a substrate of AP used as a competitive inhibitor for the ELF reaction, concentration-dependently inhibited all AP activities (Fig. 4E).
ELF fluorescence of rat duodenum in vivo.
To examine BBM AP activity in vivo, we incubated the perfused, chambered duodenal mucosa of living rats with ELF. ELF heterogeneously stained the mucus gel (Fig. 5A) with little fluorescence observed in the villi, probably due to trapping or consumption of the ELF compound by the mucus gel. Note that despite the presence of the mucus gel layer, SNARF successfully stained the underlying villous cells. After the gel layer was removed by gentle suction, ELF application successfully stained the BBM of the duodenal mucosa (Fig. 5B) in a pattern consistent with extracellular, and not intracellular, fluorescence as previously demonstrated with BCECF (7). Surface ELF fluorescence intensity was linearly increased up to 2 min (Fig. 5C). Since AP activity is pH dependent, we predicted that augmentation of the HCO3− secretory rate would increase AP activity, presumably by increasing the local pH at its catalytic site. We increased the HCO3− secretory rate by injecting PGE2 intravenously before ELF application. Since PGE2 also rapidly increases mucus secretion (5), the mucus gel was removed before ELF application. PGE2 increased the fluorescence intensity more rapidly than did the saline control, reaching a plateau at 1 min (Fig. 5C). In contrast, coincubation of the IAP inhibitor l-cysteine (10 mM) with ELF markedly decreased the fluorescence intensity, confirming the specificity of ELF fluorescence for IAP activity. Villous surface AP activity was enhanced by PGE2 intravenous injection but inhibited by luminal l-cysteine (Fig. 5D), suggesting that ELF fluorescence is applicable to the duodenum in vivo. Although PGE2 was injected intravenously, to eliminate the possible topical effect of PGE2 on BBM AP activity, we confirmed that PGE2 (1 mg/ml) added to the frozen section had no effect on ELF-based BBM AP activity and kinetics in vitro.
We then examined the effect of luminal pH and CFTR inhibitor on AP activity measured in vivo. We incubated the mucosa with acid (pH 2.2) containing ELF (167 μM) in rats pretreated with vehicle or with the selective CFTR inhibitor CFTRinh-172 (1 mg/kg ip). Acid (pH 2.2) reduced villous surface AP activity by ∼60% of control, whereas AP activity measured at pH 2.2 was abolished in CFTRinh-172-pretreated rats (Fig. 5E). Note that CFTRinh-172 treatment itself had no effect on the basal AP activity, consistent with our earlier observation that CFTRinh-172 inhibits only stimulated HCO3− secretion (6).
Effect of AP inhibitors on duodenal HCO3− secretion in vivo.
As previously reported (1, 2, 6), perfusion with an acid solution (pH 2.2) increased duodenal HCO3− secretion during the post acid-stress recovery period (Fig. 6, A–C). As determined using pH-stat measurements, l-cysteine (10 mM) coperfused with the acid solution enhanced (Fig. 6A) and l-phenylalanine (10 mM) partially augmented acid-induced duodenal HCO3− secretion (Fig. 6B), whereas d-phenylalanine (10 mM) had no additive effect, again confirming the stereospecific effect of l-phenylalanine. Since an amino acid solution might affect titratable alkalinity measured using the pH-stat method, total CO2 output measured by flow-through pH and CO2 electrodes also confirmed the inhibitory effect of l-cysteine (data not shown). In contrast, the TNAP inhibitor levamisole (10 mM) had no effect on acid-induced duodenal HCO3− secretion (Fig. 6A). Furthermore, the nonselective P2 receptor antagonist suramin (1 mM) reduced acid-induced HCO3− secretion (Fig. 6C). Since incubation with luminal acid reduced villous surface AP activity measured in vivo (Fig. 5E), this result suggests that acid-induced duodenal HCO3− secretion involves AP-related activation of P2 receptors.
To further test whether IAP inhibition increases HCO3− secretion via P2 receptor signaling, we examined two distinct AP inhibitors, a noncompetitive IAP inhibitor, l-cysteine (10 mM), and a competitive AP inhibitor, glycerol phosphate (10 mM), with or without suramin (1 mM). Since l-cysteine is acidic and glycerol phosphate is alkaline, l-cysteine or glycerol phosphate was perfused with Krebs buffer adjusted to pH 7.0, with HCO3− secretion measured as total CO2 output using pH and CO2 electrodes. Krebs buffer (pH 7.0) had no effect on basal total CO2 output, whereas l-cysteine (Fig. 7A) and glycerol phosphate (Fig. 7B) increased total CO2 output. l-Cysteine-induced HCO3− secretion was reduced and glycerol phosphate-induced HCO3− secretion was abolished by the coperfusion of suramin (Fig. 7, A and B), suggesting that AP inhibition-induced HCO3− secretion is mediated via P2 receptor activation in rat duodenum.
We have demonstrated that detection and kinetics analysis of BBM AP activity in situ in intact rat duodenum and other segments of the gastrointestinal tract using ELF fluorescence at alkaline pH is a useful, convenient, and quick method for quantifying AP activity. This method, compared with published reports using biochemical or histochemical techniques, has the advantages of ease of use, high resolution, and continuous measurement, with little ambiguity regarding tissue localization (10, 45). Furthermore, the method can be applied to the mucosa of living rats, enabling us to examine the physiological role of AP activity in intact tissue in living animals. This is the first study to measure IAP activity in vivo in situ in rat duodenal mucosa. Using this technique, we correlated AP activity with the HCO3− secretory rate, yielding the novel observation that augmentation of HCO3− secretion enhances AP activity, whereas inhibition of HCO3− secretion suppresses activity, supporting our hypothesis that IAP activity is affected by the pH at its catalytic site, which in turn is dependent on the HCO3− secretory rate. IAP inhibition enhanced acid-augmented HCO3− secretion, consistent with the hypothesis that AP hydrolyzes a stimulatory luminal phosphate. Since the nonselective P2 receptor antagonist suramin reduced HCO3− secretion induced by luminal acid and by AP inhibitors, the stimulatory luminal phosphate is likely a purine, probably ATP, in accord with other extracellular purinergic-dependent HCO3− secreting epithelia (49, 65, 69).
The high intestinal expression of IAP combined with its apparently nonphysiological pH optimum has long intrigued investigators. IAP facilitates fat absorption; luminally shed IAP is converted to SLP that facilitates transcellular triacylglyceride movement (19, 20, 61). We speculate that the mucus AP activity that we observed is related to SLP secretion into the mucus. Nevertheless, in IAP knockout mice under fat loading, fat absorption was accelerated, suggesting that IAP may also negatively regulate enterocyte fat absorption (46). IAP also has anion-stimulated ATPase activity, known as HCO3−-ATPase, which is considered by most to be an alternative activity of IAP (16, 29, 64). The presence of HCO3− lowers the pH optimum of IAP to near neutral (34, 35). Since the duodenum is the site of high HCO3− secretory rates and the highest IAP activity in the gastrointestinal tract, IAP activity is implicated in HCO3− secretion in duodenum (57). Nevertheless, no one has previously examined the effect of IAP inhibitors on HCO3− secretion in intestine. Our study demonstrated that l-cysteine is a relatively specific IAP inhibitor, because of its inhibitory pattern resembling the distribution of IAP, that further augments stimulated duodenal HCO3− secretion indirectly via P2 receptor activation.
The Km values that we reported, 0.99 and 0.81 mM, are similar to Km values previously reported for AP activity measured in cryostat sections, 0.94 mM in duodenal BBM (45) and 0.8 mM in jejunum (28). Other publications, in which different preparations, methods, and substrates were used, have reported Km values between 0.11 and 8.2 mM (10, 13).
Some discrepancies are present between AP inhibitor profiles measured in vitro and in vivo. l-Cysteine inhibited AP activity in vitro and in vivo and enhanced acid-induced duodenal HCO3− secretion in vivo, confirming its efficacy as an IAP inhibitor. l-Phenylalanine was less inhibitory of BBM AP activity in vitro and partially enhanced the acid-induced duodenal HCO3− secretion, whereas d-phenylalanine had no effect on AP activity in vitro and HCO3− secretion in vivo, suggesting that l-phenylalanine is a weak but specific IAP inhibitor. Levamisole, a well-known TNAP inhibitor (12), inhibited BBM AP activity ∼50% in vitro but had no effect on acid-induced duodenal HCO3− secretion, suggesting that TNAP is also present on the enterocyte BBM but does not participate in the regulation of duodenal HCO3− secretion. Coexpression of AP paralogs in the intestinal brush border has been advanced as an explanation for the lack of overt phenotype in IAP null mice (46).
Our data suggest that l-cysteine is the most potent IAP inhibitor, on the basis of in vivo and in vitro studies, and should probably be used in preference to l-phenylalanine (10), although its lack of stereospecificity is somewhat problematic. One possibility is that l-cysteine directly affects duodenal HCO3− secretion through the production of H2S, synthesized by cystathionine β-synthase and cystathionine γ-lyase (22), or through spontaneous production of H2S during the reduction of l-cysteine to l-cystine, which stimulates duodenal HCO3− secretion (37) and accelerates gastric ulcer healing (63). Nevertheless, our in vitro AP inhibitory data, combined with the observation that l-cysteine-induced HCO3− secretion was reduced by P2 receptor antagonism, do not support this interpretation. AP activity inhibition by the competitive AP inhibitor glycerol phosphate, which also concentration-dependently inhibited BBM AP activity in vitro and increased HCO3− secretion via P2 receptor activation, further supports our contention that the primary action of l-cysteine was IAP inhibition. The possible involvement of H2S production related to l-cysteine must still, however, be considered when interpreting the data.
The presence of apical AP activity correlates well with the presence of CFTR-dependent electrogenic HCO3− secretion (57). HCO3−-secreting organs such as the duodenum, bile duct, pancreatic duct, airways, and vas deferens coexpress CFTR and AP (17, 23, 31, 33, 36, 56). The lack of measured gastric epithelial AP activity may relate to its CFTR-independent HCO3− secretory mechanism (66, 67).
Although ingested phosphorylated compounds likely serve as substrates for IAP ectophosphorylase, no endogenously secreted luminal substrate has been conclusively identified. Since AP hydrolyzes organic phosphates, a possible candidate is ATP. Duodenal BBM have high ATPase activity, due to the presence of ectonucleoside triphosphate diphosphohydrolase (ecto-NTDPase), 5′-nucleotidase, and IAP (57, 69). In rat duodenal BBM, 50% of ATP is hydrolyzed by ecto-ATPase and the other 50% by IAP (54). Since extracellular ATP regulates and modifies cellular function through P2Y receptors (41), which are expressed on the intestinal apical membrane (27, 65), released ATP may regulate epithelial function via activation of P2Y receptors. Indeed, almost every known mammalian HCO3−-secreting epithelium expresses apical P2Y receptors, secreting ATP into the lumen, which then increases the HCO3− secretory rate (14, 38, 65). If the ATP ecto-signaling system is also present in duodenum, we speculate that active HCO3− secretion increases the catalytic rate of AP by increasing its local pH, as we observed in our in vivo experiments, and that AP hydrolyzes luminal ATP, decreasing P2Y receptor-mediated HCO3− secretion, explaining the observed augmentation of HCO3− secretion during AP inhibition. Inhibition of acid-induced and IAP inhibition-induced duodenal HCO3− secretion by the nonselective P2 receptor antagonist suramin further supports our hypothesis.
In conclusion, we developed a simple in situ duodenal AP activity assay based on the fluorogenic substrate ELF. Stimulation of duodenal HCO3− secretion augments, whereas inhibition of HCO3− secretion suppresses IAP activity, which is consistent with the HCO3− secretory rate altering the local pH of the AP catalytic site. IAP inhibition further augments stimulated duodenal HCO3− secretion, inhibited by suramin, consistent with the presence of a luminal purinergic signaling system regulating HCO3− secretion.
This work was supported by a Department of Veterans Affairs Merit Review Award, National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Grant R01 DK54221 (to J. Kaunitz), and NIDDK Animal Core Grant P30 DK0413 (to J. E. Rozengurt).
We thank Dr. Takanari Nakano, Saitama Medical University, Japan, for helpful suggestions, and Rebecca Cho for assistance with manuscript preparation.
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 © 2007 the American Physiological Society