INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE DUODENAL MUCOSA IS REGULARLY exposed to intermittent pulses of gastric acid, with luminal pH varying rapidly between 2 and 7 (23). Without protective mechanisms in place, the duodenal cells, like other cells in the upper gastrointestinal tract, are believed to irreversibly acidify in the presence of acidic luminal contents, injuring the epithelium (4, 21). With the measurement of robust epithelial HCO secretion and a neutral pH in the juxtamucosal mucus gel despite the presence of luminal acid, the "bicarbonate hypothesis" was developed, wherein HCO, secreted by the epithelial cells, completely neutralized luminal acid diffusing through the mucus gel toward the epithelium (9). Correlation of HCO secretion with mucosal protection from acid-related injury further bolstered this hypothesis (10, 12, 30).

One means of defending the mucosa against rapid shifts of pH is the phenomenon of acute adaptive protection, wherein exposure to a low concentration of acid or other substance decreases injury from a subsequent challenge with a higher concentration of the same or other substance. This phenomenon, although extensively studied in experimental gastroprotection models (13), has been investigated only twice in the duodenum (16, 17). Furthermore, both prior studies addressing this phenomenon used injury, and not alterations of defensive factors, as an endpoint. It is not known, for example, whether duodenal adaptive protective mechanisms primarily result from an enhancement of preepithelial mechanisms such as increased HCO secretion, from thickening of the overlying mucus gel, from postepithelial mechanisms such as hyperemia, or from an augmentation of intrinsic cellular mechanisms such as cellular buffering power. Furthermore, it was previously assumed that preexposure to a low concentration of injurious substance was required for inducing adaptive changes. An alternative explanation is that rapid shifts of luminal pH per se, and not the "mild-strong" exposure sequence, is of primary importance in inducing acid-related adaptive changes.

HCO secretion is the most studied duodenal defense mechanism. Despite its obvious appeal as a means of neutralizing gastric acid, most studies have addressed its measurement in the absence of luminal acid, when its acid neutralizing effect is unnecessary. One purpose of this study was therefore to measure HCO secretion during acid exposure and to determine its relative importance in mucosal defense in general and adaptive protection in particular.

Our laboratory has recently developed a technique in which the intracellular pH (pH_i) and duodenal blood flow are measured simultaneously in the rat duodenum (3). We demonstrated that a 10-min pulse of strongly acidic perfusate (pH 2.2) increased cellular buffering by a DIDS-sensitive mechanism, suggesting that rapid shifts of perfusate pH enhanced resistance to a subsequent acid challenge and, by extrapolation, injury. We hypothesized that rapid shifts of perfusate pH enhance intrinsic cellular defense mechanisms and further speculated that these pH shifts may underlie the phenomenon of acute adaptive cytoprotection. We have also established a technique for measurement of duodenal mucus gel thickness (MGT) in our system, which we showed was affected by the balance between mucus secretion and exudation (2).

We studied the effects of varying perfusate pH on duodenal mucosal defense mechanisms to test the hypothesis that adaptive changes are produced by rapid shifts of perfusate pH and that a major duodenal protective mechanism is an increase of cellular buffering power induced by the activation of epithelial ion transport.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Chemicals

Experimental Protocol for In Vivo Microscopic Study

Acid perfusion. To examine the effect of sustained luminal acid on pH_i, blood flow, and MGT, the duodenal mucosa was perfused with pH 7.0 Krebs buffer for 15 min, followed with either pH 7.0, 4.5, 3.5, or 2.2 solution for an additional 30 min (single acid challenge period), followed by a 15-min recovery period with pH 7.0 Krebs buffer.

Sequential perfusion of acids of different concentrations. Mild (pH 4.5) and strong (pH 2.2) acid concentrations were perfused over the mucosa in 15-min increments. After a 15-min perfusion with pH 7.0 Krebs buffer, we exposed the mucosa to mild  strong or strong  mild acid concentrations, with exposure at each pH lasting 15 min, followed by a 15-min recovery period at pH 7.0. Thus exposure to mild  strong acid would be as follows: pH 7.0 from t = 0-15 min (baseline); pH 4.5 from t = 15-30 min (the 1st acid challenge period); pH 2.2 from t = 30-45 min (the 2nd acid challenge period); and pH 7.0 from t = 45-60 min (recovery period). Adaptive changes are defined as differences in pH_i, MGT, and blood flow in a group in which perfusate pH is changed at 15-min intervals with respect to groups in which the same perfusate pH was held constant for 30 min.

Repeated acid exposure and effects of ion transport inhibitors. In another experimental series, two pulses of strong acid were used to provoke adaptive responses. Perfusate pH was changed from pH 7.0 for 10 min (t = 0-10 min) to pH 2.2 for 15 min (t = 10-25 min; the 1st acid challenge period), followed by pH 7.0 for 10 min (t = 25-35 min; the 1st recovery period), followed by pH 2.2 for 15 min again (t = 35-50 min; the 2nd acid challenge period), and returned to pH 7.0 for 15 min (t = 50-65 min; the 2nd recovery period). To determine the role of epithelial ion transport in regulation of pH_i, blood flow, and MGT, DMA (0.1 mM), which inhibits Na⁺/H⁺ exchange, or DIDS (0.5 mM), which inhibits Na⁺-HCO cotransport and HCO/Cl exchange, was added with the pH 2.2 perfusion during the first acid challenge period. Both inhibitors exert their effects on the epithelial cells primarily on the serosal membrane transporters (3). Adaptive changes are defined as differences in pH_i, MGT, and blood flow during the second acid challenge compared with those during the first acid challenge.

Measurement of Duodenal Loop HCO Secretion

Preparation of the duodenal loop. In a separate experiment, a duodenal loop was prepared and perfused to measure duodenal HCO secretion, as modified from previously described methods (26, 27). In urethane-anesthetized rats, the stomach and duodenum were exposed and the forestomach wall was incised 0.5 cm using a miniature electrocautery. 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 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. The resultant closed proximal duodenal loop (perfused length 2 cm) was perfused with prewarmed saline by using a peristaltic pump (Cole-Parmer Instrument, Vernon Hills, IL) at 1 ml/min. Input (perfusate) and effluent of the duodenal loop were circulated through a reservoir in which the perfusate was bubbled with 100% O₂. The pH of the perfusate was kept constant at pH 7.0 with a pH stat (models PHM290 and ABU901; Radiometer Analytical, Lyon, France).

Back titration. HCO secretion was measured by two complementary methods: back titration and direct CO₂ measurement. For back titration measurements, the amount of 0.01 N HCl added to maintain constant pH was considered equivalent to duodenal HCO secretion. For pH 7.0 perfusion, a pH stat technique was used with perfusate recirculation. Manual back titration was used for pH 2.2 perfusates. Preliminary in vitro studies indicated an excellent correlation between perfusate HCO concentration and titratable base measured in the effluent (Fig. 1A). For duodenal HCO measurement, a 30-min stabilization with pH 7.0 saline (t = 35 to 5) was followed by baseline measurements with pH 7.0 saline (t = 5 to 10). Acid solution was perfused with a Harvard infusion pump at 1 ml/min. For experiments involving repeated acid exposure, solutions were perfused identically to the protocol described in Repeated acid exposure and effects of ion transport inhibitors as follows: pH 7.0 saline was perfused for 10 min (t = 0-10; baseline), followed by pH 2.2 saline for 15 min (t = 10-25; 1st acid challenge period), followed by pH 7.0 saline for 10 min (t = 25-35; 1st recovery period), followed by pH 2.2 saline for 15 min (t = 35-50; 2nd acid challenge period), and then pH 7.0 saline for 15 min (t = 50-65; 2nd recovery period). O₂ gas-bubbled pH 7.0 saline was recirculated with a peristaltic pump, whereas pH 2.2 saline was perfused via syringe pump. The duodenal loop solution was gently flushed with 5 ml of perfusate to rapidly change the perfusate composition at t = 10, 25, 35, and 50 min. Samples from acid exposure periods were collected in tubes every 5 min and analyzed for HCO secretion by back titration to pH 2.2 with 0.1 N HCl.

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Calibration of HCO measurement systems. A: pH stat. Standard HCO solutions were circulated through the perfusion system tubes. Alkalinity was inferred from the amount of HCl added to maintain starting pH. B: CO2 electrode. A CO2 electrode as described in MATERIALS AND METHODS measured total CO2 content of HCO solutions. Lines denote the range of measured HCO concentrations. Note that at the lowest concentration, a 33% overestimate occurred, which disappeared when HCO concentration exceeded 0.5 mM, compared with the line of identity (dotted). All data are expressed as means of 2-3 experiments. C: stability of CO2-containing solutions in an open system. 10 mM NaHCO3 solutions were prepared, and total dissolved CO2 concentration ([CO2]t) was measured with a CO2 electrode. At t = 0, HCl (0.1 M) was added to titrate to pH 2.0, 4.5, or 7.0. [CO2]t was measured at t = 5, 30, and 60 min.

CO₂ measurement. To determine the relative contributions of HCO secretion and H⁺ back diffusion during acid exposure, total dissolved CO₂ was measured with a CO₂ electrode (FK1501CO2; Radiometer America, Westlake, OH) connected to a pH meter (PHM 62; Radiometer, Copenhagen, Denmark). The duodenal loops were prepared and perfused as described in Preparation of the duodenal loop. Every 5 min, effluent pH was measured immediately after collection. One-half milliliter of 1 M citrate buffer (pH 4.5) was then added to the sample (5 ml) to convert free HCO to CO₂, followed by measurement of electrode potential (mV) with the CO₂ electrode. Total dissolved CO₂ concentration ([CO₂]_t) was calculated according to a calibration curve using freshly prepared 0.1, 1, and 10 mM NaHCO₃ solutions as standards, which generate 0.1, 1, and 10 mM [CO₂]_t, respectively, at pH 4.8-5.0. Since the calibration curve had a slope ~56 mV/log [CO₂]_t within the range of 0.1-10 mM at 20-25°C, all samples were analyzed at 25°C. [CO₂]_t and pH in the perfusate provide CO₂ concentration ([CO₂]) and HCO concentration ([HCO]) from the Henderson-Hasselbach formula

(1)

(2)

Thus

(3)

(4)

According to Henry's law

(5)

Substituting Eq. 3 into Eq. 5, PCO₂ can be calculated from [CO₂]_t and pH

(6)

where pK_a, the first dissociation constant for carbonic acid in saline, is 6.2 at 25°C (25), and K_h, the Henry solubility constant for subsaturating CO₂ concentrations in weak electrolyte solutions, is 0.047 mM/mmHg at 25°C (15). To examine the accuracy of CO₂ measurements through the perfusion system, [CO₂]_t of the perfusate was measured in O₂-bubbled saline containing 0.1, 1, or 10 mM NaHCO₃ that was perfused through the system using the pump and tubing used for duodenal perfusions. Neither addition nor loss of CO₂ was observed when the afferent [HCO] was above ~0.5 mM, but there was a 33% overestimation at the lowest measured CO₂ concentration (0.21 mM), with no difference between measured and predicted CO₂ at the highest measured concentration (1.06 mM; Fig. 1B). Further experiments were done to test the stability of dissolved CO₂ over time in an open system. Ten-millimolar HCO solutions were prepared, their [CO₂]_t was measured, they were titrated to pH 2.0, 4.5, or 7.0 at t = 0, and they were kept at 25°C in the same containers used to collect the effluent. Aliquots were removed at t = 5 min, 30 min, and 60 min. Figure 1C reveals that there was an initial decrease of [CO₂]_t in the first 5 min, probably reflecting liberation of CO₂ during acid titration, but [CO₂]_t was stable for 55 min thereafter.

1

1

Statistics

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Effect of Sustained Acid Perfusion

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Effect of sustained acid perfusion on intracellular pH (pHi), blood flow, and mucosal gel thickness (MGT). pHi was measured at baseline (t = 0) and 5 min after acid perfusion (t = 20). pHi decreased rapidly and significantly at t = 20, the magnitude relative to perfusate pH (A). Blood flow (BF; B) and MGT (C) increased during pH 3.5 or 2.2 perfusion but not perfusion at pH 4.5. *P < 0.05 vs. pH 7.0 Krebs group. Values are means ± SE from 6 rats.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Effect of mild  strong or strong  mild acid exposure on pHi. In the mild  strong group (), pHi during the 2nd acid challenge period was higher than in the constant pH 2.2 group, followed by pHi recovery to baseline and subsequent overshoot at 60 min (A). In the strong  mild group (), pHi during the 2nd acid challenge period gradually increased; an overshoot is observed at 45 min (B). Constant pH 4.5 (pH 4.5-4.5; ) and constant pH 2.2 (pH 2.2-2.2; ) perfusion groups are shown for comparison. *P < 0.05 vs. the corresponding constant pH group. Values are means ± SE from 6 rats.

Effect of Rapid Shifts of Luminal Acid

In the strong  mild acid group, pH_i gradually increased after the perfusate was changed to pH 4.5, and alkalinization to above the predicted levels occurred during the second acid challenge period (Fig. 3B). No adaptive changes occurred in blood flow and MGT in the strong  mild acid group (data not shown).

Effect of Repeated Acid Exposure With or Without DMA or DIDS

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Effect of repeated acid exposure with or without 5-(N,N-dimethyl)-amiloride (DMA) or DIDS on pHi. Two acid pulses separated by a 10-min recovery period () produced an overshoot after the 1st and 2nd acid challenges at 35 and 55 min compared with the pH 7.0 Krebs group (). An adaptive response of pHi during the 2nd acid challenge at 40-50 min was observed. Although DMA () and DIDS () with pH 2.2 further deceased pHi during the 1st acid challenge at 20 min, DIDS but not DMA abolished the overshoot after the 1st acid challenge and pHi adaptation during the 2nd acid challenge. *P < 0.05 vs. pH 7.0 Krebs group; P < 0.05 vs. repeated pH 2.2 alone group by ANOVA; P < 0.05 vs. the corresponding time during the 1st acid challenge period by paired t-test.

View larger version (23K): [in this window] [in a new window]   Fig. 5.   Effect of repeated acid exposure with or without DMA or DIDS on blood flow. Blood flow increased during the 1st and 2nd acid challenge periods (). DMA (), but not DIDS (), abolished the hyperemic response during both acid challenge periods. *P < 0.05 vs. pH 7.0 Krebs group; P < 0.05 vs. repeated pH 2.2 alone group by ANOVA; P < 0.05 vs. the corresponding time during the 1st acid challenge period by paired t-test. Values are means ± SE from 6 rats.

View larger version (23K): [in this window] [in a new window]   Fig. 6.   Effect of repeated acid exposure with or without DMA or DIDS on MGT. Acid exposure increased MGT during the 1st acid challenge and further increased during the 2nd acid challenge (). Although DMA () has no effect on MGT, DIDS () reduced MGT response during the 2nd acid challenge. *P < 0.05 vs. pH 7.0 Krebs group; P < 0.05 vs. repeated pH 2.2 alone group by ANOVA; P < 0.05 vs. the corresponding time during the 1st acid challenge period by paired t-test. Values are means ± SE from 6 rats.

Spontaneous HCO secretion was measured during pH 7.0 saline perfusion (Fig. 7A). Acid exposure increased titratable alkalinity to extreme levels during the first and second acid challenges. During the recovery periods, acid loss in the pH 2.2 saline group was higher than in the pH 7.0 saline group, consistent with post-acid augmented HCO secretion. This stimulated secretion was reduced by DIDS but not by DMA. Although DIDS had no effect on acid loss during the acid challenges, DMA reduced acid loss during the second acid challenge. Figure 7B depicts total CO₂ output in the perfusate of pH 7.0 saline and repeated acid exposure groups. Total CO₂ output stabilized during pH 7.0 saline perfusion but decreased during both acid challenges and progressively increased during both recovery periods. Decreased total CO₂ output compared with the extreme increase of titratable alkalinity during acid exposures suggests that loss of luminal acidity during acid perfusion may be due to H⁺ back diffusion rather than HCO secretion or CO₂ loss. In contrast, increased total CO₂ output during the recovery periods was consistent with post-acid augmented HCO secretion.

View larger version (26K): [in this window] [in a new window]   Fig. 7.   Effect of repeated acid exposure with or without DMA or DIDS on effluent alkalinization (acid disappearance) and [CO2]t. A: Back titration. Acid increased titratable alkalinity during both challenges and during both recovery periods (). Although DMA () and DIDS () had no effect on acid loss during the 1st challenge period, DMA reduced acid loss during the 2nd acid challenge and DIDS, but not DMA, reduced HCO secretion during the recovery periods. B: CO2 measurements. Separate experiments indicated that total CO2 output was somewhat higher during neutral perfusion than was estimated by back diffusion. During the 1st and the last point of the 2nd acid challenge periods, CO2 output was suppressed, indicating that the 90-95% acid loss during this period resulted mostly from H+ back diffusion. Inset: PCO2 was calculated from [CO2]t and pH (see eq. 6). *P < 0.05 vs. pH 7.0 Krebs group; P < 0.05 vs. repeated pH 2.2 alone group by ANOVA. Values are means ± SE from 6 rats.

H+ and CO₂ Back Diffusion in Duodenal Loops

View larger version (25K): [in this window] [in a new window]   Fig. 8.   Changes of H+ concentration ([H+]) and [CO2]t in perfusates and effluents of different pH and [CO2]t solutions. Measured pH, measured [CO2]t, and calculated PCO2 of perfusates and effluents for solutions of nominal pH 7.0, 2.2, and 6.4 are shown. The numbers over the closed columns indicate %net loss of [CO2]t. *P < 0.05 vs. perfusate in pH and [CO2]t by ANOVA. Values are means ± SE from 3 rats.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Simultaneous and parallel measurements of rat duodenal defenses in vivo provided a useful means of examining the response of the mucosa to luminal acid. Alteration of perfusate pH or a brief acid challenge induced cellular base uptake, protecting the epithelial cells from acidification during subsequent acid exposure. Responses of MGT and blood flow to the acid challenges showed little such adaptation to the second acid challenge or to a change of perfusate pH. Adaptive pH_i changes were abolished by DIDS, which had no effect on blood flow but did prevent the increase of MGT during the second acid challenge. Furthermore, DIDS inhibited the post-acid challenge increases of HCO secretion. Conversely, DMA inhibited the blood flow increase during the first and second acid challenges but had no effect on MGT or on HCO secretion following acid challenge. Measurement of HCO secretion by back titration and total CO₂ measurement revealed that with pH 2.2 perfusate, most acid disappearance was due to H⁺ back diffusion and not alkaline secretion. Furthermore, DMA decreased the rate of acid back diffusion, presumably by inhibiting the hyperemic response to acid (3).

Augmented HCO secretion following acid perfusion is currently accepted as the most important duodenal defense mechanism (7, 30). The mechanism by which HCO secretion protects the mucosa has generally been assumed to be complete neutralization of luminal acid diffusing toward the mucosa within the mucus gel. Since duodenal epithelial cells readily acidify in the presence of even moderate concentrations of perfused acid, it is doubtful that complete neutralization of back-diffusing acid occurs in the preepithelial mucus gel. If acid is able to enter the cells, perhaps an alternate mucosal protective mechanism exists: intracellular buffering or pH_i regulation. If pH_i regulation is important, then HCO secretion may not need to increase in the presence of luminal acid. This prospect prompted our measurement of HCO secretion during the same repeated acid exposure that produced pH_i adaptation. We hence endeavored to ascertain the relative contributions of acid back diffusion, of CO₂ back diffusion, and of HCO secretion toward the disappearance of luminal acid in duodenal perfusions. With the use of back titration, the amount of titratable alkalinity increased 10-fold during acid perfusion, in agreement with Flemström and Kivilaakso (8), although the relative increase was greater than that found by Nylander and colleagues (18, 19). To further study this increase of titratable alkalinity during acid perfusion, we measured [CO₂]_t. Measurement of [CO₂]_t is a more direct means of measuring HCO secretion than is back titration since it has the advantage of not being affected by secretion of other alkali or loss of H⁺ due to back diffusion. The concordant increase of effluent [CO₂]_t and titratable alkalinity in PGE₂-treated rats confirms the validity of these measurements in the absence of luminal acid. As long as total PCO₂ is <10 mmHg at 25°C (the highest [CO₂]_t was <2 mM in this study), the aqueous solubility of CO₂ is such that little loss of CO₂ to the atmosphere occurs over the measurement period, as we demonstrated in preliminary studies. The rapid perfusion technique has been applied successfully by Olbe and co-workers (5), who found that the accuracy of clinical gastric HCO secretion measurements with a CO₂ electrode was improved by increasing the perfusion rate, which ensured that CO₂ concentrations would remain well below the Henry solubility limit. Furthermore, the 20-30% loss of [CO₂]_t and ~12-28% loss of H⁺ between perfusate and effluent of duodenal perfusions is in close agreement with published measurements of CO₂ loss from duodenally perfused HCO solutions (6), confirming that the duodenum has a measurable, finite permeability to CO₂ and H⁺. The rapid increase of titratable alkalinity measured during acid perfusion was not accompanied by a commensurate increase of effluent [CO₂]_t, strongly consistent with our hypothesis that HCO secretion is not augmented during acid perfusion but rather that luminal acid disappears via H⁺ back diffusion. To explain these data, we hypothesized that, during luminal acid perfusion, either: 1) HCO secretion increases ~1,000% within 5 min, HCO is converted by acid to CO₂ gas, >90% of the CO₂ is lost due to back diffusion, and HCO secretion decreases to near baseline within 5 min, or 2) HCO secretion is unchanged, ~25% of H⁺ is lost, and a further ~25% CO₂ is lost due to conversion to CO₂ gas with back diffusion. Since the data depicted in Fig. 8 and the work of others indicate that CO₂ loss, even in solutions with high PCO₂, in a rapidly perfused system is ~25% and that increases and decreases of the rate of HCO secretion generally develop over a 30- to 60-min time period, the latter H⁺ back diffusion model fits most closely with the data. Furthermore, we demonstrated that, even in an open system, CO₂ loss cannot account for the discrepancy between [CO₂]_t measurements and acid disappearance. These measurements add confidence to our contention that HCO secretion was not augmented, although H⁺ back diffusion was substantially increased during luminal acid perfusion.

Importantly, since the only time that HCO secretion is necessary for mucosal protection is during acid exposure, its lack of increase by acid diminishes its importance as a mucosal defense mechanism. During acid stress, it would seem more logical for the cell to retain its protective HCO buffer rather than release it into the lumen, as would be predicted by models implicating HCO secretion during acid exposure as a major defensive factor. From these studies, we could conclude that most of the luminal acid loss during acid challenge is from H⁺ back diffusion, suggesting that HCO secretion of itself cannot defend pH_i against acid challenge and may merely serve as an easily measurable sequelum to prior cellular base uptake.

The pH_i response to repeated acid challenge differed between the first and second challenges. During the first challenge, cells acidified more in the presence of DIDS and DMA. Following acid challenge, DIDS attenuated pH_i recovery and overshoot, coincident with its inhibition of cellular base uptake. During the second acid challenge, DIDS also abolished the adaptive pH_i response, although an early overshoot was observed. These observations are in full agreement with those of Paimela et al. (20), who noted that addition of the DIDS-related stilbene derivative SITS or removal of HCO from the serosal bathing solution of isolated Necturus duodenal mucosa increased cellular acidification during luminal acid challenge. Our interpretation differs from theirs, however, in that we believe that the increased susceptibility to acidification more likely represents suppression of cellular base uptake than it does inhibition of HCO secretion. In the absence of perfused acid, augmented HCO secretion during the first and second recovery periods was attenuated by DIDS but not by DMA, indicating that DIDS-inhibitable HCO secretion is correlated with DIDS-inhibitable cellular HCO uptake.

As shown previously (1, 3), the duodenum has a predictable and robust hyperemic response to perfused acid, even though there is no evidence of augmentation of this response to repeated acid challenges. Although the hyperemic response has been unquestionably implicated in the reduction of gastric injury susceptibility (11), the protective role of blood flow in the duodenum is controversial (14, 24, 30). One explanation for the lower amount of acid back diffusion in DMA-treated rats during the second acid challenge is that suppression of acid-related hyperemia by DMA reduced acid back diffusion during the second acid challenge. The most comprehensive study of the relationship between duodenal blood flow and injury susceptibility was published by Lugea et al. (16, 17), who found that a 1-ml intraduodenal bolus of 100 mM HCl (pH 1) significantly reduced damage due to a bolus of 400 mM HCl (pH 0.4) given 30 min later. Although they measured a significant increase of duodenal blood flow in response to a bolus of 100 mM HCl, they did not measure blood flow during the mild  strong acid sequence. It is thus difficult to ascertain the contribution of blood flow to the observed adaptive protection from injury observed in that study. The suppression of the large increase of titratable alkalinity by DMA, which also suppressed acid-related hyperemia (3), underscores the putative role of blood flow in removing mucosal (luminal) acid via back diffusion. In other words, if the function of acid-related hyperemia is to carry away back-diffusing acid, one would predict that suppression of the hyperemic response would decrease the amount of acid back diffusion, as was observed in this study.

MGT also increased during acid perfusion, as we have previously demonstrated. Augmentation of MGT during acid perfusion indirectly supports its defensive role against acid. It is likely that blood flow, MGT increase, and cellular base uptake, all of which are increased by luminal acid, act in concert to increase duodenal resistance to acid.

One of the most interesting aspects of this study was the demonstration that rapid shifts of perfusate pH per se, and not necessarily the mild  strong exposure sequence, induced the adaptive responses observed. To our knowledge, this possibility has heretofore never been explored, although our data, in which adaptive responses of pH_i were observed after reversing the mild  strong sequence or exposing the mucosa to repeated pulses of strong acid, convincingly suggests that, at least in terms of HCO uptake, the rapid shifts of perfusate pH are all that are necessary. These shifts simulate the physiological state of the duodenum, wherein the mucosa is exposed alternately to peristaltically conveyed waves of gastric acid and bursts of pancreatic HCO after meal ingestion. Measurements of duodenal pH indicate shifts of 4-5 pH units over <1 min (23). Although we have not proven that the observed resistance to acidification truly reflects protection from injury, numerous studies of gastric epithelium have confirmed that resistance to mucosal acidification during acid challenge consistently correlates with decreased injury susceptibility in standard models (22, 28, 29).

In summary, we found that acute adaptive responses occurred when perfusate pH was shifted every 10-15 min, regardless of the sequence of acid concentrations used. Furthermore, the only consistent acute adaptive alteration of a defense mechanism was presumably increased cellular HCO uptake, which produced intracellular alkalinization and resistance to acidification from perfused acid. Significant back diffusion of perfused acid correlated with increased mucosal blood flow. We propose that intracellular HCO uptake via a DIDS-sensitive basolateral Na⁺-HCO cotransporter, which temporally precedes HCO secretion in response to rapid shifts of luminal pH, is an early acute duodenal response to luminal acid. Furthermore, we postulate that HCO secretion may not be the primary mucosal defense mechanism. The reason why HCO secretion correlates well with duodenal acid resistance may signify its role in restoring epithelial pH_i after acid-induced base uptake. We conclude that cellular HCO uptake is likely to be an important duodenal defense mechanism that is induced by physiological postprandial rapid shifts of luminal pH and precedes active HCO secretion.