Heterogeneity of acid secretion induced by carbachol and histamine along the gastric gland axis and its relationship to [Ca2+]i

Jose Francisco Perez-Zoghbi, Adriana Mayora, Marie Christine Ruiz, Fabian Michelangeli

Abstract

The gastric glands of the mammalian fundic mucosa are constituted by different cell types. Gastric fluid is a mixture of acid, alkali, ions, enzymes, and mucins secreted by parietal, chief, and mucous cells. We studied activation of acid secretion using LysoSensor Yellow/Blue in conjunction with fluo 3 to measure changes in pH and Ca2+ in isolated rabbit gastric glands. We evidenced a spatial heterogeneity in the amplitude of acid response along the gland axis under histamine and cholinergic stimulation. Carbachol induced a transitory pH increase before acidification. This relative alkalinization may be related to granule release from other cell types. Omeprazole inhibited the acid component but not the rise in pH. Histamine stimulated acid secretion without increase of lumen pH. We studied the relationship between Ca2+ release and/or entry and H+ secretion in glands stimulated by carbachol. Ca2+ release was associated with a fast and transient components of H+ secretion. We found a linear relationship between Ca2+ release and H+ secretion. Ca2+ entry was associated with a second slow and larger component of acid secretion. The fast component may be the result of activation of Cl and K+ channels and hence H+/K+ pumps already present in the membrane, whereas the slow component might be associated with translocation of H+/K+ pumps to the canaliculi. In conclusion, with cholinergic stimulation, gastric glands secrete a mixture of acid and other product(s) with a pH above 4.2, both triggered by Ca2+ release. Maintenance of acid secretion depends on Ca2+ entry and perhaps membrane fusion.

  • gastric secretion
  • calcium
  • LysoSensor Yellow/Blue
  • fluo 3
  • fluorescence microscopy
  • thapsigargin
  • intracellular Ca2+ concentration

the mammalian gastric gland in the fundus of the stomach secretes a variety of products to facilitate food digestion. These products are secreted by different cell types in the gastric gland. Gastric acid is produced by parietal cells, pepsinogen by chief cells, and mucus by neck cells. The physiological control of gastric secretion is assured by a complex interplay of neural, endocrine, and paracrine systems that coordinately activate these specialized cells in the gland (24, 46). This control is performed through the activation of specific G protein-coupled receptors, which are differentially expressed on the various cell types. Therefore, the composition of primary gastric juice in the gland may vary according to the stimulus. Pepsin secretion is stimulated by agonists acting on β-adrenergic, muscarinic (M1 and M3), and cholecystokinin (CCK1) receptors (16, 25, 47, 48). Chief cells are not stimulated by histamine since they do not bear H2 receptors. Mucous neck cells are stimulated by muscarinic and β-adrenergic agonist (18).

Acetylcholine and histamine directly stimulate the parietal cell to secrete acid (5, 8, 29, 33). Histamine is released from enterochromaffin-like (ECL) cells by gastrin in mammalian and amphibian stomach (4, 38, 40, 42). Acetylcholine coming from enteric nerve terminals in the vicinity of the gland also releases histamine in the amphibian stomach and perhaps in mammals (38, 39, 42). However, acetylcholine and histamine activate acid secretion differently. Acetylcholine binds to M3-muscarinic receptors leading to increases in intracellular Ca2+ concentration ([Ca2+]i) (1, 7). Histamine binds to H2 receptors resulting in the elevation of both [Ca2+]i and cyclic AMP (cAMP) (8, 10, 27, 36).

The isolated gastric gland is the preparation of choice to study the mechanism of gastric acid secretion by parietal cells because 1) the intracellular signals regulating secretion, such as agonist-induced Ca2+ signaling, can be studied in individual cells by fluorescence microscopy; 2) parietal cells preserve their orientation and relationship to other cells in the intact gland; 3) HCl and other products are secreted and preserved in a sealed lumen, allowing their measurement; and 4) the preservation of multiple cells in the gland is important to understand the integrated secretory process.

Gastric glands have been widely used to characterize agonists-induced dynamic [Ca2+]i changes in parietal cells by fluorescence microscopy (9, 27, 3133). However, measurements of gastric secretion were determined in separate experiments using a radiolabel method with low temporal resolution. To improve the measurement of the kinetics of acid secretion and its relationship to [Ca2+]I, we previously developed a methodology using fluorescent dyes to simultaneously measure [Ca2+]i and secretion (36). In our previous work we characterized the temporal relationship in real time between acid secretion and Ca2+ changes stimulated by carbachol and histamine observing a striking difference in the kinetics of activation of H+ secretion. The acid response to a pulse of carbachol was fast, small, transient, and synchronized with the Ca2+ spike, whereas histamine induced a slow, large, and sustained H+ secretion with a long lag time (140 s), which was unrelated to the Ca2+ signal (36). Since the activation of HCl secretion by both types of secretagogues involves the translocation of H+/K+ pumps contained in tubulovesicular structures to the plasma membrane and the activation of K+ and Cl channels (41, 46), this raises the question as to which are the coupling mechanisms between the elevation of different second messengers and the activation of secretion.

In the course of our studies we detected with carbachol stimulation variability in the acid response measured in parietal cells and gland lumen despite similar Ca2+ increase recorded in parietal cells. Taking advantage of our methods we have characterized the spatiotemporal correlates of this response. In this work, we present evidence, for the first time, of a spatial heterogeneity of the pH changes in response to carbachol and histamine along the gland axis, which may be the result of a mixed secretion originating from multiple cell types. In addition, we have analyzed the amplitude of the Ca2+ changes and H+ secretion during cholinergic stimulation in trying to understand their relationship. Furthermore, given the complexity of the Ca2+ response to cholinergic agents, it became also necessary to evaluate the contribution of the Ca2+ release and entry phases in triggering and maintaining acid secretion.

METHODS

Gland preparation and dye loading.

Isolation of gastric glands was performed by collagenase digestion as described by Berglindh (3) and modified in our laboratory (29). All experiments were carried out in accordance with the guiding principles for the care and use of laboratory animals of the bioethics committee at Instituto Venezolano de Investigaciones Científicas (Ministry of Science and Technology, Venezuela). Briefly, New Zealand White rabbits were anesthetized and killed by intravenous pentobarbital injection. The stomach was perfused through the descending aorta under positive pressure. The mucosa was scraped from the underlying muscularis, minced, and digested with type IA collagenase at 37°C for 30 min. Glands were separated by sedimentation and resuspended in Ringer medium.

Fluorescence staining of isolated gastric glands.

Isolated gastric glands previously loaded with 5 μM LysoSensor Yellow/Blue (LYB) for 30 min at room temperature were adhered to polylysine coated coverslips, mounted in a flow-through perfusion microscope chamber. For the staining of secretory granules, glands were superfused with acridine orange (50–100 μM) for 15 min at room temperature. The images in Fig. 1 were taken with a Nikon Coolpix 990 digital camera at 3.3 megapixel resolution or Nikon D40X at 10 megapixel resolution. Superposition of images was performed with Photoshop.

Fig. 1.

Heterogeneity in the cell type distribution along the gastric gland. Images of an isolated rabbit gastric gland in the resting state (AD) and after stimulation with 10 μM carbachol for 5 min (EH). The gland was stained with LysoSensor Yellow/Blue (LYB; see methods). A: phase contrast image of the gland. B and F: LYB fluorescence images of resting (B) or stimulated (F) gland. C and G: combination of phase contrast and LYB fluorescence images of resting (C) and after carbachol stimulation (G). D and H: details of a region of the gland presented in C and G, respectively (denoted by squares). At the end of stimulation period, acridine orange was superfused. E: superposition of 3 images corresponding to phase contrast, acridine orange, and LYB fluorescence images in stimulated states. I: fluorescence image of a resting gastric gland stained with acridine orange (50 μM, 15 min). Images were taken with a Nikon Coolpix 990 digital camera at 3.3 megapixel resolution (AH) or with a Nikon D40X digital camera at 10 megapixel resolution. Superposition of images was performed by using Photoshop. PC, parietal cell; CC, chief cell or mucous neck cell; L, lumen.

Fluorimetric measurements and imaging.

Simultaneous measurements of [Ca2+]i and H+ secretion was achieved by using the methodology previously described (36). Gland suspensions were incubated for 30 min at room temperature with 5 μM fluo 3-AM and 5 μM LYB. After loading, glands were washed two times with Ringer medium (RM). Loaded glands were adhered to polylysine coated coverslips and mounted in a flow-through perfusion microscope chamber. Fluorescence imaging and photometry of stained glands was made by use of a fluorescence microscope (Nikon TE-300) equipped with a cooled charge-coupled device camera (Newcastle Photometric Systems), shutters, and a filter wheel for excitation wavelength changes. Fluorescence was measured from up to 16 regions of interest at 16-bit resolution. Images were recorded at 8-bit resolution and displayed in pseudocolors. Glands were alternatively illuminated at excitation wavelengths of 387 and 500 nm for LYB and fluo 3, respectively. Fluorescence emission was recorded at 535 nm for both probes. All measurements were made at room temperature. Fluorescence data were normalized according to the equation ΔF/Fo = (F − Fb)/(Fb − background), where F is the fluorescence measured at any given time, Fb is the fluorescence just before secretagogue addition, and Fo is Fb minus background fluorescence. All experiments were performed in the presence of 100 μM cimetidine to block the possible effects on parietal cells of histamine released during carbachol stimulation.

Solutions and materials.

RM contained, in mM, 145 NaCl, 5 KCl, 1 MgSO4, 1.8 CaCl2, 11 glucose, 20 HEPES, pH 7.45; 310 mosM. LYB and fluo 3-AM were purchased from Molecular Probes-Invitrogen. All other reagents were from Sigma.

RESULTS

Heterogenous distribution of cells types along the isolated gastric gland.

Isolated rabbit gastric glands are long tubular structures consisting of several cell types, including parietal, chief, ECL, and mucous cells. The phase contrast images show the high density of cells congregated along the gland axis. Parietal cells can be readily recognized bulging out of the gland (Fig. 1A). LYB accumulates exclusively in the intracellular canaliculi of parietal cells and the lumen of the glands, both of which communicate by a thin isthmus (Fig. 1, BH). Superposition of the phase contrast image with that of LYB fluorescence puts in evidence the presence of numerous cells interdispersed among the parietal cells (Fig. 1, C and D). The addition of acridine orange stained secretory granules accumulated at the apical pole of these cells, around the lumen, which may correspond to chief and/or mucous cells (Fig. 1E).

Stimulation by carbachol increased LYB fluorescence mainly in the gland lumen and the isthmus, whereas small changes were observed in the intracellular canaliculi (Fig. 1, FH). This localized increase in LYB fluorescence in the gland may be due to the drainage of small amounts of acid from individual parietal cells that converge in a luminal space of larger magnitude whose signal can be better observed. Alternatively, it may be due to the activation of H+/K+ pumps present in the apical membranes of the parietal cell lining the gland lumen.

In isolated glands stained with acridine orange we can recognize zones with a differential distribution of parietal and granule-containing cells (Fig. 1I). Glands that have retained their long structure show a high density of parietal cells in the middle region, which emit red fluorescence indicating accumulation of acid. Toward one extreme we can appreciate a zone rich in cells with a granular orange fluorescence that may correspond to chief and/or mucous cells (bottom area). In this region parietal cell appeared less red fluorescence, suggesting less acidic intracellular canaliculi. In the upper region the parietal and orange granule-containing cells were less abundant. These images confirm a differential distribution of parietal and other cell types along the gland axis in our preparation.

Carbachol stimulated mixed secretions in isolated gastric glands.

In previous work we showed that the response to carbachol was characterized by an increase of LYB fluorescence in the gland lumen, synchronized with an increase of cell fluo 3 signal (36). However, close examination of all the data of cholinergic response revealed different patterns of LYB signal depending on the gland or region of it. In nearly 40% of the glands (from a total of 70 glands from 15 animals) we observed a complex secretory response composed of an initial decrease followed by an increase of the LYB fluorescence (Fig. 2). This initial decrease of LYB fluorescence is compatible with a relative alkalinization of the gland lumen. In this phase, carbachol induced an immediate turnoff of LYB fluorescence in the entire lumen and a slower decrease in parietal cells where the intracellular canaliculi were still visible (Fig. 2, image 2). This was followed by a prompt stimulation of the acidic component observable in all parietal cells and along the gland lumen. A second pulse of carbachol in this case induced only acidification of these spaces. Since LYB fluorescence senses pH below 4.2 we cannot determine the absolute pH of the second component and can only infer that the drop in fluorescence corresponds to a relative alkalinization of the gland lumen. For the sake of clarity in this paper we will refer to this component as alkaline secretion or alkalinization.

Fig. 2.

Acid and alkaline responses of an isolated gland elicited by 2 pulses of 10 μM carbachol. General experimental details are given in methods. Glands were double stained with LYB and fluo 3 and then washed by continuous perfusion with Ringer medium (RM). Fluorescence was excited at 387 and 500 nm for LYB and fluo 3, respectively, and recorded at 530 nm. Gastric glands were continuously perfused with RM in the presence of H2 receptor antagonist cimetidine (100 μM) throughout the experiment and stimulated at the indicated times with 10 μM carbachol (CCh). Traces correspond to fluo 3 fluorescence (blue trace) and LYB fluorescence in an individual parietal cell (red trace) and its adjacent lumen (green trace), normalized as defined in methods. Sampled areas are indicated on the LYB fluorescence image. A sequence of images taken before and after stimulation is presented. Note that the changes occur in the entire lumen and cells and are synchronized. In this and following figures LYB fluorescence is shown in pseudocolors.

Variability of the acid response along the gland.

We examined the spatial distribution of the LYB and Ca2+ responses along the gland axis under stimulation by repetitive pulses of carbachol (Fig. 3). Each pulse induced a synchronized Ca2+ spike in the cells whose amplitude progressively decreased with each new stimulation (Fig. 3, left). Measurements of LYB fluorescence made along the lumen of the gland revealed a complex pattern of response in both magnitude and shape (Fig. 3, right). In luminal areas A1 to A3 each stimulation generated a step increase of LYB fluorescence corresponding to a transient pulse of H+ secretion. This initial increase was followed by a plateau or a slow and small decrease in fluorescence. In the subsequent luminal regions (areas A4 to A7) the measurements showed the progressive appearance of a second transitory component of LYB fluorescence decrease which may correspond to alkaline or less acidic secretion as shown above in Fig. 2. In areas A4 and A5 the LYB fluorescence was initiated by a mixture of both acidic and alkaline secretions (mixed response). The alkaline transitory component gradually took over the acid response up to the A7 region. This relative alkalinization was followed by a rapid acidification to reach a plateau above the basal value. Repetitive stimulation with carbachol elicited the same pattern of luminal LYB response with each pulse, however, attenuated. At the parietal cell level, the recorded secretory response was similar to that observed in the adjacent luminal regions but of a much smaller amplitude (Fig. 3, left). The change of pH induced by carbachol (acid or alkaline) was synchronized with the Ca2+ spike measured in the cells, independently of the region of the gland, with a delay in the range of 2–4 s (Table 1), not different from what we had reported before for a purely acidic response (36).

Fig. 3.

Variability of intracellular Ca2+ concentration ([Ca2+]i) response and acid secretion along the gland under stimulation by repetitive pulses of carbachol. General experimental details as in Fig. 2. On the right side, curves correspond to the normalized measurements of LYB fluorescence performed in areas A1 to A7 along the lumen of the isolated gland. On the left side, curves correspond to normalized fluo 3 fluorescence (blue traces) and normalized LYB fluorescence (black traces) in individual parietal cells (A8–A11). Sampled areas are indicated on the LYB fluorescence image.

View this table:
Table 1.

Comparative kinetic parameters of acid and alkaline accumulation in both the intracellular canaliculi of parietal cells and the lumen of gastric glands stimulated with carbachol

The alkaline component registered in some regions of the lumen could be explained by the secretion of relatively basic products. The amplitude of the response does not give sufficient information as to where the two secretions may be originating. In Table 1 we compare the time needed to reach maximal amplitude in the cell vs. the lumen for three types of response in an attempt to dissociate the acid and basic components. For the response showing only an acid phase, the maximum was attained in the cell at 16 s after the addition of carbachol, whereas it took 37 s for the lumen. This is in accordance with acid being originated at the parietal cell that diffused into the lumen. In the case of the response with a large initial alkaline component, the maximal alkalinization is attained faster in the lumen (33 s) than in the parietal cell (42 s). In the intermediate type of response again the maximal alkaline pH was attained firstly in the lumen (23 s) and then in the cell (34 s). Even in this type of response the maximal acidification is attained faster in the cell (70 s) than in the lumen (81 s). These results are compatible with the interpretation that the alkaline component is originated in a cell different from the parietal cell. Alkali would accumulate first in the lumen, buffering the H+ in the lumen and later in the parietal cell canaliculi.

Since histamine H2 receptors are present only in parietal cells we tested whether histamine would elicit the same type of response as carbachol along the gland (Fig. 4). Figure 4 shows recordings of [Ca2+]i and acid secretion in two representative cells and six adjacent gland lumina along the gland axis under stimulation by 100 μM histamine. Parietal cells within the gland responded to histamine stimulation with nonsynchronous [Ca2+]i oscillations (Fig. 4, A7 and A8), as has been reported previously (27, 36). Histamine induced an increase of LYB fluorescence in all areas of the gland despite differences in intensity. Acid secretion was initiated in all luminal areas with a lag time of about 120 s. Increases in LYB fluorescence intensity in gland lumen were progressively larger going from areas A1 to A6 indicating a heterogeneity of the acid secretory response to histamine along the gland. Maximal LYB fluorescence in the lumen was reached around 480 s after stimulation with a magnitude significantly larger than that produced by a pulse of carbachol, confirming our previous results (36). Subsequent addition of carbachol gave rise to a typical Ca2+ spike measured in parietal cells synchronized with a transitory decrease of LYB fluorescence in the cells and in the lumen. The amplitude of the alkaline response varied along the gland axis and decreased as acidification brought about by histamine increased. These experiments show that the alkaline response is stimulated by carbachol but not by histamine. This suggests that the basic component does not originate in the parietal cell but rather in other cell types, which do not bear histamine but cholinergic receptors.

Fig. 4.

Heterogeneity of [Ca 2+]i response and acid secretion along the gland under stimulation by histamine and carbachol. General experimental details as in Fig. 3. On the right side, curves correspond to the normalized measurements of LYB fluorescence performed in areas A1 to A6 along the lumen of the isolated gland. On the left side, curves correspond to normalized fluo 3 fluorescence (blue traces) and normalized LYB fluorescence (black traces) in individual parietal cells (A7–A8). Sampled areas are indicated on the LYB fluorescence image. Histamine (100 μM) stimulated only the acid component, whereas subsequent carbachol addition elicited a combined response. Note heterogeneity of the responses along the gland.

We have used omeprazole to block the H+/K+ pump and dissect the different components of the cholinergic response (Fig. 5). Omeprazole inhibited the acidification induced by carbachol observing only the basic component of secretion. This confirms that the increase of LYB fluorescence is due to the activity of the H+/K+ pump of parietal cells and suggests that the alkaline component is a different secretory process independent of the H+-K+-ATPase activity.

Fig. 5.

Effect of omeprazole on acid and alkaline responses elicited by carbachol in isolated glands. Glands, double stained with LYB and fluo 3, were pretreated with 100 μM omeprazole for 10 min and then stimulated with 10 μM carbachol. Traces correspond to normalized LYB fluorescence corresponding to the lumen region of the omeprazole-treated gland (green trace). Normalized fluo 3 fluorescence (blue trace) corresponds to an individual parietal cell adjacent to the sampled luminal region. A control trace of LYB fluorescence obtained in the lumen of a gland (same preparation) stimulated by carbachol (in the absence of omeprazole) is included for comparison (red trace). Arrows correspond to the time when images were captured.

In an attempt to identify the source of alkaline secretion we sought evidence for secretory granule exocytosis from chief or mucous cells. Exocytosed proteins and mucins could explain relative alkalinization by buffering of luminal acid contents. In acridine orange-stained glands carbachol induced an apparent decrease of fluorescence associated to secretory granules in nonparietal cells within 1 min after stimulation (Fig. 6). This supports the view that carbachol induced a mixed secretion by stimulating acid and protein secretion from different cell types.

Fig. 6.

Release of secretory granules in an isolated gland by carbachol. Glands were adhered to polylysine-coated coverslips, mounted in a flow-through microscope chamber, and superfused with a solution containing acridine orange (50 μM, 15 min). Images were taken before (A) and 1 min after (B) carbachol stimulation (10 μM) with a Nikon D40X digital camera at 10 megapixel resolution. Arrows show areas of degranulation.

Relationship between [Ca2+]i and H+ secretion induced by carbachol.

The relationship between [Ca2+]i and H+ secretion was analyzed in those glands presenting only the acid component of stimulation. We used different approaches to evaluate the contribution of the different components of the Ca2+ response, intracellular release and extracellular Ca2+ entry, during cholinergic stimulation. In Fig. 7, we present the relationship between the [Ca2+]i increase and H+ secretion in glands stimulated by three consecutive pulses of 10 μM carbachol (20 s/pulse). The time course reveals that each pulse induced a transient increase of the two parameters whose amplitude progressively decreased. The plot of amplitude of the change in LYB fluorescence as a function of the maximal Ca2+ increase induced by sequential pulses of carbachol indicated a linear relationship between the two parameters with a high correlation coefficient (Fig. 7B; r = 0.99). Cholinergic stimulation has been associated to the activation of the phospholipase C-inositol triphosphate pathway resulting in a release of Ca2+ from the endoplasmic reticulum (ER) and a secondary entry of extracellular Ca2+. To evaluate the contribution of the Ca2+ released from the ER in the acid response we used thapsigargin, an inhibitor of the sarco(endo)plasmic reticulum Ca2+-ATPase pump, to deplete ER pools (Fig. 8). After a first pulse of carbachol, which induced the Ca2+ and acid responses already described, treatment with 1 μM thapsigargin for 6 min abrogated both Ca2+ and acid responses elicited by a second carbachol pulse. In this condition the treatment with thapsigargin was sufficient to deplete ER pools, precluding further Ca2+ release by carbachol. This suggests that intracellular Ca2+ released from the ER is related to the activation of acid secretion during a short pulse of carbachol. Stimulation of acid secretion by secretagogues involves the translocation of the H+/K+ pumps by fusion of the tubulovesicular structures with the canalicular membrane and the translocation and activation of Cl and K+ channels at this membrane (46). In a previous study we showed that a pulse of carbachol induced a rapid, small, and short increase of acid secretion, which we interpreted as suggesting activation of Cl and K+ channels and H+/K+ pumps already present at the apical membrane, without inducing membrane fusion (36). In Fig. 9, we compare the effects of sustained stimulation with 10 μM carbachol (6 min) with those evoked by a short (20 s) pulse. In both cases the Ca2+ response was biphasic involving a release phase and Ca2+ entry, which returned to basal levels when the agent was removed. Regarding the LYB signal, we observed for both kinds of stimulation a rapid and small increase that was synchronized with the spike of Ca2+ increase (Fig. 9, A and B). This was followed by a slow and large increase in acid output when carbachol was maintained in the medium (Fig. 9B) but not in the case of pulse stimulation. This was observed both in the lumen and in the cell, being more important in the luminal space. This suggests that cholinergic agents may activate more than one mechanism that can be dissociated by varying the length of the stimulus. To investigate the role of extracellular Ca2+ entering during cholinergic stimulation we manipulated the Ca2+ concentration of the extracellular medium. In resting glands, the step change of extracellular Ca2+ concentration from 1.8 to 5 mM slightly increased [Ca2+]i, without stimulating acid secretion (Fig. 10A). Upon returning to 1.8 mM Ca2+ in the extracellular medium, [Ca2+]i came back to basal levels. In this condition, sustained stimulation by carbachol elicited the Ca2+ and acid response in the cell and lumen already described in Fig. 9. In the continuous presence of the agonist the change to a medium containing 5 mM Ca2+ induced a large and sustained increase in [Ca2+]i probably due to the entry of Ca2+ through store operated channels. This increase in [Ca2+]i was accompanied by an increase in the rate of acidification of both the lumen and the cell. Removal of extracellular Ca2+ lead to a decrease of [Ca2+]i (Fig. 10B). Carbachol stimulation in Ca2+-free medium induced a Ca2+ spike that returned to basal level without plateau. This corresponds to the intracellular release component. Acid response was characterized by a step augmentation in the lumen but not in the cell, followed by a second slower increase in both lumen and cell. Restitution of 1.8 mM Ca2+ in the medium in the continuous presence of carbachol provoked a fast increase in [Ca2+]i due to Ca2+ entry. This was accompanied by a concomitant increase in acid secretion in the lumen followed by a slower increase in the cell. These results indicate that carbachol is able to trigger the acid response in the absence of extracellular Ca2+. However, the entry of Ca2+ permits to induce a response of larger amplitude. Another strategy to study the effect of [Ca2+]i increase was to induce the activation of capacitative Ca2+ entry by using thapsigargin in Ca2+-free medium to deplete ER pools (Fig. 11). Addition of 2 μM thapsigargin in Ca2+-free medium induced a barely detectable rapid increase in [Ca2+]i. Restitution of extracellular Ca2+ to 1.8 mM and then 5 mM led to Ca2+ entry. This increase in [Ca2+]i was followed by a slow increase in HCl secretion in the lumen and in the cell. However, the acid response was not homogeneous in the different cells (C1, C2). Altogether, this set of results indicates that the acid secretory apparatus is able to respond to elevations in [Ca2+]i due to either release or Ca2+ entry during cholinergic stimulation.

Fig. 7.

Relationship between the [Ca2+]i increase and H+ secretion in glands stimulated by 3 consecutive pulses of carbachol. General experimental details are as given in Fig. 2. Gastric glands were continuously perfused with RM and stimulated at the indicated times with 10 μM carbachol (CCh). A: representative traces of fluo 3 fluorescence recorded in a parietal cell and LYB fluorescence in an adjacent luminal region. B: the maximal changes in fluo 3 fluorescence of 7 parietal cells are plotted against the maximal changes in LYB signal obtained in 7 adjacent luminal areas. Points correspond to the average ±SE normalized to the values obtained during the first stimulus (Stim.). The straight line is the linear regression calculated for the 21 points (correlation coefficient R = 0.99). The results are representative of a series of 3 independent experiments from 3 different animals.

Fig. 8.

Effect of endoplasmic reticulum Ca2+ depletion by thapsigargin on the [Ca2+]i increase and net acid output in glands stimulated by carbachol. Gastric glands were continuously perfused with RM and stimulated with 2 pulses of 10 μM CCh at the indicated times (A and B). In B, the second pulse of carbachol was applied 5 min after treatment with 1 μM thapsigargin. Representative traces of fluo 3 fluorescence (blue) recorded in a parietal cell and LYB fluorescence (red) measured in an adjacent luminal region.

Fig. 9.

Comparison of the effect of pulsed vs. sustained stimulation with carbachol on the [Ca2+]i increase and net acid output in isolated gastric glands. Glands were stimulated with 10 μM carbachol during a pulse of 20 s (A) or during 6 min (B). Representative traces of fluo 3 fluorescence (blue) recorded in a parietal cell and LYB fluorescence measured in the same parietal cell (red) and its adjacent luminal region (green).

Fig. 10.

Role of extracellular Ca2+ ([Ca2+]o) entering during cholinergic stimulation on the [Ca2+]i increase and net acid output in isolated gastric glands. Gastric glands were continuously perfused with RM containing different Ca2+ concentrations, as indicated in the box. Glands were stimulated with 10 μM carbachol in either normal (1.8 mM) (A) or Ca2+-free medium (B). Representative traces of fluo 3 fluorescence (blue) recorded in a parietal cell and LYB fluorescence measured in the same parietal cell (red) and its adjacent luminal region (green). C, cell; L, lumen.

Fig. 11.

Effect of capacitative Ca2+ entry on the [Ca2+]i increase and net acid output in isolated gastric glands. Gastric glands were continuously perfused with RM containing different Ca2+ concentrations, as indicated in the box. Glands were treated with 2 μM thapsigargin in Ca2+-free medium and then the extracellular Ca2+ concentration was restituted to 1.8 mM and then raised to 5 mM. Representative traces of fluo 3 fluorescence (blue) recorded in a parietal cell and LYB fluorescence measured in the parietal cell (red, purple) and its adjacent luminal region (green).

DISCUSSION

The gastric glands of the mammalian fundic mucosa are constituted by different cell types. Gastric fluid is a mixture of acid, alkali, ions, enzymes, and mucins secreted by parietal, chief, and mucous cells. These are polarized cells with an apical membrane facing the gland lumen and a basolateral side bathed by serosal fluid and bearing the receptors to different regulatory molecules. Interdispersed along the gland, endocrine and paracrine cells are to be found. All cell types are subject to fine regulation by numerous bioactive molecules originated from the neurocrine, endocrine, and paracrine systems.

In the present work we could evidence in the isolated gland preparation a diversity of cell types based on differential staining with acridine orange and LYB. Moreover, we found heterogeneity in the distribution of parietal and granule-containing cells along the gland axis. This could be further correlated with heterogeneity in the acid response to histamine and carbachol stimulation observed in different luminal regions of the gland. We observed with both agonists a progressive increase in the amplitude of LYB fluorescence change as we moved along the gland axis. This may be related to an increase in the number of parietal cells and/or the degree of cell differentiation.

Heterogeneity in the secretory response along the gland was also found with cholinergic stimulation, which displayed a complex pattern of changes in LYB fluorescence. Upon stimulation, concomitant with the Ca2+ spike, we observed an increase in acidity in the lumen. As we moved along the gland a second component of LYB fluorescence decrease progressively appeared until it totally took over the acid component (see Fig. 3). Since the probe is sensitive to acidification below pH 4, this drop in fluorescence should be due to the secretion of a product with a pH above this value and may correspond to a relative alkalinization (2, 36). Omeprazole, which inhibited the acid component of cholinergic stimulation, permitted to uncouple the two secretions, revealing only the increase in pH. This indicates that this component is independent of the activation of the H+/K+ pump in parietal cells.

The alkaline response is induced by carbachol but it was never observed with histamine stimulation, even in regions of the gland that showed the relative alkalinization with carbachol (see Fig. 4). Since carbachol stimulates both parietal and nonparietal cells and histamine H2 receptors are exclusively expressed by parietal cells in the gland, we think this alkaline secretion could be originated in nonparietal cells (44). Alternatively, carbachol but not histamine may induce alkaline secretion by the parietal cell. In this sense, an interesting and controversial apical protein, the Cl/HCO3 exchanger PAT-1 (SLC26A6), has been localized at the apical surface of the parietal cell in close proximity to the gastric H+-K+-ATPase (37). The authors postulate that the protein may act to buffer the secretory vesicles when they are reinternalized following acid secretion. In a new cycle of stimulation the release of vesicle contents would alkalinize the lumen and the intracellular canaliculus. However, the lack of alkalinization with histamine stimulation observed in the present work and the fact that histamine elicits Ca2+ spikes during oscillations of similar amplitude to those observed with carbachol does not support the interpretation of an alkaline secretion by the parietal cell.

It is difficult to assess at this point which cell types, other than parietal, are bordering the lumen in those regions. These cells accumulate apical secretory granules that stained with acridine orange but did not take up LYB and should correspond to chief and/or mucous neck cells (Fig. 1). In some cases, as presented in Fig. 2, the alkaline component appeared to be a releasable product, which could be exhausted by repetitive stimulation. This effect may be due to secretion of a product that has a pH above 4, buffering an acid fluid already present in the lumen. Likely candidates are the secretion of pepsinogen and/or mucin. The degranulation observed upon cholinergic stimulation of acridine orange-stained glands supports this interpretation. Unfortunately this method did not permit quantitative measurement of granule release and acid secretion alone or in conjunction with LYB fluorescence. On the other hand, the secretion of an alkaline fluid from nonparietal cells may contribute to acid buffering in the gland lumen (2). The gradual change of the acid/alkaline component along the gland detected in our system may explain the observation of a pH gradient in the gastric gland of the acid-secreting guinea pig mucosa (43).

Another contribution of this work to the understanding of the mechanisms of acid secretion stimulated by cholinergic agonists has been the study of the relationship of the different components of the [Ca2+]i response (intracellular release and Ca2+ entry) and the acid secretory response, taking advantage of our simultaneous measurements with high spatiotemporal resolution. The repetitive pulses of carbachol revealed a linear relationship between the amplitudes of the change in LYB fluorescence and the Ca2+ response. The dependence of the acid response on [Ca2+]i is further supported by the experiments using thapsigargin to release Ca2+ sequestered in the ER. A pulse of carbachol after 5 min in the presence of thapsigargin induced a residual release of Ca2+ that was not sufficient to trigger acid secretion.

The comparison between pulse and sustained stimulation with carbachol revealed different phases in the increase of LYB fluorescence that suggests that cholinergic agents activate more than one mechanism. Whereas a pulse induced a rapid and small increase of acid secretion synchronized with the intracellular Ca2+ spike, the sustained stimulation induced a complex response. This was characterized by a first component corresponding to the small and transient increase of LYB signal (as in the pulse) and a second slow and larger elevation in acid output, which occurred both in the lumen and in the intracellular canaliculi. This last phase appeared after the Ca2+ spike, coinciding with Ca2+ entry. We have interpreted the small burst of acid secreted under a pulse of carbachol as an activation of K+ and Cl channels and hence of H+/K+ pumps already present in the apical or canalicular membrane without fusion to translocate new pumps (36).

We further investigated the importance of Ca2+ entry from the extracellular space in the cholinergic response. Carbachol stimulation in Ca2+-free medium induced only a Ca2+ spike due to intracellular release from the ER. The entry component was recovered upon restitution of 1.8 mM Ca2+ in the medium, revealing a capacitative pathway activated by depletion of ER Ca2+. These components of the Ca2+ response had been described before (31). The Ca2+ spike induced by carbachol in the absence of Ca2+ was accompanied by a step increase of acid output. Upon restitution of extracellular Ca2+, acid secretion augmented both in the lumen and in the cell, generally with a slow rate. The slow activation of acid secretion was also observed with all manipulations of extracellular Ca2+ that led to Ca2+ entry. This occurred during step changes of extracellular Ca2+ in the presence of carbachol or thapsigargin. Our results suggest that released Ca2+ is able to trigger HCl secretion; however, extracellular Ca2+ is required to express the maximal potential of cholinergic stimulation (6, 12).

The slow and larger increase observed with sustained stimulation with carbachol may involve the Ca2+-induced translocation of H+/K+ pumps and the translocation and activation of Cl and K+ channels by fusion of the tubulovesicular structures with the canalicular membrane. This would imply that an increase in [Ca2+]i could induce membrane fusion. This is supported by the finding that Ca2+ and ATP can trigger homotypic membrane fusion between tubulovesicles or between tubulovesicles and liposomes (17). In parietal cells histamine is a potent stimulator of acid secretion and membrane fusion, activating mainly the cAMP-PKA pathway (22, 49). Ca2+ might also induce fusion by increasing cAMP concentration through the regulation of adenylate cyclase (13, 14). In parietal cells, fusion has been associated to PKA-dependent phosphorylation of cellular proteins involved in intracellular vesicular traffic (11, 34, 35, 49). However, to our knowledge, no changes in cAMP concentration have been reported in parietal cell stimulated by cholinergic agents.

In addition to membrane fusion, the activation of K+ and Cl channels is required for the production of HCl. There has been considerable controversy as to the nature of the K+ and Cl efflux pathways. The ClC2 Cl channel activated by PKA and low pH has been identified in parietal cells and has been associated to the stimulation of HCl secretion (45). However, the role of this protein in gastric acid secretion has been questioned (23). On the other hand, CLIC6, an intracellular Cl channel, has been identified in parietal cells in association with regions containing the H+-K+-ATPase and has been proposed to be the Cl efflux channel (30, 41). In addition, several K+ channel types have been identified in the apical pole of parietal cells from human and animals. Among these are the KCNQ1/KCNE2 (15, 20, 21, 26), the Kir 2.1 and Kir 4.1 K+ channels (19, 28). The KCNQ1/KCNE2 and Kir 2.1 channels are activated by low pH and PKA phosphorylation. In addition, the KCNQ1/KCNE2 channel appears to be also activated by Ca2+ (21). It is not unreasonable to think that more than one of these channels or even others not yet discovered may be present in the apical membrane of parietal cells. The diversity of the K+ channels present and the capacity of both Ca2+ and cAMP to induce fusion may be the key elements for the differential stimulation and synergism between cAMP and Ca2+-dependent secretagogues.

In conclusion, the primary composition of the gastric juice in the lumen of the gland is a mixture of HCl, proteins, and maybe other components whose contribution to overall pH of the fluid could be for the first time evidenced in an isolated gland preparation with high temporal resolution. Furthermore, localized changes of pH in the gland lumen were detected and interpreted to be a consequence of the heterogeneity of the cell distribution along the gland axis. Mucus secretion by mucous neck cells may serve to buffer acidity and protect gland cells. During cholinergic stimulation Ca2+ release and entry in the parietal cell are associated with the transient and sustained phases of acid response, respectively, suggesting the involvement of different mechanisms in the activation of secretion.

Acknowledgments

Present address of J. F. Perez-Zoghbi: Dept. of Cell Physiology & Molecular Biophysics, Texas Tech University Health Sciences Center, 3601 4th St., STOP 6551, Lubbock, TX 79430-6551.

Footnotes

  • * M. C. Ruiz and F. Michelangeli contributed equally to this work.

  • 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.

REFERENCES

View Abstract