A freeze-fracture study of bullfrog gastric oxyntic cells

The membrane-recruitment-and-recycling hypothesis of gastric HCl secretion

In the unstimulated oxyntic (or parietal) cell, the primary pump for gastric HCl secretion, the H+/K+-ATPase, is retained within the cytoplasm in a membranous compartment of tubulovesicles. Neural or hormonal stimulation of acid secretion induces extensive membrane transformations consistent with a fusion and recruitment of tubulovesicles to the apical plasma membrane. The consequent placement of H+/K+-ATPase in parallel with K(+) and Cl(-) channels provides the necessary ionic flow and ATP-driven exchange for net HCl secretion. Current evidence is consistent with a recruitment and recycling of membrane transporters, such as H+/K+-ATPase, through docking/fusion machinery analogous to that in many other systems.

Three-dimensional reconstruction of cytoplasmic membrane networks in parietal cells

There is general agreement that stimulation and consequent secretion of gastric parietal cells result in a great expansion of the apical canalicular membrane at the expense of an extensive intracellular network of membranes rich in the gastric proton pump (H,K-ATPase). However, there is ongoing controversy as to the precise nature of the intracellular membrane network,conventionally called tubulovesicles. At the heart of this controversy lies the question of whether tubulovesicles are a distinct membrane compartment or whether they are continuous with the apical plasma membrane.

To address this controversy we used high-pressure, rapid freezing techniques to fix non-stimulated (resting) rabbit gastric glands for electron microscopy. Ultra-thin (60-70 nm) serial sections were used for conventional TEM; 400-500 nm sections were used for tomography. Images were digitized and models constructed using Midas and Imod software(http://bio3d.colorado.edu). Images were aligned and contours drawn on specific cellular structures. The contours from a stack of serial sections were arranged into objects and meshed into 3D structures. For resting parietal cells our findings are as follows:(1) The apical canaliculus is a microvilli-decorated, branching membrane network that extends into and throughout the parietal cell. This agrees well with a host of previous studies. (2) The plentiful mitochondria form an extensive reticular network throughout the cytoplasm. This has not previously been reported for the parietal cell, and the significance of this observation and the dynamics of the mitochondrial network remain unknown. (3)H,K-ATPase-rich membranes do include membrane tubules and vesicles; however,the tubulovesicular compartment is chiefly comprised of small stacks of cisternae. Thus a designation of tubulocisternae seems appropriate; however,in the resting cell there are no continuities between the apical canaliculus and the tubulocisternae or between tubulocisternae. These data support the recruitment-recycling model of parietal cell stimulation.

Scanning EM of resting gastric parietal cells reveals a network of cytoplasmic tubules and cisternae connected to the intracellular canaliculus

Using a recently developed fixation technique for parietal cells (Sugai et al., Acta Anat Nippon 1995:70:S79, 1999:74:S101), we have reinvestigated the organization of the cytoplasmic membrane system in the resting stomach by ultra-high-resolution scanning electron microscopy (SEM). Rat gastric mucosae were microwave-fixed in cacodylate buffer [334 milliosmoles/kg H(2)O (mOsm)], to which 1.0% glutaraldehyde and 0.5% formaldehyde were added. Specimens examined by transmission electron microscopy (TEM) of thin sections revealed cytoplasm packed with tubular membranes similar to images detected by rapid-freeze/freeze-substitution fixation which is generally considered to cause minimal structural alterations. To render the cytoplasmic membranes visible by SEM, fixed mucosae were frozen, fractured, and the exposed cytoplasm of parietal cells was macerated by the aldehyde-osmium-DMSO-osmium procedure. With much of the cell matrix and filaments removed, SEM revealed numerous 30-60 nm tubules which formed a meshwork and also small cisternae. The cytoplasmic surface of the tubules was smooth while some cisternal areas had attached polyribosomes. Vesicles or isolated tubules were not found in appropriately macerated parietal cells. The cytoplasmic surface of the intracellular canaliculus was smooth except for round openings representing the bases of macerated microvilli. In favorable sites connections of the tubular membranes to the canaliculi were clearly visible. Stereo pair views were particularly useful to demonstrate these continuities. Connections between these two membrane compartments suggest the probability of rapid membrane transposition.

Ultra-high-resolution scanning electron microscopy of the continuity of cytoplasmic and luminal membranes in frog oxyntic cells

BACKGROUND

Despite numerous previous studies, the presumed continuity of the luminal and tubulovesicular membranes in frog oxyntic (oxyntico-peptic) cells remains to be convincingly demonstrated. This study was undertaken to clarify this question by ultra-high-resolution scanning electron microscopy of specially prepared frog stomach specimens before and during histamine stimulation.

METHODS

Fasted Japanese meadow frogs stimulated with histamine were used. Oxyntic cell cytoplasmic matrix was removed by the aldehyde-osmium-DMSO-osmium maceration procedure and impregnated with osmium-hydrazine. Specimens were examined without metal coating.

RESULTS

In the resting oxyntic cell, the luminal membrane had closely packed surface folds forming a rather flat surface with a few short microvilli. In the cytoplasm, flattened 200-500 nm vesicles were interconnected by numerous slender 20-60 nm tubules forming the tubulovesicular network. Occasional slender tubular branches were found in continuity with the luminal membrane. After histamine stimulation, the number and length of microvilli and surface folds increased, whereas the tubulovesicular membrane system decreased. Sites of clear continuity between the luminal and tubulovesicular membranes were not abundant but were clearly demonstrated in histamine-stimulated oxyntic cells. The small size of the tubules connecting the tubulovesicular system to the plasma membrane renders this observation by transmission electron microscopy very difficult. In these specimens, the clear continuity of the tubulovesicular network to the luminal plasma membrane became more evident.

CONCLUSIONS

This study confirms previous findings of increased luminal membrane and depletion of the tubulovesicle system. The demonstration of continuity between these two compartments in our SEM preparations supports the hypothesis of direct transfer of tubulovesicular membrane to oxyntic cell luminal secretory membrane.

Ultrastructural changes during stimulation of amphibian oxyntic cells viewed by scanning and transmission electron microscopy

Amphibian oxyntic cells exposed by cryofracture were examined by field emission scanning electron microscopy. Comparisons were made between the structure thus revealed and those seen in thin-sectioned material from the same mucosas examined by transmission electron microscopy. Resting oxyntic cells had apical surfaces which were relatively smooth with some short microvilli. Apical cytoplasm was filled with smooth membrane tubules (so-called vesicotubules). Stimulation with a combination of histamine, dibutyryl cyclic AMP, and isobutylmethylxanthine (a phosphodiesterase inhibitor) led to a dramatic elaboration (i.e., increased membrane surface area) and a decrease in number of vesicotubules in the apical cytoplasm. The surface morphology of the stimulated oxyntic cell was much different from that reported for the mammalian parietal cell. Two types of surface elaboration were observed. Most commonly the surface was formed of flattened microplicae or lingulae. An irregular surface formed by the swelling of enlarged spaces near the apical surface was also observed. These new data have been used to evaluate the models which have been proposed to explain the nature of the transition from resting to stimulated morphology. A new model, which incorporates fusion of intracellular vesicotubules with each other and also with apical membrane, is proposed. The proposed fusion process may cause an increase in membrane area open to the extracellular (luminal)solution within the cell (rather than the eversion of membranes into the gastric lumen). Expansion of spaces between the microplicae may be caused by hydroosmotic pressures developed during active HCI secretion.

Structure of oxyntic cell membranes during conditions of rest and secretion of HCl as revealed by freeze-fracture

The density and distribution of membrane associated particles of piglet oxyntic cell tubulovesicular and apical surface membranes were investigated during resting (nonsecreting) and secreting conditions. For the resting oxyntic cell, the abundant tubulovesicles showed a highly asymmetrical distribution of particles between fracture faces, with the P face heavily studded by particles and the E face particle deficient. The apical surface, however, had a relatively symmetrical distribution of particles on both membrane fracture faces. In contrast to the resting state, the apical surface of the stimulated oxyntic cell showed a marked asymmetry of membrane particles; the P face had a high density of particles, while there was a scarcity of particles on the E face. The observed changes in apical surface membrane particle distribution support the hypothesis that, following the initiation of acid secretion, the tubulovesicles fuse with and become an integral part of the apical surface. Thus, the apical membrane P face of the stimulated cell is enriched and the E face is diluted by the incorporation of tubulovesicular membranes.

Muscle proteins and the changes in shape of avian oxynticopeptic cells in relation to secretion

A combined ultrastructural and biochemical study of the avian oxynticopeptic cell was performed. Scanning electron microscopy demonstrates that this cell undergoes great changes in the shape of its apical pole in relation to secretory activity. These changes are confirmed by transmission electron microscopy and by freeze-fracture images. The biochemical finding of actin- and myosin-like proteins in high-speed supernatants of homogenates of these cells as well as the ultrastructural and cytochemical localization of actin-like filaments in their apical poles suggest a possible participation of these proteins in the above-mentioned changes. Thus, the study of cytoplasmic matrix elements and of their organization may be highly relevant in the search for a correlation between structure and function in these cells.

Potassium-stimulated ATPase activity and hydrogen transport in gastric microsomal vesicles

The Mg2+-dependent, K+-stimulated ATPase of microsomes from pig gastric mucosa has been studied in relation to observed active H+ transport into vesicular space. Uptake of fluorescent dyes (acridine orange and 9-aminoacridine) was used to monitor the generated pH gradient. Freeze-fracture electron microscopy showed that the vesicular gastric microsomes have an asymmetric distribution of intramembraneous particles (P-face was particulate; E-face was relatively smooth. Valinomycin stimulated both dye uptake and K+-ATPase (valinomycin-stimulated K+-ATPase); stimulation by valinomycin was due to increased K+ entry to some intravesicular activating site, which in turn depends upon the accompanying anion. Using the valinomycin-stimulated K+-ATPase and H+ accumulation as an index, the sequence for anion permeation was NO-3 greater than Br- greater than Cl- greater than I- greater than acetate approximately isethionate. When permeability to both K+ and H+ was increased (e.g using valinomycin plus a protonophore or nigericin), stimulation of K+-ATPase was much less dependent on the anion and the observed dissipation of the vesicular pH gradient was consistent with an 'uncoupling' of ATP hydrolysis from H+ accumulation. Thiocyanate interacts with valinomycin inhibiting the typical action of the K+ ionophore. But stimulation of ATPase activity was seen by adding 10 mM SCN- to membranes preincubated with valinomycin. From the relative activation of the valinomycin-stimulated K+-ATPase, it appears that SCN- is a very permeant anion which can be placed before NO-3 in the sequence of permeation. Valinomycin-stimulated ATPase and H+ uptake showed similar dependent correlations, including: dependence on [ATP] and [K+], pH optima, temperature activation, and selective inhibition by SH- or NH2-group reagents. These results are consistent with a pump-leak model for the gastric microsomal K+-ATPase which was simulated using Nernst-Planck conditions for passive pathways and simple kinetics for the pump. The pump is a K+/H+ exchange pump requiring K+ at an internal site. Rate of K+ entry would depend on permeability to K+ as well as the counterion, either (1) the anion to accompany K+ or (2) the H+ efflux path as an exchange ion. The former leads to net accumulation of H+ and anion, while the latter results in non-productive stimulation of ATP hydrolysis.

Orthogonal arrays of particles in plasma membranes of the gastric parietal cell

Freeze-fracture of rat gastric mucosa revealed a specific set of intramembranous particles in the plasma membrane of the parietal cells. The particles were small and of square shape and formed orthogonal arrays in the P-face with corresponding orthogonal arrays of pits in the E-face. Arrays, scattered among usual globular particles, were particularly numerous at the basal pole of the cell and less concentrated on the lateral side. They were not present in the apical microvillar membrane nor in the membranes of intracellular tubulovesicles. As in other cell types in which similar arrays were described previously (e.g., astrocytes, "light" cells of the kidney collecting tubule), their presence in parietal cell membranes suggest some specialized function of these membranes not shared by plasma membranes showing only a population of globular particles. This function has yet to be identified.

Effects of secretagogues on oxygen consumption, aminopyrine accumulation and morphology in isolated gastric glands

A recently developed method to isolate gastric glands from the rabbit gastric mucosa (Berglindh and Obrink 1976) was used to study the effects of some common gastric secretagogues. Three parameters were investigated: 1) Respiratory activity; 2) Intraglandular accumulation of the weak base aminopyrine; 3) Quantitative morphology of the parietal cells. The following substances were tested: Histamine, cAMP, db-cAMP, aminophylline, carbachol and pentagastrin. The strongest effect was obtained with db-cAMP which dose-dependently stimulated the respiration up to 200%, increased the aminopyrine accumulation 80% and altered the parietal cell morphology from a typically resting to a typically stimulated state. cAMP also stimulated the respiration but was about 10 times less effective on a molar basis than the dibutyryl form. Histamine, like db-cAMP, stimulated the respiration in a dose-dependent manner and strongly increased the aminopyrine accumulation. The morphological changes were, however, not of the same magnitude as after db-cAMP. Aminophylline, tested only for respiratory activity, stimulated the oxygen consumpation moderately. Carbachol induced a transient increase in both the oxygen consumption and in the aminopyrine accumulation with a peak value after approximately 15 minutes for both, but gave no significant morphological alterations. Pentagastrin, finally, was incapable of inducing changes in any of the three parameters. Aminopyrine was also found to accumulate approx. 50 times in unstimulated, morphologically resting glands. This seems to indicate that there might be acid sites already in resting glands.

Electrical impedance, ultrastructure and ion transport in foetal gastric mucosa

1. The voltage transient developed across rabbit foetal gastric mucosa in response to rectangular current pulses was measured under in vitro conditions. The short-circuit current was also measured.2. The voltage transients were used to calculate equivalent circuit parameters. The mucosa could be represented as two large capacitances in series, each shunted by a high conductance.3. Electron microscopy was carried out on some preparations after the electrical measurements had been made. An extensive smooth membrane tubular system was observed on the apical sides of the cells, with openings into the gastric lumen.4. Anoxia resulted in a reversible reduction in value of the large capacitance and short-circuit current: the relative changes in these two quantities were very closely linked. Electron microscopy showed a large and reversible decline in the extent of the apical tubular system during anoxia.5. It was concluded that the apical tubular system plays an important role in ion transport in this preparation. Considerations of the magnitude of the capacitances and of membrane area suggest that the smaller capacitor represents the plasma membrane. The dieletric constant of the smooth membrane of the tubular system is an order of magnitude lower than that of the plasma membrane.