THE ANATOMICAL RECORD 216:373-380 (1986) Stereological Investigations on Human Gastric Mucosa: 1. Normal Oxyntic Mucosa HERBERT F. HELANDER, ROBERT LETH, AND LARS OLBE Hassle Gastrointestinal Research Laboratories, Department of Biology, S 4 3 1 83 Molndal (H.F H.), and Department of Surgery, University of Goteborg, ,541345 Giiteborg (R.L, LO.), Sweden ABSTRACT Quantitative morphological data on normal human oxyntic mucosa were obtained from endoscopic biopsies in ten healthy male volunteers. Corpus mucosa was biopsied in the resting state and during maximal acid secretion and then processed for light and electron microscopy. Stereological analyses were carried out on sections comprising the entire thickness of the epithelial layer. About onethird of the mucosal volume was taken up by lamina propria and 15% by parietal cells. Counts of cells that displayed their nucleus in the sections revealed that a n average of 12% of the epithelial cells were parietal cells, 43% were mucous cells, 40% were zymogen cells, and 4% were endocrine cells. Parietal cells displaying two nuclei were twice as large a s those with only one nucleus. Six percent of the parietal cell volume was taken up by the nucleus, and 33% of the cytoplasmic volume was occupied by mitochondria. Stimulation of acid secretion resulted in a 76% increase in the secretory surface density; simultaneously there was a slight decrease in the mean size of the parietal cells and a n increase in the relative volume of the nucleus. During maximal stimulation of acid the parietal cells from the superficial mucosal layers displayed a 40% larger secretory surface than those from the deeper parts of the mucosa. The data, which will serve as a basis for studies of pathological mucosae, are compared with those obtained in other species. Quantitative light microscopic studies of the human Sweden. The subjects were fasted overnight, and the gastric mucosa have previously been carried out by i.a., following morning a soft double-lumen nasogastric tube Berger (19341, Card and Marks (19601, Cheng et al. was placed with its tip in the antrum. The position of (19771, Cox (1952),Naik et al. (19711, Oi et al. (19581, and the tip was controlled fluoroscopically. Residual gastric Scott (19251. Although much of the research was carried contents were discarded, and for the next 2 hours gastric out on pathological specimens, the results have been of juice was collected in 15-minute aliquots by means of a fundamental importance to our understanding of the suction pump giving negative pressure once per second. morphology and function of the normal human gastric Through a separate tube, a n aqueous solution of phenol mucosa, and they have served as a basis for further red (8 mgiliter) was continuously pumped into the stomresearch into the field of gastric pathology. ach at a rate of 225 m1/15 minutes. The amount of acid Several authors have published quantitative electron in each 15-minute sample was determined by titration microscopical data on the structure of the human gastric with 0.1 mmoliliter NaOH to pH 7.0. Using spectrophomucosa, in particular on parietal cells (Aase et al., 1983; tometric determination of the phenol red concentration, Ivey et al., 1980; Tarnawski et al., 1980, 1982; Morozov it was possible to calculate the loss of gastric juice into et al., 19751, but more comprehensive stereological infor- the duodenum; the titration data were corrected for such mation is lacking. losses. Basal acid output (BAO) was determined during a 60The aim of the present investigation was to provide further data on the structure of the normal human ox- minute period. Maximal acid secretion (MA01 was then yntic mucosa. Our interest has been focused on the cel- induced during 1 hour by a continuous intravenous inlular composition of the epithelium and on various fusion of betazole (Histalop, 162 mgh), and the two aspects of the parietal cell ultrastructure during secre- consecutive 15-minute samples of gastric juice with the highest acid content were considered to represent the tory rest and stimulation. maximal acid output. SUBJECTS AND METHODS A few weeks later, following a n overnight fast, a gasTen healthy male volunteers, aged 19-35 (Median age troscopy was performed using a n Olympus GIF I T gas23), were included in the study, which was approved by Received March 13, 1986; accepted June 9, 1986. the Ethical Committee at the University of Goteborg, 0 1986 ALAN R. LISS, INC. 374 H.F. HELANDER, R. LETH, AND L. OLBE troscope after anesthesizing the pharynx with Xylocain@ spray. The mucosa had a normal, pale colour, and there were no pathological findings either in the stomach or in the duodenum. Four mucosal biopsies were taken with a n Fb 13k forceps from the greater curvature about 10 cm distal to the cardia. The gastroscope was then removed, and Histalop was given subcutaneously (2.0 m g k g body weight) to obtain maximal acid secretion. A second gastroscopy was performed 90 minutes later, and four new biopsies were taken from the same area. Bleeding was minimal. Except for Histalop no other medication was given. The biopsies were put into the fixative within a few seconds of being taken from the mucosa. The fixative consisted of 3% formaldehyde (freshly prepared from paraformaldehyde), 4% glutaraldehyde, 0.05%picric acid in 0.16 M potassium phosphate buffer (final pH 7.2-7.4, about 1,600 mOsm/liter, 23°C). Fixation continued at room temperature for 2-3 hours, and following a short rinse in the buffer, the biopsies were postfixed for 2 hours in 1% OsO4 in the same buffer (room temperature). After another rinse, the tissue specimens were dehydrated in rising concentrations of ethanol and embedded in Polybed@. One-micron-thick sections were cut in a n ultramicrotome, stained with toluidine blue, and used for orientation in the light microscope. Although orientation of the biopsy specimens was attempted during the embedding in plastic, it was sometimes necessary to turn the embedded tissue around in order to obtain sections perpendicular to the mucosal surface. Employing a 121point test grid mounted in the eyepiece of the light microscope, the volume densities of the parietal cells, the endocrine cells, and the lamina propria were determined. For this purpose a n x 100 oil immersion objective lens was used, and measurements were carried out within a randomly chosen swath about 100 pm wide, starting from the bottom of the glands and extending to the mucosal surface. Sections were then cut about 80 nm thick for electron microscopy. These sections, which comprised the entire thickness of the mucosa, were picked up on striped copper grids and contrasted with uranyl acetate and lead hydroxide. The technically best section was selected from each biopsy specimen for electron microscopical analysis, and survey micrographs were taken at a magnification of ~220 (Fig. 1).By using the electron microscope at 40 kV, s&icient contrast was obtained so that the parietal cells could be readily identified at this low magnification. In the surveys, the volume densities for lamina propria, gland and pit lumina, epithelial cells, and parietal cells, respectively, were calculated in relation to the total volume of the mucosa (excluding the muscularis mucosae). The mucosa was then divided into five levels, parallel with the mucosal surface; level I was closest to the surface of the mucosa, and level V comprised the bottom of the glands (Fig. 1).Within each of these levels all epithelial cells that displayed their nucleus in the Fig. 1. Survey electron micrograph of oxyntic mucosa from healthy volunteer. Morphometry was carried out separately in each of the five levels that are indicated. The parietal ceils from the superficial layers of the mucosa (S) appear to have larger secretory canaliculi than those from the deeper layers (D). B, blood vessels; M, Muscularis mucosae. x 200. 375 STEREOLOGY OF NORMAL HUMAN GASTRIC MUCOSA section were assigned to one of the following categories: error. Since the nuclear volume density may vary bemucous cells (surface epithelial cells mucous neck tween the different types of epithelial cells, the nuclearcells), parietal cells, zymogen cells, endocrine cells, and biased sampling may also introduce a small error in the unidentified cells. The surface mucous cells and mucous determination of the percentage of each type of epitheneck cells were grouped into the same mucous cell cate- lial cell. The surface density of the secretory membrane was gory, since it was often not possible to distinguish between them with certainty. Zymogen cells dominated assessed in relation to the cell volume. Finally, the data the base of the glands, but were also found high up in for cell profile areas were calculated. For more specific the gland necks. In this position they were sometimes information on the stereological methods and the errors difficult to separate from mucous cells. However, the associated with them, see Weibel (1979) and Helander two types of cells were distinguished by their secretory (1976).Systematic errors owing to section thickness were granules: biphasic granules (Spicer et al., 1978) were not corrected. Student’s t-test was used to assess the considered characteristic of the mucous cells, whereas significance of any differences; whenever possible pairmonophasic granules were typical of the zymogen cells. wise testing was employed. In the electron microscopical survey an area was then RESULTS chosen randomly, limited by the bottom of the glands, The ten subjects had a mean BAO of 3.8 mmol/30 min the mucosal surface, and two lines 50-250 pm apart and roughly parallel to the glands. Within this swath each (range 0.5-8.2) and a mean MA0 of 20.1 mmol/30 min parietal cell that displayed the nucleus in the section (range 12.4-33.4). The biopsies comprised the entire was photographed; there were usually 10-15 parietal thickness of the epithelial layer and, in most cases, also cells within such an area. The procedure was repeated parts of the muscularis mucosae. There were no signs of with each biopsy so that the subsequent analysis com- poor fixation, such as fragmentation of the mitochonprised 598 parietal cells: 45-71 in each of the ten per- drial cristae or gross confluence of zymogen granules. sons. About half of these cells were from resting The lamina propria contained normal amounts of cellumucosae; the other half was from stimulated mucosae. lar elements, and there were no obvious indications of Normally, three biopsies were analyzed from each rest- gastritis. The general structure of the mucosa was similar to that previously described in man and in other ing and stimulated mucosa. The stereological analysis was carried out on coded mammals (Helander, 1981). The data on the composition of the oxyntic mucosa are paper prints (final magnification about 11,000). Using Weibel‘s multipurpose grid (about 140 test points per summarized in Table 1,together with comparable data 100 pm2) as an overlay, the mitochondria1 volume den- from some other studies. The lamina propria occupied sity was estimated in relation to the cytoplasmic vol- about one-third of the mucosal volume, and the parietal ume. The nuclear volume density was estimated in cells occupied 15%; analysis of variance demonstrated relation to the cell volume; because of the nuclear-biased that the variance was larger between the subjects than sampling the data for nuclear volume density is overes- within the subjects (F values 2.5 and 3.2, respectively). timated; the obtained volumes can be corrected using About 12% of the epithelial cells that exhibited the the method described by Konwinski and Kozlowski nucleus were parietal cells. Studies on the distribution (1972).Thus, a measured nuclear volume density of 11% of the parietal cells showed that most of these cells are would, after correction, correspond to some 6%. This found in the middle layers of the mucosa (Fig. 2). A few of the parietal cells (- 4%) and of the zymogen method, however, was intended for spherical cells and nuclei, and for this reason the corrected figures for nu- cells (- 1%)exhibited two nuclear profiles in the secclear volume density may also be subject to a slight tions. The profile areas of these “binucleate” parietal + TABLE 1. Comparative data on the composition of the oxyntic mucosa in several species (means) Percentage of: Man Mucosal Mucosal No. of volume* volume1* epithelial unstimulated stimulated cells1* Parietal cells Mucoid cells Zymogen cells Endocrine cells Lamina propria Gland and pit lumina Reference 15.0 f 0.7l 15.7 0.72 0.3 f 0.l2 No. of gland cells2 (mouse) Mucosal volume’ (rat) Mucosal volume’ (rat) No. of gland cells2 (cow) 19.3 30 - 19 51 15.2 f 0.9 11.6 k 0.9 19.7 15.1 - 42.9 +_ 2.0 40.0 k 2.8 4.0 k 0.5 - - 13.8 - - - + 2.1 33.6 + 1.9’ 29.7 + 1.5l 1.9 0.21 32.9 This study This study This study Chen and Withers (1975) + ‘By electron microscopy. 2By light microscopy. *Means f SEM, n = 10. 3.0 f 0.4 - - - 48.1 34.3 Helander (1976) Blom and Helander (1981) - Kapp (1967) H.F. HELANDER, R. LETH, AND L. OLBE 376 T level the mucosa in the stimulated state demonstrated that the secretory surface density was significantly higher (P< .002) in the two superficial layers than in the three more basal layers (2.11 f 0.15 vs. 1.58 f 0.06, means k SEM, n f 10, see also Fig. 5). No such difference was recorded in the resting mucosae. The cell profile area increased slightly with the distance from the mucosal surface. Thus regression analysis showed a positive correlation between cell profile area and mucosal levels, but the correlation coefficents were only + 0.54 (P < .001) for the resting and + 0.33 (P = .04) for the stimulated mucosae, respectively. No significant level differences were recorded for nuclear volume density or mitochondria1volume density. No significant correlations were recorded between BAO and the morphological data from the resting mucosae, neither were any of the morphological data from the stimulated biopsies significantly correlated to the MAO. Also the increase in secretory surface density and the increase in acid production, which occur upon stimulation, were not significantly correlated. DISCUSSION % Fig. 2. Block diagram showing the distribution of parietal cells (above abscissa) and of the epithelial cells (below the abscissa) in the five levels of the gastric mucosa. Level I comprises those 20% of mucosa that are closest t o the gastric surface, and level V comprises the 20% at the bottom of the glands. The ordinate shows the parietal cells in percentage of the number of epithelial cells in the respective level (above abscissa), and the epithelial cells in percentage of the total number of epithelial cells (below abscissa). cells averaged 272 pm2 (vs. 172 pm2 in the mononucleate parietal cells) in the resting mucosae and 259 pm2 (vs. 169pm2for nonnucleate parietal cells) in the stimulated mucosae. The mean values of the bi- and mononucleate cells are significantly different (P < .001). In the resting state the mean cell profile area averaged 180 pm2; in the stimulated state there was a small, but significant decrease to 169 pm2 (Table 2). Upon stimulation, the measured nuclear volume density increased slightly, but significantly, from 10.4% in the resting state to 11.6% of the parietal cell volume (after correction for nuclear-biased sampling these figures are 5.6% and 6.6%, respectively). Also the secretory surface density rose significantly from 1.01 in the resting to 1.78 pm2/pm3in the stimulated state. (see also Figs. 3 and 4). Closer analysis of the data from the different levels of Considerable experimentation with various fixation procedures was carried out before adopting the techniques used in the present study. Glutaraldehyde alone, in buffered solution, did not result in adequate fixation. Also fixation at temperatures below 20°C gave unsatisfactory results. The fixative that was finally adopted was modified only slightly from that suggested by It0 and Karnovsky (1968), and it has an osmolarity more than five times higher than that of blood. For this reason, the cells might shrink during the fixation, and this must be borne in mind when comparing the stereological data with those from other investigations. It deserves to be emphasized that the processing of the tissue specimens results in net changes in the cell volumes. The magnitude of these changes has not been estimated, but in all likelihood it is similar in the stimulated and in the unstimulated mucosae. Therefore, the reader should pay more attention to relative changes between the stimulated and unstimulated mucosae than to the absolute values. The most important difference between the resting and the stimulated parietal cells relates to the secretory surface. An increase in this area during stimulation has previously been observed both in man and in other mammals. The magnitude of this increase, however, varies between different species: in dogs the increase is between 4 and 10 times (Helander and Hirschowitz, 1972; Zalewsky and Moody, 1977), and in the rat it is 60% (Helander, 1976). In the present study the recorded increase averaged 76%. It seems probable that the rates of increase are inversely correlated to the basal acid secretion: in the dog it is low, and in man and the rat it is fairly high. Following stimulation of acid secretion there is also a slight, but significant, reduction in the size of the parietal cells. This would be expected, since an exocytotic secretory mechanism has been postulated for these cells (Kurosumi et al., 1958; Helander and Hirschowitz, 1972), whereby tubulovesicles in the cytoplasm would merge with the secretory membrane and release their contents in the lumen. Differences between parietal cells from different levels of the oxyntic mucosa have previously been described in STEREOLOGY OF NORMAL HUMAN GASTRIC MUCOSA 377 Fig. 3. Survey of parietal cell from a n unstimulated person. This cell was found in the bottom half of the mucosa. C, secretory canaliculus; L, lumen of oxyntic gland; LY, lysosome-like body; M, mitochondrion; N, cell nucleus; V, tubulovesicles. x9.000. several species (Helander, 1981). In the rat the superficial parietal cells-referred to as neck parietal cells-are significantly larger and have a larger secretory surface than the parietal cells at the base of the glands (Helander, 1976; Helander and Sundell, 1984; Jacobs and Sturdevant, 1982). In contrast, Morozov (1976), who studied healthy human subjects, observed the largest secretory surface in the parietal cells from the midlevel of the mucosa, with significantly smaller values for those closer to the mucosal surface and those from the bottom of the glands. He also found the highest mitochondrial volume density in parietal cells of the midlevel of the mucosa. Morozov’s findings cannot be confirmed in the present study. The differences between parietal cells from the neck and from the base presumably reflect differences in age: new parietal cells arise from mitoses in stem cells or mucous cells located in the neck of the glands. Most of these cells then move towards the bottom of the glands, where they die (Willems et al., 1972).The half-life of the Darietal cells has been calculated in mice, where it ;mounts to about 23 days (Ragins et al., 1968). In man the replacement of the parietal cell population may take 1to several years (Lipkin, 1972). The large secretory surface observed in the neck parietal cells of the stimulated mucosae may indicate a higher capacity of these cells to produce acid than in the base parietal cells. This hypothesis is supported by histochemical data from mice (Coulton and Firth, 1983), where higher activities of malate dehydrogenase (MDHase), nicotine adenine dinucleotide-linked isocitrate dehydrogenase (NAD-ICDHase)(these two enzymes are located in the matrix of the mitochondria), and pnitrophenyl phosphatese (pNPPase) are observed in the superficial parietal cells than in those from the bottom of the glands. It should be added, though, that for succinic dehydrogenase (SDHase, an enzyme located to the inner mitochondrial membranes) the results were the opposite. From a methodological point of view, it is important to note the differences between the parietal cells from different levels of the oxyntic mucosa. Sampling for morphometry should include parietal cells from all the lev&; this has not been the case in most previous studies, 378 H.F. HELANDER, R. LETH, AND L. OLBE Fig. 4. Parietal cell from the same person as in Figure 3 during maximal acid secretion. This cell was observed in the bottom half of the mucosa. The secretory canaliculi (C) are much more expanded than in the unstimulated state, and there appear to be more microvilli. L, gland lumen; N, nucleus. X9,OOO. and therefore the mean stereological data from these studies are biased. This should be borne in mind when the comparative data in Tables 1and 2 are studied. Polynucleate parietal cells have been described previously (Card and Marks, 1960), but their size has not been compared with those of the mononucleate parietal cells. Assuming that bi- and mononucleate parietal cells are similarly shaped, a rough comparison of their mean sizes would be obtained by dividing their mean cell profile areas and raising the quotient to the 3/2 power. For the resting mucosae the volume proportion would be 272/176)3/2= 1.92, and for the stimulated mucosae (259/168)3/2 = 1.91. It would thus seem that the “binu- cleate” parietal cells are about twice as large as the mononucleate ones. A similar difference in size between mono- and binucleate cells has previously been observed for the zymogen cells of the rat gastric mucosa (Helander, 1978). In previous light microscopic studies, a correlation was noted between the capacity to produce acid and the “parietal cell mass.” Card and Marks (1960) calculated that, in their patients with gastroduodenal disease, there was a highly significant correlation between the number of parietal cells and the estimated maximal acid output of the resected portion of the stomach. In that study, lo9 parietal cells would produce -23 mmol/h of 379 STEREOLOGY OF NORMAL HUMAN GASTRIC MUCOSA acid. A similar study on Chinese duodenal ulcer patients demonstrated a corresponding correlation with the acid secreting capacity of 20 m m o m per lo9 parietal cells (Cheng et al., 1977). In the present study, we observed no significant correlation between the secretory and the morphological data, either in the stimulated or in the unstimulated mucosae. This does not mean that such correlation is entirely absent: as in other species, there is a significant increase in the secretory surface in parallel with a n increased acid output as a result of stimulation. However, within the group of stimulated mucosae no morphological variable was identified that correlated with MAO, and for the resting mucosae no measured microscopic structure correlated significantly with BAO. This lack of significant correlation leads us to speculate that the acid-producing capacity in normal healthy men might instead be related to the total number of parietal cells, in the same way as Card and Marks (1960) and Cheng et al. (1977) found for patients with gastroduodenal disease. Until more data can be obtained on the parietal cell population, further discussion will remain purely speculative. Studies in progress will elucidate if, and how, acid secretory capacity in patients with gastro-duodenal disease correlates with parietal cell ultrastructure. - ACKNOWLEDGMENTS This study was supported by grants from the Swedish Medical Research Council (project N. 17 x -7601, from the Medical Faculty, University of Goteborg, and from Go- *W +I Q) 2 (0 +I 0 2 I I I I I I I II 111 IV v Fig. 5. Diagram showing the secretory surface density (in pm2/pm3 of parietal cell volume, means k SEM, n = 10) for the stimulated mucosae ( 0 )and the unstimulated mucosae (A).Level I is closest to the mucosal surface, level V is at the bottom of the glands. 380 H.F. HELANDER, R. LETH, AND L. OLBE teborg Medical Society (211/83). Part of this work has been presented previously (Gastroenterology 1983; 84:1184). LITERATURE CITED Aase, S., E. Dahl, R. Roland, and R. Hars (1983)Morphometric studies of parietal cells during basal conditions and during stimulation with pentagastrin in healthy subjects. Scan. J. Gastroenterol, 28t919-923. Berger, E.H. (1934)The distribution of parietal cells in the stomach: A histotopographic study. Am. J. Anat., 54:87-114. Blom, H., and H.F. Helander (1981) Quantitative ultrastructural studies on parietal cell regeneration in experimental ulcers in rat gastric mucosa. Gastroenterology, 80:334-343. Card, W.I., and I.N. Marks (1960)The relationship between acid output of the stomach following “maximal” histamine stimulation and the parietal cell mass. Clin. Sci., 29t147-163. Chen, K.Y., and H.R. Withers (1975)Proliferative capability of parietal and zymogen cells. Am. J. Anat., 120~421-432. Cheng, F.C.Y., S.K. Lam, and G.B. Ong (1977) Maximum acid output to graded doses of pentagastrin and its relation to parietal cell mass in Chinese patients with duodenal ulcer. Gut, 28t827-832. Coulton, G.R., and J.A. Firth (1983) Cytochemical evidence for functional zonation of parietal cells within the gastric glands of the mouse. Histochem. J., 25~1141-1150. Cox, A.J. (1952) Stomach size and its relation to chronic peptic ulcer. AMA Arch. Pathol., 54t407-422. Frexinos, J., M. Carballido, A. Louis, and A. Ribet (1971) Effect of pentagastrin-stimulation on human parietal cells. Digest. Dist. Sci., 26:1065-1074. Helander, H.F. (1976) Stereologic changes in rat parietal cells after vagotomy and antrectomy. Gastroenterology, 71:lOlO-1018. Helander, H.F. (1978) Quantitative ultrastructural studies on rat gastric zymogen cells under different physiological and experimental condition. Cell Tissue Res. 189.287-303. Helander, H.F. (1981) The cells of the gastric mucosa. Int. Rev. Cyto., 70:217-289. Helander, H.F. and B.I. Hirschowitz (1972) Quantitative ultrastructural studies on gastric parietal cells. Gastroenterology, 87t10641071. Helander, H.F., M. Vandenvee, R. Leth, and L. Olbe (1983) Quantitative electron microscopical studies of normal human oxyntic mucose. Gastroenterology, 84t1184. Helander, H.F., and G.W. Sundell (1984) Ultrastructure of inhibited parietal cells in the rat. Gastroenterology, 87t1064-1071. Ito, S., and M.J. Karnovsky (1968)Formaldehyde fixatives containing trinitro compounds. J. Cell Biol., 39:168a. Ivey, K.J., A. Tarnawski, D. Sherman, W.J. Krause, K. Ackman, M. Burks, and J. Hewett (1980)Quantitative ultrastructural analysis of the human parietal cell during acid inhibition and increase of gastric potential difference by glucagon. Gut, 21.3-8. Jacobs, D., and R.P. Sturdevant (1982) Circadian ultrastructural changes in rat gastric parietal cells under altered feeding regimens: A morphometric study. Anat. Rec., 203~101-113. Kapp, J.P. (1967) Zur Kenntnis der Fundusdrusen im Labmagen des Hausrindes unter besondere Berucksichtigung des Lebensalters. Acta Anat. (Basel), 67r113-134. Konwinski, M., and T. Kozlowski (1972)Morphometric study of normal and phythemagglutinin-stimulated lymphocytes. Z. Zellforsch., 129.500-507. Kurosumi, K., S., Shibasaki, G. Uchida, and Y.Tanaka, (1958)Electron microscope studies on the gastric mucosa of normal rats. Arch. Histol. Japn., 25t587-624. Lipkin, M. (1972) Proliferation and differentiation of mucuscontaining cells in normal and diseased mucosa. Digestive mucous secretions. Biol. Gastroenterol., 5t500c. Morozow, I.A. (1976)Ultrastructural topography of parietal cells of the gastric mucosa (Morphometric investigation) Biull. Eksp. Biol. Med. 821390-1394. Morozow, I.A., L.A. Kowanova, V.D. Vodolagin, and M. I. Minyalienko (1975) Ultrastructure of parietal cells of the stomach and their functional activity. Biull. Eksp. Biol. Med., 79t14-18. Naik, S.R., S.C. Bajaj, R.K. Goyal, D.N. Gupta, and H.K. Chuttani (1971) Parietal cell mass in healthy human stomach. Gastroenterology, 62:682-685. Oi, M., S.Hoshiko, and S. Funatsu 91958) A studv on the distribution of parietal cells in human stomach. Jikei Med: J., 5.10-66. Ragins, H., F. Wincze, and S.M. Liu (1968) The origin and survival of parietal cells in the mouse. Anat. Rec., 262:99-110. Scott, G.H. (1925) Growth of crypts and glands of the human stomach Am. J. Dis. Child., 30:147-173. Spicer, S.S., T. Katsuyama, and P.L. Sannes (1978) Ultrastructural carbohydrate cytochemistry of gastric epithelium. Histochem. J. ZOt309-333. Tarnawski, A., K.J. Ivey, W.J. Krause, D. Sherman, M. Burks, and J. Hewett (1980) Quantitative analysis of human parietal cells after pentagastrin. Lab. Invest., 42~420-426. Tarnawski, A., K.J. hey, W.J. Krause, J. Stachura, J.E. McGuigan, B.E. Kolts, D. Sherman, and M. Burks (1982) Effect of secretin on gastric parietal cell ultrastructure in man. Lab. Invest., 46:33-38. Weihel, E.R. (1979) Stereological Methods, Vol. 1. Academic Press, London. Willems, G., P. Galand, Y. Vansteenkiste, and P. Zeitoun (1972) Cell population kinetics of zymogen and parietal cells in the stomach of mice. 2. Zellforsch., 234~505-518. Zalewsky, C., and F.G. Moody (1977) Stereological analysis of the parietal cell during acid secretion and inhibition. Gastroenterology, 73t66-74.