- Email: [email protected]
Studies on human gastric mucosal zymogen granules and their zymogens
BIOCHEMICAL
MEDICINE
30,
284-294
(1983)
Studies on Human Gastric Mucosal Zymogen Granules and Their Zymogens FARHAD NAVAB,*.’
CLINTON
B. LILLIBRIDGE,
AND MICHAEL
D. TURNER?
*Division of Gastroenteroiogy, University of Arkansas for Medical Sciences, Little Arkansas 72205, and TDepartment of Medicine, Brown University, Providence, Rhode Island 02908 Received
November
Rock.
8. 1982
Pepsinogen is stored in the chief cells in the form of zymogen granules (ZG). Animal studies indicate that the release of pepsinogens from ZG occurs after appropriate stimuli. This release is controlled in animals and man by neural and hormonal mechanisms (I ). Many agents stimulate both pepsin and hydrogen ion secretion, but others do not act in this way. For instance, secretin, which inhibits acid secretion, is a potent stimulant of pepsin secretion. In human gastric mucosa two populations of ZG have been described-one composed of small granules surrounded by a trilaminar membrane and the other composed of large granules with a unilaminar membrane (2). Samloff (3) separated human pepsinogens into two groups which appeared immunologically distinct when studied with polyclonal xenoantisera. In immunofluorescent studies, Samloff (4) localized Group I pepsinogens (PG I) in the oxyntic gland area and only found isolated groups of fluorescent chief cells in the proximal antrum. In contrast Group II pepsinogens (PC II) were found in both oxyntic and pyloric areas and also in the proximal duodenum (5). In a later study using rocket immunoelectrophoresis, PG II were found in small concentrations in the oxyntic gland area (6); however, in these studies PG I appeared to be absent from the distal pyloric gland area. Previous studies of the electrophoretic pattern of gastric acid protease zymogens have used homogenized mucosal preparations in which the granules were disrupted (7,8). Moreover, any granules escaping rupture ’ To whom reprint requests should be addressed: Division of Gastroenterology. of Arkansas for Medical Sciences, 4301 West Markham. Mail Slot 567. Little 72205 284 0006-2944/83 Copyright All rights
$3.00
c: 1983 by Academic Prrsr. Inc. of rcproductmn in any form rexrved
Ilniversit) Rock, Ark.
HUMAN
GASTRIC
MUCOSAL
ZYMOGEN
GRANULES
285
from homogenization would have been broken by the relatively low osmolality of the buffers used. It therefore appeared of interest to study preparations of isolated, and as far as possible, intact ZG. Studies of pancreatic ZG in animals show that these granules rupture easily if not prepared under ideal conditions (9). These conditions have not been worked out for human gastric ZG; thus, preliminary studies may not yield purified preparations. Such preparations would be likely to contain more refined pepsinogens with a higher specific activity than crude mucosal homogenates. The characteristics of such a preparation could be useful in increasing our understanding of the nature of human gastric mucosal acid proteases. MATERIALS
AND METHODS
Preparation of zymogen granules. Surgical specimens of stomachs removed from three patients-one with gastric cancer (Case 1) and two with benign gastric ulcer (Cases 2 and 3)-were collected in chilled solutions of 0.2111 Tris/0.2 N HCl (pH 7.4) containing 0.44 M sucrose. Mucosa was separated by blunt dissection from the body of the stomach and, in Case 3, from the distal antrum as well. Subsequently, histological examination was used to confirm the location of the mucosa. Each mucosal specimen weighed approximately 1 g. It was washed five times in the Tris buffer before being cut into small pieces and homogenized in a Teflon pestle tissue grinder (Tri-R Instruments Inc., Rockville Center, N.Y.) at 1750 rpm using 10 ml of buffer per gram wet weight of tissue. The homogenate was then filtered through four layers of cheesecloth. The filtrate was centrifuged at 500g for 10 min in a Sorvall type RC2B centrifuge (Ivan Sorvall Inc., Norwalk, Conn.); the supernatant was then removed and centrifuged at 2000g for 20 min. This was repeated twice, the temperature within the centrifuge being maintained at 4°C. All three sediments were pooled, resuspended in the Tris buffer, and passed through a scintered glass filter of medium porosity (lo-15 pm maximum porosity). One milliliter of this suspension was layered on top of a l2.5 M gradient of sucrose prepared in a Beckman gradient-forming apparatus. The material was centrifuged for 30 min at 72,000g in a Beckman ultracentrifuge, Model L2-65 (Beckman Instruments, Palo Alto, Calif.) at a temperature of 4°C. A Beckman fractionator was used to separate the sample, and each fraction was examined for the presence of ZG under the phase-contrast microscope. Assay of enzyme activity. The protease activity was determined using hemoglobin as substrate. The proteolytic activity was defined as milligrams of “tyrosine-like material” released during 1 hr of incubation at 37°C from 5 ml of a 2% acid hemoglobin substrate by 1 ml of enzyme solution (mgT/ml). The proteolytic activity of whole mucosal extracts was measured
286
NAVAB.
LILLIBRIDGE.
AND
TURNER
after mucosa had been homogenized in a 0.1 M phosphate buffer at pH 7.3 by a modification of the Anson and Mirsky method ( lo,1 1). Agar-gel electrophoresis. To determine the pepsinogen pattern. preparations of ZG and whole mucosal extracts were submitted to agar-gel electrophoresis. The buffer used was an equal mixture of 0.1 M Trisglycine and 0.1 M sodium barbital, adjusted to pH 8.3 with 0. l Y HCI. Electrophoresis was carried out for 3.5 hr using a current of 70 mA at 350 v. Electron
microscopy. The ultracentrifuge fractions which appeared to contain ZG on phase-contrast microscopy were fixed in 1.2% glutaraldehyde in 0.13 M phosphate buffer, for 36 hr at 4°C. treated with chilled 5% 0~0, in phosphate buffer for 1 hr, dehydrated in ethyl alcohol exchanged with propylene oxide, and embedded in Epon resin. Sections were made using an ultramicrotome (MT, Ivan Sorvall Inc., Norwalk, Corm. 1 placed on carbon-coated 200-mesh grids, and examined in a Phillips EM 300 transmission electron microscope at 100 kV (Philips Electronics Instruments, Inc., Mahwah, N.J.). Records were obtained of the images on 70-mm Kodalith film. RESULTS
The pathological features of the gastric resections were as follows: Case 1, a small adenocarcinoma in the distal antrum; Case 2. a single benign ulcer near to the pyloric openings: Case 3. three small ulcers on a line approximating the oxynto-pyloric junction. In each case, mucosa in juxtaposition to that used in the study was examined histologically. No evidence of significant gastritis or mucosal atrophy could be detected in these specimens. Preparation
of Zymogen
Granules
Mucosa from the oxyntic gland area of the stomach was examined from three subjects; in one, pyloric mucosa was also available. The precipitated residue obtained after the initial centrifugations at 500 and 2000g consisted of supernatant from the nuclei and cell debris; this was discarded. Following centrifugation of the supernatant from the oxyntic gland preparation at 72,000g in the sucrose gradient, several bands were visible in the tube. In contrast no bands were visible in preparations from the pyloric gland area. Oxyntic gland preparation from Cases 1 and 2 were separated into 10 fractions after centrifugation in sucrose; in Case 3 eight fractions were obtained. The fractions were numbered successively beginning with fraction 1 and the bottom of the tube. Each fraction was inspected by phasecontrast microscopy, and highly refractile granules exhibiting Brownian movement were seen in fractions 4 and 5 from Cases 1 and 2. and in fractions 5 and 6 from Case 3. These granules were not observed in
HUMAN
GASTRIC
MUCOSAL
ZYMOGEN
287
GRANULES
pyloric gland material and were also absent from oxyntic gland preparations that had been treated with Triton X prior to centrifugation. Electron micrographs taken of fractions 4 and 5 from Case 1 and fractions 5 and 6 from Case 3 showed that these fractions contained intact and fragmented ZG. However, the fractions were not pure and contained particles and fragments of mitochondria and smooth and rough endoplasmic reticulum. Two types of ZG were seen: large ZG surrounded by a single membrane (Fig. la) and smaller granules incompletely surrounded by a trilaminar membrane (Fig. lb). Enzyme Content of Zymogen
Granules
The acid proteolytic activity of fractions from the sucrose gradient centrifugation of oxyntic gland mucosa from Cases 1 and 2 is shown in Table 1. A peak of activity was observed in fractions 4 and 5. In both cases a large amount of activity was detected in the soluble fraction (fractions 8-10). Acid proteolytic activity of crude mucosal extracts of oxyntic area taken from the same stomachs as Cases 1 and 2 are shown in Table 1. Activity of normal oxyntic mucosa in gastric cancer (Case 1) was approximately three times higher than corresponding mucosa in gastric ulcer (Case 2). This ratio was maintained in material recovered in the respective soluble fractions. Activity of material obtained from the oxyntic gland area in Case 3 showed a peak in the region of fractions 4 and 5, with activities of 110 and 194 mgT/ml, respectively (Fig. 2). Pyloric mucosal extracts processed similarly had a comparatively small amount of soluble enzyme activity, 11 and 9.4 mgT/ml (Fig. 2). Agglutinated
Granules
An attempt was made to store fractions rich in isolated ZG to see if ZG retained their shape and proteolytic activity. The oxyntic gland mucosal homogenate from Case 3 was stored for 2 weeks at 4°C. After this interval a sediment was observed which, when inspected under the phase-contrast microscope, was found to contain granules with Brownian movement. TABLE ACID
PROTEOLYTIC
ACTIVITY
1
OF MUCOSAL EXTRACTS AND ZYMOGEN-CONTAINING OF OXYNTIC GLAND AREA (mgT/ml)
FRACTIONS
Fraction Case 1 2
Disease Cancer Ulcer
Mucosal extracts
1
2
3
4
5
6
7
8
9
10
3348 1324
0.3 0.2
0.6 2.6
29 9.3
73 38
46 26
19 7.6
22 13
62 7.1
12 10.8
294 80
288
NAVAB,
LILLIBRIDGE.
AND
TURNER
FIG. 1. Electron micrograph of material from fraction 4 in Case 1. Magnification = 56,000~. Note the coarse granular appearances of an intact zymogen granule. (a) Large granule partly covered with single membrane. (b) Small granule with approximately the upper half covered with a trilaminar membrane.
290
NAVAB.
1
2
LILLIBRIDGE.
3
4
AND TURNER
5
6
7
a
ULTRACENTRIFUSE FRACTIUNS FIG. 2. Concentration of acid protease mgT/ml in each fraction collected after ultracentrifugation: oxyntic gland extract (closed circles) and pyloric gland extract (open circles) in Case 3. The more dense sucrose was in fractions with the lower numbers. Note the peak in activity in fraction 5 in oxyntic mucosa, and the absence of a peak and very low activity in pyloric mucosa.
The presence of ZG in this material was later confirmed on electron microscopy. The sediment agglutinated when an attempt was made to resuspend it in Tris buffer, but following centrifugation in a sucrose gradient the preparation separated into visible bands with a peak of acid proteolytic activity in fractions 5 and 6 (Fig. 3). The soluble enzyme activity was found to have decreased by a factor of 3 when compared with the upper fractions of fresh material from the same stomach. Agar-Gel Electrophoresis Material containing ZG from the oxyntic gland area of Case 3 was examined by agar-gel electrophoresis. Each of the eight fractions collected after ultracentrifugation was placed in an individual slot. No zones of proteolytic activity were seen in ultracentrifuge fractions I and 2 from
HUMAN
GASTRIC
MUCOSAL
ZYMOGEN
GRANULES
291
ULTRACENTRIFU6E FRACTIONS FIG. 3. The concentration of acid protease mgT/ml in each fraction collected after ultracentrifugation in a sucrose density medium. The material consists of agglutinated zymogen granules (closed circles) from oxyntic mucosa and granules treated with Triton X (open circles). Note absence of peak of activity following disruption with release of soluble enzyme.
the bottom of the tube which contained the more dense sucrose. All the other fractions contained protease activity. Pepsinogens were numbered from 1 to 7 beginning with the zone with maximal anodal mobility. Enzyme activity was greatest in ZG containing fractions 4 and 5 where the zones were more prominent than soluble enzyme in fractions 7 and 8. There was a striking absence of slow-moving protease (SMP) in fractions 4 and 5; however, this zone was present in trace amounts in fractions 7 and 8. Apart from the absence of SMP the electrophoretic pattern of fractions 4 and 5 was similar to that of whole mucosal extract from the same region of the stomach, and in all pepsinogen 1 appeared prominent. The electrophoretic pattern of agglutinated ZG was the same as unagglutinated ZG. The ZG treated with Triton X showed evidence of
292
NAVAB.
LILLIBRIDGE.
AND
TURNER
proteolysis only in fractions 7 and 8. and two proteases with cathodal migration now appeared. Pyloric mucosa prepared in Case 3 in a similar manner to oxyntic mucosa snowed zones of proteolytic activity only in fractions 7 and 8. Upon addition of Triton X and recentrifugation the pattern of electrophoresis was unchanged. This was quite different from the alteration in oxyntic mucosal patterns after treatment with Triton X. The electrophoretogram of the pyloric mucosal specimen which did not contain ZG was quite different from the oxyntic specimen. Pepsinogins 2-6 and also SMP were clearly defined. The striking differences were absence of Pepsinogen I and presence of SMP. DISCUSSION
Pepsin was initially detected within the gastric mucosa by estimation of peptic activity of cell slices prepared from pig stomachs. It was found in chief cells and (in small amounts) in mucous neck cells, in both oxyntic and pyloric mucosa (12). In cat stomach, chief cells were found primarily in oxyntic mucosa. but similar cells were occasionally seen in pyloric mucosa ( 13). Zymogen granules have been detected in chief cells of bat gastric mucosa by electron microscopy (14). They appear to be discharged from chief cells after feeding fasted animals (15). In the rat apart from fasting, administration of pilocarpine and operations which destroy the normal gastroduodenal junction have been found to reduce the size of ZG (16). Ultrastructural evidence has been presented that secretin stimulates chief cells: the area occupied by ZG in samples of healthy human gastric mucosa was decreased after intravenous secretin (17). Differential centrifugation was first used to isolate ZG in canine pancreas (9). These granules were observed under the phase microscope as highly refractile bodies in Brownian motion. Material from guinea pig pancreas has also been studied by differential centrifugation. and a heterogeneous ZG fraction was seen under the electron microscope (18). Some purification of fractions was obtained using successive washings or by means of separation in a discontinuous density gradient. The only previous study involving human gastric mucosa is that of Mangla et al. (19). Using differential centrifugation they detected only one size of ZG in a patient with duodenal ulcer. In this communication we have separated intact ZG from human mucosal oxyntic gland area from surgical resections of one patient with gastric adenocarcinoma and two with benign gastric ulcer. It was necessary to use resected specimens of gastric mucosa as relatively Large amounts of mucosa are required in the preparation of ZG. The histology of mucosa used appeared normal and it seems unlikely that the characteristics of ZG from stomachs unaffected by disease would be different. It is difficult to obtain ZG free of contamination (19) and small amounts
HUMAN
GASTRIC
MUCOSAL
ZYMOGEN
GRANULES
293
of subcellular organelles may be present in fractions containing ZG. Two populations of ZG were clearly present in the centrifuged fraction on electron microscopy, confirming observations previously made by Lillibridge with sections of human gastric mucosa (2). He found one population of ZG with a small diameter and trilaminar membrane, and another larger diameter population with a single membrane in human oxyntic mucosa. Different sizes of ZG have also been reported in mice. Kataoka (20) found smaller ZG in immature chief cells located in the upper parts of the gastric glands and larger ZG in more mature chief cells in the deeper parts of the glands. In this study both large and smaller ZG were observed in the same fractions from the denser sucrose gradient. It would be of considerable interest to separate these two. Clearly this would only be possible in more purified fractions than those obtained in this study. A consistent difference between ZG and whole mucosal preparations was absent from SMP; this zone was consistently present in whole oxyntic mucosal extracts. Only small amounts of this protease were detected in fractions containing soluble enzyme. This is in agreement with the suggestion of Mangla et al. (19) that SMP, believed to be a cathepsin, originates in mucous neck cells. Using electron microscopy few mucous neck cells were found in specimens that showed little SMP. In two patients with surgical resection for benign gastric ulcer pepsinogen 1 appeared to be prominent. If this zymogen is the precursor of pepsin 1, the findings would appear to agree with Taylor (21), who showed increased amounts of this pepsin in gastric juice of both gastric and duodenal ulcer patients. Storage of ZG resulted in their agglutination which was not associated with a change of the electrophoretic pattern of the zymogens. But upon disruption two zones with cathodal migration became apparent. The significance of this finding is not certain but a similar zone has been seen in the upper intestinal mucosa, and two cathodal zones may appear in concentrated extracts of gastric adenocarcinoma (21). The only other published study regarding agglutination involves pancreatic ZG which agglutinated in frozen isotonic sucrose, but retained their morphological appearance (9). Zymogen granules were not found in pyloric mucosa in Case 3, and the electrophoretogram was strikingly different from oxyntic mucosa by the absence of pepsinogen 1 and presence of SMP. Five cell types containing secretory granules have been distinguished in guinea pig stomach by Sato and Spicer (22) using cytochemical methods. In human gastric mucosa, Samloff has detected pepsinogens in four cell types (4,5). Both PG I and PG II were found in oxyntic mucosa and were localized to chief cells and mucous neck cells. But PG I were consistently absent in the distal stomach and proximal duodenum where cells, in the pyloric glands and in Brunner’s glands, contained only PG II. Only one sample of pyloric mucosa was evaluated in this study. However,
294
NAVAB,
LILLIBRIDGE,
AND TURNER
the appearance of pepsinogens 2-5 (constituents of PG 1) as well as pepsinogens 6 and 7 (constituents of PC II) suggests that proteases, included in PG I, can be present in pyloric mucosa. It is also in accordance with our findings in mucosal biopsies for healthy individuals (23). Clearly more studies are required before this finding can be generalized. However, constituents of PG I have been reported to be present in pyloric mucosa by other investigators (24,25). SUMMARY
Intact human gastric mucosal zymogen granules (ZG) were detected in specimens from surgical resections of one patient with gastric adenocarcinoma and two with benign gastric ulcer. Both large ZG with unilaminar membranes and smaller ZG with trilaminar membranes were identified by electron microscopy. The zymogens in the ZG and in mucosal extracts were separated by gel electrophoresis. Slow-moving Protease (SMP) was seen in the whole mucosal extracts but was absent from ZG. One specimen of pyloric mucosa showed a striking absence of ZG. Despite the absence of ZG, pyloric mucosa showed Pepsinogens 2-5 (constituents of PG I) as well as Pepsinogens 6 and 7 (constituents of PG II) and SMP. REFERENCES 1. Hirschowitz, B. 1.. in “Handbook of Physiology” (C. F. Code and W. Heidel, Eds.). Sect. 6, Vol. 2. p. 889. Williams & Wilkins, Baltimore, 1967. 2. Lillibridge, E. C., J. Eiophys. B&hem. C~fol. 10, 145 (1961). 3. Samloff. I. M.. Gasrroenrerology 60, 586 (1971). 4. Samloff. I. M., Gasrroenterology 66, 185 (1971). 5. Samloff. I. M., and Liebman, W. N.. GuAtroenferology 63, 36 (1973). 6. Libman, L. J., and Samloff, I. M., Gastroenrerology 74, 1132 (1978). 7. Samloff, I. M., Gastroenterology 57, 6S9 (1969). 8. Etherington, D. J.. and Taylor, W. H.. B&hem. J. 118, 587 (1970). 9. Hokin, L. E., Biochim. Biophys. Acra 18, 379 (1955). 10. Seijffers, M. J.. Segal. H. L., and Miller, L. L.. Amer. J. Physiol. 207, x (1964). 11. Anson. M. L.. and Mirsky, A. E.. J. Gen. Physiol. 16, 59 (1932). 12. Holter, H., and Linderstrom-Lang, K.. Hoppe-Seyler’s Z. Physio/. C’hem. 226 149 (1934). 13. Bowie. D. J., and Vineberg. A. M., Anaf. Ret,. 64, 357 (1935). 14. Ito. S., and Winchester, R. G.. J. Cell. Biol. 16, 541 (1963). 15. Menzies, G., J. Pathol. Bacterial. 83, 475 (1962). 16. Helander, H. F.. Cell Tis. Res. 189, 287 (1978). 17. Stachura. J., ivey, K. J., Tarnawski, A., Krause. W. J.. and Stogsdill. P.. Scu&. J. Gastroenterol. 16, 713 (1981). 18. Siekevitz. P., and Palade. G. E., J. Biophys. Biochem. gyro/. 4, 203 (1958). 19. Mangla, J. C., Navab, F., Lillibridge, C. B.. and Turner, M. D.. B&hem. Med. 13, 184 (1975). 20. Kataoka, K. Arch. Histol. Japan. 32, 251 (1970). 21. Taylor, W. H., Nature (London) 227, 76 (1970). 22. Sato, A., and Spicer, S. S.. Amer. J. Anat. 159, 307 (1980). 23. Navab, F., Rubio. C. E.. and Turner. M. D.. Gastroenteroloyy 84, 1257 (1983). 24. Whitecross, D. P.. Armstrong, C.. Clarke, A. D.. and Piper, D. W., Cur 14, 850 (1973). 25. Kojima. K.. and Moriga, M.. Gastroenleroi Jq~on 13, 421 (1978).
MEDICINE
30,
284-294
(1983)
Studies on Human Gastric Mucosal Zymogen Granules and Their Zymogens FARHAD NAVAB,*.’
CLINTON
B. LILLIBRIDGE,
AND MICHAEL
D. TURNER?
*Division of Gastroenteroiogy, University of Arkansas for Medical Sciences, Little Arkansas 72205, and TDepartment of Medicine, Brown University, Providence, Rhode Island 02908 Received
November
Rock.
8. 1982
Pepsinogen is stored in the chief cells in the form of zymogen granules (ZG). Animal studies indicate that the release of pepsinogens from ZG occurs after appropriate stimuli. This release is controlled in animals and man by neural and hormonal mechanisms (I ). Many agents stimulate both pepsin and hydrogen ion secretion, but others do not act in this way. For instance, secretin, which inhibits acid secretion, is a potent stimulant of pepsin secretion. In human gastric mucosa two populations of ZG have been described-one composed of small granules surrounded by a trilaminar membrane and the other composed of large granules with a unilaminar membrane (2). Samloff (3) separated human pepsinogens into two groups which appeared immunologically distinct when studied with polyclonal xenoantisera. In immunofluorescent studies, Samloff (4) localized Group I pepsinogens (PG I) in the oxyntic gland area and only found isolated groups of fluorescent chief cells in the proximal antrum. In contrast Group II pepsinogens (PC II) were found in both oxyntic and pyloric areas and also in the proximal duodenum (5). In a later study using rocket immunoelectrophoresis, PG II were found in small concentrations in the oxyntic gland area (6); however, in these studies PG I appeared to be absent from the distal pyloric gland area. Previous studies of the electrophoretic pattern of gastric acid protease zymogens have used homogenized mucosal preparations in which the granules were disrupted (7,8). Moreover, any granules escaping rupture ’ To whom reprint requests should be addressed: Division of Gastroenterology. of Arkansas for Medical Sciences, 4301 West Markham. Mail Slot 567. Little 72205 284 0006-2944/83 Copyright All rights
$3.00
c: 1983 by Academic Prrsr. Inc. of rcproductmn in any form rexrved
Ilniversit) Rock, Ark.
HUMAN
GASTRIC
MUCOSAL
ZYMOGEN
GRANULES
285
from homogenization would have been broken by the relatively low osmolality of the buffers used. It therefore appeared of interest to study preparations of isolated, and as far as possible, intact ZG. Studies of pancreatic ZG in animals show that these granules rupture easily if not prepared under ideal conditions (9). These conditions have not been worked out for human gastric ZG; thus, preliminary studies may not yield purified preparations. Such preparations would be likely to contain more refined pepsinogens with a higher specific activity than crude mucosal homogenates. The characteristics of such a preparation could be useful in increasing our understanding of the nature of human gastric mucosal acid proteases. MATERIALS
AND METHODS
Preparation of zymogen granules. Surgical specimens of stomachs removed from three patients-one with gastric cancer (Case 1) and two with benign gastric ulcer (Cases 2 and 3)-were collected in chilled solutions of 0.2111 Tris/0.2 N HCl (pH 7.4) containing 0.44 M sucrose. Mucosa was separated by blunt dissection from the body of the stomach and, in Case 3, from the distal antrum as well. Subsequently, histological examination was used to confirm the location of the mucosa. Each mucosal specimen weighed approximately 1 g. It was washed five times in the Tris buffer before being cut into small pieces and homogenized in a Teflon pestle tissue grinder (Tri-R Instruments Inc., Rockville Center, N.Y.) at 1750 rpm using 10 ml of buffer per gram wet weight of tissue. The homogenate was then filtered through four layers of cheesecloth. The filtrate was centrifuged at 500g for 10 min in a Sorvall type RC2B centrifuge (Ivan Sorvall Inc., Norwalk, Conn.); the supernatant was then removed and centrifuged at 2000g for 20 min. This was repeated twice, the temperature within the centrifuge being maintained at 4°C. All three sediments were pooled, resuspended in the Tris buffer, and passed through a scintered glass filter of medium porosity (lo-15 pm maximum porosity). One milliliter of this suspension was layered on top of a l2.5 M gradient of sucrose prepared in a Beckman gradient-forming apparatus. The material was centrifuged for 30 min at 72,000g in a Beckman ultracentrifuge, Model L2-65 (Beckman Instruments, Palo Alto, Calif.) at a temperature of 4°C. A Beckman fractionator was used to separate the sample, and each fraction was examined for the presence of ZG under the phase-contrast microscope. Assay of enzyme activity. The protease activity was determined using hemoglobin as substrate. The proteolytic activity was defined as milligrams of “tyrosine-like material” released during 1 hr of incubation at 37°C from 5 ml of a 2% acid hemoglobin substrate by 1 ml of enzyme solution (mgT/ml). The proteolytic activity of whole mucosal extracts was measured
286
NAVAB.
LILLIBRIDGE.
AND
TURNER
after mucosa had been homogenized in a 0.1 M phosphate buffer at pH 7.3 by a modification of the Anson and Mirsky method ( lo,1 1). Agar-gel electrophoresis. To determine the pepsinogen pattern. preparations of ZG and whole mucosal extracts were submitted to agar-gel electrophoresis. The buffer used was an equal mixture of 0.1 M Trisglycine and 0.1 M sodium barbital, adjusted to pH 8.3 with 0. l Y HCI. Electrophoresis was carried out for 3.5 hr using a current of 70 mA at 350 v. Electron
microscopy. The ultracentrifuge fractions which appeared to contain ZG on phase-contrast microscopy were fixed in 1.2% glutaraldehyde in 0.13 M phosphate buffer, for 36 hr at 4°C. treated with chilled 5% 0~0, in phosphate buffer for 1 hr, dehydrated in ethyl alcohol exchanged with propylene oxide, and embedded in Epon resin. Sections were made using an ultramicrotome (MT, Ivan Sorvall Inc., Norwalk, Corm. 1 placed on carbon-coated 200-mesh grids, and examined in a Phillips EM 300 transmission electron microscope at 100 kV (Philips Electronics Instruments, Inc., Mahwah, N.J.). Records were obtained of the images on 70-mm Kodalith film. RESULTS
The pathological features of the gastric resections were as follows: Case 1, a small adenocarcinoma in the distal antrum; Case 2. a single benign ulcer near to the pyloric openings: Case 3. three small ulcers on a line approximating the oxynto-pyloric junction. In each case, mucosa in juxtaposition to that used in the study was examined histologically. No evidence of significant gastritis or mucosal atrophy could be detected in these specimens. Preparation
of Zymogen
Granules
Mucosa from the oxyntic gland area of the stomach was examined from three subjects; in one, pyloric mucosa was also available. The precipitated residue obtained after the initial centrifugations at 500 and 2000g consisted of supernatant from the nuclei and cell debris; this was discarded. Following centrifugation of the supernatant from the oxyntic gland preparation at 72,000g in the sucrose gradient, several bands were visible in the tube. In contrast no bands were visible in preparations from the pyloric gland area. Oxyntic gland preparation from Cases 1 and 2 were separated into 10 fractions after centrifugation in sucrose; in Case 3 eight fractions were obtained. The fractions were numbered successively beginning with fraction 1 and the bottom of the tube. Each fraction was inspected by phasecontrast microscopy, and highly refractile granules exhibiting Brownian movement were seen in fractions 4 and 5 from Cases 1 and 2. and in fractions 5 and 6 from Case 3. These granules were not observed in
HUMAN
GASTRIC
MUCOSAL
ZYMOGEN
287
GRANULES
pyloric gland material and were also absent from oxyntic gland preparations that had been treated with Triton X prior to centrifugation. Electron micrographs taken of fractions 4 and 5 from Case 1 and fractions 5 and 6 from Case 3 showed that these fractions contained intact and fragmented ZG. However, the fractions were not pure and contained particles and fragments of mitochondria and smooth and rough endoplasmic reticulum. Two types of ZG were seen: large ZG surrounded by a single membrane (Fig. la) and smaller granules incompletely surrounded by a trilaminar membrane (Fig. lb). Enzyme Content of Zymogen
Granules
The acid proteolytic activity of fractions from the sucrose gradient centrifugation of oxyntic gland mucosa from Cases 1 and 2 is shown in Table 1. A peak of activity was observed in fractions 4 and 5. In both cases a large amount of activity was detected in the soluble fraction (fractions 8-10). Acid proteolytic activity of crude mucosal extracts of oxyntic area taken from the same stomachs as Cases 1 and 2 are shown in Table 1. Activity of normal oxyntic mucosa in gastric cancer (Case 1) was approximately three times higher than corresponding mucosa in gastric ulcer (Case 2). This ratio was maintained in material recovered in the respective soluble fractions. Activity of material obtained from the oxyntic gland area in Case 3 showed a peak in the region of fractions 4 and 5, with activities of 110 and 194 mgT/ml, respectively (Fig. 2). Pyloric mucosal extracts processed similarly had a comparatively small amount of soluble enzyme activity, 11 and 9.4 mgT/ml (Fig. 2). Agglutinated
Granules
An attempt was made to store fractions rich in isolated ZG to see if ZG retained their shape and proteolytic activity. The oxyntic gland mucosal homogenate from Case 3 was stored for 2 weeks at 4°C. After this interval a sediment was observed which, when inspected under the phase-contrast microscope, was found to contain granules with Brownian movement. TABLE ACID
PROTEOLYTIC
ACTIVITY
1
OF MUCOSAL EXTRACTS AND ZYMOGEN-CONTAINING OF OXYNTIC GLAND AREA (mgT/ml)
FRACTIONS
Fraction Case 1 2
Disease Cancer Ulcer
Mucosal extracts
1
2
3
4
5
6
7
8
9
10
3348 1324
0.3 0.2
0.6 2.6
29 9.3
73 38
46 26
19 7.6
22 13
62 7.1
12 10.8
294 80
288
NAVAB,
LILLIBRIDGE.
AND
TURNER
FIG. 1. Electron micrograph of material from fraction 4 in Case 1. Magnification = 56,000~. Note the coarse granular appearances of an intact zymogen granule. (a) Large granule partly covered with single membrane. (b) Small granule with approximately the upper half covered with a trilaminar membrane.
290
NAVAB.
1
2
LILLIBRIDGE.
3
4
AND TURNER
5
6
7
a
ULTRACENTRIFUSE FRACTIUNS FIG. 2. Concentration of acid protease mgT/ml in each fraction collected after ultracentrifugation: oxyntic gland extract (closed circles) and pyloric gland extract (open circles) in Case 3. The more dense sucrose was in fractions with the lower numbers. Note the peak in activity in fraction 5 in oxyntic mucosa, and the absence of a peak and very low activity in pyloric mucosa.
The presence of ZG in this material was later confirmed on electron microscopy. The sediment agglutinated when an attempt was made to resuspend it in Tris buffer, but following centrifugation in a sucrose gradient the preparation separated into visible bands with a peak of acid proteolytic activity in fractions 5 and 6 (Fig. 3). The soluble enzyme activity was found to have decreased by a factor of 3 when compared with the upper fractions of fresh material from the same stomach. Agar-Gel Electrophoresis Material containing ZG from the oxyntic gland area of Case 3 was examined by agar-gel electrophoresis. Each of the eight fractions collected after ultracentrifugation was placed in an individual slot. No zones of proteolytic activity were seen in ultracentrifuge fractions I and 2 from
HUMAN
GASTRIC
MUCOSAL
ZYMOGEN
GRANULES
291
ULTRACENTRIFU6E FRACTIONS FIG. 3. The concentration of acid protease mgT/ml in each fraction collected after ultracentrifugation in a sucrose density medium. The material consists of agglutinated zymogen granules (closed circles) from oxyntic mucosa and granules treated with Triton X (open circles). Note absence of peak of activity following disruption with release of soluble enzyme.
the bottom of the tube which contained the more dense sucrose. All the other fractions contained protease activity. Pepsinogens were numbered from 1 to 7 beginning with the zone with maximal anodal mobility. Enzyme activity was greatest in ZG containing fractions 4 and 5 where the zones were more prominent than soluble enzyme in fractions 7 and 8. There was a striking absence of slow-moving protease (SMP) in fractions 4 and 5; however, this zone was present in trace amounts in fractions 7 and 8. Apart from the absence of SMP the electrophoretic pattern of fractions 4 and 5 was similar to that of whole mucosal extract from the same region of the stomach, and in all pepsinogen 1 appeared prominent. The electrophoretic pattern of agglutinated ZG was the same as unagglutinated ZG. The ZG treated with Triton X showed evidence of
292
NAVAB.
LILLIBRIDGE.
AND
TURNER
proteolysis only in fractions 7 and 8. and two proteases with cathodal migration now appeared. Pyloric mucosa prepared in Case 3 in a similar manner to oxyntic mucosa snowed zones of proteolytic activity only in fractions 7 and 8. Upon addition of Triton X and recentrifugation the pattern of electrophoresis was unchanged. This was quite different from the alteration in oxyntic mucosal patterns after treatment with Triton X. The electrophoretogram of the pyloric mucosal specimen which did not contain ZG was quite different from the oxyntic specimen. Pepsinogins 2-6 and also SMP were clearly defined. The striking differences were absence of Pepsinogen I and presence of SMP. DISCUSSION
Pepsin was initially detected within the gastric mucosa by estimation of peptic activity of cell slices prepared from pig stomachs. It was found in chief cells and (in small amounts) in mucous neck cells, in both oxyntic and pyloric mucosa (12). In cat stomach, chief cells were found primarily in oxyntic mucosa. but similar cells were occasionally seen in pyloric mucosa ( 13). Zymogen granules have been detected in chief cells of bat gastric mucosa by electron microscopy (14). They appear to be discharged from chief cells after feeding fasted animals (15). In the rat apart from fasting, administration of pilocarpine and operations which destroy the normal gastroduodenal junction have been found to reduce the size of ZG (16). Ultrastructural evidence has been presented that secretin stimulates chief cells: the area occupied by ZG in samples of healthy human gastric mucosa was decreased after intravenous secretin (17). Differential centrifugation was first used to isolate ZG in canine pancreas (9). These granules were observed under the phase microscope as highly refractile bodies in Brownian motion. Material from guinea pig pancreas has also been studied by differential centrifugation. and a heterogeneous ZG fraction was seen under the electron microscope (18). Some purification of fractions was obtained using successive washings or by means of separation in a discontinuous density gradient. The only previous study involving human gastric mucosa is that of Mangla et al. (19). Using differential centrifugation they detected only one size of ZG in a patient with duodenal ulcer. In this communication we have separated intact ZG from human mucosal oxyntic gland area from surgical resections of one patient with gastric adenocarcinoma and two with benign gastric ulcer. It was necessary to use resected specimens of gastric mucosa as relatively Large amounts of mucosa are required in the preparation of ZG. The histology of mucosa used appeared normal and it seems unlikely that the characteristics of ZG from stomachs unaffected by disease would be different. It is difficult to obtain ZG free of contamination (19) and small amounts
HUMAN
GASTRIC
MUCOSAL
ZYMOGEN
GRANULES
293
of subcellular organelles may be present in fractions containing ZG. Two populations of ZG were clearly present in the centrifuged fraction on electron microscopy, confirming observations previously made by Lillibridge with sections of human gastric mucosa (2). He found one population of ZG with a small diameter and trilaminar membrane, and another larger diameter population with a single membrane in human oxyntic mucosa. Different sizes of ZG have also been reported in mice. Kataoka (20) found smaller ZG in immature chief cells located in the upper parts of the gastric glands and larger ZG in more mature chief cells in the deeper parts of the glands. In this study both large and smaller ZG were observed in the same fractions from the denser sucrose gradient. It would be of considerable interest to separate these two. Clearly this would only be possible in more purified fractions than those obtained in this study. A consistent difference between ZG and whole mucosal preparations was absent from SMP; this zone was consistently present in whole oxyntic mucosal extracts. Only small amounts of this protease were detected in fractions containing soluble enzyme. This is in agreement with the suggestion of Mangla et al. (19) that SMP, believed to be a cathepsin, originates in mucous neck cells. Using electron microscopy few mucous neck cells were found in specimens that showed little SMP. In two patients with surgical resection for benign gastric ulcer pepsinogen 1 appeared to be prominent. If this zymogen is the precursor of pepsin 1, the findings would appear to agree with Taylor (21), who showed increased amounts of this pepsin in gastric juice of both gastric and duodenal ulcer patients. Storage of ZG resulted in their agglutination which was not associated with a change of the electrophoretic pattern of the zymogens. But upon disruption two zones with cathodal migration became apparent. The significance of this finding is not certain but a similar zone has been seen in the upper intestinal mucosa, and two cathodal zones may appear in concentrated extracts of gastric adenocarcinoma (21). The only other published study regarding agglutination involves pancreatic ZG which agglutinated in frozen isotonic sucrose, but retained their morphological appearance (9). Zymogen granules were not found in pyloric mucosa in Case 3, and the electrophoretogram was strikingly different from oxyntic mucosa by the absence of pepsinogen 1 and presence of SMP. Five cell types containing secretory granules have been distinguished in guinea pig stomach by Sato and Spicer (22) using cytochemical methods. In human gastric mucosa, Samloff has detected pepsinogens in four cell types (4,5). Both PG I and PG II were found in oxyntic mucosa and were localized to chief cells and mucous neck cells. But PG I were consistently absent in the distal stomach and proximal duodenum where cells, in the pyloric glands and in Brunner’s glands, contained only PG II. Only one sample of pyloric mucosa was evaluated in this study. However,
294
NAVAB,
LILLIBRIDGE,
AND TURNER
the appearance of pepsinogens 2-5 (constituents of PG 1) as well as pepsinogens 6 and 7 (constituents of PC II) suggests that proteases, included in PG I, can be present in pyloric mucosa. It is also in accordance with our findings in mucosal biopsies for healthy individuals (23). Clearly more studies are required before this finding can be generalized. However, constituents of PG I have been reported to be present in pyloric mucosa by other investigators (24,25). SUMMARY
Intact human gastric mucosal zymogen granules (ZG) were detected in specimens from surgical resections of one patient with gastric adenocarcinoma and two with benign gastric ulcer. Both large ZG with unilaminar membranes and smaller ZG with trilaminar membranes were identified by electron microscopy. The zymogens in the ZG and in mucosal extracts were separated by gel electrophoresis. Slow-moving Protease (SMP) was seen in the whole mucosal extracts but was absent from ZG. One specimen of pyloric mucosa showed a striking absence of ZG. Despite the absence of ZG, pyloric mucosa showed Pepsinogens 2-5 (constituents of PG I) as well as Pepsinogens 6 and 7 (constituents of PG II) and SMP. REFERENCES 1. Hirschowitz, B. 1.. in “Handbook of Physiology” (C. F. Code and W. Heidel, Eds.). Sect. 6, Vol. 2. p. 889. Williams & Wilkins, Baltimore, 1967. 2. Lillibridge, E. C., J. Eiophys. B&hem. C~fol. 10, 145 (1961). 3. Samloff. I. M.. Gasrroenrerology 60, 586 (1971). 4. Samloff. I. M., Gasrroenterology 66, 185 (1971). 5. Samloff. I. M., and Liebman, W. N.. GuAtroenferology 63, 36 (1973). 6. Libman, L. J., and Samloff, I. M., Gastroenrerology 74, 1132 (1978). 7. Samloff, I. M., Gastroenterology 57, 6S9 (1969). 8. Etherington, D. J.. and Taylor, W. H.. B&hem. J. 118, 587 (1970). 9. Hokin, L. E., Biochim. Biophys. Acra 18, 379 (1955). 10. Seijffers, M. J.. Segal. H. L., and Miller, L. L.. Amer. J. Physiol. 207, x (1964). 11. Anson. M. L.. and Mirsky, A. E.. J. Gen. Physiol. 16, 59 (1932). 12. Holter, H., and Linderstrom-Lang, K.. Hoppe-Seyler’s Z. Physio/. C’hem. 226 149 (1934). 13. Bowie. D. J., and Vineberg. A. M., Anaf. Ret,. 64, 357 (1935). 14. Ito. S., and Winchester, R. G.. J. Cell. Biol. 16, 541 (1963). 15. Menzies, G., J. Pathol. Bacterial. 83, 475 (1962). 16. Helander, H. F.. Cell Tis. Res. 189, 287 (1978). 17. Stachura. J., ivey, K. J., Tarnawski, A., Krause. W. J.. and Stogsdill. P.. Scu&. J. Gastroenterol. 16, 713 (1981). 18. Siekevitz. P., and Palade. G. E., J. Biophys. Biochem. gyro/. 4, 203 (1958). 19. Mangla, J. C., Navab, F., Lillibridge, C. B.. and Turner, M. D.. B&hem. Med. 13, 184 (1975). 20. Kataoka, K. Arch. Histol. Japan. 32, 251 (1970). 21. Taylor, W. H., Nature (London) 227, 76 (1970). 22. Sato, A., and Spicer, S. S.. Amer. J. Anat. 159, 307 (1980). 23. Navab, F., Rubio. C. E.. and Turner. M. D.. Gastroenteroloyy 84, 1257 (1983). 24. Whitecross, D. P.. Armstrong, C.. Clarke, A. D.. and Piper, D. W., Cur 14, 850 (1973). 25. Kojima. K.. and Moriga, M.. Gastroenleroi Jq~on 13, 421 (1978).