Gut, 1970, 11, 851-854

Histological localization of plasminogen activatorand proteolytic activity in the human stomachand duodenum

P. ERAS, P. HARPEL, AND S. J. WINAWERFrom the Department ofMedicine, The New York Hospital, Cornell Medical Center, and the MemorialHospitalfor Cancer and Allied Diseases, New York

SUMMARY Plasminogen activator activity has been localized in gastric and duodenal tissueusing a histological technique. This activity could be separated from proteolytic activity byincorporating E-aminocaproic acid into the fibrin substrate or by heat treatment. Plasmin-ogen activator activity was found in relation to mucosal and submucosal blood vessels in thetissues studied. Proteolytic activity was identified in the surface epithelium, the gastric antraland fundal glands, and in the antral submucosal fat. Proximal duodenal tissue containedproteolytic activity in relation to Brunner's glands.

Fibrinolytic activator activity (plasminogenactivator) has been found in extracts of gastric andduodenal mucosa from patients with duodenalulcer disease (Cox, Poller, and Thompson, 1967).The present study reports on the histologicallocalization of this activity in human gastric andduodenal tissue. Proteolytic activity could also belocalized and distinguished from plasminogenactivator activity. This was accomplished by theuse of E-aminocaproic acid and heat treatmentof the fibrin substrate (Fig. 1).

EACAPlasmi nogen 1 lsioeactivators > Plasminogen

Heat +

Nonspecific . vNonesecific ---...~..---..> Fibrin film lysisproteases 0Fig. 1. Outline of the pathway by which c-amino-caproic acid (,-ACA) and heat treatment of the fibrinfilm substrate distinguish plasminogen activatorfrom non-specific proteolytic activity.

Patients and Methods

Tissue from the body of the stomach was obtainedat surgery from three patients operated on forduodenal ulcer disease. In one patient, additionalspecimens were obtained from the gastric fundus,antrum, and duodenum. These specimens weretaken from areas which appeared grossly normal.Microscopic examination of haematoxylin andeosin-stained sections of these tissues showed anormal appearance without evidence of in-flanmmation. Surgical specimens half an inchsquare were washed in 0.15M sodium chlorideand were quick-frozen in an International-Harriscryostat microtome, and then stored at -25°Cand assayed within three weeks.

Plasminogen-rich fibrinogen was prepared fromcitrated bovine plasma according to the methodof Kekwick, MacKay, Nance, and Record (1955).The Kekwick F. 1 W low-ionic strength, cold-insoluble fibrinogen fraction was reconstitutedin citrate-saline buffer at a concentration of3.0 g% and stored at -20°C for use in preparingfibrin films.

FIBRIN FILMSTwenty NIH units of bovine thrombin (Parke-Received for publication 9 March 1970.

on March 31, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.11.10.851 on 1 O

ctober 1970. Dow

nloaded from

http://gut.bmj.com/

P. Eras, P. Harpel, and S. J. Winawer

Davis Co.) were added to 10 ml bovine fibrinogenwhich was then diluted in Michaelis veronal buf-fer, pH 7.35, to a final concentration of 1.5 g %.The fibrinogen was allowed to clot on a levelcellophane strip 8 x 17 cm in a humid chamberat 4°C. Squares of fibrin clot, each 2 cm, oncellophane were cut and inverted (fibrin facedown) on a glass slide and the cellophane wasremoved. Tissue sections from each specimenwerethen cut 8 microns thick on the cryostat, placedon fibrin films and the fibrin slides were in-cubated at 37°C in a moist chamber for 0, 1, 3, 5,10, 30, 45, and 60 minutes. The reaction wasstopped by immersion in 10% formaline-salinefor five minutes. The fibrin film was stained with amodification of the haematoxylin and eosin stain.Excess fibrin was trimmed with a sharp knife,and the preparation covered with permount anda coverslip. The prepared slides were viewedwith a stereoscopic microscope to determine thepresence of depressed colourless areas in thefibrin film which represented areas of fibrindigestion. The shortest incubation time requiredto produce a clearly visible zone of lysis wasnoted and served as a semi-quantitative index offibrinolytic activity.

PLASMINOGEN-FREE FIBRIN FILMSFibrin films were prepared in the manner de-scribed, following which they were heated at85°C for 30 minutes in a moist chamber to destroyfibrin-associated plasminogen (Lassen, 1952).

E-AMINOCAPROIC ACID-CONTAINING FIBRINFILMSFibrin films were prepared as described with1 x 10-3M E-aminocaproic acid added to thefibrinogen solution to inhibit selectively fibrinol-ysis produced by plasminogen activator activity(Alkjaersig, Fletcher, and Sherry, 1959).

Results

The gastric and duodenal surface epitheliumshowed proteolytic activity which was notinhibited by E-aminocaproic acid or by heat

treatment of the fibrin substrate. The data onlocation of fibrinolytic activity in the tissuesstudied (other than surface epithelial activity),the earliest time of onset of lysis, and the effect ofadding E-aminocaproic acid and heating thefibrin substrates are summarized in the Table.The gastric fundus showed proteolytic activity inthe surface epithelium and glands. Plasminogenactivator activity was found in mucosal and sub-mucosal blood vessels (Fig. 2). Similar findingswere present in the gastric antrum. In sectionstaken from the transitional zone, the antralglands showed greater non-inhibitable proteolyticactivity than did the fundal glands (Fig. 3). Inaddition, the antral submucosal fat caused lysisof the fibrin films which was not inhibited byE-aminocaproic acid or by heat treatment. Theproximal duodenum showed proteolytic activityin the surface epithelium, crypts, and Brunner'sglands. Activity of plasminogen activator waslocalized to the mucosal and submucosal bloodvessels (Fig. 4).

Discussion

Fibrinolytic activity of tissue may be causedeither by proteases which can directly hydrolysefibrin or by tissue activators which convertplasminogen to plasmin, the fibrinolytic enzyme.Albrechtsen (1959) found that most tissue con-tained plasminogen activator, but he did notexamine gastrointestinal tissue. Histological tech-niques, developed by Todd (1959 and 1964), madeit possible to examine plasminogen-activatoractivity of tissue in relation to histologicalstructure. His studies and those of others showedthat the activity of plasminogen activator ismainly associated with the blood vessels of thevarious tissues examined (Kwaan, 1966).

Studies of fibrinolytic activity in the gastro-intestinal tract are few. Kwaan, Cocco, andMendeloff (1964) have shown activity of plasmin-ogen activator at sites of crypt abscesses inpatients with active ulcerative colitis. Cox et al(1967) found fibrinolytic activity in gastric veinblood and no activity in samples obtained fromthe gastric artery. Saline extracts of the gastricand duodenal mucosa contained plasminogen

Patient Tissue Histological Location and Time of Earliest LysisFibrin E-Aminocaproic Acid Fibrin Heated Fibrin

1 Gastric fundus Mucosal and submucosal blood No lysis (60 min) No lysis (60 min)vessels (20 min)2 Gastric fundus Mucosal and submucosal blood No lysis (60 min) No lysis (60 min)vessels (15 min)3 Gastric fundus Mucosal and submucosal blood No lysis (60 min) No lysis (60 min)vessels (10 min)

Gastric antrum Mucosal and submucosal blood No lysis in blood vessels at 60 No lysis in blood vessels at 60vessels, submucosal fat (10 min) min; fat cells active min; fat cells activeDuodenum Mucosal and submucosal blood No lysis in blood vessels at 30 No lysis in blood vessels at 30vessels, Brunner's glands (5 min) min; Brunner's glands active min; Brunner's glands active

Table Localization offibrinolytic activity in stomach and duodenal tissue

852 on M

arch 31, 2021 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.11.10.851 on 1 October 1970. D

ownloaded from

http://gut.bmj.com/

Histological localization ofplasminogen activator and proteolytic activity in stomach and duodenum

Fig. 2. Specimen from gastric fundus following Fig. 3. Section ofbody of the stomach (patient 3) at15-minute incubation with fibrin film (patient 2). transition zone offundal and antral gland incubatedRepresentative lytic zones, indicated by arrows, 60 minutes with E-ACA fibrin film. The clear lyticappear as light, unstained areas in relation to mucosal zones in relation to mucosal and submucosal bloodand submucosal blood vessels. vessels shown in Fig. 2 are now not present (arrows

A and B) indicating that the lysis was produced byplasminogen activator. Lytic zones are seen inrelationship to submucosal fat cells (arrow C)indicating proteolytic activity. Antral glands on leftshow greater proteolytic activity than do fundalglands on right.

activator activity as shown by the ability to lyseplasminogen-rich fibrin plates but not to lyseheated or E-aminocaproic acid-containing plates.Cox, Poller, and Thomson (1969) have also foundthat digital gastric compression at laparotomyresulted in the development of plasmin-likeactivity in peripheral venous and in gastricvenous blood. The highest fibrinolytic activitywas found in patients with peptic ulcers.The present study demonstrates that both

plasminogen activator activity and proteolyticactivity can be identified in human gastric andduodenal tissue by a histochemical technique.Although the tissues obtained were from patientsundergoing surgery for duodenal ulcer diseaseno attempt can be made in this preliminary studyto correlate the enzymatic findings with thepresence of duodenal ulcer disease. In the speci-mens examined, plasminogen activator activitywas found in superficial mucosal and submucosalblood vessels in the stomach and duodenum.That this is plasminogen activator was shown bythe inhibition of this fibrinolytic activity by usinga heated fibrin substrate or one containingE-aminocaproic acid in a concentration whichwould inhibit plasminogen activator but notplasmin itself or other proteases (Alkjaersiget al. 1959). In the specimen in which comparison

Fig. 4. Lytic zones in relationship to mucosal bloodvessels (A) in the proximal duodenum (patient 3)after five minutes' incubation. Proteolytic activity, notinhibited on heated or E-ACA fibrin films, is associatedwith surface (B) and Brunner's glands (C) ( x 100).

853 on M

arch 31, 2021 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.11.10.851 on 1 October 1970. D

ownloaded from

http://gut.bmj.com/

854 P. Eras, P. Harpel, and S. J. Winawer

was possible, there appeared to be more plasmin-ogen activator activity associated with the vesselsof the fundus than with the antrum.

Epithelial proteolytic activity, which was notinhibited by E-aminocaproic acid or heat treat-ment of the fibrin film, may be due to residualacid-pepsin. Pepsin in the presence of hydro-chloric acid has been shown to cause fibrin platelysis (Cox et al, 1967). In the present study,gastric tissue was washed extensively before assayin physiological saline, a procedure which mightbe expected to eliminate most HCI and surfacepepsin. The assay was performed at neutral pH,conditions which would favour the inactivationof both pepsin and peptic enzymes. The epithelialsurface proteolytic activity observed might there-fore be due to proteolytic enzymes active in theneutral pH range. Taylor (1959) has identified aprotease in gastric mucosal tissue which is activeat neutral pH. The mucosal cells responsible forthe production of this enzymatic activity have notbeen identified. It was of interest to find pro-teolytic activity associated with both the fundaland antral glands. Proteolytic activity wasassociated with the fundal, antral, and theBrunner's glands of the duodenum. The activityof the Brunner's and antral glands was greaterthan was the activity of the fundal glands. Theproduction of pepsin occurs principally in thechief cells of the fundal glands. Small amountsof peptic-like activity have been reported induodenal juice and duodenal mucosal extracts(Seijffers, Miller, and Segal, 1963; Cooke andGrossman, 1966) and in extracts from the pyloricgland area of the stomach (Grossman andMarks, 1960). Linderstr0m-Lang, Holtzer, andOhlsen (1935) found acid protease activity of thegastric mucosa in the pig. Whereas highestactivity correlated with peptic cells of the fundicglands, activity was found in mucosal cells of thepyloric and antral glands as well. In the presentassay system the glands with the greatest pro-teolytic activity have been reported to have lowlevels of peptic activity. The activity observed maytherefore be due to other as yet unidentifiedproteolytic enzymes.

Fat cells in the submucosa of the antrum alsohave proteolytic activity. To our knowledge,proteolytic activity associated with fat-containingcells has not been reported previously.The presence of plasminogen activator activity

and proteolytic activity in the stomach andduodenum could have implications in gastricand duodenal haemorrhage associated with avariety of disease states.

This work was supported by a US Public HealthService traineeship CST 562A67 from the CancerControl Program; a US Public Health Servicespecial fellowship 2-F3-CA-3A, 497-02, fromthe NIH, National Cancer Institute; by NIHgrant NB-0334605 (National Institute of Neuro-logical Diseases and Blindness), and by anAmerican Cancer Society research scholaraward.

References

Albrechtsen, 0. K. (1959). Fibrinolytic activity in the organism.Acta physiol. scand., 47, Suppl. 165, 1-112.

Alkjaersig, N., Fletcher, A. P., and Sherry, S. (1959). e-Amino-caproic acid: an inhibitor of plasminogen activation.J. biol. Chem., 234, 832-837.

Cooke, A. R., and Grossman, M. I. (1966). Studies on thesecretion and motility of Brunner's gland pouches. Gastro-enterology, 51, 506-514.

Cox, H. T., Poller, L., and Thomson, J. M. (1967). Gastricfibrinolysis. A possible aetiological link with peptic ulcerLancet, 1, 1300.1302.

Cox, H. T., Poller, L., and Thomson, J. M. (1969). Evidence forthe release of gastric fibrinolytic activity into peripheralblood. Gut, 10, 404-407.

Grossman, M. I., and Marks, I. N. (1960). Secretion of pepsinogenby the pyloric glands of the dog, with some observationson the histology of the gastric mucosa. Gastroenterology,38, 343-352.

Kekwick, R. A., MacKay, M. E., Nance, M. H., and Record,B. R. (1955). The purification of human fibrinogen.Biochem. J., 60, 671-683.

Kwaan, H. C. (1966). Tissue fibrinolytic activity studied by ahistochemical method. Fed. Proc., 25, 52.56.

Kwaan, H. C., Cocco, A., and Mendeloff, A. I. (1964). Histologicdemonstration of plasminogen activation in rectal biopsiesfrom patients with active ulcerative colitis. J. Lab. clin.Med., 64, 877.

Lassen, M. (1952). Heat denaturation of plasminogen in thefibrin plate method. Acta physiol. scand., 27, 371-376.

Linderstr0m-Lang, K., Holtzer, H., and Ohlsen, A. S. (1935).Studies on enzymatic histochemistry. XIII. The distributionof enzymes in the stomach of pigs as a function of itshistologic structure. C.R. Lab. Carlsberg, Ser physiol., 20,66-127.

Seijffers, M. J., Segal, H. L., and Miller, L. L. (1963). Separationof pepsin I, pepsin IIA and pepsin IIB, III fromhuman gastric mucosa. Amer. J. Physiol., 205, 1099-1105.

Taylor, W. H. (1959). Studies on gastric proteolysis. 4. Proteinaseactivity of gastric juice and gastric mucosal extracts atpH 6 to 8. Biochem. J., 71, 626-632.

Todd, A. S. (1959). The histological localization of fibrinolysinactivator. J. Path. Bact., 78, 281-283.

Todd, A. S. (1964). Localization of fibrinolytic activity in tissuesBrit. med. Bull., 20, 210-212.

on March 31, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.11.10.851 on 1 O

ctober 1970. Dow

nloaded from

http://gut.bmj.com/