PROSTAGLANDINS

RELEASE OF PROSTAGLANDINS BY SMALL INTESTINAL

TISSUE OF MAN AND RAT IN VITRO AND THE EFFECT

OF ENDOTOXIN IN THE RAT IN VIVO

B.M. Peskar, H. Weiler, E.E. Kriiner* and B.A. Peskar*

Med. Klinik and Department of Pharmacology*,

University of Freiburg, D-7800 Freiburg, FRG

ABSTRACT

Rat jejunal tissue in vitro synthesizes large amounts of

prostaglandin (PG) D2 and smaller amounts of 6-keto-PGFl, and

PGE2, whereas human small intestinal mucosa synthesizes much

smaller amounts of the three PG determined with about equal

amounts of PGE2 and PGD2. Intraperitoneal administration of

bacterial endotoxin to rats induces fluid accumulation in the

small intestine and increases significantly the release of PGD2,

PGE2 and 6-keto-PGFl, into the small intestinal lumen in viva.

Endotoxin-induced stimulation of PG release is particularly

pronounced for PGD2. Fluid accumulation and PG output are in-

hibited by indomethacin. It seems possible that the different

total amounts of PG'synthesized by small intestinal tissue of

man and rat as well as the different pattern of PG released

might contribute to species-specific responses of the gastro-

intestinal tract to various pathophysiological stimuli.

INTRODUCTION

Various effects of bacterial endotoxins such as abortion,

fever and diarrhea in a number of species have been attributed

to an increased synthesis and also a decreased enzymatic inacti-

vation of PG (for review see 1). Skarnes and Harper (2) suggested

that increased formation of PGFza might be responsible for the

diarrheogenic effect of Salmonella enteritidis endotoxin in mice.

Mice pretreated with indomethacin did not show diarrhea after

endotoxin administration. Herman and Vane (3) described enhanced

release of PG in vitro by rabbit jejunum taken from endotoxin -

injected animals as compared to controls. Recently Robert et al.

(4) found that exogenous prostacyclin (PGI2), contrary to other

PG like PGE2, does not induce diarrhea and PGD2 is only weakly

active, but both PG inhibit the enteropooling effects of PGE2

and related compounds in rats. It was, therefore, of interest

to compare the intestinal synthesis in viva of the diarrheogenic

PGE2 with that of the anti-enteropooling PGD2 and PGI2 (deter- mined as the stable hydration product 6-keto-PGFl,) after ad-

ministration of bacterial endotoxin to rats. As a measure of PG

synthesis in viva the amounts of PG in the small intestinal con-

tents were determined. Finally, we have compared the pattern

of PG synthesized by human small intestinal mucosa with that

of rat jejunum incubated in vitro.

SUPPLEMENT TO VOL. 21 9

PROSTAGLANDINS

METHODS

Groups of female Wistar rats (body weight 220-300 g) were fasted for 24 hours with free access to water. They were injec- ted i.p. with either 25 mg/kg endotoxin (Salmonella minnesota gift of Dr. E. Rietschel, Max-Planck-Institut f& Immunbiologie, Freiburg, FRG) or the corresponding volume (2.5 ml/kg) of saline. Other groups were pretreated with indomethacin (10 mg/kg i.p.) one hour before endotoxin. All rats were sacrificed 2 hours after the endotoxin injection. The volume of the contents of the small intestine was determined as de'scribed by Robert (5).

The intestinal contents were diluted to 2.0 ml with 0.001 N HCl, the pH adjusted to 3.5 with HCl, and then PG were extracted into 5 vol. of ethyl acetate. PG in the evaporated extracts were purified by silicic acid column chromatography as described by Jaffe et al. (6). PGE2 and PGD2 were recovered in fraction II of the column eluates (benzene: ethyl acetate: methanol 60:40:2), while 6-keto-PGFl, was determined in fraction II plus fraction III (benzene: ethyl acetate: methanol 60:40:20). The recovery of PG in these fractions as determined using tritiated compounds (10000 cpm) added to the intestinal contents before extraction was 41 + 396 (mean 2 S.E.M., n=5) for PGE2, 34 ? 4% (n=7) for PGD2 and 64 f 4% (n=9) for 6-keto-PGFla. The results were cor- rected for the different average recoveries. PG in the evaporated column eluates were determined radioimmunologically. The radio- immunoassays for PGE2, PGD2 and 6-keto-PGFl, have been described previously (7-9).

For determination of PGD2 it is important to consider the fact, that under various conditions PGD2 is easily converted to a compound, which also binds to our anti-PGD2 antiplasma, but does not inhibit ADP-induced aggregation of human platelets. The degradation product can be separated from authentic PGD2 by thin layer chromatography in several solvent systems (10). Com- plete conversion of PGD2 to at least one compound with Rf values identical to those of the degradation product can be achieved by incubation of PGD2 at pH 12.0 at 37OC for one hour. The de- gradation of PGD2 to a compound recognized in the radioimmuno- assay but inactive in the bioassay may cause discrepancies in PGD2 levels determined by these two methods with higher levels obtained by radioimmunoassay.

In other experiments tissue of rat jejunum (90 mg wet weight) or human jejunal mucosa (15-48 mg wet weight) obtained by suc- tion biopsy from healthy volunteers (3 male, 3 female, age 23-44) was cut into small pieces and incubated as whole cell prepara- tion in Krebs-Henseleit bicarbonate buffer as described by Knapp et al. (11). PGE2, PGD metabolites 6,15-Piketo-13,l

g, 6-keto-PGFla as well as the -dihydro-PGFl, and 15-keto-13,14-

dihydro-PGE2 were determined radioimmunologically (7-9, 12, 13) in aliquots of the incubates. For determination of the PGE2 metabolite 15-keto-13,14-dihydro-PGE2 was converted to the more stable ll-deoxy-15-keto-13,l4-dihydro-ll,l6-cyclo-PGE2 (14, 15) by incubation at pH 12 at 37OC for 2 hours and then measured using a sensitive radioimmunoassay for the bicyclic degradation product (13).

10 SUPPLEMENT TO VOL. 21

PROSTAGLANDINS

RESULTS

Table 1 shows that bacterial endotoxin induces, as expected (4), significant fluid accumulation in the small intestine of rats as compared to control rats. Furthermore, the amounts of immunoreactive PGE2, PGD2 and 6-keto-PGFla in the intestinal contents are increased by endotoxin with the largest increase in PGD2. The presence of large amounts of PGD2 (220 + 95 ng, n=8,mean + S.E.M., corrected for the recovery of 34% after silicic acid column chromatography) in the intestinal contents

Table 1. Effect of endotoxin and pretreatment with indo- methacin on fluid accumulation in rat small intestine and the amounts of PG in the intestinal contents. Results represent the mean f S.E.M.; *** p < 0.001, ** p < 0.005, * p < 0.02; values of endotoxin treatment were compared with controls and the effect of indomethacin pretreatment was compared with the effect of endotoxin alone.

Intestinal fluid volume/

PGE2 PGD2 6-keto-PGFla

100 g body (ng/total intestinal volume)

weight (ml)

0.23 13.85 26.79 6.25 Control (n) + 0.01 + 3.17 * 7.47 * 1.34

(19) (16) (16) (16)

0.78 43.76 533.59 31.14

Endotoxin (n) f 0.05*** * 9.85s' 2111.88*** f 2.63***

(21) (7) (7) (7)

Endotoxin 0.33 9.54 8.41 5.98 + Indome- ? 0.03"' ? 2.76' + 2.56** f 0.78**'

thacin (n) (9) (5) (5) (5)

of endotoxin-treated rats was confirmed using the inhibition of

ADP-induced aggregation of human platelet rich plasma as a bio-

assay method. Using thin layer chromatography (solvent system

ethyl acetate/H20/isooctane/acetic acid llO:lOO:5O:2O, by volume,

upper phase) it was shown that the antiaggregatory activity co-

chromatographs exclusively with authentic PGD2, but not with

other antiaggregatory PG like PGEl (16) or the weakly active

6-keto-PGFlo (17). Pretreatment of the rats with indomethacin inhibits the fluid accumulation and simultaneously reduces the

PG content in the intestinal lumen (table 1).

The synthesis of PGE2, PGD2 and 6-keto-PGFla by whole cell

preparations of rat jejunum and human small intestinal mucosa

is shown in table 2. Rat jejunum synthesizes large amounts of

PGD2 and smaller amounts of PGE2 and 6-keto-PGFla. Human in-

testinal mucosa synthesizes, under identical conditions, much

SUPPLEMENT TO VOL. 21 11

PROSTAGLANDINS

Table 2. Release of PG and PG metabolites by whole cell pre- parations of biopsy specimens of human jejunal mucosa and rat jejunum. Results are the mean f S.E.M. of picogram PG formed/ mg wet weight/6 min. n.d.: not determined.

15-keto-13, 6-keto-

6,15-diketo- PGE2 14-dihydro- PGD2 13,14-di-

PGE2 PGFla hydro-PGFl,

human 150 20 106 48 28 jejunal ? 58 +8 k 47 + 18 k8 mucosa (n) (6) (6) (6) (6) (6)

rat 1145

jejunum (n) * 373 n.d. (4)

5983 + 868 (19)

1128 * 235

(6)

131 + 25 (6)

smaller quantities of all 3 PG measured. Furthermore, contrary to rat jejunum, human intestinal mucosa synthesizes about equal amounts of PGE2 and PGD2. The biologically inactive metabolites 6,15-diketo-13,14-dihydro-PGFla and 15-keto-13,14-dihydro-PGE2 are also formed in only small quantities by the human small in- testinal tissue (table 2).

DISCUSSION

Increased release of PG by jejunum of endotoxin-treated rabbits incubated in vitro has been described (3). Our results on PG in small intestinal contents indicate that after admini- stration of endotoxin,synthesis of PG by rat small intestine is also increased in viva. Synthesis of large amounts of PGD2 and smaller amounts of PGE2 and PGF2, by rat gastrointestinal tissue in vitro has been found by Knapp et al. (11). Whereas in that study hypertonic environment stimulated the output of PGE2 and PGF2a by rat jejunum without significant effect on PGD2 release, our results show that after endotoxin treatment release of all PG measured is enhanced with a particularly pronounced stimu- lation of PGD2 formation. It seems possible that endogenous PGD2 and PGI2, like the exogenous compounds, exert an anti- enteropooling effect. Such an effect could then influence the sensitivity of rats to the diarrheogenic effect of endotoxins.

Although PGD2 in the intestinal contents probably reflects mucosal biosynthesis, contribution by other sources like bile or pancreatic secretion cannot be excluded in our experiments. However, irrespective of the site of production, PG that have entered the intestinal lumen in an active form may act locally on the mucosa. On the other hand, PGD2 synthesized by intesti- nal mucosa does not necessarily reach the intestinal lumen, but might also enter the circulation and exert systemic effects. However, above,

in view of the relative instability of PGD2 described correct determination of endogenous PGD2 plasma levels

may be as difficult as determinations of the PGE2 metabolite 15-keto-13,14-dihydro-PGE2 (14, 15).

12 SUPPLEMENT TO VOL. 21

PROSTAGLANDINS

Human small intestinal mucosa incubated in vitro synthe- sizes considerably smaller amounts of PG than human .gastric mucosa (18) and rat jejunum (table 2). The latter difference is probably not due to the fact that the rat tissue, contrary to the human bioptic material, includes the smooth muscle layer. Bennett et al. (19) have shown that human small intestinal smooth muscle synthesizes relatively small amounts of PG and the main site of PG synthesis in human small intestine seems to be the mucosa. Furthermore, the small amounts of PG found in incubates of human intestinal mucosa are obviously not due to rapid enzymatic inactivation within the tissue. Although significant amounts of PG metabolites are formed by human je- junal mucosa (table 2), they cannot account for the difference in PG release compared to the rat tissue. The physiological and/ or pathophysiological consequences of the different amounts and the different pattern of PG synthesis by small intestinal tissue of different species are not known so far. The possibility that such differences contribute to species-specific responses of the gastrointestinal tract to various stimuli, e.g. the sen- sitivity to bacterial endotoxins, remains to be investigated.

ACKNOWLEDGEMENTS

This work was supported by the Deutsche Forschungsgemein- schaft. We thank Mrs. M. Paravicini and Mrs. A. Holland for excellent technical assistance.

1.

2.

3.

4.

5.

6.

REFERENCES

Rietschel, E.T., U. Schade, 0. Liideritz, H. Fischer, and B.A. Peskar. Prostaglandins in endotoxicosis. Microbiology 1980: 66. 1980.

Skames, R.C., and M.J.K. Harper. Relationship between endotoxin-induced abortion and the synthesis of prosta- glandin F. Prostaglandins 1: 191. 1972.

Herman, A.G. and J.R. Vane. Effect of indomethacin on endotoxin-induced production of prostaglandins in the isolated rabbit jejunum. In: Advances in Prostaglandin and Thromboxane Research. Vol. 2. (B. Samuelsson and R. Paoletti, eds.) Raven Press, New York, 1976. p. 557.

Robert, A., A.J. Hanchar, C. Lancaster, and J.E. Nezamis. Prostacyclin inhibits enteropooling and diarrhea. In: Prostacyclin. (J.R. Vane and S. Bergstriim, eds.) Raven Press, New York, 1979. p. 147.

Robert, A., J.E. Nezamis, C. Lancaster, A.J. Hanchar, and M.S. Klepper. Enteropooling assay: A test for diarrhea produced by prostaglandins. Prostaglandins 2: 809. 1976.

Jaffe, B.M., H.R. Behrman, and C.W. Parker. Radioimmuno- assay measurement of prostaglandins E, A and F in human plasma. J. Clin. Invest. 52: 398. 1973. -

SUPPLEMENT TO VOL. 21 13

PROSTAGLANDINS

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

14

Jobke, A., B.A. Peskar, and B.M. Peskar. On the specificity of antisera against prostaglandins A2 and E2. FEBS Letters

7: 192. 1973.

Anhut, H., B.A. Peskar, W. Wachter, B. Gr;ibling, and B.M. Peskar. Radioimmunological determination of prostaglandin D2 synthesis in human thrombocytes. Experientia 34: 1494. 1978. -

Peskar, B.A., Ch. Steffens, and B.M. Peskar. Radioimmuno- assay of 6-keto-prostaglandin Fla in biological material. In: Radioimmunoassay of Drugs and Hormones in Cardiovascular Medicine. (A.Albertini, M.DaPrada and B.A.Peskar, eds.) Elsevier/North-Holland Biomedical Press, Amsterdam. 1979. P. 239.

Anhut, H., B.A. Peskar, and W. Bernauer. Release of 15-keto- 13,14-dihydro-thromboxane B2 and prostaglandin D2 during anaphylaxis as measured by radioimmunoassay. Naunyn-Schmiede- berg's Arch.Pharmacol. 305: 247. 1978.

Knapp, H.R., 0. Oelz, B.J. Sweetman, and J.A. Oates. Syn- thesis and metabolism of prostaglandins E2, F2a and D2 by the rat gastrointestinal tract. Stimulation by a hypertonic

environment in vitro. Prostaglandins 15: 751. 1978. -

Peskar, B.M., H. Weiler, P. Schmidberger, and B.A. Peskar. On the occurrence of prostacyclin metabolites in plasma and vascular tissue as determined radioimmunologically. FEBS Letters 121: 25. 1980.

Peskar, B.M., B. Giinter, Ch. Steffens, E.E. Krb;ner, and B.A. Peskar. Antibodies against dehydration products of 15-keto-13,14-dihydro-prostaglandin E2. FEBS Letters 115: 123. 1980.

Granstrom, E., M. Hamberg, G. Hansson, and H. Kindahl. Chemical instability of 15-keto-13,14-dihydro-PGE2 - the reason for low assay reliability. Prostaglandins 19: 933. 1980.

Fitzpatrick, F.A., R. Aguirre, J.E. Pike, and F.H. Lincoln. The stability of 13,14-dihydro-15-keto-PGE2. Prostaglandins 19: 917. 1980. -

Kloeze,J . Influence of prostaglandins on platelet ad- hesiveness and platelet aggregation. In: Nobel Symposium No. 2. (S.BergstrZm and B.Samuelsson, eds.) Almqvist and Wiksell, Stockholm. 1967. p. 241.

Pace-Asciak, C.R., and M. Nashat. Mechanistic studies on the biosynthesis of 6-keto-prostaglandin Fla. Biochim. Biophys. Acta 487: 495. 1977.

Peskar, B.M., H.W. Seyberth, and B.A. Peskar. Synthesis and metabolism of endogenous prostaglandins by human gastric mucosa. In: Advances in Prostaglandin and Thromboxane Re- search. Vol. 8. (B.Samuelsson, P.W.Ramwell and R.Paoletti, eds.) Raven Press, New York. 1980. p. 1511.

Bennett, A., I.F. Stamford, and H.L. Stockley. Estimation and characterization of prostaglandins in human gastro- intestinal tract. Br.J.Pharmac. 61: 579. 1977. -

SUPPLEMENT TO VOL. 21