- Email: [email protected]
Prostaglandin synthesis and metabolism in isolated pancreatic islets of the rat
PROSTAGLANDINS
PROSTAGLANDIN SYNTHESIS AND METABOLISM IN ISOLATED PANCREATIC ISLETS OF THE RAT K. L. Kelly1 S. G. Laychock Department of Pharmacology Medical College of Virginia Richmond, Virginia 23298
ABSTRACT Isolated pancreatic islets of Langerhans of the rat which were sonicated and incubated with radiolabeled arachidonic acid for 1 hr synthesized several species of prostaglandins (PGs). Both thin-layer and high-performance liquid (HPLC) chromatographic techniques demonstrated the synthesis by islet sonicates of PGF2o and PGE2 equivalents, in addition to the 15-keto-13, 14-dihydro metabolites of these primary PGs. In addition, HPLC allowed the identification of 6-keto-PGFlo (the metabolite of prostacyclin) as a major PG synthesized from arachidonate by this tissue. Islet vascular elements, as well as endocrine cells, may contribute to the synthesis of the latter compound. Lesser amounts of arachidonate were incorporated into PG-like compounds eluting as thromboxane. The synthesis of PGs was sensitive to the protein concentration of islet sonicate, and a five-fold dilution of protein resulted in a comparable reduction in arachidonate incorporation into PGs. Labeled arachidonate was also incorporated into compounds which elute as hydroxy or hydroperoxyeicosatetraenoic acids on HPLC. Thus, isolated pancreatic islets synthesize a variety of PGs which may have a physiological role in hormone secretion from this endocrine organ.
INTRODUCTION Prostaglandins (PGs) have been postulated to be mediators or modulators of endocrine secretion from various organs including the hypothalamus and pituitary gland, the adrenal gland, the thyroid gland, and the ovary (1). In addition, a modulatory role for PGs in secretion of hormones from the pancreatic islet of
lPre-doctoral trainee supported by NIH training grant 5T32GM 07111-05, awarded to the Department of Pharmacology. 2To whom reprints should be addressed. This work was supported by a Young Investigator Research Award to SGL from the NIAMDD (R23 AM25705).
MAY
1981 VOL. 21 NO. 5
759
PROSTAGLANDINS
Langerhans has recently been proposed, based largely upon the evidence that inhibitors of PG synthesis induce changes in the secretion of insulin and glucagon, and that infusion, perfusion or perifusion of the pancreas or isolated islets with PGs also alters hormone secretion from this organ (2,3). The evidence in the literature describes PGs as being both stimulatory (4,5) and inhibitory (6,7) to insulin secretion -in vivo or in vitro, depending upon variables such as the PG or PG synthetase inhibitor used, and the animal species studied. However, exogenous application of PGs to pancreatic preparations designed to study the physiological role of PGs in islet secretion may not mimic the actions of endogenous To date, few studPGs participating as intracellular messengers. ies have attempted to identify the PGs synthesized by the pancreas or isolated islets. PGE2 and PGF2a have been identified in homogenates of whole rat pancreas (8) as well as in extracts of human (9), dog, cat, rabbit and chicken pancreata (lo), as determined by thin layer chromatographic and bio-assay methods; PGEl and PGE2 were also identified in guinea-pig islets by radioimmunoassay (11). In contrast, pancreatic acinar cells have been described as devoid of cyclooxygenase activity (12). The purpose of the present investigation was to obtain a profile of PGs synthesized by homogenates of rat pancreatic islets, and to determine the PG metabolizing ability of rat islets. In the pursuit of as complete a profile of radiolabeled PGs synthesized by islet cycle-oxygenase as possible, we have employed both thin layer and high performance liquid chromatography techniques. MATERIALS PGD PGF GKPGF le;c;;;;zE2 and 15KH2F2, were gifts PGE Tritiated prostaglandins, from the 2' pjohn2~o.,K%.%~azoo, arachidonic acid and 14C-labeled arachidonic acid were obtained from New England Nuclear, Boston, MA. Hank's Balanced Salt Solution was from Grand Island Biological Co., Grand Island, N.Y. Collagenase (CLS IV) was from Worthington Biochemical Corp., Freehold N. J. KH2P04 (reagent grade) was from J.T. Baker, Phillipsburg, N.J. Reduced glutathione was from Sigma Chemical Co., St. Louis, MO. NAD and NADP were from Boehringer Mannheim. W. Germany. EDTA, and all organic solvents (HPLC Grade) were from Fisher Scientific Co., Silver Springs, MD. Silicic acid (100-200 mesh) was from Rio-Rad, Richmond, CA. Silica Gel G plates were from Analtech, Newark, Delaware. Microporasil silicic acid and fatty acid analysis HPLC columns, pumps and programmer were from Waters Associates, Milford, MA. Budget solv and 3a20 scintillation cocktails were from Research Products International, Grove Village, Illinois. METHODS Isolation of Islets. Male Sprague Dawley rats (ZOO-250 g) were fasted overnight prior to decapitation and distention of each pancreas with 10 ml Hank's Balanced Salt bicarbonate buffer (pH 7.4)
760
MAY
1981 VOL. 21 NO. 5
PROSTAGLANDINS
containing 3 mg collagenase The pancreas was then excised, minced, suspended in a final volume of 20 ml oxygenated Hank's buffer containing 6-8 mg collagenase and digested for 15 min, 37 C, at 250 rpm in a gyro-rotatory shaking water bath (New Brunswick Scientifit). Following this incubation, 30 ml cold Krebs-Ringer-bicarbonate (KRB) buffer (pH 7.4) was added to the pancreatic digest; the sedimenting tissue was washed three times. The final tissue sediment was examined by stereomicroscopy (15 x mag) and individual islets were isolated manually. Groups of 150-200 islets were washed with 1 ml KH2P04-NaOH buffer (0.05 M), pH 7.4, and resuspended in 525 ~1 KH2P04-NaOH buffer. Islets were then sonicated for 10 set with a micro-ultrasonic cell disrupter (Kontes). Protein was determined by the method of Lowrv et. al. (13). Incubation. Islet sonicates were diluted to a final mean protein concentration of 1 + 0.4 mgfml in 0.5 ml KH2P04-NaOH buffer. This final incubation buffer contained tritiated or 14carbon radiolabeled arachidonic acid (0.4 ? 0.1 nici); certain experiments also contained reduced glutathione (GSH) (1 mM) and sodium (tetra) ethylenediamine tetraacetate (EDTA) (4 mM) as indicated in the text. The final concentration of ethanol, used as the vehicle for arachidonate, was 0.01%. Islet sonicate mixtures were incubated in a shaking water bath (100 rpm) at 37 C for 1 hr, in an atmosphere of o,/co2 (95:5). Radiolabeled arachidonate was also incubated in buffer in the absence of islet sonicate (blank samples) in order to monitor spontaneous oxidative fatty acid conversion to PG-like compounds. Blank values were subtracted from the radioactive PG profiles obtained from incubations containing islet sonicate. In experiments designed to determine the PG metabolizing capability of islets, tritium-labeled PGE2 or PGF20 (0.5 nCi> was incubated with islet sonicate for 1 hr, in the presence of GSH and EDTA, and either a-nicotinamide adenine dinucleotide (NAD) or B-nicotinamide adenine dinucleotide phosphate (NADP) (4 mM). Tritium-labeled PG standards were incubated in buffer in the absence of sonicate in order to estimate the percent recovery during incubation and extraction procedures. The incubations were terminated with the addition of 500 i.rlof cold KH2P04-NaOH buffer, and acidification to pH3 with formic acid. Acidified aqueProstaglandin Isolation and Identification. ous islet samples were extracted 3 times with 10 ml ethyl acetate. The organic phases of extraction were pooled for analvsis using thin-layer (TLC) or high-performance liquid (HPLC) chromatography. Recovery of PGs averaged 70%. PG extracts were prepared for TLC analysis by prepurification on columns of silicic acid as described previously (14) and employing 10 ml benzene to remove unincorporated arachidonate from PGs. PGs recovered during this purification step were applied to Silica Gel G plates and developed in the solvent system chloroform-methanolacetic acid-deionized water (90:8:1:0.Bv/v) as described by Bailey et al. (15). PG standards were processed in parallel with experi-
MAY
1981 VOL. 21 NO. 5
761
PROSTAGLANDINS
mental samples in order to determine Rf values. Following plate development, the PGs were visualized with iodine vapor. TLC plates were scraped at 1 cm intervals from the origin to the solvent front. The scraped zones were suspended in scintillant and radioactivity levels were determined by liquid scintillation spectrometry (52% efficiency). PGs in tissue extracts were also analyzed by HPLC silicic acid and reversed phase (fatty acid) systems, as described in detail elsewhere (16). Fractions collected during elution of the various PGs were analyzed for radioactivity by liquid scintillation spectrometry. Statistical Methods.
Values shown are mean ?: S.E. RESULTS ____
Thin-layer Chromatographic Analysis of PGs. Rat islet sonicate when incubated with radiolabeled arachidonic acid synthesized several primary PGs and PG metabolites. Fig. 1A is a representative thin-layer chromatographic profile of the PG compounds synthesized by disrupted rat islets. The tritiated arachidonate substrate was largely converted to PGs migrating in three distinct zones on the thin-layer plates. PGF2o (F20) and thromboxane B2 (TXB2) demonstrate similar Rf values and dominate the first peak of PG material appearing nearest to the origin. The next broad peak of radiolabeled material corresponds to Rf values for the coincident PGE2 (E2)/ 6-keto-PGFL (6KFLo) and the closely migrating but distinct pair PGD2 (D2)/1?-keto-13,14-dihydro-PGF2a (15KH2F2,). And lastly, the PGE metabolite 15-keto-13,14-dihydro-PGE2 (15KH2E2) migrates as a discrete zone. The relative proportions of the different PGs and PG metabolites synthesized by islet sonicates are illustrated in Fig. 1B. After an hour of incubation, the percentage of arachidonic acid incorporated into islet PGs is largely distributed among the PGE2/ 6-keto-PGFlo, PGD2/ 15-keto-13,14-dihydro-PGF20 and 15-keto-13,14dihydro-PGE2 species of unsaturated fatty acids. A slightly lesser amount of label was found in the PGF2,/TXB2 zone when the incorporation of arachidonate into PGs for several experiments were averaged. The total incorporation of labeled substrate into PGs varied considerably among experiments due to variation in protein concentration of the sonicates, as well as animal variability. Thus, the individual PG data are best expressed as relative percent of total labeled arachidonate found in the PG zones. When islet sonicate was diluted to 20% (0.34 mg protein/ml) of the original protein concentration and incubated with radiolabeled arachidonic acid the total amount of radioactivity (4876 cpm) incorporated into PGs (PGE2/ 6-keto-PGFlo, PGD2/ 15-keto-13,14-dihydro PGF2o, and 15-keto-13,14-dihydro PGE2) was 27% of the total label incorporated into PGs in the undiluted sonicate (18,003 cpm). The addition of GSH and EDTA to the incubation did not change the incorporation of radiolabel into PGs by islet sonicates.
762
MAY
1981 VOL. 21 NO. 5
PROSTAGLANDINS
4
E2 TXB2
6K F,a
15KH2F2_
5
6
‘5W2E2
02
Fig. 1. (A) Representative thin-layer chromatographic profile of Incubation media contained PGs synthesized by rat islet sonicates. 3H-arachidonic acid, GSH, and EDTA. Cross-hatched areas represent areas of Silica Gel G plates scraped and counted to determine total cpm in PGs, which in (B) was used to calculate percent arachidonate incorporation in each PG zone. ( n = 5 or 6).
MAY
1981 VOL. 21 NO. 5
763
PROSTAGLANDINS
High Performance Liquid Chromatographic (HPLC) Analysis of PGs. Extracts of islet sonicates previously incubated with radiolabeled arachidonic acid were subject to analysis for PGs using both silicic acid and reversed phase (fatty acid) HPLC. Enhanced separation and resolution of PGs on columns of silicic acid made it possible to identify PGF2, in islet sonicate extracts as a separate entity from TXB2 (Fig. 2A). PGE2, 6-keto-PGFl, and TXB2 elute together on silicic acid columns, and constitute a distinct peak of labeled compounds synthesized by islets. The 15-keto-13, 14-dihydro PGE2 metabolite elutes just prior to 15-keto-13,14-dihydro PGF20 and PGD2. The latter two PGs have similar retention times and constitute a broad peak of labeled islet PG, with PGD2 perhaps appearing as a very minor shoulder of radioactivity on the PGF2, metabolite (Fig. 2A). The other labeled compounds identified early in the elution profile are arachidonic acid (A.A.) and unspecified hydroxy- or hydroperoxy-eicosatetraenoic acids (ETE). When the relative amounts of label incorporated into different PGs on HPLC were determined as a percent of the total number of counts incorporated into islet PGs, PGF2o incorporated 13% of the label (Fig.2B). The combined PGE2/6-keto-PGFla/TXB2 fraction incorporated almost 28% of the labeled arachidonate, as did the 15-keto-13,14-dihydro PGE2 fraction. The combined 15-keto-13,14dihydro PGF20/PGD2 fraction incorporated 24% of the labeled precursor found in PGs. These percentages are roughly predicted by the data in Fig.2A, however, the variability among experiments minimized somewhat the difference between different PGs. The reversed phase HPLC system helped to further define and identify PGs synthesized by islet sonicates since it resolves 6keto-PGFlo and TXB2 from PGE2. Fig. 3A shows the profile of islet PGs obtained using reversed phase chromatography. The profile supports the data obtained using silicic acid HPLC showing that PGF2o is turning over in islets less rapidly than is PGE2 and 6-ketoPGFl,. In four experiments analyzed by reversed phase HPLC, PGF2o arachidonate incorporation was equivalent to only 19 ?r 2% of the radiolabel incorporated into PGE2, 6-keto-PGFlo and TXB2; in addition, the amount of radiolabel found in PGF2o was only 48 rf 3% of the arachidonate activity found in the PGE2 fraction alone. In several experiments, the silicic acid HPLC fraction containing PGE2/6-keto-PGFl,/TXB2 was subjected to additional reversed phase HPLC. These experiments showed (Fig. 3B) that 6-keto-PGFlo contained 51% of the radiolabeled arachidonic acid originally eluted in the silicic acid fraction, whereas PGE2 and TX52 comprised only 36% and 16%, respectively, of the labeled compounds. Metabolism of PGE2 and PGF2, In Islet Sonicates. The incubation of tritiated PGE2 and PGF2o with islet sonicates containing GSH,EDTA and either NAD or NADP resulted in the conversion of each of the primary PGs to its 15-keto-13,14-dihydro metabolite. The conversion of labeled PGE2 to its 15-keto-13,14-dihydro metabolite in the presence of NAD and islet sonicate was 12%. Whereas NAD has been
764
MAY 1981 VOL. 21 NO. 5
PROSTAGLANDINS
Fig. 2. (A) Representative silicic acid HPLC profile of PGs synthesized by rat islet sonicates. Tissue was incubated with 3H-arachidonic acid, GSH, and EDTA. Cross-hatched areas delineate fractions counted and pooled to determine total cpm in PGs, which in (B) was used to calculate percent of arachidonate incorporation in each PC fraction (n = 6).
MAY 1981VOL. 21 NO. 5
765
PROSTAGLANDINS
210 K2
0
pzza aem ,..,....,....,.*..,....,....,...., 7
10
16
20
limo Flow=
M 25
30 (min) 2 ml/min
36
40
60 r
=2
50
0
5
40
(P30 P : ;; 0 I:
20 IO 0
Fig. 3. (A) Representative reversed phase HPLC profile of PGs synthesized by rat islet sonicates. Values represent 50% of PGs prepurified and eluted from silicic acid HPLC. Incubation conditions were as described in Fig. 1. (B) Percent distribution of radiolabeled arachidonic acid incorporation into PGs separated using reversed phase HPLC. PGs had been previously eluted in the E7/6-ketoFlo/TXB* fraction from silicic acid HPLC, and were re-chromatographed (n = 4).
766
MAY 1981 VOL. 21 NO. 5
PROSTAGLANDINS
described as the preferred cofactor for dehydrogenase enzymes metabolizing PGE2, NADP is preferred by enzymes metabolizing PGF2o (17). Thus, in the presence of NADP and tritiated PGF2o, islet sonicates metabolized 36% of the parent compound to the 15-keto-dihydro derivative. DISCUSSION PGs have been proposed as mediators of secretion in a number of cell types in which their synthesis has been confirmed (18). Analysis of pancreatic homogenates obtained from several species of animals, including man, have indicated that at least certain PGs are found in this tissue (8,9,10). However, the pancreas is composed of heterogeneous cell types, including exocrine acinar tissue, duct and vascular elements, as well as pancreatic islets which are composed of at least three endocrine cell types and vascular tissue (19). Isolated rat islets allow the advantage of localizing PG identification to the pancreatic endocrine system which, although composed of diverse cell types, is characterized as a paracrine secretion system which loses major portions of its secretory products to adjacent tissue elements. Thus, PGs synthesized and released from any one cell type might influence the physiology of neighboring cells, making the elucidation of islet PGs relevant to the entire tissue's function. Our studies have identified several primary PGs, as well as their metabolites, which are synthesized by rat islet sonicates. Sonicated islet preparations were chosen over intact islet incubations in these studies since the incorporation of labeled arachidonit acid into prostaglandins is limited in intact tissue. We have likewise found that the specific activity of the fatty acid in other lipid fractions of intact islets is also low (unpublished observations). The incorporation of arachidonic acid into prostaglandin and other fractions of intact islets may be limited by the physical impedencc of the capsular membrane or the biochemistry of the islet generally. Using disrupted tissue, however, TLC and HPLC methods have illustrated that PGF2n and PGE2 equivalent compounds are synthesized by rat islet cycle-oxygenase. These results corroborate previous findings that arachidonate is converted to PGF2 and PGE2, as identified in extracts of the perfused rat pancreas ? 8). In addition, we have demonstrated the presence in islets of 15-hydroxy PG dehydrogenase and 13,14-dihydro PG reductase (20) through the metabolism of both endogenous and exogenous PGE2 and PGF2o to their 15-keto-13,14-dihydro metabolites. Complete characterization of these latter enzymes will be important in determining the physiological role of PGs in islet secretory function, since the metabolism of PGs may be intimately related to the nucleotide levels of cells which are altered by changes in secretory activity (21). Resolution of TX-like compounds on HPLC showed a low level of arachidonate incorporation into these products as well. Based upon the data, the synthesis of PGD2 by islet sonicate would appear to be of modest proportions, at best.
MAY 1981 VOL. 21 NO. 5
767
PROSTAGLANDINS
The other major PG synthesized by islet sonicates, and clearly identified by HPLC methods as distinct from other PGs, was 6-ketoPGFlo, the metabolite of prostacyclin (PG12) (22). Thus, pancreatic islets possess in addition to cycle-oxygenase, prostacyclin synthetase which converts PG endoperoxides to PGI2 (23). The metabolite 6-keto-PGFlo has previously been identified in isolated rat adrenocortical cells (24); this is a system where vascular elements have been eliminated, thus allowing for the identification of PG12 synthesis solely by endocrine cells. Unfortunately, there are vascular elements in islets in addition to the endocrine cells, and either or both of these elements may be contributing to the PGs synthesized by islet sonicates. Future studies utilizing isolated islet cells will help to elucidate the origin of the islet PGs. The extent to which the small, unavoidable acinar tissue contamination of isolated islets contributes to the PGs identified in these preparations is difficult to assess. In rat acinar tissue sonicates processed in parallel with our islet preparations we have identified arachidonate incorporation into PGF2o and PGE2 equivalents as well as into the metabolites of these compounds, although no PG synthesis was identified in exocrine tissue by other investigators (12). However, the incorporation of labeled arachidonate into islet PGs was more than 2-fold higher than that appearing in exocrine tissue PGs as calculated per mg of protein in the respective sonicates (unpublished observation). Thus, the trace amounts of exocrine tissue adhering to isolated islets would not be expected to greatly contribute to the profile of islet PGs reported. Analysis of arachidonate incorporation into PG products by HPLC also revealed that this fatty acid is incorporated into non-PG hydroperoxy compounds which chromatograph on silicic acid as the hydroxyeicosatetraenoicacid (HETE) (25). These hydroxy fatty acids are products of an enzyme, lipoxygenase (26),which has heretofore not been demonstrated in pancreatic tissue. A biological role for the HETE compounds in endocrine function remains to be determined, however, some of the other PGs identified in this study have been shown to influence islet secretion. Exogenous PGE2 and PGF2, can enhance insulin and glucagon release from the perfused pancreas, whereas PGD2 is a more potent secretagogue for glucagon than for release of insulin, and PG12 and TX appear to have no effect on secretion of either hormone (27). Thus, the rat pancreatic islet synthesizes a variety of PGs which may have a physiological role in hormone secretion. Future studies will determine whether or not stimulation of secretory function in intact cells alters the synthetic profile of islet PGs. REFERENCES 1. 2. 3.
768
Flack, J.D. In: The Prostaglandins (P.W. Ramwell, ed.), Plenum Press N.Y., 1973, Vol. 1, pp. 327-345. Pek, S., Lands, W.E.M., Akpan, J., and Hurley, M. Adv. Prostaglandin and Thromboxane Res. 8: 1295-1298, 1980. Robertson, R.P. Diabetes -28: F43-948, 1979.
MAY 1981 VOL. 21 NO. 5
PROSTAGLANDINS
4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.
20. 21. 22. 23. 24. 25. 26. 27.
Burr, I.M., and Sharp, R. Endocrinology 94: 835-839, 1974. Johnson, D.G., Fujimoto, W.Y., and Williams, R.H. Diabetes -22: 658-663, 1973. Robertson, R.P., Gavareski, D.J., Porte, D.,Jr., and Bierman, E.L. J. Clin. Invest. 54: 310-315, 1974. Sacca, L., Rengo, F., Chiariello, M., and Condorelli, M. Endocrinology 92: 31-34, 1973. Hamamdzic, Mx and Malik, K. Am. J. Physiol. 232: 201-209, 1977. Karim, S.M.M., Sandler, M., and Williams, E.D. Br. J. Pharmac. Chemother. 31: 340-344, 1967. Karim, S.M.E, Hillier, K., and Devlin, J. J. Pharm. Pharmac. 20: 749-753, 1968. zyckx, A.S., and Lefebvre, P.J. Adv. Prostaglandin and Thromboxane Res. 8: 1299-1302, 1980. Chauvelot, L., Heisler, S., Huot, J., and Gagnon, D. Life Sci. 25: 913-920, 1979. zwry, O.H., Rosebrough, N.J., Farr, A.L., and Randall, R.J. J. Biol. Chem. 193: 265, 1951. Laychock, S.G., and Rubin, R.P. Prostaglandins -10: 529-540, 1975. Bailey, J.M., Bryant, R.W., Feinmark, S.J., and Makkeja, A.N. Prostaglandins 13: 479-492, 1977. Whorton, A.R., Szgel, M., Oates, J.A., and Fralich, J.C. Biochim. Biophys. Acta 529: 176-180, 1978. Pace-Asciak, C.R. In: Renal Prostaglandins (James B.Lee, ed.) Eden Press, Vermont, 1978, Vol. 1, pp. 55-83. Rubin, R.P. and Laychock, S.G. In: Calcium in Drug Action (G. Weiss, ed.) Plenum Press, N.Y. 1978, pp. 135-155. Wellmann, K.F., and Volk, B.W. In: The Diabetic Pancreas. (Bruno W. Volk and Klaus F. Wellmann, eds.). Plenum Press, N.Y., 1977, pp. 99-128. Anggard, E., and Larsson, C. Eur. J. Pharmacol. -14: 66-70, 1971. Ammon, H.P.T., and Verspohl, E. Endocrinology -99: 1469-1476, 1976. 17: 4096-4101, 1978. Sun, F.F., and Taylor, B.M. Biochemistry Sun, F.F., Chapman, J.P., and McGuire, J.C.Prostaglandins 14: 1055-1074, 1977. zychock, S.G., and Walker, L. Prostaglandins -18: 793-811, 1979. Gr&en, K., Hamberg, M., Samuelsson, B., and Frijlich, J.C. Adv. Prostaglandin and Thromboxane Res. 5: 15-38, 1978. Samuelsson, B., Borgeat, P., Hammarstrgm, S., and Murphy, R.C. Adv. Prostaglandin and Thromboxane Res. 6: l-18, 1980. Akpan, J.O., Hurley, M.C., Pek, S., and Lands, W.E. Can. J. Biochem. -57: 540-547, 1979. Editor:
Harold R. Behrman
MAY 1981VOL. 21 NO. 5
Rec'd: 9-80
Acpt: 3-81
769
PROSTAGLANDIN SYNTHESIS AND METABOLISM IN ISOLATED PANCREATIC ISLETS OF THE RAT K. L. Kelly1 S. G. Laychock Department of Pharmacology Medical College of Virginia Richmond, Virginia 23298
ABSTRACT Isolated pancreatic islets of Langerhans of the rat which were sonicated and incubated with radiolabeled arachidonic acid for 1 hr synthesized several species of prostaglandins (PGs). Both thin-layer and high-performance liquid (HPLC) chromatographic techniques demonstrated the synthesis by islet sonicates of PGF2o and PGE2 equivalents, in addition to the 15-keto-13, 14-dihydro metabolites of these primary PGs. In addition, HPLC allowed the identification of 6-keto-PGFlo (the metabolite of prostacyclin) as a major PG synthesized from arachidonate by this tissue. Islet vascular elements, as well as endocrine cells, may contribute to the synthesis of the latter compound. Lesser amounts of arachidonate were incorporated into PG-like compounds eluting as thromboxane. The synthesis of PGs was sensitive to the protein concentration of islet sonicate, and a five-fold dilution of protein resulted in a comparable reduction in arachidonate incorporation into PGs. Labeled arachidonate was also incorporated into compounds which elute as hydroxy or hydroperoxyeicosatetraenoic acids on HPLC. Thus, isolated pancreatic islets synthesize a variety of PGs which may have a physiological role in hormone secretion from this endocrine organ.
INTRODUCTION Prostaglandins (PGs) have been postulated to be mediators or modulators of endocrine secretion from various organs including the hypothalamus and pituitary gland, the adrenal gland, the thyroid gland, and the ovary (1). In addition, a modulatory role for PGs in secretion of hormones from the pancreatic islet of
lPre-doctoral trainee supported by NIH training grant 5T32GM 07111-05, awarded to the Department of Pharmacology. 2To whom reprints should be addressed. This work was supported by a Young Investigator Research Award to SGL from the NIAMDD (R23 AM25705).
MAY
1981 VOL. 21 NO. 5
759
PROSTAGLANDINS
Langerhans has recently been proposed, based largely upon the evidence that inhibitors of PG synthesis induce changes in the secretion of insulin and glucagon, and that infusion, perfusion or perifusion of the pancreas or isolated islets with PGs also alters hormone secretion from this organ (2,3). The evidence in the literature describes PGs as being both stimulatory (4,5) and inhibitory (6,7) to insulin secretion -in vivo or in vitro, depending upon variables such as the PG or PG synthetase inhibitor used, and the animal species studied. However, exogenous application of PGs to pancreatic preparations designed to study the physiological role of PGs in islet secretion may not mimic the actions of endogenous To date, few studPGs participating as intracellular messengers. ies have attempted to identify the PGs synthesized by the pancreas or isolated islets. PGE2 and PGF2a have been identified in homogenates of whole rat pancreas (8) as well as in extracts of human (9), dog, cat, rabbit and chicken pancreata (lo), as determined by thin layer chromatographic and bio-assay methods; PGEl and PGE2 were also identified in guinea-pig islets by radioimmunoassay (11). In contrast, pancreatic acinar cells have been described as devoid of cyclooxygenase activity (12). The purpose of the present investigation was to obtain a profile of PGs synthesized by homogenates of rat pancreatic islets, and to determine the PG metabolizing ability of rat islets. In the pursuit of as complete a profile of radiolabeled PGs synthesized by islet cycle-oxygenase as possible, we have employed both thin layer and high performance liquid chromatography techniques. MATERIALS PGD PGF GKPGF le;c;;;;zE2 and 15KH2F2, were gifts PGE Tritiated prostaglandins, from the 2' pjohn2~o.,K%.%~azoo, arachidonic acid and 14C-labeled arachidonic acid were obtained from New England Nuclear, Boston, MA. Hank's Balanced Salt Solution was from Grand Island Biological Co., Grand Island, N.Y. Collagenase (CLS IV) was from Worthington Biochemical Corp., Freehold N. J. KH2P04 (reagent grade) was from J.T. Baker, Phillipsburg, N.J. Reduced glutathione was from Sigma Chemical Co., St. Louis, MO. NAD and NADP were from Boehringer Mannheim. W. Germany. EDTA, and all organic solvents (HPLC Grade) were from Fisher Scientific Co., Silver Springs, MD. Silicic acid (100-200 mesh) was from Rio-Rad, Richmond, CA. Silica Gel G plates were from Analtech, Newark, Delaware. Microporasil silicic acid and fatty acid analysis HPLC columns, pumps and programmer were from Waters Associates, Milford, MA. Budget solv and 3a20 scintillation cocktails were from Research Products International, Grove Village, Illinois. METHODS Isolation of Islets. Male Sprague Dawley rats (ZOO-250 g) were fasted overnight prior to decapitation and distention of each pancreas with 10 ml Hank's Balanced Salt bicarbonate buffer (pH 7.4)
760
MAY
1981 VOL. 21 NO. 5
PROSTAGLANDINS
containing 3 mg collagenase The pancreas was then excised, minced, suspended in a final volume of 20 ml oxygenated Hank's buffer containing 6-8 mg collagenase and digested for 15 min, 37 C, at 250 rpm in a gyro-rotatory shaking water bath (New Brunswick Scientifit). Following this incubation, 30 ml cold Krebs-Ringer-bicarbonate (KRB) buffer (pH 7.4) was added to the pancreatic digest; the sedimenting tissue was washed three times. The final tissue sediment was examined by stereomicroscopy (15 x mag) and individual islets were isolated manually. Groups of 150-200 islets were washed with 1 ml KH2P04-NaOH buffer (0.05 M), pH 7.4, and resuspended in 525 ~1 KH2P04-NaOH buffer. Islets were then sonicated for 10 set with a micro-ultrasonic cell disrupter (Kontes). Protein was determined by the method of Lowrv et. al. (13). Incubation. Islet sonicates were diluted to a final mean protein concentration of 1 + 0.4 mgfml in 0.5 ml KH2P04-NaOH buffer. This final incubation buffer contained tritiated or 14carbon radiolabeled arachidonic acid (0.4 ? 0.1 nici); certain experiments also contained reduced glutathione (GSH) (1 mM) and sodium (tetra) ethylenediamine tetraacetate (EDTA) (4 mM) as indicated in the text. The final concentration of ethanol, used as the vehicle for arachidonate, was 0.01%. Islet sonicate mixtures were incubated in a shaking water bath (100 rpm) at 37 C for 1 hr, in an atmosphere of o,/co2 (95:5). Radiolabeled arachidonate was also incubated in buffer in the absence of islet sonicate (blank samples) in order to monitor spontaneous oxidative fatty acid conversion to PG-like compounds. Blank values were subtracted from the radioactive PG profiles obtained from incubations containing islet sonicate. In experiments designed to determine the PG metabolizing capability of islets, tritium-labeled PGE2 or PGF20 (0.5 nCi> was incubated with islet sonicate for 1 hr, in the presence of GSH and EDTA, and either a-nicotinamide adenine dinucleotide (NAD) or B-nicotinamide adenine dinucleotide phosphate (NADP) (4 mM). Tritium-labeled PG standards were incubated in buffer in the absence of sonicate in order to estimate the percent recovery during incubation and extraction procedures. The incubations were terminated with the addition of 500 i.rlof cold KH2P04-NaOH buffer, and acidification to pH3 with formic acid. Acidified aqueProstaglandin Isolation and Identification. ous islet samples were extracted 3 times with 10 ml ethyl acetate. The organic phases of extraction were pooled for analvsis using thin-layer (TLC) or high-performance liquid (HPLC) chromatography. Recovery of PGs averaged 70%. PG extracts were prepared for TLC analysis by prepurification on columns of silicic acid as described previously (14) and employing 10 ml benzene to remove unincorporated arachidonate from PGs. PGs recovered during this purification step were applied to Silica Gel G plates and developed in the solvent system chloroform-methanolacetic acid-deionized water (90:8:1:0.Bv/v) as described by Bailey et al. (15). PG standards were processed in parallel with experi-
MAY
1981 VOL. 21 NO. 5
761
PROSTAGLANDINS
mental samples in order to determine Rf values. Following plate development, the PGs were visualized with iodine vapor. TLC plates were scraped at 1 cm intervals from the origin to the solvent front. The scraped zones were suspended in scintillant and radioactivity levels were determined by liquid scintillation spectrometry (52% efficiency). PGs in tissue extracts were also analyzed by HPLC silicic acid and reversed phase (fatty acid) systems, as described in detail elsewhere (16). Fractions collected during elution of the various PGs were analyzed for radioactivity by liquid scintillation spectrometry. Statistical Methods.
Values shown are mean ?: S.E. RESULTS ____
Thin-layer Chromatographic Analysis of PGs. Rat islet sonicate when incubated with radiolabeled arachidonic acid synthesized several primary PGs and PG metabolites. Fig. 1A is a representative thin-layer chromatographic profile of the PG compounds synthesized by disrupted rat islets. The tritiated arachidonate substrate was largely converted to PGs migrating in three distinct zones on the thin-layer plates. PGF2o (F20) and thromboxane B2 (TXB2) demonstrate similar Rf values and dominate the first peak of PG material appearing nearest to the origin. The next broad peak of radiolabeled material corresponds to Rf values for the coincident PGE2 (E2)/ 6-keto-PGFL (6KFLo) and the closely migrating but distinct pair PGD2 (D2)/1?-keto-13,14-dihydro-PGF2a (15KH2F2,). And lastly, the PGE metabolite 15-keto-13,14-dihydro-PGE2 (15KH2E2) migrates as a discrete zone. The relative proportions of the different PGs and PG metabolites synthesized by islet sonicates are illustrated in Fig. 1B. After an hour of incubation, the percentage of arachidonic acid incorporated into islet PGs is largely distributed among the PGE2/ 6-keto-PGFlo, PGD2/ 15-keto-13,14-dihydro-PGF20 and 15-keto-13,14dihydro-PGE2 species of unsaturated fatty acids. A slightly lesser amount of label was found in the PGF2,/TXB2 zone when the incorporation of arachidonate into PGs for several experiments were averaged. The total incorporation of labeled substrate into PGs varied considerably among experiments due to variation in protein concentration of the sonicates, as well as animal variability. Thus, the individual PG data are best expressed as relative percent of total labeled arachidonate found in the PG zones. When islet sonicate was diluted to 20% (0.34 mg protein/ml) of the original protein concentration and incubated with radiolabeled arachidonic acid the total amount of radioactivity (4876 cpm) incorporated into PGs (PGE2/ 6-keto-PGFlo, PGD2/ 15-keto-13,14-dihydro PGF2o, and 15-keto-13,14-dihydro PGE2) was 27% of the total label incorporated into PGs in the undiluted sonicate (18,003 cpm). The addition of GSH and EDTA to the incubation did not change the incorporation of radiolabel into PGs by islet sonicates.
762
MAY
1981 VOL. 21 NO. 5
PROSTAGLANDINS
4
E2 TXB2
6K F,a
15KH2F2_
5
6
‘5W2E2
02
Fig. 1. (A) Representative thin-layer chromatographic profile of Incubation media contained PGs synthesized by rat islet sonicates. 3H-arachidonic acid, GSH, and EDTA. Cross-hatched areas represent areas of Silica Gel G plates scraped and counted to determine total cpm in PGs, which in (B) was used to calculate percent arachidonate incorporation in each PG zone. ( n = 5 or 6).
MAY
1981 VOL. 21 NO. 5
763
PROSTAGLANDINS
High Performance Liquid Chromatographic (HPLC) Analysis of PGs. Extracts of islet sonicates previously incubated with radiolabeled arachidonic acid were subject to analysis for PGs using both silicic acid and reversed phase (fatty acid) HPLC. Enhanced separation and resolution of PGs on columns of silicic acid made it possible to identify PGF2, in islet sonicate extracts as a separate entity from TXB2 (Fig. 2A). PGE2, 6-keto-PGFl, and TXB2 elute together on silicic acid columns, and constitute a distinct peak of labeled compounds synthesized by islets. The 15-keto-13, 14-dihydro PGE2 metabolite elutes just prior to 15-keto-13,14-dihydro PGF20 and PGD2. The latter two PGs have similar retention times and constitute a broad peak of labeled islet PG, with PGD2 perhaps appearing as a very minor shoulder of radioactivity on the PGF2, metabolite (Fig. 2A). The other labeled compounds identified early in the elution profile are arachidonic acid (A.A.) and unspecified hydroxy- or hydroperoxy-eicosatetraenoic acids (ETE). When the relative amounts of label incorporated into different PGs on HPLC were determined as a percent of the total number of counts incorporated into islet PGs, PGF2o incorporated 13% of the label (Fig.2B). The combined PGE2/6-keto-PGFla/TXB2 fraction incorporated almost 28% of the labeled arachidonate, as did the 15-keto-13,14-dihydro PGE2 fraction. The combined 15-keto-13,14dihydro PGF20/PGD2 fraction incorporated 24% of the labeled precursor found in PGs. These percentages are roughly predicted by the data in Fig.2A, however, the variability among experiments minimized somewhat the difference between different PGs. The reversed phase HPLC system helped to further define and identify PGs synthesized by islet sonicates since it resolves 6keto-PGFlo and TXB2 from PGE2. Fig. 3A shows the profile of islet PGs obtained using reversed phase chromatography. The profile supports the data obtained using silicic acid HPLC showing that PGF2o is turning over in islets less rapidly than is PGE2 and 6-ketoPGFl,. In four experiments analyzed by reversed phase HPLC, PGF2o arachidonate incorporation was equivalent to only 19 ?r 2% of the radiolabel incorporated into PGE2, 6-keto-PGFlo and TXB2; in addition, the amount of radiolabel found in PGF2o was only 48 rf 3% of the arachidonate activity found in the PGE2 fraction alone. In several experiments, the silicic acid HPLC fraction containing PGE2/6-keto-PGFl,/TXB2 was subjected to additional reversed phase HPLC. These experiments showed (Fig. 3B) that 6-keto-PGFlo contained 51% of the radiolabeled arachidonic acid originally eluted in the silicic acid fraction, whereas PGE2 and TX52 comprised only 36% and 16%, respectively, of the labeled compounds. Metabolism of PGE2 and PGF2, In Islet Sonicates. The incubation of tritiated PGE2 and PGF2o with islet sonicates containing GSH,EDTA and either NAD or NADP resulted in the conversion of each of the primary PGs to its 15-keto-13,14-dihydro metabolite. The conversion of labeled PGE2 to its 15-keto-13,14-dihydro metabolite in the presence of NAD and islet sonicate was 12%. Whereas NAD has been
764
MAY 1981 VOL. 21 NO. 5
PROSTAGLANDINS
Fig. 2. (A) Representative silicic acid HPLC profile of PGs synthesized by rat islet sonicates. Tissue was incubated with 3H-arachidonic acid, GSH, and EDTA. Cross-hatched areas delineate fractions counted and pooled to determine total cpm in PGs, which in (B) was used to calculate percent of arachidonate incorporation in each PC fraction (n = 6).
MAY 1981VOL. 21 NO. 5
765
PROSTAGLANDINS
210 K2
0
pzza aem ,..,....,....,.*..,....,....,...., 7
10
16
20
limo Flow=
M 25
30 (min) 2 ml/min
36
40
60 r
=2
50
0
5
40
(P30 P : ;; 0 I:
20 IO 0
Fig. 3. (A) Representative reversed phase HPLC profile of PGs synthesized by rat islet sonicates. Values represent 50% of PGs prepurified and eluted from silicic acid HPLC. Incubation conditions were as described in Fig. 1. (B) Percent distribution of radiolabeled arachidonic acid incorporation into PGs separated using reversed phase HPLC. PGs had been previously eluted in the E7/6-ketoFlo/TXB* fraction from silicic acid HPLC, and were re-chromatographed (n = 4).
766
MAY 1981 VOL. 21 NO. 5
PROSTAGLANDINS
described as the preferred cofactor for dehydrogenase enzymes metabolizing PGE2, NADP is preferred by enzymes metabolizing PGF2o (17). Thus, in the presence of NADP and tritiated PGF2o, islet sonicates metabolized 36% of the parent compound to the 15-keto-dihydro derivative. DISCUSSION PGs have been proposed as mediators of secretion in a number of cell types in which their synthesis has been confirmed (18). Analysis of pancreatic homogenates obtained from several species of animals, including man, have indicated that at least certain PGs are found in this tissue (8,9,10). However, the pancreas is composed of heterogeneous cell types, including exocrine acinar tissue, duct and vascular elements, as well as pancreatic islets which are composed of at least three endocrine cell types and vascular tissue (19). Isolated rat islets allow the advantage of localizing PG identification to the pancreatic endocrine system which, although composed of diverse cell types, is characterized as a paracrine secretion system which loses major portions of its secretory products to adjacent tissue elements. Thus, PGs synthesized and released from any one cell type might influence the physiology of neighboring cells, making the elucidation of islet PGs relevant to the entire tissue's function. Our studies have identified several primary PGs, as well as their metabolites, which are synthesized by rat islet sonicates. Sonicated islet preparations were chosen over intact islet incubations in these studies since the incorporation of labeled arachidonit acid into prostaglandins is limited in intact tissue. We have likewise found that the specific activity of the fatty acid in other lipid fractions of intact islets is also low (unpublished observations). The incorporation of arachidonic acid into prostaglandin and other fractions of intact islets may be limited by the physical impedencc of the capsular membrane or the biochemistry of the islet generally. Using disrupted tissue, however, TLC and HPLC methods have illustrated that PGF2n and PGE2 equivalent compounds are synthesized by rat islet cycle-oxygenase. These results corroborate previous findings that arachidonate is converted to PGF2 and PGE2, as identified in extracts of the perfused rat pancreas ? 8). In addition, we have demonstrated the presence in islets of 15-hydroxy PG dehydrogenase and 13,14-dihydro PG reductase (20) through the metabolism of both endogenous and exogenous PGE2 and PGF2o to their 15-keto-13,14-dihydro metabolites. Complete characterization of these latter enzymes will be important in determining the physiological role of PGs in islet secretory function, since the metabolism of PGs may be intimately related to the nucleotide levels of cells which are altered by changes in secretory activity (21). Resolution of TX-like compounds on HPLC showed a low level of arachidonate incorporation into these products as well. Based upon the data, the synthesis of PGD2 by islet sonicate would appear to be of modest proportions, at best.
MAY 1981 VOL. 21 NO. 5
767
PROSTAGLANDINS
The other major PG synthesized by islet sonicates, and clearly identified by HPLC methods as distinct from other PGs, was 6-ketoPGFlo, the metabolite of prostacyclin (PG12) (22). Thus, pancreatic islets possess in addition to cycle-oxygenase, prostacyclin synthetase which converts PG endoperoxides to PGI2 (23). The metabolite 6-keto-PGFlo has previously been identified in isolated rat adrenocortical cells (24); this is a system where vascular elements have been eliminated, thus allowing for the identification of PG12 synthesis solely by endocrine cells. Unfortunately, there are vascular elements in islets in addition to the endocrine cells, and either or both of these elements may be contributing to the PGs synthesized by islet sonicates. Future studies utilizing isolated islet cells will help to elucidate the origin of the islet PGs. The extent to which the small, unavoidable acinar tissue contamination of isolated islets contributes to the PGs identified in these preparations is difficult to assess. In rat acinar tissue sonicates processed in parallel with our islet preparations we have identified arachidonate incorporation into PGF2o and PGE2 equivalents as well as into the metabolites of these compounds, although no PG synthesis was identified in exocrine tissue by other investigators (12). However, the incorporation of labeled arachidonate into islet PGs was more than 2-fold higher than that appearing in exocrine tissue PGs as calculated per mg of protein in the respective sonicates (unpublished observation). Thus, the trace amounts of exocrine tissue adhering to isolated islets would not be expected to greatly contribute to the profile of islet PGs reported. Analysis of arachidonate incorporation into PG products by HPLC also revealed that this fatty acid is incorporated into non-PG hydroperoxy compounds which chromatograph on silicic acid as the hydroxyeicosatetraenoicacid (HETE) (25). These hydroxy fatty acids are products of an enzyme, lipoxygenase (26),which has heretofore not been demonstrated in pancreatic tissue. A biological role for the HETE compounds in endocrine function remains to be determined, however, some of the other PGs identified in this study have been shown to influence islet secretion. Exogenous PGE2 and PGF2, can enhance insulin and glucagon release from the perfused pancreas, whereas PGD2 is a more potent secretagogue for glucagon than for release of insulin, and PG12 and TX appear to have no effect on secretion of either hormone (27). Thus, the rat pancreatic islet synthesizes a variety of PGs which may have a physiological role in hormone secretion. Future studies will determine whether or not stimulation of secretory function in intact cells alters the synthetic profile of islet PGs. REFERENCES 1. 2. 3.
768
Flack, J.D. In: The Prostaglandins (P.W. Ramwell, ed.), Plenum Press N.Y., 1973, Vol. 1, pp. 327-345. Pek, S., Lands, W.E.M., Akpan, J., and Hurley, M. Adv. Prostaglandin and Thromboxane Res. 8: 1295-1298, 1980. Robertson, R.P. Diabetes -28: F43-948, 1979.
MAY 1981 VOL. 21 NO. 5
PROSTAGLANDINS
4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.
20. 21. 22. 23. 24. 25. 26. 27.
Burr, I.M., and Sharp, R. Endocrinology 94: 835-839, 1974. Johnson, D.G., Fujimoto, W.Y., and Williams, R.H. Diabetes -22: 658-663, 1973. Robertson, R.P., Gavareski, D.J., Porte, D.,Jr., and Bierman, E.L. J. Clin. Invest. 54: 310-315, 1974. Sacca, L., Rengo, F., Chiariello, M., and Condorelli, M. Endocrinology 92: 31-34, 1973. Hamamdzic, Mx and Malik, K. Am. J. Physiol. 232: 201-209, 1977. Karim, S.M.M., Sandler, M., and Williams, E.D. Br. J. Pharmac. Chemother. 31: 340-344, 1967. Karim, S.M.E, Hillier, K., and Devlin, J. J. Pharm. Pharmac. 20: 749-753, 1968. zyckx, A.S., and Lefebvre, P.J. Adv. Prostaglandin and Thromboxane Res. 8: 1299-1302, 1980. Chauvelot, L., Heisler, S., Huot, J., and Gagnon, D. Life Sci. 25: 913-920, 1979. zwry, O.H., Rosebrough, N.J., Farr, A.L., and Randall, R.J. J. Biol. Chem. 193: 265, 1951. Laychock, S.G., and Rubin, R.P. Prostaglandins -10: 529-540, 1975. Bailey, J.M., Bryant, R.W., Feinmark, S.J., and Makkeja, A.N. Prostaglandins 13: 479-492, 1977. Whorton, A.R., Szgel, M., Oates, J.A., and Fralich, J.C. Biochim. Biophys. Acta 529: 176-180, 1978. Pace-Asciak, C.R. In: Renal Prostaglandins (James B.Lee, ed.) Eden Press, Vermont, 1978, Vol. 1, pp. 55-83. Rubin, R.P. and Laychock, S.G. In: Calcium in Drug Action (G. Weiss, ed.) Plenum Press, N.Y. 1978, pp. 135-155. Wellmann, K.F., and Volk, B.W. In: The Diabetic Pancreas. (Bruno W. Volk and Klaus F. Wellmann, eds.). Plenum Press, N.Y., 1977, pp. 99-128. Anggard, E., and Larsson, C. Eur. J. Pharmacol. -14: 66-70, 1971. Ammon, H.P.T., and Verspohl, E. Endocrinology -99: 1469-1476, 1976. 17: 4096-4101, 1978. Sun, F.F., and Taylor, B.M. Biochemistry Sun, F.F., Chapman, J.P., and McGuire, J.C.Prostaglandins 14: 1055-1074, 1977. zychock, S.G., and Walker, L. Prostaglandins -18: 793-811, 1979. Gr&en, K., Hamberg, M., Samuelsson, B., and Frijlich, J.C. Adv. Prostaglandin and Thromboxane Res. 5: 15-38, 1978. Samuelsson, B., Borgeat, P., Hammarstrgm, S., and Murphy, R.C. Adv. Prostaglandin and Thromboxane Res. 6: l-18, 1980. Akpan, J.O., Hurley, M.C., Pek, S., and Lands, W.E. Can. J. Biochem. -57: 540-547, 1979. Editor:
Harold R. Behrman
MAY 1981VOL. 21 NO. 5
Rec'd: 9-80
Acpt: 3-81
769