Proc. Nati. Acad. Sci. USA Vol. 76, No. 3, pp. 1155-1159, March 1979
Biochemistry
Prostaglandin synthesis in isolated rat kidney glomeruli (prostacyclin/prostaglandin F2%/prostaglandin E2/prostaglandin D2/thromboxane B2)
Aviv HASSID*t, MARTHA KONIECZKOWSKI*, AND MICHAEL J. DUNN* *Nephrology Division, Department of Medicine and tDepartment of Biochemistry, Case Western Reserve University, Cleveland, Ohio 44106
Communicated by Oscar D. Ratnoff, December 14, 1978
ABSTRACT Isolated rat kidney glomeruli converted octatritiated arachidonic acid to several prostaglandins whose production was inhibited by meclofenamate. These were, in order of decreasing abundance, prostaglandin F2a, prostaglandin E2, 6-oxo-prostaglandin Fla, thromboxane B2, and prostaglandin D2. These products were identified by thin-layer chromatography, before and after treatment with potassium hydroxide or sodium borohydride. Prostaglandins F2a and E2 were also determined by radioimmunoassay. The major product made by glomeruli was an unidentified substance(s), whose appearance was partially inhibited by meclofenamate, and was likely to be a hydroxylated fatty acid(s). The specific activity of glomerular fatty acid cyclo-oxygenase (EC 1.14.99.1), based on radioimmunoassay for prostaglandins E2 and F2,, was 10- to 40-fold higher than that of cortical tubular enzyme. These data demonstrate that glomeruli have the capability of synthesizing an array of end-products from arachidonic acid. These prostaglandins may exert important physiologic effects, because renin secretion and arteriolar resistance are regulated by the glomerulus and the afferent and efferent arterioles.
Increasing evidence exists which indicates that renal prostaglandins exert important influence on renal cortical functions such as renin secretion and renal vascular resistance as well as on medullary functions such as urine concentration (1). Initial work on the localization of prostaglandin synthesis within the kidney stressed medullary and papillary sites (2, 3). More recent efforts, using cortical slices or membranes, have clearly demonstrated synthesis of prostaglandin (PG) E2 and PGF2a (4-6). Using [1-14C]PGH2 as substrate, Zenser et al. (7) have reported that rat renal cortical membranes synthesize thromboxane (Tx) B2 and 6-oxo-PGFi, in addition to PGD2, PGE2, and PGF2a. The syntheses of TxB2 and 6-oxo-PGFia were greater in cortex than in the inner medulla (7). Whorton et al. (8) have demonstrated the conversion of 14C-labeled arachidonic acid (C20:4), by rabbit renal cortical microsomes, to 6-oxo-PGFia, PGE2, PGF2a, and PGD2. Medullary tissue produced significant amounts of PGE2, but no 6-oxo-PGFia was detected (8). Needleman and coworkers (9), using perfused rabbit kidneys, have also observed synthesis of PGI2 [6(9)-oxido-11a-15Sdihydroxyprosta-5,13-dienoic acid (prostacyclin)], determined as the appearance of the stable metabolite 6-oxo-PGFi,. The-precise anatomical localization of prostaglandin production within the kidney is not well known. Histochemical and immunofluorescent techniques have detected synthetic enzymes (fatty acid cyclo-oxygenase, EC 1.14.99.1) in collecting tubules and medullary interstitial cells (10, 11). Cortical sites of prostaglandin synthesis have been defined less precisely. Smith and Bell (12), using immunohistochemical microscopy, have variably found fatty acid cyclo-oxygenase in glomeruli as well as in arterioles and cortical collecting tubules. Terragno et al. (13) have reported synthesis of PGI2, measured as 6oxo-PGFia, in arteries and arterioles that were microdissected The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "ad-
and removed from pig kidney. There have been no reports of the capabilities of isolated glomeruli to convert C20:4 to prostaglandin or thromboxane end-products. We hypothesized that the glomerulus might synthesize prostaglandins and thromboxane because of the growing evidence that afferent and efferent arteriolar resistance (1, 14), renin production (15-17), and glomerular filtration (18) may be altered by inhibition of prostaglandin synthesis or by infusion of prostaglandins. We have determined the identity and abundance of the prostaglandins generated from endogenous and exogenous C20:4 by isolated rat glomeruli and nonglomerular cortical tissue. METHODS Isolation of Glomeruli and Incubation with [3HJC20:4. Kidneys were removed from four to nine rats (Sprague-Dawley of either sex, 250-500 g) under ether anesthesia. Cortex was separated from medulla and was finely minced. The glomeruli were separated from other cortical tissue by passing the minced tissue through sieves of different sizes, according to the technique of Misra (19). Glomeruli were obtained from the 105- and 74-am sieves, and tubules were removed from the 210-,m sieve. The purities of the isolated glomerular and nonglomerular fraction were determined microscopically, by counting the number of glomerular and nonglomerular particles suspended in a given volume. The purity of the glomerular fraction varied between 90 and 98%. Virtually none of the isolated glomeruli contained the afferent and efferent arterioles. The nonglomerular fraction, consisting of cortical tubular fragments, was essentially free of contaminating glomeruli. The glomerular and nonglomerular fractions were preincubated for 15 min at room temperature in the presence of 50.9 AM dexamethasone (Sigma), in Earle's balanced salt solution (GIBCO), in order to inhibit deacylation of endogenous phospholipids and maximize conversion of added C20:4 (20, 21). After 15 min, the tissues were precipitated by gentle centrifugation and were divided into two. The first half was resuspended and incubated for 10 min at 37°C in Earle's balanced salt solution containing meclofenamate (26.8 AM) (gift of Warner Lambert/Parke Davis, Ann Arbor, MI) and dexamethasone (50.9 ,M), whereas the second half was incubated with dexamethasone in the absence of meclofenamate. In the experiment shown in Fig. 1, the protein content per tube of glomerular incubation was 1.32 mg. Protein was determined by the method of Lowry et al. (22) with bovine serum albumin as standard. The number of glomeruli per tube was approximately 1.4 X 104, as determined by counting the glomeruli in an aliquot. [3H]C20o4 (1-3 ,Ci, specific activity 80 Ci/mmol; 1 Ci = 3.7 X 1010 becquerels) or nonradioactive C20:4 (final concentration 16.4 ,M) was added to initiate the reaction in a final volume of 0.9 ml. After incubation for 20 min at 37°C, 0.4-ml aliquots were removed and meclofenamate was added Abbreviations: PG, prostaglandin; PGI2, 6(9)-oxido-lla-15S-dihydroxyprosta-5,13-dienoic acid (prostacyclin); Tx, thromboxane; TLC, thin-layer chromatography; RIA, radioimmunoassay; C20:4, arachidonic
vertfsement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.
acid.
1155
1156
Proc. Natl. Acad. Sci. USA 76 (1979)
Biochemistry: Hassid et al.
to those aliquots that had not received it before the incubation (final concentration 26.8 ,M). After 40 min, second and final aliquots were taken and treated similarly to the first samples. Identical incubations were carried out in the absence of glomeruli and tubules to assess nonenzymatic transformation of C204. Nonradioactive and radioactive C20:4 (purchased from Sigma and New England Nuclear, respectively) was purified before use by silica gel column chromatography according to the procedure of Flower et al. (23). Purified C20:4 was stored in hexane under nitrogen at -35°C in the dark. Extraction of Prostaglandins and Thin-Layer Chromatography (TLC). Extraction of acidic lipids, including prostaglandins, was achieved by lowering the pH to 3.0-3.5 with 0.5 M HC1 followed by extraction twice with 3 vol of ethyl acetate. An aliquot of the extracted radioactive materials was spotted on silicic acid thin-layer plates (silica gel 60, E. Merck, Darmstadt, Germany) together with 5-10 ,g of prostaglandin standards, and developed twice in the organic phase of ethyl acetate/iso-octane/acetic acid/water (11:5:2:10, vol/vol). The prostaglandin and thromboxane standards were generously provided by J. Pike (Upjohn). The standards were visualized by exposing the thin-layer plates to iodine vapor. The plates were divided into 26-30 segments, including a segment for each of the standards. The silica gel from each segment was transferred to a scintillation vial, and the radioactivity was determined by scintillation counting in an aqueous counting mixture (Formula-963, New England Nuclear). Counting efficiency was 40-50% and was determined with a radioactive external standard. Prostaglandins were separated by TLC and eluted from the silica gel with acetone. Individual prostaglandins were then treated with 1 M KOH in methanol at room temperature for 1 hr. An identical incubation was carried out in methanol in the absence of KOH. After acidification and extraction with ethyl acetate, treated and untreated materials were spotted on TLC together with standards and developed in the solvent described above. The separated prostaglandins were also reduced with excess sodium borohydride in methanol at room temperature for 2 hr. The products were then determined by TLC and scintillation counting as described above. Radioimmunoassay (RIA). RIA for PGE2 and PGF2, was performed with specific antibodies purchased from Pasteur Institute, Paris, France. Anti-PGF2a does not distinguish between PGFia and PGF2a. In practice, this is not a problem because PGF2a is far more abundant than PGFia. 6-OxoPGFi,, PGE1, and PGE2 cross react with anti-PGF2a 1.6%, 0.3%, and 0.8%, respectively. PGAI, PGA2, PGBI, PGB2, and TxB2 crossreact less than 0.1%. PGE1, PGF2a, 13,14-dihydro-PGE2, and 13,14-dihydro15-keto-PGE2 crossreact with anti-PGE2 2.7%, 0.12%, 0.12%, and 0.10%, respectively. PGFia, PGA1, PGA2, PGB2, TxB2, and 6-oxo-PGFia crossreact less than 0.04% with anti-PGE2. [3H]PGE2 and [3H]PGF2a (100-200 Ci/mmol) were purchased from New England Nuclear. PGE2 and PGF2a were determined by RIA directly without purification. RESULTS Isolated glomeruli converted octatritiated C20:4 to products that were separable on silica gel TLC. In the experiment shown in Fig. 1, in the absence of meclofenamate, the radioactive peak corresponding to PGF2a had 1.2% of the total radioactivity recovered from the thin-layer plate. The peaks corresponding to PGE2, 6-oxo-PGFia, TxB2, and PGD2, had 0.47%, 0.47%, 0.30%, and 0.28%, respectively, of the total recovered radioactivity; 83.1% of the radioactivity was recovered unchanged
as C20:4 (not shown in Fig. 1). In other experiments, the radioactivity corresponding to the PG products ranged from 3 to 9% of recovered radioactivity. TLC of radioactive products in a solvent system of 99:1 (vol/vol) ethyl acetate/acetic acid gave results consistent with those of Fig. 1. Experiments performed in the absence of dexamethasone gave results qualitatively similar to those of Fig. 1. Cortical tubules incubated with octatritiated C20:4 also appeared to synthesize radioactive products that comigrated with prostaglandin standards. However, the yield for all prostaglandin products was much lower than that for glomeruli, even in the face of severalfold more tubular protein (results not shown). Because of the lower yield, tubular prostaglandin production could not be accurately quantitated. It appeared that the specific activity for glomerular prostaglandin synthesis was at least 20-fold higher than that for cortical tubular synthesis. In addition to comigration with authentic prostaglandins, the glomerular radioactive products were identified by the following criteria. (i) The amounts of radioactive materials comigrating with the prostaglandin standards were reduced to 10-40% when glomeruli were incubated in the presence of meclofenamate, a potent fatty acid cyclo-oxygenase inhibitor (Fig. 1). (ii) On treatment with KOH in methanol, the radioactivity corresponding to PGE2 and PGD2 was selectively decreased, relative to a control sample that was treated identically except for the absence of KOH (results not shown). Most of the radioactivity corresponding to PGE2 and PGD2 was lost. KOH treatment converted the radioactive material comigrating with PGE2 to a product comigrating with PGB2, thus confirming the identity of the original material as PGE2 (results not shown). We were unable to show significant increases in a new product on treatment of PGD2 with KOH, in spite of decreased radioactivity comigrating with PGD2 (results not shown). Therefore, the identification of PGD2 remains tentative. Radioactive materials corresponding to 6-oxo-PGFia, PGF2a, and TxB2 were essentially unchanged on treatment with KOH (compared 1400 PG F=
1200 1000
E 800 600
6-oxo-1
400
PGE2
6000
PGD2
TxB2
4000E 2000
200 2
4
6
8
10 12
14 16" 16
l
18
Distance from origin, cm
FIG. 1. TLC of products of octatritiated C20:4 metabolism in glomeruli incubated with or without meclofenamate. Incubation time was 20 min. Solid bars are in the absence of meclofenamate. Hashed bars are in the presence of 28.6 MM meclofenamate. dpm are corrected for nonenzymatic conversion of C20:4 by subtracting dpm obtained from an incubation and TLC performed under identical conditions, but in the absence of glomeruli. Note difference in dpm scale to the right of broken horizontal axis. C20:4 migrated between 17.4 and 18.1 cm (83,556 dpm in the absence and 96,438 dpm in the presence of meclofenamate) and is not shown in this figure. dpm values in the presence of meclofenamate were multiplied by a factor such that total radioactivity recovered equalled that of the experiment lacking meclofenamate, 100,511 dpm.
Hassid et al.
Biochemistry:
6
8
Proc. Natl. Acad. Sci. USA 76 (1979)
12
10
14
6 8 10 12 14 Distance from origin, cm
16
Distance from origin, cm
FIG. 2. TLC of radioactive glomerular 6-oXO-PGFIa before and after treatment with sodium borohydride. Glomerular 6-oxo-PGF1a was purified by TLC as described in text. Equal amounts of 6-oxoPGFia, were treated with 200 ,ud of methanol or with 4 mg of sodium borohydride in 200 ,ul of methanol. After 1 hr the products were recovered by acidification, extraction with ethyl acetate, and rechromatography on silica gel thin-layer. (Upper) Radiochromatogram generated by 6-oxo-PGFia incubated in methanol. (Lower) Radiochromatogram of borohydride-treated 6-oxo-PGFl,,.
to a control incubation in methanol alone), thus confirming the absence of a f3-hydroxy ketone moiety in these materials (results not shown). (iii) Borohydride treatment of the radioactive product corresponding to 6-oxo-PGFia resulted in partial conversion to products that had different chromatographic mobilities than the presumed starting material, 6oxoPGFIa, (Fig. 2). However, the conversion to reduced products was not complete, presumably because 6-oxo-PGFia is predominantly in the cyclic lactol form (24). We were also unable to reduce authentic 6oxo-PGFIa with sodium borohydride. These results confirm those of Pace-Asciak (24), who also found it difficult to reduce 6-oxo-PGFi, with borohydride. Borohydride treatment of TxB2 Control
400C TxB2
300C
E M
200C 1 00C
6-oxo- PGPG FILa F2C
PG E2
C20:4
PG- Tx B2 2< PGE2
C20:4
4000 3000
E
0.
2000 6-oxo-
1000
PGFia
2
4
6
8
10
12
14
16
18
Distance from origin, cm
FIG. 3. TLC of radioactive glomerular TxB2 before and after treatment with sodium borohydride. The protocol was similar to that described in text and Fig. 2. (Upper) Chromatogram of TxB2 treated with methanol. (Lower) Chromatogram of TxB2 treated with boro-
hydride.
1157
FIG. 4. TLC of radioactive glomerular PGD2 before and after treatment with sodium borohydride. The protocol was similar to that described in text and Fig. 2. (Upper) Chromatogram of PGD2 treated with methanol. (Lower) Chromatogram of PGD2 treated with borohydride.
gave a single more polar product (Fig. 3). This is consistent with the structure of TxB2 which, in part, exists as the acyclic aldehyde tautomer (25).t As expected, PGE2 was converted to two more polar products with sodium borohydride, one of which corresponded to PGF2a. The other product was not identified, but may be PGF2, (results not shown). PGD2 was converted by borohydride treatment to a radioactive product identical in chromatographic mobility to PGF2a (Fig. 4). PGF2a, as expected, did not generate significant new products when treated with sodium borohydride (results not shown). (iv) Glomeruli and cortical tubules incubated with nonradioactive C20:4 (17.3 ,uM) synthesized immunoreactive PGE2 and PGF2a (Table 1). Meclofenamate (27 ,uM) inhibited this synthesis by 60-80%. Glomeruli synthesized 40-fold more PGF2a and 10-fold more PGE2, per mg of protein, than tubules (Table 1). The ratio of glomerular PGF2a to PGE2 determined by RIA (Table 1) was similar to that determined by TLC (Fig. 1). Glomeruli incubated in the absence of exogenous C20:4 produced levels of immunoreactive PGF20 and PGE2 lower than those produced in its presence (Table 1). The appearance of immunoreactive PGE2 and PGF2a was inhibited by mepacrine, a cyclo-oxygenase and phospholipase A2 inhibitor. The appearance of the most abundant glomerular C20:4 metabolite or metabolites (migrating between 16.1 cm and 17.0 cm, Fig. 1) was inhibited less than 50% by meclofenamate. This partial inhibition may be due to comigration of a cyclo-oxygenaserelated product (e.g., 12L-hydroxy-5,8,10-heptadecatrienoic acid) with a product unrelated to cyclo-oxygenase (e.g., 12Lhydroxy-5,8,10,14-eicosatetraenoic acid). Alternatively, meclofenamate may inhibit a lipoxygenase product. The chromatographic mobility of this material or materials is consistent with that of hydroxylated fatty acid(s) (23). Cortical tubular enzyme also synthesized a product having similar chromatographic mobility (data not shown). * Note Added in Proof. Using specific antiserum to TxB2 (Pasteur Institute), we have verified the release of TxB2 by normal glomeruli incubated for 20 min with 10 nM C20:4.
1158
Biochemistry:
Hassid et al.
Proc. Nati. Acad. Sci. USA 76 (1979)
Table 1. RIA of PGF2a and PGE2 in incubations of glomerular and nonglomerular fractions in the presence or absence of C20:4 or inhibitors of prostaglandin synthesis Immunoreactive Inhibitor prostaglandin, Incubation of prostaglandin pmol/mg proteint C20:4, time, min AM* synthesis PGF2a PGE2 Incubations with glomerular fraction 20 17.3 None 88.9 46.5 20 17.3 18.3 Meclofenamate (27 MM) 8.79 40 17.3 89.8 59.3 None 40 17.3 Meclofenamate (27 MM) 21.0 12.5 60 0 19.5 None 16.0 60 0 1.80 0.82 Mepacrine (4.9 mM) ND 120 0 29.6 None 120 0 ND Mepacrine (4.9 mM) 2.86 Incubations with nonglomerular fraction 20 17.3 None 20 17.3 Meclofenamate (27 MM) 40 17.3 None 40 17.3 Meclofenamate (27 MM)
1.98 0.82 3.94 1.15
3.71 2.00 6.77 2.35
* C20:4 represents externally added amounts only. Total C20:4 available to fatty acid cyclo-oxygenase, which includes that liberated from endogenous sources by phospholipase A2, was not determined. Incubation conditions are described in the text in detail. t Values expressed for total time of incubation. ND, not determined.
DISCUSSION Our results demonstrated the synthesis of significant quantities of PGF2a, PGE2, 6-oxo-PGFIa, TxB2, and PGD2 by rat glomeruli. PGE2 and PGD2 have a f3-hydroxy ketone structure in the cyclopentane ring, and were therefore susceptible to dehydration by a base-catalyzed reaction (26). 6-oxo-PGFIa, PGF2a, and TxB2 were not altered by base treatment. These results are consistent with the initial identification of radioactive glomerular prostaglandins based on comigration with prostaglandin standards on TLC (Fig. 1). Sodium borohydride treatment resulted in decreased amounts of PGE2, PGD2, 6oxo-PGFia, and TxB2, but not of PGF2a, consistent with reduction of keto groups to alcohol groups. RIA data for PGE2 and PGF2a further strengthened the structural designations based on comigration with standards on TLC. The relative abundance of the prostaglandins was quite similar to that obtained in rat cortical membranes and in rabbit cortical microsomes (7, 8). This suggests the absence of significant conversion of PGE2 to PGF2, in glomeruli by the soluble enzyme prostaglandin-9-ketoreductase (27) and that PGF2a is most likely produced directly from C20:4. The lower specific activity of cyclo-oxygenase in nonglomerular cortical tissues, 1/10th to 1/40th, cannot be due to increased metabolism of prostaglandins by 15-hydroxyprostaglandin dehydrogenase and A-13-prostaglandin reductase, because no significant radioactive product whose formation was inhibited by meclofenamate could be detected in regions of the chromatograms expected to contain prostaglandin metabolites (15-keto prostaglandins or 13,14-dihydro 15-keto prostaglandins). This suggests that a significant and perhaps major portion of renal cortical prostaglandin synthesis occurs in glomeruli, despite larger amounts of nonglomerular tissue. Until recently, it was assumed that prostaglandins synthesized in the renal medulla were transported to the renal cortex via the tubular fluid and thereby controlled some cortical physiologic functions (28). However, our results with isolated glomeruli and the data of others using cortical membranes and microsomes (7, 8) verify the capacity of the renal cortex to synthesize a spectrum of prostaglandins. It seems most likely that important cortical functions, such as renin secretion and control
of renal blood flow, may be influenced by prostaglandins produced locally in glomeruli and~afferent and efferent arterioles rather than by prostaglandins produced in the renal medulla. Our results demonstrate that glomeruli possess much greater prostaglandin synthetic capacity than cortical tubules do. The evidence supporting important renal cortical actions of prostaglandins is good. C20,4 and prostaglandin endoperoxides released renin from rat, rabbit, and dog kidneys studied in vivo or as slices in vitro (29-32). Furthermore, PGE2 and PGI2 stimulated renin secretion in both the filtering and nonfiltering dog kidney in vivo (15, 16, 33), presumably because of a direct effect on the juxtaglomerular cells. Most prostaglandins, and especially PGE2 and PGI2, vasodilate renal arterioles, and TxA2 vasoconstricts these vessels. Vasoconstriction mediated by either a adrenergic agonists or by angiotensin II is dramatically potentiated after inhibition of the fatty acid cyclo-oxygenase (34, 35). This undoubtedly is the result of diminished synthesis of PGI2 and PGE2 by glomeruli and their arterioles. Conversely, TxA2 probably mediates the renal vasoconstriction observed after ureteral obstruction (36, 37). The glomeruli may be the locus of synthesis of TxA2 after ureteral obstruction, because elevated ureteral pressure immediately alters glomerular hydrodynamics. Finally, micropuncture experiments have shown that infusion of prostaglandins (18) or inhibition of fatty acid cyclo-oxygenase (14) alters glomerular ultrafiltration and afferent and efferent arteriolar resistances. These diverse actions of prostaglandins and thromboxanes on renal cortical function in health and disease reinforce the potential importance of our observations that rat glomeruli can synthesize PGF2a,, PGE2, 6-oxo-PGFia, TxB2, and PGD2. We acknowledge the expert technical assistance of Leila Jacobs, Martin Brett, and Dr. Alice S. Petrulis and the secretarial help of Wendy Ingaran and Illein Youngman. This work was supported by National Institutes of Health Grant HL-22563 and the American Heart Association Grant AHA-77916.
1. Dunn, M. J. & Hood, V. L. (1977) Am. J. Physiol. 233, F169F184. 2. Anggard, E.,
Bohman, S. O., Griffin, J. E., Larsson, C. & Maunsbauch, A. B. (1972) Acta Physiol. Scand. 84, 231-246.
Biochemistry: Hassid et al. 3. Crowshaw, K. & McGiff, J. C. (1973) in Mechanisms of Hypertension, ed. Sambhi, M. P. (Elsevier, New York), pp. 254-273. 4. Larsson, C. & Anggard, E. (1973) Eur. J. Pharmacol. 21, 3036. 5. Larsson, C. & Anggard, E. (1976) J. Pharm. Pharmacol. 28, 326-328. 6. Pong, S. S. & Levine, L. (1976) Res. Commun. Chem. Pathol. Pharmacol. 13, 115-123. 7. Zenser, T. V., Herman, C. A., Gorman, R. R. & Davis, B. B. (1977) Biochem. Biophys. Res. Commun. 79,357-363. 8. Whorton, A. R., Smigel, M., Oates & Frolich, J. C. (1978) Biochim. Biophys. Acta 529, 176-180. 9. Needleman, P., Bronson, S. D., Wyche, A., Sivakoff, M. & Nicolaou, K. C. (1978) J. Clin. Invest. 61, 839-849. 10. Janszen, F. H. & Nugteren, D. H. (1971) Histochemistry 27, 159-164. 11. Smith, W. L. & Wilkin, G. P. (1978) Am. J. Physiol. 235, F451-F457. 12. Smith, W. L. & Bell, T. G. (1978) Prostaglandins 15, 715 (abstr.). 13. Terragno, N. A., McGiff, J. C. & Terragno, A. (1978) Clin. Res. 26, 545 (abstr.). 14. Bayliss, C. & Brenner, B. M. (1977) Kidney Int. 12, 550
(abstr.). 15. Data, J. L., Gerber, J. G., Crump, W. J., Frolich, J. C., Hollifield, J. W. & Nies, A. S. (1978) Circ. Res. 42, 454-458. 16. Bolger, P. M., Eisner, G. M., Ramwell, P. W. & Corey, E. J. (1978) Nature (London) 271, 467-469. 17. Yun, J., Kelly, G., Bartter, F. & Smith, H. (1977) Circ. Res. 40, 459-464. 18. Bayliss, C., Deen, W. M., Myers, B. D. & Brenner, B. M. (1976) Am. J. Physiol. 230, 1148-1158. 19. Misra, R. P. (1972) Am. J. Clin. Pathol. 58, 135-139. 20. Flower, R. J. (1978) in Advances in Prostaglandins and
Proc. Natl. Acad. Sci. USA 76 (1979)
1159
Thromboxane Research, eds. Galli, C., Galli, G. & Porcellati, G. (Raven, New York), Vol. 3, pp. 105-112. 21. Hong, S. & Levine, L. (1976) Proc. Natl. Acad. Sci. USA 73, 1730-1734. 22. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem. 193,265-275. 23. Flower, R. J., Cheung, H. S. & Cushman, D. W. (1973) Prostaglandins 4,325-341. 24. Pace-Asciak, C. (1976) Experientia 32, 291-292. 25. Hamberg, M. & Samuelsson, B. (1974) Proc. Natl. Acad. Sci. USA 71,3400-3404. 26. Bergstrom, S., Rhyage, R., Samuelsson, B. & Sj6vall, J. (1963) J. Biol. Chem. 238,3555-3564. 27. Hassid, A. & Levine, L. (1977) Prostaglandins 13,503-516. 28. Frolich, J. C., Wilson, T. W., Sweetman, B. J., Smigel, M., Nies, A. S., Carr, K., Watson, J. T. & Oates, J. A. (1975) J. Clin. Invest. 55,763-770. 29. Weber, P., Holzgreve, H., Stephan, R. & Herbst, R. (1975) Eur. J. Pharmacol. 34,299-304. 30. Larsson, C., Weber, P. & Anggard, E. (1974) Eur. J. Pharmacol. 28,391-394. 31. Bolger, P. M., Eisner, G. M., Ramwell, P. W. & Slotkoff, L. M. (1976) Nature (London) 259, 244-245. 32. Bolger, P. M., Eisner, G. M. & Slotkoff, L. M. (1978) Fed. Proc. Fed. Am. Soc. Exp. Biol. 37,306 (abstr.). 33. Aiken, J. W. & Vane, J. R. (1975) J. Pharmacol. Exp. Ther. 184, 678-687. 34. Satoh, S. & Zimmerman, B. G. (1975) Circ. Res. Suppl. 1, 8996. 35. Henrich, W. L., Berl, T., McDonald, K. M., Anderson, R. J. & Schrier, R. W. (1978) Am. J. Physiol. 235, F46-F51. 36. Morrison, A. R., Nishikawa, K. & Needleman, P. (1977) Nature (London) 267, 259-260. 37. Morrison, A. R., Nishikawa, K. & Needleman, P. (1978) J. Pharmacol. Exp. Ther. 205, 1-18.
Biochemistry
Prostaglandin synthesis in isolated rat kidney glomeruli (prostacyclin/prostaglandin F2%/prostaglandin E2/prostaglandin D2/thromboxane B2)
Aviv HASSID*t, MARTHA KONIECZKOWSKI*, AND MICHAEL J. DUNN* *Nephrology Division, Department of Medicine and tDepartment of Biochemistry, Case Western Reserve University, Cleveland, Ohio 44106
Communicated by Oscar D. Ratnoff, December 14, 1978
ABSTRACT Isolated rat kidney glomeruli converted octatritiated arachidonic acid to several prostaglandins whose production was inhibited by meclofenamate. These were, in order of decreasing abundance, prostaglandin F2a, prostaglandin E2, 6-oxo-prostaglandin Fla, thromboxane B2, and prostaglandin D2. These products were identified by thin-layer chromatography, before and after treatment with potassium hydroxide or sodium borohydride. Prostaglandins F2a and E2 were also determined by radioimmunoassay. The major product made by glomeruli was an unidentified substance(s), whose appearance was partially inhibited by meclofenamate, and was likely to be a hydroxylated fatty acid(s). The specific activity of glomerular fatty acid cyclo-oxygenase (EC 1.14.99.1), based on radioimmunoassay for prostaglandins E2 and F2,, was 10- to 40-fold higher than that of cortical tubular enzyme. These data demonstrate that glomeruli have the capability of synthesizing an array of end-products from arachidonic acid. These prostaglandins may exert important physiologic effects, because renin secretion and arteriolar resistance are regulated by the glomerulus and the afferent and efferent arterioles.
Increasing evidence exists which indicates that renal prostaglandins exert important influence on renal cortical functions such as renin secretion and renal vascular resistance as well as on medullary functions such as urine concentration (1). Initial work on the localization of prostaglandin synthesis within the kidney stressed medullary and papillary sites (2, 3). More recent efforts, using cortical slices or membranes, have clearly demonstrated synthesis of prostaglandin (PG) E2 and PGF2a (4-6). Using [1-14C]PGH2 as substrate, Zenser et al. (7) have reported that rat renal cortical membranes synthesize thromboxane (Tx) B2 and 6-oxo-PGFi, in addition to PGD2, PGE2, and PGF2a. The syntheses of TxB2 and 6-oxo-PGFia were greater in cortex than in the inner medulla (7). Whorton et al. (8) have demonstrated the conversion of 14C-labeled arachidonic acid (C20:4), by rabbit renal cortical microsomes, to 6-oxo-PGFia, PGE2, PGF2a, and PGD2. Medullary tissue produced significant amounts of PGE2, but no 6-oxo-PGFia was detected (8). Needleman and coworkers (9), using perfused rabbit kidneys, have also observed synthesis of PGI2 [6(9)-oxido-11a-15Sdihydroxyprosta-5,13-dienoic acid (prostacyclin)], determined as the appearance of the stable metabolite 6-oxo-PGFi,. The-precise anatomical localization of prostaglandin production within the kidney is not well known. Histochemical and immunofluorescent techniques have detected synthetic enzymes (fatty acid cyclo-oxygenase, EC 1.14.99.1) in collecting tubules and medullary interstitial cells (10, 11). Cortical sites of prostaglandin synthesis have been defined less precisely. Smith and Bell (12), using immunohistochemical microscopy, have variably found fatty acid cyclo-oxygenase in glomeruli as well as in arterioles and cortical collecting tubules. Terragno et al. (13) have reported synthesis of PGI2, measured as 6oxo-PGFia, in arteries and arterioles that were microdissected The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "ad-
and removed from pig kidney. There have been no reports of the capabilities of isolated glomeruli to convert C20:4 to prostaglandin or thromboxane end-products. We hypothesized that the glomerulus might synthesize prostaglandins and thromboxane because of the growing evidence that afferent and efferent arteriolar resistance (1, 14), renin production (15-17), and glomerular filtration (18) may be altered by inhibition of prostaglandin synthesis or by infusion of prostaglandins. We have determined the identity and abundance of the prostaglandins generated from endogenous and exogenous C20:4 by isolated rat glomeruli and nonglomerular cortical tissue. METHODS Isolation of Glomeruli and Incubation with [3HJC20:4. Kidneys were removed from four to nine rats (Sprague-Dawley of either sex, 250-500 g) under ether anesthesia. Cortex was separated from medulla and was finely minced. The glomeruli were separated from other cortical tissue by passing the minced tissue through sieves of different sizes, according to the technique of Misra (19). Glomeruli were obtained from the 105- and 74-am sieves, and tubules were removed from the 210-,m sieve. The purities of the isolated glomerular and nonglomerular fraction were determined microscopically, by counting the number of glomerular and nonglomerular particles suspended in a given volume. The purity of the glomerular fraction varied between 90 and 98%. Virtually none of the isolated glomeruli contained the afferent and efferent arterioles. The nonglomerular fraction, consisting of cortical tubular fragments, was essentially free of contaminating glomeruli. The glomerular and nonglomerular fractions were preincubated for 15 min at room temperature in the presence of 50.9 AM dexamethasone (Sigma), in Earle's balanced salt solution (GIBCO), in order to inhibit deacylation of endogenous phospholipids and maximize conversion of added C20:4 (20, 21). After 15 min, the tissues were precipitated by gentle centrifugation and were divided into two. The first half was resuspended and incubated for 10 min at 37°C in Earle's balanced salt solution containing meclofenamate (26.8 AM) (gift of Warner Lambert/Parke Davis, Ann Arbor, MI) and dexamethasone (50.9 ,M), whereas the second half was incubated with dexamethasone in the absence of meclofenamate. In the experiment shown in Fig. 1, the protein content per tube of glomerular incubation was 1.32 mg. Protein was determined by the method of Lowry et al. (22) with bovine serum albumin as standard. The number of glomeruli per tube was approximately 1.4 X 104, as determined by counting the glomeruli in an aliquot. [3H]C20o4 (1-3 ,Ci, specific activity 80 Ci/mmol; 1 Ci = 3.7 X 1010 becquerels) or nonradioactive C20:4 (final concentration 16.4 ,M) was added to initiate the reaction in a final volume of 0.9 ml. After incubation for 20 min at 37°C, 0.4-ml aliquots were removed and meclofenamate was added Abbreviations: PG, prostaglandin; PGI2, 6(9)-oxido-lla-15S-dihydroxyprosta-5,13-dienoic acid (prostacyclin); Tx, thromboxane; TLC, thin-layer chromatography; RIA, radioimmunoassay; C20:4, arachidonic
vertfsement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.
acid.
1155
1156
Proc. Natl. Acad. Sci. USA 76 (1979)
Biochemistry: Hassid et al.
to those aliquots that had not received it before the incubation (final concentration 26.8 ,M). After 40 min, second and final aliquots were taken and treated similarly to the first samples. Identical incubations were carried out in the absence of glomeruli and tubules to assess nonenzymatic transformation of C204. Nonradioactive and radioactive C20:4 (purchased from Sigma and New England Nuclear, respectively) was purified before use by silica gel column chromatography according to the procedure of Flower et al. (23). Purified C20:4 was stored in hexane under nitrogen at -35°C in the dark. Extraction of Prostaglandins and Thin-Layer Chromatography (TLC). Extraction of acidic lipids, including prostaglandins, was achieved by lowering the pH to 3.0-3.5 with 0.5 M HC1 followed by extraction twice with 3 vol of ethyl acetate. An aliquot of the extracted radioactive materials was spotted on silicic acid thin-layer plates (silica gel 60, E. Merck, Darmstadt, Germany) together with 5-10 ,g of prostaglandin standards, and developed twice in the organic phase of ethyl acetate/iso-octane/acetic acid/water (11:5:2:10, vol/vol). The prostaglandin and thromboxane standards were generously provided by J. Pike (Upjohn). The standards were visualized by exposing the thin-layer plates to iodine vapor. The plates were divided into 26-30 segments, including a segment for each of the standards. The silica gel from each segment was transferred to a scintillation vial, and the radioactivity was determined by scintillation counting in an aqueous counting mixture (Formula-963, New England Nuclear). Counting efficiency was 40-50% and was determined with a radioactive external standard. Prostaglandins were separated by TLC and eluted from the silica gel with acetone. Individual prostaglandins were then treated with 1 M KOH in methanol at room temperature for 1 hr. An identical incubation was carried out in methanol in the absence of KOH. After acidification and extraction with ethyl acetate, treated and untreated materials were spotted on TLC together with standards and developed in the solvent described above. The separated prostaglandins were also reduced with excess sodium borohydride in methanol at room temperature for 2 hr. The products were then determined by TLC and scintillation counting as described above. Radioimmunoassay (RIA). RIA for PGE2 and PGF2, was performed with specific antibodies purchased from Pasteur Institute, Paris, France. Anti-PGF2a does not distinguish between PGFia and PGF2a. In practice, this is not a problem because PGF2a is far more abundant than PGFia. 6-OxoPGFi,, PGE1, and PGE2 cross react with anti-PGF2a 1.6%, 0.3%, and 0.8%, respectively. PGAI, PGA2, PGBI, PGB2, and TxB2 crossreact less than 0.1%. PGE1, PGF2a, 13,14-dihydro-PGE2, and 13,14-dihydro15-keto-PGE2 crossreact with anti-PGE2 2.7%, 0.12%, 0.12%, and 0.10%, respectively. PGFia, PGA1, PGA2, PGB2, TxB2, and 6-oxo-PGFia crossreact less than 0.04% with anti-PGE2. [3H]PGE2 and [3H]PGF2a (100-200 Ci/mmol) were purchased from New England Nuclear. PGE2 and PGF2a were determined by RIA directly without purification. RESULTS Isolated glomeruli converted octatritiated C20:4 to products that were separable on silica gel TLC. In the experiment shown in Fig. 1, in the absence of meclofenamate, the radioactive peak corresponding to PGF2a had 1.2% of the total radioactivity recovered from the thin-layer plate. The peaks corresponding to PGE2, 6-oxo-PGFia, TxB2, and PGD2, had 0.47%, 0.47%, 0.30%, and 0.28%, respectively, of the total recovered radioactivity; 83.1% of the radioactivity was recovered unchanged
as C20:4 (not shown in Fig. 1). In other experiments, the radioactivity corresponding to the PG products ranged from 3 to 9% of recovered radioactivity. TLC of radioactive products in a solvent system of 99:1 (vol/vol) ethyl acetate/acetic acid gave results consistent with those of Fig. 1. Experiments performed in the absence of dexamethasone gave results qualitatively similar to those of Fig. 1. Cortical tubules incubated with octatritiated C20:4 also appeared to synthesize radioactive products that comigrated with prostaglandin standards. However, the yield for all prostaglandin products was much lower than that for glomeruli, even in the face of severalfold more tubular protein (results not shown). Because of the lower yield, tubular prostaglandin production could not be accurately quantitated. It appeared that the specific activity for glomerular prostaglandin synthesis was at least 20-fold higher than that for cortical tubular synthesis. In addition to comigration with authentic prostaglandins, the glomerular radioactive products were identified by the following criteria. (i) The amounts of radioactive materials comigrating with the prostaglandin standards were reduced to 10-40% when glomeruli were incubated in the presence of meclofenamate, a potent fatty acid cyclo-oxygenase inhibitor (Fig. 1). (ii) On treatment with KOH in methanol, the radioactivity corresponding to PGE2 and PGD2 was selectively decreased, relative to a control sample that was treated identically except for the absence of KOH (results not shown). Most of the radioactivity corresponding to PGE2 and PGD2 was lost. KOH treatment converted the radioactive material comigrating with PGE2 to a product comigrating with PGB2, thus confirming the identity of the original material as PGE2 (results not shown). We were unable to show significant increases in a new product on treatment of PGD2 with KOH, in spite of decreased radioactivity comigrating with PGD2 (results not shown). Therefore, the identification of PGD2 remains tentative. Radioactive materials corresponding to 6-oxo-PGFia, PGF2a, and TxB2 were essentially unchanged on treatment with KOH (compared 1400 PG F=
1200 1000
E 800 600
6-oxo-1
400
PGE2
6000
PGD2
TxB2
4000E 2000
200 2
4
6
8
10 12
14 16" 16
l
18
Distance from origin, cm
FIG. 1. TLC of products of octatritiated C20:4 metabolism in glomeruli incubated with or without meclofenamate. Incubation time was 20 min. Solid bars are in the absence of meclofenamate. Hashed bars are in the presence of 28.6 MM meclofenamate. dpm are corrected for nonenzymatic conversion of C20:4 by subtracting dpm obtained from an incubation and TLC performed under identical conditions, but in the absence of glomeruli. Note difference in dpm scale to the right of broken horizontal axis. C20:4 migrated between 17.4 and 18.1 cm (83,556 dpm in the absence and 96,438 dpm in the presence of meclofenamate) and is not shown in this figure. dpm values in the presence of meclofenamate were multiplied by a factor such that total radioactivity recovered equalled that of the experiment lacking meclofenamate, 100,511 dpm.
Hassid et al.
Biochemistry:
6
8
Proc. Natl. Acad. Sci. USA 76 (1979)
12
10
14
6 8 10 12 14 Distance from origin, cm
16
Distance from origin, cm
FIG. 2. TLC of radioactive glomerular 6-oXO-PGFIa before and after treatment with sodium borohydride. Glomerular 6-oxo-PGF1a was purified by TLC as described in text. Equal amounts of 6-oxoPGFia, were treated with 200 ,ud of methanol or with 4 mg of sodium borohydride in 200 ,ul of methanol. After 1 hr the products were recovered by acidification, extraction with ethyl acetate, and rechromatography on silica gel thin-layer. (Upper) Radiochromatogram generated by 6-oxo-PGFia incubated in methanol. (Lower) Radiochromatogram of borohydride-treated 6-oxo-PGFl,,.
to a control incubation in methanol alone), thus confirming the absence of a f3-hydroxy ketone moiety in these materials (results not shown). (iii) Borohydride treatment of the radioactive product corresponding to 6-oxo-PGFia resulted in partial conversion to products that had different chromatographic mobilities than the presumed starting material, 6oxoPGFIa, (Fig. 2). However, the conversion to reduced products was not complete, presumably because 6-oxo-PGFia is predominantly in the cyclic lactol form (24). We were also unable to reduce authentic 6oxo-PGFIa with sodium borohydride. These results confirm those of Pace-Asciak (24), who also found it difficult to reduce 6-oxo-PGFi, with borohydride. Borohydride treatment of TxB2 Control
400C TxB2
300C
E M
200C 1 00C
6-oxo- PGPG FILa F2C
PG E2
C20:4
PG- Tx B2 2< PGE2
C20:4
4000 3000
E
0.
2000 6-oxo-
1000
PGFia
2
4
6
8
10
12
14
16
18
Distance from origin, cm
FIG. 3. TLC of radioactive glomerular TxB2 before and after treatment with sodium borohydride. The protocol was similar to that described in text and Fig. 2. (Upper) Chromatogram of TxB2 treated with methanol. (Lower) Chromatogram of TxB2 treated with boro-
hydride.
1157
FIG. 4. TLC of radioactive glomerular PGD2 before and after treatment with sodium borohydride. The protocol was similar to that described in text and Fig. 2. (Upper) Chromatogram of PGD2 treated with methanol. (Lower) Chromatogram of PGD2 treated with borohydride.
gave a single more polar product (Fig. 3). This is consistent with the structure of TxB2 which, in part, exists as the acyclic aldehyde tautomer (25).t As expected, PGE2 was converted to two more polar products with sodium borohydride, one of which corresponded to PGF2a. The other product was not identified, but may be PGF2, (results not shown). PGD2 was converted by borohydride treatment to a radioactive product identical in chromatographic mobility to PGF2a (Fig. 4). PGF2a, as expected, did not generate significant new products when treated with sodium borohydride (results not shown). (iv) Glomeruli and cortical tubules incubated with nonradioactive C20:4 (17.3 ,uM) synthesized immunoreactive PGE2 and PGF2a (Table 1). Meclofenamate (27 ,uM) inhibited this synthesis by 60-80%. Glomeruli synthesized 40-fold more PGF2a and 10-fold more PGE2, per mg of protein, than tubules (Table 1). The ratio of glomerular PGF2a to PGE2 determined by RIA (Table 1) was similar to that determined by TLC (Fig. 1). Glomeruli incubated in the absence of exogenous C20:4 produced levels of immunoreactive PGF20 and PGE2 lower than those produced in its presence (Table 1). The appearance of immunoreactive PGE2 and PGF2a was inhibited by mepacrine, a cyclo-oxygenase and phospholipase A2 inhibitor. The appearance of the most abundant glomerular C20:4 metabolite or metabolites (migrating between 16.1 cm and 17.0 cm, Fig. 1) was inhibited less than 50% by meclofenamate. This partial inhibition may be due to comigration of a cyclo-oxygenaserelated product (e.g., 12L-hydroxy-5,8,10-heptadecatrienoic acid) with a product unrelated to cyclo-oxygenase (e.g., 12Lhydroxy-5,8,10,14-eicosatetraenoic acid). Alternatively, meclofenamate may inhibit a lipoxygenase product. The chromatographic mobility of this material or materials is consistent with that of hydroxylated fatty acid(s) (23). Cortical tubular enzyme also synthesized a product having similar chromatographic mobility (data not shown). * Note Added in Proof. Using specific antiserum to TxB2 (Pasteur Institute), we have verified the release of TxB2 by normal glomeruli incubated for 20 min with 10 nM C20:4.
1158
Biochemistry:
Hassid et al.
Proc. Nati. Acad. Sci. USA 76 (1979)
Table 1. RIA of PGF2a and PGE2 in incubations of glomerular and nonglomerular fractions in the presence or absence of C20:4 or inhibitors of prostaglandin synthesis Immunoreactive Inhibitor prostaglandin, Incubation of prostaglandin pmol/mg proteint C20:4, time, min AM* synthesis PGF2a PGE2 Incubations with glomerular fraction 20 17.3 None 88.9 46.5 20 17.3 18.3 Meclofenamate (27 MM) 8.79 40 17.3 89.8 59.3 None 40 17.3 Meclofenamate (27 MM) 21.0 12.5 60 0 19.5 None 16.0 60 0 1.80 0.82 Mepacrine (4.9 mM) ND 120 0 29.6 None 120 0 ND Mepacrine (4.9 mM) 2.86 Incubations with nonglomerular fraction 20 17.3 None 20 17.3 Meclofenamate (27 MM) 40 17.3 None 40 17.3 Meclofenamate (27 MM)
1.98 0.82 3.94 1.15
3.71 2.00 6.77 2.35
* C20:4 represents externally added amounts only. Total C20:4 available to fatty acid cyclo-oxygenase, which includes that liberated from endogenous sources by phospholipase A2, was not determined. Incubation conditions are described in the text in detail. t Values expressed for total time of incubation. ND, not determined.
DISCUSSION Our results demonstrated the synthesis of significant quantities of PGF2a, PGE2, 6-oxo-PGFIa, TxB2, and PGD2 by rat glomeruli. PGE2 and PGD2 have a f3-hydroxy ketone structure in the cyclopentane ring, and were therefore susceptible to dehydration by a base-catalyzed reaction (26). 6-oxo-PGFIa, PGF2a, and TxB2 were not altered by base treatment. These results are consistent with the initial identification of radioactive glomerular prostaglandins based on comigration with prostaglandin standards on TLC (Fig. 1). Sodium borohydride treatment resulted in decreased amounts of PGE2, PGD2, 6oxo-PGFia, and TxB2, but not of PGF2a, consistent with reduction of keto groups to alcohol groups. RIA data for PGE2 and PGF2a further strengthened the structural designations based on comigration with standards on TLC. The relative abundance of the prostaglandins was quite similar to that obtained in rat cortical membranes and in rabbit cortical microsomes (7, 8). This suggests the absence of significant conversion of PGE2 to PGF2, in glomeruli by the soluble enzyme prostaglandin-9-ketoreductase (27) and that PGF2a is most likely produced directly from C20:4. The lower specific activity of cyclo-oxygenase in nonglomerular cortical tissues, 1/10th to 1/40th, cannot be due to increased metabolism of prostaglandins by 15-hydroxyprostaglandin dehydrogenase and A-13-prostaglandin reductase, because no significant radioactive product whose formation was inhibited by meclofenamate could be detected in regions of the chromatograms expected to contain prostaglandin metabolites (15-keto prostaglandins or 13,14-dihydro 15-keto prostaglandins). This suggests that a significant and perhaps major portion of renal cortical prostaglandin synthesis occurs in glomeruli, despite larger amounts of nonglomerular tissue. Until recently, it was assumed that prostaglandins synthesized in the renal medulla were transported to the renal cortex via the tubular fluid and thereby controlled some cortical physiologic functions (28). However, our results with isolated glomeruli and the data of others using cortical membranes and microsomes (7, 8) verify the capacity of the renal cortex to synthesize a spectrum of prostaglandins. It seems most likely that important cortical functions, such as renin secretion and control
of renal blood flow, may be influenced by prostaglandins produced locally in glomeruli and~afferent and efferent arterioles rather than by prostaglandins produced in the renal medulla. Our results demonstrate that glomeruli possess much greater prostaglandin synthetic capacity than cortical tubules do. The evidence supporting important renal cortical actions of prostaglandins is good. C20,4 and prostaglandin endoperoxides released renin from rat, rabbit, and dog kidneys studied in vivo or as slices in vitro (29-32). Furthermore, PGE2 and PGI2 stimulated renin secretion in both the filtering and nonfiltering dog kidney in vivo (15, 16, 33), presumably because of a direct effect on the juxtaglomerular cells. Most prostaglandins, and especially PGE2 and PGI2, vasodilate renal arterioles, and TxA2 vasoconstricts these vessels. Vasoconstriction mediated by either a adrenergic agonists or by angiotensin II is dramatically potentiated after inhibition of the fatty acid cyclo-oxygenase (34, 35). This undoubtedly is the result of diminished synthesis of PGI2 and PGE2 by glomeruli and their arterioles. Conversely, TxA2 probably mediates the renal vasoconstriction observed after ureteral obstruction (36, 37). The glomeruli may be the locus of synthesis of TxA2 after ureteral obstruction, because elevated ureteral pressure immediately alters glomerular hydrodynamics. Finally, micropuncture experiments have shown that infusion of prostaglandins (18) or inhibition of fatty acid cyclo-oxygenase (14) alters glomerular ultrafiltration and afferent and efferent arteriolar resistances. These diverse actions of prostaglandins and thromboxanes on renal cortical function in health and disease reinforce the potential importance of our observations that rat glomeruli can synthesize PGF2a,, PGE2, 6-oxo-PGFia, TxB2, and PGD2. We acknowledge the expert technical assistance of Leila Jacobs, Martin Brett, and Dr. Alice S. Petrulis and the secretarial help of Wendy Ingaran and Illein Youngman. This work was supported by National Institutes of Health Grant HL-22563 and the American Heart Association Grant AHA-77916.
1. Dunn, M. J. & Hood, V. L. (1977) Am. J. Physiol. 233, F169F184. 2. Anggard, E.,
Bohman, S. O., Griffin, J. E., Larsson, C. & Maunsbauch, A. B. (1972) Acta Physiol. Scand. 84, 231-246.
Biochemistry: Hassid et al. 3. Crowshaw, K. & McGiff, J. C. (1973) in Mechanisms of Hypertension, ed. Sambhi, M. P. (Elsevier, New York), pp. 254-273. 4. Larsson, C. & Anggard, E. (1973) Eur. J. Pharmacol. 21, 3036. 5. Larsson, C. & Anggard, E. (1976) J. Pharm. Pharmacol. 28, 326-328. 6. Pong, S. S. & Levine, L. (1976) Res. Commun. Chem. Pathol. Pharmacol. 13, 115-123. 7. Zenser, T. V., Herman, C. A., Gorman, R. R. & Davis, B. B. (1977) Biochem. Biophys. Res. Commun. 79,357-363. 8. Whorton, A. R., Smigel, M., Oates & Frolich, J. C. (1978) Biochim. Biophys. Acta 529, 176-180. 9. Needleman, P., Bronson, S. D., Wyche, A., Sivakoff, M. & Nicolaou, K. C. (1978) J. Clin. Invest. 61, 839-849. 10. Janszen, F. H. & Nugteren, D. H. (1971) Histochemistry 27, 159-164. 11. Smith, W. L. & Wilkin, G. P. (1978) Am. J. Physiol. 235, F451-F457. 12. Smith, W. L. & Bell, T. G. (1978) Prostaglandins 15, 715 (abstr.). 13. Terragno, N. A., McGiff, J. C. & Terragno, A. (1978) Clin. Res. 26, 545 (abstr.). 14. Bayliss, C. & Brenner, B. M. (1977) Kidney Int. 12, 550
(abstr.). 15. Data, J. L., Gerber, J. G., Crump, W. J., Frolich, J. C., Hollifield, J. W. & Nies, A. S. (1978) Circ. Res. 42, 454-458. 16. Bolger, P. M., Eisner, G. M., Ramwell, P. W. & Corey, E. J. (1978) Nature (London) 271, 467-469. 17. Yun, J., Kelly, G., Bartter, F. & Smith, H. (1977) Circ. Res. 40, 459-464. 18. Bayliss, C., Deen, W. M., Myers, B. D. & Brenner, B. M. (1976) Am. J. Physiol. 230, 1148-1158. 19. Misra, R. P. (1972) Am. J. Clin. Pathol. 58, 135-139. 20. Flower, R. J. (1978) in Advances in Prostaglandins and
Proc. Natl. Acad. Sci. USA 76 (1979)
1159
Thromboxane Research, eds. Galli, C., Galli, G. & Porcellati, G. (Raven, New York), Vol. 3, pp. 105-112. 21. Hong, S. & Levine, L. (1976) Proc. Natl. Acad. Sci. USA 73, 1730-1734. 22. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem. 193,265-275. 23. Flower, R. J., Cheung, H. S. & Cushman, D. W. (1973) Prostaglandins 4,325-341. 24. Pace-Asciak, C. (1976) Experientia 32, 291-292. 25. Hamberg, M. & Samuelsson, B. (1974) Proc. Natl. Acad. Sci. USA 71,3400-3404. 26. Bergstrom, S., Rhyage, R., Samuelsson, B. & Sj6vall, J. (1963) J. Biol. Chem. 238,3555-3564. 27. Hassid, A. & Levine, L. (1977) Prostaglandins 13,503-516. 28. Frolich, J. C., Wilson, T. W., Sweetman, B. J., Smigel, M., Nies, A. S., Carr, K., Watson, J. T. & Oates, J. A. (1975) J. Clin. Invest. 55,763-770. 29. Weber, P., Holzgreve, H., Stephan, R. & Herbst, R. (1975) Eur. J. Pharmacol. 34,299-304. 30. Larsson, C., Weber, P. & Anggard, E. (1974) Eur. J. Pharmacol. 28,391-394. 31. Bolger, P. M., Eisner, G. M., Ramwell, P. W. & Slotkoff, L. M. (1976) Nature (London) 259, 244-245. 32. Bolger, P. M., Eisner, G. M. & Slotkoff, L. M. (1978) Fed. Proc. Fed. Am. Soc. Exp. Biol. 37,306 (abstr.). 33. Aiken, J. W. & Vane, J. R. (1975) J. Pharmacol. Exp. Ther. 184, 678-687. 34. Satoh, S. & Zimmerman, B. G. (1975) Circ. Res. Suppl. 1, 8996. 35. Henrich, W. L., Berl, T., McDonald, K. M., Anderson, R. J. & Schrier, R. W. (1978) Am. J. Physiol. 235, F46-F51. 36. Morrison, A. R., Nishikawa, K. & Needleman, P. (1977) Nature (London) 267, 259-260. 37. Morrison, A. R., Nishikawa, K. & Needleman, P. (1978) J. Pharmacol. Exp. Ther. 205, 1-18.
