Metabolism of arachidonic acid by rat adrenal glomerulosa cells: synthesis of hydroxyeicosatetraenoic acids and epoxyeicosatrienoic acids.

0013-7227/91/1284-2183$03.00/0 Endocrinology Copyright© 1991 by The Endocrine Society

Vol. 128, No. 4 Printed in U.S.A.

Metabolism of Arachidonic Acid by Rat Adrenal Glomerulosa Cells: Synthesis of Hydroxyeicosatetraenoic Acids and Epoxyeicosatrienoic Acids* WILLIAM B. CAMPBELL, MILTON T. BRADY, LORI J. ROSOLOWSKY AND J. R. FALCK The University of Texas Southwestern Medical Center at Dallas, Departments of Pharmacology and Molecular Genetics, Dallas, Texas 75235-9041

cyclooxygenase inhibitor, indomethacin, the lipoxygenase inhibitors, nordihydroguaiaretic acid, baicalein and AA861, and the combined cyclooxygenase/lipoxygenase inhibitors, BW755C and eicosatetrayenoic acid, inhibited the formation of the [uC]PGs, the [14C]HETEs, and the [14C]EETs. Metyrapone and clotrimazole, inhibitors of cytochrome P45o, increased the synthesis of [14C]PGs and [14C]HETEs and reduced the synthesis of [14C] EETs. Superoxide dismutase did not alter arachidonic acid metabolism. In contrast, arachidonic acid metabolism was increased in cells pretreated with catalase. These data indicate that adrenal glomerulosa cells metabolize exogenous arachidonic acid to a number of oxygenated metabolites including PGs, HETEs, and EETs. From studies with inhibitors, the EETs appear to be synthesized by a cytochrome P450 epoxygenase and the HETEs by lipoxygenases. (Endocrinology 128: 2183-2194, 1991)

ABSTRACT. Metabolites of arachidonic acid have been implicated in the regulation of aldosterone release. To form a basis for further investigations in this area, the present study has isolated and identified the metabolites formed from exogenous arachidonic acid by adrenal zona glomerulosa cells and characterized the effects of several inhibitors on the synthesis of these eicosanoids. Rat adrenal glomerulosa cells metabolized exogenous [uC]arachidonic acid to products comigrating with the prostaglandins (PGs), hydroxyeicosatetraenoic acids (HETEs) and epoxyeicosatrienoic acids (EETs). The metabolites were found in the cells and the incubation media; however, none of the metabolites were found esterified to cellular lipids. The major metabolites were identified as 6-keto PGFlo, PGE2, PGF2n, PGD2,12(S)-HETE, 15(S)-HETE, 14,15-EET, 11,12-EET, 8,9EET, and 5,6-EET. The identities of the HETEs and EETs were confirmed by gas chromatography/mass spectrometry. There was no evidence for the synthesis of leukotrienes. The

T

HE METABOLISM of arachidonic acid has been extensively studied in a number of tissues and cells. These investigations have revealed that metabolism occurs by three pathways: 1) the cyclooxygenase pathway giving prostaglandins (PGs) and thromboxane, 2) the lipoxygenase pathway giving hydroperoxyeicosatetraenoic acids (HPETEs), hydroxyeicosatetraenoic acids (HETEs), and leukotrienes (LTs), and 3) the cytochrome P45o pathway giving epoxyeicosatrienoic acids (EETs), dihydroxyeicosatrienoic acids, and HETEs (1-7). Many studies indicate that arachidonic acid metabolites are involved in the regulation of aldosterone secretion (8Received November 7,1990. Address requests for reprints to: Dr. William B. Campbell, Department of Pharmacology, The University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Boulevard, Dallas, Texas 752359041. * These studies were supported by grants from the National Heart, Lung and Blood Institute (HL-21066) and National Institute of Medical Sciences (GM-31278). The Finnigan mass spectrometer was purchased with funds provided by the National Institutes of Health (Grants GM27506 and GM-16488-S1). This work was performed while Lori Rosolowsky was a National Institutes of Health Pre-Doctoral Fellow (5T32-GM-07062).

17). Indomethacin, an inhibitor of cyclooxygenase, was found by several laboratories to inhibit basal and angiotensin II (All)-stimulated aldosterone release, in vivo and in vitro, in concentrations that reduced adrenal PG synthesis (8, 9,11,12,14). These findings suggested that PGs may mediate or modulate All-induced aldosterone release. In contrast, other investigators failed to observe any inhibitory effect with indomethacin or meclofenamate (10, 13, 15, 16). Using radiochemical methods and RIA, rat adrenal glomerulosa cells were found to synthesize PGI2) PGE2, and PGF2« as the major metabolites of the cyclooxygenase pathway (13, 14). The release of aldosterone was stimulated by exogenous PGI 2 and PGE2, but not by PGF2« (9-12, 14). When added in concentrations that promoted steroidogenesis, All stimulated the release of PGI2, but not PGE 2 or PGF 2a , from adrenal cells (14). However, the amounts of PGI2 that were released were much less than the concentrations that promoted aldosterone release. ACTH and potassium failed to stimulate the release of PGs from these cells (11, 13). It may be concluded from these investigations

2183

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 22:53 For personal use only. No other uses without permission. . All rights reserved.

2184

ARACHIDONATE METABOLISM IN ADRENAL CELLS

that cyclooxygenase metabolites of arachidonic acid were not involved in the regulation of aldosterone release. Other pathways of arachidonic acid metabolism have been implicated in the regulation of aldosterone secretion. Inhibition of phospholipase A2 with mepacrine or bromophenacyl bromide or lipoxygenases with eicosatetrayenoic acid (ETYA) or BW755C blocked All-induced aldosterone release in bovine adrenal cells (15). Similarly, in rat and human adrenal cells, lipoxygenase inhibitors reduced aldosterone release when stimulated by All but not by ACTH or potassium (16, 17). Furthermore, exogenous 12-HETE or 12-HPETE stimulated aldosterone release and restored the response to All in cells treated with lipoxygenase inhibitors. 15-HETE and 15HPETE were without effect. Adrenal cells were found to contain metabolites with the chromatographic properties of 12-HETE and 15-HETE (16). These results suggested that 12-HETE was involved in All-stimulated aldosterone release. A prerequisite to the further understanding of the role of eicosanoids in the regulation of aldosterone biosynthesis is a knowledge of arachidonic acid metabolism by adrenal glomerulosa cells. The present studies were designed to isolate and identify the metabolites that are synthesized from exogenous arachidonic acid by the adrenal zona glomerulosa and investigate the synthetic pathways involved.

Endo • 1991 Vol 128 • No 4

Suspensions of adrenal glomerulosa cells were prepared from adrenal capsules as described previously (14). Rats were killed by decapitation. Their adrenals were removed and defatted, and the capsules were removed. The capsules were placed in potassium-free medium 199 containing 1 mg/ml collagenase, 0.05 mg/ml deoxyribonuclease, 0.05 mg/ml ribonuclease, and 2 mg/ml BSA and minced. The tissue was then incubated for 60 min in a Dubnoff metabolic shaker under 95% O2 and 5% CO2 at 37 C at 60 rpm. The cells were dispersed by repeated pipetting (30 times) every 20 min. The cell suspension was washed twice with medium 199 containing 2 mg/ml BSA. The cells were resuspended in medium 199 containing 4.5 meq/liter potassium and 2 mg/ml BSA. The cells were counted in a hemacytometer and adjusted to 500,000 to 700,000 cells/ml. The number of viable cells was determined by either exclusion staining with trypan blue or differential fluorescent staining with fluorescein diacetate and ethidium bromide, and the cell purity was based on cell morphology. Viability averaged 94%, and purity was

greater than 95%. Metabolism of ^C] arachidonic acid by adrenal cells

Medium 199 (potassium free) was purchased from GIBCO (Grand Island, NY) and collagenase from Worthington (Freehold, NJ). Indomethacin, metyrapone, clotrimazole, deoxyribonuclease, ribonuclease, superoxide dismutase, catalase, nordihydroguaiaretic acid (NDGA), ACTH, and arachidonic acid were purchased from Sigma (St. Louis, MO). BSA (fatty acid free) was obtained from Calbiochem (La Jolla, CA). The HETEs, HPETEs, and baicalein were purchased from Biomol (Philadelphia, PA). The EETs were synthesized as previously described (18). PGs were provided by the Upjohn Co. (Kalamazoo, MI), LTs by Merck-Frosst (Montreal, Canada), ETYA by Hoffman-LaRoche (Nutley, NJ), and BW755C by Burroughs-Wellcome (Beckenham, England). All solvents were HPLC grade and purchased from Burdick and Jackson (Muskegan, MI). Sprague-Dawley rats were obtained from Simonsen Laboratories (Simonsen, CA). [14C(U)]Arachidonic acid (1 Ci/mmol) and [35S]cysteine (650 Ci/mmol) were obtained from New England Nuclear (Boston, MA). Octadecasilyl silica (ODS) extraction columns were obtained from Analytichem (Harbor City, CA).

Suspensions of adrenal cells (10 ml) were incubated with 1 ixM [14C]arachidonic acid (1 /LtCi) for 30 min at 37 C in an atmosphere of 95% O2/5% CO2. In an identical manner, [I4C] arachidonic acid was incubated in the absence of cells to assess the extent of nonenzymatic oxidation of the fatty acid. In some experiments, cells were pretreated with cyclooxygenase, lipoxygenase, and/or cytochrome P450 inhibitors or superoxide dismutase or catalase for 15 min before [14C]arachidonic acid was added. After the incubation, the media and cells were sonicated, extracted with ODS extraction columns and analyzed by HPLC. These experiments were repeated on 3 or 4 separate incubations. In other studies, the media and cells were separated by centrifugation. The media were then extracted and analyzed by HPLC. The cell pellet was suspended in 3 ml distilled water, sonicated, extracted, and analyzed in a similar manner. To analyze the radioactive products in the lipids of the cells, cell pellet was extracted by the method of Bligh and Dyer (19). This extract was treated with 0.1 N NaOH and methanol (1:1) overnight to hydrolyze the lipids. The lipids were then extracted and analyzed by HPLC as described above. In some studies of LT synthesis, cells were incubated in a similar fashion with [35S]cysteine (1 /xCi) instead of [14C]arachidonic acid. Additional studies were designed to identify the metabolites by UV absorbance spectroscopy and gas chromatography-mass spectrometry (GC-MS). In these experiments, the experimental design was modified to obtain an adequate mass of the metabolites for these analyses. Cells (100 ml) from 50 rats were incubated with 10 ^M [14C]arachidonic acid (3 /iCi) for 30 min at 37 C. The media and cells from 4 of these incubations were pooled, sonicated, extracted, and purified by HPLC. These analyses were repeated on 4 separate sets of large incubations, each consisting of a different adrenal cell preparation.

Preparation of isolated rat adrenal glomerulosa cells

Extraction of uC-metabolites of arachidonic acid

Male Sprague-Dawley rats (200-300 g) were maintained on a standard Wayne Rat Chow Diet containing 142 meq/kg sodium and 290 meq/kg potassium and tap water ad libitum.

The incubation media or media and cells were extracted with ODS extraction columns (6 ml)(20, 21). For extracting eicosanoids, the ODS columns were washed with 5 ml each of ethanol

Materials and Methods Materials


ARACHIDONATE METABOLISM IN ADRENAL CELLS and water. Immediately before addition to the column, the sample (10 ml) was acidified to pH 3 with glacial acetic acid, and ethanol was added to a final amount of 15%. The column was washed sequentially with 5 ml aliquots of 15% ethanol, water, and petroleum ether. The arachidonic acid metabolites were eluted with 2 ml ethyl acetate. For the large incubations, 20 ml extraction columns were used and the volume of solvents and sample was increased 10-fold. The recoveries from the columns were between 92-98% for PGs, HETEs, and EETs. LTs were also extracted with ODS extraction columns that had been prewashed sequentially with 5 ml methanol, water, and 0.1% EDTA in water (22). The sample was added to the column, washed with water, and the LTs eluted with 10 ml methanol. The recovery of LTs was 85-90%. Separation of P^CJ arachidonic acid metabolites by HPLC

Five HPLC methods were used to isolate the metabolites from adrenal cells. For all of the separations, a Beckman microprocessor-controlled gradient chromatograph was used. Method A was a reverse-phase separation used to resolve the metabolites of arachidonic acid into groups (5, 21). The separation used a Nucleosil (Phenomenex, Rancho Palos Verdes, CA) or LiChrosorb (Unimetrics, Anaheim, CA) C-18 column (5 fim, 4.6 X 250 mm). Solvent A was water, and solvent B was acetonitrile containing 0.1% glacial acetic acid. The flow rate was 1 ml/min. The metabolites were eluted over 40 min with a linear gradient from 50% B in A to 100% B. Absorbance was monitored at 235 nm using a Beckman model 155 UV detector. The column eluate was collected in 0.2 ml fractions by a fraction collector. Liquid scintillation fluid was added to the fractions, and the radioactivity was measured with a Beckman liquid scintillation spectrometer. For large scale purifications, aliquots of the fractions (50 fi\) were analyzed for radioactivity, and the fractions containing the radioactive metabolites were pooled for further analysis. Method B was a reverse phase separation used to separate the PGs (14, 21). PG standards (5 fig of each) were added to the sample before HPLC analysis. A LiChrosorb or Nucleosil C-18 column was used at a flow rate of 1 ml/min. Solvent A was 0.025 M phosphoric acid in water, and solvent B was acetonitrile. The elution was accomplished over 40 min with 31% B in A. Absorbance was measured at 192 nm to determine the elution times of added PG standards. The eluate was collected in 0.5 ml fractions and analyzed as described above to detect the products synthesized by the cells. The identities of the PGs were determined by comigration of radioactivity with a known PG standard after simultaneous injection. Method C, a normal phase separation for resolving the HETEs, used an Ultrasphere (Beckman-Altex, Palo Alto, CA) silica gel column (5 nm, 4.6 X 250 mm)(6, 21). Solvent A was hexane containing 0.1% glacial acetic acid, and solvent B was hexane containing 2% isopropanol and 0.1% glacial acetic acid. The flow rate was 3 ml/min. The solvent program consisted of a 40 min linear gradient from 25% B in A to 75% B in A. Absorbance was monitored at 235 and 280 nm with a HewlettPackard diode array detector. The column eluate was monitored for absorbance between 200 and 350 nm every 6 sec, and the data were stored on a Hewlett-Packard HP-85 computer. The eluate was collected in 0.6 ml fractions, and 0.1 ml aliquots

2185

were processed as described above for radioactivity. The fractions containing radioactive products were pooled for mass spectrometric analysis. Method D was a normal phase system used to separate the EETs (21). The column was an Ultrasphere silica gel column. The solvent was hexane containing 0.4% isopropanol and 0.1% glacial acetic acid. The flow rate was 2 ml/min. The elution was carried out over 25 min. Absorbance of the column eluate was monitored at 235 and 280 nm as described for method C. The eluate was collected in 0.4 ml fractions, and 0.1 ml aliquots were analyzed for radioactivity. Radioactive metabolites were pooled for further analysis by GC-MS. Method E, a reverse phase system for resolving LTs, used a Nucleosil C-18 column (5 fim, 4.6 X 250 mm) and a flow rate of 1 ml/min (22). The elution solvent was methanol-waterglacial acetic acid (70:30:0.03) with the pH adjusted to 4.5 with ammonium hydroxide. Metabolites were eluted over 40 min. Absorbance was monitored at 280 nm. The column eluate was collected in 0.2 ml fractions and analyzed for radioactivity as described above. Separation of HETE stereoisomers by chiral HPLC Adrenal cells were incubated with [14C]arachidonic acid as described above. The cells and incubation media were extracted, and the major HETE metabolites were isolated by sequential reverse phase (method A) and normal phase (method C) HPLC. Racemic unlabeled 12-HETE (5 fig) was added to the 12-HETE fraction and racemic unlabeled 15-HETE to the 15-HETE fraction. The HETEs were converted to their methyl esters by treatment with 0.5 ml of a fresh solution of ethereal diazomethane and 50 fd methanol. After incubation at room temperature for 30 min, the solvent was removed under N2. Method F was used to resolve 12- and 15-HETE into their R- and Sisomers by chiral-phase HPLC. The column was a Chiralcel OC column (4.6 x 250 mm, J.T. Baker, Phillipsburg, NJ). The elution solvent was hexane containing 1% isopropanol and 0.1% glacial acetic acid, and the flow rate was 1 ml/min. The 12(R)HETE eluted after 26.2 min and the 12(S)-HETE after 30.5 min. The 15(R)-HETE eluted after 31.4 min and the 15(S)HETE after 35.1 min. The column eluate was collected in 0.5 ml aliquots and measured for radioactivity. It was also monitored at 235 nm to detect the elution of the isomers of the HETE standard. The radioactivity eluting with the S- and Risomers was used to calculate the per cent S-isomer and per cent R-isomer that compose the HETE fractions. GC-MS The pooled radioactive material from the HETE analysis was derivatized as previously described (20). To prepare the methyl ester derivative, the dried sample was treated with 0.5 ml of a fresh solution of ethereal diazomethane and 50 fd methanol for 30 min at room temperature. The ether was evaporated under a stream of N2. The sample was transferred to a dry 0.3 ml Reactivial in 50 fil acetonitrile and silylated using 0.1 ml bis(trimethylsilyl)trifluoroacetamide for 12 h at room temperature. The derivitized samples and standards were analyzed by electron impact and positive ion chemical ionization MS using a Finnigan 4500 quadrupole mass spectrometer interfaced with a Finnigan gas chromatograph. The ionization


2186


potential was 70 eV. For the HETEs, GC was performed using a 3% SP-2100 glass 6 foot packed column. The oven temperature was 210 C, and the injector temperature was 250 C. The carrier gas was N2 with a flow rate of 20-30 ml/min. For chemical ionization MS, the ionization gas was methane. The data were stored on a Finnigan INCOS 2000 data system. Pooled samples of radioactive products from the EET analyses and standards were converted to their pentafluorobenzyl esters and analyzed by negative ion chemical ionization MS. The sample was transferred to a 0.3 ml Reactivial, and the solvent was removed under argon. The sample or standard was then dissolved in 50 n\ acetonitrile and treated with 5 /A pentafluorobenzyl bromide and 10 MI iV,iV-diisopropylethylamine (23). This mixture was incubated under argon at room temperature for 30 min. The solvent was removed under argon, redissolved in methylene chloride, filtered through Unisil silica gel, and washed from the silica gel with methylene chloride. The methylene chloride was removed under argon. Chromatographic separation of the pentafluorobenzyl esters of the EETs was performed on a DB-5 capillary column (0.25 /xm X 15 m, J and W Scientific, Folsom, CA) using N2 as the carrier gas. The injector temperature was 220 C, and the column temperature was 190 C for 2 min and then increased from 190 C to 340 C at a rate of 5 C /min. Negative ion chemical ionization MS was performed with methane as the ionization gas and an ionization potential of 70 eV.

EndoM991 Voll28-No4

2500

2000 1500 1000 500

o 2500 5 2000 Z

MEDIA

a: 1500 Q.

w 1000 z

§ 50°

Vw

1500r

Results Isolation and identification of adrenal ^C] arachidonic acid metabolites When isolated adrenal glomerulosa cells are incubated with [14C]arachidonic acid, a number of radioactive products are formed that are more polar than arachidonic acid. Figure 1A illustrates the profile of 14C-metabolites that are extracted from the cells and media (HPLC method A). The first radioactive peak eluting from the column was a broad peak (fractions 18-40) that comigrated with the PGs. Other 14C-labeled metabolites eluted as a group of peaks comigrating with the HETEs (fractions 100-130) and the EETs (fractions 130-160). Fractions 160-180 and fractions 180-200 also contained major radioactive peaks. The latter peak (fraction 180200) was found to comigrate with [14C] arachidonic acid. When [14C]arachidonic acid was incubated with media under identical conditions in the absence of cells, only two radioactive peaks were observed (data not shown). One peak eluted in fractions 160-180, and the other eluted with arachidonic acid in fractions 180-200. Thus, the peak eluting in fraction 160-180 appears to be an autooxidation product of arachidonic acid rather than a metabolite of the cells. Of the extracted radioactivity, 4.8% was found in the PG fraction, 2.6% in the HETE fraction, and 1.9% in the EET fraction. The media from the cells contained a similar profile of metabolites as the extract from the cells and media (Fig. IB). In contrast,

1000 CELLS

500

°C)

50

100

150

200

FRACTION

FIG. 1. Separation of the [14C]arachidonic acid metabolites of rat adrenal glomerulosa cells. Cells were incubated with [14C]arachidonic acid. The metabolites were extracted from the cells and media (top), media alone (middle), or cell lipids (bottom). The metabolites were resolved by reverse-phase HPLC (method A). Migration times of known standards are indicated with arrows and labels.

when the cell lipids were extracted and hydrolyzed the extract contained a single radioactive peak which comigrated with arachidonic acid. Additional experiments were performed to isolate and identify these adrenal metabolites of arachidonic acid. Media and cells from 200 rats were incubated with arachidonic acid, pooled, and extracted. The extract was purified by HPLC using method A. The radioactive peaks comigrating with the PGs (fractions 18-40), HETEs (fractions 100-130), EETs fraction A (fractions 130140), EETs fraction B (fractions 140-160), and arachidonic acid (fractions 180-200) were collected, extracted, and analyzed individually. The extract of the PG fraction was resolved into its components using HPLC method B. Four major radioactive peaks were isolated (Fig. 2). These radioactive peaks were identified as 6-keto PGF ln , PGF2ft, PGE2, and PGD2 based on comigration with



700 r

T

It

2187

t

7

2 3

700 350-

600

JU

£ 500 2

4 | 7,

50

100 FRACTION

150

10

20 TIME(min)

30

2 400

200

QC

in

0.2 r

2

O 200

0,1

O

100 0 0

10

20 30

40 50

TIME (minutes) FIG. 2. Separation of the PGs synthesized by rat adrenal glomerulosa cells. The PG fraction (fractions 18-40) was collected from the reversephase separation illustrated in Fig. 1. This fraction was rechromatographed with another reverse-phase system (method B) to resolve the PGs into their components. Unlabeled PGs were added to the extract. The elution of metabolites was followed with radioactivity and standard PGs by UV absorbance at 192 nm. Identity of the PGs was established by comigration with known standards after coinjection.

known standards after coinjection. These results are in agreement with our previous findings (14). The HETE fraction (fractions 100-130) was further purified by normal phase HPLC using method C. Figure 3 illustrates the radioactivity in aliquots of the column eluate as well as the UV absorbance of the eluate. Two major radioactive peaks were observed, peaks 2 and 3. Peak 2 comigrated with 12-HETE and peak 3 with 15HETE. When UV absorbance was continuously monitored at 235 nm, absorbance was observed that coincided with these radioactive peaks, indicating the presence of a conjugated diene structure. Five other minor radioactive peaks were also observed in the chromatogram that also had corresponding UV absorbance. Peaks 4, 5, 6, and 7 comigrated with standards for 11-HETE, 9-HETE, 8-HETE, and 5-HETE, respectively. Fractions containing the individual radioactive metabolites were collected and pooled. Each of these metabolites had a UV absorbance maximum at 235 nm (Table 1). The metabolites were then derivatized to their methyl ester and trimethylsilyl ethers and analyzed by GC/MS. Positive chemical ionization MS of the metabolite in peak 1 indicated the presence of contaminating material; therefore a clear mass spectrum was not possible. The metabolites from peaks 2-7 gave similar mass spectra with major ions at 435 (M + CH3CH2), 407 (M + H), 391 (M - CH3) and

FIG. 3. Separation of the HETEs synthesized by rat adrenal glomerulosa cells. The HETE fraction (fractions 100-130) was collected from the reverse-phase separation illustrated in Fig. 1. This fraction was rechromatographed on normal phase HPLC to resolve the HETEs (method C). The column eluate was monitored for radioactivity (top) and UV absorbance at 235 nm (bottom). Elution times of known HETE standards are indicated above the chromatogram.

317 [M - (CH3)3Si0H] (Table 1). These data are consistent with a methyl ester and trimethylsilyl ether with a mol wt of 406. Authentic derivatized HETEs had identical retention times by GC and gave similar mass spectra. Several of these metabolites were also analyzed by electron impact MS. The mass spectrum of the metabolite from peak 2 which comigrated with 12-HETE is shown in Fig. 4 (top). The following ions were found: 406 (M); 391 (M - CH3); 375 (M - 0CH 3 ); 316 [M (CH3)3Si0H]; and 295 [M - ( C H 2 - C H = C H - ( C H 2 ) 4 CH3]. A similar fragmentation pattern was observed with the methyl ester-trimethylsilyl ether of 12-HETE. These data are consistent with peak 2 being 12-HETE. The metabolite in peak 3, which comigrated with 15-HETE, gave diagnostic ions at 406 (M); 391 (M - CH3); 375 (M - OCH3); 335 [M - (CH 2 ) 4 -CH 3 )]; 316 [M (CH3)3Si0H]; and 225 [M - ( C H 2 - C H = C H - C H 2 CH=CH-(CH 2 ) 3 -COOCH 3 )] (Fig. 5, top). A similar fragmentation pattern was obtained with a derivatized 15-HETE standard indicating that peak 3 is 15-HETE. Peak 4 comigrated with 11-HETE, and the mass spectrum of that metabolite is shown in Fig. 4 (bottom). Characteristic ions were found at 406 (M); 391 (M CH3); 375 (M - 0CH 3 ); 316 [M - (CH3)3Si0H]; and 225 [M - ( C H 2 - C H = C H - C H 2 - C H = C H - ( C H 2 ) 3 COOCH3]. The same fragments were observed with authentic derivatized 11-HETE indicating that peak 4 is 11-HETE. Peak 7 comigrated with 5-HETE. The mass


2188

Endo • 1991 Voll28«No4


TABLE 1. Positive chemical ionization mass spectra and UV absorbance maxima of HETEs isolated from adrenal glomerulosa cells Peak

Structure

2

12-HETE

UV absorbance maximum (nm) 235

3

15-HETE

235

4

11-HETE

235

5

9-HETE

235

6

8-HETE

235

7

5-HETE

235

Major ions (mass/charge) 435 (M + 317 (M 435 (M + 335 (M 435 (M + 317 (M 435 (M + 317 (M 435 (M + 317 (M 435 (M + 317 (M -

29); 407 (M + 90); 295 (M 29); 407 (M + 71); 317 (M 29); 407 (M + 90); 285 (M 29); 407 (M + 90); 285 (M 29); 407 (M + 90); 285 (M 29); 407 (M + 90); 285 (M -

1); 391 (M - 15); 111); 285 (M - 121) 1); 391 (M - 15); 90); 285 (M - 121) 1); 391 (M - 15); 121); 225 (M - 181) 1); 391 (M - 15); 121); 255 (M - 151) 1); 391 (M - 15); 121); 265 (M - 141) 1); 391 (M - 15); 121); 203 (M - 203)

Radioactive metabolites from Fig. 3 were collected, derivatized, and analyzed by positive ion chemical ionization mass spectrometry.

spectrum (Fig. 5, bottom) contained major ions of 406 (M); 391 (M - CH3); 375 (M - OCH3); 316 [M (CH3)3Si0H); 305 [M - ((CH3)3-COOCH3)]; 255 [M (CH2-CH=CH-CH2-CH=CH-(CH2)4-CH3)]; and 203 [M - (CH=CH—CH=CH—CH 2 —CH=CH— CH 2 -CH=CH-(CH 2 ) 4 -CH 3 )1. The same ions were found with the derivatized standard for 5-HETE. Thus, peak 7 is 5-HETE. The R- and S-isomers of 12- and 15-HETE could be completely resolved by chiral-phase HPLC, as indicated in Materials and Methods. When adrenal [14C] 12-HETE was analyzed in this manner, a single radioactive peak was observed that eluted with 12-S-HETE at 30.5 min. Similarly, adrenal [14C] 15-HETE eluted as a single radioactive peak which comigrated with 15-S-HETE (35.1 min). These data indicate that the 12- and 15-HETE formed by the adrenal are the S-configuration. The EET fraction (fractions 130-160) eluted as two peaks, EET fraction A (fractions 130-140) and EET fraction B (fractions 140-160). These were collected separately and further analyzed by normal phase HPLC using method D. Figure 6A illustrates the further purification of EET fraction A. One major radioactive peak was obtained which eluted in fractions 50-58 and comigrated with authentic 14,15-EET. EET fraction B yielded four radioactive peaks (Fig. 6B). Peaks 2, 3, and 4 comigrated with standards for 11,12-EET, 8,9-EET, and 5,6EET, respectively. The fractions containing these radioactive peaks were pooled individually. These metabolites had no UV absorbance at wavelengths greater than 210 nm. They were then derivatized to their pentafluorobenzyl esters and analyzed by negative ion chemical ionization MS. The mass spectra for peaks 1-4 are presented in Fig. 7. The mass spectrum of each product had a major ion at 319 (M - H) indicating a mol wt of 320 for the parent molecule. When authentic pentafluorobenzyl esters of the EETs were analyzed, similar mass spectra

were obtained, and the standards had the same retention times by GC as the derivatized unknowns. Based on comigration on reverse-phase HPLC, normal-phase HPLC, capillary GC, and a typical mass spectrum indicating a mol wt of 320, peaks 1-4 appear to be 14,15-, 11,12-, 8,9-, and 5,6-EET, respectively. In a separate set of experiments, cells were incubated with [14C]arachidonic acid or [35S]cysteine, and the cells and media were extracted using a method previously shown to extract LTs (22). The extract was then analyzed by reverse phase HPLC using method E. Figure 8 indicates that there were no radioactive peaks from [14C1 arachidonic acid or [35S]cysteine, respectively, that comigrated with authentic standards for LTB4, LTC4, LTD4, LTE4, or 14,15-LTD4. These data indicate that while adrenal cells synthesize 5-HETE and 15-HETE these products are not further metabolized to LTs of the 5- or 15-series. Effect of inhibitors on the metabolism of /14C7arachidonic acid

Cells were treated with drugs reported to inhibit cyclooxygenase, lipoxygenase and/or cytochrome P450 before the addition of [14C]arachidonic acid. The cells and media were extracted, and the extract was analyzed by HPLC using method A. As in Fig. 1, Fig. 9A indicates that adrenal cells produce radioactive metabolites that comigrate with the PGs, HETEs, and EETs. Pretreatment of the cells with indomethacin, a cyclooxygenase inhibitor, decreased the synthesis of [14C]PGs, [14C] HETEs, and [14C]EETs (Fig. 9). A similar nonselective inhibition of arachidonic acid metabolism occurred with the lipoxygenase inhibitors, baicalein, NDGA (Fig. 9), and AA861 (data not shown), and the combined cyclooxygenase/lipoxygenase inhibitors, BW755c (Fig. 9) and ETYA (data not shown). In contrast, the cytochrome P450 inhibitors, clotrimazole (Fig. 9) and metyrapone


2189

ARACHIDONATE METABOLISM IN ADRENAL CELLS 100

I^HETE

7 5

,

=

s

/ = v \ / COOCH3

A—A OTMS

-10x

50-

0)

o c a •o c

352 335 |

< 50 a> 100 n

100

406 3 9 1

JvUUU 150 200 250 300 350 400 450

500

5-HETE

}g a>

[—10x OTMS, COOCH,

50-

305

100 150

200

250 300 350 400 450

500

M/E

50

Jl

316

406 1415

100 150 200 250 300 350 400

450

M/E

FlG. 4. Electron impact mass spectra of methyl ester-trimethylsilyl esters of HETEs from rat adrenal glomerulosa cells. Radioactive metabolites [peak 2 (top) and peak 4 {bottom)] from Fig. 3 were collected, derivitized, and analyzed by GC/MS.

FIG. 5. Electron impact mass spectra of methyl ester-trimethylsilyl ethers of HETEs from rat adrenal glomerulosa cells. Radioactive metabolites [peak 3 (top) and peak 7 (bottom)] from Fig. 3 were collected, derivitized, and analyzed by GC/MS.

(data not shown), inhibited the synthesis of the [14C] EETs and increased the synthesis of [14C]PGs and [14C] HETEs. Pretreatment with superoxide dismutase failed to alter the synthesis of the HETEs and EETs; however, the synthesis of PGs was reduced (Fig. 10). Arachidonic acid metabolism to PGs, HETEs, and EETs was increased in the presence of catalase.

dogenesis (8-10). Subsequently, lipoxygenase inhibitors were reported to inhibit All-induced aldosterone release implicating these metabolites as important mediators (15-17). None of these studies attempted to determine whether the inhibitors selectively blocked cyclooxygenase or lipoxygenase. Despite these numerous studies suggesting the importance of adrenal eicosanoids, there has been no attempt to define the metabolites of arachidonic acid that are produced by adrenal glomerulosa cells. The present study identifies the major adrenal metabolites of arachidonic acid and characterizes the effects of several inhibitors on the synthesis of these eicosanoids. Adrenal glomerulosa cells have been found previously to metabolize arachidonic acid to several PGs including PGI2 and PGE2 (13,14). In addition, there is chromatographic evidence that the cells also synthesize 12-HETE and 15-HETE (16). We have found that exogenous ara-

Discussion Metabolites of arachidonic acid have been implicated in the regulation of adrenal steroidogenesis (8-17). To a great extent, this conclusion was based on experiments with inhibitors of phospholipase, cyclooxygenase, or lipoxygenases. For example, inhibitors of PG synthesis reduced the basal and All-stimulated aldosterone release suggesting that a PG mediated a portion of the steroi-


500 •

r

250 •

5,6-

14,15 11,12

A.

8,9


2190

I j 0 • 60

120

60

120

500

250 -

FRACTION

FIG. 6. Separation of the EETs synthesized by rat adrenal glomerulosa cells. The EETs were collected as two fractions, fraction A (130-140) and B (141-160), from the reverse-phase separation illustrated in Fig. 1. These fractions were rechromatographed on normal phase HPLC using method D. Fraction A is illustrated in the top chromatogram and fraction B in the bottom.

chidonic acid is metabolized to several oxygenated metabolites that comigrate on HPLC with the PGs, HETEs, and EETs. These eicosanoids appear to be products of adrenal cell metabolism since their presence was not observed when media without cells were incubated under identical conditions. Analysis of the PGs confirmed our previous findings that the cells synthesize PGI2, PGE2, PGF2«, and PGD2 (14). The identification of these PGs was based on comigration of radioactive metabolites with known PG standards after coinjection on HPLC. The synthesis of PGs is catalyzed by cyclooxygenase and is inhibited by indomethacin and related drugs (1). As previously reported, we found that indomethacin reduced the synthesis of PGs by adrenal cells (9-14). Similar results were obtained with the cyclooxygenase/lipoxygenase inhibitors ETYA and BW755c. Additionally, the adrenal cells synthesized a group of metabolites with the chromatographic properties of the HETEs. These metabolites were also isolated, and the major products were identified as 12-HETE and 15HETE. The identification of the structures of the HETEs was based on 1) comigration with known standards on reverse phase and normal phase HPLC, 2) UV absorbance maximum of 235 nm indicating a conjugated diene, 3) comigration of the methyl ester-trimethylsilyl

Endo • 1991 Voll28«No4

ether derivatives on GC with known standards, 4) a mol wt of 406 of the derivatized metabolite on analysis by chemical ionization MS and 5) the presence of a typical fragmentation pattern and molecular ion of the derivatized metabolite when analyzed by electron impact MS. The synthesis of the HETEs may occur by several different pathways: lipoxygenases (2-4), cytochrome P450 (7), or lipid peroxidation by oxygen free-radicals (24, 25). We tested the effect of several inhibitors on the adrenal metabolism of arachidonic acid to gain an insight into the biosynthetic pathway involved in HETE production. The lipoxygenase inhibitors, NDGA, baicalein, and AA861, were not selective in their inhibition of HETE synthesis when used in concentrations that inhibit angiotensin-induced aldosterone release (16, 17). All three drugs reduced the synthesis of PGs, HETEs, and EETs. Similar results were obtained with the combined cyclooxygenase/lipoxygenase inhibitors, ETYA and BW755c. Other investigators have also reported this lack of specificity for these inhibitors (21, 26, 27). In contrast, clotrimazole and metyrapone, cytochrome P450 inhibitors, increased HETE synthesis. Superoxide and hydrogen peroxide are synthesized and released by several types of cells and are byproducts of oxidations by cyclooxygenase and cytochrome P450. Peroxidation of extracellular lipids by superoxide and hydrogen peroxide may be blocked by enzymes that degrade these oxidative intermediates (5). Superoxide dismutase failed to alter HETE synthesis while catalase increased it. These findings eliminate the possibilities that HETE synthesis occurs by mechanisms involving cytochrome P450 or oxygen free-radical lipid peroxidation. By a process of elimination, it appears that HETE synthesis involves a 12- and 15-lipoxygenase; however, this conclusion would be strengthened by the availability of a specific lipoxygenase inhibitor. In agreement with this conclusion, Nadler and co-workers (16) also reported that the synthesis of 12-HETE and 15HETE was inhibited by BW755C but not by metyrapone or SKF525A and suggested that these HETEs were products of lipoxygenase and not cytochrome P450. However, unlike the present study, the effect of these drugs on the synthesis of PGs and EETs was not determined. These data suggest that the 12- and 15-lipoxygenase pathways exist in adrenal cells. The initial product of lipoxygenase metabolism is the HPETE which may degrade spontaneously or through the action of a peroxidase to the HETE (2, 3). In addition, the HPETE may undergo further metabolism to LTs, hepoxilins and trioxilins, diHETEs, or short chain aldehydes (4, 28-30). Using the methods described, we failed to detect the production of HPETEs. This may reflect spontaneous degradation to the HETEs in the incubation, extraction, and purification procedures, or the adrenal cell may contain peroxidases that rapidly degrade the HPETEs.


2191

ARACHIDONATE METABOLISM IN ADRENAL CELLS A. PEAK 1-14,15-EET-PBF

B. PEAK 2-11,12-EET-PBF 3

•

•

1

' • » * i»

i.

. ll «r. I T , '

'

'

* i m '

C. PEAK 3-8,9-EET-PBF

T>

113

•

1481

w

,*-.,.TTE.J

D. PEAK4-5.6-EET-PBF

T -i- ,«

I

FIG. 7. Negative ion chemical ionization mass spectra of the pentafluorobenzyl (PFB) esters of the EETs from rat adrenal glomerulosa cells. The radioactive metabolites from Fig. 6 were collected, derivitized, and analyzed by GC/MS.

The rapid degradation of the HPETE to the HETE would be consistent with the absence of secondary metabolites that may be derived from further metabolism of the HPETEs. For example, we failed to detect the production of LTs of the 5- or 15-series, and no metabolites were found that comigrated with the diHETEs, hepoxilins, or trioxilins. Unlike these results in zona glomerulosa cells, Hirai et al. (31) reported that cells from the adrenal zona fasciculata synthesized 5-HETE, LTB4, and the 6-trans isomer of LTB4. However, the identification of these metabolites was based only on comigration with known standards on HPLC. A more rigorous approach to the identification of these metabolites will be required to confirm these findings. A group of metabolites was also observed that comigrated with the EETs. These metabolites were isolated and identified as 14,15-EET, 11,12-EET, 8,9-EET, and 5,6-EET. The structure of the EETs was based on 1) comigration with known standards on reverse phase and normal phase HPLC, 2) absence of UV absorbance, 3) comigration of the pentafluorobenzyl ester of the metabolite with known standards on GC and 4) a mol wt of the metabolites of 320 when analyzed by negative ion chemical ionization MS. Since electron impact mass

spectra of the four regioisomers of the EETs are similar and provide no insight into the position of the epoxy group, this analysis was not performed on the biological samples. These data are the first demonstration of EET synthesis by adrenal glomerulosa cells. The only biosynthetic pathway reported for the EETs involves a cytochrome P450 epoxygenase (6, 7). The involvement of this pathway is supported by our results that the cytochrome P450 inhibitors, metyrapone and clotrimazole, decreased the synthesis of the EETs but increased the synthesis of the HETEs and PGs. The finding of EET synthesis by the adrenal may have been anticipated since the adrenal is a rich source of cytochrome P450 enzymes that are involved in steroidogenesis. When used in concentrations that block EET synthesis, cytochrome P450 inhibitors also block steroidogenesis. The possibility, however, exists that EET synthesis occurs via a cytochrome P450 that is not involved in steroidogenesis. This possibility is supported by the findings that ETYA, indomethacin, and baicalein block EET synthesis but only have small effects on aldosterone release (14, 16). In addition to EETs synthesis, cytochrome P450 is involved in omega and omega-1 oxidation of fatty acids and eicosanoids (32, 33). In the adrenal zona glomerulosa



2192

epoxidation of arachidonic acid. We examined the distribution of these metabolites between the cells and the incubation media. All of the metabolites appeared to be present in both the cells and media. There was no indication that any of the metabolites were concentrated in the cells. Several studies have shown that HETEs are incorporated into the cellular lipids of some cells (34-36). We found that exogenously added 5-HETE was incorporated into the diglycerides and 12-HETE was taken up into phospholipids of adrenal glomerulosa cells (37). Along these lines, when liver homogenates or platelets were incubated with arachidonic acid, EETs were recovered from the cell phospholipids as well as the incubation medium (38, 39). The largest amount of the EETs was located in cell phospholipids, principally phosphatidylinositol. However, in the present study, there was no indication that the endogenously synthesized EETs or HETEs were incorporated into lipids of adrenal cells. The only radioactive product that was isolated from the adrenal cell lipids was unmetabolized arachidonic acid. Similar results were obtained in cells treated with All or ACTH, i.e., only [14C]arachidonic acid was esterified to the cellular lipids. In the initial studies on the role of arachidonic acid metabolites in the regulation of aldosterone release, indomethacin was found to inhibit basal and All-stimulated aldosterone release (8, 9, 11, 12, 14). It was concluded that PGs mediate steroidogenesis. However, subsequent studies failed to substantiate this conclusion (10, 11, 13-15). Instead, Nadler and co-workers (16, 17) suggested that 12-HETE was the mediator of All-induced aldosterone release. They reported that 12-HETE, but not 15-HETE, stimulated aldosterone release and that

"C-ARACHIDONIC ACID

350 300 250 200 150 100

£ z2 c

LLJ Q.

50 0

Cytochrome P450 arachidonic acid epoxygenase: stereochemical characterization of epoxyeicosatrienoic acids.

Epoxyeicosatrienoic Acids and 20-Hydroxyeicosatetraenoic Acid on Endothelial and Vascular Function.

Synthesis of epoxyeicosatrienoic acids and heteroatom analogs.

Biosynthesis, biological effects, and receptors of hydroxyeicosatetraenoic acids (HETEs) and oxoeicosatetraenoic acids (oxo-ETEs) derived from arachidonic acid.

Arachidonic acid epoxygenase. Stereochemical analysis of the endogenous epoxyeicosatrienoic acids of human kidney cortex.

Metabolism of arachidonic, eicosapentaenoic, and docosahexaenoic acids in HepG2 cells and rat hepatocytes.

Collisionally induced dissociation of epoxyeicosatrienoic acids and epoxyeicosatrienoic acid-phospholipid molecular species.

Metabolism of 12(R)-hydroxyeicosatetraenoic acid by rat liver microsomes.

H+ exchange and are mitogenic in cultured rat glomerular mesangial cells.

Inside epoxyeicosatrienoic acids and cardiovascular disease.

Cyclooxygenase-derived proangiogenic metabolites of epoxyeicosatrienoic acids.

Production of the polyunsaturated fatty acids arachidonic acid and eicosapentaenoic acid by the fungus Pythium ultimum.

Arachidonic acid metabolism in polymorphonuclear leukocytes: unstable intermediate in formation of dihydroxy acids.

Uptake and metabolism of circulating fatty acids by rat intestine.

Dual effects of dopamine in rat adrenal glomerulosa cells.

Stereospecificity of the hydroxyeicosatetraenoic and hydroxyoctadecadienoic acids produced by cultured bovine endothelial cells.

Effects of arachidonic acid on the concentration of hydroxyeicosatetraenoic acids in culture media of mesenchymal stromal cells differentiating into adipocytes or osteoblasts.

Metabolism of dicarboxylic acids in rat hepatocytes.

Effect of ACTH on 125I-labeled angiotensin II binding and response by rat adrenal glomerulosa cells.

Interaction of epoxyeicosatrienoic acids and adipocyte fatty acid-binding protein in the modulation of cardiomyocyte contractility.

Regiospecific and enantioselective metabolism of 8,9-epoxyeicosatrienoic acid by cyclooxygenase.

Epoxyeicosatrienoic acids and glucose homeostasis in mice and men.

Arachidonic acid stimulates phosphoinositide metabolism and catecholamine release from bovine adrenal chromaffin cells.

Epoxyeicosatrienoic acids protect cardiac cells during starvation by modulating an autophagic response.