Journal of Neurochemistry. Raven Press, Ltd., New York. 1994 International Society for Neurochemistry. Neurosteroidogenesis in Rat Retinas. P. Guarneri, R.
Journal of Neurochemistry Raven Press, Ltd ., New York 1994 International Society for Neurochemistry
Neurosteroidogenesis in Rat Retinas P. Guarneri, R. Guarneri, *C . Cascio, tP . Pavasant, *F. Piccoli, and tV. Papadopoulos Institute of Experimental Medicine, CNR, *Institute of Neuropsychiatry, University of Palermo, Palermo, Italy ; and tDepartment of Anatomy and Cell Biology, Georgetown University Medical Center, Washington, D.C., U.S.A .
Abstract: Neurosteroids (steroids synthesized in the CNS) function by modulating neurotransmission. To establish an experimental model for investigation of neurosteroid synthesis and regulation, independent of blood-borne steroids, we examined the steroidogenic activity of isolated rat retinas. We identified progesterone, pregnenolone, dehydroepiandrosterone, desoxycorticosterone, 3a,5a-tetrahydrodesoxycorticosterone, 3ahydroxy-5a-dihydroprogesterone, 17-hydroxyprogesterone, and 17-hydroxypregnenolone together with their esterified forms. As pregnenolone is the precursor of all steroids, its formation was studied in detail as an index of a steroid-synthesizing tissue . Pregnenolone was identified further by gas chromatography coupled to mass spectrometry, and its in vitro synthesis was inhibited by lovastatin, an inhibitor of mevalonolactone and cholesterol biosynthesis . We then examined pregnenolone synthesis in the presence of mevalonolactone as a precursor of sterol formation together with lovastatin, which reduces endogenous mevalonolactone synthesis, as well as with inhibitors of pregnenolone metabolism . The incorporation of mevalonolactone into pregnenolone and its sulfate ester was time- and concentration-dependent and blocked by aminoglutethimide, a competitive inhibitor of cytochrome P450 side-chain cleavage (P450sj enzyme . Immunocytochemical studies with a specific antibody to P450scc revealed a primary localization of the enzyme at the retinal ganglion cell layer. A less pronounced immunostaining was also seen at cells of the inner nuclear layer. Compounds known to stimulate cyclic AMP content also stimulated pregnenolone formation by rat retinas. These results demonstrate that rat retinas synthesize steroids and, for the first time, they reveal the steroidogenic ability of neuronal cells. We propose rat retinas as an in vitro model system to study neurosteroidogenesis in the CNS. Key Words: Neurosteroidogenesis-NeurosteroidsP450 s,, enzyme-Cyclic AMP regulation-Rat retina . J. Neurochem. 63, 86-96 (1994) .
trolled by an in situ mechanism independent of peripheral endocrine glands (Baulieu and Robel, 1990 ; Paul and Purdy, 1992) . Indeed, it has been demonstrated that certain structures in the CNS have the capacity to metabolize cholesterol to pregnenolone, the parent compound of progesterone metabolites, corticosteroids, androgens, and estrogens (Baulieu and Robel, 1990, Guarneri et al ., 1992, Papadopoulos et al ., 1992) . Moreover, the cytochrome P450 side-chain cleavage (P450, ee) enzyme, which cleaves cholesterol to form pregnenolone, has been localized in the cerebral white matter (Le Goascogne et al ., 1987), and glial cells have been shown to be the primary site for pregnenolone synthesis (Hu et al ., 1987 ; Jung-Testas et al ., 1989 ; Guarneri et al ., 1992 ; Papadopoulos et al ., 1992) . Furthermore, enzymes responsible for steroid metabolism have been located in both neurons and glial cells (for references, see Paul and Purdy, 1992) . In light of these findings, steroid synthesis in the CNS may constitute a local regulatory mechanism of specific steroid levels that act as modulators of inhibitory and excitatory neurotransmission (Majewska, 1992 ; Paul and Purdy, 1992) . Thus, the term neurosteroids was proposed referring to steroids of central origin with action in the CNS (Baulieu et al ., 1987) . The concept of neurosteroidogenesis, however, is still in its infancy, because questions on the structure of the steroids synthesized, as well as on the regulation of their synthesis and their metabolism, remain to be answered .
Received September 10, 1993 ; revised manuscript received November 15, 1993 ; accepted November 22, 1993 . Address correspondence and reprint requests to Dr . P . Guarneri at Institute of Experimental Medicine, CNR, c/o Institute of Neuropsychiatry, University of Palermo, Via G . La Loggia, 1-90129 Palermo, Italy . Abbreviations used: (BU)ZcAMP, dibutyryl cyclic AMP ; DHEA, dehydroepiandrosterone ; DOC, desoxycorticosterone ; GC-MS, gas chromatography-mass spectrometry ; MVA, DI .-mevalonolactone; 3a-OH-DHP, 3a-hydroxy-5a-dihydroprogesterone ; P450,_ cytochrome P450 side-chain cleavage ; PBS, phosphate-buffered saline; RIA, radioimmunoassay ; SU 10603, 1,2,3,4-tetrahydro-4-oxo-7chloro-2-naphthylpyridine ; THDOC, 3a,5a-tetrahydrodesoxycorticosterone ; trilostane, (2a,4a,5a,17fi)-4,5-epoxy-17-hydroxy-3-oxoandrostane-2-carboni tri le .
It is well known that the brain possesses the complete enzymatic machinery for the biosynthesis of the steroid precursor, cholesterol, as well as enzymes involved in the metabolism of blood-borne steroids . However, only recently several lines of evidence have revealed that brain steroid concentrations are con-
86
NEUROSTEROIDOGENESIS IN RAT RETINAS In studying neurosteroidogenesis, we have to challenge the complexity and heterogeneity of brain organization, as well as obviate the peripheral endocrine steroid sources. Thus, most of the qualitative and quantitative conclusions concerning neurosteroids have been based upon in vivo and in vitro data obtained using gonadectomized/adrenalectomized rats (Rubel et al ., 1987 ; Purdy et al ., 1991) or cell cultures grown in serum-free conditions (Jung-Testas et al ., 1989 ; Guarneri et al ., 1992). In the present work, we developed a rat retina model for studying neurosteroid synthesis and regulation, as well as neurosteroid interaction with neurons. The retina possesses structural properties that offer advantages for analyzing CNS mechanisms . As a peripheral extension of the CNS, it is readily accessible and it can be easily dissected and incubated in vitro where intact membranes surrounding retinal living cells can be kept in an appropriate extracellular environment and removed from blood-retina barrier influences . Moreover, the capacity of rat retinas to synthesize steroids represents a novel and unexpected finding. MATERIALS AND METHODS Materials [7-'H(N)]Pregnenolone (25 .0 Ci/mmol) and 11,2,6,7_ 3 H]progesterone (93 .5 Ci/mmol) were obtained from DuPontNew England Nuclear, whereas the other radioactive steroids were purchased from ICN (Milan, Italy) . Lovastatin was a gift from Merck Sharp and Dohme. (2a,4(Y,5a,17ß)4,5-Epoxy-l7-hydroxy-3-oxoandrostane-2-carbonitriIe (trilostane) was a gift from the Sterling-Winthrop Group. 1,2,3,4-Tetrahydro-4-oxo-7-chloro-2-naphthylpyridine (SU 10603) was a gift from CIBA-Geigy . i3r.-Mevalonolactone (MVA), aminoglutethimide, forskolin, dibutyryl cyclic AMP I(BO2cAMP], and isoproterenol were obtained from Sigma. All chemicals were of HPLC grade and were purchased from commercial sources. Isolation of rat retinas Male Wistar rats weighing 160-180 g were kept at 18°C under a 12-h light/dark cycle (0800-2000). All experiments were performed between 1500 and 1700 . Retinas were quickly dissected in ice-cold Krebs-HEPES buffer (buffer A, in mM : NaCl, 128 ; CaCI Z 2 .7 ; MgS04 , 1 .2 ; Na2HPO4 , 1 ; glucose, 16 ; HEPES, 20, pH 7.4). Each pair of retinas was placed in flasks containing 3 ml of buffer A, incubated in a shaker-water bath at 37°C, and equilibrated with 95% 0=/5% COZ . Retinas were kept under these conditions during the entire incubation time established for each experiment. Incorporation of MVA into pregnenolone and its sulfate ester Retinas were allowed to equilibrate in buffer A for 5 min under the conditions described above. After this period, 20 p.M lovastatin, the competitive inhibitor of 3-hydroxy-3methylglutaryl-CoA reductase (Brown et al ., 1978), was added for 75 min. Furthermore, to prevent pregnenolone metabolism, retinas were also incubated in the presence of 20 pM trilostane and 10 pM SU 10603, which competitively inhibit 3(3-hydroxysteroid dehydrogenase (Putts et al ., 1978)
87
and 17a-hydroxylase (Skikita et al ., 1965) activities, respectively . The synthesis was initiated by the addition of MVA (20 pM) made up in buffer A containing 15 mM maleic acid and 0.5 mM NADP (pH 7.0), which were used as an NADPH-generating system . The lactone form of mevalonate was preferred, because it is rapidly taken up and readily incorporated into cell membranes (Edwards et al ., 1977 ; Faust et al ., 1979 ; Guarneri et al ., 1992) . In some experiments, 0.76 mM aminoglutethimide, a potent inhibitor of P450,,, (Hall, 1985), was added 15 min before MVA. All the reagents were prewarmed at 37°C before use. At the indicated time intervals, 2 ml of ice-cold deionized water was added and the flasks were transferred to an ice-cold bath . The samples were homogenized, and aliquots were removed for protein determinations . In order to monitor the eventual procedural losses occurring during the steps of pregnenolone extraction and chromatography, [7-'H(N)lpregnenolone (5,000 cpm, 25 .0 Ci/mmol) was used as internal standard . Then ethyl acetate (1 :3, vol/vol) was added to the samples, which were kept at -80°C until the following day . Extraction procedure The samples were thawed and extracted with ethyl acetate by mixing repeatedly for 20 min in a rotor mixer. The two phases were separated by centrifugation at 2,000 g for 5 min, then the organic phase was removed, and the extraction steps were repeated two more times. The pooled organic extracts contained unesterified pregnenolone . The aqueous phase was treated to extract the esterified pregnenolone by a technique that preferentially provides for the mild solvolytic splitting of sulfate groups (Burstein and Lieberman, 1958 ; Morfin et al ., 1992). Then the solvolysis reaction was performed by adding sulfuric acid, to drop pH to 1, and 20% (wt/vol) NaCI and by extracting three times with ethyl acetate as described before . An overnight solvolysis of the pooled extracts was carried out at 37°C, and [;Hldehydroepiandrosterone (['H]DHEA) sulfate was used as internal standard . The hydrolysis of pregnenolone sulfate allowed the analysis of this compound by the same procedures used for the unesterified form . Finally, both extracts were collected separately and evaporated to dryness . In some experiments in which the contents of unesterified and esteritied forms of progesterone, 3a-hydroxy-5a-dihydroprogesterone (3a-OH-DHP), DHEA, desoxycorticosterone (DOC), 3a,5a-tetrahydrodesoxycorticosterone (THDOC), 17a-hydroxyprogesterone, and 17a-hydroxypregnenolone were studied, we used the same extraction procedure by adding the radioactive form of each steroid to monitor the recoveries . Chromatography The sample extracts were taken up in n-hexane and applied to Sep-Pak silica cartridges (Waters, Milford, MA, U.S .A .) . Pregnenolone and the other steroids of interest were eluted with n-hexane/isopropyl alcohol (95:5, vol/vol) . Then they were separated by HPLC on a Hibar Lichrosorb Diol column (10 pm ; Merck, Rahway, NJ, U.S .A .) . The HPLC column was equilibrated with n-hexane, and steroids were eluted, at the flow rate of 0.5 ml/min, with an n-hexane/isopropyl alcohol (90 :10) gradient, which linearly increased up to 10% in 35 min, then from 10% up to 45% in 80 min, and finally reached 100% in 10 min. Pregnenolone and the other steroids were identified by the retention times observed with authentic radiolabeled steroids . Moreover, two alternative analyses
J . Neurohem ., Vol. 6_3, No. 1, 1994
88
P. GUARNERI ET AL.
were also used : (a) chromatography on a Lichrosorb Si 60 HPLC column (5 pm) developed under isocratic conditions with n-hexane/isopropyl alcohol (97:3, vol/vol) at the flow rate of 0.5 ml/min ; and (b) chromatography on thin-layer silica gel plates (F-254, Merck) using chloroform/ethyl acetate (4 :1, vol/vol) as the developing system (pregnenolone: R,- = 0.51) . Some of the chromatographic conditions used in this study were reported pr+..,r~iously (Guarneri et al ., 1992). Gas chromatography-mass spectrometry (GC-MS) Pregnenolone separated after HPLC was analyzed as methyloxime trimethylsilyl derivatives (Shackleton, 1986). Quantitative analysis Following the chromatographic procedures, the material corresponding to pregnenolone and to the other steroids studied was evaporated to dryness, dissolved in phosphate-gelatin buffer, and left at room temperature for 20 min . Different aliquots of each steroid were grantified by radioimmunoassay (RIA) performed with specific antibodies obtained from ICN (Milan, Italy) and used under the conditions recommended by the supplier . As anti-3a-OH-DHP and antiTHDOC antibodies are not commercially available, the two steroids were tested by the cross-reaction with the antipregnenolone antibody and anti-DOC antibody, respectively . An aliquot was also taken and counted by scintillation spectroscopy for recovery measurements, which averaged 7080%. Therefore, all the data were corrected for the procedural losses . RIA data were analyzed with the IBM-PC RIA Data Reduction program (version 4.1) obtained from Jaffe and Associates (Silver Spring, MD, U.S .A .) . Immunocytochemistry Rat eyes were fixed in 3 .7% formaldehyde, in phosphatebuffered saline (PBS) supplemented with 0.5 mM CaC1 2 and 1 .0 MM MgC12 in order to preserve better the integrity of the tissues. The following day, they were washed with Cat'/ Mgt'-free PBS for 24 h to remove formaldehyde . Tissues were embedded in polyester wax (Gal lard -Schlesinger Chemical Manufacturing Corp ., Carle Place, NY, U.S .A .) according to the method of Kusakabe et al . (1984)-. The tissues were dehydrated through graded steps of ice-cold methanol and then equilibrated with 50% (vol/vol) polyester wax in absolute ethanol at 37°C for 12 h. Then they were incubated further in 90% polyester wax in ethanol at 37°C for another 4 h before being embedded in 90% polyester wax/l0% ethanol solution using plastic embedding trays (Polysciences, Inc., Warrington, PA, U.S .A .) . Sections (7 pm) were then made at 0°C and picked up on egg albumincoated slides . The sections were allowed to dry slowly at 4°C and stored at 4°C until used . The polyester wax sections were dewaxed through a graded series of ethanol solutions (5 min for each step), followed by a 5-min incubation in distilled water. The sections were then incubated for 5 min in 10% hydrogen peroxide solution to inhibit endogenous peroxidase activity and washed with distilled water and Ca"/ Mg -+ -free PBS. Next, the sections were incubated with nonimmune goat serum (Zymed, San Francisco, CA, U .S .A .) in a moisture chamber at room temperature for 30 min, followed by P450 s« antibody (1 :100 dilution in 10% calf serum) for 4 h. After three washes in Ca2+/Mg2 '-free PBS (5 min each), the sections were incubated with horseradish peroxidase conjugated goat anti-rabbit IgG (1 :2,000 dilution) for
.1.
Neurochem., Vol. 63, No . 1, 1994
FIG. 1 . Histograms of unesterified and esterified steroids chromatographed on a Hibar Lichrosorb Diol HPLC column (10 pm) as described in Materials and Methods. Each peak was tested with specific antibodies by RIA. The bars correspond to the amounts of steroids eluted with a gradient of hexane/2-propanol (90:10) in hexane (see Materials and Methods) . The elution position of each steroid was identified using radiolabeled steroid standards. Data shown are means -- SD (n = 6) from two independent experiments .
I h, followed by three washes in the Ca2 '/Mg 2+ -free PBS. Finally, peroxidase activity was demonstrated by 3-amino9-ethylcarbazolc, a substrate for peroxidase, for 30 min . The reaction was stopped by washing with distilled water. The chromogen was then preserved by mounting in Crystal mount (Biomeda Corp ., Foster City, CA, U.S .A .) and coverslip using Permount (Fisher Scientific, Columbia, MD, U.S .A .) . Pictures were taken by using an Olympus microscope (model BH-2) with Kodak Gold 400 ASA film . Cyclic AMP stimulation In some experiments, we studied the in vitro effects of different concentrations of forskolin, (Bu) ZcAMP, and isoproterenol on endogenous pregnenolone formation. Retinas kept in buffer A at 37°C and exposed to gas mixture were incubated with 20 pM trilostane and 10 pM SU 10603 for 45 min, during which time stimulants were added. The incubation was interrupted as already described, and pregnenolone and its sulfate ester were extracted and tested by RIA. Protein determination Protein content was determined according to the method of Lowry et al . (1951) using bovine serum albumin as standard. To eliminate the interferences of buffer A with protein assay, proteins were precipitated with 0.25 M perchloric acid and then dissolved with 0.1 M NaOH .
NEUROSTEROIDOGENESIS IN RAT RETINAS
89
Statistical analysis Student's t test was used for the analysis of experimental data, which are expressed as means ± SD . RESULTS Neurosteroids in rat retinas The study on the steroidogenic ability of rat retinas was approached by determining the content of both unesterified and esterified forms of pregnenolone and its derivatives . Retinal extracts containing the unesterified and esterified steroids were analyzed separately by HPLC, and the peaks were identified with the corresponding steroids by coelution with radiolabeled steroid standards and by their immunoreactivities with specific antibodies . Figure l shows a representative chromatogram of unesterified and esterified steroids obtained using the conditions described under Materials and Methods.The quantitative determinations of each peak based on the immunoreactivities with specific antibodies for each steroid are shown in Table 1 . The concentrations of the unesterified and esterified forms of pregnenolone were considerably higher with respect to the other steroids tested, whereas the levels of DOC and 3a-OH-DHP were the lowest . Moreover, we found that in retinas the levels of all sulfate esters tested were greater than those of the unesterified forms . As antibodies directed against THDOC and 3a-OHZ, are not commercially available, their corresponding HPLC peaks were analyzed by anti-DOC and antipregnenolone antibodies, respectively . The relative quantitative analysis of THDOC revealed that levels of both unesterified and esterified forms present in retinas were quite high, even higher than those of DOC itself. The peak matched with 3a-OH-DHP lightly crossreacted with the anti -pregnenolone antibody . Despite the low concentration detected, this peak was uniTABLE 1 . Steroid levels in rat retinas Unesterified forms (ng/mg of protein) Progesterone Pregnenolone 3a-OH-DHP DHEA DOC THDOC 17-Hydroxyprogesterone 17-Hydroxypregnenolone
0 .08 0 .17 0 .01 0 .09 0 .05 0 .13 0 .11 0 .12
± -! !± ±± ±
0.01 0.03 0 .003 0 .01 0 .01 0 .03 0 .02 0 .06
Esterified forms (ng/mg of protein) 0.17 0.68 0.02 0.25 0.09 0.33 0.18 0 .22
± ± ± ±± ± !±
0 .04 0 .11 0 .002 0 .02 0 .03 0 .05 0 .02 0 .04
Both unesterified and esterified steroids extracted from rat retinas were applied separately onto a Lichrosorb Diol HPLC column under the conditions described in Materials and Methods . Before chromatography, esterified steroids were solvolyzed in order to split the sulfate groups as described in Materials and Methods . The peaks obtained from HPLC separation were identified by coelution with the authentic radiolabeled steroids and tested through immrmoreactivity with specific antibodies for each steroid . Data represent the means ± SD (n = 4) from four different experiments .
FIG . 2. In vitro metabolic fate of pregnenolone and pregnenolone sulfate in intact retinas . Following dissection, retinas were incubated in Krebs-HEPES solution at 37°C under gas exposure . At different times of the incubation, pregnenolone and its esterified form were extracted, chromatographed, and analyzed as described in Materials and Methods . Data represent the means ±- SD (n = 6) from three independent experiments . .p < 0 .05 ; ""p
