Receptor-mediated gonadotropin action in the ovary. Rat luteal cells preferentially utilize and are acutely dependent upon the plasma lipoprotein-supplied sterols in gonadotropin-stimulated steroid production.

cells responded to luteinizing hormone, human chorionic gonadotropin, isoproterenol, and cholera toxin with an increase in progesterone and cyclic AMP responses. Administration of 4-aminopyrazolo(3,4~)p~midine (4-UP) (12.5 mgfkg body weight) for 3 consecutive days reduced the circulating cholesterol level from 47.32 2 1.12 mg/dl to 8.16 & 0.43 mg/dl and plasma progesterone from 280 f 26 ng/ml to 96 k 6 ng/ ml. Cellular sterol ester content was reduced to 25% following 4-APP injection. Injection of drug also produced a small but significant decrease in luteal cell free cholesterol. Treatment of rats with 4-APP also reduced the in vitro basal and hormone-stimulated progesterone production. This treatment showed no effect on luteal cell ‘251-human chorionic gonadotropin binding or gonadotropin and cholera toxin stimulated cyclic AMP synthesis. Exposure of luteal cells isolated from control and 4-APP injected groups to homologous or heterologous lipoproteins (low and high density lipoproteins) produced significant stimulation in steroidogenesis, both under basal conditions and in response to gonadotropins. Similarly, luteal cells isolated from 4APP-injected rats incorporated [3H]cholesterol from [3~~cholesteryl linoleate into progesterone at a much faster rate compared to control cells, and this effect was further enhanced in response to human chorionic gonadotropin. Injection of rats with Triton WR 1339 (1 g/kg body weight) resulted in a 10-fold increase in circulating cholesterol level. This treatment, however, produced a significant decrease in plasma progesterone and cyclic AMP and progesterone responses in isolated cells. The inhibitory effect of Triton injection on steroidogenesis could not be reversed by lipoprotein addition. These studies suggest that rat luteal cells are acutely dependent upon exogenous lipoprotein cholesterol for the maximum expression of steroidogenic response.

These studies were aimed at evaluating the role of steroid precursor and circulating plasma lipoproteins in gonadotropin induced steroidogenesis. Rat luteal cells responded to luteinizing hormone, human chorionic gonadotropin, isoproterenol, and cholera toxin with an increase in progesterone and cyclic AMP responses. Administration of 4-aminopyrazolo(3,4-~) p~m i d i n e ( 4 -U P ) (12.5 mgfkg body weight) for 3 consecutive days reduced the circulating cholesterol level from 47.32 2 1.12 mg/dl to 8.16 & 0.43 mg/dl and plasma progesterone from 280 f 26 ng/ml to 96 k 6 ng/ ml. Cellular sterol ester content was reduced to 25% following 4-APP injection. Injection of drug also produced a small but significant decrease in luteal cell free cholesterol. Treatment of rats with 4-APP also reduced the in vitro basal and hormone-stimulated progesterone production. This treatment showed no effect on luteal cell '251-human chorionic gonadotropin binding or gonadotropin and cholera toxin stimulated cyclic AMP synthesis. Exposure of luteal cells isolated from control and 4-APP injected groups to homologous or heterologous lipoproteins (low and high density lipoproteins) produced significant stimulation in steroidogenesis, both under basal conditions and in response to gonadotropins. Similarly, luteal cells isolated from 4-APP-injected rats incorporated [3H]cholesterol from [3~~cholesteryl linoleate into progesterone at a much faster rate compared to control cells, and this effect w a s further enhanced in response to human chorionic gonadotropin. Injection of rats with Triton WR 1339 (1 g/kg body weight) resulted in a 10-fold increase in circulating cholesterol level. This treatment, however, produced a significant decrease in plasma progesterone and cyclic AMP and progesterone responses in isolated cells. The inhibitory effect of Triton injection on steroidogenesis could not be reversed by lipoprotein addition. These studies suggest that rat luteal cells are acutely dependent upon exogenous lipoprotein cholesterol for the maximum expression of steroidogenic response.
The steroidogenic tissues including ovary are unique in the sense that they require cholesterol not only for membrane biogenesis but also as a precursor for steroid hormone synthe-* This work was supported by Grant HD06656 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. . These studies are compatible with the idea that steroidogenic tissues could take up and utilize blood cholesterol. However, little information is available regarding the role of cholesterol and its utilization in isolated cell system. Studies reported by Brecher and Hyun (lo) suggest that adrenal cells isolated from 4-APP'-treated rats secrete less corticosterone in response to adrenocorticotropic hormone. The dependence of adrenocortical cells (11,12) granulosa cells (9), and choriocarcinoma cells (13) on exogenous cholesterol source has been established following prolonged incubation of these cells in lipoprotein-de~cient medium. While this manuscript was in review, Gwynne and Hess f 14) have reported that adrenal cells isolated from 4-APP-injected rats respond to HDL with an increase in corticosterone production. The purpose of this study was to determine the regulation of steroidogenesis in the ovarian cells by exogenously supplied lipoproteins and gonadotropins without prolonged culture in cholesterol-deficient medium. We also determined the effect of lowering and increasing blood cholesterol with 4-APP (15-18) and Triton WR-1339 (19-221, respectively, on subsequent steroidogensis in isolated cells, and reversal of their effects by lipoproteins in vitro. Our results show that the ovarian cells are unique among various steroid-producing cells (11)(12)(13) in their ability to take up and metabolize lipoprotein cholesterol upon isolation from tissue with no apparent need for culturing for prolonged periods of time. and pregnant mare's serum gonadotropin were purchased from Grand Island Biological Co., Grant Island, NY, and Organon, Oss, Holland, respectively. '"I-choriogonadotropin was prepared according to Catt et al. (23). All other reagents used were of analytical reagent grade.

Methods
Animals a n d Hormonal Treatment-Twenty-two-to 24-day-old female Sprague-Dawley rats (Spartan Research Inc., Hazlett, MI) were employed in the present studies. Highly luteinized ovaries from these rats were obtained following a regimen described by Parlow (24). Rats were injected subcutaneously with 50 I.U. of PMSG and followed 56 h later with 25 I.U. of hCG. Day 0 was taken as the day of hCG injection. Where required, rats were injected with 4-APP (intraperit.oneally, 12.5 mg/kg body weight) in PBS, pH 3.0, between 09.00 to 10.00 h on days 3, 4, and 5 after hCG administration, and animals were killed on day 6. In studies with Triton WR-1339, intraperitoneal injections of 1 g/kg body weight were given every 12 h, beginning on day 4 after hCG administration. Animals were killed 12 h after the last Triton WR-1339 injection. Control animals received the vehicle.
Preparation of Collagenase-dispersed Luted Cells-Collagenasedispersed luteal cells were prepared by a slight modification (25, 26) of the procedure described earlier from this laboratory (27)(28)(29). DNA content of the cells was measured by the colorimetric procedure of Burton (30).
Incubation Conditions for Progesterone Measurements-Unless otherwise stated, aliquots of luteal cells (1.5 to 2 X lob cells in 0.1 ml) were transferred into plastic tubes (12 X 75 mm) containing 0.3 ml of Medium 109,0.1% BSA, and, where required, appropriate concentrations of hCG, LH, cholera enterotoxin, 8-Br-cAMP, or various lipoprotein fractions were also added. The incubations were carried out at 37 "C in a Dubnoff metabolic shaking incubator gassed with 02COz (95:5, v/v). Following incubation, the reaction was stopped by placing the sample tubes in a boiling water bath for 3 min. The samples were extracted with light petroleum ether and assayed for progesterone by radioimmunoassay as described earlier from this laboratory (28). Measurements of CAMP Accumulation in Luteal Cells-Luteal cells (approximately 2 X 10") were dispersed in a final volume of 0.4 ml of Medium 109 containing 0.1% BSA, 0.5 mM 1-methyl-3-isobutyhanthine, and, where necessary, various hormones at the concentrations specified in each experiment were also added. Incubations were performed at 3 i "C under 95% 02. 5% CO, with shaking at 100 cycles/min. After incubation, usually for 2 h, the reaction was stopped by placing tubes in a boiling water bath for 5 min and then transferred to ice. The samples were processed and assayed for CAMP by the procedure of Gilman (31) as described earlier (32).
De~erminution of '"I-hCG Binding to Luteal Celis-The binding of "'I-hCG was determined by a minor modification of a procedure of Clark and Menon (28) as described under Table 11. Briefly, cells (2 X 10" cells) were incubated with "'I-hCG (approximately 150,000 cpm, 25 to 35,000 cpm/ng) in 0.4 ml of Medium 109/0.1$ BSA and in the presence or absence of 10 pg/ml of unlabeled hCG. Aft.er incubation at 37 "C for 3 h, the tubes were centrifuged at 500 X g for 30 min. The supernatant was withdrawn and the radioactivity of washed pellet was determined in an automatic gamma counter, The specific binding was calculated from the difference of total binding to that observed in the presence of a 1000-fold excess of unlabeled hCG.
Miscellaneous Procedures-Total plasma or serum cholesterol was determined by the procedure of Zak (37). For lipid analysis, lipids were extracted from the lipoprotein fraction in chloroform/methanol (2:1, v/v) (38). Lipoprotein phospholipid phosphorus was determined by the procedure described previously (39,40). The relative contribution of individual phospholipids to total phospholipids and free cholesterol and cholesterol ester to total cholesterol was determined after the separation of individual lipid by thin layer chromato~aphy (41). The separation of neutral lipids, including cholesterol and cholesterol esters, was accomplished by single dimension, two-step development thin layer chromatography (41), using isopropyl ether/ acetic acid (96:4, v/v) as the fvst developing solvent and petroleum ether/diethyl ether/acetic acid (9010:1, v/v) or diethyl ether/hexane (694, v/v) as the second solvent. Cholesterol and cholesterol esters were eluted successively with diethyl ether (20 ml); chloroform/methanol (10 ml; 4:1, v/v); followed by chloroform/methanol (10 ml; 2:l. v/v). Cholesterol esters after saponification in alcoholic KOH and free cholesterol were quantitated by the colorimetric procedure of Zak (37) and, if necessary, by the micromethod of Glick et 01. (42). Protein content of lipoproteins was determined by a modification of the procedure of Lowry et ul. (43) as described by Markwell et al. (44).
Incorporation of fHJCholestero1 from j3H/CholesteryE Linoleate-Human LDL into Progesterone by Luteal Cells-Rat luteal cells (approximately 2 X 10") were incubated in a final volume of 1 ml of Eagle's Minimum Essential Medium/Medium 109 containing 1 mg/ ml of BSA, 20 pg of protein/ml of ["H]cholesteryl linoleate-human LDL (60,000 cpmfnmol, cholesteryl linoleate), and, where required, 10 ng/ml of hCG were also added. After incubation for 4 h in an atmosphere of O,/CO, (955, viv), I ml of phosphate-buffered saline was added to each tube and transferred to boiling water bath for 3 min. ["CCfProgesterone (50 pg, 1000 cpm), 20 a-hydroxypregn-4-ene-%one (50 pg), and pregnenolone (50 pg) were added as carriers, and steroids were extracted two times with 5 volumes of light petroleum ether and two times with 5 volumes of ethyl acetate. The petroleum ether and ethyl acetate fractions were dried separately under nitrogen. The residues from two fractions were dissolved in chloroform/methanol (2:1), combined, and redried under a stream of nitrogen. The residues were redissalved in 3 ml of 9056 aqueous methanol and partitioned three times against 1 ml of hexane (9,45). Steroid in methanolic phase was separated by thin layer chromatography using Silica Gel G glass plates. The plates were developed either one time in a solvent system of chloroformidiethyl ether (9:1, v/v) (46), three successive times in a system of isopropyl ether/diethvl ether/acetic acid (70302, v/v) (9), or three times in a solvent system of isopropvl ether/petroleum ether/acetic acid (70:201, v/v) (47). Authentic standards. cholesteryl oleate, cholesterol, progesterone. 20 cu-hydroxypregn-4-en-3-one, and pregnenolone were always run simultaneously, and steroids were visualized by exposure to iodine vapor. Individual steroids were eluted in chloroform/methanol (2:l) and counted for radioactivity determinations.

RESULTS
Luteal C~l~~t e r o i d o~e n e s i s in Response to ~o~a d o t r o~i n~, Cholera Enterotoxin, and Cyclic Nucleotide Derivative-The results presented in Fig. 1 (A-D) show progesterone production by luteal cells in response t.o various stimulators. Sensitivity of the luteal cells to physiological concentrations of hCG is shown in Fig. 1A. Concentrations of hCG as low as 0.01 ng/ ml significantly stimulated progesterone production and this effect was enhanced with increasing concentrations of hCG reaching a maximum at 0.1 ng/ml of hormone. The concentration of hormone required for half-maximal stimulation (EDw) was 0.02 ng/ml. The dose-dependent stimulatory effect of bovine LH (lutropin) on steroid synthesis by luteal cells is shown in Fig. 1B. T h e EDio of LH was around 0.5 ng/ml. Similarly, cholera toxin, a universal stimulator of adenylate cyclase in various mammalian systems (44)(45)(46) also effectively enhanced steroidogenesis in a dose-related manner, with EDm equivalent of 4 n g / d (Fig. 1C). Results presented in Fig. 1D show the stimulatory action of 8-Br-CAMP on luteal cell steroid production. These cells were also very responsive to 8-Br-CAMP and maximum stimulatory effect of the nucleotide was at a concentration of 0.5-1 m~ with EDw 0.1 mM. These results present the stimulatory conditions of luteal cell steroidogenesis.

Lipoproteins and Ovarian Steroidogenesis
The results presented in Fig. 2 show time-dependent increases in progesterone production under basal conditions and in response to hCG (10 ngiml). In the presence of hormone, progesterone synthesis increased almost linearly with increase in incubation time reaching a maximum at 120 min. Basal production of progesterone also increased up to 60-90 min.
Effect of Various Hormones on cAMP Response in Luteal Cells-The results presented in Fig. 3 compare the ability of hCG, LH (NIH-LHBS), and cholera toxin to stimulate CAMP accumulation in luteal cells. LH and hCG produced about 8to 10-fold stimulation of cAMP accumulation. Cholera toxin also effectively enhanced cAMP synthesis, although maximum effect was of a lower magnitude compared to that observed with gonadotropin. The concentration of hCG, LH, and choleratoxin required for half-maximal stimulation of cAMP production (EDa) were found to be 5.5,20, and 10 ne/ ml, respectively. Addition of 1-methyl-3-isobutylxanthine (0.5 mM), an inhibitor of cyclic nucleotide phosphodiesterase, greatly potentiated the stimulatory effect of tropic hormone and cholera toxin. Other hormones, including follicle-stimu" lating hormones, th~oid-stimulating hormone, growth hormone, and prolactin did not affect cyciic AMP synthesis (data not presented). Further experiments reported below were performed to examine the effect of modulation of blood cholesterol levels by pharmacological agents on cAMP and progesterone responses.
Effect of I-APP Treatment for Rat on Subsequent Normonal Steroidogenesis in Isolated Luteal Cells-Results presented in Table I show the effect of 4-APP injection on plasma cholesterol, progesterone, luteal cell cholesterol, and cholesterol esters and in uztro steroidogenic response of luteal cells. 4-APP injection (12.5 mg/kg body weight) reduced piasma cholesterol level from 47.32 ~t 1.12 mg/dl (control) to 8.16 2 0.43 mg/dl (4-APP). This treatment also reduced the plasma progesterone level from 280 -I 26 ng/ml to 96 r+-6 ng/ml. Similarly, cellular cholesterol ester content was reduced to about 25% in luteal cells isolated from 4-APP-injected rats. Free cellular cholesterol content, however, was less affected by this treatment.
Results presented in Table IA Table I1 show that 4-APP treatment did not affect the subsequent ability of luteal ceus to bind '"I-hCG and express gonadotropin-and cholera toxin-stimulated cyclic AMP accumulation. These data clearly demonstrate that the inhibitory action of 4-APP on progesterone synthesis was at a point distal to hormone-receptor interaction and cyclic AMP accumulation.
Effect of Various Lipoprotein Fractions on Luteal Cell Steroidogenesis-Since 4-APP treatment reduced circulating blood cholesterol level and inhibited ovarian steroidogenesis, it was of interest to delineate the underlying mechanism, Initial attempts were made to reverse the inhibitory action  Effect of 4-APP injection of '251-hCG binding and hormone-stimulated cyclic AMP accumulation in isolated luteal cells Female rats pretreated with PMSG and hCG were injected with ' presence of Oz/C02 (955, v/v). At the end of incubation, binding was PBS or 4-APP (12.5 mg/kg body weight) under the conditions de-stopped by the rapid addition of 2 ml of incubation medium and scribed under "Experimental Procedures." The luteal cells isolated j centrifugation at 250 X g for 10 min. The sedimented cells after an from both groups of ovaries were used for the determination of lz5I-~ additional wash in medium were counted for radioactivity. The results hCG binding and cyclic AMP accumulation in response to various , are expressed as specific binding, computed from the difference of stimulators. For Iz5I-hCG binding, luteal cells (approximately 2 X IO6) : total binding with that observed in the presence of unlabeled hCG. were incubated with '251-hCG (180,000 cpm, 30 cpm/pg) in the pres-, Conditions for cyclic AMP production were the same as described in ence or absence of 2.5 pg/ml of unlabeled hCG in 0.4 ml of Medium , Table I

FIG. 5. Effect of increasing concentrations of human HDL (h-HDL) on gonadotropin-and 8-Br-CAMP-induced steroidogenesis in luteal cells of PBS (Panel A)-or 4-APP (Panel B)injected rats.
The experimental details were the same as described in Fig. 2

Effect of Triton WR-1339 Injection on Progesterone and Cyclic AMP Responses in Rat Luteal Cells-The studies
reported thus far suggest that reduction in circulating blood cholesterol levels leads to subsequent inhibition of luteal cell steroidogenesis. We then evaluated the possibility of increasing blood cholesterol levels on the luteal cell steroidogenic capacity. To accomplish this, rats were injected with Triton WR-1339, which is known to increase circulating blood cholesterol level (19-22). Injection of rats with Triton WR-1339 (1 g/kg body weight) every 12 h for 2 days raised the plasma

Effect of Triton WR-1339 injection on "~I-chor~ogonado~rop~n binding and hormone-stimulated CAMP accumulation in isolated luteal cells
Pseudopregnant rats were injected with Triton WR-1330 under the conditions described in Table IV. Isolated cells were assessed for "'11-hCG binding and CAMP production as described in Table 111.  (Table IV). Finally, Triton WR-1339 injection produced only a slight decrease in luteal cell cholesterol ester and free cholesterol contents.
Results of Table V show the effect of TritonWR-1339 on '"I-hCG binding and cyclic AMP accumulation in luteal cells in response to various stimulators. As evident, there was no change in '"I-hCG binding to luteal cells of Triton WR-1339 injected rats compared to saline-injected rats. However, this treatment significantly reduced the extent of cyclic AMP synthesis by luteal cells in response to hCG or cholera enterotoxin.

DISCUSSION
The present studies were aimed at evaluating the role of precursor and circulating plasma lipoproteins in gonadotropin induced steroidogenesis in ovary. Luteal cells isolated from PMSG/hCG-primed pseudopregnant rat ovaries were used as a model system. These cells were chosen because of their extreme sensitivity to physiological doses of gonadotropin. Studies reported here demonstrate that these cells respond to LH, hCG, and isoproterenol with an increase in progesterone production (48)(49)(50). Further studies were undertaken to study systematically the effect of varying circulating blood cholesterol on progesterone s-ynthesis. The 4-APP, which is an adenine analog, has previously been shown to inhibit secretion of all major classes of plasma lipoproteins and causes a decrease in circulating levels of plasma cholesterol (17)(18)(19). In the present studies, 4-APP-mediated decrease in circulating level of cholesterol was closely correlated with the decrease in plasma progesterone levels. Interestingli, administration of 4-APP to rats greatly reduced the in vitro steroidogenic capacity of isolated luteal cells both under basal conditions and in response to various stimulators. While maximum steroid production was decreased considerably, cells isolated from 4-APP-injected rats still retained the ability to respond to gonadotropins. The lack of effect of 4-APP on '2'II-hCG binding and gonadotropin stimulated cyclic AMP production, along with the decreased responsiveness of cells to 8-Br-CAMP and Bt2cAMP, further support the notion that the 4-APP effect was primarily at a step after the hormone/receptor/ adenylate cyclase system and, more specifically, at the precursor level. Indeed, in the present studies, 4-APP administration resulted in a major decrease in luteal cell sterol esters and a small but significant decrease in free cholesterol content. Our next approach was to see if 4-APP-mediated reduction in steroidogenic response could be reversed in rritro by lipoproteins. To o w surprise, incubation of luteal cells from PBS (control)-injected rats, with human LDL or human HDL significantly enhanced the steroidogenesis over that observed in the absence of lipoprotein addition. These cells are, thus, clearly different from those of adrenocortical cells (11, la), granulosa cells (9) and choriocarcinoma cells (13) which require prolonged exposure to lipoprotein-deficient serum before becoming dependent on exogenous lipoprotein cholesterol for steroid synthesis. Luteal cells are therefore unique in their ability to take up and metabolize lipoprotein cholesterol immediately upon the isolation from the tissue with no apparent need for prolonged culture. Various lipoproteins effectively reversed the inhibition of steroidogenesis seen in cells isolated from 4-APP-pretreated rats. It was also demonstrated that cells of 4-APP-injected rats incorporated a higher amount of [SH]cholesteryl linoleate-LDL delivered cholesterol into progesterone compared to cells of PBS-injected rats. The extensive in uivo studies reported by Anderson and Dietschy (6) suggest that three major steroid producing tissues, ovary, adrenal, and testis, preferentially take up and incorporate cholesterol from HDL rather than LDL into steroid. Further, infusion of HDL from human or rat plasma to 4-APP-treated rats was reported to block the [l-I4CJacetate incorporation into digitonin precipitable sterols in ovary and adrenal gland (18). However, in contrast to the adrenal gland, infusion of LDL failed to exert any effect on [1-'"C]acetate incorporation into digitonin precipitable sterols in the ovary. In contrast, mouse adrenal glands were shown to obtain cholesterol in uiuo from two ~ipoprotein systems, one specific for LDL and the other specific for HDL (7). The present studies, as well as those reported earlier for cultured granulosa cells (9), suggest that either LDL or HDL can supply cholesterol for steroidogenesis in cells of ovarian origin, although possibly not to the same extent.
The elegant work of Brown and Goldstein suggests that, in many LDL-responsive cells, the lipoprotein is bound and taken up by a saturable membrane-associated process (51)(52)(53)(54). Following internalization and degradation of LDL particle, the released cholesterol could be utilized for membrane biogenesis or steroid synthesis or stored as cholesterol ester. Relatively higher intracellular concentrations of cholesterol lead to suppression of activity of hydroxymethylglutaryl COA reductase and in some instances, such as in human fibroblasts, reduction in the number of binding sites on cell surface, SO as to further reduce the uptake of lipoprotein cholesterol (53,54). We have recently demonstrated the presence of both LDL and HDL receptors in the luteal cells.' Thus, it is conceivable that a similar mechanism is operating in the luteal cells.
Injection of rats with Triton WR-1339 produced a IO-fold ' J. Hwang and K. M. J. Menon, unpublished results.
increase in plasma cholesterol, an effect observed earlier and notably by Anderson and Dietschy (18) and by Goidfarb (21). However, detergent administration leads to a significant reduction in circulating progesterone level, an effect in contrast to an increase in plasma cholesterol. Since Triton WR-1339 causes a marked increase in hepatic cholesterol synthesis (22, 55, 56) and an increase in circulating level of cholesterol (19-22), we were interested to test whether plasma cholesterol had any effect on luteal cell steroidogenesis. Administration of Triton WR-1339 significantly reduced the steroidogenic capacity of luteal cells in response to gonadotropin. Our efforts to reverse the inhibitory effect by the addition of various lipoproteins were unsuccessful. Further, since detergent injection also resulted in a decrease in cyclic AMP response, it is possible that detergent produced a general decrease in overall metabolic activity of luteal cells. However, the possibility that detergent treatment may change the metabolism of lipoproteins in luteal cells cannot be ruled out.