Purification and characterization of equine testicular cytochrome P-450 aromatase: comparison with the human enzyme.

Cytochrome P-450 aromatase was purified by five chromatographic steps from adult stallion testis. It was first separated from NADPH-cytochrome P-450 reductase (reductase) on omega-aminohexyl-Sepharose 4B then purified to homogeneity on concanavalin A-Sepharose 4B, hydroxyapatite-Sepharose 4B, DEAE-Sepharose CL-6B and on a second hydroxyapatite-Sepharose 4B. On the other hand, purifications of the equine testicular and rat liver reductases, which allowed the reconstitution of aromatase activity in vitro, were achieved for each species in one chromatographic step on an adenosine 2',5'-diphosphate-agarose affinity column. Analysis on SDS/PAGE indicated single bands with apparent molecular masses of 53, 82, and 80 kDa for purified equine testicular cytochrome P-450 aromatase (eAROM), equine testicular reductase and rat liver reductase respectively. eAROM shows a time- and concentration-dependent activity that was stable for at least 2 months when stored at -78 degrees C. It is a highly hydrophobic protein composed from 505 residues and direct sequencing of its N-terminal part showed good homology when compared with human aromatase. When deglycosylated by N-glycosidase-F the apparent molecular mass of eAROM was decreased from 53 to 51 kDa as revealed by electrophoresis, its activity, however, was not impaired. eAROM exhibits much higher affinity for androgens than for 19-norandrogens, Km values were approximately 3, 16 and 170 nM for androstenedione (A), testosterone (T) and 19-nortestosterone (19-NT) respectively. However, it aromatizes 19-norandrostenedione (19-NA) slightly more efficiently than A, the estrone (E1) formed was 4.27 vs 3.54 pmol min-1 micrograms-1 respectively (P < 0.01). After incubation of eAROM with radiolabelled A and separation of steroids on HPLC, E1, 19-hydroxyandrostenedione (19-OHA) and 19-oxoandrostenedione (19-oxoA) were accumulated in the incubation medium in a time-dependent manner. The presence of 4-hydroxyandrostenedione (4-OHA), a suicide inhibitor of aromatase, cause a time-dependent inactivation of the enzyme. Whereas the activity of eAROM was unchanged in the presence of K+ (up to 250 mM), it was increased in the presence of EDTA (up to 50 mM) and decreased in the presence of DTT or Mg2+ (from 25 mM). We conclude that: (a) eAROM is a glycoprotein, however, deglycosylation by N-glycosidase-F does not appear to impair its activity, (b) eAROM aromatizes really both androgens and 19-norandrogens having a higher affinity for androgens, (c) the intermediary compounds of aromatization 19-OHA and 19-oxoA appear to be synthesized by the same active site that synthesizes E1 as the final product, (d) the inhibition of eAROM by increasing concentrations of Mg2+ and the stimulation of its activity by EDTA, taken together, indicate the importance of negatively charged residues in the polypeptide chain of equine aromatase, which play a role in enzymatic activity.

The synthesis of 19-norandrostenedione from dehydroepiandrosterone in equine placenta is inhibited by aromatase inhibitors 4-hydroxyandrostenedione and fadrozole.

19-Norandrostenedione was synthesized in vitro from dehydroepiandrosterone by explants of equine full-term placenta. The synthesis of 19-norandrostenedione was inhibited by two specific aromatase inhibitors, 4-hydroxyandrostenedione and fadrozole.

Immunohistochemical localization of cytochrome P450 aromatase in equine gonads.

Estrogens are the major steroids produced by equine gonads. To identify the cells responsible for estrogen synthesis, an antiserum against purified equine testicular cytochrome P450 aromatase was produced in rabbits. The reactivity and specificity of the antiserum were assessed by ELISA, immunoblot analysis, and immunoneutralization studies. Immunofluorescence microscopy demonstrated that in the male gonad, cytochrome P450 aromatase (P450arom) was localized in the interstitial tissue, whereas, under the experimental conditions used, the Sertoli and germ cells did not show any specific staining. In the ovary, the granulosa cells of small follicles exhibited faint immunofluorescent staining for P450arom and the granulosa cells of large, viable more follicles showed a high degree of immunoreactivity. In the corpus luteum, all the luteinized cells showed immunoreactivity. No immunoreactivity was detected in other cells of small and large viable follicles. Immunolocalization of P450arom in the equine testicular Leydig cells and in ovarian granulosa and luteinized cells indicates that these cells have the ability to metabolize androgens to estrogens and possibly to catechol estrogens.

In vitro 19-norandrogen synthesis by equine placenta requires the participation of aromatase.

Explants of equine full-term placenta have been shown to synthesize 19-norandrogens from labelled androgens. Steroid metabolites were purified by silica-gel column chromatography then analysed and quantified by c18-reverse-phase HPLC coupled to a radioactive flow detector. 19-Norandrostenedione was subsequently recrystallized to constant specific activity, providing unequivocal evidence of its synthesis by the equine placenta. 19-Norandrostenedione synthesis appeared to be localized in the microsomal fraction. Regardless of the substrate used, formation of 19-norandrogens was far weaker than that of oestrogens; moreover, the yield of 17-oxosteroids produced was much greater than that of 17 beta-hydroxysteroids, suggesting the presence of a dehydrogenase with predominant oxidative activity. Sulphoconjugated steroids formed were less than 0.5% of total steroids. Although 19-nortestosterone could not be generated by equine purified aromatase incubated with labelled testosterone, the synthesis of 19-norandrogens and oestrogens by equine placental explants was blocked by two specific aromatase inhibitors, 4-hydroxyandrostenedione and fadrozole. Our results provide evidence for a placental origin of at least a part of the 19-norandrogens previously identified in the blood of the pregnant mare. Furthermore, it is suggested that 19-norandrogen biosynthesis would involve the enzymatic metabolism of 19-oxygenated androgens formed by equine aromatase.

Aromatase activity in the mare ovary during estrous cycle. Measurement of endogenous steroids and of their in vitro inhibitory effect.

This present study was undertaken to clarify estrogen synthesis in the mare ovary. First of all, an evaluation of endogenous steroid contents was carried out in the follicular fluid and in the luteal tissue at different stages of the luteal phase. Radioimmunoassays were performed after separation and purification of each hormone by chromatography. High amounts of conjugated (0.9 mg/l) and unconjugated (4 mg/l) estradiol-17 beta were found in the follicular fluid of the large follicules (50 mm). These concentrations of estrogens decreased drastically in the luteal tissue, and only low levels of circulating estrogens are found during the luteal phase. On the other hand, a high aromatization ability has been evidenced in the cyclic corpus luteum in vitro. In an attempt to clarify the regulation of estrogen synthesis, we have tested the inhibitory effect of several endogenous steroids on equine ovarian aromatase activity. 5 alpha-Dihydrotestosterone appeared to be the most potent competitive inhibitor (Ki = 181 nmol/l) of aromatase activity, while the addition of a 3-sulfate group induced a slump in the inhibitory potency of estrone (Ki = 397 nmol/l vs 2206 nmol/l) and dehydroepiandrosterone (Ki = 291 nmol/l vs 6157 nmol/l). The physiological role of these conjugated steroids has not been known until now; we suggest that they would play a role in protecting aromatase from inhibition, in vivo. The high amounts of progesterone found in the luteal tissue (1.3 g/kg of proteins) might play a role in the regulation of estrogen production either by suppressing the induction of aromatase synthesis or by inhibiting the activity of the enzyme complex.

Different in vitro metabolism of 7 alpha-methyl-19-nortestosterone by human and equine aromatases.

The ability of human and equine placental microsomes to aromatize 7 alpha-methyl-19-nortestosterone (MNT) was studied. Kinetic analysis indicates that MNT shares the androgen-binding site of human and equine placental microsomal aromatases. Human placental microsomal estrogen synthetase had about a 2.5-fold higher relative affinity for MNT than the equine placental enzyme (KiMNT/Km androstenedione of 32 versus 87). However, MNT was not metabolized by human placental microsomes, whereas it was very actively metabolized by equine placental microsomes. Further studies using purified equine cytochrome P-450arom indicated that the presence of a 7 alpha-methyl group and the absence of a C19 methyl group did not impair its conversion by the purified enzyme. The product of this reaction was separated and identified as 7 alpha-methylestradiol by gas chromatography coupled to mass spectrometry.

Equine ovarian aromatase: evidence for a species specificity.

Mare granulosa cells and cyclic corpus luteum microsomes are reported to aromatize 19-norandrogens more efficiently than androgens. However, 16 alpha-hydroxytestosterone and epitestosterone were not aromatized by the equine corpus luteum microsomal estrogen synthetase. These results indicate that the equine aromatase system would be different from the human placental microsomal estrogen synthetase, which aromatizes 16 alpha-hydroxyandrogens and epitestosterone but not 19-norandrogens. Furthermore, our data show that the rates of aromatization of androgens and 19-norandrogens were not additive and that 19-norandrogens competitively inhibited the aromatization of androgens, suggesting that a single enzymic system would be involved in the aromatization of androstenedione, 19-norandrostenedione, testosterone, and 19-nortestosterone. Our findings, which are identical to those previously reported for stallion testis and mare placental estrogen synthetases, provide evidence for a strong species specificity of the equine aromatase system.

Binding of estrogen-3-sulfates to stallion plasma and equine serum albumin.

The binding of estrone-3-sulfate (E1-3-S) and estradiol-3-sulfate (E2-3-S) to adult stallion plasma was determined and compared with the binding to equine serum albumin (ESA). On the ESA molecule, two binding sites for E1-3-S with an association constant of 1.3 x 10(5) M-1 and several sites of weaker affinity were found; the data for E2-3-S showed the existence of four binding sites of moderate affinity (1 x 10(5) M-1) and several sites of weaker affinity. The removal of albumin from the stallion plasma resulted in the absence of binding of E1-3-S or E2-3-S, whereas the removal of glycoproteins resulted in binding parameters similar to those obtained with whole plasma. These results indicate that ESA is the only estrogen sulfate binder in horse plasma. Under physiological conditions, 95% of E1-3-S was bound to ESA.

Estrogen synthetase in the horse. Comparison of equine placental and rat liver NADPH-cytrochrome c (P-450) reductase activities.

NADPH-cytochrome c (P-450) reductases from horse placenta and rat liver were purified and their biological activities compared using cytochrome c as substrate. Rat liver reductase was purified to electrophoretic homogeneity in one chromatographic step on 2',5'-ADP agarose, and had a relative mass of 85,000 Da as estimated by SDS-PAGE. Equine placental reductase was separated from cytochrome P-450 on aminohexyl-Sepharose 4B and further purified on 2',5-ADP agarose; this preparation exhibited two bands, one of 85,000 and one of 80,000 Da, on SDS-PAGE. The lower molecular weight form was assumed to be a proteolytic product of the higher molecular weight form. A high retention of activity was obtained in both preparations. Equine placenta and rat liver enzymes were found to exhibit very similar Vmax and Km, suggesting that they are not species specific.

Concentration decrease of corticosteroid binding globulin (CBG) in plasma of the mare throughout pregnancy.

A significant decrease of CBG binding capacity in plasma of the mare throughout pregnancy was demonstrated using equilibrium dialysis and gel equilibration methods. As indicated with immunoelectrophoresis experiments, the pregnancy related fall of CBG binding capacity was linked to an actual decrease in blood CBG concentration. This result contrasts sharply with data on most other mammalian species, with the exception of the gestating rhesus monkey.

Androgen and 19-norandrogen aromatization by equine and human placental microsomes.

The ability of equine and human placental microsomes to aromatize testosterone and 19-nortestosterone was studied. When 3 microM [1 beta,2 beta-3H]testosterone was used as substrate, the specific activity of equine placental microsomal aromatase was 2.5 times higher than that of the human microsomal enzyme. Although 19-nortestosterone was aromatized 67 times more rapidly by equine than by human aromatase, we found that equine aromatase exhibited a markedly weaker affinity for this substrate than did the human enzyme. Competitive inhibition of testosterone aromatization by 19-nortestosterone occurred with both equine and human aromatases. While having no effect on mare placental microsomes, Na+ and K+ (500 mM) stimulated testosterone aromatization by human placental microsomes by 73 and 52% respectively. If indeed a single enzyme is responsible for the aromatization of testosterone and 19-nortestosterone, which seems to be the case in both equine and human placental aromatase, our results show that differences in the structure of the active sites exist between equine and human aromatases.

Distribution, composition and image analysis of plasma lipoproteins in steroid-treated calves.

Six 8-day-old female calves were treated with a subcutaneous implant of 200 mg testosterone + 20 mg estradiol-17 beta. Thirty-five days following implantation, plasma lipoproteins were compared to those in control calves of the same age. The LDL exhibited a slight change in protein and lipid concentrations and no change in particle size. The effects of steroid therapy on HDL and particularly on the lighter density HDL were characterized by a reduction of densities associated with a decrease in protein content, and by a rise in lipids and an increase in particle size. The changes in HDL composition but not in LDL alterations were consistent with those associated with sexual maturation described previously. Although testosterone is the predominant component of our combined preparation, the effects of our treatment on young female calves is not consistent with the data reported for human lipoproteinemia. The high levels of urinary estradiol in treated calves suggest that these effects result more likely from the aromatization of the injected testosterone.

Androgen synthesis and aromatization by equine corpus luteum microsomes.

Whereas mare corpus luteum does not produce androgens or estrogens in vivo, the incubation of mare corpus luteum microsomes with progesterone and NADPH resulted in 17 alpha-hydroxyprogesterone and estrogen production with a small yield of androstenedione. In the presence of an aromatase inhibitor (4-hydroxyandrostenedione), 17 alpha-hydroxyprogesterone and androstenedione were accumulated. Aromatization of testosterone and androstenedione occurred via stereospecific loss of the 1 beta, 2 beta hydrogen atoms and was inhibited by MgCl2, KCl, and EDTA. The Km of estrogen synthetase from equine corpus luteum for testosterone was 18.5 +/- 2.7 nM and for androstenedione was 11.5 +/- 1.5 nM. 19-Norandrogens were aromatized with a slightly higher efficiency than were androgens, but the affinity of the aromatase was lower for 19-norandrogens than for androgens. Our results suggest that aromatases from equine testis and corpus luteum are closely related enzymes. On the other hand, the question arises as to the relationship among the cell origin, the synthetizing abilities, and in vivo production of the corpus luteum in different mammalian species.

Synthesis and aromatization of 19-norandrogens in the stallion testis.

The results of the measurement of 19-nortestosterone in the testiscular artery and vein of the stallion, the very low levels of this steroid in the peripheral blood of geldings and the similar patterns of increase in the peripheral levels of 19-nortestosterone and testosterone after hCG stimulation, show that 19-nortestosterone, like testosterone, is essentially synthesized in the testis. This testicular origin was confirmed by the ability of testicular tissue to synthesize 19-norandrogens from [4-14C]androgens in vitro. 19-Nortestosterone was 50% conjugated in the peripheral blood and almost entirely conjugated after biosynthesis in vitro. The sequence of appearance of steroids in the peripheral blood after a single injection of 10,000 IU hCG suggests that, in the equine testis, 19-norandrogens are produced by a specific C10-19 desmolase (estrene synthetase), stimulable by hCG. 19-Nortestosterone was aromatized into estradiol-17 beta by stallion testicular microsomes. The affinity of the aromatase for 19-nortestosterone was very low compared to that for testosterone. At low and presumably physiological levels, and at a high testosterone/19-nortestosterone ratio, testosterone did not inhibit 19-nortestosterone aromatization by more than 53%. Thus, 19-nortestosterone may be aromatized in vivo in the testis in spite of the endogenous concentrations of androgens. However, the low velocity of 19-nortestosterone aromatization by testicular microsomes at roughly physiological concentrations suggests that 19-norandrogen aromatization may only participate slightly in the testicular estrogen production. These results suggest that in the equine testis, two aromatizing enzyme systems may exist: one which aromatizes both androgens and 19-norandrogens, and a minority system more specific for 19-norandrogens.

Androgen and oestrogen response to a single injection of hCG in cryptorchid horses.

Androgen (testosterone and androstenedione) and oestrogen (oestradiol -17 beta and oestrone) concentrations were measured by radio-immunoassay in the peripheral plasma of two geldings (five-years-old), three bilateral cryptorchids (two, two and a half, and five-years-old) and three normal intact stallions (four, five and five and a half-years-old) before and after a single injection of 10,000 iu human chorionic gonadotrophin (hCG). In the stallions, hCG administration resulted in an immediate sharp increase of conjugated oestrogens and a more gradual increase of unconjugated androgens. In the cryptorchids, the unconjugated androgens increased following a similar pattern to that observed in the stallions, but reached lower peak values, whereas the conjugated oestrogens showed only a very slight increase. In the stallions and cryptorchids, the maximum oestrogen levels were reached two days after injection, whereas the maximal levels for androgens were reached a day later. In the geldings, hCG injection had no effect on plasma steroid levels. It is suggested that the measurement of unconjugated androgens (testosterone or/and androstenedione) before and three days after intravenous injection of 10,000 iu hCG may prove useful for the diagnosis of cryptorchidism or exploration of testicular function in stallions.

Down-regulation of testicular aromatization in the horse.

A single i.m. injection of testosterone (750 mg of testosterone bexahydrobenzoate) or i.v. injection of human chorionic gonadotrophin (hCG) (10,000 IU) was given to geldings and stallions. Levels of unconjugated and conjugated (after solvolysis) androgens and estrogens were measured in blood and urine samples taken daily from the day of injection (D0) to the tenth day post-injection (D10). In the stallion, both treatments resulted in a sharp increase of plasma estrogens, which peaked one day before the androgen levels. Our results confirmed the testicular localization of a potent aromatase, which is able to aromatize androgens from endogenous as well as exogenous origin into conjugated estrogens. The very similar patterns of estrogen increase following testosterone or hCG administration suggest that the estrogen rise induced by hCG results at least partly from increased availability of testosterone. The abrupt drop in plasma estrogen levels cannot be explained by a lack of substrate, since two successive androgen injections did not succeed in maintaining the high estrogen levels. Since estrogens were unable to inhibit the aromatase activity in vitro, the drop in estrogen levels suggests a down-regulation of the aromatase synthesis.

Age-related morphological and functional changes in the Leydig cells of the horse.

Two ultrastructurally distinct types of Leydig cells were observed in the equine testis. Whereas the adult testis exhibited both postpubertal and adult Leydig cells, the testis of the pubertal horse contained only the postpubertal type, and that of the aged horse contained only the adult type. However, Percoll-purified testicular preparations from pubertal, adult, and aged horses all exhibited two distinct Leydig cell populations. The quantitative distribution and the functional characteristics of these Leydig cell populations (ability to bind human chorionic gonadotropin [hCG] and increase of testosterone production after hCG stimulation) evolved with the age of the horse. It is concluded that equine Leydig cells derive from two redundant successive postnatal generations and that there is no strict correlation between the functional properties and the morphological characteristics of these cells.

Aromatization of testosterone and 19-nortestosterone by a single enzyme from equine testicular microsomes. Differences from human placental aromatase.

A single enzyme in the stallion testis was able to aromatize both testosterone and nortestosterone. This enzyme had a much lower affinity for nortestosterone than for testosterone. In contrast to human placental estrogen synthetase, this enzyme aromatized testosterone and 19-nortestosterone with similar efficiency. The differences observed (effects of monovalent cations, inhibition of androstenedione aromatization by testosterone and 19-nortestosterone and, above all, rate of norandrogen aromatization) suggest that the aromatase in the horse testis is not the same as that in the human placenta.