Porcine ovary samples were prepared for histochemical and ultrastructural analyses. In situ analysis of DNA fragmentation was performed on histological sections of follieles using the terminal deoxynucleotidyl transferase-mediated biotinylated deoxyuridine triphosphate nick end-labeling TUNEL) method. No apoptotic cells were observed in healthy follicles. In atretic follicles, apoptotic TUNEL staining was seen in scattered granulosa cells located on the inner surface of the follicular wall, but not in cumulus cells, internal or external theca cells, or oocytes. Nuclear condensation, a typical feature of apoptosis was seen only in scattered granulosa cells. The neutral Ca2+/Mg2+-dependent endonuclease is involved in granulosa cell apoptosis. No endonuclease activity was detected in cumulus cells. An IgM monoclonal antibody (PFG-1) capable of inducing granulosa cell apoptosis was then produced against granulosa cells prepared from healthy antral follicles. Two-dimensional (2D) Western blotting analysis revealed that PFG-1 specifically recognized a cell-membrane protein (PFG-1 antigen, 55 kD, pl 5.9). PFG-1 immunohistochemically reacted with granulosa cells of healthy follicles but not those of atretic follicles. When the isolated granulosa cells prepared from healthy follicles were cultured in medium containing 0.1 μg/ml of PFG-1, the cells underwent apoptosis. These observations indicated that apoptosis occurs in granulosa cells but not cumulus cells in the atretic follicles and that the PFG-1 antigen, a novel cell death receptor, is different from the apoptosis-mediating receptors Fas antigen or tumor neerosis factor receptor 1 (TNFR-1).

We prepared an IgM monoclonal (PAG-1) antibody against follicular granulosa cells of porcine ovaries. Immunohistochemical reaction of the antibody was detected only in follicular granulosa cells, not in any other ovarian tissues or organs. Western blotting analysis demonstrated that 38- and 42-kD bands were present in the granulosa cell membrane samples prepared from healthy follicles. In the samples from early atretic and progressed atretic follicles, 38- and 64-kD bands were shown. During follicular atresia, the 64-kD band increased. After glycosidase treatment, the 64-kD band shifted to 42-kD, but no change was seen in the 38- and 42-kD bands. The present results indicate that the 42-kD band membrane protein is glucoconjugated during atresia, resulting in a glycoconjugated 64-kD protein appearing in the granulosa cells of the atretic follicles. We hypothesize that the glycoconjugated cell membrane protein (64-kD glycoprotein) may act as a trigger for phagocytosis in neighboring granulosa cells in atretic follicles.

The sugar chains in cellular glycoconjugates have many biological functions. Extensive morphological development and remodeling occur in the ovary of female animals. This caused us to study glycobiological characteristics of ovarian cells, particularly granulosa cells that undergo apoptosis during follicular atresia. The lectin Sambucus sieboldiana agglutinin (SSA) specific for Sia alpha 2,6Gal/GalNAc showed positive staining for granulosa cells only in atretic follicles of porcine ovaries by lectin histochemistry. Lectin blot analysis for SSA demonstrated specific glycoproteins only in atretic follicles. Furthermore, we performed analysis of backbone structures of SSA-positive glycans carried by granulosa cell glycoproteins increased during atresia by glycosidase treatment. Most of these structures were Sia alpha 2,6Gal beta 1,4GlcNAc on complex-type N-glycans, suggesting that only ST6Gal I of four distinct alpha 2,6-sialyltransferases catalyzes alpha 2,6-sialic acid transfer in most of the increased glycoproteins of granulosa cells during follicular atresia. Reverse transcription-polymerase chain reaction analysis demonstrated that the expression of ST6Gal I mRNA was up-regulated in granulosa cells during atresia. These results suggested that the alteration of glycoconjugates by ST6Gal I in granulosa cells during atresia is involved in some processes of ovarian follicular atresia and granulosa cell apoptosis.

To determine exactly when progesterone (P) acts as a most important mediator in the ovulatory process, a dose of 10mg/kg of RU486, an antiprogesterone, was administered to PMSG/hCG treated immature female rats (22 days old) at 0 (RU0 groups), 2 (RU2 group), 4,6,8 or 10 hours after an hCG injection, respectively. In the RU0 groups, the ovulatory effect of P was investigated at 0.2 (P2 group), 4 (P4 group), 6 or 8 hours after the hCG injection. Ovulation rates were calculated 24 hours after the hCG injections. Serum P and estradiol (E2) concentrations in the RU0 groups, the P2 group and the control rats (C group), were determined. The results were as follows. 1) The numbers of ova in the RU0, RU2 and RU4 group rats were significantly (p less than 0.01, p less than 0.01 and p less than 0.01) lower than that in the C group rats. 2) In the P2 and P4 groups, the numbers of ova were significantly (p less than 0.01 and p less than 0.01) increased compared with the RU0 group and returned to the control range. 3) In the RU0 group, the serum P and E2 concentrations within 8 hours after an hCG injection did not show any significant differences compared with the C group. In contrast, in the P2 group, the P concentrations at 4 and 6 hours after hCG increased significantly (p less than 0.01 and p less than 0.01) compared to the corresponding RU0 group. These results clarified an indispensable role of P in the ovulatory process within 6 hours after an hCG injection.

In an effort to determine the reliability of the midluteal progesterone(P)/estradiol(E2) ratio as an index of the potential for conception, we measured the midluteal P and E2 levels in 76 infertile women who had been treated at our infertility clinic. This parameter in conception cycles was compared with that in non-conception cycles. Eighty cycles of the 76 women were classified into two groups, depending upon whether pregnancy occurred or not. Group 1 and group 2 were composed of 31 conception cycles and 49 drug-induced cycles, respectively. Midluteal concentrations of P and E2 did not show any significant differences between the two groups. The P/E2 ratios were 106.4 +/- 71.3 (mean +/- SD) and 71.5 +/- 44.16, respectively. This difference was statistically significant (p less than 0.05). Discriminant function analysis showed that the smallest probability of misclassification between the two groups decreased from 44% when using P only to 36% when using the combined P and E2. The following equation: Y = 0.0727X1 - 0.00456X2 - 0.130 was obtained (when X1 = P, X2 = E2). These results suggest that the midluteal P/E2 ratio gives clinicians the best indication of luteal function for the achievement of pregnancy.

To elucidate the effect of the antiprogesterone steroid RU 486 (RU) on endometrial glycogen metabolism, a dose of 30 mg/kg of the agent was administered to pregnant rats on Day 2 (group 1) or Day 4 (group 2) of pregnancy, and glycogen content, glycogen synthetase (GS) and glycogen phosphorylase (GP) activities in the endometrium were investigated on Day 6. In addition, serum ovarian steroid hormones and the number of implantation sites were evaluated. The glycogen content in the endometrium in group 1 and group 2 decreased significantly (p less than 0.02 and p less than 0.05) compared with the control group. The total GS activity in the endometrium in group 1 decreased significantly (p less than 0.05). On the other hand, the independent GS activity in group 2 increased significantly (p less than 0.05) compared with the control group. The active GP activity in group 1 increased markedly, and the total GP activity increased significantly (p less than 0.05) compared with the control value. In group 2, the active and total GP activities increased significantly (p less than 0.01 and p less than 0.02) compared with the control values. Serum progesterone (P) concentrations in group 1 and group 2 were significantly (p less than 0.05 and p less than 0.05) lower than those in the control groups 2 days after the RU administration, and implantations were significantly (p less than 0.01 and p less than 0.01) inhibited. These results suggest that RU 486 affects endometrial glycogen metabolism, resulting in the prevention of implantation.

A dose of 30 mg RU 486/kg, an antiprogesterone, was administered to pregnant rats on Day 2 (Group 1) or Day 4 (Group 2) of pregnancy. RU 486 significantly changed serum progesterone and oestradiol concentrations and luteal 3 beta-HSD and 20 alpha-HSD activities in Group 1, and implantation was significantly inhibited. The luteal 3 beta-HSD activity in Group 2 rats on Day 6 was significantly (P less than 0.01) lower than the control value (7.5 +/- 0.6 and 10.1 +/- 0.6 mU/mg protein respectively). This decline in the 3 beta-HSD activity was followed by a marked decrease in the serum progesterone concentration, resulting in a significant decrease of the progesterone/oestradiol ratio and implantation was completely inhibited. The 20 alpha-HSD activity, which could not be detected on Day 6 in the control rats, was twice as great in Group 2 than in Group 1 rats (17.5 +/- 1.2 and 7.4 +/- 3.1 mU/mg protein respectively). Ultrastructural examination of corpora lutea of Group 2 rats confirmed luteolysis. These results suggest that RU 486 has a luteolytic effect and its anti-implantation effect is concomitant with luteolysis of the corpora lutea of pregnancy.

Serial serum determinations of α-fetoprotein (αFP) as a tumor marker were carried out in the management of six patients with endodermal sinus tumor (EST). Histological examinations in all six patients revealed a typical EST pattern, and in one of them another germ cell tumor, a malignant teratoma, was also found. All patients were treated postoperatively with combination chemotherapy. The serum αFP concentrations before treatment, using radioimmunoasssay, were abnormally high and ranged from 2500 to 100,000 ng/ml. One patient having Stage Ia neoplasm is living at 50 months after diagnosis with a normal αFP concentration and without clinical evidence of recurrence. The other five patients with Stage III disease died from 4 to 9 months after surgery, respectively, although the serum αFP in all these cases during chemotherapy decreased markedly but temporarily to a normal level (less than 20 ng/ml) or to 3000 ng/ml and was paralleled by a certain improvement in the conditions of the disease. These results suggest that serial serum determination of αFP may be useful as a marker and prognostic indicator of endodermal sinus tumor.

Tamoxifen at a dose of 10 mg/day for 5 days was given to five infertile women in the luteal phase. Daily serum samples were obtained during the luteal phase for radioimmunoassay of progesterone (P), estradiol (E2), follicle-stimulating hormone, luteinizing hormone (LH), and prolactin levels. The integrated luteal phase concentrations of serum P and E2 before and after cycles of tamoxifen treatment increased from 87.8 +/- 16.2 ng/ml and 1120 +/- 164.4 pg/ml to 131.6 +/- 18.9 ng/ml and 1461 +/- 205.2 pg/ml, respectively (P less than 0.01 and P less than 0.05). No apparent increase in circulating LH levels was seen in one of the five cases, but this patient's serum P and E2 levels rose nonetheless. This suggests that the significant increase in circulating P and E2 induced by tamoxifen is not consistently associated with an increase in serum LH concentration.

Twenty-five infertile women conceived while taking tamoxifen (TMX). Daily serum profiles of 5 of the 25 TMX-induced conception cycles were elucidated and compared with those found in 5 normal cycles. In spite of lower levels of follicle-stimulating hormone and luteinizing hormone during the follicular phases, estradiol concentrations were higher in the TMX-induced conception cycles. It is suggested that this may be due to a direct ovarian effect of TMX as one of its major mechanisms in the course of folliculogenesis. On the other hand, progesterone concentrations on days 6 and 7 during the luteal phases were also higher in the TMX-induced conception cycles. It is suggested that this may be due to a luteotropic influence at the blastocyst stage.

A group of 22 women with anovulatory cycles was treated with tamoxifen in 46 cycles. We elucidated daily hormone profiles in 5 of the 33 tamoxifen-induced ovulatory cycles in order to make a comparison with those found in 5 normal cycles. Serum follicle-stimulating hormone and luteinizing hormone (LH) rose during or a few days after the tamoxifen therapy, and thereafter serum estradiol levels rose gradually over the normal ranges during the follicular phase and reached an extremely high preovulatory peak which triggered an LH surge. This increased estradiol production exerted its usual negative feedback on LH secretion. Following ovulation, serum progesterone rose and fell in a manner similar to that in the normal cycles, with occasional values that exceeded the normal range. The endocrine significance of these findings is discussed.

A group of 17 patients with suspected luteal phase deficiency was treated with tamoxifen. Tamoxifen therapy was found to lengthen the luteal phase in all patients and resulted in pregnancy in 6 of 17 patients. The integrated luteal phase progesterone (P) concentration in the nontreatment cycle of seven patients was significantly lower (P less than 0.01) than that of five normal women. Therapy with tamoxifen increased the P concentration to 186.0 +/- 24.4 ng/ml/cycle (mean +/- standard error of the mean), i.e., twice that of the control cycle. The mean estradiol (E2) concentration at the midcycle peak was about twice that observed during the nontreatment cycle. The glycogen content of the endometrial tissue at the midluteal phase in the tamoxifen cycle was significantly higher (P less than 0.025) than that of endometrial tissue in the nontreatment cycle, indicating improvement of the endometrial function.