A novel method for purifying dispersed porcine theca cells, with less than 3% granulosa cell contamination, was developed by the repeated use of mechanical and enzymatic procedures. The steroidogenic criteria used for the identification and purity evaluation of both theca and granulosa cells were also improved. Purified theca and granulosa cells from medium-sized follicles displayed steroidogenic differences when they were cultured in the presence of 10% fetal bovine serum: (1) the theca cells synthesized oestradiol (239.1 ± 35.1 pg ml⁻¹ per 2.5 × 10⁵ cells in 40 h), but the granulosa cells did not synthesize it unless aromatizable androgens were added; (2) theca cells synthesized androstenedione (73.2 ± 14.4 ng ml⁻¹ per 2.5 × 10⁵ cells in 40 h), but granulosa cells did not; (3) FSH did not affect progesterone production in theca cells; (4) the theca cells secreted androstenedione for up to 48 h; and (5) FSH significantly stimulated progesterone production in granulosa cells during a culture for 40 h (P < 0.05), but not during culture for 12 h. The lack of response to FSH was used as a reliable, functional indicator of the purity of porcine theca cells. However, this criterion proved not to be useful for cells cultured for 12 h; porcine FSH had no effect on the progesterone production of theca cells co-cultured for this time with as many as 20% granulosa cells. However, after co-culturing for 40 h, this criterion resulted in the detection of only 3% granulosa cell contamination. Lack of response to FSH is a sensitive and reliable criterion for evaluating the purity of porcine theca cells, as long as FSH responsiveness of granulosa cells is fully confirmed.

The activity of GnSI/AF was measured in pig follicular fluid (pFF) from 58 individual follicles of various sizes, by bioassay using rat pituitary cells, to investigate the relationship between gonadotrophin surge inhibiting/attenuating factor (GnSI/AF) activity and follicular development. In addition, the correlation between GnSI/AF and inhibin activities and the content of sex steroids (oestradiol, progesterone and testosterone) of follicles was examined. The activity of GnSI/AF in pFF varied significantly (0.155–1.69 U μl⁻¹) with size of the follicle. The activities (mean ± sem) were intermediate and constant in follicles with diameters from 3 to 5 mm (0.583 ± 0.080 U μl⁻¹, n = 24), were higher and reached the highest value in follicles with diameters between 6 and 8 mm (0.863 ± 0.068 U μl⁻¹, n = 21), and were lower, reaching the lowest value in follicles with diameters of 9 and 10 mm (0.401 ± 0.089 U μl⁻¹, n = 13). In contrast, inhibin activity was almost constant during the development of follicles, although individual values varied from 0.9 to 2.5 U μl⁻¹. For follicles with diameters of 4–8 mm, inhibin activity was 1.754 ± 0.042 U μl⁻¹ (n = 39); activity was higher in the smallest follicles with diameters of 3 mm (2.063 ± 0.015 U μl⁻¹, n = 6) and was lower in follicles with diameter of 9 mm, reaching the lowest value in follicles with diameter of 10 mm (1.176 ± 0.068 U μl⁻¹, n = 7); inhibin activity was not significantly correlated with GnSI/AF activity. Steroid concentration increased in a similar pattern to that of GnSI/AF activity, but no marked decrease was noticed in the large follicles. These characteristic changes of the activity of GnSI/AF of follicles during development, especially in the preovulatory stage, suggest that GnSI/AF plays a role in the ovulatory process as an ovarian regulator for the timely occurrence of the LH surge.

The effect of intraoviductal embryos on endometrial receptivity was studied by intraendometrial and intrauterine embryo transfer. Five-week-old female ICR mice were mated after superovulation; a vaginal plug confirmed day 1 of pregnancy. On day 4 (90 h after hCG injection), blastocysts were collected and transferred to pseudopregnant female mice and to recipient mice in which the uterotubal junction had been ligated bilaterally on day 1 of pregnancy. Three embryos per uterine horn, a total of six embryos per recipient mouse at days 1–6, were transferred to the endometrium or uterine cavity and implantation and pregnancy rates were calculated. The implantation rate for intraendometrial embryo transfer to recipients of days 3, 5 and 6 was significantly higher for uterotubal junction-ligated mice (72.2, 20.8 and 9.7%, respectively) than for pseudopregnant mice (55.0, 8.3 and 0.0%, respectively). The implantation rate for intrauterine embryo transfer to recipients at days 2, 5 and 6 was significantly higher for uterotubal junction-ligated mice (11.1, 25.0 and 8.3%, respectively) than for pseudopregnant mice (0.0, 3.3 and 0.0%, respectively). Uterotubal junction-ligated mice achieved implantation and bore neonates by intrauterine embryo transfer on days 2 and 6, whereas no implantation was achieved in pseudopregnant mice. The difference in implantation rate could not be explained by a difference in progesterone concentration between the groups. The distribution of proliferating cells in the endometrium was also studied immunohistochemically by use of anti-proliferating cell nuclear antigen (PCNA) antibody in the recipient mice. PCNA-positive cells were more abundant in uterotubal junction-ligated mice and demonstrated a marked extension from the epithelium to the stroma over time, in contrast to those in pseudopregnant mice. These findings indicate that an intraoviductal embryo exerts a biological effect by sending a signal to the endometrial epithelium and stroma, thus facilitating endometrial receptivity to the embryo and improving the rate of implantation.

The objective was the immunocytochemical localization of steroidogenic enzymes in the corpus luteum of Hokkaido brown bears during the period of delayed implantation. Cholesterol side-chain cleavage cytochrome P450 (P450scc), 3β-hydroxysteroid dehydrogenase (3βHSD), 17α-hydroxylase cytochrome P450 (P450c17) and aromatase cytochrome P450 (P450arom) were localized as biosynthetic sites of pregnenolone, progesterone, androgens, and oestrogens, respectively. Ovaries containing corpora lutea were obtained from three mature bears during the expected delayed implantation period and ovarian sections were immunostained by the avidin–biotin–peroxidase complex method using polyclonal antibodies generated against steroidogenic enzymes of mammalian origin. P450scc and 3βHSD were localized in all luteal cells, whereas P450c17 (0.4–5.1% of 1000 cells) and P450arom (7.1–11.2% of 1000 cells) were localized in only a few luteal cells. These data suggest that luteal cells contain steroidogenic enzymes required for progesterone synthesis but also have a minimum capability for synthesizing androgen and oestrogen during the delayed implantation period in Hokkaido brown bears.