There is evidence that insulin-like growth factor binding protein-3 (IGFBP-3) is a part of the intrinsic IGF system in rat CL. Here, we examined when during luteogenesis the IGFBP-3 gene is expressed. IGFBP-3 messenger RNA (mRNA) was characterized by Northern analysis and in situ hybridization techniques. Animals were selected at proestrus (P 1000 h and P 2000 h), estrus (E 0200 h and E 1000 h), diestrus I (DI 1100 h), and diestrus II (DII 1100 h), and in pregnancy (day 12, before luteolysis and day 21, during luteolysis). A single 2.6 kilobase IGFBP-3 transcript was identified at each stage of the estrous cycle; however, the amounts of message varied markedly, being most abundant at P 1000 h, least abundant at P 2000 h, E 0200 h, E 1000 h, and DI, and then more abundant at DII. Corroborating our earlier report, IGFBP-3 mRNA was limited solely to corpora lutea (CL). Newly-formed CL-I at E 0200 h and E 1000 h revealed no IGFBP-3 hybridization. This is the period of early luteinization when cells undergo hypertrophy and capillaries and lymphatics penetrate the granulosa lutein layer. At DI 1100 h, a few cells (12.2 +/- 3.4%) near the central cavity of the CL-I showed a positive hybridization signal for IGFBP-3; this period is commensurate with active luteinization when the vascular tissue develops a distinctly sinusoidal character and progesterone secretion by CL-I increases. At DII 1100 h, more cells in the central area were positive for IGFBP-3 (55.2 +/- 6.4%); this is the period of active luteolysis when P4 secretion has fallen to basal levels. At P 1000 h, a positive IGFBP-3 hybridization signal was detected in most CL-1 cells (85.3 +/- 2.8%), and the signal was particularly intense in subtypes of endothelial cells lining venous sinusoids and/or lymphatics and some perivascular cells; this is the period when patches of pyknotic cells appear in the central area of CL-I. At P 2000 h, 45.9 +/- 1.6% of the CL-I cells showed a positive signal; however, the intensity of the signal was much weaker when compared to P 1000 h. During the next cycle, the CL-I become the CL of the second generation (CL-II), which show increased necrosis. Between estrus and diestrus II of the next cycle, a large number of the CL-II cells (approximately 75%) were positive for IGFBP-3 and the signal was very strong in some groups of cells. By P 1000 h of the next cycle, IGFBP-3 mRNA was not detected in CL-II. During the next cycle, CL-II become third generation CL (CL-III). No hybridization signal was detected in atropic CL-III. Active CL-I during mid-pregnancy revealed no hybridization signal; however, a moderate IGFBP-3 signal was detected in the CL-I at parturition. These results show that IGFBP-3 mRNA expression in adult rat ovaries is restricted to involuting corpora lutea of the cycle and pregnancy. Most of the IGFBP-3 is concentrated in subtypes of vascular (sinusoidal and/or lymphatic) and perivascular (resident macrophages?) cells, rather than in the lutein cells themselves, which show moderate or weak activity. It is possible that intrinsic IGFBP-3 may play a role in the normal mechanisms of luteolysis of rat CL, perhaps by mechanisms involving vascular elements.
