The insulin-like growth factor system in human peritoneal fluid: Its effects on endometrial stromal cells and its potential relevance to endometriosis

Peritoneal fluid (PF) lines the abdomen and pelvis and is believed to contain growth factors that stimulate endometriosis, a benign gynecological condition associated with pelvic pain and infertility, in which endometrial cells proliferate and differentiate on the pelvic peritoneum, outside of their normal location within the uterus. In this study, we examined the insulin-like growth factor (IGF) system in seven paired samples of PF and serum from normally cycling women and examined the mitogenic potential of this fluid on cultured endometrial stromal cells. IGF-I, IGF-II, and IGF- binding protein-1 (IGFBP-1), -2, -3, and -4 were identified in PF by immunoassays. PF IGF levels, determined by RIA, were approximately 60% of paired serum levels, and PF levels of IGFBP-2 and IGFBP-3, determined by Western ligand blotting and RIA, respectively, were approximately half of their serum concentrations. IGFBP-4 was barely detectable by Western ligand blotting in PF, and levels of IGFBP-1, determined by immunoassay, were not appreciably different in PF and serum. Incubation of [¹²⁵I]IGF-II with serum and PF and subsequent size-exclusion chromatography at neutral pH revealed approximately equal incorporation of radiolabel in the IGFBP regions of 150 and 44 kilodaltons (kDa) in serum and primarily in the 44-kDa region in PF. RIA of IGFBP-3 in the IGFBP regions of column effluent revealed that the majority of IGFBP-3 was in the 150-kDa region in both serum and PF, suggesting the presence of the ternary complex in PF. Western ligand blotting of column effluent samples revealed 37-/43-kDa IGFBP-3 primarily in the 150- kDa complex in serum and a marked reduction in the amount of the 37-/43-kDa IGFBP in PF. Western immunoblotting of column effluent with IGFBP-3 antiserum revealed immunoreactive IGFBP-3 forms of 37-43 kDa (major) and 28 kDa (minor) in serum and almost exclusively the 28-kDa band in PF, suggesting that IGFBP- 3 in PF may be proteolytically processed. The presence of an IGFBP-3 protease was confirmed using [¹²⁵I]IGFBP-3 as substrate and was not appreciably present in paired serum samples. Inhibitor profiles demonstrated that this protease is a metal-independent serine protease, and its approximate relative molecular mass was estimated to be 69 kDa, determined by size-exclusion chromatography. The mitogenic potential of IGF peptides and PF was assessed on cultured endometrial stromal cells to test the hypothesis that IGFs in PF may stimulate the growth of endometrium in the pelvic cavity, for example in the disorder of endometriosis. IGF-I and IGF-II were mitogenic to endometrial stromal cells, and maximum growth stimulation occurred at 10 and 50 ng/mL, respectively. PF was also mitogenic to endometrial stromal cells in a dose- dependent fashion, and this mitogenic effect was inhibited up to 30% in the presence of αIR3, a blocking antibody to the type IIGF receptor. These data demonstrate that the IGF system (IGF peptides, IGFBPs, and an IGFBP protease) is present in human PF. Furthermore, they suggest that the IGF system may be one of several growth factor systems in PF that has the capacity to stimulate endometrial cellular proliferation and may participate in the growth of ectopic endometrium on the pelvic peritoneum, as in the disorder of endometriosis.