Insulin-like growth factor binding proteins in sera of pregnant nonhuman primates

Insulin-like growth factors (IGFs) are mitogenic peptides that are important for fetal and maternal tissue growth during pregnancy. They circulate complexed primarily with a serum binding protein, IGFBP-3, which regulates the availability of the IGFs to their target tissues. We previously reported that in pregnant women, serum IGFBP-3 levels, assessed by Western ligand blotting, decline markedly beginning at 6 weeks gestation due to a circulating protease that cleaves IGFBP-3 into a 29-kilodalton (kDa) protein and lower mol wt (M_r) fragments. In the current study, we compared IGFBP profiles, IGFBP-3 and IGFBP-1 levels, and IGFBP protease activities in sera from pregnant and nonpregnant women, baboons, and rhesus monkeys, using Western ligand blotting, IGFBP-specific immunoassays, IGFBP-3 protease assay, and zymographic gel electrophoresis. Serum IGFBP profiles in nonpregnant human and nonhuman primates were similar and were not cycle-dependent. IGFBP-3 (37-43 kDa), IGFBP-2 (31kDa), and IGFBP-1 (28 kDa) were identified in all three species using IGFBP-specific human antisera. A 24-kDa IGFBP was also present and is believed to be IGFBP-4. Serum IGFBP-1 levels increased throughout gestation in human and nonhuman primates. Serum IGFBP-2 and putative IGFBP-4 were barely detectable in all three species from midgestation to term, but increased several days postpartum. In contrast, serum IGFBP-3 profiles differed markedly between species during gestation. Rather than the decrease seen in human pregnancy serum, there was an increase in circulating IGFBP-3 levels in nonhuman primates. Furthermore, for both baboon and rhesus monkey, the M_r of serum IGFBP-3 was about 2 kDa greater in pregnant than in nonpregnant animals, and deglycosylation studies suggested that the higher M_r forms may be alternatively glycosylated or may have a unique primary structure. As in nonpregnant women, serum IGFBP-3 protease activity was absent in nonpregnant and pregnant baboons. However, rhesus monkey serum contained a calcium-dependent protease that cleaved recombinant human IGFBP-3 into unique fragments, compared to the human pregnancy enzyme. Unlike human pregnancy serum, which proteolyzes IGFBP-3, in human nonpregnancy serum, rhesus serum incubated under similar conditions did not result in proteolysis of rhesus IGFBP-3, suggesting that the IGFBP-3 protease in human pregnancy serum is not present in the circulation of the rhesus monkey. To assess proteolytic activity in these sera, zymographic polyacrylamide gel analysis, using gelatin as a substrate, was performed. A minor band of proteolytic activity (72 kDa) was observed in all three species throughout gestation. However, in human pregnancy serum, a major proteolytic band (96 kDa) was evident early in gestation, rose markedly by 18 weeks, was calcium dependent, and was inhibited by EDTA and phenanthroline, suggesting a tentative identification of this protease as the circulating IGFBP-3 protease. Minor bands of proteolysis at 72 kDa in nonhuman primate sera and 96 kDa in rhesus monkey sera were also calcium-dependent and inhibited by EDTA and phenanthroline. Aprotinin and soybean trypsin inhibitor did not inhibit proteolytic activity in any of the sera. We propose that proteolysis of IGFBP-3 during human pregnancy, believed to be by a protease of decidual origin, may provide a mechanism to enhance the bioavailability of IGFs during human gestation. Although phylogenetically related to humans, nonhuman primates have highly divergent mechanisms of implantation and decidualization and may have developed alternate mechanisms for cellular targeting of the IGFs during pregnancy.