Increased Proteolysis of Insulin-Like Growth Factor-Binding Protein-3 (IGFBP-3) in Noninsulin-Dependent Diabetes Mellitus Serum, with Elevation of a 29-Kilodalton (kDa) Glycosylated IGFBP-3 Fragment Contained in the Approximately 130- To 150-kDa Ternary Complex

Insulin-like growth factor-I (IGF-I) in serum is predominantly bound in a ternary complex, consisting of IGF peptide, IGF-binding protein-3 (IGFBP-3), and an acid-labile subunit, or a binary complex, consisting of IGF peptide and any of the six IGFBPs. In the binary complex, IGF-I is more bioavailable and has a faster turnover rate. Proteolysis of IGFBP-3 may alter the distribution of IGF-I between these complexes by reducing IGFBP-3 affinity for IGF-I and/or acid-labile subunit and may offer an additional mechanism for regulation of IGF availability. In the present study, sera from patients with noninsulin-dependent diabetes mellitus (NIDDM) were found to have significantly higher IGFBP-3 proteolytic activity than sera from age-matched healthy subjects (188 ± 12% vs. 104 ± 6% of a control serum pool; P < 0.001). The mean (±SE) of serum IGFBP-3 levels determined by Western ligand blotting was lower in NIDDM patients than in healthy control subjects (61.5 ± 5% and 79 ± 5% of a control serum pool, respectively; P < 0.01). However, IGFBP-3 concentrations determined by RIA did not differ. This discrepancy could be explained by IGFBP-3 proteolysis, resulting in IGFBP-3 fragments that are detectable by RIA, but not by Western ligand blotting. Western immunoblotting of sera with or without prior treatment with endoglycosidase-F demonstrated that a glycosylated 29-kilodalton (kDa) IGFBP-3 form with a protein core of 20 kDa was present in sera from healthy controls, and this fragment was increased in NIDDM and term pregnancy sera, suggesting that it is produced by endogenous proteolysis. The presence of the 29-kDa IGFBP-3 proteolytic fragment at about 130-150 kDa after neutral size chromatography of pooled sera may suggest that 29-kDa IGFBP-3 participates in ternary complex formation. Further studies are required to determine whether the avidity of ternary complex formation with the 29-kDa IGFBP-3 fragment is reduced and whether the resulting increased IGF turnover can explain the reduced IGF-I levels (z scores) observed in NIDDM patients compared to healthy subjects (-0.81 ± 0.32 SD vs. +0.26 ± 0.17 SD; P < 0.001). Neutral size-chromatography of sera demonstrated that IGFBP-3 protease activity in the approximately 130- to 150-kDa mol wt range is regulated by NIDDM and pregnancy in parallel with that of unfractionated sera. In addition, all sera tested, including sera from healthy subjects, demonstrated high IGFBP-3 proteolytic activity in the approximately 40- to 50-kDa mol wt range. It may be hypothesized that lack of an inhibitor of the approximately 40- to 50-kDa protease(s), which is normally present, may contribute to increased IGFBP-3 protease activity in NIDDM and term pregnancy sera. The increased IGFBP-3 proteolytic activity in NIDDM serum was partially cation dependent and inhibited by serine protease inhibitors, similar to that of term pregnancy sera.