PROTEASE-INDUCED ALTERATION OF INSULIN-LIKE GROWTH-FACTOR BINDING PROTEIN-3 AS DETECTED BY RADIOIMMUNOASSAY - AGREEMENT WITH LIGAND BLOTTING DATA

Citation
C. Lassarre et al., PROTEASE-INDUCED ALTERATION OF INSULIN-LIKE GROWTH-FACTOR BINDING PROTEIN-3 AS DETECTED BY RADIOIMMUNOASSAY - AGREEMENT WITH LIGAND BLOTTING DATA, Growth regulation, 4(2), 1994, pp. 48-55
Citations number
26
Categorie Soggetti
Endocrynology & Metabolism
Journal title
ISSN journal
0956523X
Volume
4
Issue
2
Year of publication
1994
Pages
48 - 55
Database
ISI
SICI code
0956-523X(1994)4:2<48:PAOIGB>2.0.ZU;2-G
Abstract
Structural alteration of insulin-like growth factor binding protein-3 (IGFBP-3) resulting from limited proteolysis by one or more serine pro teases in vivo was first described in the serum of pregnant women and in certain pathological conditions. Western immunoblotting has since b een employed to detect the phenomenon in normal serum, using a polyclo nal antibody raised against recombinant human IGFBP-3 and a highly sen sitive technique of visualization by chemiluminescence. The major prot eolytic fragment of 30 kDa, which fails to be detected in native serum by ligand blotting owing to its weak affinity for IGFs, has proved cl early visible in all serum samples tested, sometimes accompanied by sm aller fragments of 20 and 16 kDa. Among the serum samples analysed, in creasing proportions of proteolysed IGFBP-3 were found in the followin g order: acromegalic patients, normal subjects, GH-deficient patients, pregnant women. In RIAs done with the same antibody, many of the seru m samples yielded dose-response curves which,were not parallel with st andard curves, with lower gradients. In the samples where measurements were possible, apparent IGFBP-3 levels proved lower in pregnant women (2.28+/-0.23mg/l, mean+/-SER) than in normal adults (4.26+/-0.33 mg/l , P<0.001). These observations, which contradict earlier reports of hi gher levels in pregnant women, suggest that the 30 kDa proteolytic fra gment has a weaker affinity for the antibody than the intact IGFBP-3 ( which in ligand- and immunoblotting appears as a characteristic 42-39 kDa doublet and which is barely or not detectable in pregnancy serum). The diminished affinity of the 30 kDa IGFBP-3 fragment was confirmed by an experiment incubating pregnancy serum with normal serum at 37 de grees C. Here, the progressive proteolysis of intact IGFBP-3 in normal serum coincided with the gradual decrease in immunoassayable IGFBP-3 concentrations. Conversely, in a women after delivery and in 2 cases o f Laron's syndrome under rhIGF-1 therapy, the reappearance or increase of the 42-39 kDa band (intact IGFBP-3) in ligand blotting coincided w ith an increase in immuno-assayable IGFBP-3 levels. Scatchard analysis of the results of competitive binding experiments provided proof of t his difference in affinity. Recombinant IGFBP-3 had approximately 10 t imes the affinity of pregnancy serum IGFBP-3 for the antibody. Further more, acromegalic serum proved to have two types of site, one of high, and one of low, affinity, which correspond to the intact and proteoly sed forms. Since the limited proteolysis of IGFBP-3 does not disrupt t he ternary 150 kDa complexes, our results show that the protease-induc ed structural changes are recognized by the antibody within the comple xes in native serum. Finally, this study shows that the results of IGF BP-3 measurement by RIA depend on the relative affinities of the antib ody for the intact and proteolysed forms of the protein and they shoul d therefore be interpreted with caution.