Preferential Measurement of Insulin-Like Growth Factor (IGF) I-Related Peptides in Serum with the Aid of IGF-Binding Proteins (IGF BPs) Produced by Rat Liver in Culture. Estimation of Serum IGF BP Levels*

Abstract

A protein-binding assay for insulin-like growth factor (IGF) was developed which preferentially measures IGF I-related peptides in serum. Binding proteins (BPs) extracted from the culture media of livers from approximately 4-week-old rats were used in the assay, with pure IGF I as tracer and a partially purified IGF preparation as standard. Serum samples were gel filtered in acetic acid to separate the IGFs from their BPs. IGFs could be detected with this assay in extracts corresponding to as little as 0.2 μl normal serum. The affinity of these liver BPs was greatest for IGF I. IGF II, somatomedin A, and multiplication-stimulating activity were found to be 2,3, and 10 times less potent, respectively, than IGF I in displacing [¹²⁵I]IGF I. There was no cross-reaction with insulin and proinsulin, structurally the most closely related peptides. There was a highly significant correlation (r = 0.98, P < 0.001) between IGF values obtained from simultaneous assays, using either 1) somatomedin C/IGF I antibodies, or 2) the liver BPs, for acromegalic, normal, and hypopituitary serum extracts. Nevertheless, the protein-binding assay yielded values 1.7-fold those of the RIA. The BPs therefore do not possess the specificity of the antibodies for IGF I, but they do have the advantage of being less species specific in that they permit measurement of IGFs in the rat (among others) with the same sensitivity as in man. The validity of the assay was demonstrated both by the studies of IGF levels as a function of age, which yielded a profile characteristic of IGF I, and by the GH-dependence of the IGF levels measured. Mean IGF levels (±SEM) were the following: 1.03 ± 0.03 U/ml in normal adults; 2.62 ± 0.10 in acromegalic patients; 0.19 ± 0.01 in patients with total GH deficiency; 0.58 ± 0.04 in patients with partial GH deficiency (the reference serum being assigned a potency of 1 U IGF/ml). Human GH administration (6 mg/m2, im) to untreated hypopituitary patients on average provoked a 2-fold increase in IGF levels within 24 h. In hypophysectomized rats there was a close relation between the IGF level attained and the dose of GH administered (P < 0.001). IGF BPs were titrated by incubating different concentrations of the serum extracts with [¹²⁵I]IGF I and comparing these with a BP preparation obtained from the reference serum used in the IGF assay and arbitrarily assigned a value of 1 U IGF BP/ml. Titration curves were parallel for the different mammalian sera tested and in man under the various physiological and pathological conditions studied. Changes in IGF BP levels as a function of age and GH secretion followed the general pattern of those in IGF levels, but there was a greater overlap between normal and pathological values. Mean IGF BP levels (±SEM) were: 0.98 ± 0.04 U/ml in normal adults; 1.88 ± 0.13 in acromegalic patients; 0.58 ± 0.06 in patients with total GH deficiency; 0.57 ± 0.08 in patients with partial GH deficiency. Positive correlations were found between IGF and BP levels in normal children and adults (as a group) (r = 0.51, P < 0.001) and in acromegalic, normal, and hypopituitary subjects (viewed together) (r = 0.74, P < 0.001). Short term administration of GH provoked no significant variations in IGF BP levels. It may be concluded that 1) BP levels less accurately reflect GH status than do IGF levels, and 2) the processes regulating the biosynthesis of IGFs and BPs are probably distinct from one another, but interdependent.