An efficient expression system in Escherichia coli for several biologically active insulin-like growth factor-I (IGF-I) fusion peptide analogues is described.

(1) Many cell types secrete insulin-like growth factor (IGF)-binding proteins that can be expected to sequester free IGF and modify the biological activities of the growth factors. (2) A binding protein purified from bovine kidney (MDBK) cells potently inhibited the ability of IGF-2 to stimulate DNA synthesis or protein accumulation as well as to reduce rates of protein breakdown in chick embryo fibroblasts. The binding protein did not influence the biological activities of des-(1-3)-IGF-1, while effects on IGF-1 were intermediate. Since the chick embryo fibroblasts contain only the type 1 IGF receptor, the MDBK-cell binding protein must have reduced the accessibility of IGF-2 and IGF-1 to that receptor. Binding to the type 2 receptor on L6 myoblasts was also inhibited. (3) Inhibiting effects on both protein breakdown responsiveness to IGF and IGF binding to cell receptors were also observed with human amniotic fluid binding protein, although here IGF-1 and IGF-2 were equipotent. These results contrast with stimulatory responses on different IGF-1 actions of the same binding protein reported previously [Elgin, Busby & Clemmons (1987) Proc. Natl. Acad. Sci. U.S.A. 84, 3254-3258]. (4) The biological potencies of IGF-1, IGF-2 and des-(1-3)-IGF-1 correlate inversely with their binding to proteins released into the medium by cells, so that the enhanced potency of des-(1-3)-IGF-1 is a consequence of it not binding to purified binding proteins or those released by cultured cells.

In order to elucidate the role of the N-terminus of insulin-like growth factor 1 (IGF-1) with respect to its biological properties, we chemically synthesized analogues of IGF-1 truncated by one to five amino acid residues from the N-terminus. In a bioassay that measured the stimulation of protein synthesis in rat L6 myoblasts, the concentrations required to produce a half-maximal response were: IGF-1, 13 ng/ml; des-(1)-IGF-1, 10 ng/ml; des-(1-2)-IGF-1, 13 ng/ml; des-(1-3)-IGF-1, 1.5 ng/ml; des-(1-4)-IGF-1, 5.1 ng/ml; des-(1-5)-IGF-1, 1200 ng/ml. When tested for their abilities to compete with 125I-IGF-1 binding to L6 myoblasts at 3 degrees C, the concentrations required for 50% competition were: IGF-1, des-(1)-IGF-1 and des-(1-2)-IGF-1, 20 ng/ml; des-(1-3)-IGF-1, 14 ng/ml; des-(1-4)-IGF-1, 40 ng/ml; des-(1-5)-IGF-1, greater than 1000 ng/ml. Receptor-binding experiments at 25 degrees C, however, gave results suggesting that the myoblasts were secreting a binding protein selective for the three longest peptides. This interpretation was confirmed by binding studies with medium conditioned by the L6 myoblasts as well as binding protein purified from MDBK-cell-conditioned medium. In both cases IGF-1, des-(1)-IGF-1 and des-(1-2)-IGF-1 competed for tracer IGF-1 binding at least 60-fold better than did the three shorter peptides. The results obtained account for the increased potency of des-(1-3)-IGF-1 and des-(1-4)-IGF-1, since their activities are not attenuated by the binding protein, and the relatively lower potency of des-(1-4)-IGF-1 is a consequence of this peptide binding less well to the L6-myoblast receptor.

The development of an efficient expression system for insulin-like growth factor-I (IGF-I) in Escherichia coli as a fusion protein is described. The fusion protein consists of an N-terminal extension made up of the first 46 amino acids of methionyl porcine GH ([Met1]-pGH) followed by the dipeptide Val-Asn. The latter two residues provide a unique hydroxylamine-sensitive link between [Met1]-pGH(1-46) and the N-terminal Gly of IGF-I. Downstream processing of the fusion proteins involved isolation of inclusion bodies, cleavage at the Asn-Gly bond, refolding of the reduced IGF-I peptide and purification to homogeneity. This expression system was also used to produce two variants of IGF-I in which Glu3 was substituted by either Gly or Arg to give [Gly3]-IGF-I and [Arg3]-IGF-I respectively. Production of milligram quantities of IGF-I peptide was readily achieved. The purity of the IGF-I, [Gly3]-IGF-I and [Arg3]-IGF-I was established by high-performance liquid chromatography and N-terminal sequence analysis. [Gly3]-IGF-I and [Arg3]-IGF-I were more potent than IGF-I in biological assays measuring stimulation of protein synthesis and DNA synthesis or inhibition of protein breakdown in rat L6 myoblasts. Both analogues bound very poorly to bovine IGF-binding protein-2 and slightly less well than IGF-I to the type-1 receptor on rat L6 myoblasts. We conclude that reduced binding to IGF-binding proteins rather than increased receptor binding is the likely explanation for the greater biological potency of the analogues compared with IGF-I.

No abstract