A personal view on the early history of the insulin-like growth factors

Salmon and Daughaday, when trying to set up an in vitro assay for Growth Hormone (GH), failed to obtain a direct effect on sulphate uptake in cartilage of hypophysectomized (hypox) rats. They recognised that this was not the consequence of poor methodology or materials, but an encrypted message from the examined system. They decided to turn around and to try and decipher it. Treatment with GH appeared to render hypox rat serum active in stimulating sulphation in hypox rat cartilage. They proposed that GH induced an intermediary substance, responsible for this biological effect: sulphation factor (SF), later renamed to Somatomedin(s) (SM). This hypothesis met with great criticism and very few took on to study this hypothetical substance. Besides disbelief, slow progress was also due to initial lack of a practical assay and to the failure to find a tissue with enriched concentration from which to extract the activity. From experimental evidence, the concept gradually evolved that SF/SM was insulin-like and might be identical to NSILA (non-suppressible insulin-like activity). This again generated controversy. This characteristic was too far away from the known effects of GH to be readily acceptable as a physiological phenomenon. The subsequent recognition of the distinct characteristics of the receptors for SM/NSILA and insulin, the discovery of the SM/NSILA binding proteins and, much later, a beginning understanding of their interactions, modifications and breakdown, have gradually resolved this apparent contradiction. When the sequence of two NSILA molecules became known, they were named IGF-I and -II. Structural similarity with proinsulin and identity of IGF-I with SM-C and -A were established and it was found that Multiplication Stimulating Activity (MSA), a growth factor isolated from fetal calf serum and subsequently from conditioned media of a rat liver cell line, was the rat equivalent of IGF-II. Structure-function relations could be studied, a quest which is not yet brought to an end. Meanwhile, the endocrine profile of SF/SM had gradually emerged by measuring plasma levels with bioassays. The main determinants were found to be age, body size, GH and the nutritional state. Later, radioimmunoassays were developed, enabling consolidation and detailing of these early observations, and allowing explorations at the tissue level. As another aspect of the endocrine paradigm, in vivo effects of IGFs were studied. The initial demonstration of an effect of crude preparations on longitudinal growth in experimental animals raised heavy scepticism, since the effect might have been an artefact caused by contaminants. It took confirmation with highly purified preparations and biosynthetic IGF-I to ease this concern. Still, not until recent years it was demonstrated, by knocking out the genes, that a true physiological and not a pharmacological effect had been induced previously. When it was found that most tissues produce SMs and are sensitive to their actions, the concept emerged that IGFs may have para- and autocrine functions. Early experiments with combinations of growth factors in cell cultures had begun to define their specific roles in the cell cycle as competence or progression factors. SM-C fell in the latter category. Still, the awareness grew that, for obtaining physiologically meaningful results on the role of IGFs in complex, dynamic and tissue-specific environments, involving interactions of many hormones and growth factors, the intactness of tissue was a prerequisite. One result of this approach was the discovery of a direct interaction of GH with cartilage, leading, in concert with IGFs, to a clonal expansion of the cartilage cells of the growth plate. The isolation and sequencing of the IGF-I and -II genes, and later, of six IFG-BPs initiated the gradual elucidation of structure and function at the DNA and RNA level and the study of natural and synthetic IGF-variants. The generation of transgenic animals became feasible