The Insulin Receptor Isoform A: A Mitogenic Proinsulin Receptor?

Abstract

The existence of a precursor (“proinsuline”) for the putative hormone “insuline” was already postulated in 1916 (before the isolation of insulin) by Sir Edward Sharpey Schafer (1). It took another half-century for proinsulin to be isolated and characterized by Don Steiner (for recent review, see Ref. 2). Historically and currently, proinsulin has been regarded as a weak-affinity ligand for the insulin receptor (IR) with a low metabolic potency (3). The availability of biosynthetic human proinsulin in the 1980s allowed the exploration of possible clinical usefulness of human proinsulin as a soluble, intermediate acting insulin agonist by the Eli Lilly Co. (Indianapolis, IN) in 1984 (3). Low receptor affinity (and hence low receptor-mediated endocytosis) translated into a longer half-life and slower metabolic clearance rate. It also showed a marginal hepatoselectivity. Lack of demonstrable clinical advantage and an increase in the occurrence of acute myocardial infarctions led to a close of the clinical trials in early 1988. Proinsulin as an active biological agent went into oblivion since, with the notable exception of the work by the lab of Flora de Pablo in Madrid, suggesting that proinsulin is an active agent of its own in early embryonic development even before the appearance of the pancreatic primordium (4). Renewed interest in the actions of proinsulin is likely to arise from the publication in this issue of an article by Malaguarnera et al. (5) in the group that has pioneered a series of elegant studies on the biochemical properties (ligand binding selectivity and signal transduction) of the two splice variants of the IR, isoforms A and B (IR-A and IR-B), and their homodimers and heterodimers with IGF-I receptors (for review, see Ref. 6). IR-A and IR-B differ by the absence or presence of a 12-amino acid peptide sequence encoded by IR exon 11. The IR-A is predominantly expressed during prenatal life, and in adult life is expressed preferentially in noninsulin target tissues. It has a greater affinity for IGFs than IR-B, especially for IGF-II. IR-A appears to be especially potent in mitogenic signaling and may play a role in cancer progression (6). In this new study, using fibroblasts devoid of IGF-IR (R ) stably transfected with similar amounts of IR-A or IR-B, the authors report that proinsulin binds to IR-A with a 7-fold higher affinity than IR-B [dissociation constant (Kd) for IR-A 4.6 nM vs. Kd for IR-B 28.3 nM] and activates IR-A with a greater potency than IR-B, behaving essentially like IGF-II except for lack of binding to the IGF-I receptor, as previously shown (3). This is the first direct comparison of the affinities of the two receptor isoforms for proinsulin. It is interesting to compare the data with historical figures using cells that were later shown to express predominantly one or the other isoform or both. The authors define binding of proinsulin to IR-A as “high affinity.” One may question this statement given a Kd of 4.6 nM. In my laboratory, the Kd for high-affinity insulin binding to IR-A has been consistently around 0.2 nM (7, 8), in agreement with Ref. 9, which is 23 times lower than the proinsulin Kd reported here. The authors do not report on the relative values of proinsulin’s Kd to insulin’s Kd in their cells. Inspection of figure 1 in Ref. 5 suggests about 10% for IR-A and 2% for IR-B. Relative binding of porcine proinsulin relative to porcine insulin in IM-9 lymphocytes, which express only IR-A, was initially reported to be 10% (10). In liver plasma membranes, which predominantly express IR-B, it was 3–5% (due to a 20-fold lower asso-