Insulin through the ages: Phylogeny of a growth promoting and metabolic regulatory hormone

Insulin Through the Ages: Phylogeny of a Growth Promoting and Metabolic Regulatory Hormone1

SYNOPSIS. Insulin was discovered in 1922 as the causative factor for a human metabolic disorder (diabetes mellitus), but it was recognized early that the hormone had a broad phylogenetic distribution. By the mid 1970s, insulin had been isolated and sequenced from all classes of vertebrates, including Agnatha. Also it was discovered that the insulin gene family in vertebrates included two closely related hormones named insulin-like growth factor (IGF)-I and -II. More recently, the application of recombinant DNA techniques have identified insulin-like peptide genes in invertebrates, including insects, molluscs and nematode and these findings clearly establish that insulin is an evolutionarily ancient hormone which is present in all metazoa. Here we briefly review the structure and function of the insulin/ IGF gene family in vertebrates and invertebrates. Although these studies are ongoing, it appears that in invertebrates the insulin-like peptides function predominately as mitogenic growth factors that act to promote tissue growth and development. However, in vertebrates the mitogenic growth function has been subsumed by IGF-I and -H while insulin has acquired the function of being primarily a metabolic regulatory hormone. The gene duplication and divergence events necessary for this development probably occurred early during vertebrate evolution in the transition from protochordates, represented by extant amphioxus, to primitive jawless vertebrates, represented by extant lamprey and hagfish.

INTRODUCTION

The discovery of insulin by Banting and Best in 1922 represented a milestone in clinical medicine but it has also contributed substantially to progress in the fields of molecular and comparative endocrinology. The importance of insulin used in the treatment of diabetes elicited great interest in this hormone and consequently resources became available for studies to characterize its structure and function. As a result, insulin was the first protein whose primary amino acid sequence was determined (Ryle et aL, 1955); the radioimmunoassay technique was developed using insulin as the model (Yalow and Berson, 1970); insulin was the first hormone for which the 3-dimensional crystal structure was solved (Adams et aL, 1969); proinsulin was the first hormone precursor discovered (Steiner et aL, 1967); and the insulin cDNA and gene were among the first to be cloned by recombinant DNA techniques (Ullrich et al., 1977). Together with these impressive "firsts" in molecular endocrinology, there has also been considerable interest to determine the phylogenetic and evolutionary origin of insulin.

Shortly after the discovery of insulin, some researchers thought that it was a unique hormone present only in mammals and birds. However, as early as 1923 Collip reported finding insulin-like activity in a bivalve mollusc, Mya arenaria. Assisted by advances in a number of biochemical techniques, including the development of an efficient method for isolating pure insulin from the pancreas, improved immunological and immunohistochemical assays for insulin, and importantly development of the Edman degradation method for amino acid sequence analysis, scientists by 1975 had isolated and sequenced insulin molecules from representative species in all classes vertebrates including hagfish (Fig. 1). The hagfish, together with lampreys, are living representatives of the jawless vertebrates (class Agnatha) and are considered to be the most evolutionarily ancient vertebrates.

In contrast to the success achieved in vertebrates, early attempts to isolate and characterize an insulin-like molecule from an invertebrate species were unsuccessful. However, in 1986 Nagasawa and coworkers in a herculean effort successfully purified and sequenced a 4 kilodalton peptide from the silkworm Bombyx mori which they originally called 4K-prothoracicotropic hormone-II and later renamed bombyxin A2 (Nagasawa et al., 1986)). The amino acid sequence of bombyxin A2 clearly showed that it was an insulin-like molecule, i.e., it consisted of two peptide chains (A and B) crosslinked by highly conserved insulin type disulfide bonds. The overall sequence identity between bombyxin A2 and human insulin was 40%.

More recently, recombinant DNA techniques have been used to clone insulin-like peptide genes in a variety of invertebrate species, including mollusc, insects and the nematode Caenorhabditis elegans (Smit et al., 1988; Lagueux et al., 1990; Duret et al., 1998; Gregoire et al., 1998). An insulin receptor-like tyrosine kinase cDNA has also been cloned from a colenterate, hydra (Steele et al., 1996) and from C. elegans (Kimura et al., 1997). Although a previous report claimed that sponges contain an insulin mRNA (Robitzki et al., 1989), this has been retracted (Duret et al., 1998). Recently, the yeast genome was completely sequenced and no insulin-like genes were found (Chervitz et al., 1998). These results suggest that insulin-like genes have coevolved with the appearance of the metazoa (multicellular animals with differentiated tissues).