Abstract

Complementary DNAs encoding a preproinsulin-like growth factor (prepro-IGF) have been cloned from a primitive vertebrate species, the Atlantic hagfish, by using a DNA amplification strategy based on the polymerase chain reaction. A composite sequence containing a 414-nucleotide open reading frame encoding 138 amino acids and 164 nucleotides in the 3'-untranslated region was obtained. The deduced partial sequence of hagfish prepro-IGF reveals that it is organized like the mammalian prepro-IGFs with an unusually large (greater than 39-amino acid) signal peptide (initiator methionine residue is missing), 29-amino acid B, 15-amino acid C, 21-amino acid A, 10-amino acid D, and 26-amino acid E domains. All the invariant residues necessary to form the correct tertiary fold of an insulin-like molecule have been conserved in hagfish IGF. Sequence comparisons revealed that the A and B domains of hagfish IGF are equally similar to those of human IGF-I (35 out of 50 amino acids) or IGF-II (37 out of 53 amino acids). In contrast, the similarity between hagfish and mammalian pro-IGFs in the C, D, and E domains is relatively low. Northern blot analysis of RNA isolated from hagfish brain, heart, liver, skeletal muscle, and islet organ, however, indicated that hagfish IGF, like mammalian IGF-I, is expressed predominantly in the liver as a 4.2-kilobase transcript. DNA blot analysis revealed that hagfish IGF is a single copy gene. The predicted sequence of hagfish prepro-IGF thus demonstrates that the divergence of the IGF and insulin genes occurred prior to the separation of the Agnatha and that the organization and tertiary structure of IGF have been well maintained throughout 550 million years of vertebrate evolution.