The structure of the messenger RNA (mRNA) encoding the precursor to mouse submaxillary epidermal growth factor (EGF) was determined from the sequence of a set of overlapping complementary DNA's (cDNA). The mRNA is unexpectedly large, about 4750 nucleotide bases, and predicts the sequence of preproEGF, a protein of 1217 amino acids (133,000 molecular weight). The EGF moiety (53 amino acids) is flanked by polypeptide segments of 976 and 188 amino acids at its amino and carboyxl termini, respectively. The amino terminal segment of the precursor contains seven peptides with sequences that are similar but not identical to EGF.

Complementary DNA clones encoding the human kidney epidermal growth factor (EGF) precursor have been isolated and sequenced. They predict the sequence of a 1,207 amino acid protein which contains EGF flanked by polypeptide segments of 970 and 184 residues at its NH2- and COOH-termini, respectively. The structural organization of the human EGF precursor is similar to that previously described for the mouse protein and there is 66% identity between the two sequences. Transfection of COS-7 cells with the human EGF precursor cDNA linked to the SV40 early promoter indicate that it can be synthesized as a membrane protein with its NH2-terminus external to the cell surface. The human EGF precursor gene is approximately 110 kilobase pairs and has 24 exons. Its exon-intron organization revealed that various domains of the EGF precursor are encoded by individual exons. Moreover, 15 of the 24 exons encode protein segments that are homologous to sequences in other proteins. Exon duplication and shuffling appear to have played an important role in determining the present structure of this protein.

The activation of protein phosphastase-1 (PP1) by insulin plays a critical role in the regulation of glycogen metabolism. PTG is a PP1 glycogen-targeting protein, which also binds the PP1 substrates glycogen synthase, glycogen phosphorylase, and phosphorylase kinase (Printen, J. A., Brady, M. J., and Saltiel, A. R. (1997) Science 275, 1475-1478). Through a combination of deletion analysis and site-directed mutagenesis, the regions on PTG responsible for binding PP1 and its substrates have been delineated. Mutagenesis of Val-62 and Phe-64 in the highly conserved (K/R)VXF PP1-binding motif to alanine was sufficient to ablate PP1 binding to PTG. Phosphorylase kinase, glycogen synthase, and phosphorylase binding all mapped to the same C-terminal region of PTG. Mutagenesis of Asp-225 and Glu-228 to alanine completely blocked the interaction between PTG and these three enzymes, without affecting PP1 binding. Disruption of either PP1 or substrate binding to PTG blocked the stimulation of PP1 activity in vitro against phosphorylase, indicating that both binding sites may be important in PTG action. Transient overexpression of wild-type PTG in Chinese hamster ovary cells overexpressing the insulin receptor caused a 50-fold increase in glycogen levels. Expression of PTG mutants that do not bind PP1 had no effect on glycogen accumulation, indicating that PP1 targeting is essential for PTG function. Likewise, expression of the PTG mutants that do not bind PP1 substrates did not increase glycogen levels, indicating that PP1 targeting glycogen is not sufficient for the metabolic effects of PTG. These results cumulatively demonstrate that PTG serves as a molecular scaffold, allowing PP1 to recognize its substrates at the glycogen particle.Protein phosphatase-1 (PP1) 1 is one of the four major serine/ threonine protein phosphatase families expressed in eukaryotic cells (2). The enzyme regulates a variety of cellular functions, including cell cycle progression, RNA splicing, vesicle fusion, ion channel function, and muscle contraction (3-8). PP1 also plays a key role in the hormonal regulation of glycogen metabolism, catalyzing the dephosphorylation of glycogen synthase, glycogen phosphorylase, and phosphorylase kinase (9). These dephosphorylation reactions promote the net synthesis of glycogen by activating glycogen synthase and inhibiting phosphorylase. A number of pharmacological (10) and biochemical (11-13) studies have suggested that insulin stimulates glycogen synthesis, at least in part, through the activation of PP1.PP1 is ubiquitously expressed and resides in most cellular compartments. However, the hormonal activation of the enzyme is restricted to discrete sites, such as the glycogen particle, suggesting that mechanisms must exist to ensure the localized regulation of the enzyme. For example, although PP1 is found in a number of cellular locations in fat, liver, and muscle cells, insulin produces the dephosphorylation of only a small fraction of phosphoproteins. The compartmentalization of PP1 is mediated by a family ...

Complementary DNAs encoding mouse liver insulin-like growth factor I (ICF-I) have been isolated and sequenced. Alternative RNA splicing results in the synthesis of two types of mouse IGF-I precursor that differ in the size and sequence of the COOH-terminal peptide.The sequences of the signal peptides, IGF-I moieties and the first 16 amino acids of the COOH-terminal peptides or E-domains of the two precursors are identical.The sequence difference results from the presence in preproIGF-IB mRNA of a 52 base insertion which introduces a 17 amino acid segment into the COOH-terminal peptide of preproIGF-IB and also causes a shift in the reading frame of the mRNA. As a consequence of this insertion, the COOH-terminal 19 and 25 amino acids of mouse preproIGF-IA and -IB, respectively, are different. The sequences of mouse and human preproIGF-IA are highly conserved and possess 94% identity.In contrast, the sequences of mouse and human preproIGF-IB are quite different in the region of the COOH-terminal peptide. A comparison of the sequences of mouse and human preproIGF-IB mRNA indicates that they are generated by different molecular mechanisms.

Overlapping recombinant clones that encompass the insulin-like growth factor (IGF) I and II genes have been isolated from a human genomic DNA library. Each gene is present once per haploid genome; the IGF-I gene spans >35 kilobase pairs (kbp) and the IFG-II gene is at least 15 kbp. The exon-intron organization of these genes is similar, each having four exons, which is one more than the related insulin gene.Comparison of the restriction endonuclease cleavage maps of the IGF-ll and insulin genes, including their flanking regions and hybridization with an IGF-II cDNA probe, revealed that they are adjacent to one another. The IGF-II and insulin genes have the same polarity and are separated by 12.6 kbp of intergenic DNA that includes a dispersed middle repetitive Alu sequence. The order of the genes is 5'-insulin-IGF-II-3'.The insulin-like growth factors (IGFs) and insulin are related polypeptides that have a high degree of sequence homology and exhibit a similar spectrum of biological activities (1-4). They produce rapid metabolic changes and have long-term growth promoting effects as well. However, the relative concentrations required to elicit these responses are different. In general, insulin is more potent in producing short-term metabolic effects, while the IGFs are more potent in promoting growth. Complementary DNAs encoding the precursors to human insulin, IGF-I and IGF-II, have been isolated and sequenced (5-8). The chromosomal location of the human insulin (INS), IGF-I (IGFI), and IGF-II (IGF2) genes has also been determined (refs. 9-15; unpublished data Maniatis. The isolation and characterization of the human INS gene and flanking regions has been described (17-21). The X phage, XhINS-2, was isolated by hybridization with nick-translated (22) human insulin cDNA, pchil (5). Phage XhINS-5 and -3 were isolated using the 2.2-and 1.8-kbp EcoRI fragments, respectively, of XhINS-2 as probes. The IGFI gene-containing phage were isolated by hybridization (21) with the human IGF-I cDNA, phigfl (7). XhIGF1-4 was isolated from a XEMBL4 (23) library of partial Sau3A fragments ofDNA from the lymphoblastoid cell line GM1416, which has the karyotype 48,XXXX. The IGF2 gene-containing phage were isolated by hybridization with the human IGF-II cDNA, phigf2 (7). Phage were plaque-purified and DNA was isolated by standard procedures (24). The order of the EcoRI fragments in the human DNA inserts was determined by single and double digests of the phage DNA as well as by digests of restriction fragments eluted from lowmelting-point agarose gels (24). The locations of the exons and repeated sequences were determined by restriction mapping, blotting, and hybridization with nick-translated cDNA and human DNA probes, respectively. As phigfl lacked sequences corresponding to the 5' untranslated region, an oligonucleotide (5' TAATTGGGTTGGAAGA-CTGC 3') derived from the sequence of a longer IGF-I cDNA reported by Jansen et al. (6) was chemically synthesized for use as a probe to locate this region of the IGFI gene.DNA Sequenc...