Coordinate control of muscle cell survival by distinct insulin-like growth factor activated signaling pathways.

Peptide growth factors control diverse cellular functions by regulating distinct signal transduction pathways. In cultured myoblasts, insulin-like growth factors (IGFs) stimulate differentiation and promote hypertrophy. IGFs also maintain muscle cell viability. We previously described C2 skeletal muscle lines lacking expression of IGF-II. These cells did not differentiate, but underwent progressive apoptotic death when incubated in differentiation medium. Viability could be sustained and differentiation enabled by IGF analogues that activated the IGF-I receptor; survival was dependent on stimulation of phosphatidylinositol 3-kinase (PI3-kinase). We now find that IGF action promotes myoblast survival through two distinguishable PI3-kinase-regulated pathways that culminate in expression of the cyclin-dependent kinase inhibitor, p21. Incubation with IGF-I or transfection with active PI3-kinase led to rapid induction of MyoD and p21, and forced expression of either protein maintained viability in the absence of growth factors. Ectopic expression of MyoD induced p21, and inhibition of p21 blocked MyoD-mediated survival, thus defining one PI3-kinase-dependent pathway as leading first to MyoD, and then to p21 and survival. Unexpectedly, loss of MyoD expression did not impede IGF-mediated survival, revealing a second pathway involving activation by PI3-kinase of Akt, and subsequent induction of p21. Since inhibition of p21 caused death even in the presence of IGF-I, these results establish a central role for p21 as a survival factor for muscle cells. Our observations also define a MyoD-independent pathway for regulating p21 in muscle, and demonstrate that distinct mechanisms help ensure appropriate expression of this key protein during differentiation.

Insulin-like growth factor binding protein-5 modulates muscle differentiation through an insulin-like growth factor-dependent mechanism.

The insulin-like growth factor binding proteins (IGFBPs) are a family of six secreted proteins which bind to and modulate the actions of insulin-like growth factors-I and -II (IGF-I and -II). IGFBP-5 is more conserved than other IGFBPs characterized to date, and is expressed in adult rodent muscle and in the developing myotome. We have shown previously that C2 myoblasts secrete IGFBP-5 as their sole IGFBP. Here we use these cells to study the function of IGFBP-5 during myogenesis, a process stimulated by IGFs. We stably transfected C2 cells with IGFBP-5 cDNAs under control of a constitutively active promoter. Compared with vector-transfected control cells, C2 myoblasts expressing the IGFBP-5 transgene in the sense orientation exhibit increased IGFBP-5 levels in the extracellular matrix during proliferation, and subsequently fail to differentiate normally, as assessed by both morphological and biochemical criteria. Compared to controls, IGFBP-5 sense myoblasts show enhanced survival in low serum medium, remaining viable for at least four weeks in culture. By contrast, myoblasts expressing the IGFBP-5 antisense transcript differentiate prematurely and more extensively than control cells. The inhibition of myogenic differentiation by high level expression of IGFBP-5 could be overcome by exogenous IGFs, with des (1-3) IGF-I, an analogue with decreased affinity for IGFBP-5 but normal affinity for the IGF-I receptor, showing the highest potency. These results are consistent with a model in which IGFBP-5 blocks IGF-stimulated myogenesis, and indicate that sequestration of IGFs in the extracellular matrix could be a possible mechanism of action. Our observations also suggest that IGFBP-5 normally inhibits muscle differentiation, and imply a role for IGFBP-5 in regulating IGF action during myogenic development in vivo.

Focal expression of insulin-like growth factor I in rat kidney collecting duct.

To address the question of insulin-like growth factor (IGF) I localization and synthesis in kidney, we used two complementary experimental approaches: immunohistochemistry of fixed paraffin-embedded rat kidney sections; and measurement of IGF I mRNA in isolated components of the rat nephron, using a highly sensitive and specific solution hybridization assay. Immunostainable IGF I was localized exclusively to principal cells of cortical and medullary collecting ducts. Administration of growth hormone to hypophysectomized rats for 8 d resulted in enhanced immunohistochemical staining of IGF I within collecting ducts, but no detectable IGF I in other portions of the nephron. The abundance of IGF I mRNA was 7-12-fold higher in isolated papillary collecting ducts than in proximal tubules or glomeruli, and was enriched 10-fold compared with whole kidney. Our data demonstrate colocalization of IGF I and IGF I mRNA in the collecting duct, consistent with focal expression of the IGF I gene at this site.

Complete thrombospondin mRNA sequence includes potential regulatory sites in the 3' untranslated region.

The nucleotide sequence of human thrombospondin (TS) mRNA has been determined from human fibroblast and endothelial cDNAs. The sequence of 5802 bp begins 110 bp upstream from the initiator codon and includes the entire 3' untranslated region (UTR) of the mRNA. The coding region (3510 bp) specifies a protein of 1170 amino acids with all of the known features of the TS subunit (Frazier, W. A. 1987. J. Cell Biol. 105:625-632). The long 3' UTR of 2166 nucleotides is extremely A/T-rich, particularly in the latter half. It contains 37 TATT or ATTT(A) sequences that have been suggested as mediators of the stability of mRNAs for cytokines, lymphokines, and oncogenes (Shaw, G., and R. Kamen. 1986. Cell. 46:659-667). Another unusual feature of the 3' UTR of TS mRNA is a stretch of 42 nucleotides of which 40 are thymidines (uridine in the mRNA) including an uninterrupted sequence of 26 thymidines. This region is flanked by two sets of direct repeats suggesting that it may be an insertion element of retrotranscriptional origin. Comparison of the 3' untranslated region of TS mRNA with the GenBank data base indicates the greatest degree of similarity with an alpha-interferon gene which contains a number of the TATT/ATTT consensus sites. The degree of similarity between the TS and interferon sequences is the same in regions of the interferon gene corresponding to its coding and noncoding regions suggesting that most of the TS 3' UTR may be derived from an interferon gene or pseudogene. The features of the TS mRNA 3' UTR provide a potential explanation for the rapid regulation of TS message observed in cultured cells in response to PDGF and suggest that TS is a member of a group of proteins which are intimately involved in the control of cell growth and differentiation.

Polymorphism in the 5'-flanking region of the human insulin gene and its possible relation to type 2 diabetes.

The arrangement of the human insulin gene in DNA from 87 individuals was analyzed by the Southern blot hybridization technique with a cloned genomic human insulin probe. Insertions of 1.5 to 3.4 kilobase pairs in the 5'-flanking region of the gene were found in DNA from 38 individuals. These insertions occurred within 1.3 kilobase pairs of the transcription initiation site. In contrast, no insertions were observed in the region 3' to the coding sequence. The prevalence of these insertions in type 2 diabetes was significantly greater than in the other groups (P less than .001). The limitation of this striking length polymorphism to a potential promoter region suggests that these insertions may play a role in insulin gene expression.

Insulin-like growth factors I and II. Peptide, messenger ribonucleic acid and gene structures, serum, and tissue concentrations.

There is currently widespread interest in the IGFs (IGF-I and IGF-II) and their roles in the regulation of growth and differentiation of an ever increasing number of tissues are being reported. This selective review focused on the current state of our knowledge about the structure of mammalian IGFs and the multiple forms of mRNAs which arise from alternative splicing and promoter sites which arise from gene transcription. Current progress in the immunological measurement of the IGF is reviewed including different strategies for avoiding binding protein interference. The results of measurements of serum IGF-I and IGF-II in fetus and mother and at various stages of postnatal life are described. Existing knowledge of the concentration of these peptides in body fluids and tissues are considered. Last, an attempt is made to indicate circumstances in which the IGFs are exerting their actions in an autocrine/paracrine mode and when endocrine actions predominate. In the latter context it was concluded that an important role for GH action on skeletal tissues via hepatic production of IGF-I and endocrine action of IGF-I on growth cartilage is likely.

Polymorphism in the 5' flanking region of the human insulin gene. Relationships with noninsulin-dependent diabetes mellitus, glucose and insulin concentrations, and diabetes treatment in the Pima Indians.

Variations in DNA sequences flanking the insulin gene were studied in relation to noninsulin-dependent diabetes mellitus (NIDDM) in 87 unrelated Pima Indians at least 35 yr of age. DNA was isolated from nuclei of peripheral blood leukocytes and digested with restriction endonucleases. Less variation in this region was found in Pima Indians than in other racial groups previously studied. Only two classes of alleles (classes 1 and 3) were found, and there was virtually no variation within classes. At least one class 3 allele was found in 47% of the 38 nondiabetic subjects and in 37% of the 49 with NIDDM (odds ratio = 0.65, P = 0.4, 95% confidence interval for the odds ratio = 0.25 to 1.67). Homozygosity for class 3 alleles, however, was found only in diabetics. There were no differences according to genotype in obesity, fasting or postload glucose or insulin concentrations, or in the relationships between insulin and glucose concentrations. 61% (11/18) of the diabetics with a class 3 allele were receiving drug treatment for diabetes compared with only 26% (8/31) of diabetics without a class 3 allele (P = 0.03). The insulin gene polymorphism probably plays no important role in the genesis of NIDDM in Pima Indians, nor does it influence the glucose or insulin concentrations or their relationship to each other, but the class 3 allele, especially when homozygous in this population, may influence the severity of the disease as indicated by need for drug treatment.

Insulin-like growth factor-mediated muscle cell survival: central roles for Akt and cyclin-dependent kinase inhibitor p21.

Polypeptide growth factors activate specific transmembrane receptors, leading to the induction of multiple intracellular signal transduction pathways which control cell function and fate. Recent studies have shown that growth factors promote cell survival by stimulating the serine-threonine protein kinase Akt, which appears to function primarily as an antiapoptotic agent by inactivating death-promoting molecules. We previously established C2 muscle cell lines lacking endogenous expression of insulin-like growth factor II (IGF-II). These cells underwent apoptotic death in low-serum differentiation medium but could be maintained as viable myoblasts by IGF analogues that activated the IGF-I receptor or by unrelated growth factors such as platelet-derived growth factor BB (PDGF-BB). Here we show that IGF-I promotes muscle cell survival through Akt-mediated induction of the cyclin-dependent kinase inhibitor p21. Treatment of myoblasts with IGF-I or transfection with an inducible Akt maintained muscle cell survival and enhanced production of p21, and ectopic expression of p21 was able to sustain viability in the absence of growth factors. Blocking of p21 protein accumulation through a specific p21 antisense cDNA prevented survival regulated by IGF-I or Akt but did not block muscle cell viability mediated by PDGF-BB. Our results define Akt as an intermediate and p21 as a critical effector of an IGF-controlled myoblast survival pathway that is active during early myogenic differentiation and show that growth factors are able to maintain cell viability by inducing expression of pro-survival molecules.

Dual control of muscle cell survival by distinct growth factor-regulated signaling pathways.

In addition to their ability to stimulate cell proliferation, polypeptide growth factors are able to maintain cell survival under conditions that otherwise lead to apoptotic death. Growth factors control cell viability through regulation of critical intracellular signal transduction pathways. We previously characterized C2 muscle cell lines that lacked endogenous expression of insulin-like growth factor II (IGF-II). These cells did not differentiate but underwent apoptotic death in low-serum differentiation medium. Death could be prevented by IGF analogues that activated the IGF-I receptor or by unrelated growth factors such as platelet-derived growth factor BB (PDGF-BB). Here we analyze the signaling pathways involved in growth factor-mediated myoblast survival. PDGF treatment caused sustained activation of extracellular-regulated kinases 1 and 2 (ERK1 and -2), while IGF-I only transiently induced these enzymes. Transient transfection of a constitutively active Mek1, a specific upstream activator of ERKs, maintained myoblast viability in the absence of growth factors, while inhibition of Mek1 by the drug UO126 blocked PDGF-mediated but not IGF-stimulated survival. Although both growth factors activated phosphatidylinositol 3-kinase (PI3-kinase) to similar extents, only IGF-I treatment led to sustained stimulation of its downstream kinase, Akt. Transient transfection of a constitutively active PI3-kinase or an inducible Akt promoted myoblast viability in the absence of growth factors, while inhibition of PI3-kinase activity by the drug LY294002 selectively blocked IGF- but not PDGF-mediated muscle cell survival. In aggregate, these observations demonstrate that distinct growth factor-regulated signaling pathways independently control myoblast survival. Since IGF action also stimulates muscle differentiation, these results suggest a means to regulate myogenesis through selective manipulation of different signal transduction pathways.

A calcineurin-NFATc3-dependent pathway regulates skeletal muscle differentiation and slow myosin heavy-chain expression.

The differentiation and maturation of skeletal muscle cells into functional fibers is coordinated largely by inductive signals which act through discrete intracellular signal transduction pathways. Recently, the calcium-activated phosphatase calcineurin (PP2B) and the family of transcription factors known as NFAT have been implicated in the regulation of myocyte hypertrophy and fiber type specificity. Here we present an analysis of the intracellular mechanisms which underlie myocyte differentiation and fiber type specificity due to an insulinlike growth factor 1 (IGF-1)-calcineurin-NFAT signal transduction pathway. We demonstrate that calcineurin enzymatic activity is transiently increased during the initiation of myogenic differentiation in cultured C2C12 cells and that this increase is associated with NFATc3 nuclear translocation. Adenovirus-mediated gene transfer of an activated calcineurin protein (AdCnA) potentiates C2C12 and Sol8 myocyte differentiation, while adenovirus-mediated gene transfer of noncompetitive calcineurin-inhibitory peptides (cain or DeltaAKAP79) attenuates differentiation. AdCnA infection was also sufficient to rescue myocyte differentiation in an IGF-depleted myoblast cell line. Using 10T1/2 cells, we demonstrate that MyoD-directed myogenesis is dramatically enhanced by either calcineurin or NFATc3 cotransfection, while a calcineurin inhibitory peptide (cain) blocks differentiation. Enhanced myogenic differentiation directed by calcineurin, but not NFATc3, preferentially specifies slow myosin heavy-chain expression, while enhanced differentiation through mitogen-activated protein kinase kinase 6 (MKK6) promotes fast myosin heavy-chain expression. These data indicate that a signaling pathway involving IGF-calcineurin-NFATc3 enhances myogenic differentiation whereas calcineurin acts through other factors to promote the slow fiber type program.

Growth hormone rapidly activates rat serine protease inhibitor 2.1 gene transcription and induces a DNA-binding activity distinct from those of Stat1, -3, and -4.

Transcriptional regulation by growth hormone (GH) represents the culmination of signal transduction pathways that are initiated by the cell surface GH receptor and are targeted to the nucleus. Recent studies have demonstrated that the activated GH receptor can stimulate Stat1, a cytoplasmic transcription factor that becomes tyrosine phosphorylated and translocates to the nucleus, where it can interact with specific DNA sequences to modulate gene expression. GH also has been found to induce protein binding to a portion of the rat serine protease inhibitor (Spi) 2.1 gene promoter that is required for GH-induced transcription of Spi 2.1. Using GH-deficient hypophysectomized rats as a model, we show that GH treatment rapidly and potently induces both nuclear Spi 2.1 mRNA expression in the liver and specific nuclear protein binding to a 45-bp segment of the Spi 2.1 gene promoter. A GH-inducible gel-shifted complex appears within 15 min of systemic hormone administration and can be inhibited by an antiphosphotyrosine monoclonal antibody but is not blocked by a polyclonal antiserum to Stat1, Stat3, or Stat4, even though the nucleotide sequence contains two gamma interferon-activated sequence-like elements that could interact with STAT proteins. By Southwestern (DNA-protein) blot analysis, approximately 41- and 35-kDa GH-inducible proteins were detected in hepatic nuclear extracts with the Spi 2.1 DNA probe. Thus, a GH-activated signaling pathway stimulates Spi 2.1 gene expression through a unique mechanism that does not appear to involve known members of the STAT family of transcription factors.

Islet beta-cell function and polymorphism in the 5'-flanking region of the human insulin gene.

The present study investigates the possible relationship between human beta-cell secretory capacity and polymorphism in the 5'-flanking region of the human insulin gene. The glucose potentiation slope was measured in normal and non-insulin-dependent diabetic subjects (NIDDM). This slope, as reported previously (Ward, W. K., et al., Am. J. Physiol. 1984; 246:E405-11), is an index of the ability of hyperglycemia to potentiate the insulin response to arginine and as such is a measure of beta-cell responsiveness to glucose. Restriction enzyme analysis using a human insulin gene probe was performed on leukocyte DNA isolated from the same individuals. We conclude that a 1.6 kb polymorphism in the 5'-flanking region of the human insulin gene in both normal and NIDDM subjects has no association with insulin secretory responses as defined here by the glucose potentiation slope.

Insulin gene analysis in a family with maturity-onset diabetes of the young.

The insulin gene locus has been studied in a large kindred with maturity-onset diabetes of the young (MODY) characterized by hypoinsulinemia. DNA was isolated from peripheral leukocytes of 42 family members and 5 spouses. A highly polymorphic region in the 5'-flanking portion of the human insulin gene provided an opportunity for linkage analysis. The presence of three different length polymorphisms of + 1600 base pairs (bp), - 50 bp, and - 150 bp different from the common size allele allowed haplotype assignment of insulin alleles. The hypothesis of linkage was tested by calculating the log of the ratio of the likelihood of the hypothesis of linkage to that of the hypothesis of nonlinkage (LOD score) at a given recombination distance between the insulin polymorphism and the diabetes locus. At a recombination frequency of 0.0, the LOD score was - 14.50 and, therefore, the hypothesis of tight linkage can be strongly rejected. This report is the third study of the relationship between the insulin locus and MODY; however, it is the first report in which a formal linkage analysis indicates with a high degree of probability no linkage between the insulin locus and hypoinsulinemia in a family. Because MODY is a heterogeneous disorder, it may be that different genotypes result in a composite phenotype. The lack of linkage between an insulin allele and MODY in a total of four families studied, however, suggests that the insulin locus is probably not a marker for the MODY phenotype. These results do not exclude the possibility that the insulin locus may be involved in the etiology of other forms of NIDDM.

Lack of association of the polymorphic locus in the 5'-flanking region of the human insulin gene and diabetes in American blacks.

A polymorphic region 5' to the human insulin gene has been associated with diabetes in earlier studies. This polymorphic region is composed of tandem repeats that fall into 3 general size classes, designated class 1 (600 base pairs), class 2 (1300 base pairs), and class 3 (2500 base pairs). Frequencies of these classes of alleles vary among racial groups. American Blacks have been underrepresented in published studies of insulin gene polymorphism and diabetes. We undertook a cooperative study between two centers (San Francisco and St. Louis) to determine geno-types at the insulin locus in 313 unrelated American Blacks (132 nondiabetic, 27 with IDDM, and 154 with NIDDM). In both centers, nondiabetic individuals were younger and leaner than NIDDM patients. Allelic and genotypic frequencies at the insulin locus were not different between the two centers. Class 1 alleles represented 60% of all alleles, class 2 alleles 11%, and class 3 29%. No class of insulin allele was associated with NIDDM in this study. Subdivision of the study population by obesity, family history, or age at diagnosis failed to detect a subgroup for which the insulin allele was associated with NIDDM. Only 27 IDDM individuals were studied, and no significant association of class 1 alleles with this group was noted. However, examination of more IDDM individuals is required before a definitive statement can be made. Fasting serum triglyceride levels were determined retrospectively in 50 NIDDM individuals. No differences in triglyceride levels among genotypes were noted. The frequency of class 3 alleles in 13 hypertriglyceridemic NIDDM subjects was not different from that of the whole group.(ABSTRACT TRUNCATED AT 250 WORDS)

Analysis of insulin gene expression in human pancreas.

The purpose of these studies was to determine whether insulin gene expression at the level of proinsulin mRNA could be studied in human pancreas. RNA was isolated from autopsy specimens and analyzed by RNA-blot hybridization with various 32P-human insulin gene probes spanning either the entire gene or the second intervening sequence. A major band 0.62 kilobases (kb) in length accounted for over 95% of the mRNA, consistent in size with presumed mature proinsulin mRNA. In addition, minor bands of 1.5 and 1.3 kb were seen, consistent with an initial gene transcript containing both intervening sequences and with a processed intermediate. The 1.5- and 1.3-kb RNA were confirmed to be proinsulin mRNA precursors by hybridization specifically with the IVS II probe. Total RNA and polyadenylated RNA from five normal pancreata and two insulinomas revealed the same pattern. This method provides a means of determining whether altered insulin gene expression is one cause of diabetes.

Structure and expression of a chicken insulin-like growth factor I precursor.

Insulin-like growth factor I (IGF-I) is a 70 amino acid growth-promoting polypeptide whose sequence and functions have been highly conserved among mammals. As an initial step in defining the role of IGF-I in other vertebrate species, we have isolated and characterized an IGF-I cDNA from the chicken. This cDNA encodes a 153 amino acid primary translation product which resembles in structure and sequence the IGF-IA protein of mammals. There is strong amino acid conservation between chicken and mammalian IGF-I throughout the entire protein. Sixty of 70 amino acids are identical in mature IGF-I among the chicken, rat, and human peptides, with five differences being localized to the C domain, and two to the D region. A comparable degree of amino acid identity is found in the COOH-terminal extension peptide (28/35 residues). At the NH2-terminus, where there is more amino acid divergence (32/48 identities), the most 5'-AUG codon is the only methionine residue conserved among all three species, suggesting that it functions as the authentic translation initiation site, an observation supported by cell-free studies of biosynthesis and cotranslational proteolytic processing. The pattern of IGF-I gene expression appears to be simpler in chickens than in mammals, since a single predominant mRNA of 2.6 kilobases can be detected in liver polyadenylated RNA on Northern blots. In the chicken, as in rats and humans, IGF-I mRNA is synthesized in multiple tissues, including liver, brain, skeletal muscle, and heart.(ABSTRACT TRUNCATED AT 250 WORDS)

Evolution of insulin-like growth factor I (IGF-I): structure and expression of an IGF-I precursor from Xenopus laevis.

By means of a cloning strategy employing the polymerase chain reaction, we have isolated and characterized cDNAs for Xenopus laevis insulin-like growth factor I (IGF-I). These cDNAs encode a primary IGF-I translation product of 153 residues that demonstrates considerable amino acid sequence similarity with IGF-IA peptides from other species. Fifty-seven of 70 residues of the mature protein are identical among human, rat, chicken, and Xenopus IGF-I, while less amino acid conservation is found at the COOH-terminus (25/35 identities) or at the NH2-terminus (24/48 identities) of the precursor protein. Despite the lower degree of structural similarity at the NH2-terminus, in vitro studies of IGF-I biosynthesis and proteolytic processing support a conserved function for the atypically long 48 residue NH2-terminal signal sequence in directing the nascent IGF-I peptide through the secretory pathway. The 5'-untranslated region of Xenopus IGF-I mRNA matches the human, rat, and chicken sequences in greater than 90% of 279 nucleotides. IGF-I mRNAs from all four species encode a conserved upstream open reading frame of 14 amino acids starting 240-250 nucleotides 5' to the translation start site, suggesting a possible role for this region in modulating IGF-I gene expression. The X. laevis IGF-I gene is transcribed and processed into three mRNAs of 1.6, 2.1, and 3.0 kilobases in liver, and IGF-I mRNAs can be detected in liver, lung, heart, kidney, and peritoneal fat of adult animals. These studies demonstrate that both the IGF-I protein precursor and potential regulatory regions of IGF-I mRNA have been conserved during vertebrate evolution, and indicate that like several other polypeptide growth factors, IGF-I may be of fundamental importance in regulating specific aspects of growth and development in all vertebrates.

Transcriptional activation of the insulin-like growth factor-II gene during myoblast differentiation.

Insulin-like growth factor-II (IGF-II) is a secreted 67-amino acid peptide that functions principally as a prenatal growth regulator in mammals. To date, the mechanisms involved in the stimulation of IGF-II expression in the embryo and its attenuation after birth are unknown. Recent studies have shown that IGF-II mRNA and protein are induced during the terminal stages of muscle development in vitro, and that IGF-II may act as an autocrine differentiation agent for skeletal myoblasts. We now have investigated the regulation of IGF-II gene expression in muscle cells. Steady state levels of IGF-II mRNA increased by more than 30-fold during terminal differentiation of the C2 mouse myoblast cell line. Transcription run-on experiments using isolated muscle cell nuclei and direct analysis of nuclear RNA each demonstrated a greater than 10-fold rise in nascent IGF-II mRNA during cellular differentiation, and ribonuclease protection experiments showed that more than 95% of IGF-II mRNAs initiated in noncoding exon 3, implying that transcriptional activation occurs principally through promoter 3, the most 3' of the three mouse IGF-II gene promoters. Analysis of chromatin structure around the IGF-II gene in C2 cells revealed four major and four minor DNase-I-hypersensitive sites, but did not provide insight into the mechanisms of gene activation, since all sites were present in proliferating and differentiating cells. Gene transfer experiments showed that promoter 3 was at least 50-fold more active than promoter 1 or 2 in C2 cells, but the functional assessment of nearly 26 kilobases of additional DNA within the IGF-II locus by an "enhancer trap" approach did not delineate any chromosomal regions capable of mediating differentiation-specific gene activation. Our results demonstrate that muscle cells encode mechanisms for activating IGF-II gene transcription and suggest that these cells may be excellent models for identifying the developmentally regulated factors that control IGF-II gene expression.

A novel human insulin-like growth factor I messenger RNA is expressed in normal and tumor cells.

We have identified and characterized a novel human insulin-like growth factor I (IGF-I) precursor from the transplantable T61 human breast cancer xenograft and from normal liver. The mRNA encoding this precursor contains a 5'-untranslated region that is 83% identical to the corresponding region of a previously described variant rat IGF-I. The nucleotide sequence of the cloned cDNA predicts an IGF-IA protein precursor of 137 amino acids, including a 32 residue signal peptide, 70 amino acid IGF-I, and a 35 residue COOH-terminal extension or E peptide. The exon encoding this variant maps in the genome between IGF-I exons 1 and 2, in a similar location to the homologous rat exon 1a. The rat and human exons 1a are 59% identical over 1443 nucleotides, with DNA sequence conservation occurring in a mosaic pattern. Human IGF-I mRNAs encoding this novel exon are expressed in liver, T61 tumor cells, and in an ovarian carcinoma cell line, NIH OVCAR3. These studies demonstrate that as in the rat, the human IGF-I gene contains six exons that are variably processed into multiple IGF-I mRNAs. The mechanisms responsible for generating different IGF-I mRNAs thus appear to be conserved among mammalian species.

Growth hormone rapidly activates insulin-like growth factor I gene transcription in vivo.

Many of the growth-promoting properties of GH are mediated by insulin-like growth factor I (IGF-I), a highly conserved circulating 70-amino acid peptide. Recent studies have shown that multiple mechanisms influence IGF-I gene expression, including transcription from two promoters, alternative RNA splicing, and variable polyadenylation. In order to determine how GH regulates IGF-I gene expression we have analyzed the response of hypophysectomized rats to a single ip injection of recombinant GH. A rise in hepatic IGF-I mRNA was detected within 2 h of GH treatment, with peak values of more than 15-fold above untreated animals by 4 h, and a decline by 16 h. A coordinate increase was seen in all IGF-I mRNA splicing and polyadenylation variants, indicating that neither alternative RNA processing nor differential poly A addition were altered by GH. Transcription run-on experiments using isolated hepatic nuclei and direct analysis of nuclear RNA demonstrated a rise in nascent IGF-I mRNA within 30 min of GH treatment, with peak levels reaching more than 10-fold above background by 2 h and declining by 6 h. As determined by RNase protection assays, transcripts directed by each promoter were coordinately and equivalently activated after GH. A single GH-responsive DNase I hypersensitive site was mapped in chromatin to the second IGF-I intron. This site exhibited rapid kinetics of induction which mirrored the pattern of transcriptional stimulation after GH treatment. These experiments show that GH enhances IGF-I expression in vivo by predominantly transcriptional mechanisms. The rapid kinetics of IGF-I gene activation and the temporally associated chromatin changes demonstrate a direct link between a GH-dependent signal transduction pathway and nuclear events.