Insulin-like growth factor-I (IGF-I) and IGF-II have been purified to homogeneity from kangaroo (Macropus fuliginosus) serum, thus this represents the first report of the purification, sequencing and characterisation of marsupial IGFs. N-Terminal protein sequencing reveals that there are six amino acid differences between kangaroo and human IGF-I. Kangaroo IGF-II has been partially sequenced and no differences were found between human and kangaroo IGF-II in the 53 residues identified. Thus the IGFs appear to be remarkably structurally conserved during mammalian radiation. In addition, in vitro characterisation of kangaroo IGF-I demonstrated that the functional properties of human, kangaroo and chicken IGF-I are very similar. In an assay measuring the ability of the proteins to stimulate protein synthesis in rat L6 myoblasts, all IGF-I proteins were found to be equally potent. The ability of all three proteins to compete for binding with radiolabelled human IGF-I to type-1 IGF receptors in L6 myoblasts and in Sminthopsis crassicaudata transformed lung fibroblasts, a marsupial cell line, was comparable. Furthermore, kangaroo and human IGF-I react equally in a human IGF-I RIA using a human reference standard, radiolabelled human IGF-I and a polyclonal antibody raised against recombinant human IGF-I. This study indicates that not only is the primary structure of eutherian and metatherian IGF-I conserved, but also the proteins appear to be functionally similar.

Kangaroo IGF-II has been purified from western grey kangaroo (Macropus fuliginosus) serum and characterised in a number of in vitro assays. In addition, the complete cDNA sequence of mature IGF-II has been obtained by reverse-transcription polymerase chain reaction. Comparison of the kangaroo IGF-II cDNA sequence with known IGF-II sequences from other species revealed that it is very similar to the human variant, [Ser29]-hIGF-II. Both the variant and kangaroo IGF-II contain an insert of nine nucleotides that encode the amino acids Leu-Pro-Gly at the junction of the B and C domains of the mature protein. The deduced kangaroo IGF-II protein sequence also contains three other amino acid changes that are not observed in human IGF-II. These amino acid differences share similarities with the changes described in many of the IGF-IIs reported for non-mammalian species. Characterisation of human IGF-II, kangaroo IGF-II, chicken IGF-II and [Ser29]-hIGF-II in a number of in vitro assays revealed that all four proteins are functionally very similar. No significant differences were observed in the ability of the IGF-IIs to bind to the bovine IGF-II/cation-independent mannose 6-phosphate receptor or to stimulate protein synthesis in rat L6 myoblasts. However, differences were observed in their abilities to bind to IGF-binding proteins (IGFBPs) present in human serum. Kangaroo, chicken and [Ser29]-hIGF-II had lower apparent affinities for human IGFBPs than did human IGF-II. Thus, it appears that the major circulating form of IGF-II in the kangaroo and a minor form of IGF-II found in human serum are structurally and functionally very similar. This suggests that the splice site that generates both the variant and major form of human IGF-II must have evolved after the divergence of marsupials from placental mammals.

Abstract

While numerous researchers have used rat models to investigate the in vivo actions of IGF-I, interpretation of the results in terms of true concentrations of rat IGF-I (rIGF-I) in plasma has been hampered by the absence of homologous reference standards. In order to overcome this we have produced recombinant rIGF-I (rrIGF-I) from Escherichia coli using procedures similar to those we have previously described for the production of other recombinant IGFs. The rrIGF-I is indistinguishable from serum-derived rIGF-I when characterized in a number of in vitro assays including ability to stimulate protein synthesis and inhibit protein degradation in cultured rat cells, as well as in interactions with the rat type-1 IGF receptor and with rat IGF-binding proteins. Moreover, both the serum-derived and the recombinant rat proteins are similar to recombinant human IGF-I (rhIGF-I) in these assays. However, differences between the human and rat IGFs are apparent when tested in immunoassays using some antibodies raised against rhIGF-I. Furthermore, the differences between rhIGF-I and rrIGF-I are even greater when rhIGF-I is used as the competing radiolabel in these assays, a situation that can lead to a two- to threefold underestimation of the actual concentration of IGF-I in rat plasma. These results indicate that, while immunoassays employing antibodies raised against rhIGF-I and rhIGF-I reference standards reliably indicate trends in IGF-I concentrations in rat plasma, the true amounts of rIGF-I present can only be assured in an assay using homologous tracer and reference peptides.

Journal of Endocrinology (1996) 149, 379–387

Abstract

The development of a homologous radioimmunoassay (RIA) for chicken insulin-like growth factor-I (cIGF-I) and its use to investigate the developmental changes in IGF-I in the chicken and turkey is described. A doubleantibody RIA has been developed using recombinantly derived cIGF-I as antigen, radiolabelled tracer and standard. The resulting immunoassay has a minimum detection limit of 0·035 ng and effective dose of 2·5 ng. Dose–response curves of chicken and turkey plasma and tissue extracts were parallel with cIGF-I standard. The antiserum is specific for IGF-I as no cross-reactivity with chicken IGF-II, insulin, glucagon, gastrin or avian pancreatic polypeptide was observed. We have also established that acid/ethanol extraction of chicken and turkey plasma reduced possible interference of IGF-binding proteins (IGFBPs) in the RIA. Comparison of IGF-I immunoactivity in unextracted and acid/ethanol-extracted samples following gel filtration under acidic and neutral conditions indicates that the cIGFBPs may be acid-labile. Analyses of samples from growing chickens and turkeys using the homologous avian reagents revealed higher IGF-I concentrations than if the IGF were quantified using heterologous mammalian-derived reagents. A similar pattern was observed when tissue extracts were assayed for IGF-I content. The application of the homologous RIA to monitor blood and tissue IGF-I levels during embryonic development and posthatch growth in avian species will provide more accurate comparisons of results from studies on the role of IGF-I in growth and metabolism of domestic birds.

Journal of Endocrinology (1994) 142, 225–234

The development of a homologous RIA for chicken insulin-like growth factor-II (cIGF-II) and its application to investigate the developmental changes in IGF-II in the chicken and turkey are described. A double-antibody RIA has been developed using recombinantly derived cIGF-II as antigen, radiolabelled tracer and standard. Serial dilutions of chicken and turkey plasma were parallel to serial dilutions of cIGF-II standard. We have also established that acid/ethanol extraction of chicken and turkey plasma reduced possible interference of insulin-like growth factor-binding proteins in the RIA. Consumption of a low-protein diet by male chickens lowered plasma IGF-I twofold, whereas IGF-II levels were unchanged. Food withdrawal evoked an increase in circulating IGF-II, while IGF-I levels were reduced. Refeeding returned both growth factors to normal circulating concentrations. During chick embryo incubation, plasma IGF-II levels were tenfold higher than those of IGF-I. In the turkey embryo, plasma IGF-II concentrations were higher than those of IGF-I. During the post-hatch period. IGF-II levels declined with age in chickens. In the growing turkey, IGF-II levels were consistently higher than IGF-I levels. The application of the homologous RIA to monitor plasma levels during embryonic development and post-hatch growth in avian species will provide more accurate comparisons of results from studies on the role of IGF-II in growth and metabolism of domestic birds.

Abstract

We have examined the influence of nutrition on plasma IGF-I, IGF-II and IGF-binding protein (IGFBP) levels and on hepatic IGF-I gene expression in young meat-type chickens. Plasma IGF concentrations were measured by using RIA with recombinant chicken IGFs as standards.

In chickens fed the control diet containing 200 g/kg dietary protein ad libitum for 7 days, plasma IGF-I concentrations increased significantly from those found in the initial control group. Food restriction for either 4 or 7 days decreased plasma IGF-I by 30% from the initial control. When chickens were refed ad libitum for 3 days after 4 days of restricted feeding, plasma IGF-I levels recovered to those of the control birds fed ad libitum. In chickens eating a low protein diet (100 g/kg protein), the plasma IGF-I tended to be lowered but the decrease was not significant. Although the intensity of IGF-I and β-actin mRNA bands protected in the RNase protection assay was changed by nutrition, no statistical effect of nutrition on the ratio of IGF-I to β-actin was observed. The nutritional treatments had no effect on plasma IGF-II concentrations.

Western ligand blot and chromatographic analyses were used to investigate the influence of nutrition on IGFBP profiles. Both IGF-I and IGF-II ligands in the Western ligand blot revealed the most intense binding at 30 kDa for plasma obtained from chickens with restricted food intake. The 30 kDa band also appeared at a lower intensity in the group fed a low protein diet but not in any other groups. These observations were confirmed by neutral gel chromatography. The chicken IGF-II ligand revealed an intensely labelled band corresponding to 75 kDa and this was not affected by nutrition.

IGF-I and IGFBP concentrations in the plasma of young broiler chickens were influenced by nutritional state but IGF-II concentrations were not. The lack of a response in circulating IGF-II levels may have been due to the presence of high concentrations of a 75 kDa specific binding protein which did not respond to nutrition in this experiment.

Journal of Endocrinology (1996) 149, 181–190

Abstract

The metabolic clearance of chicken IGF-I (cIGF-I), cIGF-II, human IGF-I (hIGF-I), and hIGF-II was examined in the chicken using ¹²⁵I-labelled growth factors. Superose-12 chromatography of plasma collected at 7·5 min post-infusion revealed peaks of radioactivity corresponding to 150 and 43 kDa and unbound tracer. Statistical analysis of trichloroacetic acid (TCA)-precipitable radioactivity in sequential plasma samples as well as following chromatography of the same samples revealed that clearance of the radiolabelled peptides followed an apparent triphasic pattern. The close similarity of the individual chromatographically defined pools in their clearance rate compared with the three components described by TCA precipitation strongly suggested their identity. Both free ¹²⁵I-labelled cIGF-II (3·11 min) and hIGF-II (3·01 min) were cleared at a greater rate than their IGF-I counterparts. Unbound hIGF-I was cleared at a greater rate than cIGF-I (4·45 vs 5·66 min respectively). A similar pattern for clearance was evident in the radiolabelled growth factors associated with the 43 kDa component, although at a longer half-life. There was no difference in the apparent clearance of the radiolabelled growth factors associated with the 150 kDa component between IGF-I or -II or between species. Analysis of the chromatographic profiles of radioactive IGF-I peptides complexed to serum proteins versus those bound to labelled IGF-II peptides revealed the presence of a large molecular mass binding protein in vivo. Ligand blotting of chicken serum determined that a binding protein with a mass of 70 kDa was detectable with ¹²⁵I-IGF-II probes only, and was not present in pig serum. In addition, tissue uptake of ¹²⁵I-cIGF-I and -II was evaluated. Similar patterns of tissue distribution and uptake were observed for ¹²⁵I-cIGF-I and -II, except that cIGF-II uptake by the liver exceeded that of ¹²⁵I-cIGF-I at 15 min post-infusion. The rank order of tissue distribution was as follows: kidney > testis > heart > liver > pancreas > small intestine> cartilage > bursa > gizzard > leg muscle > breast muscle > brain. We conclude from these studies that the clearance of IGFs from the compartments identified in blood and the potential target tissues is dependent on their interactions with IGF-binding proteins and receptors.

Journal of Endocrinology (1996) 150, 149–160