F. John Ballard

Des(1–3)IGF-I: a truncated form of insulin-like growth factor-I

Des(1-3)IGF-I, a truncated variant of human IGF-I with the tripeptide Gly-Pro-Glu absent from the N-terminus, has been isolated from bovine colostrum, human brain and porcine uterus. This protein probably results from post-translational cleavage of IGF-I. Des(1-3)IGF-I generally is about 10-fold more potent than IGF-I at stimulating hypertrophy and proliferation of cultured cells, a consequence of much reduced binding to IGF-binding proteins, in turn caused by the absence of the glutamate at position 3. The increased potency is retained in part when the variant is administered in vivo, with selective anabolic effects particularly evident in gut tissues. Clinical opportunities for des(1-3)IGF-I have not yet been evaluated, but could apply in catabolic states as well as for the treatment of inflammatory bowel diseases.

Milk-derived growth factors as serum supplements for the growth of fibroblast and epithelial cells

SummaryWe have investigated the response of several epithelial and fibroblastic cells to a mitogenic extract of bovine milk. Cation exchange chromatography was used to produce a mitogen-rich fraction from an industrial whey source that, although comprising only 0.5% of total whey protein, contained the bulk of the growth factor activity. This fraction was a source of potent growth promoting activity for all mesodermal-derived cells tested, including human skin and embryonic lung fibroblasts, Balb/c 3T3 fibroblasts, and rat L6 myoblasts. Maximal growth of all these cell types exceeded that observed in 10% fetal bovine serum. Feline kidney and baby hamster fibroblasts and Chinese hamster ovary cells were less responsive, achieving a maximal growth response of 50–75% that observed in 10% fetal bovine serum. Maximal growth achieved in whey-extract-supplemented cultures of Balb/c 3T3 and human skin fibroblasts, and L6 myoblast cultures exceeded that seen in response to recombinant acidic or basic fibroblast growth factor, platelet-derived growth factor, insulin-like growth factor, or epidermal growth factor. Importantly, addition of low concentrations of fetal bovine serum to the whey-derived mitogenic fraction produced an additive response. However, concentrated milk-derived factors were found to be inhibitory to the growth of all epithelial lines tested, including rat intestinal epithelial cells, canine kidney epithelial cells, and mink lung cells. It is concluded that industrial whey extracted in this form constitutes an important source of potent growth-promoting agents for the supplementation of mesodermal-derived cell cultures.

Probing the disulfide folding pathway of insulin-like growth factor-I

The crucial step of folding of recombinant proteins presents serious challenges to obtaining the native structure. This problem is exemplified by insulin-like growth factor (IGF)-I which when refolded in vitro produces the native three-disulfide structure, an alternative structure with mispaired disulfide bonds and other isomeric forms. To investigate this phenomenon we have examined the refolding properties of an analog of IGF-I which contains a 13-amino acid N-terminal extension and a charge mutation at position 3 (Long-[Arg3]IGF-I). Unlike IGF-I, which yields 45% of the native structure and 24% of the alternative structure when refolded in vitro, Long-[Arg3]IGF-I yields 85% and 10% of these respective forms. To investigate the interactions that affect the refolding of Long-[Arg3]IGF-I and IGF-I, we acid-trapped folding intermediates and products for inclusion in a kinetic analysis of refolding. In addition to non-native intermediates, three native-like intermediates were identified, that appear to have a major role in the in vitro refolding pathway of Long-[Arg3]IGF-I; a single-disulfide Cys18-Cys61 intermediate, an intermediate with Cys18-Cys61 and Cys6-Cys48 disulfide bonds and another with Cys18-Cys61 and Cys47-Cys52 disulfide bonds. Furthermore, from our kinetic analysis we propose that the Cys18-Cys61, Cys6-Cys48 intermediate forms the native structure, not by the direct formation of the last (Cys47-Cys52) disulfide bond, but by rearrangement via the Cys18-Cys61 intermediate and a productive Cys18-Cys61, Cys47-Cys52 intermediate. In this pathway, the last disulfide bond to form involves Cys6 and Cys48. Finally, we apply this pathway to IGF-I and conclude that the divergence in the in vitro folding pathway of IGF-I is caused by non-native interactions involving Glu3 that stabilize the alternative structure.

Clearance of IGFs and insulin from wounds: effect of IGF-binding protein interactions

We have examined the role binding proteins have in regulating the clearance of exogenous growth factors from wounds. Hunt-Schilling chambers were subcutaneously implanted in rats, and the clearance of insulin-like growth factor (IGF) I from the chamber wound fluid was compared with IGF-II, LR3-IGF-I, which binds poorly to IGF-binding proteins (IGFBP), or insulin. Elimination rate constants of the slow phase of the decay curves did not differ between IGF-I and IGF-II. However, LR3-IGF-I and insulin were cleared more rapidly from wound fluid than IGF-I so that the half-lives for IGF-I, IGF-II, LR3-IGF-I, and insulin were 872, 861, 563, and 324 min, respectively. In wound fluid, minimal degradation of the IGFs occurred, whereas insulin was degraded considerably. The increased clearance of LR3-IGF-I and insulin equated with a reduced association with wound fluid IGFBPs, and increased amounts of radioactivity of these peptides were detected in the circulation and urine. These results show that this model of wound repair may be of use in examining the kinetics of growth factors and other bioactive molecules in extravascular spaces and support the hypothesis that IGFBPs can be significant regulators of IGF bioavailability in vivo.We have examined the role binding proteins have in regulating the clearance of exogenous growth factors from wounds. Hunt-Schilling chambers were subcutaneously implanted in rats, and the clearance of insulin-like growth factor (IGF) I from the chamber wound fluid was compared with IGF-II, LR3-IGF-I, which binds poorly to IGF-binding proteins (IGFBP), or insulin. Elimination rate constants of the slow phase of the decay curves did not differ between IGF-I and IGF-II. However, LR3-IGF-I and insulin were cleared more rapidly from wound fluid than IGF-I so that the half-lives for IGF-I, IGF-II, LR3-IGF-I, and insulin were 872, 861, 563, and 324 min, respectively. In wound fluid, minimal degradation of the IGFs occurred, whereas insulin was degraded considerably. The increased clearance of LR3-IGF-I and insulin equated with a reduced association with wound fluid IGFBPs, and increased amounts of radioactivity of these peptides were detected in the circulation and urine. These results show that this model of wound repair may be of use in examining the kinetics of growth factors and other bioactive molecules in extravascular spaces and support the hypothesis that IGFBPs can be significant regulators of IGF bioavailability in vivo.

Growth hormone but not insulin-like growth factor-I improves wound strength in pigs.

Systemic growth hormone and locally administered insulin‐like growth factor‐I have been shown in a number of studies to improve the breaking strength of incisional wounds, especially in compromised animals. The objective of the present study was to compare these two agents when administered subcutaneously distant from an incisional wound site in pigs, as well as to examine effects of a combination growth hormone/insulin‐like growth factor treatment. Growth hormone was shown to increase wound breaking strength in two experiments, whereas insulin‐like growth factor‐I or a more potent analog had no effect. Moreover, breaking strength was only minimally improved above the vehicle groups by the combination of growth hormone and insulin‐like growth factor‐I. These effects could not be explained by changes in plasma insulin‐like growth factor‐I concentrations which were highest in the combination groups, nor by plasma insulin‐like growth factor binding protein‐3 which was raised equally whenever growth hormone was administered. We conclude that systemic growth hormone but not insulin‐like growth factor‐I improves wound strength in normal pigs, whereas insulin‐like growth factor‐I reduces the magnitude of the growth hormone effect by an unknown mechanism.

Optimization of the hydroxylamine cleavage of an expressed fusion protein to produce recombinant human insulin-like growth factor (IGF)-I.

The application of gene fusion technology for the production of heterologous proteins in Escherichia coli has required the development of specific cleavage methods to separate the coexpressed fusion protein partner from the protein of interest. When hydroxylamine is used to cleave Asn–Gly fusion protein linkages, undesirable chemical modification of asparagine and glutamine amino acids can also occur. In this study, hydroxylamine cleavage conditions were modified to minimize unwanted chemical heterogeneity that occurred during the cleavage of the fusion protein [Met1]‐pGH(1‐11)‐Val‐Asn‐IGF‐I (Long‐IGF‐I). The cleavage reaction was shown to be dependent on the hydroxylamine concentration, temperature, and pH. Optimal cleavage conditions were identified that resulted in very low levels of chemical heterogeneity, but under these mild conditions that cleavage of the labile Asn–Gly bond was reduced. Therefore, the reaction was further modified to improve the yield of IGF‐I while minimizing chemical heterogeneity. The yield of unmodified IGF‐I was improved from less than 25% to greater than 70%. Analysis of the heterogeneity produced using the modified cleavage technique showed that Asn26 was converted to a hydroxamate. This variant was characterized in refolding and biological assays where it was equivalent to IGF‐I. To further assess the effectiveness of the modified cleavage technique and to evaluate the potential for process scale‐up, a gram‐scale cleavage reaction of Long‐IGF‐I was carried out. The process yielded IGF‐I with a low level of chemical heterogeneity that was easily removed by ion‐exchange chromatography. Moreover, this work shows that the production of unmodified IGFs using hydroxylamine cleavage of fusion proteins is facilitated using the mild cleavage reaction.

In vitro characterization and in vivo clearance of recombinant barramundi (Lates calcarifer) IGF-I

Little is known about fish insulin-like growth factors (IGFs) as only small amounts have been isolated from native sources or indeed produced recombinantly. This report describes the production of milligram quantities of recombinant barramundi IGF-I (bIGF-I) and its subsequent characterization. Recombinant bIGF-I was produced in Escherichia coli using a gene fusion system similar to that previously described for the production of other non-mammalian IGFs. Recombinant bIGF-I was similar to human IGF-I (hIGF-I) in stimulating protein synthesis and in competing for binding of labelled hIGF-I to IGF receptors whether tested in rat myoblasts or in salmon embryo fibroblasts. However, recombinant bIGF-I differed from its human counterpart in its affinity for a polyclonal antibody raised against hIGF-I, with at least 200-fold more bIGF-I required to obtain 50% displacement of labelled hIGF-I from the antibody. Hence, the recombinant protein will be essential for developing a specific homologous immunoassay for measuring IGF-I concentrations in barramundi during growth and development. In addition, studies investigating the clearance of labelled bIGF-I and hIGF-I in vivo reveal that the human protein is cleared from the circulation of juvenile barramundi almost twice as fast as the barramundi protein, thus providing the first in vivo evidence that there are functional differences between fish and human IGF-Is. Neutral gel chromatography of serum from the clearance study suggest that this is due to differences in the affinities of the labelled human and fish IGF-I for the IGFBPs present in barramundi.