To explore the structural differences in various animal somatomedins, we examined the sera of 20 animal species using two region-specific radioimmunoassays, a radioimmunoassay specific for somatomedin-C and a placental membrane radioreceptor assay. Using small peptide portions of the insulin-like growth factor-I (IGF-I) C chain and D chain made by solid-state methods, we generated antisera in rabbits and developed radioimmunoassays to these regions. For a radioimmunoassay directed against intact somatomedin-C, we used the antibody distributed by the NIH (NIH-radioimmunoassay). For the detection of somatomedin peptide content, we used a placental membrane radioreceptor assay. ¹²⁵I-Labelled IGF-I was used as the radioligand in all assays and an insulin-free, partially purified somatomedin preparation was used for the standard curves. All samples were chromatographed in 0·25 m-formic acid to remove any somatomedin-binding protein before assay.

The correlation between the radioreceptor assay and the NIH-radioimmunoassay was good (r ² = 0·84) but the slope was significantly (P<0·001) different from a value of 1 which would be expected if the two assays were detecting equal amounts of somatomedin activity. In all the species, the somatomedin level measured by radioreceptor assay was greater than that measured by the NIH-radioimmunoassay, suggesting the presence of a somatomedin-like protein which has retained its ability to bind to the somatomedin receptor but which is sufficiently different to give it reduced affinity in the NIH-radioimmunoassay.

The somatomedins of rat, hamster and mouse (all rodents of the superfamily Myomorpha) were not detectable using the radioimmunoassays for the C and D chains of IGF-I, although they were high in the radioreceptor assay. This suggests that these regions have a different structure from the analogous regions in the guinea-pig (a rodent of the superfamily Caviomorpha) and implies a significant structural change in a somatomedin-like molecule during the evolution of the order Rodentia. This is analogous to the marked differences in the known sequence of the pro-insulins in these two superfamilies.

Abstract

We have previously reported the presence of IGF-I and IGF-binding proteins (IGFBP-2, -3 and -4) in rat milk. Herein, the potential sources of rat milk IGF-I and IGFBPs were investigated. Lactating dams (day 14 postpartum) were separated from their pups and injected intraperitoneally with 0·45 μCi ¹²⁵I-IGF-I or ¹²⁵I-IGFBP-3. After 3 h, serum and milk of rats receiving ¹²⁵I-IGF-I contained 7642 ± 3121 and 14 455 ± 7837 c.p.m./ml respectively. Serum and milk of rats given ¹²⁵I-IGFBP-3 contained 7232 ± 1366 and 10 371 ± 4091 c.p.m./ml respectively. Sephacryl S-200 gel filtration chromatography demonstrated that the ¹²⁵I-IGF-I in both serum and milk was primarily in the 150 kDa IGF-binding complex, whereas the distribution of ¹²⁵I-IGFBP-3 differed between serum and milk. In serum, most of the ¹²⁵I-IGFBP-3 was in the 150 kDa fraction, while most ¹²⁵I-IGFBP-3 in milk was in the 40 kDa fraction. Northern analysis of liver showed IGFBP-1 and -3 mRNA expression, with variable expression of IGFBP-2 and -4 mRNA. In contrast, mammary tissue expressed only IGFBP-2 and -4 mRNA, suggesting that these IGFBPs in milk may arise from de novo synthesis within the mammary gland. The lack of detectable IGFBP-3 mRNA in mammary tissue and the translocation of ¹²⁵I-IGFBP-3 from the serum suggest that milk IGFBP-3 arises from the maternal circulation.

Journal of Endocrinology (1995) 145, 569–578

ABSTRACT

In order to obtain a phenotypically stable cell population of chondrocytes, high density primary monolayer cultures of bovine articular chondrocytes were established. Using these cultures, a specific insulin-like growth factor-I/somatomedin-C (IGF-I/SM-C) receptor was demonstrated and characterized. At 15 °C steady-state binding was attained by 5 h, and averaged 25% per 2·2 × 10⁶ cells. Fifty per cent displacement of ¹²⁵I-labelled IGF-I/SM-C by unlabelled IGF-I/SM-C occurred at concentrations of only 2·3 ng/ml, whereas IGF-II and porcine insulin were approximately 15-and 1000-fold less potent respectively. Scatchard analysis gave a linear plot, with a calculated association constant of 2·26 × 10⁹ l/mol and a receptor number of 15 400 sites per cell.

Preincubation of chondrocyte monolayers with either IGF-I/SM-C or porcine insulin at 37 °C for 20 h resulted in reduction of ¹²⁵I-labelled IGF-I/SM-C binding in a dose-dependent manner, although higher concentrations were required with insulin. More than 40% down-regulation of the receptor occurred with IGF-I/SM-C at concentrations of 10 nmol/l and nearly 70% reduction at 50 nmol/l. Interestingly, after preincubation with either human (h) or bovine (b)GH, 40% down-regulation of ¹²⁵I-labelled IGF-I/SM-C binding was observed at concentrations of 10 μmol/l. Local production of IGF-I/SM-C by chondrocytes in response to GH stimulation may have occurred, but, because only 120 pmol IGF-I/SM-C and < 30 pmol IGF-I/SM-C per litre were recovered from serum-free conditioned media preincubated with bGH and hGH respectively, this was not established.

These studies demonstrate that cultured bovine articular chondrocytes possess a highly specific IGF-I/SM-C receptor, and that this receptor population is regulated not only by IGF-I/SM-C and insulin but also by high concentrations of either hGH or bGH. These results are consistent with the growth-promoting action of IGF-I/SM-C on skeletal tissues, and suggest the possibility that GH itself may play its own role to modulate IGF-I/SM-C receptors on chondrocytes.

J. Endocr. (1985) 107, 275–283

ABSTRACT

Castrated prepubertal lambs were hypophysectomized and then treated with GH and testosterone either alone or in combination over a series of 3-week treatment periods. Hypophysectomy resulted in a rapid reduction in skeletal growth rate which could be reversed by the administration of either GH (4IU three times a week for 3 weeks) or testosterone propionate (10 mg daily for 3 weeks). When GH or testosterone treatment was withdrawn, skeletal growth fell to the post-operative rate. Combined treatment with both GH and testosterone was no more or less effective than either hormone given singly. The order of administration did not have any effect on the growth rate. Circulating concentrations of insulin-like growth factor-I (IGF-I) were reduced by hypophysectomy, but neither GH nor testosterone treatment, alone or in combination, had any effect on IGF-I concentrations. Concentrations of IGF-II rose following hypophysectomy, and again were not affected by any of the hormonal replacement treatments.

In conclusion, both GH and testosterone could stimulate skeletal growth in the hypophysectomized lamb without any alteration of circulating IGF concentrations, and testosterone can clearly stimulate skeletal growth in the complete absence of GH.

Journal of Endocrinology (1989) 121, 563–570

ABSTRACT

Two studies were designed to examine the differences in galactopoietic potency of molecular variants of pituitary- and recombinant-derived bovine GH (bGH). The recombinant bGH molecules included aminoterminal and position-127 amino acid substitutions which are representative of two of the four natural pituitary variants or of partially degraded bGH molecules. Amino-terminal variants of bGH included methionine (Met¹), alanine (Ala¹), serine (Ser¹) or deletion of four amino acids (Δ¹⁻⁴). The Δ^1–4 variants were representative of degradation products previously isolated in pituitary bGH preparations. In the first study, 54 lactating Holstein cows received i.m. injections of a buffer solution (control), pituitaryderived bGH, or recombinant-derived [Met¹, Leu¹²⁷]-bGH, [Met¹, Val¹²⁷]-bGH, [Ala¹, Leu¹²⁷]-bGH, or [Ala¹, Val¹²⁷]-bGH. Cows received 25 mg bGH/day for 21 days. Substitution of the amino-terminal alanyl residue with methionine did not affect milk response. GH variants with Val¹²⁷ elicited a greater milk response (8·5 kg/day) than Leu¹²⁷ bGH variants (6·5 kg/day). The average milk response to the four recombinant bGH variants was 7·5 kg/day greater than controls compared with 4·4 kg/day for pituitary-derived bGH. In contrast, blood bGH concentrations were equivalent for pituitary and recombinant bGH treatments, approximately 20 μg/l more than control levels at 3 h after injection. Blood free fatty acid concentrations were increased, but insulin and glucose levels were unaffected by bGH treatment. In the second study, 54 lactating Holstein cows received i.m. injections of a buffer control solution or recombinant-derived [Met¹,Leu¹²⁷]-bGH, [Ser¹,Leu¹²⁷]-bGH, [Ser¹,Val¹²⁷]-bGH, [Δ^1–4,Leu¹²⁷]-bGH or [Δ^1–4,Val¹²⁷]-bGH. Cows received 25 mg bGH/day for 28 days. The milk response to full-length bGH variants was 6·6 kg/day greater than the response to the amino-terminal deletion variants (P<0·05). Substitution of valine for leucine did not affect milk response to either the deletion (Δ^1–4) or full-length (Met¹ or Ser¹) bGH molecules. In conclusion, the lowered galactopoietic potency of pituitary bGH preparations was demonstrated, at least in part, to be due to the presence of amino-terminal amino acid deletions rather than differences in amino acid sequences of recombinant bGH. Ala¹ bGH variants with valine at postiion 127 elicited a greater milk response than Leu¹²⁷ variants.

Journal of Endocrinology (1992) 132, 47–56

The IGFs are mitogenic agents which are closely linked to regulatory processes in carbohydrate metabolism. Because limited information is available on the occurrence of the IGF system in the pancreatic β-cell milieu, we evaluated the presence of IGFs, IGF receptors, and IGF-binding proteins (IGFBPs) in the β-cell lines βTC3 and HIT T-15. Serum-free conditioned media (SFCM) from βTC3 cells contained IGF-II at concentrations greater than 100 ng/ml. High (15 kDa) and low (7·5 kDa) molecular weight IGF-II were detected both by column chromatography followed by RIA and by immunoblotting. GH (10–1000 ng/ml) conditioning of βTC3 cells stimulated IGF-II secretion in a dose-dependent manner. IGF-II mRNA was detected in βTC3 cells using Northern blots, and also showed a GH-dependent relationship. IGF-II peptide was detected in SFCM from HIT cells, albeit at lower concentrations. To evaluate the presence of IGF receptors in β-cell lines, affinity cross-linking studies were performed on βTC3 cells, demonstrating type I IGF receptors which bound iodinated IGF-II with high affinity, iodinated IGF-I with lesser affinity, and had minimal appreciable binding to iodinated insulin. Type II IGF receptors were not detected. SFCM from βTC3 and HIT cells was subjected to Western ligand blotting, which disclosed the presence of two major IGFBPs of 29 kDa and 24 kDa, characteristic of IGFBP-2 and IGFBP-4. The identity of the specific IGFBPs was confirmed by immunoprecipitation and Northern blotting. Varying the glucose concentration had no significant effect on the levels of IGFBPs, nor did preconditioning with GH, IGF-I, IGF-II, insulin, or glucagon. Levels of both IGFBPs in βTC3 cell-conditioned media increased in the presence of dexamethasone at concentrations of 10⁻⁶ m or greater. In summary, we present evidence that β-cell lines comprise an environment for GH and IGF action. We speculate that IGFs, their receptors and binding proteins function as a complex interactive system which regulates β-cell growth and function.

Journal of Endocrinology (1997) 152, 455–464

ABSTRACT

Two studies were designed to examine the pharmacokinetic and galactopoietic potency of three molecular variants of recombinant-derived bovine GH (rbGH): [Met¹, Leu¹²⁷]-bGH, [Ala¹, Val¹²⁷]-bGH and [Ala¹, Val^127, His¹³³]-bGH. Histidine substitution for arginine at residue 133 of rbGH was shown to impart thrombin resistance. In a Latin square design, nine lactating Holstein cows received a 25 mg rbGH bolus infusion via the jugular vein followed by frequent blood sampling over the next 12 h. The serum GH concentration data were found to fit to a two-compartment open model. Neither primary nor secondary kinetic parameter estimates differed significantly (P>0·05) among the three rbGH variants. Thus, the disposition of GH concentration at time t was described by the equation C(t)=(1295·5 μg/l) (e^{−(0·11/min)(t)}) + (317·3 μg/l)(e^{−(0·03/min)(t)}). Overall averages were: area under the curve=27·1 mg · min per 1, clearance=0·15 litres/min per 100 kg and volume of distribution of the central compartment =2·59 litres/100 kg. The t _1/2 for the two compartments averaged 8·2 and 29·1 min. In the second study, 36 lactating Holstein cows received i.m. injections of one of four oil-based formulation treatments: control vehicle or 500 mg of one of the three rbGH variants every 14 days for 42 days. Average and maximum serum GH concentrations and area under the curve estimates were increased by approximately 3–6 μg/l, 5–15 μg/l and 40–90 μg · day per 1 respectively. Ala¹, Val¹²⁷ rbGH treatments elicited greater blood GH concentrations than [Met¹, Leu¹²⁷]-bGH when administered in an oil-based formulation. Blood GH responses did not directly translate into milk response differences, possibly due to differences in biopotency or receptor availability. Thrombin resistance resulting from substitution of histidine at position 127 of rbGH did not affect blood GH pharmacokinetic parameters or milk response over other rbGH variants.

Journal of Endocrinology (1993) 139, 441–450