Radioimmunoassay For Insulin-like Growth factor-I Research Articles

Production and characterization of recombinantly derived peptides and antibodies for accurate determinations of somatolactin, growth hormone and insulin-like growth factor-I in European sea bass ( Dicentrarchus labrax)

A specific radioimmunoassay (RIA) for European sea bass ( Dicentrarchus labrax) growth hormone (GH) was developed and validated. For this purpose, a stable source of GH was produced by means of recombinant DNA technology in a bacteria system. The identity of the purified protein (ion exchange chromatography) was demonstrated by Western blot and a specific GH antiserum was raised in rabbit. In Western blot and RIA system, this antiserum recognized specifically native and recombinant GH, and it did not cross-react with fish prolactin (PRL) and somatolactin (SL). In a similar way, a specific polyclonal antiserum against the now available recombinant European sea bass SL was raised and used in the RIA system to a sensitivity of 0.3 ng/ml (90% of binding of tracer). Further, European sea bass insulin-like growth factor-I (IGF-I) was cloned and sequenced, and its high degree of identity with IGF-I peptides of barramundi, tuna, and sparid fish allowed the use of a commercial IGF-I RIA based on barramundi IGF-I antiserum. These assay tools assisted for the first time accurate determinations of SL and GH–IGF–I axis activity in a fish species of the Moronidae family. Data values were compared to those found with gilthead sea bream ( Sparus aurata), which is currently used as a Mediterranean fish model for growth endocrinology studies. As a characteristic feature, the average concentration year round of circulating GH in growing mature males of European sea bass was higher than in gilthead sea bream. By contrast, the average concentration of circulating SL was lower. Concerning to circulating concentration of IGF-I, the measured plasma values for a given growth rate were also lower in European sea bass. These findings are discussed on the basis of a different energy status that might allowed a reduced but more continuous growth in European sea bass.

Reliability of the oral glucose tolerance test in the early postoperative assessment of acromegaly remission.

The suppression of growth hormone (GH) to less than 1 microg/L during the oral glucose tolerance test (OGTT) is generally considered to be the standard for the assessment of biochemical remission of GH excess following surgery for GH-secreting pituitary adenomas. In this study the authors examine the reliability of the results of the early postoperative OGTT (epOGTT) in indicating remission or persistence of active acromegaly. Data from the case files of 67 consecutive patients who underwent surgery for the first time for GH-secreting pituitary adenomas were reviewed retrospectively. Definitive remission of acromegaly was considered to be present if, without adjuvant therapy and at the most recent follow-up examination, GH was suppressed to less than 1 microg/L during the OGTT, the level of insulin-like growth factor-I (IGF-I) was within normal limits, and there was no clinical or magnetic resonance imaging evidence of persisting disease. The results of the epOGTT (obtained during the 2nd postoperative week) and the 3-month-postoperative OGTT (3mpOGTT) were compared with the patient's outcome at the most recent follow-up examination. A highly sensitive (< or = 0.3 microg/L) immunoradiometric assay for GH and a highly sensitive (< or = 32 microg/L) radioimmunoassay for IGF-I were used. Correct epOGTT findings were noted in 83.6% of the patients: correct normal results (definitive remission of acromegaly) in 55.2% and correct pathological results (persisting acromegaly) in 28.3% of the patients. The rate of false findings was 16.4%: false normal results in 1.5% and false pathological results in 14.9% of the patients. The rate of correct 3mpOGTT findings increased to 98.5%: correct normal results in 68.6% and correct pathological ones in 29.8% of the patients. A false (false pathological) 3pmOGTT result occurred in only one patient (1.5%). At the most recent follow-up examinations (median 3.6 years) all OGTT findings were correct: correct normal results in 70.1% and correct pathological results in 29.9% of the patients. An intact adenopituitary function was associated (p = 0.04) with the occurrence of false epOGTT findings. The high rate of false results, 16.4% for the epOGTT, declined significantly to 1.5% 3 months postoperatively and to 0% at the most recent follow-up examination. The OGTT appears to be more reliable at 3 months postoperatively. Unless there is obvious evidence of persisting disease following surgery for GH-secreting pituitary adenomas, adjuvant therapy should be delayed for 3 months postoperatively to avoid subjecting the patient to superfluous treatment.

Characterization of Two Forms of Recombinant Salmon Insulin-Like Growth Factor-I: Activities and Complexing with Insulin-Like Growth Factor-I Binding Proteins

Coho salmon insulin-like growth factor I (IGF-I) with N-terminal glycine residue ([Gly]-rsIGF-I) was derived from methionine-extended recombinant salmon IGF-I ([Met]-rsIGF-I) produced in Escherichia coli according to [18]. Purified [Met]-rsIGF-I was treated with methionine aminopeptidase (MAP) to remove the N-terminal Met residue and resultant [Gly]-rsIGF-I was further purified by HPLC on a reverse-phase C4 column. The partial N-terminal amino acid sequence (residues 1–25) of [Gly]-rsIGF-I was identical to that of the native salmon IGF-I. [Gly]-rsIGF-I appeared on sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) as a single band with a molecular weight of 7 kDa and cross-reacted with an antibody against [Met]-rsIGF-I in immunoblot and radioimmunoassay (RIA). In the salmon IGF-I RIA, the dilution curves of [Gly]-rsIGF-I were parallel to the [Met]-rsIGF-I standard curves and the two peptides appeared to be equally potent in displacing label. At concentrations 50 and 100 ng/ml, [Gly]rsIGF-I significantly stimulated the sulfate uptake by the cultured salmon branchial cartilage in a dose-dependent manner. The stimulatory effect of [Gly]-rsIGF-I was equipment to that of [Met]-rsIGF-I. Used as probes in a Western ligand blotting, both [Gly]-rsIGF-I and [Met]-rsIGF-I were able to bind to two IGF binding proteins (IGFBPs) of 23 kDa and 28 kDa from salmon hepatocyte culture medium. The binding could be prevented by preincubation with either unlabeled [Gly]-or [Met]-rsIGF-I. These results indicate that [Gly]-rsIGF-I, which is identical to the native form of IGF-I, and the Met-extended form of recombinant salmon IGF-I, are equally biologically and immunologically active and may substitute for each other in physiological experiments.

Immunoradiometric assay measurements of insulin-like growth factor-I (IGF-I): Comparison with radioimmunoassays using native or des (1–3) IGF-I as radioligands

Insulin-like growth-factor-binding proteins (BPs) in serum interfere with the measurement of insulin-like growth factor-I (IGF-I). Various assays have been developed to overcome this interference. We evaluated an immunoradiometric (IRMA) assay and compared it with the radioimmunoassay (RIA) using both native IGF-I and a truncated form of IGF-I [des (1-3) IGF-I] as radioligands. The IRMA was simpler (one step assay) and faster (3 hr incubation) than RIA(s) (overnight incubation). Sera were extracted with acid ethanol (AE) before all three assays. Analysis of serum samples (n = 78) performed by use of the two different radioligands in the RIA assays were highly correlated (r = 0.967, P < 0.0001). Measurements of serum IGF-I by IRMA in same samples were also highly correlated with those of the RIA assays (r = 0.952 for RIA and 0.947 for triGF-I RIA, P < 0.0001 for both). To assess the effect of binding protein-3 (BP-3) levels (after AE extraction) on these assays, BP-3 levels were measured in sera from 36 healthy women. The mean BP-3 level was 3.6 +/- 0.79 (S.D.) mg/L (range 1.3-5.0), and there was no significant difference in IGF-I levels measured by the three assays. Also, BP-3 levels were inversely correlated with IGF-I levels as measured by all three methods (r = 0.73 for IRMA, 0.71 for triGF-I, and 0.75 for IGF-I RIA). To assess the effect of binding protein-1 (BP-1) levels on these assays, IGF-I was also measured by IRMA and trlGF-I RIA in 19 women with advanced breast cancer. Women with breast cancer had significantly higher (P < 0.001) BP-1 levels than age matched healthy women. IGF-I levels measured by IGF-I IRMA were slightly lower than those measured by trlGF-I RIA in breast cancer patients. However, this difference was not statistically significant (P = 0.56). These findings suggest that variations in BP-3 or BP-1 levels after AE extraction have no significant effect in any of these assays. "We conclude that trlGF-I as a radioligand provides no added advantage over the standard IGF-I RIA. We also conclude that the IRMA assay is valid for measuring IGF-I and is faster and more convenient than RIA.

Development of a Homologous Radioimmunoassay for Coho Salmon Insulin-like Growth Factor-I

A specific homologous radioimmunoassay (RIA) for measurement of insulin-like growth factor-I (IGF-I) in plasma of salmonid and a few non-salmonid fish species was developed using recombinant coho salmon IGF-I (rsIGF-I) as tracer and standard, and antiserum against this peptide raised in rabbits. The minimum detection level of IGF-I was 1.5 ng/ml and linearity was obtained in a range from 1.5 to 23 ng/ml. No cross-reaction was detected in the salmon IGF-I RIA with mammalian growth factors, salmon pituitary hormones, salmon or mammalian insulin, or any peptide in rat plasma. Although salmon IGF-I has high sequence similarity to mammalian IGF-I, it did not cross-react with anti-human IGF-I serum in human RIA and serial dilutions of plasma from salmon were not parallel to the human IGF-I standards in this assay system. In contrast, dilution curves for plasma of salmonids, such as coho (Oncorhynchus kisutch), Atlantic (Salmo salar), and sockeye (O. nerka) salmon, rainbow trout (O. mykiss), some other teleost fish, such as tilapia (Oreochromis mossabmica), carp (Cyprus carpio), eel (Anguilla rostrata), Atlantic halibut (Hippoglossus hippoglossus), and agnathan, the sea lamprey (Petromyzon marinus), assessed in salmon IGF-I RIA were parallel to the rsIGF-I standards. Acid-ethanol extraction of plasma samples altered the molecular weight, but not the quantity, of immunoreactive IGF-I, implying that IGF-I binding proteins in salmon plasma do not affect the performance of the salmon IGF-I RIA. Gel filtration of nonacidified plasma on a Sephadex G-75 superfine column produced two immunoreactive IGF-I peaks of molecular weights of approximately >70 k and 7 kDa, whereas acidification of plasma increased the relative amount of the 7-kDa peak (IGF-I) and the >70-kDa peak disappeared. The recoveries of rsIGF-I added to extracted or nonextracted plasma were 97.4 and 94.9%, respectively. Inter- and intraassay coefficients of variation were 3.6 and 3.3%, respectively. Plasma IGF-I levels in coho salmon smolts were 117.4 ± 19.1 ng/ml as compared to IGF-I levels in parr (45.3 ± 2.5 ng/ml) or in adult fish (45.2 ± 5.4 ng/ml) measured in the same assay. Injection of salmon growth hormone, but not prolactin or somatolactin, caused a significant and dose-dependent elevation of plasma IGF-I levels, while either fasting or injection of streptozotocin led to a significant decline in systemic IGF-I. We conclude that the salmon IGF-I RIA reported in this study is valid for measurement of IGF-I levels in salmonids and, likely, in a variety of other fish species.

Immunoreactive and receptor-active insulin-like growth factor-I (IGF-I) and IGF-binding protein in blood plasma from the freshwater fish Macquaria ambigua (golden perch).

The presence of insulin-like growth factor-I (IGF-I)-related molecules and IGF-binding factors in blood from golden perch, Macquaria ambigua, an Australian native freshwater fish, was investigated. Serum was acidified to dissociate IGF and IGF-binding protein complexes that might be present, and fractionated by size-exclusion high-performance liquid chromatography at pH 2.8. Fractions were neutralized and their activities assessed by (i) an immunoassay for mammalian IGF-I which also detects chicken IGF-I but in which all known forms of IGF-II react very poorly, (ii) a receptor assay for IGF-II in which all known forms of IGF-I react poorly, and (iii) a type-I IGF receptor assay in which mammalian IGF-I and IGF-II polypeptides are amost equivalent. No IGF-II-like activity was detected. Three peaks of IGF-I-like activity were detected by IGF-I immunoassay and type-I IGF receptor assay. The major peak of activity was similar in molecular size to human IGF-binding protein-3, 45-55 kDa ('large IGF'), and a minor peak of activity which was similar in size to mammalian IGFs, 7.5 kDa. A third peak of activity was observed eluting at a time which indicates that it is a smaller molecule than any previously described IGF. The large IGF was temperature-sensitive, but was not a binding protein for 125I-labelled human IGF-I (hIGF-I). This material therefore was able to bind to anti-hIGF-I antibodies and to human type-I IGF receptors, and may represent the fish equivalent of mammalian prepro-IGFs.(ABSTRACT TRUNCATED AT 250 WORDS)

Sequestration and secretion of insulin‐like growth factor‐I by bovine aortic endothelial cells

Endothelial cells elaborate growth promoting activities in culture medium that support limited smooth muscle cell and fibroblast growth in vitro in the absence of serum. We investigated whether insulin-like growth factor-I (IGF-I) was synthesized and secreted by bovine aortic endothelial cells in vitro. Subconfluent endothelial cell cultures in serum-free medium secreted severalfold higher IGF-I levels than confluent cultures by acid-sizing chromatography and IGF-I radioimmunoassay. The IGF-I secretory level was not sustained during a second serum-free incubation. In contrast, secretion of IGF binding proteins persisted and was maintained at constant levels throughout the same observation periods. Analysis of poly(A+)RNA by northern blots revealed hybridization of an IGF-I cDNA to a 7.5- to 7.0-kb transcript and superinduction of the 7.5-7.0-kb mRNA by the translational inhibitor, cyclohexamide. However, no endogenously labeled IGF-I was detected in conditioned media after incubation of cultures with [35S]cysteine or [3H]leucine. When cultures were incubated in the presence of serum supplemented with IGF-I, subconfluent cultures sequestered and released more IGF-I than confluent cultures. We concluded that the majority of IGF-I secreted in vitro was sequestered from serum.

Radioimmunoassay for insulin-like growth factor-I in bovine milk

This experiment was designed (1) to study the effectiveness of acid-ethanol extraction and gly-gly HCl incubation to dissociate the IGF-binding protein complex before IGF-I RIA measurements in bovine milk, and (2) to determine the influence of freezing/thawing and defatting of milk samples on IGF-I concentration of milk. Acidification of milk with acid-ethanol (acid-ETOH) liberated more IGF-I from its binding proteins than with gly-gly HCl. Milk samples obtained from three control cows and three rbST-treated cows (41–44 wk postpartum) were treated with acid-ethanol and assayed by a radioimmunoassay. IGF-I concentrations in the milk were 3.30 ± 1.40 for the control cows and 2.77 ± 1.36 ng mL−1 for the rbST-treated cows. Fat removal before or after freezing did not affect milk IGF-I measurements. The milk concentration of IGF-I after freezing/thawing twice was not different from that after freezing/thawing once. The radioimmunoassay resulting from this study should provide a tool for monitoring IGF-I concentrations in milk throughout the entire lactation. Key words: Milk, IGF-I, radioimmunoassay, cows

Evaluation of interference by insulin-like growth factor I (IGF-I) binding proteins in a radioimmunoassay for IGF-I in serum from dairy cows

Insulin-like growth factor I (IGF-I) circulates in serum bound to a number of different binding proteins (BPs). With antibodies currently available, BPs must be dissociated and inactivated or removed from serum prior to measurement of IGF-I by radioimmunoassay (RIA). Serum samples which spanned a 13-fold range in IGF-I concentration were obtained from lactating dairy cows and used to develop conditions for assay of IGF-I with minimal interference from BPs. Removal of BPs from serum by acid-ethanol extraction resulted in interference in the RIA. Therefore, serum was incubated with 0.1 M glycyl-glycine HCl to inactivate BPs as suggested by Underwood et al. (11). Time, temperature and pH were optimum when serum was incubated for 48 hr at 37 C, pH 3.7. Binding protein inactivation was evaluated by ability of glycyl-glycine incubated serum to reassociate with 125I-IGF-I. In addition, BPs isolated by gel filtration of glycyl-glycine incubated serum were tested for interference in the RIA. The concentration of IGF-I in serum where inactivated BPs were removed by acid gel filtration was compared to corresponding glycyl-glycine incubated serum. There was a 1:1 relationship which intersected at zero indicating that total IGF-I could be measured. Therefore, incubation of serum with glycyl-glycine is a reliable method for measuring total IGF-I in serum from dairy cows.

Radioimmunoassay for insulin-like growth factor-I: solutions to some potential problems and pitfalls

This report describes essential requirements for the validation of a radioimmunoassay (RIA) for insulin-like growth factor-I (IGF-I) and presents solutions to some problems and pitfalls commonly observed. The preparation of IGF-I to be used as radioligand or standard has to be selected carefully since some IGF-I preparations are contaminated with variants which demonstrate different potencies for different antisera used in the RIA. Accurate assessment of IGF-I levels in blood plasma requires an efficient extraction method for the IGF-binding proteins (IGFBPs). Extraction methods to remove the influence of IGFBPs in the RIA were compared using blood plasma of considerable differences in IGF-I/IGFBP ratios. Acidification of plasma before column chromatography on Sephadex G-75 (G75) is generally considered to be the most reliable extraction method, but it is very time-consuming. The acid-ethanol extraction (AE) of plasma is not valid in many situations. Non-parallel displacement to the IGF-I standard was observed with AE-extracted plasma samples in the RIA. In addition, a comparison of IGF-I values obtained in the RIA after AE or G75 extraction of fetal ovine plasma has shown no significant correlation. We report an extraction technique based on a modified AE extraction followed by cryo-precipitation (AEC). AEC extraction on blood plasma reduced residual IGFBPs to a level that did not interfere in the assay. Furthermore, AEC-extracted plasma samples showed parallel displacement in the RIA to highly purified preparations of authentic IGF-I. We observed high correlations, with a slope close to unity, of IGF-I values obtained in the RIA using the AEC or G75 extraction for plasma from different species including adult and fetal sheep, rat, mouse and man. The AEC extraction provides a rapid and simple alternative to G75 extraction for blood plasma from a variety of species provided that high-affinity antisera are used for the RIA.

Chicken insulin-like growth factor-I: Amino acid sequence, radioimmunoassay, and plasma levels between strains and during growth

Insulin-like growth factor-I (IGF-I) has been purified from chicken serum and sequenced. The peptide has eight amino acid substitutions when compared with human IGF-I: serine 26, leucine 38, histidine 39, histidine 40, lysine 41, glutamine 50, isoleucine 64, and proline 67. Chicken IGF-I (cIGF-I) has been measured using a radioimmunoassay with a human IGF-I (hIGF-I) standard and an antibody raised against hIGF-I. In this assay the cross-reactivity of cIGF-I was approximately 50% with respect to hIGF-I and the cross-reactivity of chicken IGF-II was 1.7% with respect to chicken IGF-I. To determine whether binding proteins in chicken plasma can artifactually interfere with IGF-I measurements as they do in mammals, chicken plasma was fractionated by molecular sieve chromatography at acid pH. When the fractions corresponding to the binding protein region were included in the IGF-I radioimmunoassay, essentially no apparent IGF-I was detected, indicating that the binding proteins did not interfere. This result, together with the finding that IGF-I in acid-ethanol extracts of chicken plasma produced parallel dose-response curves to pure cIGF-I and hIGF-I, allows the reliable measurement of cIGF-I in such extracts. The concentrations of IGF-I in plasma from male birds increased two-to threefold between 1 and 7 weeks after hatching to achieve 30–45 ng/ml. Smaller increases were found in female chickens from a higher value at 1 week. No diurnal pattern of IGF-I levels could be detected. In 4-week-old birds, the plasma concentration of the peptide fell from nearly 40 to 15 ng/ml after 24 hr of starvation and to 9 ng/ml 20 hr later. These effects are very similar to those described for mammals and strongly suggest that the regulation of IGF-I is conserved during evolution, notwithstanding the lower plasma concentrations of the growth factor in chickens.

Purification, amino acid sequences and assay cross-reactivities of porcine insulin-like growth factor-I and -II.

Porcine insulin-like growth factor-I (IGF-I) and IGF-II have been characterized to help define the roles of these peptides in the growth process. The amino acid sequence of porcine IGF-I was found to be identical to the human and bovine peptides. Porcine IGF-II was more similar to human IGF-II than to forms of this growth factor in other mammalian species, differing only in the replacement of asparagine for serine at residue 36. In a biological assay that measures the stimulation of protein synthesis in rat L6 myoblasts, porcine IGF-I was approximately ninefold more potent than porcine IGF-II or bovine IGF-II, while recombinant human IGF-I and IGF-II had half the potency of the respective natural peptides. Porcine and recombinant human IGF-I showed essentially equal competition for binding in a human IGF-I radioimmunoassay while between 0.6 and 1.5% cross-reactivity was observed with human, bovine or porcine IGF-II. A receptor assay for IGF-II demonstrated similar potencies for the three IGF-II peptides, while the cross-reactivity of recombinant human IGF-I was only 0.05%. Porcine IGF-I exhibited a higher cross-reactivity, presumably due to very slight contamination with IGF-II.

Insulin-like growth factor I activity is stored in the yolk of the avian egg

Growth factors of maternal origin may be incorporated into the vertebrate egg and play a role in early phases of embryo growth and differentiation. An avian insulin-like growth factor I (IGF-I) activity from unfertilized chicken egg-yolk has been partially purified by HPLC. The material is slightly more hydrophobic than recombinant human IGF-I. It reacts in a human IGF-I radioimmunoassay and is specifically depleted by anti-human IGF-I antibodies. Like authentic IGF-I, the extracts enriched in IGF-I activity stimulated the accumulation of δ-crystallin mRNA in epithelial cells from chick embryo lens with a potency approximately equivalent to its IGF-I immunoactivity.

Validation of a heterologous radioimmunoassay for insulin-like growth factor-I in bovine serum.

A quantitative, repeatable, heterologous radioimmunoassay (RIA) for insulin-like growth factor-I (IGF-I) was developed for bovine serum. Untreated serum could not be assayed due to interference by IGF-I-binding protein. Serum acidified to pH 3.6 in glycine-HCl (0.1 mol/l) for 24 h at 37 degrees C and neutralized with NaOH produced inhibition curves non-parallel to the [Thr59]-IGF-I standard. Neutralization by 40-fold dilution of acidified serum samples with assay buffer produced inhibition curves nearly parallel to the IGF-I standard. Complete parallelism was achieved by utilizing preprecipitated normal rabbit serum-sheep anti-rabbit gamma-globulin to separate antibody-bound 125I-labelled IGF-I from free 125I-labelled IGF-I. Recovery of IGF-I (1.3-52.3 fmol) added to serum was quantitative. The sensitivity of the RIA (n = 6) was 8.25 +/- 0.17 (S.E.M.) fmol. Intra- and interassay coefficients of variation were 3.03 and 4.95% respectively. Serum IGF-I levels measured in beef calves at weaning were positively correlated with weaning weight, total weight gain from birth to weaning and average daily weight gain. In conclusion, a heterologous RIA for IGF-I in beef serum which is sensitive, accurate, precise and repeatable has been developed.

Insulin-like growth factors (IGF) I and II in diabetic pregnancy: suppression of normal pregnancy-induced rise of IGF-I.

The concentrations of somatomedins/insulin-like growth factors were measured by a specific radioimmunoassay for insulin-like growth factor-I and a specific radioreceptor assay for insulin-like growth factor-II in sera of term normal and Type 1 (insulin-dependent) diabetic pregnant women and in various cord sera of their newborn infants. Serum insulin-like growth factor-I levels in normal (non-diabetic) maternal serum were higher than in non-pregnant women (486 +/- 26 versus 215 +/- 26 ng/ml). The normal pregnancy-induced increment of insulin-like growth factor-I was markedly reduced in diabetic pregnancy. It was not different in patients with good or poor glycaemic control, as judged by normal or elevated blood levels of haemoglobin A1c content. Insulin-like growth factor-I levels in cord serum of infants of diabetic women with good glycaemic control (86 +/- 11 ng/ml) and poor glycaemic control (91 +/- 19 ng/ml) were significantly higher (p less than 0.01) than in infants of non-diabetic women (43 +/- 42 ng/ml). The fetal birth weight ratios were not significantly correlated with insulin-like growth factor-I levels in cord serum. Serum insulin-like growth factor-II levels in maternal and cord serum in diabetic and normal pregnancy were not different from each other or from normal non-pregnant women. The increment in insulin-like growth factor-I levels in maternal serum in pregnancy may influence placental structure and function. Lack of this increment in maternal diabetes may have direct implication in placental abnormalities in diabetes and indirect implications on fetal development and metabolism.(ABSTRACT TRUNCATED AT 250 WORDS)

Insulin-like growth factor I (IGF-I) in healthy children, adolescents and adults as determined by a radioimmunoassay specific for the synthetic 53-70 peptide region.

A radioimmunoassay (RIA) specific for the synthetic 53-70 peptide region of human insulin-like growth factor I (IGF-I) was used to determine IGF-I in the serum of 191 healthy newborns, children and adolescents and in 26 adults. The results compare favourably with reported values obtained using RIA systems for the native IGF-I molecule. Intra- and inter-assay CV were 3.3 and 7.2% respectively. In childhood, mean +/- SD IGF-I levels rise from 6.0 +/- 3.5 nmol/l in newborns to 16.5 +/- 4.0 nmol/l at 8-11 years in both sexes. At the onset of puberty, IGF-I levels in females (24.9 +/- 6.6 nmol/l) are significantly (P greater than 0.005) higher than in males (17.2 +/- 4.2 nmol/l). With further pubertal development IGF-I levels continue to rise, reaching peak values at pubertal stage P4 (40.6 +/- 4.5 nmol/l in males, 42.8 +/- 5.1 nmol/l in females) and decline thereafter to lower values during adulthood: 16.5 +/- 5.8 nmol/l (males) and 24.2 +/- 7.0 nmol/l (females) (P greater than 0.001). In pubertal males, IGF-I correlates significantly with height (r = 0.66, P less than 0.001), bone age (r = 0.69, P less than 0.001) and growth velocity (r = 0.64, P = 0.025) as well as with testosterone levels (r = 0.69, P less than 0.001). In pubertal females a significant correlation is found between IGF-I and height (r = 0.55, P less than 0.020). The ready availability of a simple, precise and reproducible IGF-I RIA, should contribute much to evaluating the importance of IGF-I measurements in normal growth and in the diagnosis and therapy of various growth disorders.