Clinical Endocrinology (1992) 37, 249-253
Effect of growth hormone on the low level of growth hormone binding protein in idiopathic short stature Manuel Fontoura't, Elizabeth Mugnier', Raja Brauner", Raphael Rappaport' and Marie-Catherine Postei-Vinay$ $Unit6 344 and *Unit6 30 of the lnstitut National de la Sent6 et de la Recherche Mddicale, Hdpital Necker-Enfants Malades, Paris, France (Received 3 January 1992; returned for revision 20 February 1992; finally revised 17 March 1992; accepted 8 April 1992)
protein. A positive correlation was found between IGF-I plasma levels and growth hormone-binding protein and also between growth velocity and growth hormone-binding protein. CONCLUSIONS The low growth hormone-binding protein and the response to high doses of OH suggest partial OH resistance at the receptor level, in thls group of children with idlopathlc short stature.
Summary OBJECTIVE Growth
hormone receptor status was assessed in children with idiopathic short stature by evaluating piasma growth hormone-binding protein before and under GH therapy. DESIGN Among 22 children presenting idiopathic short stature, 15 were randomly selected to be treated with OH (1.2 IUlkglweek); they were studied before and under GH therapy. Untreated patients served as a control group for age and OH effect. PATIENTS Twenty-two prepubertal children, aged 5-11 years, were studied. They presented growth retardationof - 2 . 8 f O . l SDS (meanf SEM). All had normal GH secretion and their mean IGF-i plasma level was normal. MEASUREMENT Growth hormone-binding protein was measured using high pressure liquid chromatographygel flitration. The specific binding of '=i-hGH to the growth hormone-bindingprotein was expressed as a percentage of the total radioactivity. RESULTS Specific binding of l%hGH to the high affinity growth hormone-binding protein was low with a mean f SEM value of 11.1 f 0.9% of radioactivity. in the treated group, growth hormone-bindingprotein increased slgnificantiy after 3 months of treatment; It reached 21.1 k 1 .O% of radioactivity (mean fSEM) in the eight children who have been treated for 18 months. In the seven untreated children, the growth hormone-binding protein value Increased to 1 6 - 2 f l - l %after 18 months; thls value is significantly lower than that found in the GH-treated children, demonstrating that the OH effect is greater than the age-related increase in the growth hormone-binding Correspondence: M. C. Postel-Vinay, INSERM Unit6 344, Facultk Necker, 156 rue de Vaugirard, 75730 Paris Cedex 15, France. ?Present address: Departmento de Pediatria, Hospital San Jolo, Porto, Portugal.
Causes of the growth defect in children with idiopathic short stature (ISS) have frequently been discussed. Impaired growth hormone (GH) secretion may be an underlying cause for short stature (Zadik et al., 1985). Recent studies indicate that the majority of these children, when treated with GH, increase their growth rate (Genentech Collaborative Study Group, 1989; Lin et al., 1989). The GH-binding protein (GHBP), which has been identified in plasma, represents a soluble short form of GH receptor: the amino acid sequences of the extracellular domain of the G H receptor and of the plasma GH-binding protein (GHBP) are identical; this identity was demonstrated in rabbits (Leung et al., 1987; Spencer et al., 1988), mice (Smith et al., 1989), rats (Baumbach et al., 1989), and is highly probable in humans. Therefore it was of interest to evaluate the GHBP in children with ISS and to examine its possible regulation under G H therapy, as a reflection of cellular GH receptor levels. Our results suggest a certain degree of GH resistance since these children with ISS have a low plasma GH binding activity which is increased by high doses of GH.
Materials and methods Subjects
Twenty-two children (14 boys, eight girls), aged 5-11 years, with idiopathic short stature were studied. All children were in good health. They presented no demonstrable causes of short stature and had normal birth weight. They came from families where both parents had normal height, except for two families with a mother of short stature (-2.3 and - 2.5 SDS). They remained prepubertal during this study. They presented with short stature: SD from normal height ranged from -3.7 to -2.0 (mean heightkSEM -2.8kO-1 SDS). Their growth velocity ranged from 3.0 to 6.1 cm/year (mean 249
250
M. Fontoura et a / .
4.7 f0.2 cm/year) when the study started. They had bone age retardation: 6.5 f0.5 us 8.3 f0.7 years for chronological age. They all had normal GH secretion as assessed by a G H peak higher than 20 pU/l (mean 41.4 f3.2 pU/l, range =2 14-77.6 pU/l; lmU/l= 0.5 pg/l) during an arginine-insulin tolerance test and during sleep (GH peak = 59.4f 5.8 mU/l, range=27.6-148 mU/l). The mean value for their IGF-I plasma level was normal: 118f 14 pg/l (meanfSEM) (Rosenfeld et al., 1986). Fifteen of the 22 children were randomly selected to be treated, in a therapeutic trial, with recombinant human GH (Saizen, Serono Laboratories), at the dosage of 1-2IU/kg/week, given S.C. 6 times a week. The protocol was approved by the ethical committee of H6pital Necker. GHBP was measured before treatment in all children. In the 15 treated children, GHBP was measured after 3 and 6 months of treatment; only eight patients had been treated long enough to allow the assay at 12 and 18 months. The remaining seven children were followed without treatment; their GHBP was measured at the beginning of the study and after 18 months.
Hormone radioimmunoassays
hGH was measured by a double antibody method; 1st IRP hGH (WHO 66/217) was used for the standard. IGF-I was measured by RIA on acid-ethanol extracted plasma as described by Breier et al. (1991).
Measurement of GHBP
Binding of lZ5I-hGHto plasma GHBP was determined as previously described by Tar et al. (1990). Briefly, plasma (1OOpl) was incubated for 20 hours at 4°C with 100 pI potassium phosphate 0.1 M, pH 7.0, BSA 0.1%, containing Iz5I-hGH(2 x lo5 c.p.m.). After filtration through 0.45 mm Millipore minifilter, the entire incubation mixture was injected onto HPLC Protein Pak 300 sw column (0-75x 30 cm).Elution was performed isocratically using a degassed buffer (Na2S040.1 M, potassium phosphate 0.1 M, pH 7.0) pumped at a rate of 0.5 ml/min. Radioactivity was recorded on line using a Berthold LB 504 gamma detector connected to an Apple IIe computer. The binding of hGH is expressed as the radioactivityin the individual peak divided by the total radioactivity in peaks I, I1 and 111. To evaluate non-specific binding to peak 11-BP, 5 pg of GH was added to the plasma incubation. On the elution profiles, no antiGH antibodies could be detected in plasmas of the GH-treated patients. Absence of antiGH antibodies was confirmed by RIA (Rougeot et al., 1991).
Clinical Endocrinology (1992) 37
Statistical analyses
Results are presented as meanf SEM. Unpaired and paired Student’s t-tests were used for statistical evaluations. Differences were considered statistically significant at P < 0.05. Correlations were examined by linear regression analysis. Results
Specific binding of Iz5I-hGHto high affinity GHBP was low in the group of children with idiopathic short stature: the mean value found in the 22 patients was 11.1 +0.9% of radioactivity, which is significantly lower ( P < 0.001) than the value of 2443&1.7% found in normal prepubertal children (Tar et al., 1990). Fifteen of the 22 patients were randomly selected to be treated with GH. Under GH therapy, the binding to GHBP increased, as shown in Table 1. After 3 months, the mean value (13.5f 1.1% of radioactivity) was significantly higher as analysed by paired t-test. After 6 months of treatment, the value reached 16.5f 1.5%. Eight of the 15 patients who have now been treated for almost 2 years had a G H binding activity at 17.9 5 1.7% after 12 months and at 21.1 1.0% after 18 months; the GHBP value found after 18 months of treatment is not different from the normal value in prepubertal children. In all treated patients, the binding of GH to the BP has increased, with the magnitude ranging from 1.4 to 3.4. The elution profiles of IZ51-hGHincubated with the plasma of one patient, before and under GH therapy, are shown on Fig. 1. The increase in binding was assessed by Scatchard plot analysis derived from competition experiments (not shown); it appears mainly related to an increase in the binding capacity which is significantly higher after 18months of treatment (22.4 f3-7 pg/l) than before treatment (12-2fO.4 pg/l). The seven children, who did not receive GH therapy, were studied during the same time period (Table 2). Their plasma GH binding activity increased significantly from 11.5 f 1.6% of radioactivity to 16-2+ 1.1% after 18 months; however, this value is significantly lower than that found in the treated group after 18 months of therapy (21.1 f 1.0%, P
Effect of growth hormone on the low level of growth hormone binding protein in idiopathic short stature Manuel Fontoura't, Elizabeth Mugnier', Raja Brauner", Raphael Rappaport' and Marie-Catherine Postei-Vinay$ $Unit6 344 and *Unit6 30 of the lnstitut National de la Sent6 et de la Recherche Mddicale, Hdpital Necker-Enfants Malades, Paris, France (Received 3 January 1992; returned for revision 20 February 1992; finally revised 17 March 1992; accepted 8 April 1992)
protein. A positive correlation was found between IGF-I plasma levels and growth hormone-binding protein and also between growth velocity and growth hormone-binding protein. CONCLUSIONS The low growth hormone-binding protein and the response to high doses of OH suggest partial OH resistance at the receptor level, in thls group of children with idlopathlc short stature.
Summary OBJECTIVE Growth
hormone receptor status was assessed in children with idiopathic short stature by evaluating piasma growth hormone-binding protein before and under GH therapy. DESIGN Among 22 children presenting idiopathic short stature, 15 were randomly selected to be treated with OH (1.2 IUlkglweek); they were studied before and under GH therapy. Untreated patients served as a control group for age and OH effect. PATIENTS Twenty-two prepubertal children, aged 5-11 years, were studied. They presented growth retardationof - 2 . 8 f O . l SDS (meanf SEM). All had normal GH secretion and their mean IGF-i plasma level was normal. MEASUREMENT Growth hormone-binding protein was measured using high pressure liquid chromatographygel flitration. The specific binding of '=i-hGH to the growth hormone-bindingprotein was expressed as a percentage of the total radioactivity. RESULTS Specific binding of l%hGH to the high affinity growth hormone-binding protein was low with a mean f SEM value of 11.1 f 0.9% of radioactivity. in the treated group, growth hormone-bindingprotein increased slgnificantiy after 3 months of treatment; It reached 21.1 k 1 .O% of radioactivity (mean fSEM) in the eight children who have been treated for 18 months. In the seven untreated children, the growth hormone-binding protein value Increased to 1 6 - 2 f l - l %after 18 months; thls value is significantly lower than that found in the GH-treated children, demonstrating that the OH effect is greater than the age-related increase in the growth hormone-binding Correspondence: M. C. Postel-Vinay, INSERM Unit6 344, Facultk Necker, 156 rue de Vaugirard, 75730 Paris Cedex 15, France. ?Present address: Departmento de Pediatria, Hospital San Jolo, Porto, Portugal.
Causes of the growth defect in children with idiopathic short stature (ISS) have frequently been discussed. Impaired growth hormone (GH) secretion may be an underlying cause for short stature (Zadik et al., 1985). Recent studies indicate that the majority of these children, when treated with GH, increase their growth rate (Genentech Collaborative Study Group, 1989; Lin et al., 1989). The GH-binding protein (GHBP), which has been identified in plasma, represents a soluble short form of GH receptor: the amino acid sequences of the extracellular domain of the G H receptor and of the plasma GH-binding protein (GHBP) are identical; this identity was demonstrated in rabbits (Leung et al., 1987; Spencer et al., 1988), mice (Smith et al., 1989), rats (Baumbach et al., 1989), and is highly probable in humans. Therefore it was of interest to evaluate the GHBP in children with ISS and to examine its possible regulation under G H therapy, as a reflection of cellular GH receptor levels. Our results suggest a certain degree of GH resistance since these children with ISS have a low plasma GH binding activity which is increased by high doses of GH.
Materials and methods Subjects
Twenty-two children (14 boys, eight girls), aged 5-11 years, with idiopathic short stature were studied. All children were in good health. They presented no demonstrable causes of short stature and had normal birth weight. They came from families where both parents had normal height, except for two families with a mother of short stature (-2.3 and - 2.5 SDS). They remained prepubertal during this study. They presented with short stature: SD from normal height ranged from -3.7 to -2.0 (mean heightkSEM -2.8kO-1 SDS). Their growth velocity ranged from 3.0 to 6.1 cm/year (mean 249
250
M. Fontoura et a / .
4.7 f0.2 cm/year) when the study started. They had bone age retardation: 6.5 f0.5 us 8.3 f0.7 years for chronological age. They all had normal GH secretion as assessed by a G H peak higher than 20 pU/l (mean 41.4 f3.2 pU/l, range =2 14-77.6 pU/l; lmU/l= 0.5 pg/l) during an arginine-insulin tolerance test and during sleep (GH peak = 59.4f 5.8 mU/l, range=27.6-148 mU/l). The mean value for their IGF-I plasma level was normal: 118f 14 pg/l (meanfSEM) (Rosenfeld et al., 1986). Fifteen of the 22 children were randomly selected to be treated, in a therapeutic trial, with recombinant human GH (Saizen, Serono Laboratories), at the dosage of 1-2IU/kg/week, given S.C. 6 times a week. The protocol was approved by the ethical committee of H6pital Necker. GHBP was measured before treatment in all children. In the 15 treated children, GHBP was measured after 3 and 6 months of treatment; only eight patients had been treated long enough to allow the assay at 12 and 18 months. The remaining seven children were followed without treatment; their GHBP was measured at the beginning of the study and after 18 months.
Hormone radioimmunoassays
hGH was measured by a double antibody method; 1st IRP hGH (WHO 66/217) was used for the standard. IGF-I was measured by RIA on acid-ethanol extracted plasma as described by Breier et al. (1991).
Measurement of GHBP
Binding of lZ5I-hGHto plasma GHBP was determined as previously described by Tar et al. (1990). Briefly, plasma (1OOpl) was incubated for 20 hours at 4°C with 100 pI potassium phosphate 0.1 M, pH 7.0, BSA 0.1%, containing Iz5I-hGH(2 x lo5 c.p.m.). After filtration through 0.45 mm Millipore minifilter, the entire incubation mixture was injected onto HPLC Protein Pak 300 sw column (0-75x 30 cm).Elution was performed isocratically using a degassed buffer (Na2S040.1 M, potassium phosphate 0.1 M, pH 7.0) pumped at a rate of 0.5 ml/min. Radioactivity was recorded on line using a Berthold LB 504 gamma detector connected to an Apple IIe computer. The binding of hGH is expressed as the radioactivityin the individual peak divided by the total radioactivity in peaks I, I1 and 111. To evaluate non-specific binding to peak 11-BP, 5 pg of GH was added to the plasma incubation. On the elution profiles, no antiGH antibodies could be detected in plasmas of the GH-treated patients. Absence of antiGH antibodies was confirmed by RIA (Rougeot et al., 1991).
Clinical Endocrinology (1992) 37
Statistical analyses
Results are presented as meanf SEM. Unpaired and paired Student’s t-tests were used for statistical evaluations. Differences were considered statistically significant at P < 0.05. Correlations were examined by linear regression analysis. Results
Specific binding of Iz5I-hGHto high affinity GHBP was low in the group of children with idiopathic short stature: the mean value found in the 22 patients was 11.1 +0.9% of radioactivity, which is significantly lower ( P < 0.001) than the value of 2443&1.7% found in normal prepubertal children (Tar et al., 1990). Fifteen of the 22 patients were randomly selected to be treated with GH. Under GH therapy, the binding to GHBP increased, as shown in Table 1. After 3 months, the mean value (13.5f 1.1% of radioactivity) was significantly higher as analysed by paired t-test. After 6 months of treatment, the value reached 16.5f 1.5%. Eight of the 15 patients who have now been treated for almost 2 years had a G H binding activity at 17.9 5 1.7% after 12 months and at 21.1 1.0% after 18 months; the GHBP value found after 18 months of treatment is not different from the normal value in prepubertal children. In all treated patients, the binding of GH to the BP has increased, with the magnitude ranging from 1.4 to 3.4. The elution profiles of IZ51-hGHincubated with the plasma of one patient, before and under GH therapy, are shown on Fig. 1. The increase in binding was assessed by Scatchard plot analysis derived from competition experiments (not shown); it appears mainly related to an increase in the binding capacity which is significantly higher after 18months of treatment (22.4 f3-7 pg/l) than before treatment (12-2fO.4 pg/l). The seven children, who did not receive GH therapy, were studied during the same time period (Table 2). Their plasma GH binding activity increased significantly from 11.5 f 1.6% of radioactivity to 16-2+ 1.1% after 18 months; however, this value is significantly lower than that found in the treated group after 18 months of therapy (21.1 f 1.0%, P
