| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Pediatric Endocrinology (C.L., D.R., A.R., P.B.), Unité 561 Institut National de la Santé et de la Recherche Médicale (C.L., C.L.S., P.B.), and Pediatric Unit of Hormonal Biochemistry (N.L.), Saint Vincent de Paul Hospital, Paris V University, 75014 Paris, France; and Service de Biostatistique (C.E.), Hôpital Necker, 75015 Paris, France
Address all correspondence and requests for reprints to: Pierre Bougnères, M.D., Ph.D., Pediatric Endocrinology, Hôpital Saint Vincent de Paul Paris V University, 82 Avenue Denfert Rochereau, 75014 Paris, France. E-mail: bougneres{at}paris5.inserm.fr.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Objectives: This study tested a cell-based assay as an index of individual IGF-I sensitivity that could help dissect GH pharmacogenetics.
Design: Akt phosphorylation (P-Akt) was quantified in response to IGF-I in fresh lymphocytes from 50 short children (25 with idiopathic short stature and 25 born short for gestational age) whose growth parameters were being prospectively monitored during the first year of GH therapy (86 ± 20 µg/kg·d).
Results: Intra-individual triplicate measurements of IGF-I-stimulated P-Akt were reasonably consistent (0.11 
SD; mean 0.23). Among the 50 children, the distribution of P-Akt in lymphocytes stimulated by 125 ng/ml IGF-I was closely associated with the growth response to GH administration (univariate P = 0.001). Both GH dosage (P = 0.006) and the fold increase in IGF-I levels (P = 0.04) in response to GH (P = 0.04) were also correlated with the growth response.
Conclusion: Lymphocytes are the only IGF-I target cells that can be easily studied in clinical research. IGF-I-stimulated P-Akt in these cells was found to be a predictor of GH efficacy, supporting a significant role of the first steps of IGF-I signaling in the individual variability of GH effects on growth.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
IGF-I is a master activator of a complex signaling network that regulates growth, development, and differentiation at the tissue level and proliferation and survival at the cellular level (10, 11, 12). Six high-affinity IGF-binding proteins occur in cell surface-associated forms and in the circulation and in extracellular fluids. The IGF-binding proteins can modulate IGF actions both positively and negatively through effects on IGF half-life and receptor interaction (10, 13, 14). The IGF-I signaling system includes the IGF-I receptor (IGF-IR) and the insulin receptor (IR). The IGF-IR mediates the majority of IGF-I biological effects (15) and is abundantly expressed in chondrocytes. IGF-I also cross-talks with the IR through hybrid receptors consisting of covalently linked IGF-IR and IR hemireceptors, but the overall contribution of hybrid receptors to IGF-I action still remains unclear (13, 16, 17, 18). At physiological levels, IGF-I also acts through both versions of the IR to preferentially activate IR substrate (IRS)-2 and downstream biological actions (19). The IGF-IR undergoes autophosphorylation after ligand activation. Tyrosine-phosphorylated residues in the juxtamembrane domain of the β-subunits of IGF-IR then recruit IRS-1 and IRS-2, which serve as scaffolding/adaptor proteins that couple the activated IGF-IR to upstream components of the phosphatidylinositol 3-kinase (PI3K) and Erk signal transduction cascades (20, 21). The PI3K pathway is activated by both IGF-IR and IR binding of IGF-I and is considered the major route for IGF-I signaling for regulating cell proliferation and differentiation (22). The phosphorylation of Akt is a major step of the PI3K pathway (22, 23) and was tested in the current study as a quantitative index of IGF-I action on target cells.
After estimating individual IGF-I sensitivity of Akt phosphorylation at the cellular level, we tested whether such an estimate was associated with the variability of the clinical growth response in 50 ISS or SGA children during the first year of GH administration. Our working hypothesis was that individual variations in IGF-I capacity to activate the PI3K-Akt pathway would possibly influence IGF-I effects on skeletal growth. We were conscious that our approach can offer only a (very) limited view of IGF-I action, because Akt phosphorylation recapitulates only the upper part of the PI3K pathway activity and leaves many downstream steps or alternate pathways of IGF-I signaling (10, 19) unstudied. We chose circulating lymphocytes as our cell model (osteogenic cells cannot be sampled) not only because of ethics and practicality but also because lymphocytes have an active PI3K-Akt pathway (24), have been used to study insulin action (25, 26), and are biologically responsive to IGF-I (24, 27). Even if lymphocytes are not taking part in growth physiology and have their own regulation, they use the same proximal molecular effectors as osteogenic cells to respond to IGF-I and could thus help unravel the genetic variability arising from the shared receptor and initial postreceptor effectors.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Patients and study design
Fifty children of European ancestry assessed by the ethnicity and birthplace of their four grandparents were recruited with height SD score (SDS) less than –2 and bone age less than 9 yr for boys and less than 8 yr for girls. None entered puberty within the year of study (using a criteria of Tanner I breasts for girls and testicular volume < 3 ml for boys). Twenty-five children were born SGA (length at birth <–2 SDS), and 25 had ISS. None had deficiencies of GH or other hormones, chromosomal disorders or syndromes, skeletal dysplasia, or metabolic disease. All children received GH as daily sc injections 6 d/wk. Children were seen by D.R. or P.B. as outpatients every 3 months for clinical examination and measurements of serum IGF-I. Height was measured in duplicate using a Harpenden stadiometer 1 yr before the onset of GH, at time of GH onset, and after 1 yr of treatment. Venous blood was sampled in the fasting state for lymphocyte studies at 9 or 12 months of GH treatment 24–48 h after the last sc injection of GH. Lymphocyte studies took place during late spring and summer to minimize the risk of intercurrent illnesses. Children were healthy at the time of study, with no sign of viral or other intercurrent infection and had normal temperature and an erythrocyte sedimentation rate smaller than 5 mm at the second hour.
In each studied subject, the growth response to GH was quantified as the value of the change (
) in height SDS (H-SDS). During the year preceding the onset of GH treatment, height SDS varied in a proportion of subjects, showing variable degrees of mild deceleration or acceleration. To more accurately evaluate the effect of GH supplementation on growth, we thus subtracted the 0H-SDS observed during the preceding year from the 1H-SDS recorded at the end of the first year of GH administration (1H-SDS – 0H-SDS).
IGF-I measurements
Serum IGF-I was measured by immunoradiometric assay after ethanol-acid extraction using DSL-5600 Active reagents (Diagnostic Systems Laboratories, Webster, TX). IGF-I SDS calculations were provided by DSL as reported (6). Intra- and interseries coefficients of variation were 3 and 1.5%, respectively at 50 ng/ml IGF-I, 1.5 and 3.7% at 260 ng/ml, and 2.5 and 3.9% at 760 ng/ml. The sensitivity was 4 ng/ml.
Akt phosphorylation in lymphocytes
Immediately (<10 min) after sampling, fresh circulating B and T lymphocytes were purified by separating EDTA blood samples on a Ficoll-Paque-Plus density gradient (Amersham, Little Chalfont, UK). Peripheral blood lymphocytes were isolated and platelets eliminated using magnetic cell separation (Miltenyi, Köln Germany). A fraction of cells was directly lysed in ice-cold buffer. The other fractions were stimulated for 15 min at 37 C with 125, 250, and 500 ng/ml human recombinant IGF-I (R&D Systems, Minneapolis, MN) and lysed. For immunoblot protein detection, 1.5 x 106 cells in extracts were electrophoresed through 9% polyacrylamide/SDS gels and transferred onto nitrocellulose membranes (Hybond ECL; Amersham), incubated with polyclonal antibodies to phospho- (P-)Akt and Akt (Cell Signaling Technology, Beverly, MA). After washing, membranes were incubated with horseradish peroxidase-conjugated secondary antibody and extensively washed in PBS/Tween 20. The proteins were visualized by the SuperSignal West Femto and West Pico chemiluminescent substrates (Pierce, Rockford, IL) and monitored in an imaging system (ChemioGenius2; Syngene, Frederick, MD). Band intensity was analyzed using GeneTools (Syngene). We tested the individuality of the results by performing 15 triplicate studies in lab students or staff sampled at monthly intervals. The SD/mean ratios for P-Akt/Akt measurements were 0.07–0.12 (intraassay) and 0.11–0.23 (intra-individual), indicating that our evaluation of Akt phosphorylation was reasonably consistent for a given individual, compared with a much greater inter-individual dispersion. The response of P-Akt/Akt ratio to 125, 250, and 500 ng/ml IGF-I were closely intercorrelated (P = 0.87). Results of Akt phosphorylation were expressed as fold increase of the P-Akt/Akt ratio at the different levels of IGF-I stimulation vs. the basal unstimulated value.
In a subset of healthy controls and patients, we tested two other methods for the measurement of P-Akt and Akt. The first was a cell-based ELISA where P-Akt and Akt were detected using indirect double labeling with an ELISA Kit (R&D Systems) and quantified with a Victor 3 fluorometer (Perkin-Elmer, Norwalk, CT). ELISA correlated with the previous chemiluminescence method (0.66 r 0.77 across tested individuals). In our hands, the limited magnitude of the response to IGF-I in several individuals did not allow a consistent degree of precision for P-Akt and Akt measurement. The other method that we tested used quantitative Western blot assays where blotted proteins were detected and quantified with the Odyssey Infrared Imaging System (LI-Cor Biosciences, Lincoln, NE). Scanning of membranes was performed at 680 nm (Akt) and 780 nm (P-Akt) with an Odyssey instrument. The LI-Cor results correlated closely with chemiluminescence (0.89 r 0.99 across tested individuals). We selected chemiluminescence (instead of LI-Cor) as the method of reference for the current study because it showed the best scores for precision and reproducibility, allows eye control of bands, and correlates closely with LI-Cor measurements and because the first GH-treated patients were studied using chemiluminescence.
Calculations and statistical analyses
Clinicians were kept blind to lymphocyte data and lab technicians to clinical data. The percent increase in serum IGF-I was calculated as the mean of three to four IGF-I measurements performed during the 6- to 12-month period of GH treatment divided by the mean IGF-I before treatment (mean of two measurements at time –6 months and at onset of GH) and was expressed as SDS. Their GH cumulative dose was calculated precisely over the first year of treatment for statistical analysis and expressed per kilogram body weight.
Results are presented as mean ± SD. The log-transformed values were used in statistical analyses for some parameters to more closely approximate a normal distribution. Relationships between growth response to GH administration and clinical and biological parameters were quantified using univariate linear regression models. Variables identified as significant in this univariate analysis were then included in a multiple regression model with a backward selection procedure. Statistical significance was considered as P < 0.05. All statistical analysis was performed using the R software package (29). Variables are expressed as mean ± SD except in Table 3
.
|
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The change in height SDS (0H-SDS) during the year preceding the start of GH treatment was –0.16 ± 0.22 SDS (range–0.6 to 0.4 SDS) and increased to 0.98 ± 0.38 SDS (1H-SDS range, 0–2.3 SDS) during the first year of GH therapy (P < 0.001). Consequently, the mean attained height changed from –2.54 ± 0.85 SDS at the start of treatment to –1.57 ± 0.91 SDS at 1 yr of treatment. Thus, the change in H-SDS induced by GH administration (1H-SDS – 0H-SDS) was 1.13 ± 0.48 SDS during the first year of treatment compared with the preceding year. This change in height velocity showed a near-normal distribution in the studied group of subjects (Fig. 1), a favorable condition for explicative statistical analyses.
|
The basal serum concentrations of IGF-I in the ISS and SGA children are shown in Table 1. The basal level of IGF-I was –1.14 ± 0.69 SDS (range, –2.5 to 1.4 SDS). IGF-I increased continuously to reach 0.51 ± 0.97 SDS after 1 yr of GH therapy, corresponding to a 2.78-fold increase from its baseline value (Table 2). No significant differences were observed between SGA and ISS subgroups.
|
|
Given the volume of blood necessary for the lymphocyte studies, we could not perform more than three levels of IGF-I stimulation per child. We found that the phosphorylation of Akt increased in response to graded concentrations of IGF-I according to dose-response kinetics that show a reasonable consistency across studied children. Half-maximal stimulatory IGF-I concentrations averaged 109 ± 6 ng/ml, and maximal stimulation of Akt phosphorylation was observed at 422 ± 29 ng/ml IGF-I. We thus used the value of Akt phosphorylation at 125 and 500 ng/ml IGF-I as approximations of these characteristics to calculate in vitro indices of IGF-I sensitivity of lymphocytes, as the fold increase in the P-Akt/Akt ratio from unstimulated baseline values.
Prediction of the growth response to GH treatment
Etiology of short stature, parental heights, age, or sex had no significant effect on changes in growth velocity during the first year of GH therapy in this sample of ISS or SGA children. Univariate analysis showed no significant correlation of 1H-SDS – 0H-SDS with basal or first-year IGF-I level but showed association with the first-year fold increase in IGF-I (P = 0.04) and with Akt phosphorylation in response to 125 ng/ml IGF-I in lymphocytes (P = 0.001). Akt phosphorylation in response to 500 ng/ml IGF-I was also predictive of the change in H-SDS during the first year of treatment (P = 0.003). The fold increase in IGF-I levels correlated with the cumulative GH dosage (r = 0.32; P = 0.025).
We found no association between IGF-I-stimulated Akt phosphorylation in lymphocytes and IGF-I levels before or during GH treatment; thus, serum IGF-I levels do not seem to reflect IGF-I sensitivity or resistance. This could, however, be different in a larger cohort of short children (6) or in the general population of children.
Results of the multivariate analysis indicate that IGF-I-stimulated Akt phosphorylation and GH cumulative dose were the two most important variables to explain the individual growth response occurring during the first year of GH treatment (Table 3
). GH dosage and the response of lymphocyte Akt phosphorylation to 125 ng/ml IGF-I were independently predictive of this growth response. Figure 2 describes the univariate correlation of GH dosage (Fig. 2A) and the response of lymphocyte Akt phosphorylation (Fig. 2B) with the growth response to GH administration. The fold increase in IGF-I level was not found to be a predictor independent of GH dosage of the growth response to GH administration.
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The current work attempted to evaluate whether a lymphocyte-based assay would reflect IGF-I sensitivity at the individual level. The phosphorylation of Akt provided a convenient IGF-I-stimulated biological parameter that recapitulates the proximal steps of IGF-I action. Our first observation was that this parameter showed individual consistency, with a coefficient of intra-individual reproducibility (SD/mean in triplicates) ranging from 0.11–0.23 in the controlled conditions of the current sampling and experiments. As expected, IGF-I was found to stimulate Akt phosphorylation in fresh lymphocytes according to a dose-response curve that again showed reasonably comparable kinetics across studied subjects.
The individuality and consistency of the IGF-I response allowed us to test whether the in vitro stimulation of lymphocytes by IGF-I could contribute as an index of IGF-I sensitivity to the variance of GH effects observed in GH-treated children. Looking for an association between IGF-I-stimulated lymphocytes and children’s growth faces the problem that lymphocytes are not involved in growth physiology. The proximal steps of the biochemical pathway of IGF-I action on cells, however, are shared by all IGF-I-sensitive cells (19, 22, 23, 24). Although there are many reasons to think that the regulation of cell responses, such as proliferation or differentiation, is closely dependent on tissue specificity, it is equally likely that the genetic variation expressed by certain effectors of IGF-I signaling induce parallel effects in different cell types. The current observation that the in vitro stimulation of Akt phosphorylation in lymphocytes was closely associated with the GH-induced gain in children’s growth rate supports that the early steps of IGF-I action are involved in the variance of IGF-I effects on growth. This observation will encourage the search for genomic and nongenomic factors capable of modulating the activity of these early steps of IGF-I signaling and hence GH pharmacogenomics mediated through IGF-I effects. Lymphocytes are a convenient source of gene expression studies (32), and it is interesting that the response of these cells to IGF-I shows some association with the clinical response to GH treatment. However, the individual variability of IGF-I effects on osteogenic cells and skeletal growth is far from being recapitulated by Akt phosphorylation, which represents at most 20% of overall variance in the GH-induced gain of growth in the current study. Other (likely numerous) sources of variability are to be expected from 1) non-IGF-I-dependent growth variability; 2) genetic variations outside the PI3K-Akt pathway, such as IGF-I transport, IGF-I signaling downstream of Akt, alternate IGF-I signaling pathways not tested here, etc.; 3) genetic factors modifying IGF-I transport and interaction with binding proteins; 4) nongenetic factors influencing the transport and action of IGF-I in vivo; and 5) factors that are active to regulate IGF-I sensitivity in chondrocytes but not in lymphocytes. In addition, patient age and long-term exposure to GH therapy are known to influence the variability of the growth response to GH, and possibly to IGF-I, adding other mechanisms of variability to those studied here during the first year of treatment.
In conclusion, we do not propose, given its experimental nature, that the stimulation of lymphocytes can be used as a practical index of IGF-I sensitivity allowing one to distinguish GH resistance from IGF-I resistance in clinical conditions. Lymphocytes can, however, serve as a tool for genetic studies attempting to associate IGF-I-induced changes in gene expression in the IGF-IR-Akt pathway with genome-wide polymorphisms, according to recently reported expression quantitative trait loci (eQTL) analyses (32). For such studies, hundreds of individuals will be needed, far beyond the size of the current sample. Our in vitro index of IGF-I sensitivity may also help dissect components of the variable clinical response to GH (skeletal growth and secondary metabolic effects) in research protocols exploring the efficacy of GH or IGF-I treatments at the individual level.
|
Footnotes |
|---|
Disclosure Statement: All of the authors (C.L., D.R., C.E., A.R., N.L., C.L.S., and P.B.) have nothing to declare.
First Published Online February 5, 2008
Abbreviations: IGF-IR, IGF-I receptor; IR, insulin receptor; IRS, IR substrate; ISS, idiopathic short stature; P-, phospho-; PI3K, phosphatidylinositol 3-kinase; SDS, SD score; SGA, short for gestational age.
Received November 20, 2007.
Accepted January 30, 2008.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
