Abstract
Stunted growth is a common complication of childhood diseases characterized by chronic inflammation or infections. We previously demonstrated that NSE/hIL-6 transgenic mice, overexpressing the inflammatory cytokine IL-6 since early phase of life, showed a marked growth defect associated with decreased IGF-I levels, suggesting that IL-6 is one of the factors involved in stunted growth complicating chronic inflammation in childhood. Here we show that NSE/hIL-6 mice have normal liver IGF-I production, decreased levels of IGF binding protein-3 (IGFBP-3) and increased serum IGFBP-3 proteolysis. Reduced IGFBP-3 levels results in a marked decrease in the circulating 150-kDa ternary complex, even in the presence of normally functional acid labile subunit. Pharmacokinetic studies showed that NSE/hIL-6 mice have accelerated IGF-I clearance. Patients with systemic juvenile idiopathic arthritis (s-JIA), a chronic inflammatory disease characterized by prominent IL-6 production and complicated by stunted growth associated with low IGF-I levels, have markedly decreased IGFBP-3 levels, increased serum IGFBP-3 proteolysis and normal acid labile subunit levels. Our data show that chronic overproduction of IL-6 causes decreased IGFBP-3 levels, resulting in a decreased association of IGF-I in the 150-kDa complex. Decreased levels of IGF-I appear to be secondary to increased clearance.
STUNTED GROWTH IS a common and well known complication of childhood diseases characterized by chronic inflammation and/or severe recurrent infections, such as systemic juvenile idiopathic arthritis (s-JIA), Crohn’s disease, cystic fibrosis, and chronic granulomatous disease (1–4). Several clinical observations show that growth impairment is related to the inflammatory activity of these diseases (1–4), suggesting that a factor (or factors) produced during chronic inflammatory responses is responsible for growth retardation. We have previously shown that chronic overproduction of the inflammatory cytokine IL-6 is at least one of the factors responsible for the impairment of linear growth in these conditions. In the IL-6 transgenic mice NSE/hIL-6 elevated circulating levels of IL-6, present since birth, cause a markedly reduced growth rate leading to adult mice that are 50–70% the size of wild-type littermates (5). While in NSE/hIL-6 mice the GH production is normal, overexpression of IL-6 is associated with a marked decrease in circulating IGF-I levels. Administration of IL-6 to nontransgenic animals induced a marked decrease in circulating IGF-I levels (5), supporting the direct effect of IL-6 on the IGF-I system. s-JIA is characterized by prominent IL-6 production that appears to explain several, if not all, the clinical and laboratory features of the disease (6, 7) and is frequently complicated by an impairment of linear growth associated with low IGF-I levels (8). In s-JIA low levels of IGF-I were found to be inversely correlated with serum IL-6 levels (5), supporting the conclusion that the relation between IL-6 and IGF-I observed in the NSE/hIL-6 model can be applied to human diseases. Indeed, other observations suggest that, in addition to s-JIA, this same mechanism may operate also in other diseases featuring inflammation and stunted growth. In patients with cystic fibrosis, GH production appears to be substantially normal (9), whereas circulating levels of IGF-I are reduced (10). Also in patients with Crohn’s disease, in which the production of IL-6 by inflamed mucosal tissue and the correlation between IL-6 levels and disease activity are well documented (11), circulating levels of IGF-I are decreased (12), whereas GH production is normal (13, 14).
In this study, we have extended the evaluation of the effects of the chronic overexpression of IL-6 on the IGF-I system and its relation with growth impairment by studying NSE/hIL-6 mice and patients with s-JIA. We report that chronic overexpression of IL-6 does not directly affect liver IGF-I production, but rather induces a marked decrease in the circulating levels of IGF binding protein-3 (IGFBP-3).
Materials and Methods
Animals and treatments
NSE/hIL-6 mice were generated using the NSE/hIL-6 construct that carries the rat neuro-specific enolase (NSE) promoter driving the expression of human IL-6 cDNA. These mice do not show histological or behavioral signs of neuron damage. Of the four stable lines generated, transgenic mice of lines 15 and 22 did not show growth impairment and had undetectable circulating levels of hIL-6 and normal levels of IGF-I. On the contrary, transgenic mice of lines 26 and 35 with high circulating levels of IL-6 since early phases of life presented a significantly reduced growth rate, that led to mice 50–70% the size of their age-matched littermates; their stunted growth was associated with a marked decrease in IGF-I levels. These mice have normal food intake, and normal hematic glucose (5). Unless otherwise indicated, all experiments involving NSE/hIL-6 mice were performed in 3-wk-old animals of line 26. Transgenic animals were identified by PCR analysis of DNA extracted from a tail segment, as described (5). Mice were maintained in standard conditions under a 12-h light, 12-h dark cycle, provided irradiated food (Mucedola, Settimo Milanese, Milan, Italy) and chlorinated water ad libitum. Procedures involving animals and their care were conducted in conformity with national and international laws and policies (EEC Council Directive 86/609, OJ L 358, 1, Dec. 12, 1987; Italian Legislative Decree 116/92, Gazzetta Ufficiale della Repubblica Italiana n. 40, Feb. 18, 1992; NIH guide for the Care and Use of Laboratory Animals, NIH Publication No. 85–23, 1985). To evaluate the effect of the administration of IL-6 to CB6F1 (C57Bl6xBalb/C) mice, 3-wk-old animals were treated ip with two doses at a 12-h interval of 10 μg/dose of human recombinant IL-6 (rhIL-6), resuspended in sterile pyrogen-free saline solution; control mice were injected ip with sterile pyrogen-free saline solution. Circulating IGFBPs were evaluated 12 h after the second rhIL-6 administration. For IGF-I clearance studies, 3-wk-old animals were treated with a single bolus injection in the caudal vein of 3 μg/g of body weight of recombinant human IGF-I (kindly provided by Pharmacia & Upjohn, Inc.). Plasma samples were collected at 15, 30, 60, 120, and 240 min after the injection.
Patients
Twenty-six patients (age range: 2–18 yr) fulfilling the diagnostic criteria for the diagnosis of s-JIA (15) were included in the study. All patients presented active disease at time of sampling as defined by the presence of synovitis on examination and were receiving nonsteroidal antiinflammatory drugs. Ten patients were not receiving oral glucocorticoids, whereas 16 were treated with low-dose glucocorticoids: 9 were receiving daily glucocorticoids (mean dose 0.32 mg/kg·day), whereas 7 were on alternate day regimen (mean dose 0.26/kg every other day). Fifteen patients were also receiving methotrexate. Since marked changes in circulating IL-6 levels occur during the febrile peak (6, 16), peripheral blood samples were collected during the morning hours, when all patients were afebrile. Thirty-five healthy children comparable for age, hospitalized for bone marrow donation or for minor surgical procedures, were used as controls. Permission for drawing of extra-blood during routine venipuncture was obtained from parents of all children.
Immunoblotting for the detection of IGFBP-3
Human serum samples (0.3 μl), diluted in nonreducing sample buffer (0.5 m Tris-HCl, pH 6.8, 20% glycerol, 10% SDS, 0.05% bromophenol blue), were electrophoresed on a 12% SDS-PAGE gel and then electroblotted onto 0.2 μm pore size Immunlite membranes (Bio-Rad Laboratories, Inc., Hercules, CA) at a constant voltage of 100 V for 2 h in blotting buffer (25 mm Tris, pH 8.3, 192 mm glycine, 20% methanol). Membranes were rinsed in 20 mm Tris, pH 7.5, 0.5 m NaCl (TBS); nonspecific binding was blocked by incubation for 1.5 h at room temperature in TBS, 5% nonfat dry milk (Bio-Rad Laboratories, Inc.). Membranes were incubated overnight at room temperature in the presence of 2 μg/ml of a rabbit polyclonal antibody to human IGFBP-3 (A. F. Schuetzdeller, Tubingen, Germany) in TBS, 0.05% Tween 20 (TTBS) or to murine IGFBP-3 (GroPep Pty. Ltd., Australia). The antibody was revealed by subsequent incubations with a biotinylated goat serum to rabbit immunoglobulin (Vector Laboratories, Inc., Burlingame, CA), at 1.4 μg/ml in TTBS for 1 h at room temperature, and with alkaline phosphatase AP-conjugated streptavidin (Roche Molecular Biochemicals, Mannheim, Germany) at 0.3 U/ml in TTBS for 1 h at room temperature. All steps were followed by three washes in TTBS, except the last which was followed by two additional washes in TBS. Membranes were then incubated with a chemiluminescent AP substrate (Immunstar, Bio-Rad Laboratories, Inc.) according to the instructions provided by the manufacturer, placed on plastic, and exposed to x-ray film (Kodak X-Omat AR, Eastman Kodak Co., Rochester, NY). Specific bands were quantified by densitometric analyses (BIO-GENE, Version 6.01). Intact IGFBP-3 was quantified by determining the intensity of the 38- to 42-kDa doublet. To evaluate in vivo proteolysis of IGFBP-3, which implies preassay exposure of endogenous serum IGFBP-3 to proteases in each sample, the major IGFBP-3 proteolytic fragment of 30 kDa was quantified. In vivo proteolysis was estimated by calculating the absorbance of the 30-kDa IGFBP-3 fragment over the sum of the intact IGFBP-3 and the 30-kDa IGFBP-3 fragment in the same lane. To correct for interassay variability, densitometric results were expressed as a percentage of a control serum, processed in parallel in each gel.
Western ligand blotting for human and murine IGFBPs
Human or murine samples (0.5 μl), diluted in nonreducing sample buffer were electrophoresed on a 12.5% SDS-PAGE gel and then electroblotted onto 0.2-μm pore size PVDF membranes (Bio-Rad Laboratories, Inc.) at a constant voltage of 100 V for 2 h in blotting buffer. Membranes were rinsed in TBS and nonspecific binding was blocked by incubation for 1.5 h at room temperature in TBS, 5% nonfat dry milk (Bio-Rad Laboratories, Inc.). Membranes were incubated overnight at room temperature in the presence of a modified biotinylated ligand for IGFBPs (A. F. Schuetzdeller) diluted 1:500 in TTBS. The binding was revealed by subsequent incubation with AP-conjugated streptavidin (Roche Molecular Biochemicals), as described above. All steps were followed by 3 washes in TTBS, except the last which was followed by two additional washes in TBS. Incubation with the AP substrate and exposure were performed as described above. For densitometric analysis, IGFBP-3 was identified according to Donahue et al. (17).
IGFBP-3 protease assay
Serum IGFBP-3 protease activity was evaluated with a nonradioactive method essentially as described by Davenport et al. (18). For human samples, 2.5 μl of test serum were mixed with 2.5 μl normal human serum and incubated for 16 h at 37 C in a total volume of 20 μl of PBS, pH 7.4. At the end of incubation the reaction was stopped by the addition of 5 μl of EDTA 100 mm. Murine serum samples were incubated for 3 h at 37 C. To evaluate the effect of protease inhibitors, similar incubations were performed in the presence of aprotinin (2 mg/ml) or EDTA (10 mm), as indicated. The amount of residual intact IGFBP-3 and of the fragments were visualized by immunoblotting as described above and quantified by densitometry.
Gelatin zimography
The presence of metalloproteinases (MMPs) was evaluated in sera by zimogram. Serum samples were dissolved in gel sample buffer in the absence of reducing agents and electrophoresed on 10% SDS-polyacrylamide substrate gels containing 0.28% gelatin type A. After electrophoresis, gels were washed twice for 30 min each time in 2.5% Triton X-100 at room temperature and then incubated in a fresh solution of 40 mm Tris pH 7.5, containing 10 mm CaCl2 and 0.2 m NaCl for 18 h at 37 C to allow enzymatic digestion of gelatin substrate. Proteolytic activity was visualized by staining gels in 0.2% Coomassie blue R-250 in 50% methanol and 10% acetic acid, followed by destaining in 50% methanol, 10% acetic acid solution.
Serum chromatographic profile
Serum pool from wild-type and transgenic mice were gel-filtrated by FPLC on HiPrep 16/60 S-200 column equilibrated in phosphate buffer saline, pH 7.2. Samples were eluted at 0.8 ml/min and fractions were collected at 0.8-min intervals. Immunoreactive IGF-I was evaluated in fractions pooled at 0.05 Kav intervals. In these conditions, the 150-kDa ternary complex elutes predominantly at 0.10–0.15 Kav, whereas the 35–45 binary complexes elute predominantly at 0.30–0.35 Kav. To evaluate the effect of the exogenous addition of IGFBP-3 and IGF-I on the formation of the ternary complex, we added 300 ng of recombinant human IGF-I and 3 μg of recombinant human IGFBP-3 (kindly provided by Dr. Mascarenhas, Celtrix Pharmaceuticals, Inc., Santa Clara, CA) to 1 ml of serum from wild-type or from transgenic NSE/hIL-6 mice. Samples were incubated overnight at 4 C and then subjected to gel-filtration.
Measurement of IGF-I
IGF-I was measured using a commercially available RIA that recognizes both murine and human IGF-I, according to instructions provided by the manufacturer (Nichols Institute Diagnostics, San Juan de Capistrano, CA). Before the assay, plasma samples (anticoagulated with EDTA, final concentration 5 mg/ml) were subjected to acid-ethanol extraction. Liver protein extracts for the measurement of IGF-I were prepared as described by Davenport et al. (19). Briefly, liver fragments were weighted, frozen in liquid nitrogen, and pulverized. Two sequential extractions with ice-cold 0.5 n HCl, followed by centrifugation at 13.000 rpm, were performed. Supernatants were collected from the two centrifugations, combined and subjected to C18 column (Waters, Division of Millipore Corp., Milford, MA) chromatography to remove IGFBPs. Eluates were exsiccated and resuspended in the assay buffer provided by the manufacturer of the IGF-I RIA.
Measurement of serum acid labile subunit (ALS) and serum IL-6 in patients with s-JIA
Serum IL-6 levels were measured with a hybridoma growth factor assay using B9 cells as previously described (6); the hybridoma growth factor activity in sera of patients with s-JIA was abolished by the addition of a goat antiserum to hIL-6 (not shown). Serum ALS levels were measured using a commercially available immunoassay, according to the instructions provided by the manufacturer (Diagnostics Systems Laboratories, Inc., Webster, TX). The detection limit of the assay was 0.7 ng/ml.
Statistical analysis
Results
Normal liver IGF-I expression in NSE/hIL-6 mice
Because IL-6 is a major inducer of the liver acute phase response (20) and because circulating IGF-I is produced essentially by the liver (21), we first hypothesized that IL-6 affected circulating IGF-I levels by directly inhibiting liver IGF-I production. To evaluate liver IGF-I production in NSE/hIL-6 mice, we measured IGF-I in liver protein extracts obtained from 20-d-old transgenic mice and their wild-type littermates and compared them with the circulating IGF-I levels in the same animals. While, as previously reported (5), transgenic mice had significantly lower (P = 0.002) circulating IGF-I levels (138.0 ± 72.4 ng/ml) than those of nontransgenic littermates (258.9 ± 80.2 ng/ml), no significant differences were observed when liver extract IGF-I levels were compared (transgenic: 97.9 ± 26.0 ng/g wet tissue; nontransgenic: 105.0 ± 39.1 ng/g wet tissue; P > 0.1) (Fig. 1). In agreement with an in vitro observation showing no effect of IL-6 on GH-induced IGF-I mRNA expression in cultured hepatocytes (22), our findings in vivo exclude a direct effect of IL-6 overexpression on liver IGF-I production, and show that the decrease in circulating IGF-I levels in NSE/hIL-6 mice cannot be explained by low liver IGF-I production.
IGF-I levels in plasma and liver extracts of transgenic (shaded bars; n = 13) and nontransgenic (white bars; n = 13) NSE/hIL-6 mice of line 26. Three-week-old animals were studied. Results are expresses as ng/ml for plasma levels and as ng/g of wet tissue for liver extracts and are shown as mean + sd, represented by the vertical bars. The P values of the difference between transgenic and nontransgenic mice are shown.
Decreased IGFBP-3 levels and IGFBP-3 proteolysis in NSE/hIL-6 mice
Because a significant portion of the circulating IGF-I is carried in a ternary complex with IGFBP-3 and a non-IGFBP, termed ALS, and because the half-life of IGF-I is markedly prolonged by its association in this ternary complex (23), it is also possible that a decrease in IGFBP-3 and/or in ALS may be responsible for the low circulating IGF-I levels.
We first evaluated IGFBP-3 levels in the NSE/hIL-6 transgenic mice comparing them with nontransgenic littermates. In agreement with the molecular mass reported by Donahue et al. (17) in their characterization of circulating IGFBPs in mice, IGFBP-3 visualized by Western ligand blotting appeared as two bands of 41.5 and 38.5 kDa, respectively. As shown in Fig. 2 for some representative samples, transgenic NSE/hIL-6 mice of line 26, with elevated circulating hIL-6 and growth defect (5), showed a marked decrease in the levels of IGFBP-3 when compared with nontransgenic littermates. Similar results were obtained with NSE/hIL-6 mice of line 35 (data not shown), the other line with elevated circulating hIL-6 and growth defect (5). Quantification by densitometric analysis in all mice tested showed that transgenic mice of line 26 have a significant decrease in IGFBP-3 to values that are approximately 50% of those of the corresponding nontransgenic littermates (Table 1). On the contrary, in mice of line 22, with undetectable circulating IL-6 and normal growth (see the materials and methods section), levels of IGFBP-3 were comparable between transgenic and nontransgenic mice. Further supporting the relation of the decrease in IGFBP-3 levels with the growth defect of NSE/hIL-6 mice, we found that circulating IGFBP-3 levels were significantly correlated with body weight and with circulating IGF-I levels both in transgenic and nontransgenic mice (Table 2).
Circulating IGFBPs of representative transgenic (T) and non-transgenic (NT) NSE/hIL-6 mice of lines 26 visualized by Western ligand blotting.
Circulating levels of IGFBP-3 in transgenic NSE/hIL-6 mice and wild-type littermates of line 26 (elevated IL-6 levels, low IGF-I levels and stunted growth) and of line 22 (undetectable IL-6 levels, normal IGF-I levels and normal growth
| (n) | IGFBP-3 (% reference serum) | ||
|---|---|---|---|
| Line 26 | Transgenics | 13 | 60.1 ± 29.1a |
| Wild-type | 13 | 136.2 ± 40.4 | |
| Line 22 | Transgenics | 5 | 104.3 ± 18.0b |
| Wild-type | 5 | 89.7 ± 23.7 |
| (n) | IGFBP-3 (% reference serum) | ||
|---|---|---|---|
| Line 26 | Transgenics | 13 | 60.1 ± 29.1a |
| Wild-type | 13 | 136.2 ± 40.4 | |
| Line 22 | Transgenics | 5 | 104.3 ± 18.0b |
| Wild-type | 5 | 89.7 ± 23.7 |
Mice were 3 wk old at time of bleeding. IGFBP-3 levels have been quantified by densitometric analysis of Western ligand blotting as described in the materials and methods section and expressed as a percentage of a reference serum of a wild-type mouse processed in parallel in each gel. Results are shown as mean ± sd.
P < 0.0001 vs. corresponding wild-type.
P > 0.1 vs. corresponding wild-type.
Circulating levels of IGFBP-3 in transgenic NSE/hIL-6 mice and wild-type littermates of line 26 (elevated IL-6 levels, low IGF-I levels and stunted growth) and of line 22 (undetectable IL-6 levels, normal IGF-I levels and normal growth
| (n) | IGFBP-3 (% reference serum) | ||
|---|---|---|---|
| Line 26 | Transgenics | 13 | 60.1 ± 29.1a |
| Wild-type | 13 | 136.2 ± 40.4 | |
| Line 22 | Transgenics | 5 | 104.3 ± 18.0b |
| Wild-type | 5 | 89.7 ± 23.7 |
| (n) | IGFBP-3 (% reference serum) | ||
|---|---|---|---|
| Line 26 | Transgenics | 13 | 60.1 ± 29.1a |
| Wild-type | 13 | 136.2 ± 40.4 | |
| Line 22 | Transgenics | 5 | 104.3 ± 18.0b |
| Wild-type | 5 | 89.7 ± 23.7 |
Mice were 3 wk old at time of bleeding. IGFBP-3 levels have been quantified by densitometric analysis of Western ligand blotting as described in the materials and methods section and expressed as a percentage of a reference serum of a wild-type mouse processed in parallel in each gel. Results are shown as mean ± sd.
P < 0.0001 vs. corresponding wild-type.
P > 0.1 vs. corresponding wild-type.
Correlation coefficient (Spearman’s R) and significance level (P) of the association of circulating IGFBP-3 levels with body weight and circulating IGF-I levels in 3-wk-old NSE/hIL-6 mice
| IGFBP-3 vs. body weight | IGFBP-3 vs. IGF-I | |
|---|---|---|
| Transgenics (n = 13) | Rs = 0.857 | Rs = 0.687 |
| P < 0.001 | P < 0.01 | |
| Wild-type (n = 16) | Rs = 0.952 | Rs = 0.636 |
| P < 0.0001 | P < 0.01 |
| IGFBP-3 vs. body weight | IGFBP-3 vs. IGF-I | |
|---|---|---|
| Transgenics (n = 13) | Rs = 0.857 | Rs = 0.687 |
| P < 0.001 | P < 0.01 | |
| Wild-type (n = 16) | Rs = 0.952 | Rs = 0.636 |
| P < 0.0001 | P < 0.01 |
Correlation coefficient (Spearman’s R) and significance level (P) of the association of circulating IGFBP-3 levels with body weight and circulating IGF-I levels in 3-wk-old NSE/hIL-6 mice
| IGFBP-3 vs. body weight | IGFBP-3 vs. IGF-I | |
|---|---|---|
| Transgenics (n = 13) | Rs = 0.857 | Rs = 0.687 |
| P < 0.001 | P < 0.01 | |
| Wild-type (n = 16) | Rs = 0.952 | Rs = 0.636 |
| P < 0.0001 | P < 0.01 |
| IGFBP-3 vs. body weight | IGFBP-3 vs. IGF-I | |
|---|---|---|
| Transgenics (n = 13) | Rs = 0.857 | Rs = 0.687 |
| P < 0.001 | P < 0.01 | |
| Wild-type (n = 16) | Rs = 0.952 | Rs = 0.636 |
| P < 0.0001 | P < 0.01 |
To evaluate proteolytic degradation of IGFBP-3, sera from transgenic and nontransgenic mice were incubated at 37 C and IGFBP-3 visualized by immunoblotting. A marked decrease in the intact IGFBP-3 and appearance of a major fragment at 20 kDa was observed in transgenic mice compared with nontransgenic mice (Fig. 3A). These changes were completely reversed by incubation in the presence of EDTA (Fig. 3A). Gelatin zymography showed the presence of a major 72-kDa band consistent with the molecular mass of MMP-2 (Fig. 3B). The additional band at 68 kDa is consistent with the molecular mass of a partially activated form of MMP-2. These data suggest that the decrease in serum IGFBP-3 in NSE/hIL-6 mice is, at least in part, due to increased proteolysis.
IGFBP-3 proteolysis in sera of NSE/hIL-6 mice. A, Sera from transgenic (T) or non-transgenic (NT) mice were incubated at 37 C in the absence (lanes 2 and 5) or in the presence of 10 mm EDTA (lanes 3 and 6) and compared with the nonincubated samples (lanes 1 and 4). IGFBP-3 was visualized by immunoblotting. The arrows indicate the 40-kDa band corresponding to intact IGFBP-3 and the 20-kDa fragment. Results from two representative samples (out of 4 tested) are shown. B, Gelatin zymography of sera from transgenic (T) or nontransgenic (NT) NSE/hIL-6 mice. Zymography was performed using 10% SDS-polyacrylamide/SDS gels containing 0.28% gelatin type A as described in Materials and Methods. Results from two representative sera (out of 4 tested) are shown.
IL-6 administration decreases serum IGFBP-3 in CB6F1 mice
To verify whether IL-6 directly induces a decrease in circulating IGFBP-3 levels in nontransgenic mice, we administered recombinant human IL-6 (rhIL-6) to mice of the same strain of the NSE/hIL-6 transgenics. CB6F1 mice were treated with rhIL-6 using a treatment scheme that we have previously shown to cause a significant decrease in IGF-I without affecting food intake (5), therefore ruling out a possible effect of fasting on IGFBP-3 levels. Administration of rhIL-6 to CB6F1 mice resulted in a significant (P < 0.001) decrease in IGFBP-3 levels 24 h after the first treatment compared with saline-treated mice (IL-6 treated: IGFBP-3 53.0 ± 18.1%; saline treated: 128.8 ± 36.5% of a reference serum) (Fig. 4). The findings in NSE/hIL-6 and CB6F1 mice show that elevated in vivo levels of IL-6 cause a significant decrease in circulating IGFBP-3 levels.
Effect of the administration of recombinant hIL-6 to CB6F1 mice on circulating IGFBP-3 levels. Three-week-old CB6F1 mice (8 mice/group) were treated with 2 ip injections, at a 12-h interval, of recombinant human IL-6 (10 μg/dose) or of sterile pyrogen-free saline. A blood sample was collected 12 h after the second injection. Circulating IGFBP-3 levels were quantified by densitometric analysis of Western ligand blotting, and expressed as a percentage of a reference serum processed in parallel in each gel. The P value of the difference between transgenic and nontransgenic mice is shown.
Decrease in the 150-kDa ternary complex and presence of functionally normal ALS in NSE/hIL-6 mice
In the NSE/hIL-6 transgenic mice the 35- to 45-kDa pool was significantly reduced but still present (Fig. 5). This finding is consistent with the presence of IGF-I bound to the low amounts of IGFBP-3 present in NSE/hIL-6 transgenic mice and/or to the IGFBP-1, -2, and -4. More importantly, we found that in the NSE/hIL-6 transgenics the amount of IGF-I recovered in the 150-kDa complex was markedly reduced compared with wild-type mice (Fig. 5).
Elution profile of immunoreactive IGF-I obtained by HiPrep 16/60 S-200 gel filtration in neutral conditions of pooled sera from nontransgenic and transgenic NSE/hIL-6 mice. IGF-I was evaluated in fractions pooled at 0.05 Kav intervals. In these conditions the 150-kDa ternary complex elutes predominantly at 0.10–0.15 Kav, whereas the 35–45 binary complexes elute predominantly at 0.30–0.35 Kav. Shown are representative results from two different chromatographies.
The marked reduction in the ternary complex may be secondary to the decrease in IGFBP-3, described in the previous paragraph, and/or to the presence of a decrease in functional circulating ALS. In the absence of available antibody that recognizes murine ALS, we evaluated the presence of functional ALS by measuring 150-kDa complex formation following exogenous addition of recombinant human IGF-I and recombinant human IGFBP-3 to serum from NSE/hIL-6 mice. In the absence of exogenous additions the amount of IGF-I recovered in the 150-kDa ternary complex was markedly lower in the NSE/hIL-6 transgenic compared with the wild-type littermates (Table 3). Following exogenous addition of IGF-I and IGFBP-3, the amount of IGF-I recovered in the 150-kDa ternary complex was comparable between transgenic NSE/hIL-6 mice and their wild-type littermates (Table 3). These data show that sera from NSE/hIL-6 mice contain a sufficient amount of functional ALS to form the ternary complex in the presence of exogenous IGFBP-3 and IGF-I. Therefore, the marked decrease in the 150-kDa ternary complex in the serum of NSE/hIL-6 mice appears to be secondary to the marked decrease in IGFBP-3 levels.
Amount of IGF-I (expressed as percentage of the total IGF-I recovered) recovered in the 150-kDa complex, in the 35- to 45-kDa complex, and as free IGF-I before and after exogenous addition of recombinant IGF-I and recombinant IGFBP-3 to serum pool from transgenic or wild-type mice
| Exogenous IGF-I IGFBP-3 | IGF-I recovered (Percentage of total recovered IGF-I) | |||||
|---|---|---|---|---|---|---|
| IGF-I in 150 kDa (0.05–0.2 Kav) | IGF-I in 35–45 kDa (0.25–0.55 Kav) | Free IGF-I (0.6–0.75 Kav) | ||||
| No | Yes | No | Yes | No | Yes | |
| Wild-type | 31.0 | 38.3 | 67.7 | 48.6 | 0.8 | 13.1 |
| Transgenics | 13.6 | 36.5 | 84.9 | 47.7 | 1.6 | 15.8 |
| Exogenous IGF-I IGFBP-3 | IGF-I recovered (Percentage of total recovered IGF-I) | |||||
|---|---|---|---|---|---|---|
| IGF-I in 150 kDa (0.05–0.2 Kav) | IGF-I in 35–45 kDa (0.25–0.55 Kav) | Free IGF-I (0.6–0.75 Kav) | ||||
| No | Yes | No | Yes | No | Yes | |
| Wild-type | 31.0 | 38.3 | 67.7 | 48.6 | 0.8 | 13.1 |
| Transgenics | 13.6 | 36.5 | 84.9 | 47.7 | 1.6 | 15.8 |
Serum pool was incubated overnight at 4C in the absence or in the presence of recombinant IGF-I (300 ng/ml) and recombinant IGFBP-3 (3μ g/ml) and then subjected to gel filtration as described in Materials and Methods. Shown are representative results of two different chromatographies.
Amount of IGF-I (expressed as percentage of the total IGF-I recovered) recovered in the 150-kDa complex, in the 35- to 45-kDa complex, and as free IGF-I before and after exogenous addition of recombinant IGF-I and recombinant IGFBP-3 to serum pool from transgenic or wild-type mice
| Exogenous IGF-I IGFBP-3 | IGF-I recovered (Percentage of total recovered IGF-I) | |||||
|---|---|---|---|---|---|---|
| IGF-I in 150 kDa (0.05–0.2 Kav) | IGF-I in 35–45 kDa (0.25–0.55 Kav) | Free IGF-I (0.6–0.75 Kav) | ||||
| No | Yes | No | Yes | No | Yes | |
| Wild-type | 31.0 | 38.3 | 67.7 | 48.6 | 0.8 | 13.1 |
| Transgenics | 13.6 | 36.5 | 84.9 | 47.7 | 1.6 | 15.8 |
| Exogenous IGF-I IGFBP-3 | IGF-I recovered (Percentage of total recovered IGF-I) | |||||
|---|---|---|---|---|---|---|
| IGF-I in 150 kDa (0.05–0.2 Kav) | IGF-I in 35–45 kDa (0.25–0.55 Kav) | Free IGF-I (0.6–0.75 Kav) | ||||
| No | Yes | No | Yes | No | Yes | |
| Wild-type | 31.0 | 38.3 | 67.7 | 48.6 | 0.8 | 13.1 |
| Transgenics | 13.6 | 36.5 | 84.9 | 47.7 | 1.6 | 15.8 |
Serum pool was incubated overnight at 4C in the absence or in the presence of recombinant IGF-I (300 ng/ml) and recombinant IGFBP-3 (3μ g/ml) and then subjected to gel filtration as described in Materials and Methods. Shown are representative results of two different chromatographies.
Increased clearance of IGF-I in NSE/hIL-6 transgenic mice
Because it has been demonstrated that plasma IGF-I clearance is accelerated in conditions characterized by low IGFBP-3 levels (19, 24–26), we evaluated plasma IGF-I half-life in NSE/hIL-6 mice. Figure 6 shows the time course of the disappearance of iv injected IGF-I from plasma of wild-type and NSE/hIL-6 transgenic mice. As shown in Table 4, the total plasma clearance of IGF-I was markedly increased (+99%) in the NSE/hIL-6 transgenic mice. This faster clearance resulted in a lower systemic exposure (AUC0-∞) and in a shorter elimination phase of IGF-I in the NSE/hIL-6 transgenics than in wild-type mice. Together with the previously shown results on liver IGF-I levels, this finding shows that in NSE/hIL-6 transgenic mice reduced circulating levels of IGF-I are secondary to accelerated clearance and not to decreased liver production.
Time course of the disappearance of iv injected IGF-I in wild-type (open circles) and NSE/hIL-6 transgenic mice (closed circles). Mice were injected with an iv bolus of 3 μg/g of body weight of recombinant human IGF-I. Results are shown as means + se of 4 animals for each time point.
Pharmacokinetics of IGF-I after a single iv bolus injection in wild-type and NSE/hIL-6 transgenic mice
| T1/2 (α) (min) | T1/2 (β) (min) | AUC0–∞ (μg/min·ml) | CL (ml/min·g) | |
|---|---|---|---|---|
| Wild-type | 21.5 | 194.8 | 204.8 | 0.015 |
| NSE/hIL-6 transgenics | 18.1 | 126.8 | 102.6 | 0.029 |
| Transgenic/wild-type (%) | 84.2 | 65.1 | 50.1 | 199 |
| T1/2 (α) (min) | T1/2 (β) (min) | AUC0–∞ (μg/min·ml) | CL (ml/min·g) | |
|---|---|---|---|---|
| Wild-type | 21.5 | 194.8 | 204.8 | 0.015 |
| NSE/hIL-6 transgenics | 18.1 | 126.8 | 102.6 | 0.029 |
| Transgenic/wild-type (%) | 84.2 | 65.1 | 50.1 | 199 |
T1/2 (α), Distribution half-life; T1/2 (β), elimination half-life; AUC0–∞, area under the concentration-time curve from time 0 to infinity; CL, total plasma clearance.
Pharmacokinetics of IGF-I after a single iv bolus injection in wild-type and NSE/hIL-6 transgenic mice
| T1/2 (α) (min) | T1/2 (β) (min) | AUC0–∞ (μg/min·ml) | CL (ml/min·g) | |
|---|---|---|---|---|
| Wild-type | 21.5 | 194.8 | 204.8 | 0.015 |
| NSE/hIL-6 transgenics | 18.1 | 126.8 | 102.6 | 0.029 |
| Transgenic/wild-type (%) | 84.2 | 65.1 | 50.1 | 199 |
| T1/2 (α) (min) | T1/2 (β) (min) | AUC0–∞ (μg/min·ml) | CL (ml/min·g) | |
|---|---|---|---|---|
| Wild-type | 21.5 | 194.8 | 204.8 | 0.015 |
| NSE/hIL-6 transgenics | 18.1 | 126.8 | 102.6 | 0.029 |
| Transgenic/wild-type (%) | 84.2 | 65.1 | 50.1 | 199 |
T1/2 (α), Distribution half-life; T1/2 (β), elimination half-life; AUC0–∞, area under the concentration-time curve from time 0 to infinity; CL, total plasma clearance.
Decreased intact IGFBP-3 and normal ALS levels in patients with s-JIA
s-JIA is a chronic inflammatory disorder that associates high levels of IL-6, stunted growth and decreased levels of IGF-I (4, 6–8). We have previously shown that in patients with s-JIA IGF-I levels were inversely correlated to IL-6 levels (5). Evaluation of serum IGFBP-3 levels by immunoblotting showed that patients with s-JIA had a marked decrease in intact IGFBP-3 (38- to 42-kDa doublet), (Fig. 7A for 5 representative patients). The intensity of the major 30-kDa proteolytic fragment did not appear to be markedly increased. Because levels of IGFBP-3 increase progressively with age, to evaluate the results obtained by densitometric analyses, patients and controls were divided in 4 age groups. As shown in Fig. 7B, in patients with s-JIA levels of the intact IGFBP-3 (38- to 42-kDa doublet) were lower than in the age-matched controls. When, by using Z scores, intact IGFBP-3 levels (38- to 42-kDa doublet) in the whole s-JIA patient population (Z score:− 0.452 ± 0.487) were compared with the levels of controls (Z score:−0.030 ± 0.099), the difference was highly significant (P < 0.0001). On the contrary, similarly to indirect evidence in NSE/hIL-6 mice, patients with s-JIA had serum levels of ALS (18.5 ± 5.9 μg/ml) that were comparable to those of controls (22.3 ± 9.2 μg/ml) (data not shown).
Circulating IGFBP-3 in healthy controls and in patients with s-JIA. A, Serum IGFBP-3 visualized by immunoblotting in 4 representative controls and 5 representative patients with s-JIA. The intact 38- to 42-kDa doublet is indicated by the arrow. B, Circulating levels of intact IGFBP-3 (38- to 42-kDa doublet) in healthy controls (white bars) and in patients with s-JIA (shaded bars) divided in the indicated age-groups. IGFBP-3 levels have been quantified by densitometric analysis of immunoblotting as described in Materials and Methods section and expressed as a percentage of a reference serum processed in parallel in each gel. The statistical significance level of the differences between patients and controls of each age-group are shown.
Evaluation of the possible relation of IGFBP-3 levels with glucocorticoid treatment shows comparable levels of intact IGFBP-3 in patients not receiving glucocorticoids (Z score: −0.534 ± 0.439) and in patients receiving them (Z score: −0.482 ± 0.515), suggesting that glucocorticoids do not directly affect IGFBP-3 levels, at least at the low doses administered in our patients. As expected, we found a significant direct correlation of serum IGFBP-3 levels with circulating IGF-I levels (Table 5). In addition, further supporting a direct relation between IL-6 production and IGFBP-3 levels, in patients with s-JIA serum levels of IL-6 were significantly inversely correlated with serum IGFBP-3. Moreover, a significant inverse correlation was also found with clinical and laboratory parameters of disease activity, such as the number of the joints with active arthritis, erythrocyte sedimentation rate values and serum C-reactive protein levels (Table 5). No significant correlation of the same parameters with serum ALS levels were found (data not shown).
Correlation coefficient (Spearman’s R) and significance level (P value) of the association of serum IGFBP-3 levels (Z score) with circulating IGF-I or IL-6 levels and with the number of joints with active arthritis (NAA), serum C-reactive protein (CRP) concentrations and erythrocyte sedimentation rate (ESR) values
| Spearman’s R | P value | |
|---|---|---|
| IGF-I | 0.449 | 0.013 |
| IL-6 | −0.514 | 0.011 |
| NAA | −0.534 | 0.011 |
| CRP | −0.603 | 0.005 |
| ESR | −0.497 | 0.011 |
| Spearman’s R | P value | |
|---|---|---|
| IGF-I | 0.449 | 0.013 |
| IL-6 | −0.514 | 0.011 |
| NAA | −0.534 | 0.011 |
| CRP | −0.603 | 0.005 |
| ESR | −0.497 | 0.011 |
Correlation coefficient (Spearman’s R) and significance level (P value) of the association of serum IGFBP-3 levels (Z score) with circulating IGF-I or IL-6 levels and with the number of joints with active arthritis (NAA), serum C-reactive protein (CRP) concentrations and erythrocyte sedimentation rate (ESR) values
| Spearman’s R | P value | |
|---|---|---|
| IGF-I | 0.449 | 0.013 |
| IL-6 | −0.514 | 0.011 |
| NAA | −0.534 | 0.011 |
| CRP | −0.603 | 0.005 |
| ESR | −0.497 | 0.011 |
| Spearman’s R | P value | |
|---|---|---|
| IGF-I | 0.449 | 0.013 |
| IL-6 | −0.514 | 0.011 |
| NAA | −0.534 | 0.011 |
| CRP | −0.603 | 0.005 |
| ESR | −0.497 | 0.011 |
IGFBP-3 proteolysis in patients with s-JIA
To evaluate whether IGFBP-3 proteolysis had occurred in vivo before the immunoblotting analyses (termed preassay IGFBP-3 proteolysis), the ratio of IGFBP-3 30-kDa fragment/IGFBP-3 30-kDa fragment + intact IGFBP-3 was calculated for each sample and expressed as a percentage of a control serum run in parallel in each gel. In patients with s-JIA (n = 35) preassay IGFBP-3 proteolysis was significantly (P = 0.003) higher than in controls (n = 27) (s-JRA: 146.1 ± 31.5%; CTRL: 120.6 ± 32.4%).
To determine if a serum proteolytic activity was present, test sera were mixed with control human serum, then incubated overnight, and IGFBP-3 visualized by immunoblotting. As shown in Fig. 8A for four representative samples (1 controls and 3 patients with s-JIA), overnight incubation at 37 C resulted in the decrease in intact IGFBP-3 and in the increase in the 30-kDa proteolytic fragment, compared with the same samples incubated at −20 C. Densitometric analysis showed that after overnight incubation at 37 C 95.4 ± 10.5% of the intact IGFBP-3 38- to 42-kDa doublet was detected in samples from healthy controls (n = 11), whereas only 68.4 ± 24.4% of the intact IGFBP-3 38-43 to 42-kDa doublet was detected in samples from patients with s-JIA (n = 13) (P < 0.001 vs. controls) (Fig. 8B). In addition, a significant (P < 0.001) increase in the 30-kDa proteolytic fragment was detected in samples from s-JIA patients (169.1 ± 78.7%), compared with samples from controls (90.7 ± 14.3%) (Fig. 8B). These changes were completely prevented by incubation in the presence of the metalloproteinase inhibitor EDTA or the serine protease inhibitor aprotinin (Fig. 8C for one representative sample).
Proteolysis of IGFBP-3 by sera from patients with systemic JIA. A, Sera from controls (CTRL) or patients with systemic JIA (s-JIA) were mixed as described in Materials and Methods section with normal human serum and incubated at 37 C (lanes 1, 3, 5, 7) or at −20 C (lanes 2, 4, 6, 8). IGFBP-3 was visualized by immunoblotting (black arrow: 38- to 42-kDa intact doublet; white arrow: 30-kDa proteolytic fragment). Results from 4 representative sera (1 controls and 3 s-JIA patients) are shown. B, Densitometric analysis of the IGFBP-3 intact 38- to 42-kDa doublet and of the 30-kDa proteolytic fragment in a total of 11 sera from controls and 13 sera from patients with s-JIA treated as described for panel A. Results are expressed as percent of the intensities of the 38- to 42-kDa doublet and of the 30 kDa proteolytic fragment in the sample incubated at 37 C relative to the intensities of the same bands in the same sample incubated at 20 C. C, Effect of EDTA or aprotinin on IGFBP-3 proteolysis of sera from patients with s-JIA. The reaction mixture was prepared as above and incubated at −20 C (lane 1) or at 37 C in the absence (lane 2) or in the presence of 10 mm EDTA (lane 3) or 2 mg/ml aprotinin (lane 4). Results from one representative serum are shown. Black arrow, 38- to 42-kDa intact doublet; white arrow, 30-kDa proteolytic fragment. *, P < 0.001 vs. controls.
Discussion
We have previously reported that NSE/hIL-6 mice, with high levels of circulating IL-6 since birth, present a marked decrease in growth rate associated with a decrease in circulating IGF-I levels (5). Because IL-6 is an inflammatory cytokine whose levels are markedly increased in a variety of chronic inflammatory diseases and because low IGF-I levels are characteristically present in these diseases (8, 10, 12), NSE/hIL-6 mice represent an animal model of the growth impairment associated with chronic inflammation in childhood, demonstrating that chronic overproduction of IL-6 is at least one of the mechanisms by which chronic inflammation affects linear growth.
In this study, we report that NSE/hIL-6 mice have normal liver IGF-I production, as shown by the comparable levels of IGF-I protein in liver extracts. This finding is in agreement with a previous in vitro observation demonstrating that IL-6 does not affect GH-induced IGF-I mRNA expression in cultured primary hepatocytes (22) and shows that the decreased circulating IGF-I levels in NSE/hIL-6 mice are not caused by decreased liver production.
A significant portion of the circulating IGF-I is carried in a 150-kDa ternary complex with IGFBP-3 and ALS. Because the half life of IGF-I is markedly prolonged by its association in this ternary complex from less than 10 min to 16 h (23), it is also possible that impaired formation of the ternary complex due to a decrease in IGFBP-3 and/or in ALS may be responsible for the low circulating IGF-I levels. Indeed, in this study we found that in NSE/hIL-6 mice the amount of IGF-I recovered in the 150-kDa ternary complex is markedly lower than in the corresponding wild-type littermates. Addition of exogenous IGFBP-3 and IGF-I to sera from NSE/hIL-6 mice led to efficient formation of the 150-kDa ternary complex, suggesting the presence of functionally normal serum ALS in NSE/hIL-6 mice. Moreover, we found normal ALS levels in patients with s-JIA. All together, these findings show that chronic overproduction of IL-6 does not affect ALS levels in vivo. This conclusion is supported by the in vitro observation that IL-6 does not inhibit spontaneous or GH-induced release of ALS from cultured hepatocytes (27).
On the contrary, we found that NSE/hIL-6 mice have markedly decreased levels of circulating IGFBP-3. This decrease in IGFBP-3 cannot be attributed to the effect of other inflammatory cytokines, such as IL-1 or TNF because their levels are undetectable in NSE/hIL-6 mice (data not shown). Further supporting a direct effect of IL-6 on IGFBP-3 levels, acute administration of IL-6 to nontransgenic CB6F1 mice resulted in a marked decrease in circulating IGFBP-3 levels. Caloric and protein restriction cause a reduction in the levels of IGFBP-3 both in humans and rodents (24, 28). However, the effect of IL-6 on IGFBP-3 levels cannot be explained by a reduced food intake because 1) NSE/hIL-6 mice have normal glucose levels and normal food intake (5); and 2) in the IL-6-treated nontransgenic CB6F1 mice, a modest decrease in food intake occurs only 24 h later than the decrease in IGFBP-3 (data not shown). Similarly to the NSE/hIL-6 mice, patients with s-JIA have markedly decreased levels of circulating IGFBP-3. Two previous studies (29, 30) reported low IGFBP-3 levels in JIA. However, in both studies the decrease in IGFBP-3 levels was much lower than that reported in our study. In one study, patients with the three different onset forms of JIA were studied together (29). The different method used for the detection of IGFBP-3 may explain this partial discrepancy. While in this study IGFBP-3 was visualized by immunoblotting and quantified by densitometric analysis of the intact 38- to 42-kDa doublet, in these two studies serum IGFBP-3 levels were measured using an ELISA, that may detect both intact IGFBP-3 and its proteolytic fragment(s) (see below).
Decreased IGFBP-3 levels have been reported in patients with Crohn’s disease or with cystic fibrosis (31, 32). We found that in patients with s-JIA levels of intact IGFBP-3 were inversely correlated with serum IL-6 levels and with clinical and laboratory parameters of disease activity. A direct relationship between increased IL-6 and reduced IGFBP-3 is also suggested by 1) the close relation of a rapid fall in IL-6 and CRP levels with a significant increase in IGFBP-3 levels following elementary diet in Crohn’s disease (31, 33); and 2) the correlation between IL-6 production and decreased levels of IGF-I and IGFBP-3 in HIV-infected children with decreased height velocity (34, 35). All together, these clinical observations, as well as our results in mice, strongly suggest that the chronic overproduction of IL-6 is responsible for the decrease in circulating IGFBP-3 in chronic inflammation/infection in childhood.
The mechanism by which IL-6 induces a decrease in IGFBP-3 levels remains to be clarified. While IGF-I and ALS are produced by hepatocytes and their production is not affected by IL-6 (see above), IGFBP-3 is produced by Kupffer cells (36). Because Kupffer cells constitutively express IL-6 receptor (37), it is conceivable to hypothesize that IL-6 may directly affect IGFBP-3 production. On the other hand, in this study we show that both the NSE/hIL-6 transgenic mice and patients with s-JIA have increased proteolysis of serum IGFBP-3, suggesting that the IL-6 induced decrease in intact IGFBP-3 levels is, at least in part secondary to IGFBP-3 proteolysis. In this respect it is noteworthy that IL-6 have been shown to induce production of several proteases including cathepsin B and L (38, 39) and, more interestingly, of MMP from a variety of cell types (40–43). It is noteworthy that the serum IGFBP-3 proteolytic activity is inhibitable by aprotinin and EDTA, a pattern consistent with metal-dependent proteases, and that gelatin zymography of serum of NSE/hIL-6 mice showed the presence of MMP-2.
Whatever the mechanism(s) of IL-6-induced decrease in IGFBP-3 levels, our findings in mice show that the decrease in intact IGFBP-3, even in the presence of functionally normal ALS, leads to defective association of IGF-I in the 150-kDa ternary complex. Incidentally, also in patients with s-JIA IGF-I is recovered essentially in the 35- to 45-kDa complex (data not shown). In addition, we report that NSE/hIL-6 mice show shortened plasma IGF-I half-life and accelerated IGF-I clearance. The latter observation is in agreement with studies in animals showing that conditions with low levels of IGFBP-3 are associated with a marked decrease in IGF-I half-life and accelerated clearance leading to low IGF-I levels even in the presence of normal liver IGF-I production (24, 44). Further supporting a direct relationship of IGFBP-3 levels with IGF-I, we found that IGFBP-3 levels were directly correlated with IGF-I levels both in NSE/hIL-6 mice and in patients with s-JIA.
In addition to IL-6, other inflammatory cytokines, such as IL-1 and TNF, have been shown to affect the IGF-I system. However, the changes induced by IL-6 appear to be different from those induced by IL-1 and TNF. Both TNF and IL-1 induce decreased circulating levels of IGF-I in rats (45, 46), but, in contrast to what found for IL-6, this effect of IL-1 and TNF appears to be secondary to decreased liver IGF-I production, possibly mediated by reduced expression of GH receptors on hepatocytes (22, 45–47). While we found that IL-6 overproduction caused a marked decreased in IGFBP-3 levels, administration of IL-1 or TNF to rats did not result in a significant decrease in IGFBP-3 levels (45, 46). In addition, while IL-1 has been shown to suppress ALS production by primary hepatocytes (27, 48), our results show that chronic overproduction of IL-6 does not affect ALS levels in vivo.
In conclusion, in this study we show that chronic overproduction of IL-6 results in decreased IGFBP-3 levels, secondary at least in part, to increased proteolysis, and in decreased association of IGF-I in the 150-kDa ternary complex, even in the presence of normal ALS. Because of the normal IGF-I liver expression and of the shortened plasma IGF-I half-life in the NSE/hIL-6 mice, decreased levels of IGF-I appear to be secondary to increased IGF-I clearance. These findings may have relevance for the identification of treatments aimed at correcting the growth defect associated with chronic inflammation/infection in childhood. GH treatment of patients with s-JIA, as well as with other childhood chronic inflammatory diseases, has given unsatisfactory responses (29, 49, 50). Our results imply that therapies aimed at correcting the abnormalities of the IGF-I system, such as administration of IGF-I and/or IGFBP-3, may be more effective in the treatment of the growth impairment associated with chronic inflammation in childhood.
Acknowledgements
We wish to thank Dr. Mario Regazzi-Bonora for his help in the analysis of phrmacokinetic data, Dr. Mohamad Maghnie for critical discussion during the preparation of the manuscript, Dr. Desmond Mascarenhas for kindly providing recombinant human IGFBP-3, and Nicola Corea for his technical assistance.
Abbreviations
- ALS
Acid labile subunit
- IGFBP-3
IGF binding protein-3
- MMP
metalloproteinase
- NSE
neuro-specific enolase
- s-JIA
systemic juvenile idiopathic arthritis.
