| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Institut National de la Santé et de la Recherche Médicale Unité 714 (N.B.L., A.B., P.M., S.B.), Institut Biomédical des Cordeliers, 75006 Paris, France; Institut National de la Santé et de la Recherche Médicale Unité 515 (D.S., Y.L.B., M.B.), Hôpital St-Antoine, 75012 Paris, France; Université Pierre et Marie Curie (Y.L.B.), Paris 6, 75005 Paris, France; and Université Denis Diderot (A.B.), Paris 7, 75005 Paris, France
Address all correspondence and requests for reprints to: Sylvie Babajko, Laboratoire de Biologie Oro-faciale et Pathologie, Institut National de la Santé et de la Recherche Médicale Unité 714, Institut biomédical des Cordeliers, 15/21 rue de l’école de Médecine, 75006 Paris, France. E-mail: Sylvie.Babajko{at}bhdc.jussieu.fr.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In all biological fluids, IGFs are noncovalently bound to high affinity binding proteins (IGFBPs), six of which have been characterized to date (4). As well as carrying IGFs in the bloodstream, these proteins modulate their half-lives and cellular bioavailability and can either potentiate or inhibit their mitogenic effects (5). In addition, some of them have intrinsic activities that are distinct from their ability to bind IGFs, which contributes to specific IGFBP activity (6, 7).
The IGF system has an important pleitrophic impact on bone and cartilage cells (8). IGF-I appears to be one of the most important growth factors affecting the anabolism of chondrocyte extracellular matrix constituents and bone remodeling. Transgenic mouse models have established the key role the IGF system plays in regulating of normal fetal growth, development, and ossification (9, 10, 11). However, such a role for the endocrine IGFBP-1 has not been studied for bone growth. To identify whether the endocrine function for IGFBP-1 is required during development, mice overexpressing human IGFBP-1 (hIGFBP-1) were produced using 
1-antitrypsin promoter, which allows a liver-specific expression from embryonic day (E)14 (12). This experimental model allows IGFBP-1 to exert its endocrine action, mimicking the physiological and pathological human situations. Thus, IGFBP-1 function and growth effects can be characterized during development, and IGFBP-1 fetal production can be distinguished from fetomaternal relations.
The maternal nutrient supply to the developing oocyte, embryo, or fetus has long been recognized as one of the principal environmental factors influencing the development of offspring. A reliable and balanced supply of amino acids, lipids, and carbohydrates is required to support the high cell proliferation rates and the key developmental processes that take place during the embryonic and fetal stages of life. It has been shown in rats, that IUGR is associated with lower hepatic glycogen storage and neonatal hypoglycemia (13).
The aim of the present study was to investigate the effects of the endocrine pathway of IGFBP-1 on fetal growth by analyzing both osteogenesis and carbohydrate resources, both of these parameters being directly related to growth. Indeed, in human pathology, serum IGFBP-1 levels are found to be markedly increased in cases of defects associated with IUGR (14), and our experimental model could mimic this pathological situation. The present study shows that endocrine overexpression of hIGFBP-1 may affect body growth and skeletal formation as well as biomineralization in vivo. Moreover, IGFBP-1 overexpression may also reduce carbohydrate resources necessary for growth and survival.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1-antitrypsin promoter fused to hIGFBP-1 cDNA have been previously described (12). Northern blotting revealed transgene expression exclusively in the liver during fetal life, through to adulthood. The mice were bred and kept under standardized laboratory conditions. Heterozygous (HT) females of 3–6 months of age were crossed with HT males, giving an average of 5.9 pups per litter. Each litter contained homozygous (HM) and HT pups overexpressing hIGFBP-1 as well as nontransgenic (NT) mice, which we used as internal controls. For genotyping, growth, and skeleton analysis, we analyzed 366 pups (242 fetuses and 124 newborns were collected for our studies) from 62 HT females. For maternal influence studies, we used 144 pups (96 fetuses and 48 newborns) from 24 HT females. We also crossed four NT females with HM males giving an average of 8.7 pups per litter and four HM females with NT males giving an average of 3.5 pups per litter. Embryonic mice were staged by designating midday on the day on which a copulatory plug was first apparent as E0.5. Animals were killed by cervical dislocation after chloral anesthesia. All animals were weighed to the nearest milligram. We determined the growth of 16-d-old embryos (E16.5) to 1-d-old newborns [postnatal day (P)1], as reflected by the body weight increase, for transgenic HM mice and compared this to NT mice obtained from same HT crossings. When growth rates were compared between NT and HM embryos, plots of whole-embryo weight vs. developmental time provided an adequate, albeit coarse, overall index of the growth processes. Because the development between different litters and between embryos from the same litter is well known to be asynchronous, we examined a large number of litters to address this problem.
These protocols received the approval of the French Ministry of Research and Technology (no. 3299B).
Biologic samples
Tails were immediately frozen in liquid nitrogen and stored at –80 C until used for DNA extraction. Genotyping by PCR and Southern blotting analysis were carried out as previously described (12). Skeletons were fixed in 4% paraformaldehyde for staining. The calvaria and septum were frozen in liquid nitrogen for RNA extraction. Dissected livers were immediately frozen in liquid nitrogen for glycogen dosages. Blood glycemia were measured on one drop of fresh blood from newborn animals using reactive strips (Glucotides; Bayer Diagnostics, Puteaux, France) and a glycemia reader (Glucometer 4; Bayer Diagnostics) according to the manufacturer’s protocol.
Alizarin red/alcian blue staining of whole embryo skeleton
Mineralization stages were visualized by staining entire skeletons with alcian blue and alizarin red S., according to the method of McLeod (15) modified as described previously (16). Alcian blue allowed the visualization of cartilage and alizarin red S. was used for the visualization of mineralized bone.
Histology
For histological examination, dissected embryos were fixed in 4% paraformaldehyde in 1x PBS, pH 7.4, at room temperature overnight, dehydrated in graded ethanol series and embedded in paraffin for sectioning. The sections were stained with alizarin red S. as follows: the sections were cleaned of paraffin and rehydrated, and the slides were stained in 1% aqueous alizarin red S. The sections were then passed through acetone, acetone/toluene (vol/vol), and toluene to remove excess of stain and to dehydrate them before mounting.
Hepatic glycogen dosages
Glycogen was prepared and purified according to Van Handel (17) and measured with anthrone using a colorimetric assay (18). After dissection, livers were weighed, cut, and kept frozen until needed. Between 20–40 mg of the livers were homogenized in 40% KOH and boiled for 15 min. Total glycogen was recovered after three series of ethanol precipitations and was resuspended in 5% H2SO4. One volume of the glycogen extract was incubated with 10 vol of 0.15% anthrone for 20 min at 90 C to produce a colorimetric reaction that allowed the glycogen to be quantified by reading the OD at 620 nm. The standard curve was obtained with rabbit glycogen (Sigma, St. Quentin Fallavier, France).
Quantitative PCR
RNAs were extracted from the calvaria and septum using TriReagent (Sigma) according to the manufacturer’s instructions. One microgram of total RNA was reverse transcribed using 0.25 U of Superscript II (Invitrogen, Carlsbad, CA), according to the manufacturer’s instructions, and 1/500 of RT reaction was subjected to quantitative PCR using the Light Cycler faststart DNA Master SYBR Green I kit (Roche, Mannheim, Germany). PCR consisted of 40 cycles of 10 sec at 94 C, 10 sec at 60 C, and 20 sec at 72 C. The primers for IGFBP-1 were 5-AAATGGAAGGAGCCCTGCC and 5-CCATTCTTGTTGCAGTTTGGC, for IGF-I were 5-TCACATCTCTTCTACCTGGC and 5-GTCCACACACGAACTGAAGA, for IGF-II were 5-GCGGCTTCTACTTCAGCAGG and 5-GAAGAACTTGCCCACGGGGT.
Plasma IGF-I and hIGFBP-1 concentrations
Plasma IGF-I concentrations were measured using Mouse/Rat IGF-I RIA kit (DSL2009; Diagnostic System Laboratories, Webster, TX) according to the manufacturer’s instructions using 10 µl serum for the IGF extraction. The sensitivity threshold was 21 ng/ml.
Human IGFBP-1 was measured in plasma samples using an immunoradiometric assay (IRMA) specific for human IGFBP-1, with no cross-reaction with murine IGFBP-1 (DSL 7200 Active IGFBP-1 IRMA Kit; Diagnostic Systems Laboratories). The sensitivity threshold was 0.25 ng/ml plasma. No immunoreactive IGFBP-1 was detected in wild-type mice.
Data analysis
Results were expressed as mean ± SE. The Mann-Whitney U test for nonparametric data were used to determine differences between NT, HT, and HM mice. Tests were considered significant at P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). After birth, this divergence from the expected percentage was even greater with the ratio being 1.26 NT/2.13 HT/0.61 HM (n = 124). This increased loss of HM transgenic pups after birth was due to an early death of the HM mutant pups (12).
|
). At this stage, the mean weight of HM fetuses was 904 ± 108 vs. 1004 ± 143 mg for NT fetuses. This growth delay persisted at E18.5 and became even more evident after birth, with the mean weight of HM pups being 1322 ± 195 vs. 1645 ± 214 mg NT pups. At birth, the growth of HM mice was retarded by 20% vs. NT mice. In the first month after birth, the HM mice could be 30% smaller than the NT mice (Ref. 12 and data not shown).
|
and 3). In the skull and in the axial and appendicular skeleton of NT mice, ossification occurs during the last third period of the gestation starting at E14.5. At E16.5, alizarin red staining revealed a delay in mineralization in the skull of HM (Fig. 2, compare B and D with A and C). At E18.5, the parietal and interparietal bones of NT mice were mineralized, whereas the gap between the parietal and interparietal bones and the lambdoid suture of HM mice were widely opened (Fig. 2, compare F and H with E and G). We extended our study to adulthood and we observed that, at P21, the overall skull morphology was different between HM mice and NT mice, particularly in the posterior compartment (data not shown). At 9 months, the skulls of HM mice were smaller and flattened in the posterior area than those of NT mice (Fig. 2, compare J and L with I and K).
|
|
). The femur and tibia, as well as vertebrae and sternum showed a lack of mineralization at E16.5 (Fig. 3A). At E18.5, alizarin red staining confirmed a clear delayed mineralization in sternebrae and ventral ribs (compare Fig. 3, C with D). HM mice had less alizarin-positive mineralized caudal vertebrae and later mineralization of the lower limbs than NT mice. At P21, this delay in mineralization was still present, with the kneecap still being less mineralized (Fig. 3, compare E with F). Among the same littermates, HM mice showed mild but defined skeletal changes at different stages of antenatal and postnatal development when compared with NT mice. We observed a delay in mineralization of about 2 d (which appeared to be compensated after birth). This delay was generalized in the cranio-facial as well as axial and appendicular skeleton.
Tissue (bone) formation in transgenic mice
We visualized these mineralization defects in frontal sections of mouse fetuses (Fig. 4). Alizarin red staining appeared more intense in NT mice than in HM mice. At E16.5 and at E18.5, there appeared to be more skull mineralization in NT mouse samples. Several skeletal sites showed a delay in mineralization, such as nasal, zygomatic, maxillary, and orbito-sphenoid bones. Other bones did not show such a delay in mineralization, but were thinner than the bones of NT mice. All bones showed less biomineralization as revealed by a decrease in the amount of apparent calcium deposits. Thus, as visualized by whole-mount skeleton staining, histological studies also showed a consistent mineralization defect.
|
Plasma IGF-I concentrations remained low and were similar in NT, HT, and HM animals during the fetal period and on the day of birth (Table 2). In 3-month-old animals, IGF-I levels in HM mice were 2.8-fold lower than in NT mice (Table 2 and Refs. 12 and 19).
|
). We also observed this decreased level of glycogen at E18.5, being 69 ± 8 µg/mg liver for HM mice vs. 89 ± 17 µg/mg liver for NT mice. This lower energetic reservoir was also found in newborn animals: glycemia was only 1.89 ± 0.91 mM in HM mice but 3.03 ± 1.14 mM in NT mice (Fig. 5B). In each experiment, HT mice showed an intermediate phenotype (Fig. 5).
|
).
|
).
The analysis of maternal plasma hIGFBP-1 concentrations showed a 2.9-fold increase in transgene expression during pregnancy (Fig. 7). The plasma hIGFBP-1 level in nonpregnant mice was 4.5 ± 0.5 ng/ml, and reached 12.9 ± 1.5 ng/ml in pregnant mice (P < 0.001).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1-antitrypsin proximal promoter (12). Our mice showed growth retardation, decreased IGF-I levels, reproduction defects, and brain and kidney abnormalities. Crossey et al. (22) established the fourth transgenic mouse line in which IGFBP-1 expression was controlled by its own promoter increasing its expression in the liver and in the endometrium. These mice showed reduced fetal growth because of a placental insufficiency essentially due to maternal IGFBP-1. The fifth and final model was developed by Watson et al. (23), with IGFBP-1 being under the control of the -fetoprotein promoter. These mice showed a decreased growth that started during the fetal period, but no decrease in the level of circulating IGF-I. In summary, most models have shown a reduced body growth beginning during the fetal period if the transgene is expressed. Each group focused their studies on different developmental aspects and suggested different hypotheses concerning the role of IGFBP-1. In the present study, we investigated the effect of IGFBP-1 on bone mineralization (reflecting growth factor activity of the IGF system) and on available carbohydrate resources (reflecting the metabolic activity of IGF system), both of these parameters being directly related to body growth.
In adult transgenic mice, plasma hIGFBP-1 concentrations, which directly reflect transgene expression, were moderately higher (>4-fold) than in NT mice. Maternal plasma human IGFBP-1 levels have been shown to increase 2.9 times during gestation (as seen for endogenous mouse IGFBP-1), and probably followed 1-antitrypsin expression, which has been shown to increase during pregnancy in humans (24).
Our results showed that HM mice were clearly under represented with respect to the Mendelian ratio. One of the main reasons suspected for HM death around birth was glycemia status. At birth, the newborns have to make several adjustments to adapt to extrauterine life, including maintaining normal glycemia levels (25). Before birth, fetal glucose levels are maintained by transplacental passage of glucose from the mother. However, there is a critical period between birth and the establishment of suckling when the newborn depends on its hepatic glycogen pool to maintain blood glucose. Thus, the presence of appropriate hepatic glucose stores at birth enhances survival during this critical transitional period. Our results clearly showed that glycemia levels were lower in HM neonates than in NT neonates. Moreover, the hepatic glycogen pool also decreased twice in HM fetuses, at E17.5 and E18.5, suggesting that these mice could not survive during the first hours/days after birth. The low levels of carbohydrate resources found in fetuses and/or neonates were even lower in pups with HM mothers. This may explain the high number of neonate deaths, ranging from 3% for HT neonates with NT mothers to 47% for HT neonates with HM mothers (12). Thus, the glycogen pool, which is essential for newborn survival, may be linked to circulating fetal and maternal IGFBP-1 levels. Nevertheless, the mechanism of action was not elucidated and liver developmental defects or placental insufficiency couldn’t be excluded.
We observed that HM mice showed growth retardation during fetal life, which was significant at E17.5, with HM mice being 10% smaller than NT mice. During fetal life and until birth, we detected no difference in plasma IGF-I levels between HM and NT animals, suggesting that IGFBP-1 has an effect during this period probably acting by IGF sequestration. After birth, this growth retardation was even more evident, reaching 20% on the day of birth, and was never compensated throughout postnatal life (30% during adulthood). These results are consistent with those recently reported by Watson et al. (23). A recent study carried out on zebrafish demonstrated the key role of IGFBP-1 in embryonic growth retardation (26). In adult HM mice, IGF-I levels were half as high as those in NT littermates, resulting in decreased GH production, which arises from IGFBP-1 action on the hypothalamo-pituitary axis (12, 19). We have shown previously that GH treatments could catch up growth retardation (19). In children with persistent postnatal growth retardation, diminished spontaneous secretion of GH is the rationale for GH treatment (27). Our findings are consistent with those from transgenic mice carrying a null mutation of the IGF-I gene in which growth retardation occurs during the fetal period and persists after birth (9, 10).
Each transgenic model that overexpresses either IGFBP-2, IGFBP-3, or IGFBP-5 under the ubiquitous cytomegalovirus promoter has shown a reduced body size due to a reduction in carcass weights (28, 29, 30). Similar inhibitory effects on growth are observed when IGFBP-6 is overexpressed in the circulation during the growth period (31). All these data were obtained in vivo and consistently show that the IGF system, particularly IGFBPs, play an important role in the growth and development of the mouse embryo. Nevertheless, IGFBP-1 is the only one that shows a strong physiological increase in the circulation during fetal life and pregnancy, suggesting it is specifically involved in fetal growth.
Another aspect of this study was the detailed analysis of bone formation and mineralization caused by the growth retardation phenotype. Whole skeletal preparations showed specific evidence of delayed mineralization of long bones as well as of skull bones (16). Histological examination of transgenic mice revealed a poor mineralization of skull bone, with this mineralization defect being observed during antenatal life and persisting at 9 months of age (site-specific abnormalities observed mainly at the posterior level of the skull resulted in an irreversible cranial dysmorphy). These findings concerned bone formation essentially during the fetal and early postnatal life, with the overall delay being in both endochondral (long bones) and membranous (skull and facial skeleton) ossifications. It has been shown that IGFs increased collagen expression in vitro and mineral apposition rates in transgenic mice overexpressing IGF-I (8, 32). Conversely, IGF-I null mice show reduced bone growth despite normal chondrocyte proliferation (33). These studies show that IGFs have a modest mitogenic activity in bone development and a more prevalent impact on the differentiated function of osteoblasts. The decreased mineral apposition rates we have shown probably resulted from sequestration of IGF-I by circulating IGFBP-1 (because hIGFBP-1 was not synthesized in bones). However, we cannot exclude IGF-independent effects of IGFBP-1 (34, 35). In IGF-I receptor null mice, ossification centers of the cranial and facial bones appear later, with a prenatal delay of 2 d. In IGF-I null mice and in mice in which the IGF-II paternal allele was deleted, the appearance of the ossification centers occurs no more than 1 d late (36, 37). The overexpression of IGFBP-2 in bones with the cytomegalovirus promoter causes decreased bone growth and mineral content (38). The overexpression of IGFBP-4 under osteocalcin promoter causes marked postnatal growth retardation associated with a reduced bone turnover (39). The overexpression of IGFBP-5 with an osteocalcin promoter also causes a significant transient decrease in bone mineral density, and trabecular bone volume secondary to reduced trabecular number and thickness (40). Our mouse model showed delayed bone ossification similar to IGF-I receptor null mice, supporting the role of the IGF system in antenatal and postnatal bone mineralization. Our transgenic model can be shown to be valid by the demonstration of an increase in bone formation by the overexpression of IGF-I under the control of the osteocalcin promoter, and confirms the known anabolic actions of IGF-I in vitro (32).
The unequivocal evidence linking the IGF system with fetal growth and development raises the intriguing possibility of understanding the IUGR. Many studies reported increased plasma levels of IGFBP-1 associated with reduced IGF-I levels in IUGR. This was initially reported by Unterman et al. (41) in rats and by Lassarre et al. (42) and by Chard (43) in humans. More recent studies have shown that IGFBP-1 expression in the cord blood during the third trimester was negatively correlated with birth weight (44, 45). The earlier this increase in IGFBP-1 expression occurs during the pregnancy, both in maternal serum and amniotic fluid, the more the birth weight is affected (43). In IUGR, high maternal IGFBP-1 concentrations may further reduce the availability of maternal IGF-I for the placenta by a still unknown mechanism. This could worsen placental function and thus adversely affect fetal growth as it has been shown using two different IGFBP-1-overexpressing transgenic mouse models (23, 46).
Small-for-gestational-age (SGA) infants show reduced bone formation and reduced bone mineral content (47, 48). In SGA infants, skeletal maturation is also delayed (49). Untreated children may have a low-normal growth velocity, a poor weight gain, and a slower maturation of their bones for their chronological age. SGA infants also have elevated blood pressure, a high prevalence of noninsulin-dependent diabetes mellitus, insulin resistance, impaired glucose tolerance, disturbances in cholesterol metabolism and blood coagulation, and a highly predictive risk of coronary heart disease (50). Thus, our transgenic mice, which showed a delay in skeletal maturation in utero, early postnatal growth retardation, perinatal mortality, and impaired glucose tolerance (data not shown), presented anomalies similar to those observed in severe cases of human SGA and IUGR.
In conclusion, our study has provided the first evidence that the overexpression of IGFBP-1 in transgenic mice is responsible for the intrauterine and early postnatal growth retardation, as well as delayed bone ossification and a reduced carbohydrate pool. We have tried to discriminate the role of maternal IGFBP-1, which is important for fetal growth because it contributes to reduce carbohydrate resources and the role of fetal IGFBP-1, which is also important for growth because of its action on bone mineralization. Our results suggest that an excess of IGFBP-1 may be one of the principal factors contributing to growth retardation and bone development in severe cases of human IUGR. Thus, the effect of GH treatment on mineralization and the relationship between maternal IGFBP-1 and IUGR risk should be further studied.
Elevated maternal IGFBP-1 levels which are one of the most frequently described parameters in IUGR may be very important in this pathology (43).
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure statement: N.B.L., D.S., Y.L.B., M.B., A.B., P.M., and S.B. have nothing to declare.
First Published Online June 29, 2006
1 N.B.L. and D.S. have equally contributed to this work. 
Abbreviations: E, Embryonic day; hIGFBP-1, human IGF binding protein-1; HM, homozygous; HT, heterozygous; IGFBP-1, IGF binding protein-1; IRMA, immunoradiometric assay; IUGR, intrauterine growth retardation; NT, nontransgenic; P, postnatal day; SGA, small for gestational age.
Received February 9, 2006.
Accepted for publication June 15, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-1-Antitrypsin and -2-macroglobulin concentration in serum during pregnancy. Acta Obstet Gynecol Scand 46:126–137[Medline]5ß1 integrin via mitogen-activated protein kinase pathway. J Clin Endocrinol Metab 86:2484–2493This article has been cited by other articles:
 |
|
 |
|
  M. P. Sperandeo, P. Annunziata, A. Bozzato, P. Piccolo, L. Maiuri, M. D'Armiento, A. Ballabio, G. Corso, G. Andria, G. Borsani, et al. Slc7a7 disruption causes fetal growth retardation by downregulating Igf1 in the mouse model of lysinuric protein intolerance Am J Physiol Cell Physiol, July 1, 2007; 293(1): C191 - C198. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
) and NT (+) progeny from the same litters resulting from HT crossings. Measurements were made from E16.5 to 1 d postnatal (P1). Each point is the mean of n measurements (all represented) for a given age.
), and HM (
), HT (
) females were weighed the day after birth (P1). Newborn animals from HM mothers were significantly smaller than animals born from HT mothers, which were smaller than those born from NT mothers. B, Glycemia levels were measured in animals the day after birth (P1) using reactive strips and a glycemia reader. Glycemia levels of newborn animals from HM mothers were significantly lower than those of animals born from HT mothers, which were lower than those born from NT mothers.