当前位置: 首页 > 医学版 > 期刊论文 > 内科学 > 内分泌学杂志 > 2005年 > 第2期 > 正文
编号:11168175
Insulin-Like Growth Factor-Binding Protein-5 Induces a Gender-Related Decrease in Bone Mineral Density in Transgenic Mice
     Laboratory of Molecular Signaling, The Babraham Institute (D.A.M.S., G.T., F.A.L., N.F.A., E.J.C., J.M.P.), Cambridge, United Kingdom CB2 4AT; and Musculoskeletal Disease Center, Jerry L. Pettis Memorial Veterans Affairs Medical Center (S.M., Y.K., J.E.W., D.J.B.), and Departments of Medicine and Biochemistry, Loma Linda University (S.M., D.J.B.), Loma Linda, California 92357

    Address all correspondence and requests for reprints to: Dr. Jennifer M. Pell, Laboratory of Molecular Signaling, The Babraham Institute, Cambridge, United Kingdom CB2 4AT. E-mail: jenny.pell@bbsrc.ac.uk. Or to: Dr. Subburaman Mohan, Musculoskeletal Disease Center (151), Jerry L. Pettis Veterans Affairs Medical Center, 11201 Benton Street, Loma Linda, California 92357. E-mail: subburaman.mohan@med.va.gov.

    Abstract

    IGF-binding protein-5 (IGFBP-5) is abundant in serum and bone during normal skeletal development, but levels decrease in osteoporosis. Studies have shown that IGFBP-5 stimulates markers of bone formation by potentiating IGF actions and by IGF-independent actions. To test the hypothesis that IGFBP-5 promotes the acquisition of bone mineral density (BMD), we generated transgenic (Tg) mice overexpressing Igfbp5 using a cytomegalovirus enhancer and ?-actin promoter (CMV/?A). Tg animals showed an increase in serum IGFBP-5 concentrations by 7.7- to 3.5-fold at 3–8 wk of age, respectively. Concentrations were 6–49% higher for males compared with females in both wild-type and Tg mice. Surprisingly, BMD decreased in a gender-dependent manner, with Tg male adults affected more severely than Tg females (31.3% vs. 19.2% reduction, respectively, compared with wild-type mice, assessed by dual energy x-ray absorptiometry). Significant gender differences in BMD were confirmed by peripheral quantitative computed tomography. Histomorphometry revealed that although the bone formation rate and mineralizing surface at the periosteum decreased in Tg mice, they increased at the endosteum, suggesting opposing effects of IGFBP-5 on periosteal and endosteal osteoblasts (by altering proliferation or survival). These findings differ from previous observations in Igf1- and Igf2-null animals. In conclusion, IGFBP-5 has a significant influence on BMD acquisition and maintenance that is dependent on gender and age. The phenotype of Igfbp5 mice cannot be explained solely by IGF inhibition; thus, this study provides the first in vivo evidence, by genetic manipulation, for IGF-independent actions of IGFBP-5 in bone function. These findings have implications for the gender-biased progression of osteoporosis.

    Introduction

    BONE MINERAL DENSITY (BMD) attained during puberty is an important indicator of the risk of developing both postmenopausal (type I) and senile (type II) osteoporosis (1, 2, 3) and is synchronized by peptide growth factors, cytokines, and steroid hormones (3, 4). In particular, the IGFs regulate bone volumetric expansion, mineralization and longitudinal growth during development, and maintain the skeleton of older animals (5, 6). The IGFs are unique among growth factors in that they have not only proliferative and antiapoptotic effects, but also enhance the differentiated function of skeletal cells (6, 7, 8, 9). They also regulate the synthesis and turnover of bone extracellular matrix (ECM) proteins, such as collagen (10). Although both IGF-I and IGF-II are important for prepubertal skeletal growth and density, IGF-I also has a key role in mediating the effects of GH during and after puberty (6, 11, 12). GH has additional direct and IGF-independent actions on skeletal cells (6).

    Mice null for either Igf1 or Igf2 have severely delayed skeletal development (13, 14) with attenuated hypertrophy of terminally differentiated chondrocytes (6, 8). Igf1-null mice (11, 12) or mice null for the IGF type 1 receptor in osteoblasts (9) exhibit a strikingly decreased mineralization rate of osteoid matrix. In complementary experiments, targeted overexpression of Igf1 in osteoblasts increased the bone formation rate (BFR) and bone volume without altering osteoblast number (15).

    The half-lives of the IGFs and their ability to activate the type 1 receptor are tightly controlled by six high affinity binding proteins (IGFBP-1 to -6). IGFBP-5 stimulates osteoblast and osteoclast proliferation and/or function in vitro and in vivo (16, 17, 18, 19, 20) and augments IGF action (21, 22). Microarray analysis revealed that Igfbp5 mRNA is up-regulated 30-fold during the latter stages of chondrogenesis (23). Indeed, IGFBP-5 has several features to support a key role in bone. It is the most abundant bone IGFBP (24) and has a high affinity for hydroxyapatite and ECM proteins, providing a mechanism to store IGFs (22, 25). ECM-associated IGFBP-5 has a decreased affinity for IGF-I, thereby stimulating IGF actions by delivering the IGFs to the type 1 receptor (22, 26, 27). The local release of sequestered IGFs and IGFBP-5 in bone could provide a means to couple osteoclastic bone resorption and osteoblastic bone formation during remodeling (28). This hypothesis is strengthened by the correlated decline in IGF-I and IGFBP-5 concentrations in the circulation and skeleton during aging (24, 29, 30, 31), with age-related impairment in the coupling of bone formation to resorption. In addition, circulating levels of free and total IGF-I and IGFBP-5 are reduced in osteoporosis patients (32, 33, 34).

    Although it is still unclear whether the anabolic effects of IGFBP-5 in bone are solely due to stimulation of IGF actions, direct effects of IGFBP-5 on osteoblastic function have been observed in Igf1-null mice (19). Systemic administration of IGFBP-5 increases bone formation parameters in mice without altering serum IGF-I levels (18). Additional in vitro studies support the idea that mitogenic and antiapoptotic effects of IGFBP-5 may in part be independent of IGFs (16, 17, 19, 35); IGFBP-5-binding sites on osteoblast cell surfaces have been reported and may be phosphorylated in response to IGFBP-5 (36, 37). IGFBP-5 also possesses a nuclear localization signal and can form transcriptional complexes in osteoblasts (38).

    Taken together, the current evidence suggests that IGFBP-5 may have a significant influence on BMD acquisition and maintenance in vivo. The aim of this investigation was therefore to examine the effects of overexpression of Igfbp5 in vivo (39) by measurement of bone parameters using dual energy x-ray absorptiometry (DEXA), peripheral quantitative computed tomography (pQCT), osteocalcin levels, and dynamic histomorphometry. We reveal significant gender- and age-related effects on skeletal development that are distinct from the Igf1- and Igf2-null phenotypes.

    Materials and Methods

    Igfbp5-overexpressing mice

    All animal experiments were conducted in accordance with the animal welfare, experimentation, and ethics committee at The Babraham Institute, under United Kingdom Home Office regulations. The generation and characterization of Tg mice overexpressing Igfbp5 from the CMV/?A regulatory element have been described previously (39, 40). Two lines of CMV/?A-Igfbp5 mice overexpressing Tg Igfbp5 at high or intermediate levels were used to produce the hemizygous (–/+) offspring described in this study and were created using Igfbp5 hemizygote fathers mated to wild-type (WT) females (C57BL/6JxCBA/CA). WT mothers were used because Igfbp5 hemizygous females demonstrate significantly reduced litter sizes, and the resultant pups exhibit intrauterine growth retardation regardless of genotype. Igfbp5 Tg animals were identified by PCR, which gave a control endogenous Igfbp5 band of 881 bp with an additional band for transgene-derived Igfbp5 of 309 bp (Igfbp5 exon 2 forward primer, 5'-AGA TGG CTG AAG AGA CCT ACT CC-3'; Igfbp5 exon 3 reverse primer, 5'-GCT TTC TCT TGT AGA ATC CTT TG-3').

    Bone and serum collection

    Hemizygous and WT littermates of both genders were analyzed at 3 wk (before puberty), 5 wk (completion of puberty), and 8 wk (adulthood) by DEXA or pQCT. Histomorphometry was performed on samples from mice at 5 wk of age when bone displays an active period of modeling; mice were injected ip with tetracycline hydrochloride (20 μg/g body weight; Sigma-Aldrich Corp., Dorset, UK) 8 d before harvesting and then with demeclocycline (20 μg/g body weight; Sigma-Aldrich Corp.) 2 d before harvesting. Mice were killed by CO2 inhalation. Blood was collected quickly by heart puncture, and serum was stored at –20 C. Bilateral femora and tibiae were dissected, cleaned of soft tissue, and stored at –70 C for DEXA or pQCT analyses. Histomorphometric analysis was performed on fresh femora fixed on ice with 4% paraformaldehyde for 4–6 h and stored in 70% ethanol. The longitudinal length of femora was measured with digital calipers (Stoelting, Wood Dale, IL).

    IGFBP-5 RIA

    Serum IGFBP-5 levels were measured using an RIA for mouse IGFBP-5 (41). Antibodies against recombinant human IGFBP-5 were raised in guinea pigs as described previously (29). IGFBP-5 antiserum from a guinea pig that demonstrated significant cross-reactivity toward mouse IGFBP-5 was selected for RIA. None of the other mouse IGFBPs displayed significant cross-reactivity. Mouse serum samples were diluted 1:10 before assay. The sensitivity of the assay was 10 ng/ml. The inter- and intraassay coefficients of variation were less than 10%.

    Bone densitometry by DEXA

    Bone mineral content (BMC), bone area, and areal BMD (aBMD) of the femora and tibiae were measured by DEXA using a PIXImus instrument (Lunar Corp., Madison, WI). The precision for BMC, area, and aBMD was ±1% for repeat measurements of the same bones.

    Volumetric BMD (vBMD) and geometric parameters determined by pQCT

    vBMD and geometric parameters at the proximal metaphysis, mid-diaphysis, and distal metaphysis of femora were determined by pQCT (Stratec XCT 960M, Norland Medical Systems, Fort Atkinson, WI). Previously, pQCT has been validated as a reliable method for determining vBMD and geometric parameters in mice (11, 12). Routine calibration was performed daily with a defined standard (cone phantom) containing hydroxyapatite embedded in Lucite (Norland Medical Systems, Inc.). Analysis of the scans was performed using the manufacturer-supplied software program (Stratec Medizintechnic, Bone Density software, version 5.40 C, Norland Medical Systems, Inc.). vBMD and the circumferences of the periosteum and endosteum were estimated by Loop analysis. The thresholds were set at 230–630 mg/cm3. The voxel size was set at 0.07 mm, and a 0.5-mm thick slice was scanned through the entire length of the bone. The reference line as a center of scanning was set at the midpoint of the femur; thereafter, nine slices were scanned symmetrically from the reference line. The data for three slices from proximal, middle, or distal femur were averaged and expressed as proximal metaphysis, mid-diaphysis, or distal metaphysis. The coefficients of variation for vBMD, periosteal circumference, and endosteal circumference for repeat measurements of four mouse femora (two to five measurements) were less than 3%, 1%, and 2%, respectively.

    Osteocalcin RIA

    Serum osteocalcin levels were measured using an RIA that has been validated for measuring mouse osteocalcin (18). The sensitivity of the assay was 0.5 ng/ml. The inter- and intraassay coefficients of variation were less than 8%.

    Dynamic histomorphometry

    The femora were defleshed and embedded in methylmethacrylate. Thick cross-sections (0.5-mm thickness) were cut from the midpoint of the shaft with a diamond wire Histo-saw (Delaware Diamond Knives, Wilmington, DE). The cross-sections were ground lightly, mounted in aqueous Fluoromount-G (Fisher Scientific, Pittsburgh, PA), and examined under an Olympus BH-2 fluorescent/brightfield microscope to measure bone areas and tetracycline labels.

    All bone histomorphometric parameters were measured with the OsteoMeasure system equipped with a digitizing tablet (OsteoMetrics, Inc., Atlanta, GA) and a color video camera (Sony Corp., Tokyo, Japan). Bone area measurements were made at x60, whereas measurements of tetracycline were made at a total magnification of x600. The total cross-sectional bone area (square millimeters), medullary area (square millimeters), periosteal circumference (millimeters), and endosteal circumference (millimeters) were measured. Cortical bone area was calculated by subtracting the medullary area from the total cross-sectional bone area. The mineralizing surface (MS; millimeters) was calculated as the sum of the length of tetracycline double labels (dL) plus half of the length of single labels (sL) along the entire endosteal or periosteal bone surfaces, so MS = dL + 0.5sL. The mineral apposition rate (MAR; microns per day) at the periosteal and endosteal surfaces was determined by dividing the mean of the width of the double fluorescent labels by the interval (6 d). Mean label width was determined indirectly by dividing the label area by the mean of the inner and outer double-label lengths. The BFRs at the periosteal and endosteal surfaces were calculated by multiplying MAR by MS (expressed in x10–3 mm2 per day). The methodological coefficients of variation for each parameter, which were determined by the ANOVA of three repeat measurements of five different bone specimens, were as follows: area, 0.78–1.86%; circumference, 0.57–1.36%; label length, 6.44%; and label width, 4.98%. Histomorphometric indexes were based on nomenclature recommended by the American Society of Bone and Mineral Research (42).

    Statistical analyses

    Mice were grouped according to age, line, sex, and genotype. Differences between genotypes, lines, ages, or genders at selected times were determined by t test (confirmed using Mann-Whitney test). ANOVA was used to test for differences between groups using genotype, gender, age, or litter as factors. Values are expressed as the mean ± SEM in all figures.

    Results

    Serum IGFBP-5 concentrations are regulated by gender and development in mice

    Three lines of Tg mice were generated that overexpressed Igfbp5 under the control of the CMV/?A regulatory sequence (39). Two lines expressing Igfbp5 at high and intermediate levels were used in this study. The transgene was expressed from early development and in all adult tissues studied to date. Previous studies have established the CMV/?A sequence drives expression in primary ossification centers, specifically in chondrocytes, osteoblasts, lining cells, and cells within the marrow (9, 43).

    Circulating IGFBP-5 concentrations increased steadily during development from 3–8 wk of age (P < 0.001) and were consistently higher in male than female mice (Fig. 1). The gender regulation of IGFBP-5 was evident in WT mice and was maintained for Tg siblings. The developmental induction observed for WT mice was accelerated for Tg siblings. WT males had IGFBP-5 levels of approximately 100 ng/ml before puberty, increasing to 239 ± 35 ng/ml in adults. In comparison, WT females displayed lower serum levels at all time points (P < 0.05, by two-way ANOVA with gender and age as the main effects; n = 5–9). Older mice showed a continued rise in IGFBP-5 concentrations (12 months; males, 339 ± 27 ng/ml; females, 305 ± 39 ng/ml; P < 0.05 vs. 8 wk; n = 5–6). Tg mice also exhibited a general increase in serum IGFBP-5 during development, although the increase appeared to plateau sooner than for WT mice. From 3–8 wk of age, male Tg mice displayed a 5.8- to 3.9-fold elevation, respectively, of IGFBP-5 over WT males, and female Tg mice displayed a 7.7- to 3.5-fold increase, respectively, over their female WT siblings. A second Tg line expressing Igfbp5 at an intermediate level demonstrated a 2.3-fold increase (P < 0.05; n = 5–10) in serum IGFBP-5 in adult males. Thus, higher serum IGFBP-5 levels were observed in both WT and Tg males over females, and the IGFBP-5 induction during WT postnatal development was accelerated in Igfbp5 lines.

    FIG. 1. Serum IGFBP-5 concentrations increase during development and are gender related. Hemizygous Tg mice overexpressing Igfbp5 and WT littermates from the highest expressing line are represented. Bars labeled with different letters are significantly different as follows: a and b, c, and d, P < 0.001 (for Tg vs. WT mice of the same gender); a and c, P < 0.05; b and d, P < 0.001 (for females vs. males). Data were analyzed by three-way ANOVA, with genotype, gender, and age as the main effects. Values shown are the mean ± SEM of five to nine mice per group. Prepubertal WT females had serum IGFBP-5 concentrations just below the sensitivity of the RIA (100 ng/ml).

    Disproportionate linear growth in Igfbp5 Tg mice

    Previously, we demonstrated that overexpression of Igfbp5 in mice resulted in retardation of postnatal development in a nonproportional fashion (39). Tg mice from the high expressing line showed maximal growth deceleration of whole body weight before 3 wk of age (body weight was 65–70% of WT mice); by 8 wk body weight gain stabilized (60–65% of WT mice; Fig. 2A). Tg males and females were growth retarded by similar proportions.

    FIG. 2. Disproportionate growth inhibition in Tg mice overexpressing Igfbp5 at 8 wk of age. A, Whole body weight gain. B, Linear growth of the axial skeleton represented by body length (nose to anus). C, Tail length. D, Linear growth of the appendicular skeleton represented by femur length. Hemizygous Tg mice () and WT siblings () of both genders from the highest expressing line are shown. The Tg to WT proportion (percentage) within genders is given above pairs of bars. ***, P < 0.001, Tg vs. WT siblings of same gender (by t test). Values shown are the mean ± SEM of six to 13 mice per group.

    Body weight does not necessarily reflect longitudinal skeletal growth; therefore, the lengths of the body, tail, and femur were determined. Body length of Igfbp5 mice (nose to anus) was reduced by 14.5% compared with that of WT siblings at 8 wk (Fig. 2B). In contrast, tail length was reduced further by 27.8% (Fig. 2C). The femoral length of Tg mice was significantly decreased by 10.1% (Fig. 2D). The differential reduction in linear growth of the body, tail, and femur (P < 0.01) suggests that the influence of IGFBP-5 is dependent on skeletal region. Reductions in body, tail, and femur length were less severe than the decrease in body weight, confirming that different target cells (e.g. myoblasts vs. chondrocytes) respond differently to IGFBP-5.

    Sexually dimorphic BMD, BMC, and volumetric expansion in Igfbp5 Tg mice

    The effects of Igfbp5 overexpression on BMD, mineral content, and bone geometry were evaluated by subjecting tibiae and femora to DEXA. A substantial decrease in aBMD for Igfbp5 mice was revealed together with a significantly different response between male and female Tg mice (Fig. 3A). Although aBMD for Tg males was strikingly reduced (31.3%), the reduction for Tg females was more modest (19.2%) in 8-wk-old tibiae. The decrease in aBMD in Tg mice was due to markedly less BMC (Fig. 3B). In contrast, the reduction in tibial area, although significant, was less than the reduction in BMC (accounting for the decreased aBMD) and was not sexually dimorphic (Fig. 3C). Femora of Tg adults responded similarly to the Tg tibiae, supporting the sexual dimorphism observed for aBMD and BMC (data not shown). Therefore, decreased aBMD in Igfbp5 mice was due to a striking and sexually dimorphic reduction in BMC despite a significant impairment in bone area.

    FIG. 3. Sexually dimorphic decrease in tibial aBMD in Igfbp5 Tg mice at 8 wk of age. A, Areal BMD. B, BMC. C, Tibial area. Tibial properties were measured by DEXA. Absolute values (left y axis) were plotted for hemizygous Tg mice () and WT siblings () of both genders from the highest expressing line. The Tg to WT proportion (percentage; right y axis) were plotted for males and females ( ). **, P < 0.01; ***, P < 0.001 (Tg vs. WT siblings of same gender, or male vs. female Tg to WT proportion, by t test). Values shown are the mean ± SEM of six mice per group.

    The dependency of skeletal phenotype on the dose of Igfbp5 was investigated by performing DEXA on males from the second Tg line overexpressing Igfbp5 at an intermediate level. As expected, femoral aBMD was reduced by a more moderate level in the intermediate expressors (16.7 ± 2.1% less than in WT mice) compared with the high expressors. BMC and femoral area also displayed the same qualitative trend as aBMD for the intermediate line, although BMC was more strongly affected (BMC, 31.5 ± 2.7% reduction; area, 18.1 ± 2.9%). Therefore, aBMD, BMC, and bone area showed a reproducible response dependent on transgene expression levels between the two Tg lines overexpressing Igfbp5 at high and intermediate levels.

    The gender differences observed for bone density of Igfbp5 mice were also examined by performing DEXA on tibiae during development from 3–8 wk of age. The Tg to WT fraction of female aBMD was essentially constant from 3–8 wk (Fig. 4A). However, male aBMD deteriorated from 3 wk to adulthood. The femoral aBMD of Igfbp5 mice exhibited similar decreases and reinforced the sexual dimorphism (data not shown). Overall, testing Tg male vs. female tibial aBMD relative to WT mice as a function of age by two-way ANOVA showed a highly significant gender difference (P < 0.001).

    FIG. 4. Sexually dimorphic response of tibial aBMD in Igfbp5-overexpressing mice during development. A, Areal BMD. B, BMC. C, Tibial area. Tibial parameters were measured by DEXA. Hemizygous Tg mice values relative to those of WT siblings from the same gender (percentage) were plotted using the highest expressing line. **, P < 0.01; ***, P < 0.001 (male vs. female Tg to WT proportion at 8 wk of age, by t test). Values shown are the mean ± SEM of six to eight mice per group.

    The aBMD reductions observed for adult Tg tibiae (Fig. 3A), and the age-related changes (Fig. 4A) were due primarily to greater decreases in BMC (Fig. 4B) than in tibial area (Fig. 4C). The femora of Tg mice were affected to a similar extent compared with the tibiae and confirmed the sexual dimorphism for BMC (data not shown). Importantly, testing Tg male against female tibial BMC or area relative to WT mice with age as a factor by two-way ANOVA showed a significant gender difference for BMC (P < 0.001), but not for area (P = 0.23).

    Areal BMD is calculated from two dimensional x-ray images and provides an estimate of density; therefore, pQCT was used to measure BMD per unit volume (vBMD) at three femoral sites (distal metaphysis, mid-diaphysis, and proximal metaphysis). Total vBMD was also reduced in a gender-related fashion for Tg mice at 8 wk (only males showed a decline by up to 24% at all three sites; Fig. 5A). Thus, pQCT confirmed that the response of vBMD to IGFBP-5 was gender- and site-dependent (e.g. metaphysis vs. diaphysis).

    FIG. 5. Femoral vBMD and geometry decrease in a gender- and bone site-dependent manner in Tg mice overexpressing Igfbp5 at 8 wk of age. A, Total vBMD. B, Periosteal circumference. C, Endosteal circumference. Femoral properties were measured by pQCT at three sites: proximal metaphysis, mid-diaphysis, and distal metaphysis. Absolute values were plotted for hemizygous Tg mice () and WT siblings () of both genders from the highest expressing line. *, P < 0.05; **, P < 0.01; ***, P < 0.001 (Tg vs. WT siblings of same gender, by t test). Pairs of bars (representing a specific site) labeled with different letters are significantly different as follows and highlight the gender differences in the Tg to WT proportion: a and b, P < 0.05; c and d, P < 0.01; e and f, P < 0.001 (females vs. males, by t test). Values shown are the mean ± SEM of six mice per group.

    The changes observed for BMD were also explored by considering the geometry of the femur by pQCT. The periosteal and endosteal circumferences are the outer and inner surfaces of the bone, respectively, when sectioned transversely (the inner surface being the one lining the marrow cavity). Both periosteal and endosteal circumferences are the net result of total osteoblastic apposition countering osteoclastic resorption. Periosteal circumference was significantly decreased for Tg mice of both genders at all three femoral sites (Fig. 5B). Interestingly, endosteal circumference was reduced to a significantly greater extent for Tg females compared with Tg males at the three femoral sites (Fig. 5C). Thus, the gender-related decreases detected in bone density for Igfbp5 mice could be attributed to contrasting effects of IGFBP-5 on the periosteal vs. endosteal surfaces and the gender context.

    Decreased total osteoblastic function in mice overexpressing Igfbp5

    The mechanism behind the significantly reduced BMD and altered geometry was investigated by measuring serum osteocalcin levels and performing dynamic histomorphometry. Serum osteocalcin decreased in Tg males by 35.2% (Fig. 6), suggesting a decrease in total osteoblastic activity (a result of osteoblast numbers and their differentiation stage). The Tg line overexpressing Igfbp5 at an intermediate level also confirmed a dose-dependent reduction in serum osteocalcin (data not shown).

    FIG. 6. Decreased serum osteocalcin levels in Tg males overexpressing Igfbp5 during development. Hemizygous Tg mice and WT siblings from the highest expressing line are shown. *, P = 0.017, Tg vs. WT siblings (by hierarchical ANOVA, with genotype, age, and litter as factors). Values are the mean ± SEM of four to seven mice per group.

    Histomorphometry revealed a significant decrease in cortical bone proportion of the femoral mid-diaphysis in Igfbp5 mice (Table 1), consistent with the decreases observed in bone area (Figs. 3C and 4C) and geometry (Fig. 5, B and C). Periosteal circumference was significantly reduced by 14.1% (Table 1), reinforcing the pQCT data (Fig. 5B). The BFR at the periosteum was strikingly compromised by 31.2% (Table 1), indicating a decrease in total osteoblastic activity. The MAR was not altered, implying no disruption to osteoblast differentiation or individual function. Instead, the compromise in BFR was due to a significant shortening of active MS, suggesting a reduction in osteoblast numbers (by perturbations to either proliferation or survival).

    TABLE 1. Dynamic histomorphometry demonstrates altered and opposing osteoblast numbers at the periosteum and endosteum of the femoral mid-diaphysis at 5 wk of age with overexpression of Igfbp5

    Histomorphometry at the endosteum confirmed a significant decrease in the endosteal circumference (Table 1), as observed with pQCT (Fig. 5C), suggesting enhanced total endosteal osteoblastic apposition or impaired osteoclastic resorption. The significant decrease in medullary area (by 16.8%) corroborated the contraction of the endosteum. Consequently, BFR was significantly increased (by 91.1% for females and 44.8% for males compared with WT; Table 1), supporting enhancement of total osteoblastic activity at the endosteum. Although the MAR tended to increase, the elevated endosteal BFR was primarily due to a substantial increase in active mineralizing surface (by 76.7% for females and 35.1% for males), indicating a greater number of osteoblasts. Therefore, even at 5 wk the gender-related compromise in cortical bone density could be partially explained by endosteal osteoblastic activity.

    Discussion

    Our findings demonstrate that overexpression of Igfbp5 has a substantial gender-related effect on the acquisition and maintenance of BMD in mice, determined by DEXA and pQCT. Two independent Tg lines overexpressing Igfbp5 at high and intermediate levels under the direction of a CMV/?A element confirmed the dependence of bone phenotype on the dose of IGFBP-5. The reduction of bone density could be attributed to a significant decrease in osteoblast numbers at the periosteal surface, whereas osteoblastic activity was not altered, as assessed by dynamic histomorphometry. In contrast, osteoblast numbers at the endosteal surface were enhanced, particularly for females. Compromises in bone development were evident before puberty and extended to adulthood. This is the first study to report a significant function for IGFBP-5 in skeletal development that is strongly dependent on gender, age, and bone site (axial vs. appendicular, metaphysis vs. diaphysis, and periosteal vs. endosteal). The implication of these findings is that fluctuation in IGFBP-5 expression and its interaction with gender-related factors could modulate the acquisition of BMD and strength before and during the critical period of bone expansion at puberty and then modulate the maintenance of BMD to old age.

    Differential roles of IGFBP-5 in osteoblastic function

    Our conclusions extend and in part contrast with those of a recent study describing Tg mice overexpressing Igfbp5 under control of the osteocalcin promoter (44). The osteocalcin-Igfbp5 mice displayed transient reductions in BMD that were prominent at 4–5 wk of age; by 8 wk of age, complete recovery was observed. In comparison, bone impairment for the CMV/?A-Igfbp5 mice was similarly strong, although the CMV/?A-Igfbp5 phenotype was sustained throughout the course of the study (3–8 wk) and was dependent on gender and age. Osteocalcin-Igfbp5 mice retained a WT number of osteoblasts, but osteoblastic function was impaired. However, our study revealed changes in osteoblast numbers at the periosteum and endosteum, whereas osteoblastic function was not affected. The osteocalcin-Igfbp5 phenotype is consistent with use of the osteocalcin promoter, which expresses transgenes in mature osteoblasts. Although the osteocalcin-Igfbp5 models were valuable in investigating the autocrine/paracrine role of IGFBP-5 in differentiated osteoblasts, they cannot be used to address questions concerning the influence of local and endocrine IGFBP-5 in developing osteoblasts. The CMV/?A regulatory sequence (9, 39, 43) induces Igfbp5 expression in both immature and mature osteoblasts. Because both immature and mature osteoblasts express Igfbp5 endogenously, transgene expression driven by the CMV/?A regulatory sequence may be more suitable than that driven by the osteocalcin promoter to evaluate the effects of IGFBP-5 derived from developing and mature osteoblasts, and the circulation, on the skeletal phenotype.

    Our study demonstrates for the first time that endogenous IGFBP-5 levels in WT mouse serum increase steadily during development to old age. This developmental induction of serum IGFBP-5 is accelerated in Igfbp5 Tg mice, so total IGFBP-5 levels increased by up to 7.7-fold during prepubertal growth compared with those in WT mice, which exhibit very low serum levels. A recent study has suggested that circulating IGF-I (particularly the ratio of free IGF-I to IGFBP-bound) may contribute to longitudinal and volumetric bone growth (5, 45). IGFBP-5 can effectively compete with IGFBP-3 to form the stable reservoir of IGF in the circulatory ternary complex (46). Similarities between the CMV/?A-Igfbp5 and osteocalcin-Igfbp5 models support a significant role for IGFBP-5 in total osteoblastic function. Differences between the two models may be due to timing, location, or level of transgene expression or to an endocrine role for IGFBP-5. Our study provides additional compelling evidence to support the idea that the response of developing osteoblasts to IGFBP-5 is dependent on age, gender, and bone site.

    Several independent reports have demonstrated that IGFBP-5 stimulates bone cell proliferation in vitro and markers of total osteoblastic function in vivo (16, 18, 19). In line with these reports we observed a significant increase in osteoblast number at the endosteum, resulting in protection of the endosteal surface, and a decrease in the medullary area, particularly for Tg females. Although other findings of our study initially appear contrary to a stimulatory role for IGFBP-5, the difference between a stimulatory or an inhibitory manifestation of IGFBP-5 actions may be due to short-term and localized administration of IGFBP-5 to mature mice vs. a constitutive and widespread increase in Igfbp5 expression from early development. Hence, the decreased periosteal osteoblast numbers in the Tg mice may be due to an excess of IGFBP-5 in the periosteal microenvironment. Posttranslational processing and ECM association play essential roles in regulating IGFBP-5 action, particularly by lowering affinity for the IGFs and delivering the IGFs to tissues and receptors, thereby increasing proliferative action (16, 26, 27). High levels of IGFBP-5 could saturate posttranslational processing systems and ECM binding sites to force excess IGFBP-5 into solution, where it sequesters IGFs with high affinity; at lower levels, IGFBP-5 could be anabolic. In addition, the form of delivery of IGFBP-5 may affect its actions in a cell-specific manner. Recently, Igfbp5 overexpression induced G2/M cell cycle arrest and apoptosis in breast cancer cell lines, whereas exogenous addition had no effect (47).

    Previously, we demonstrated that widespread overexpression of Igfbp5 in mice influenced individual tissue development in different ways, so the severe decline in whole body weight gain was partially due to a significant decrease in skeletal muscle mass relative to body weight (39). Alternatively, the significant increase in fractional brain and liver weights demonstrated that these tissues were protected from the actions of elevated IGFBP-5 in comparison with muscle. Similarly in this study, decreases observed in femoral BMD, area, geometry, and length were significantly less than the decrease in body weight of Tg mice compared with WT mice, suggesting that the total function of osteoblasts and chondrocytes was protected relative to that of myoblasts.

    IGF-dependent and -independent actions of IGFBP-5

    The mechanism accounting for the reduced osteoblast numbers at the periosteum of Igfbp5 Tg mice could be due to sequestration of IGF-I and/or -II in the bone environment, IGF-independent actions, or a combination of IGF-dependent and -independent actions. If IGFBP-5 inhibited IGF action, the phenotype of Igfbp5 mice should approach those of Igf1-, Igf2-, and Igf1r-null mice (9, 12), whereas IGFBP-5 actions unrelated to the IGFs would be revealed as phenotypic characteristics diverging from these null models. Using the methodology described in this study, it has been demonstrated that Igf1- and Igf2-null mice exhibit clearly distinct bone phenotypes (11, 12). The prepubertal compromise of the Igfbp5-overexpressing bone implies a greater interaction with IGF-II than -I, as does the body weight growth pattern (39), consistent with IGFBP-5 possessing a greater affinity for IGF-II. However, overall, the extent and timing of the bone density changes observed in Igfbp5 mice did not entirely conform to either Igf1- or Igf2-null animals, in which gender differences could not be detected. Although comparison between these loss of function vs. our reduced function models cannot tell us about the quantity of IGF-independent actions of IGFBP-5, it can provide qualitative information about the sites and timing of IGF-independent actions. The pQCT, osteocalcin, and histomorphometry analyses highlighted additional differences in bone geometry and osteoblastic function between both null models and Igfbp5 mice. Intriguingly, there was no significant difference in endosteal BFR for the Igf1-null mice at a similar age (11); hence, IGFs and IGFBP-5 may regulate osteoblastic function using multiple mechanisms. Thus, although part of the bone phenotype may be due to partial inhibition of IGF activity, we observed striking differences from Igf null models, suggesting IGF-independent actions for IGFBP-5.

    Previous studies demonstrated that administration of recombinant IGFBP-5 modulated total osteoblastic function in Igf1-null models in vitro and in vivo (19). A recent study revealed that a non-IGF-binding form of IGFBP-5 retained the ability to protect against apoptosis (35). Enhanced survival provided by transgene-derived IGFBP-5 might account for the increase in endosteal osteoblast number, whereas Igf-null models characteristically display a decrease in the number of mature cells (hypoplasia;6, 9, 14). IGFBP-5 contains a nuclear localization sequence and can form an interaction with the four and a half LIM domain 2 transcription factor in osteoblasts (38), suggesting that IGFBP-5 has the ability to form transcription-regulating complexes. Additional work is required to partition the molecular mechanisms mediating the IGF-dependent and -independent actions of IGFBP-5.

    Gender differences in skeleton of Igfbp5 mice

    BMD was affected to a significantly greater extent in male than in female Igfbp5 adult mice, so the absolute values of bone density and other parameters were similar in Tg males and females. Hence, Igfbp5 overexpression could be said to have had a gender-neutralizing effect. The differential response between genders was probably due to a significantly greater rate of bone formation at the endosteum in females, whereas bone formation at the periosteum was reduced equally in both genders. Femoral length was decreased similarly in both genders, suggesting that the role of IGFBP-5 in total chondrocyte action is not gender related. However, regression analysis showed that although many of the other bone parameters correlated with serum IGFBP-5 levels, and despite the reproducible elevation in serum IGFBP-5 in Tg males compared with females, the correlation coefficients were different between genders, implying an additional effect of gender-dependent factors. In addition, BMD was not significantly different between genders until several weeks after the release of pubertal hormones. Hence, despite a similar fold increase in IGFBP-5 levels between genders in adult mice, the bone phenotype of Igfbp5 mice was highly gender dependent.

    It has been established for a long time that IGF-I and -II are major mediators of the actions of sex steroids, such as androgen, estrogen, and progesterone (48, 49, 50). Additionally, GH, which is secreted in a gender-dependent, pulsatile secretion pattern, mediates the majority of its effects via IGF-I (51) and is a key activator of Igfbp5 expression (52). Surprisingly, several studies could not detect significant gender differences in the bones of Igf1-, Igf2-, and GH-null models by DEXA and pQCT (12, 45), suggesting the extent of IGF action mediated by sex steroids is equal in males and females. Several studies have inferred that IGFBP-5 can mediate the effects of sex steroids, e.g. estrogen (53), GnRH (54), progesterone (55), and androgen (56). Additional studies are required to determine the molecular mechanisms by which IGFBP-5 modulates the actions of gender-related factors and whether these mechanisms are IGF dependent or independent.

    Serum IGFBP-5 concentrations in WT mice were consistently higher in males and increased steadily from puberty to adulthood. Moreover, the gender-related changes in endogenous IGFBP-5 in the circulatory system were maintained in Tg mice, and the developmental induction was accelerated. There is no evidence to suggest that transcription from the CMV/?A element is regulated differently between genders from numerous other Tg models (e.g. Ref.57); thus, the sexually dimorphic bone observed in the Igfbp5 mice is unlikely to be due to gender differences in transgene expression. Endogenous Igfbp5 mRNA has unusually long untranslated regions (58), suggesting putative elements for posttranscriptional regulation. Evidence does exist for the enhancement of Igfbp5 mRNA stability in osteoblasts by retinoic acid (59) and prostaglandin E2 (60). Prostaglandin E2 is a local factor that mediates specific effects induced by parathyroid hormone, cytokines (i.e. IL-1 and TGF-?), and mechanical strain.

    In summary: 1) Igfbp5 overexpression decreases BMD, periosteal BFR, and serum osteocalcin levels in mice; 2) Igfbp5 overexpression increases endosteal BFR; 3) IGFBP-5 actions in Tg mice are gender dependent; and 4) the phenotype of Igfbp5 mice cannot be explained solely on the basis of IGF sequestration, providing the first in vivo evidence for IGF-independent effects of IGFBP-5 in bone.

    Acknowledgments

    We thank the Small Animal Barrier Unit staff of The Babraham Institute for their support. Histomorphological analysis of bone samples was performed at the Musculoskeletal Disease Center of the J. L. Pettis Veterans Affairs Medical Center in facilities provided by the U.S. Department of Veterans Affairs.

    References

    Bass S, Delmas PD, Pearce G, Hendrich E, Tabensky A, Seeman E 1999 The differing tempo of growth in bone size, mass, and density in girls is region-specific. J Clin Invest 104:795–804

    Rizzoli R, Bonjour JP, Ferrari SL 2001 Osteoporosis, genetics and hormones. J Mol Endocrinol 26:79–94

    Riggs BL, Khosla S, Melton III LJ 2002 Sex steroids and the construction and conservation of the adult skeleton. Endocr Rev 23:279–302

    Karsenty G, Wagner EF 2002 Reaching a genetic and molecular understanding of skeletal development. Dev Cell 2:389–406

    Yakar S, Rosen CJ, Beamer WG, Ackert-Bicknell CL, Wu Y, Liu JL, Ooi GT, Setser J, Frystyk J, Boisclair YR, LeRoith D 2002 Circulating levels of IGF-1 directly regulate bone growth and density. J Clin Invest 110:771–781

    Lupu F, Terwilliger JD, Lee K, Segre GV, Efstratiadis A 2001 Roles of growth hormone and insulin-like growth factor 1 in mouse postnatal growth. Dev Biol 229:141–162

    Mochizuki H, Hakeda Y, Wakatsuki N, Usui N, Akashi S, Sato T, Tanaka K, Kumegawa M 1992 Insulin-like growth factor-I supports formation and activation of osteoclasts. Endocrinology 131:1075–1080

    Wang J, Zhou J, Bondy CA 1999 Igf1 promotes longitudinal bone growth by insulin-like actions augmenting chondrocyte hypertrophy. FASEB J 13:1985–1990

    Zhang M, Xuan S, Bouxsein ML, von Stechow D, Akeno N, Faugere MC, Malluche H, Zhao G, Rosen CJ, Efstratiadis A, Clemens TL 2002 Osteoblast-specific knockout of the insulin-like growth factor (IGF) receptor gene reveals an essential role of IGF signaling in bone matrix mineralization. J Biol Chem 277:44005–44012

    McCarthy TL, Centrella M, Canalis E 1989 Regulatory effects of insulin-like growth factors I and II on bone collagen synthesis in rat calvarial cultures. Endocrinology 124:301–309

    Kasukawa Y, Baylink DJ, Wergedal JE, Amaar Y, Srivastava AK, Guo R, Mohan S 2003 Lack of insulin-like growth factor I exaggerates the effect of calcium deficiency on bone accretion in mice. Endocrinology 144:4682–4689

    Mohan S, Richman C, Guo R, Amaar Y, Donahue LR, Wergedal J, Baylink DJ 2003 Insulin-like growth factor regulates peak bone mineral density in mice by both growth hormone-dependent and -independent mechanisms. Endocrinology 144:929–936

    Baker J, Liu JP, Robertson EJ, Efstratiadis A 1993 Role of insulin-like growth factors in embryonic and postnatal growth. Cell 75:73–82

    Liu JP, Baker J, Perkins AS, Robertson EJ, Efstratiadis A 1993 Mice carrying null mutations of the genes encoding insulin-like growth factor I (Igf-1) and type 1 IGF receptor (Igf1r). Cell 75:59–72

    Zhao G, Monier-Faugere MC, Langub MC, Geng Z, Nakayama T, Pike JW, Chernausek SD, Rosen CJ, Donahue LR, Malluche HH, Fagin JA, Clemens TL 2000 Targeted overexpression of insulin-like growth factor I to osteoblasts of transgenic mice: increased trabecular bone volume without increased osteoblast proliferation. Endocrinology 141:2674–2682

    Andress DL, Birnbaum RS 1992 Human osteoblast-derived insulin-like growth factor (IGF) binding protein-5 stimulates osteoblast mitogenesis and potentiates IGF action. J Biol Chem 267:22467–22472

    Mohan S, Nakao Y, Honda Y, Landale E, Leser U, Dony C, Lang K, Baylink DJ 1995 Studies on the mechanisms by which insulin-like growth factor (IGF) binding protein-4 (IGFBP-4) and IGFBP-5 modulate IGF actions in bone cells. J Biol Chem 270:20424–20431

    Richman C, Baylink DJ, Lang K, Dony C, Mohan S 1999 Recombinant human insulin-like growth factor-binding protein-5 stimulates bone formation parameters in vitro and in vivo. Endocrinology 140:4699–4705

    Miyakoshi N, Richman C, Kasukawa Y, Linkhart TA, Baylink DJ, Mohan S 2001 Evidence that IGF-binding protein-5 functions as a growth factor. J Clin Invest 107:73–81

    Kanatani M, Sugimoto T, Nishiyama K, Chihara K 2000 Stimulatory effect of insulin-like growth factor binding protein-5 on mouse osteoclast formation and osteoclastic bone-resorbing activity. J Bone Miner Res 15:902–910

    Rajaram S, Baylink DJ, Mohan S 1997 Insulin-like growth factor-binding proteins in serum and other biological fluids: regulation and functions. Endocr Rev 18:801–831

    Schneider MR, Wolf E, Hoeflich A, Lahm H 2002 IGF-binding protein-5: flexible player in the IGF system and effector on its own. J Endocrinol 172:423–440

    Sekiya I, Vuoristo JT, Larson BL, Prockop DJ 2002 In vitro cartilage formation by human adult stem cells from bone marrow stroma defines the sequence of cellular and molecular events during chondrogenesis. Proc Natl Acad Sci USA 99:4397–4402

    Nicolas V, Mohan S, Honda Y, Prewett A, Finkelman RD, Baylink DJ, Farley JR 1995 An age-related decrease in the concentration of insulin-like growth factor binding protein-5 in human cortical bone. Calcif Tissue Int 57:206–212

    Bautista CM, Baylink DJ, Mohan S 1991 Isolation of a novel insulin-like growth factor (IGF) binding protein from human bone: a potential candidate for fixing IGF-II in human bone. Biochem Biophys Res Commun 176:756–763

    Clemmons DR 2001 Use of mutagenesis to probe IGF-binding protein structure/function relationships. Endocr Rev 22:800–817

    Mohan S, Baylink DJ 2002 IGF-binding proteins are multifunctional and act via IGF-dependent and -independent mechanisms. J Endocrinol 175:19–31

    Hayden JM, Mohan S, Baylink DJ 1995 The insulin-like growth factor system and the coupling of formation to resorption. Bone 17:93S–98S

    Mohan S, Libanati C, Dony C, Lang K, Srinivasan N, Baylink DJ 1995 Development, validation, and application of a radioimmunoassay for insulin-like growth factor binding protein-5 in human serum and other biological fluids. J Clin Endocrinol Metab 80:2638–2645

    Langlois JA, Rosen CJ, Visser M, Hannan MT, Harris T, Wilson PW, Kiel DP 1998 Association between insulin-like growth factor I and bone mineral density in older women and men: the Framingham Heart Study. J Clin Endocrinol Metab 83:4257–4262

    Karasik D, Rosen CJ, Hannan MT, Broe KE, Dawson-Hughes B, Gagnon DR, Wilson PW, Visser M, Langlois JA, Mohan S, Kiel DP 2002 Insulin-like growth factor binding proteins 4 and 5 and bone mineral density in elderly men and women. Calcif Tissue Int 71:323–328

    Jehle PM, Jehle DR, Mohan S, Bohm BO 1998 Serum levels of insulin-like growth factor system components and relationship to bone metabolism in type 1 and type 2 diabetes mellitus patients. J Endocrinol 159:297–306

    Boonen S, Mohan S, Dequeker J, Aerssens J, Vanderschueren D, Verbeke G, Broos P, Bouillon R, Baylink DJ 1999 Down-regulation of the serum stimulatory components of the insulin-like growth factor (IGF) system (IGF-I, IGF-II, IGF binding protein [BP]-3, and IGFBP-5) in age-related (type II) femoral neck osteoporosis. J Bone Miner Res 14:2150–2158

    Jehle PM, Schulten K, Schulz W, Jehle DR, Stracke S, Manfras B, Boehm BO, Baylink DJ, Mohan S 2003 Serum levels of insulin-like growth factor (IGF)-I and IGF binding protein (IGFBP)-1 to -6 and their relationship to bone metabolism in osteoporosis patients. Eur J Intern Med 14:32–38

    Cobb LJ, Salih DA, Gonzalez I, Tripathi G, Carter EJ, Lovett F, Holding C, Pell JM 2004 Partitioning of IGFBP-5 actions in myogenesis: IGF-independent anti-apoptotic function. J Cell Sci 117:1737–1746

    Andress DL 1995 Heparin modulates the binding of insulin-like growth factor (IGF) binding protein-5 to a membrane protein in osteoblastic cells. J Biol Chem 270:28289–28296

    Andress DL 1998 Insulin-like growth factor-binding protein-5 (IGFBP-5) stimulates phosphorylation of the IGFBP-5 receptor. Am J Physiol 274:E744–E750

    Amaar YG, Thompson GR, Linkhart TA, Chen ST, Baylink DJ, Mohan S 2002 Insulin-like growth factor-binding protein 5 (IGFBP-5) interacts with a four and a half LIM protein 2 (FHL2). J Biol Chem 277:12053–12060

    Salih DA, Tripathi G, Holding C, Szestak TA, Gonzalez MI, Carter EJ, Cobb LJ, Eisemann JE, Pell JM 2004 Insulin-like growth factor-binding protein 5 (Igfbp5) compromises survival, growth, muscle development, and fertility in mice. Proc Natl Acad Sci USA 101:4314–4319

    Niwa H, Yamamura K, Miyazaki J 1991 Efficient selection for high-expression transfectants with a novel eukaryotic vector. Gene 108:193–199

    Mohan S, Masinde G, Li X, Baylink DJ 2003 Mapping quantitative trait loci that influence serum insulin-like growth factor binding protein-5 levels in F2 mice (MRL/MpJ x SJL/J). Endocrinology 144:3491–3496

    Parfitt AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H, Meunier PJ, Ott SM, Recker RR 1987 Bone histomorphometry: standardization of nomenclature, symbols, and units. Report of the ASBMR Histomorphometry Nomenclature Committee. J Bone Miner Res 2:595–610

    Lobe CG, Koop KE, Kreppner W, Lomeli H, Gertsenstein M, Nagy A 1999 Z/AP, a double reporter for Cre-mediated recombination. Dev Biol 208:281–292

    Devlin RD, Du Z, Buccilli V, Jorgetti V, Canalis E 2002 Transgenic mice overexpressing insulin-like growth factor binding protein-5 display transiently decreased osteoblastic function and osteopenia. Endocrinology 143:3955–3962

    Sjogren K, Sheng M, Moverare S, Liu JL, Wallenius K, Tornell J, Isaksson O, Jansson JO, Mohan S, Ohlsson C 2002 Effects of liver-derived insulin-like growth factor I on bone metabolism in mice. J Bone Miner Res 17:1977–1987

    Baxter RC, Meka S, Firth SM 2002 Molecular distribution of IGF binding protein-5 in human serum. J Clin Endocrinol Metab 87:271–276

    Butt AJ, Dickson KA, McDougall F, Baxter RC 2003 Insulin-like growth factor-binding protein-5 inhibits the growth of human breast cancer cells in vitro and in vivo. J Biol Chem 278:29676–29685

    Ernst M, Heath JK, Rodan GA 1989 Estradiol effects on proliferation, messenger ribonucleic acid for collagen and insulin-like growth factor-I, and parathyroid hormone-stimulated adenylate cyclase activity in osteoblastic cells from calvariae and long bones. Endocrinology 125:825–833

    Gori F, Hofbauer LC, Conover CA, Khosla S 1999 Effects of androgens on the insulin-like growth factor system in an androgen-responsive human osteoblastic cell line. Endocrinology 140:5579–5586

    Fournier B, Gutzwiller S, Dittmar T, Matthias G, Steenbergh P, Matthias P 2001 Estrogen receptor (ER)-, but not ER-?, mediates regulation of the insulin-like growth factor I gene by antiestrogens. J Biol Chem 276:35444–35449

    Liu JL, Yakar S, LeRoith D 2000 Mice deficient in liver production of insulin-like growth factor I display sexual dimorphism in growth hormone-stimulated postnatal growth. Endocrinology 141:4436–4441

    Thoren M, Hilding A, Brismar T, Magnusson P, Degerblad M, Larsson L, Saaf M, Baylink DJ, Mohan S 1998 Serum levels of insulin-like growth factor binding proteins (IGFBP)-4 and -5 correlate with bone mineral density in growth hormone (GH)-deficient adults and increase with GH replacement therapy. J Bone Miner Res 13:891–899

    Adesanya OO, Zhou J, Bondy CA 1996 Cellular localization and sex steroid regulation of insulin-like growth factor binding protein messenger ribonucleic acids in the primate myometrium. J Clin Endocrinol Metab 81:2495–2501

    Onoda N, Li D, Mickey G, Erickson G, Shimasaki S 1995 Gonadotropin-releasing hormone overcomes follicle-stimulating hormone’s inhibition of insulin-like growth factor-5 synthesis and promotion of its degradation in rat granulosa cells. Mol Cell Endocrinol 110:17–25

    Boonyaratanakornkit V, Strong DD, Mohan S, Baylink DJ, Beck CA, Linkhart TA 1999 Progesterone stimulation of human insulin-like growth factor-binding protein-5 gene transcription in human osteoblasts is mediated by a CACCC sequence in the proximal promoter. J Biol Chem 274:26431–26438

    Gregory CW, Kim D, Ye P, D’Ercole AJ, Pretlow TG, Mohler JL, French FS 1999 Androgen receptor up-regulates insulin-like growth factor binding protein-5 (IGFBP-5) expression in a human prostate cancer xenograft. Endocrinology 140:2372–2381

    Katsuno M, Adachi H, Kume A, Li M, Nakagomi Y, Niwa H, Sang C, Kobayashi Y, Doyu M, Sobue G 2002 Testosterone reduction prevents phenotypic expression in a transgenic mouse model of spinal and bulbar muscular atrophy. Neuron 35:843–854

    James PL, Jones SB, Busby Jr WH, Clemmons DR, Rotwein P 1993 A highly conserved insulin-like growth factor-binding protein (IGFBP-5) is expressed during myoblast differentiation. J Biol Chem 268:22305–22312

    Zhou Y, Mohan S, Linkhart TA, Baylink DJ, Strong DD 1996 Retinoic acid regulates insulin-like growth factor-binding protein expression in human osteoblast cells. Endocrinology 137:975–983

    McCarthy TL, Casinghino S, Mittanck DW, Ji CH, Centrella M, Rotwein P 1996 Promoter-dependent and -independent activation of insulin-like growth factor binding protein-5 gene expression by prostaglandin E2 in primary rat osteoblasts. J Biol Chem 271:6666–6671(Dervis A. M. Salih, Subbu)