Transgenic Models of Metabolic Bone Disease: Impact of Estrogen Receptor Deficiency on Skeletal Metabolism

Connective Tissue Research, 44(Suppl. 1): 250–263, 2003 c 2003 Taylor & Francis Copyright ° 0300-8207/03 $12.00 + .00 DOI: 10.1080/03008200390181744

Transgenic Models of Metabolic Bone Disease: Impact of Estrogen Receptor Deficiency on Skeletal Metabolism L. K. McCauley,1,2 T. F. T¨ozum, ¨ 1,3 K. M. Kozloff,4 A. J. Koh-Paige,1 C. Chen,1 1 M. Demashkieh, H. Cronovich,1 V. Richard,4 E. T. Keller,2 T. J. Rosol,5 and S. A. Goldstein3 1

Departments of Periodontics/Prevention/Geriatrics, University of Michigan, Ann Arbor, Michigan, USA Department of Pathology, Medical School, University of Michigan, Ann Arbor, Michigan, USA 3 Department of Periodontology, Faculty of Dentistry, Hacettepe University, Ankara, Turkey 4 Department of Orthopedic Surgery, Medical School, University of Michigan, Ann Arbor, Michigan, USA 5 Department of Veterinary Biosciences, Ohio State University, Columbus, Ohio, USA 2

Estrogen has protective effects on the skeleton via its inhibition of bone resorption. Mechanisms for these effects and the selectivity to the estrogen receptor α (ERα) or ERβ are unclear. The purpose of our study was to determine the impact of the ERα on skeletal metabolism using murine models with targeted disruption of the ERα and β. Mice generated by homologous recombination and Cre/loxP technology yielding a deletion of the ERα exon 3 were evaluated and also crossed with mice with a disruption of the exon 3 of the ERβ to result in double ERα and ERβ knockout mice. Skeletal analysis of long bone length and width, radiographs, dual X-ray absorptiometry, bone histomorphometry, micro computerized tomography, biomechanical analysis, serum biochemistry, and osteoblast differentiation were evaluated. Male ERα knockout mice had the most dramatic phenotype consisting of reduced bone mineral density (BMD), and bone mineral content (BMC) of femurs at 10 and 16 weeks and 8–9 months of age. Female ERα knockout mice also had reduced density of long bones but to a lesser degree than male mice. The reduction of trabecular and cortical bone in male ERα knockout mice was statistically significant. Male double ERα and ERβ knockouts had similar reductions in bone density versus the single ERα knockout mice suggesting that the ERα is more protective than the ERβ in bone. In vitro analysis revealed no differences in osteoblast differentiation or mineralized nodule formation among cells from ERα genotypes. These data suggest that estrogens are important in skeletal metabolism in males; the ERα plays an important role in estrogen protective effects; osteoblast differentiation is not altered with loss of the ERα; and compensatory mechanisms are present in the absence of the ERα and/or another receptor for Received 9 November 2001; accepted 7 February 2002. Address correspondence to L. K. McCauley, 1011 N. University Ave. Ann Arbor, MI 48109-1078, USA. E-mail: [email protected]

250

estrogen exists that mediates further effects of estrogen on the skeleton. Keywords

Bone, Bone Density, Estrogen, Estrogen Receptor, Osteoblasts.

INTRODUCTION A critical tool for understanding metabolic bone disease is the use of gene-targeted murine models. Inactivation or modification of a gene and the consequences of the mutation facilitate the understanding of gene function during development, normal physiology, and disease states. The generation of mice with ablation of the estrogen receptor α (ERα) or β (ERβ) have provided insight into estrogen actions, but have been confounding as they have not demonstrated phenotypes that would have been explained by a loss of ER function and do not typify the disorder of osteoporosis. The incidence of osteoporosis is strongly correlated with the decline in estrogen production in postmenopausal females; however, recent evidence suggests that estrogen deficiency contributes to bone loss in men as well [1–3]. Although significant advances have been made with mechanisms of action of estrogens, it is still unclear exactly how estrogen acts in bone. The largest source of estrogens is the ovary, but the corpus luteum, placenta, adrenal gland, and testes also produce estrogens. Estrogens are thought to influence bone growth and its ultimate cessation by closure of epiphyseal plates, and the inhibition of bone resorption [2, 4]. Eliminating estrogen synthesis in the

SKELETAL PHENOTYPE OF ESTROGEN RECEPTOR-DEFICIENT MICE

ovary, as in ovariectomized (OVX) animals, has confirmed the skeletal impact of estrogen, but it has not been able to provide enough information to clarify the specific mechanisms responsible for estrogen effects in bone. Models targeting the estrogen receptor have been generated and provide the potential for new information to address this critical problem. Estrogens diffuse into target cells and are retained with high affinity and specificity by the intranuclear estrogen receptor (ER) [5]. Once bound, the ER undergoes a conformational change that allows the receptor to bind to chromatin and modulate transcription of target genes. The ER is a member of the steroid hormone receptor family that includes receptors such as the glucocorticoid receptor, the thyroid hormone receptor, and the retinoid receptors [6]. Steroid receptors consist of a hypervariable N-terminal domain that contributes to the transactivation function, a highly conserved central domain responsible for specific DNA binding, dimerization, and nuclear localization, and a C-terminal domain responsible for ligand binding and the ligand-dependent transactivation function [7, 8]. The human estrogen receptor α (ERα) originally cloned in 1986 was found to have high homology between species and, until recently, was thought to be the primary mediator of estrogen action [9]. The ERβ was more recently cloned from rat prostate and found to be highly homologous to the rat ERα, especially in the DNA-binding domain (DBD) (>90%) and the ligand-binding domain (LBD) (55%) [10–12]. The presence of two receptors was thought to provide an explanation for the selective actions of estrogens and antiestrogens in different target tissues, but their precise roles in skeletal and other tissues are still unclear and the possibility of a third receptor has not yet been ruled out [13, 14]. Evidence is building that suggests differential ligand activation may occur via the ERα and ERβ [13–18]. The ER mediates transactivation via a classical estrogen response element (ERE) and also an AP-1 enhancer element that requires ligand and the AP-1 transcription factors Fos and Jun [19]. Interestingly, the ERα and ERβ were found to signal in opposite ways when complexed with estradiol from an AP-1 site in transfected HeLa, breast, and uterine cells [19], but the selectivity of ER action in osteoblasts is unknown. Estrogens have been shown to have direct effects on osteoclasts and osteoblasts in vitro; both of which have been reported to have estrogen receptors [20]. It is largely recognized that the effects of estrogen on bone are due to interactions with the osteoblast/stromal cell in the bone marrow microenvironment to inhibit osteoclast activity, and therefore the decline in estrogen production results in an increase in osteoclastic bone resorption [21]. Although expression of estrogen receptors in osteoblasts has been characterized during differentiation in vitro, the functional activity of the ERα versus ERβ in osteoblasts is still unknown. The purpose of our study was to determine the impact of the ERα on skeletal metabolism using an animal model with targeted mutations in ERα and β resulting in a loss of function.

251

METHODS AND MATERIALS Generation of Estrogen Receptor Mutant Mice The generation of the ERα and ERβ mutant mice has been previously described [22]. Heterozygous ERα +/− mice were bred to yield a Mendelian distribution of ERα +/+, +/−, and −/− mice for the described studies. Heterozygous ERα +/− mice were also crossed with ERβ +/− mice to obtain double ERα −/− β −/− mice. Genotyping on tail-biopsy DNA was performed by PCR using the primers 50 -TTGCCCGATAACAATAACAT-30 , 50 ATTGTCTCTTTCTGACAC-30 , and 50 -GGCATTACCACTTC TCCTGGGAGTCT-30 for ERα and 50 -TATCCCTAGCTCTGG AAGGC-30 , 50 -ACATTTATATCAGATCATCTCTGC-30 , and 50 -AAAGCGCATGCTCCAGACTGC-30 for ERβ. Standard PCR protocols and conditions previously described were used [22]. Faxitron and Gross Analysis Measurement of femur length and width was performed using electronic calipers with length reported as the longest measurement from proximal to distal and the width as the narrowest measurement at the mid-diaphysis. Radiographic analysis was performed using a microradiography system (Faxitron X-Ray Corporation, Wheeling, IL). Dual X-Ray Absorptiometry Scanning densitometry of femurs and tibias of ER mutant mice was performed by dual energy X-Ray absorptiometry (DEXA) using a Norland SABRE pDEXA machine (Norland, Ft. Atkinson, WI, USA) and the Small Animal Research Software. Dissected bones were scanned at 3 mm/s with a sensitivity of 0.5 mm × 0.5 mm. Whole bone scans of femur and standardized area scans of the distal femur were performed. Longitudinal QA consisted of 25 scans of a frozen mouse carcass obtained prior to the study initiation and at various times during data acquisition. Precision (as determined by 16 scans with repositioning of the subject between each scan) was 0.8% for the femur. Bone Histomorphometry Static bone histomorphometry was performed on the proximal tibias and lumbar vertebrae. Briefly, tibias and vertebrae were removed from the mice, trimmed of musculature, and fixed in neutral-buffered formalin for 48 hr. Bones were decalcified in 10% EDTA (pH 7.4) and embedded in paraffin. Sections were cut at 5 µm and stained with hematoxylin and eosin. Measurements were performed with an Olympus microscope and Bioquant Image Analysis System with Nova software and Optronics DEI-750 camera (R&M Biometrics, Nashville, TN, USA) and standardized according to Parfitt et al. [23]. Measurements included total bone area, trabecular bone area (excluding the primary spongiosa), and cortical width (at mid-diaphysis).

252

L. K. MCCAULEY ET AL.

Serum Biochemistry Serum osteocalcin was determined by radioimmunoassay (Biomedical Technologies, Stoughton, MA, USA). Microcomputerized Tomography Three dimensional microcomputerized tomography (MCT) analyses were performed as described to determine the volume of bone and detect differences in the architecture of bone among the genotypes [24, 25]. Vertebrae (10 weeks) were dissected and frozen at −20◦ C until analysis. The MCT scanner digitizes mineralized biologic specimens in three dimensions. Image reconstruction was performed at a scale of 12 microns per voxel. The reconstructed image was then subjected to an adaptive thresholding algorithm so that each voxel now at a 25 micron resolution was assigned a bone or nonbone value [24, 25]. Utilizing a series of image processing algorithms, the geometric, structural, and architectural features of the bones were analyzed. Bone volume (BV/TV), bone surface (BS/BV), trabecular number (TbN), trabecular spacing (TbSp), and trabecular thickness (TbTh) were calculated for the vertebrae. Biomechanical Analyses Biomechanical testing using four-point bending was performed as described [26]. Briefly, femurs from male and female ERα mutant mice at 10 weeks-of-age were dissected, cleaned of remaining tissue, and fixed in neutral-buffered formalin for 48 hr. Femur diameters were measured with calipers immediately before mechanical testing with three values recorded in the anterior-posterior direction and three values in the medial-lateral direction. Femurs were tested to failure in four-point bending on an MTS servohydraulic testing system at a constant displacement rate of 0.55 mm/sec. The following variables were analyzed; stiffness (N/mm), load at yield (N), maximum load (N), and load at failure (N). A total of 8–10 femurs from individual mice of each gender and genotype were evaluated during the course of two independent experiments and results combined to yield a total n of 12–20/genotype. Primary Osteoblast Isolation and Culture Primary calvarial osteoblasts were isolated as previously described [27]. Briefly, mice were euthanized, calvaria dissected

from periosteum and subjected to sequential digestions in collagenase A (2 mg/ml, Roche Biochemicals, Indianapolis, IN, USA) and 0.25% trypsin (Gibco BRL) for 20, 40, and 90 min. Cells from the third digest were washed, counted, and plated in phenol red-free α-MEM with 10% charcoal-stripped fetal bovine serum containing 100 units/ml of penicillin and streptomycin. Primary cultures were used without passage. Primary calvarial cells were plated at 50,000/cm2 and induced to differentiate and form a mineralized matrix with the addition of ascorbic acid (50 µg/ml) and β-glycerophosphate (100 mM) as described [28]. After a 24-day culture period, calcium accumulation in cell layers was determined by extraction with 15% trichloroacetic acid followed by colorimetric determination with cresolphlthalein complexone (Sigma, St. Louis, MO, USA) [27]. Statistical Analyses All experiments were performed 2–4 times to confirm accuracy and reproducibility of results. Data are expressed as mean ±SEM of samples from multiple mice as indicated. ANOVA or Student’s t-test was performed using GraphPad Instat with significance at p < .05. RESULTS Microradiography and Gross Analysis The vertebrae of the tail in mice undergo a well-characterized pattern of mineralization during development that can be quantified radiographically. Mice at postnatal day 4 and 6 were sacrificed and tail vertebrae were radiographed. The number of vertebrae that were mineralized from a fixed reference were quantified. There was no significant difference in ERα +/+, +/−, and −/− mice (data not shown). This indicated that there was no alteration in the rate of mineralization during development, as measured at a fixed time point, associated with a loss of the ERα. At 10 weeks-of-age there was no difference in the length of the femurs of ERα −/− compared with ERα +/+ mice for either males or females; but there was a significant reduction in femur width in ERα −/− males at 10 weeks (Table 1). There

TABLE 1 Gross femur measurements: ERα mutant mice, 10 and 16 weeks-of-age, mean (mm) ± SEM, (n) Genotype Male ERα +/+ Male ERα +/− Male ERα −/− Female ERα +/+ Female ERα +/− Female ERα −/− ∗

p < .05 vs. +/+.

Femur length (10 wks)

Femur length (16 wks)

Femur width (10 wks)

Femur width (16 wks)

15.73 ± 0.15 (8) 15.69 ± 0.12 (6) 15.35 ± 0.14 (10) 14.72 ± 0.15 (7) 15.17 ± 0.15 (7) 14.76 ± 0.26 (7)

15.79 ± 0.12 (8) 16.15 ± 0.15 (8) 15.96 ± 0.23 (6) 16.39 ± 0.33 (3) 15.99 ± 0.24 (4) 15.62 ± 0.27 (6)

1.35 ± 0.02 (8) 1.33 ± 0.04 (6) 1.25 ± 0.08 (10)∗ 1.22 ± 0.03 (7) 1.32 ± 0.03 (7) 1.23 ± 0.03 (7)

1.28 ± 0.01 (7) 1.25 ± 0.03 (8) 1.23 ± 0.03 (7) 1.24 ± 0.018 (6) 1.25 ± 0.023 (6) 1.20 ± 0.023 (7)

SKELETAL PHENOTYPE OF ESTROGEN RECEPTOR-DEFICIENT MICE

were no differences in length or width of femurs among any of the genotypes at 16 weeks; but there was a trend toward a reduction in the width of the femurs in both male and female ERα −/− mice. Radiographic analysis of mouse long bones at 12 weeks-ofage revealed decreased density of cancellous and cortical bone in the distal metaphysis of male ERα −/− mice compared with ERα +/+ mice (Figure 1).

Figure 1. Radiographic analysis of femurs from ERα mutant mice. (A) Representative Faxitron X-ray of 12-week-old male ERα −/− (left) vs. ERα +/+ (middle and right). Arrow indicates relative lack of trabeculation in the ERα −/− and cortical thinning. (B) Representative dual X-ray absorptiometry (DEXA) scanning of a femur from a male ERα −/− (left), ERα +/− (middle), and ERα +/+ (right) mouse, age 10–11 weeks. Areas of highest density are indicated by yellow-white coloration whereas areas of lowest density are indicated by purple coloration.

253

Dual X-Ray Absorptiometry Areal bone mineral density (BMD) and bone mineral content (BMC) were measured on whole femurs of ER mutant mice at various ages. These parameters also were measured in the distal femur to isolate an area with greater bone volume. Figure 1B is a representative scan from ERα mutant mice at 12 weeksof-age when there was reduced bone density (demonstrated by the intensity of coloration, yellow-white is highest density) in the ERα −/− femur. Figure 2A and 2B show the BMD and BMC of male ERα mutant mice at 10 weeks, 16 weeks, and 8– 9 months. There was a reduction in BMD and BMC of the whole femur and the distal femur in the 10-week-old male ERα −/− mice versus +/+ or +/− mice. At 16 weeks-of-age, the male ERα −/− still had significantly reduced BMD and BMC for both the whole femur and the distal femur. At 8–9 months-ofage male ERα −/− mice had similar reductions in bone density. Statistically significant reductions in BMD of the distal femur and BMC and area of both the whole femur and distal femur were present, and a trend toward a reduction in the whole femur BMD was present. In contrast, the female ERα mutant mice did not have altered BMD of the whole femur at any age (Figure 2C, 2D). In female mice at 16 weeks-of-age there were significant reductions in the BMC and the BMD of the distal femur; however, the BMD of the whole femur was not statistically reduced compared with ERα +/+ mice. Further, the female ERα mutant mice had no change in BMD or BMC at 8–9 months-of-age. DEXA analysis was performed on the double ERα −/− ERβ −/− mice to characterize the relative contribution of the ERα to the phenotype and determine whether loss of ERβ modified the results. Figure 3A, 3B demonstrate representative DEXA scans and BMD values for male ERα−/−β−/− versus ERα+/−β+/− at 10, 16, 19 weeks, and 8–9 months. At all time points the BMD levels were similarly reduced for the double-mutant mice compared with heterozygotes and were statistically significant at 10, 16, and 19 weeks-of-age. Interestingly, it appeared as if the ‘peak’ bone mass was achieved later in ERα−/−β−/− mice (19 weeks) compared with the ERα+/−β+/− mice (16 weeks). Bone Histomorphometry Static histomorphometric analysis of the proximal tibia metaphysis revealed a trend toward reduced total bone in male ERα mice (Figure 4). Reductions in percent trabecular bone were statistically significant in the proximal tibia of male ERα-nullmutant mice compared with wild-type or heterozygous mice. Female ERα-null mice had no difference in total bone or trabecular bone area. Cortical thickness, measured at the mid-diaphysis of the tibia, was significantly reduced in male ERα −/− mice (Figure 5). This reduction did not appear to be associated with an alteration in bone formation as the numbers of osteocytes/mm2 of cortical bone were unchanged. There were no alterations in the trabecular bone area of the vertebrae of 10-week-old male or female ERα mutant mice (Figure 6).

254

L. K. MCCAULEY ET AL.

Figure 2. DEXA analysis of femurs from male (A, B) or female (C, D) mice, at 10, 16, and 8–9 months. Bone mineral density (A, C), bone mineral content (B, D) were evaluated. Data are expressed as mean ± SEM; n = 5–10/group. (Continued)

SKELETAL PHENOTYPE OF ESTROGEN RECEPTOR-DEFICIENT MICE

Figure 2. (Continued)

255

256

L. K. MCCAULEY ET AL.

trabecular number and reduced trabecular spacing, which likely contributed to the overall increased bone volume. The biomechanical analyses revealed a trend toward reduced stiffness in the male ERα −/− mice versus the +/+ or +/− mice (Table 3); however, these differences were not significant when the size and shape of the bones were taken into account, suggesting that there was no inherent alteration in tissue material properties. There were no other significant differences in biomechanical parameters in either gender among the genotypes. Serum Biochemistry Osteocalcin (OCN) is the most abundant noncollagenous protein produced by osteoblasts, and serum levels reflect bone formation. Using a mouse osteocalcin radioimmunoassay, OCN levels were determined in ERα mutant mice at 10 weeks-of-age. There was no significant difference in serum OCN concentration in the ERα −/− versus the ERα +/+ mice (Figure 8). This further supported the data that there was no alteration in bone formation with loss of the ERα. Primary Osteoblast Differentiation Assays to evaluate primary osteoblast-like cell differentiation were performed on ERα mutant mice. Primary mouse calvarial cells were isolated from mouse pups at day 4. After a 20–24 day differentiation period, cells were harvested for quantitative and qualitative measurement of calcium incorporation and mineralized nodule formation. There was no difference in the ability of ERα −/− osteoblastic cells to form a mineralized matrix in vitro (Figure 9). These data indicate that the ERα present in wild-type osteoblastic cells is not required for normal differentiation and bone formation in vitro, although it is possible, but unlikely, that compensatory mechanisms are present in the in vitro culture system.

Figure 3. DEXA analysis of femurs from double ER mutant mice. (A) Representative DEXA scanning of a femur from a male ERα −/− ERβ −/− (left), ERα +/− ERβ +/− (middle), and ERα +/+ ERβ +/− (right) mouse, age 10–11 weeks. (B) Bone mineral density of ERα −/− ERβ −/− and ERα +/− ERβ+/− mouse femurs at 10, 16, 19 weeks, and 8–9 months of age. Data are expressed as mean ± SEM.

Microcomputerized Tomography and Biomechanical Analyses Microcomputerized tomographic analysis of the vertebral bone of male and female ERα mutant mice revealed increased bone volume in vertebrae from ERα −/− mice at 10 weeks (Figure 7 and Table 2). ERα −/− female vertebrae had increased

DISCUSSION Estrogen has long been recognized for its protective skeletal effects, but little is known regarding the subtype of the estrogen receptor responsible for effects in bone. Extensive work has been performed with animal models of surgically induced menopause (OVX) and bone mass [29–32], but little data exist for animals with a genetically defined ER status. In the present study, male and female ERα mutant mice displayed abnormal but not dramatically altered phenotypes in bone. The most prominent finding was a reduction in bone density in male ERα −/− mice. Similar findings have been reported for male mice by Vidal et al. [33], but some of their results were different from ours and they suggested that IGF-1 was responsible for the effects of estrogen receptor ablation in male mice. In-depth investigation of putative factors responsible for the reduction in bone density found in ER mutant mice is still lacking and hinders the progress of understanding the effects of estrogens in bone. In contrast to the few studies characterizing the skeleton after ER ablation, many studies have focused on the reproductive phenotype of ER knockout mice [34–36]. Signaling through the ERs

SKELETAL PHENOTYPE OF ESTROGEN RECEPTOR-DEFICIENT MICE

257

Figure 4. Bone histomorphometry of 10-weeks-old ERα mutant mice. (A) Representative proximal tibia of male mice, (B) total bone area of male proximal tibias, (C) trabecular area expressed as a percent of total bone area of male proximal tibiae, (D) total bone area of female proximal tibiae, (E) trabecular area expressed as a percent of total bone area of female proximal tibiae. Data are expressed as mean ± SEM. The n = 8 ERα +/+, n = 4-5 ERα +/−, and n = 8-9 ERα −/− tibias for males and n = 7 for each female genotype.

258

L. K. MCCAULEY ET AL.

Figure 5. Cortical bone of 10-weeks male ERα mutant mice tibias, (A) Representative photo of cortical bone from mid-diaphysis of tibiae from each genotype, (B) cortical thickness of tibiae, (C) osteocyte numbers/mm2 cortical bone. Data are expressed as mean ± SEM.

Figure 6. Bone histomorphometry of vertebrae at 10 weeks. (A) Male or (B) female ERα mutant mice show trabecular bone area expressed as a percent of total bone area of a lumbar vertebra. Data are expressed as mean ± SEM. The n = 8 ERα +/+, n = 5 ERα +/−, n = 9 ERα −/− for male vertebrae. The n = 8 ERα +/+, n = 8 ERα +/− and n = 9 ER −/− for female vertebrae.

259

SKELETAL PHENOTYPE OF ESTROGEN RECEPTOR-DEFICIENT MICE

Figure 7. Representative micro-CT scan of vertebrae from male and female ERα mutant mice at 10 weeks.

is indispensable for folliculogenesis as evidenced by an arrest in follicle growth in mice with a double knockout of the ERα and β [22], and mice are anovulatory [37]. Furthermore, in the female ERα mutant mice, there are 10-fold higher circulating estradiol concentrations and higher androgen concentrations compared with their wild-type littermates [34]. This may account for the differences in the bone phenotype due to compensatory factors present in the female ERα −/− mice as compared with the males. For example, higher androgen levels in the female ERα −/− mice may contribute to the increase in bone volume found in the vertebrae. The impact of ERα deletion was evaluated in-depth in mice at three ages. At 10 weeks-of-age, mice are sexually mature but still have not reached their adult body size. At 16 weeksof-age, the murine skeleton is considered mature and does not significantly increase in size after this time. At 8–9 months-

of-age the cumulative effects of aging on the skeleton can be investigated. The DEXA analysis revealed changes in the ERα mutant mice relative to their wild-type and heterozygous littermates at all three ages. Analysis was performed on the whole femur to evaluate bone as an organ (composed of cortical and trabecular bone) and the distal femur that has a higher ratio of trabecular to cortical bone than the entire bone. In male mice the ER mutation affected both trabecular and cortical bone in the proximal tibia. This contrasts to the recent report of Vidal et al. who did not find alterations in the trabecular bone [33] of male ERα −/− mice. Indeed, histomorphometric analysis of the vertebrae did not reveal differences in trabecular bone of male or female ERα −/− mice, and micro-CT indicated no alteration in bone volume in male vertebrae. This may be due to differences in appendicular versus axial skeleton and reflect site-specific alterations in the hormonal responses for bones

TABLE 2 Micro-CT analysis: ERα mutant mouse vertebrae, 10 weeks-of-age (mean ± SEM) Genotype (n) Male ERα +/+ (9) Male ERα +/− (6) Male ERα −/− (10) Female ERα +/+ (7) Female ERα +/− (7) Female ERα −/− (7) ∗

BV/TV

BS/BV

Tb N

Tb Sp

Tb Th

0.63 ± 0.12 0.63 ± 0.09 0.69 ± 0.08 0.47 ± 0.04 0.48 ± 0.09 0.56 ± 0.06∗

16.6 ± 3.98 16.2 ± 3.54 16.1 ± 3.94 22.1 ± 1.70 22.6 ± 2.84 25.4 ± 5.12

4.62 ± 0.41 4.36 ± 0.67 5.05 ± 0.91 4.91 ± 0.22 5.10 ± 0.42 7.08 ± 1.02∗∗

0.057 ± 0.02 0.058 ± 0.01 0.045 ± 0.01 0.086 ± 0.01 0.086 ± 0.02 0.054 ± 0.01∗∗

0.145 ± 0.04 0.145 ± 0.04 0.170 ± 0.09 0.092 ± 0.01 0.088 ± 0.01 0.082 ± 0.01

p < .05 vs. +/+; ∗∗ p < 0.1 vs. +/+, +/−.

260

L. K. MCCAULEY ET AL.

TABLE 3 Biomechanical analysis: ERα mutant mouse femurs, 10 weeks-of-age (mean ± SEM) Genotype (n) Male ERα +/+ (20) Male ERα +/− (12) Male ERα −/− (20) Female ERα +/+ (17) Female ERα +/− (16) Female ERα −/− (16) ∗

Stiffness (N/mm)

Yield load (N)

Diameter (mm)

Maximum load (N)

Failure load (N)

194.5 ± 54.0 202.7 ± 47.2 147.4 ± 35.3∗ 175.4 ± 50.6 160.8 ± 61.6 185.8 ± 48.3

17.43 ± 7.48 17.88 ± 6.92 13.67 ± 5.84 15.18 ± 7.74 14.59 ± 6.24 11.99 ± 4.53

1.45 ± 0.08 1.43 ± 0.15 1.34 ± 0.11∗ 1.31 ± 0.10 1.60 ± 0.09 1.35 ± 0.11

29.82 ± 10.7 29.46 ± 5.88 22.94 ± 5.50 21.87 ± 6.70 23.04 ± 6.25 21.43 ± 4.20

27.98 ± 5.76 29.33 ± 5.94 22.80 ± 5.60 21.00 ± 6.60 22.60 ± 6.41 20.55 ± 4.38

p < .01 vs. +/+.

Figure 8.

Serum osteocalcin levels from male (A) and female (B) ERα mutant mice at 10 weeks-of-age. Data are expressed as mean ± SEM.

Figure 9. Differentiation of primary osteoblastic cells from ERα mutant mice. (A) Von Kossa stain of mineralization from ERα +/+ (left) or ERα −/− (right) primary calvarial cells; (B) calcium accumulation in nodules of primary calvarial cells (24 days in culture). Data are expressed as mean ± SEM for triplicate samples.

SKELETAL PHENOTYPE OF ESTROGEN RECEPTOR-DEFICIENT MICE

that develop differently or are subjected to different mechanical forces. The BMD measured by the DEXA represents the grams of mineral per area of bone and depends on the area that is detected as bone. If the sizes of the bones are the same, the BMC may be a more accurate representation of mineral content since it is an actual measurement versus a derived measurement. Although our gross measurements of the femurs of the ERα mutant mice revealed no statistical differences in the length or width of the bones, there was a trend toward a reduced length at 10 weeks and a reduction in width of the femurs in male ERα −/− mice at 16 weeks that likely represents a difference in size when translated into three dimensions. In agreement with this, the area (as a two dimensional transformation of the three dimensions) calculated by the DEXA was reduced in the male ERα −/− mice at most ages (data not shown). The reduction in BMD in the male ERα mutant mice ranged from ∼10–30% that of wild-type, and was not progressive with age. Male mice with ablation of both ERα and ERβ displayed a similar 10–30% reduction in BMD, suggesting that the ERβ does not play a significant additional role in determination of bone mass. Interestingly, the addition of a 19-week group in the double knockout group suggests that these mice experienced a delay in attaining their “peak” bone mass. This raises the possibility that the action of estrogen with both the ERα and β may facilitate bone growth. The female ERα mutant mice were generally much less affected by loss of the ERα; however, differences were apparent. The first suggestion (at 10 weeks-of-age) was the reduction in BMD of the distal femur. At 16 weeks-of-age the BMC was reduced in the whole femur and both BMD and BMC were reduced in the distal femur. By 8–9 months-of-age there were no longer any detectable changes in the bone density of female ERα mutant mice. It is clear that compensatory factors in the male and female ERα mutant mice are different and likely contribute to the phenotype. To limit these compensatory factors and gain a better understanding of the cellular and molecular actions of the ER in bone, an in vitro evaluation of osteoblast function and differentiation was performed. Estrogen receptors have been detected in osteoblasts suggesting a role for estrogen in the regulation of osteoblast function [18, 38–41]. Published data on the effects of estrogen on osteoblasts are varied and often contradictory. Estrogen has been reported to increase, decrease, or have no effect on osteoblast proliferation and differentiation [42]. Different model systems and cell culture techniques may be responsible for variable results. Alternatively, the direct effects of estrogen in osteoblast function may not be fully realized using only an in vitro model system. The estrogen receptor knockout mice are an ideal model system to investigate these effects. The findings reported here that osteoblasts from the ERα knockout mice differentiate normally in vitro and the ERα knockout mice have normal circulating osteocalcin concentrations suggest that the ERα does not play a central role in osteoblast differentiation. In other studies, the administration of estrogen to ovariectomized ERα knockout mice resulted in a stimulation

261

of bone formation at the endocortical surface suggesting that osteoblasts may respond to estrogen by a mechanism other than the ERα [43]. Osteoclasts also have been reported to have estrogen receptors, and estrogen has well-characterized effects on inhibiting bone resorption and osteoclast activity. Whether this is a direct effect on the osteoclasts, such as stimulation of apoptosis, or by an indirect mechanism acting through the osteoblast is unclear; however, much data support the latter. Several major cytokines have been suggested to mediate an indirect effect of estrogen on osteoclast activity. These include IL-1, IL-6, TGFβ, and more recently osteoprotegerin [42, 44–46]. SUMMARY The ERα knockout mice have reduced bone density that is, at least in part, due to a reduction in trabecular and cortical bone. Males and females are both affected, with males having a more dramatic phenotype. Parameters of osteoblast differentiation were unaltered with ERα ablation both in vitro and in vivo. Double ERα and β knockout mice display a similar skeletal phenotype as the ERα knockout mice. These data suggest that the ERα is at least partially responsible for the protective effects of estrogen in mice and may be due in part to the prevention of osteoclastic bone resorption. These data also suggest that there are either compensatory mechanisms to protect against loss of the ERα or that another receptor exists that also plays a role in estrogen actions in bone. The subtlety of the phenotype in comparison to the skeletal effects of ovariectomy or orchidectomy continues to be an enigma, but it also emphasizes the difference in developmental versus age-associated hormonal alterations. Female ERα knockout mice have increased circulating levels of estradiol and testosterone that may confound the data. Furthermore, recent reports of a cell membrane estrogen receptor raise the possibility that the lack of a more profound phenotype may be attributed to actions of estrogen at this nongenomic receptor [47, 48]. If this is true, our data would suggest that this receptor is not the same gene product as the nuclear ERα. ACKNOWLEDGMENTS The authors express appreciation to A. Kr¨ust, S. Dupont, and P. Chambon for providing the mice for these studies and Vicky Li for technical assistance. This work was partially supported by the Nathan Shock Center for the Basic Biology of Aging (NIA AG 13283), the NIA (AG15904) to E.T. Keller, the Center for Craniofacial Regeneration, the Michigan Arthritis Center Biomechanics and Image Processing Core (AR 20577), the National Cancer Institute (CA77911), the National Center for Research Resources (RR00168), the Glenn Barber Funds at the Ohio State University, TUBITAK (Scientific and Technical Research Council of Turkey) Integrated Ph.D. Program, and the Program in Comparative Integrative Genomics. REFERENCES [1] Riggs, B.L., Khosla, S., and Melton, J. (1998). A unitary model for involutional osteoporosis: Estrogen deficiency causes both type I and type

262

[2]

[3]

[4] [5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18] [19]

[20] [21]

L. K. MCCAULEY ET AL. II osteoporosis in postmenopausal women and contributes to bone loss in aging men. J. Bone Miner. Res. 13:763–773. Cutler, G.B. (1997). The role of estrogen in bone growth and maturation during childhood and adolescence. J. Steroid Biochem. Mol. Biol. 61:141– 144. Takagi, M., Miyashita, Y., Koga, M., Ebara, S., Arita, N., and Kasayama, S. (2000). Estrogen deficiency is a potential cause for osteopenia in adult male patients with Noonan’s syndrome. Calci. Tiss. Inter. 66:200–203. Compston, J.E. (2001). Sex steroids and bone. Physiol. Rev. 81:419–447. Nilsson, S., M¨akel¨a, S., Treuter, E., Tujague, M., Thomsen, J., Andersson, G., Enmark, E., Pettersson, K., Warner, M., and Gustafsson, J. (2001). Mechanisms of estrogen action. Physiol. Rev. 81:1535–1565. Mangelsdorf, D.J., Thummel, C., Beato, M., Herrlich, P., Schutz, G., Umesono, K., Blumberg, B., Kastner, P., Mark, M., and Chambon, P. (1995). The nuclear receptor superfamily: The second decade. Cell 83:835–839. Murphy, L.C., Dotzlaw, H., Leygue, E., Douglas, D., Coutts, A., and Watson, P.H. (1997). Estrogen receptor variants and mutations. Mol. Biol. 62:363–372. Resche-Rigon, M., and Gronemeyer, H. (1998). Therapeutic potential of selective modulators of nuclear receptor action. Curr. Opin. Chemi. Biol. 2:501–507. Green, S., Walter, P., Kumar, V., Krust, A., Bornert, J.M., Argos, P., and Chambon, P. (1986). Human estrogen receptor cDNA: Sequence, expression and homology to v-erb-A. Nature 320:134–139. Kuiper, G.G.J.M., Enmark, E., Pelto-Huikko, M., Nilsson, S., and Gustafsson, J.A. (1996). Cloning of a novel estrogen receptor expressed in rat prostate and ovary. Biochemistry 93:5925–5930. Kuiper, G.G.J.M., and Gustafsson, J.A. (1997). The novel estrogen receptor-ß subtype: Potential role in the cell- and promotor-specific actions of estrogens and anti-estrogens. FEBS Lett. 410:87–90. Nilsson, S., Kuiper, G.G.J.M., and Gustafsson, J.A. (1998). ERß: A novel estrogen receptor offers the potential for new drug development. TEM 9:387–395. Bodenner, D.L., Kozlowski, M., and Manolagas, S.C. (1998). Switch from transcriptional inhibition to stimulation of the IL-6 gene depending upon the estrogen receptor form (α vs. β) and type of ligand (estradiol vs. hydroxytamoxifen). J. Bone. Miner. Res. 23:S186 Coleman, K.M., Gustafsson, J.A., and Smith, C.L. (1999). Activation of estrogen receptor-alpha and estrogen receptor-beta by ligand-dependent and ligand-independent pathways. Abstracts Endocr. Soc. P1:234(abstr.) Paige, L.A., Christensen, D.J., Grøn, H., Norris, J.D., Gottlin, E., Padilla, K.M., Chang, C.-Y., Ballas, L.M., Hamilton, P.T., McDonnell, D.P., and Fowlkes, D.M. (1999). Estrogen receptor (ER) modulators each induce distinct conformational changes in ER α and ER β. Med. Sci. 96:3999– 4004. Zou, A., Marschke, K.B., Arnold, K.E., Berger, E.M., Fitzgerald, P., Mais, D.E., and Allegretto, E. (1999). Estrogen receptor ß activates the human retinoic acid receptor α-1 promoter in response to tamoxifen and other estrogen receptor antagonists, but not in response to estrogen. Mol. Endocrinol. 13:418–430. Webb, P., Nguyen, P., Valentine, C., Lopez, G., Kwok, G., McInerney, E., Katzenellenbogen, B., Enmark, E., Gustafsson, J., Nilsson, S., and Kushner, P. (1999). The estrogen receptor enhances AP-1 activity by two distinct mechanisms with different requirements for receptor transactivation functions. Mol. Endocrinol. 13:1672–1685. Bland, R. (2000). Steroid hormone receptor expression and action in bone. Clin. Sci. 98:217–240. Paech, K., Webb, P., Kuiper, G.G.J.M., Nilsson, S., Gustafsson, J.A., Kushner, P.J., and Scanlan, T.S. (1997). Differential ligand activation of estrogen receptors ERα and ERβ at AP1 sites. Science 277:1508–1510. Oursler, M.J. (1998). Estrogen regulation of gene expression in osteoblasts and osteoclasts. Crit. Rev. Eukaryotic Gene Expres. 8:125–140. Pacifici, R. (1998). Estrogen and bone loss. Adv. Organ Biol. 5C:641– 659.

[22] Dupont, S., Kr¨ust, A., Gansmuller, A., Dierich, A., Chambon, P., and Mark, M. (2000). Effect of single and compound knockouts of estrogen receptors alpha (ERα) and beta (ERβ) on mouse reproductive phenotypes. Development 127:4277–4291. [23] Parfitt, A.M., Drezner, M.K., Glorieux, F.H., Kanis, J.A., Malluche, H., Meunier, P.J., Ott, S.M., and Recker, R.R. (1989). Bone histomorphometry: Standardization of nomenclature, symbols, and units. J. Bone. Miner. Res. 2:595–610. [24] Goulet, R.W., Goldstein, S.A., Ciarelli, M.J., Kuhn, J.L., Brown, M.B., and Feldkamp, L.A. (1994). The relationship between the structural and orthogonal compressive properties of trabecular bone. J. Biomech. 27:375– 389. [25] Kuhn, J.L., Goldstein, S.A., Feldkamp, L.A., Goulet, R.W., and Jesion, G. (1990). Evaluation of a microcomputed tomography system to study trabecular bone structure. J. Orthop. Res. 8:833–842. [26] Xu, T., Bianco, P., Fisher, L.W., Longenecker, G., Smith, E., Goldstein, S., Bonadio, J., Boskey, A., Heegaard, A.M., Sommer, B., Satomura, K., Dominguez, P., Zhao, C., Kulkarni, A.B., Robey, P.G., and Young, M.F. (1998). Targeted disruption of the biglycan gene leads to an osteoporosislike phenotype in mice. Nat. Gen. 20:78–82. [27] Koh, A.J., Beecher, C.A., Rosol, T.J., and McCauley, L.K. (1999). 30 ,50 -Cyclic adenosine monophosphate activation in osteoblastic cells: Effects of parathyroid hormone-1 receptors and osteoblastic differentiation, in vitro. Endocrinology 140:3154–3162. [28] McCauley, L.K., Koh, A.J., Beecher, C.A., Cui, Y., Decker, J.D., and Francheschi, R.T. (1995). Effects of differentiation and transforming growth factor beta on PTH/PTHrP receptor mRNA levels in MC3T3-E1 cells. J. Bone Miner. Res. 10:1243–1255. [29] Kimble, R.B., Bain, S., and Pacifici, R. (1997). The functional block of TNF but not of IL-6 prevents bone loss in ovariectomized mice. J. Bone Miner. Res. 12:935–941. [30] Kimmel, D.B., Bozzato, R.P., Kronis, K.A. Coble, T., Sindrey, D., Kwong, P., and Recker, R.R. (1993). The effect of recombinant humen (1-84) or synthetic human (1-34) parathyroid hormone on the skeleton of adult osteopenic ovariectomized rats. Endocrinology 132:1577–1584. [31] Lane, N.E., Haupt, D., Kimmel, D.B., Modin, G., and Kinney, J.H. (1999). Early estrogen replacement therapy reverses the rapid loss of trabecular bone volume and prevents further deterioration of connectivity in the rat. J. Bone Miner. Res. 14:206–214. [32] Shen, V., Birchman, R., Xu, R., Otter, M., Wu, D., Lindsay, R., and Dempster, D. (1995). Effects of reciprocal treatment with estrogen and estrogen plus parathyroid hormone on bone structure and strength in ovariectomized rats. J. Clin. Invest. 96:2331–2338. [33] Vidal, O., Lindberg, M.K., Hollberg, K., Baylink, D.J., Andersson, G., Lubahn, D.B., Mohan, S., Gustafsson, J.A., and Ohlsson, C. (2000). Estrogen receptor specificity in the regulation of skeletal growth and maturation in male mice. Proc. Natl. Acad. Sci. USA 97:5474–5479. [34] Couse, J.F., and Korach, K.S. (1999). Estrogen receptor null mice: What have we learned and where will they lead us? Endocr. Rev. 20:358–417. [35] Couse, J.F., Curtis, S.W., Washburn, T.F., Lindzey, J., Golding, T.S., Lubahn, D.B., Smithies, O., and Korach, K.S. (1995). Analysis of transcription and estrogen insensitivity in the female mouse after targeted disruption of the estrogen receptor gene. Mol. Endocrinol. 9:1441– 1454. [36] Curtis, S.W., Clark, J., Myers, P., and Korach, K.S. (1999). Disruption of estrogen signaling does not prevent progesterone action in the estrogen receptor α knockout mouse uterus. Biochemistry 96:3646–3651. [37] Schomberg, D.W., Couse, J.F., Mukherykee, A., Lubahn, D.B., Sar, M., Mayo, K.E., and Korach, K.S. (1999). Targeted disruption of the estrogen receptor α (ERα) gene in female mice: Characterization of ovarian responses and phenotype in the adult. Endocrinology 140:2733–2744. [38] Komm, B.S., Terpening, C.M., Benz, D.J., Graeme, K.A., Gallegos, A., Korc, M., Greene, G.L., O’Malley, B.W., and Haussler, M.R. (1988). Estrogen binding, receptor mRNA and biologic response in osteoblastlike osteosarcoma cells. Science 241:81–84.

SKELETAL PHENOTYPE OF ESTROGEN RECEPTOR-DEFICIENT MICE [39] Eriksen, E.F., Colvard, D.S., Berg, N.J., Graham, M.L., Mann, K.G., Spelsberg, T.C., and Riggs, B.L. (1988). Evidence of estrogen receptors in normal human osteoblast-like cells. Science 241:84–86. [40] Arts, J., Kuiper, G.G.J.M., Janssen, J.M.M.F., Gustafsson, J.A., Lowik, C.W., Pols, H.A.P., and van Leeuwen, J.P. (1997). Differential expression of estrogen receptors α and β mRNA during differentiation of human osteoblast SV-HFO cells. Endocrinology 138:5067–5070. [41] Bodine, P.V., Henderson, R.A., Green, J., Aronow, M., Owen, T., Stein, G.S., Lian, J.B., and Komm, B.S. (1998). Estrogen receptor-alpha is developmentally regulated during osteoblast differentiation and contributes to selective responsiveness of gene expression. Endocrinology 139:2048– 2057. [42] Oursler, M.J., Landers, J.P., Riggs, B.L., and Spelsberg, T.C. (1993). Oestrogen effects on osteoblasts and osteoclasts. Ann. Med. 25:361– 371. [43] Vandenput, L., Ederveen, A.G.H., Erben, R., Stahr, K., Swinnen, J.V., Van Herck, E., Verstuyf, A., Boonen, S., Bouillon, R., and Vanderschueren, D. (2001). Testosterone prevents orchidectomy-induced bone loss in estrogen receptor-α knockout mice. Biochem. Biophys. Res. Comm. 285:70–76.

263

[44] Srivastava, S., Weitzmann, M., Cenci, S., Ross, F. P., Adler, S., and Pacifici, R. (1999). Estrogen decreases TNF gene expression by blocking JNK activity and the resulting production of c-Jun and JunD. J. Clin. Invest. 104:503–513. [45] Oursler, M.J., Cortese, C., Keeting, P., Anderson, M.A., Bonde, S.K., Riggs, B.L., and Spelsberg, T.C. (1991). Modulation of transforming growth factor-beta production in normal human osteoblast-like cells by 17 beta-estradiol and parathyroid hormone. Endocrinology 129:3313–3320. [46] Ng, K.W., and Martin, T.J. (1998). Hormonal regulation of bone remodeling. Adv. Organ Biol. 5A:65–100. [47] Kousteni, S., Bellido, T., Plotkin, L.I., O’Brien, C.A., Bodenner, D.L., Han, L., Han, K., DiGregorio, G.B., Katzenellenbogen, J.A., Katzenellenbogen, B.S., Roberson, P.K., Weinstein, R.S., Jilka, R.L. and Manolagas, S.C. (2001). Nongenotropic, sex-nonspecific signaling through the estrogen or androgen receptors: Dissociation from transcriptional activity. Cell 104:719–730. [48] Watson, C.S., and Gametchu, B. (1999). Membrane-initiated steroid actions and the proteins that mediate them. Proc. Soc. Exp. Biol. Med. 220: 9–19.