Scholarly article on topic 'Hypercholesterolemia Promotes an Osteoporotic Phenotype'

Hypercholesterolemia Promotes an Osteoporotic Phenotype Academic research paper on "Health sciences"

CC BY-NC-ND
0
0
Share paper
Academic journal
The American Journal of Pathology
OECD Field of science
Keywords
{}

Abstract of research paper on Health sciences, author of scientific article — Kristine Pelton, Jaclynn Krieder, Danese Joiner, Michael R. Freeman, Steven A. Goldstein, et al.

A role for hypercholesterolemia in the development of osteoporosis has been suggested in published reports. However, few studies contain direct evidence of a role for maintenance of cholesterol homeostasis in bone health. Using isocaloric high-fat/high-cholesterol and low-fat/no-cholesterol diets in a 4-month feeding study combined with micro computed tomography analysis, we demonstrated in two different mouse strains that mice with hypercholesterolemia lose cortical and trabecular bone in the femurs and vertebrae (bone mineral density was decreased on average by ≈90 mg/mL in the cortical vertebrae in one strain) and cortical bone in the calvariae (bone mineral density was decreased on average by ≈60 mg/mL in one strain). Mechanical testing of the femurs demonstrated that loss of bone in the mice with hypercholesterolemia caused changes in the mechanical properties of the bone including loss of failure load (failure load was decreased by ≈10 N in one strain) and energy to failure. Serologic and histomorphologic analyses suggested that hypercholesterolemia promotes osteoclastogenesis. These studies support a role for hypercholesterolemia in the development of osteoporosis and provide a model with which to test intervention strategies to reduce the effects of hypercholesterolemia on bone health.

Academic research paper on topic "Hypercholesterolemia Promotes an Osteoporotic Phenotype"

The American Journal of Pathology, Vol. 181, No. 3, September 2012 Copyright © 2012 American Society for Investigative Pathology. Published by Elsevier Inc. All rights reserved. http://dx .doi .org/10.1016/j .ajpath .2012.05.034

Metabolic, Endocrine, and Genitourinary Pathobiology

Hypercholesterolemia Promotes an Osteoporotic Phenotype

Kristine Pelton,* Jaclynn Krieder,+ Danese Joiner,+ Michael R. Freeman,*®11 Steven A. Goldstein,+ and Keith R. Solomon+*11

From the Department of Orthopaedic Surgery * and the Urological Diseases Research Center? Children's Hospital Boston, Boston, Massachusetts; the Department of Orthopaedic Surgery? University of Michigan, Ann Arbor, Michigan; and the Departments of Surgery? Biological Chemistry and Molecular Pharmacology^ and Orthopaedic Surgery," Harvard Medical School, Boston, Massachusetts

A role for hypercholesterolemia in the development of osteoporosis has been suggested in published reports. However, few studies contain direct evidence of a role for maintenance of cholesterol homeostasis in bone health. Using isocaloric high-fat/high-cholesterol and low-fat/no-cholesterol diets in a 4-month feeding study combined with micro computed tomography analysis, we demonstrated in two different mouse strains that mice with hypercholesterolemia lose cortical and trabecular bone in the femurs and vertebrae (bone mineral density was decreased on average by ^90 mg/mL in the cortical vertebrae in one strain) and cortical bone in the calvariae (bone mineral density was decreased on average by ^60 mg/mL in one strain). Mechanical testing of the femurs demonstrated that loss of bone in the mice with hypercholesterolemia caused changes in the mechanical properties of the bone including loss of failure load (failure load was decreased by ^10 N in one strain) and energy to failure. Serologic and histomorphologic analyses suggested that hyper-cholesterolemia promotes osteoclastogenesis. These studies support a role for hypercholesterolemia in the development of osteoporosis and provide a model with which to test intervention strategies to reduce the effects of hypercholesterolemia on bone health. (Am J Pathol 2012, 181:928-936; http://dx. doi.org/ 10.1016/j. ajpath.2012.05.034)

Osteoporosis is the most common bone disorder, and is characterized by systemic reduction in bone mass and

changes in the bone microarchitecture that result in fragility fractures.1 The increased risk of spontaneous and traumatic bone fractures can lead to disability, often severe and permanent; placement in institutional care; and death. Osteoporotic fractures of the hip and spine result in mortality of up to 20% for 1 year after the fracture event.2

Recent evidence links the risk of developing osteoporosis to both cardiovascular disease and metabolic syn-drome3-10 (defined as central obesity with insulin resistance, and/or dyslipidemia and high blood pressure), which suggests that a common element between these conditions may also be operating to increase the risk of osteoporosis. Although not always present in individuals with cardiovascular disease or metabolic syndrome, hypercholesterolemia is a common factor in those with either disease.

Osteoblasts (cells that make bone) and osteoclasts (cells that resorb bone) synthesize cholesterol endoge-nously via the mevalonate pathway and absorb cholesterol-containing lipoproteins from the circulation. Consequently, control of cellular cholesterol content is a balance between metabolic processes intrinsic to the cell and the regulation of cholesterol distribution by the organism. Maintenance of the overall cholesterol content of cell membranes is determined by a complex set of processes including synthetic pathways in the endoplasmic reticulum, regulation of lipogenic genes by cholesterol-regulated sterol-response element binding proteins, receptor-mediated internalization, intracellular transport mechanisms, and cholesterol efflux via lipoprotein com-plexes.11-13 Despite the elaborate mechanisms for main-

Supported by a grant from the NIH (NIH CA101046 to K.R.S.) and a pilot project grant from Children's Hospital Boston (K.R.S.).

Accepted for publication May 14, 2012.

A guest editor acted as editor-in-chief of the manuscript. No person at Thomas Jefferson University was involved in the peer review process or final disposition of this article.

Supplemental material for this article can be found on http://ajp. amjpathol.org or at http://dx.doi.org/10.1016/j.ajpath.2012.05.034.

Address reprint requests to Keith R. Solomon, Ph.D., Department of Orthopaedic Surgery, Children's Hospital Boston, Enders 248, 300 Longwood Ave., Boston, MA 02115. E-mail: keith.solomon@childrens. harvard.edu.

taining correct cholesterol balance, osteoblasts and osteoclasts, like other cells, are subject to disruptions in cholesterol homeostasis as a consequence of dysregu-lation in endogenous synthetic pathways, poor cholesterol efflux, and/or systemic changes in serum cholesterol concentrations. These changes may result in altered cellular responses to specific growth factors and/or cytokines, as well as changes in the ability of bone cells to mature, differentiate, and stabilize bone.

Epidemiologic evidence indicates elevated serum cholesterol concentration as a risk factor in development of osteoporosis.5,6,14-21 Other studies have demonstrated that statin drugs, which inhibit the rate-limiting step in cholesterol synthesis,22-24 increase bone mineral density (BMD) in humans, lower fracture risk, and cause changes in bone cell (osteoblast) biochemistry in vivo and in tissue culture.25-39 When provided in typical dosages in humans, statins work almost exclusively in the liver, where they suppress cholesterol synthesis, leading to reduced serum cholesterol concentrations. Simply stated, the level of extrahepatic statin is too small and for too short a time for the purported direct extrahepatic tissue effects of statins to occur.22-24,40-50 Consequently, if statins alter risk of developing osteoporosis, this phenomenon likely stems from their effects on circulating cholesterol concentrations, and not from direct effects on osteoblasts and osteoclasts.

Because of the suggested links between cardiovascular disease, metabolic syndrome, and osteoporosis,3-10 previous studies that have suggested an effect of elevated cholesterol concentration and statins on BMD and fracture risk,5,6,14-21,25-39 and our understanding of the roles of cholesterol in bone cell biology, we hypothesized that altering the level of cholesterol in membranes of osteoblasts and/or osteoclasts, as a consequence of elevated serum cholesterol concentrations, will contribute to development of osteoporosis.

In a previous study, an atherogenic diet (high-fat/ high-cholesterol [HFHC] with sodium cholate) fed to mice led to significantly reduced femoral mineral content (43% lower) and mineral density (15% lower) when compared with a low-fat/no-cholesterol (LFNC) diet fed to control animals.51 A more recent study showed that rats fed a HFHC diet also demonstrated loss of femoral BMD when compared with rats fed normal chow.52 Neither of these studies used isocaloric diets, and, therefore, did not rule out energy effects, resulting in inconclusive interpretation of a specific role for cholesterol in reducing BMD.

The objective of the present study was to test the effect of hypercholesterolemia on bone integrity. Using a novel diet scheme that included isocaloric diets lacking sodium cholate, we demonstrated that hypercholesterolemia is associated with alterations in BMD, bone volume fraction (BVF; bone volume/total volume), number of trabeculae, trabecular spacing, and other physical measures, accompanied by changes in the mechanical properties of the bone and an increase in bone resident osteoclasts.

Materials and Methods

Mice and Diets

All animal experiments were approved by the Institutional Animal Care and Use Committee of Children's Hospital Boston (protocol A07-06-084). Five-week-old male C57BL/6 mice (The Jackson Laboratory, Bar Harbor, ME) and C.B-17/Icr-SCID/Sed-Prkdcscid mice (Massachusetts General Hospital, Boston, MA) were fed an LFNC diet (catalog No. D12102; Research Diets, Inc., New Brunswick, NJ) for 2 weeks, after which blood was drawn from the saphenous tail vein, and the serum cholesterol concentration was determined using the Infinity Cholesterol Liquid Stable Reagent (Thermo Electron Corp., Wal-tham, MA). The mice were then divided into groups that received either the HFHC diet (catalog No. D12108; Research Diets) or the LFNC diet for 4 months. Although these diets differ in composition, they are used isocalo-rically, ie, by feeding mice the same number of calories in each diet. Terminal blood samples were obtained via cardiac puncture for serologic assay (pyridinoline crosslinks assay; Quidel Corp., San Diego, CA) and osteocalcin assay (Biomedical Technologies, Inc., Stoughton, MA). Assays were performed according to the manufacturers' instructions. Triglyceride, insulin, and bilirubin concentrations, and other liver function tests were performed at the Yale University Mouse Phenotyping Center (New Haven, CT), and androgen level tests were performed at the Hormone Assay Core, Vanderbilt University (Nashville, TN).

Histomorphologic Analyses

Femurs were fixed in 10% neutral buffered formalin and decalcified for 2 weeks at 4°C using 10% EDTA (pH 7.4). The bones were dehydrated in graded ethanol (50%, 70%, 90%, and 95%) and twice in 100% ethanol for 1 hour each, cleared twice in xylene for 1 hour each, and embedded in paraffin. Five-micrometer sections were cut, deparaffinized, and stained using the TRAP kit (Acid Phosphatase, Leukocyte kit; Sigma Corp., St. Louis, MO) according to the manufacturer's instructions, and were counterstained with hematoxylin. TRAP-stained cells with three nuclei or more were counted as osteoclasts for quantification.

Bones, Micro Computed Tomography, and Mechanical Testing

The calvaria, right femurs, and L4 vertebrae were removed immediately after the animals were euthanized, and were washed in lactated Ringer's solution. All specimens were scanned and then reconstructed onto 18-^m voxels using a cone beam micro-computed tomography (microCT) system (GE Healthcare BioSciences Corp., Piscataway, NJ) to assess geometric and morphologic properties. A single threshold was applied to the three-dimensional dataset, as determined from evaluation of the calibrated density histograms. Cortical mineralization and geometric analyses were performed over a 3-mm mid-diaphyseal segment in the femurs, and trabecular

morphologic analysis over a standardized volume of bone segmented using a spline algorithm, starting at the distal growth plate and moving proximally to 20% of the total femur length. The elements of the spline were interpolated to generate a three-dimensional region of interest. Data from each femur were averaged over the entire region of interest to provide representative values for the entire three-dimensional mid-diaphysis. Outcome measures included cross-sectional area, cortical thickness, and bending moment of inertia. In addition to the femoral properties, two standardized volumes of trabecular bone were segmented from the proximal and distal ends of vertebral bodies. Volumes were determined by identifying the largest volume of trabecular bone possible by digitizing each section to define a series of splines that interpolated the volume of interest. The trabecular bone in both the femoral metaph-yses and the vertebrae were evaluated to determine BVF, trabecular thickness, degree of anisotropy, number of tra-beculae, and trabecular spacing. The analyses were performed using the advanced analysis programs and MicroView software suites in the GE MicroCT systems (GE Healthcare Biosciences). Morphologic analyses of the cal-variae were performed on segments created using a region of interest growing tool. All of the algorithms have been validated in the Orthopaedic Research Laboratories at the University of Michigan (Ann Arbor).

Biomechanical Properties

Before biomechanical testing, all specimens were maintained frozen at -20°C. Testing was performed at room temperature, and the specimens were maintained moist using Ringer's solution. Whole-bone mechanical properties of the femurs to failure were performed in four-point bending using a servohydraulic testing machine (858 Mini Bionix II; MTS Systems Corp., Eden Prairie, MN) at a constant displacement rate of 0.5 mm/second. Femors were loaded in the anteroposterior direction so that the posterior side of the bone was in tension and the anterior side was in compression. All four loading points were placed in contact with the bone by adjusting the height of the upper two points independently. Regions loaded in four-point bending corresponded to those measured using microCT. Displacement was monitored using a linear variable differential transducer (Lucas Schaevitz 050 MHR; Measurement Specialties, Inc., Hampton, VA), and load data were collected using a load cell (Sensotec, Inc., Columbus, OH) in series with the actuator. Load and displacement data were sampled at 2000 Hz using the TestStar IIs system (version 2.4; MTS Systems). Load-displacement curves were then analyzed for whole-bone yield load, ultimate load, stiffness, failure energy, and displacement ratio (ultimate displacement:yield displacement) using MATLAB software (version 6.5; The MathWorks, Inc., Natick, MA). The predicted tissue modules and yield strength were calculated using standard mechanics of materials equations and geometric properties calculated during microCT analysis. These equations predict inherent material properties of the extracellular matrix, assuming a bending mode of failure.

Statistical Analysis

Statistical significance was determined using Student's t-test. P < .05 was considered significant.

Results

To understand how hypercholesterolemia affects various measures related to bone quality, we used two cohorts of mice fed isocaloric HFHC and LFNC diets, respectively, for 4 months. Previous testing of these diets indicated that to raise cholesterol concentrations in mice, a diet required both additional fat and cholesterol, ie, HFHC diet,53 whereas the LFNC diet is well-defined and is composed essentially of normal mouse chow. Cohorts of mice fed these diets did not differ by weight even after 4 months (mean ± SD, 37.66 ± 3.68 g LFNC versus 38.61 ± 5.30 g HFHC for C57BL/6 mice, and 31.5 ± 2.93 g LFNC versus 31.31 ± 2.95 g HFHC for SCID mice) (Figure 1A; see also Supplemental Table S1 at http://ajp. amjpathol.org), but demonstrated significant differences in serum cholesterol concentrations (122.23 ± 22.53

Figure 1. Effects of diet on weight and blood chemistry. Mice (14 to 16 per group) fed an isocaloric high-fat/high-cholesterol diet (black bars) or a low-fat/no-cholesterol diet (white bars) for 4 months were weighed (A), and serum cholesterol (B), triglyceride (C), and testosterone (D) concentrations were determined. A: Body mass. Data are plotted as mean ± SD versus diet group. There is no statistically significant difference between the 2 groups in either strain. B—D: Serologic measures. Mice were bled via cardiac puncture, and the serum was subjected to various tests. B: Serum cholesterol concentration was measured in the collected serum via Infinity colorimetric assay. Data are plotted as mean ± SD cholesterol concentration versus group. C: Triglyceride concentration was measured as described in Materials and Methods. Data are plotted as triglyceride concentration versus group. D: Testosterone concentration was measured as described in Materials and Methods. Data are plotted as mean ± SE testosterone concentration versus diet group. In all cases, the Student's t-test was used to calculate significance. Data were considered significant at P < 0.05. *P < 0.05, **P < 0.01, and ***P < 0.005.

A 13501

4 1300 -I 1250 -- -1200-| ^ 1150 -

i £ 1100 -

I 1050"

I® 1000 -

F 1350 3. 1300 -

I 1250 1

Q - 1200

1150 1100 -

0.35 f 0.3

0.2 0.15

f 13 " £ 12 -

8 1.1 -<

® 1 05 -

i- 0.25

Figure 2. Effects of diet on femoral cortical bone. Mice (14 to 16 per group) fed an isocaloric high-fat/high-cholesterol diet (black bars) or a low-fat/no-cholesterol diet (white bars) for 4 months were sacrificed, and their femors were removed, cleaned, and analyzed using microCT. Measurements included tissue mineral density (TMD) (A), thickness (B), cortical surface area (CSA)(C), moment of inertia (Iyy) (D), and bone mineral density (BMD). Data are given as mean ± SD versus group. In all cases, the Student's t-test was used to calculate significance. Data were considered significant at P < 0.05. *P < 0.05, **P < 0.01, and ***P < 0.005.

mg/dL LFNC versus 152.37 ± 36.10 mg/dL HFHC for C57BL/6 mice, and 158.11 ± 22.76 mg/dL LFNC versus 181.53 ± 24.82 mg/dL HFHC for SCID mice) (Figure 1B; see also Supplemental Table S1 at http://ajp.amjpathol. org). Mice fed the HFHC diet had hypercholesterolemia, whereas the other group had normal cholesterol concentrations. The HFHC diet did raise triglyceride concentrations significantly in SCID mice, but not in C57BL/6 mice (Figure 1C). Liver toxicity, as determined by liver function test results, was never detected, and insulin levels were not significantly different between the diet cohorts, although they trended higher in the C57BL/6 HFHC diet cohort (data not shown). There were no significant differences in androgen levels (Figure 1D).

To determine whether elevated serum cholesterol concentrations affected bone health, the calvariae, femors, and vertebral bodies of the mice were removed, cleaned, and subjected to microCT.

The hypercholesterolemic mice demonstrated significant changes in their femors when compared with mice fed a normocholesterolemic diet, ie, significantly reduced tissue mineral density (TMD) in the cortical bone regions (1163.5 ± 20.4 mg/mL LFNC versus 1147.5 ± 16.1 mg/mL HFHC) in the C57BL/6 mice, which also trended similarly but did not reach statistical significance in the SCID mice (1273.0 ± 25.7 mg/mL LFNC versus 1249.8 ± 36.9 mg/mL HFHC), and significantly thinner diaphyseal cortical bone with a smaller cortical surface area in both strains (Figure 2, A-C; see also Supplemental Table S1 at http://ajp.amjpathol.org). We also measured significant differences in the moment of inertia (Iyy) and BMD in LFNC versus HFHC diet cohorts of C57BL/6 mice (0.27 ± 0.03 mm4 LFNC versus 0.21 ± 0.04 mm4 HFHC) (Figure

2, D and E; see also Supplemental Table S1 at http:// ajp.amjpathol.org). These values trended similarly between the diets in the SCID mice, but did not reach statistical significance. The trabecular bone in the femors of the hypercholesterolemic cohorts also demonstrated statistically significant changes as compared with normocholesterolemic mice, with a smaller BVF (0.43 ± 0.06 LFNC versus 0.27 ± 0.05 HFHC for the C57BL/6 mice, and 0.46 ± 0.10 LFNC versus 0.37 ± 0.08 HFHC for the SCID mice), larger trabecular spacing (0.10 ± 0.02 mm LFNC versus 0.16 ± 0.03 mm HFHC for the C57BL/6 mice, and 0.09 ± 0.02 mm LFNC versus 0.12 ± 0.07 mm HFHC for the SCID mice), fewer trabeculae in both strains (5.59 ± 0.42 LFNC versus 4.51 ± 0.49 HFHC for the C57BL/6 mice, and 6.21 ± 0.53 LFNC versus 5.53 ± 1.0 HFHC for the SCID mice) (Figure 3, A-C; see also Supplemental Table S1 at http://ajp.amjpathol.org), and significantly less trabecular thickness and TMD in the LFNC diet cohort (691.05 ± 25.43 mg/mL) versus the HFHC diet cohort (717.37 ± 15.16 mg/mL) of the C57BL/6 strain. These values trended similarly in the SCID mice (Figure 3, D and E; see also Supplemental Table S1 at http://ajp.amjpathol.org).

Calvariae from mice fed the HFHC diet versus those fed the LFNC diet demonstrated significant reductions in BMD (763.5 ± 55.12 mg/mL LFNC versus 700.98 ± 45.53 mg/mL HFHC in the C57BL/6 mice, and 713.16 ± 69.13 mg/mL LFNC versus 664.5 ± 42.6 mg/mL HFHC in the SCID mice) and mean thickness in both strains (0.21 ± 0.02 mm LFNC versus 0.20 ± 0.02 mm HFHC in the C57BL/6 mice, and 0.33 ± 0.05 mm LFNC versus 0.29 ± 0.04 mm HFHC in the SCID mice) (Figure 4,A and B; see also Supplemental Table S1 at http://ajp.

i- 750 | _"*>'

E 0.56 T

ic"]nhj_rV" }"' * i r Hrhij-. !<H rf T r-h'""

in* r ■ I :: li f Hi Hi

*::___■ ■ hJULX ___■___■ h__

" RTÜIT P.S7RI i'fi 1 u 1 SCID C57BL.B 1 * SCID Cb/BL.6 1 >~ " SCID C57BDG

SCID C57BL/6

SCID C57BL/6 1

Figure 3. Effects of diet on femoral trabecular bone. Mice (14 to 16 per group) fed an isocaloric high-fat/high-cholesterol diet (black bars) or a low-fat/no-cholesterol diet (white bars) for 4 months were sacrificed, and their femors were removed, cleaned, and analyzed using microCT. Measurements included bone volume fraction (BVF) [bone volume/total volume (BV/TV)] (A), trabecular spacing (distance between trabeculae) (B), number of trabeculae (C), trabecular thickness (D), and tissue mineral density (TMD) (E). Data are given as mean ± SD versus group. In all cases, the Student's t-test was used to calculate significance. Data were considered significant at P < 0.05. *P ^ 0.05, **P ^ 0.01, and ***P < 0.005.

с 800

ä 8" 700

J ё, боо

1 500 400

Г+1* I

■ I Щ\ I ■

CID C57BL'6 1 'SCID C57BL6

iHt НМЛ

CID C&7BL6 1 w SCID CbiHL.S

Figure 4. Effects of diet on the calvariae. Mice (14 to 16 per group) fed an isocaloric high-fat/high-cholesterol diet (black bars) or a low-fat/no-cholesterol diet (white bars) for 4 months were sacrificed, and their calvariae were removed, cleaned, and analyzed using microCT. Measurements included bone mineral density (BMD) (A), thickness (B), tissue mineral density (TMD) (C), bone mineral content (BMC) (D), and tissue mineral content (TMC) (E). Data are given as mean ± SD versus group. In all cases, the Student's t-test was used to calculate significance. Data were considered significant at P < 0.05. *P < 0.05, ***P < 0.005.

amjpathol.org). The calvariae from the C57BL/6 mice fed the LFNC diet also had significantly higher TMD (826.6 ± 18.21 mg/mL versus 802.44 ± 17.65 mg/mL), bone mineral content (1.08 ± 0.13 mg versus 0.97 ± 0.11 mg), and tissue mineral content (0.98 ± 0.09 mg versus 0.82 ± 0.09 mg) when compared with mice fed the HFHC diet, and these measures also trended similarly in the SCID mouse cohort (Figure 4, C—E; see also Supplemental Table S1 at http://ajp.amjpathol.org).

The trabecular bone of the vertebral bodies from the C57BL/6 and SCID HFHC cohorts demonstrated significantly fewer trabeculae (7.71 ± 0.19 LFNC versus 7.21 ± 0.46 HFHC for the C57BL/6 mice, and 6.90 ± 0.78 LFNC versus 5.96 ± 0.79 HFHC for the SCID mice) and larger trabecular spacing (0.05 ± 0.01 mm LFNC versus 0.08 ± 0.01 mm HFHC for the C57BL/6 mice, and 0.09 ± 0.02 mm LFNC versus 0.11 ± 0.03 mm HFHC for the SCID mice) than in the LFNC cohorts (Figure 5,A and B; see also Supplemental Table S1 at http://ajp.amjpathol.org). In addition, there were also significant differences between the C57BL/6 HFHC diet mice and LFNC diet mice in both trabecular thickness (0.08 ± 0.01 mm LFNC versus 0.06 ± 0.01 mm HFHC) and BVF (0.61 ± 0.05 LFNC versus 0.45 ± 0.06 HFHC) (Figure 5, C and D; see also Supplemental Table S1 at http://ajp.amjpathol.org).

In contrast to the overall similarity of the effects of diet on the C57BL/6 and SCID mice, the effects of diet on cortical vertebral bone varied between the two strains. In the C57BL/6 strain, bone mineral content (0.52 ± 0.08 mg LFNC versus 0.45 ± 0.07 mg HFHC), BMD (479.62 ± 25.59 mg/mL LFNC versus 389.53 ± 41.74 mg/mL HFHC), TMC (0.42 ± 0.07 mg LFNC versus 0.31 ± 0.06 mg HFHC), and TMD (621.43 ± 17.22 mg/mL LFNC versus 597.03 ± 19.46 mg/mL HFHC) were all significantly reduced, whereas in the SCID mice, there were

no significant differences in any of these measurements between the LFNC and HFHC groups, although bone mineral content, BMD, and TMC measurements all trended similarly to those of the C57BL/6 strain (Figure 6, A-D; see also Supplemental Table S1 at http://ajp.amjpathol.org).

Because the HFHC versus LFNC diet was associated with significant changes in both mineral content and microarchitecture of the femors, we next determined whether these changes resulted in alterations in the mechanical properties of the bone (Figure 7; see also Supplemental Table S1 at http://ajp.amjpathol.org). The femors from the C57BL/6 HFHC cohort were less stiff (323.08 ± 35.36 N/mm LFNC versus 244.0 ± 42.43 N/mm HFHC), had significantly smaller ultimate and failure loads (38.41 ± 5.71 N LFNC versus 27.94 ± 4.24 N HFHC), and less energy to failure (4.28 ± 1.96 N/mm LFNC versus 2.76 ± 1.29 N/mm HFHC) than did the femors from the C57BL/6 LFNC group. The mechanical properties of the femors did not differ in the SCID HFHC group when compared with the SCID LFNC group (Figure 7, A-D; see also Supplemental Table S1 at http://ajp. amjpathol.org).

Serologic testing indicated that there were no differences in osteocalcin levels between the hypercholester-olemic HFHC cohort and the normocholesterolemic LFNC cohort, which suggested that altered osteoblast activity was unlikely to explain the observed changes in the bone (data not shown). However, we did observe differences in the level of pyridinoline cross-links found in collagen I fragments, a measure of osteoclast activity (Figure 8A). This suggested that increased osteoclast activity contributes to the osteopenic phenotype of the hypercholesterolemic mice. Consistent with an effect of the HFHC diet on osteoclastogenesis, histomorphometric

Figure 5. Effects of diet on the trabecular bone of vertebrae. Mice (14 to 16 per group) fed an isocaloric high-fat/high-cholesterol diet (black bars) or a low-fat/no-cholesterol diet (white bars) for 4 months were sacrificed, and L4 was removed, cleaned, and analyzed using microCT. Measurements included number of trabeculae (A), trabecular spacing distance between trabeculae (B), trabecular thickness (C), and bone volume fraction (BVF) [bone volume/total volume (BV/TV)] (D). Data are given as mean ± SD versus group. In all cases, the Student's t-test was used to calculate significance. Data were considered significant at P < 0.05. *P < 0.05, ***P < 0.005.

Figure 6. Effects of diet on the cortical bone of the vertebrae. Mice (14 to 16 per group) fed an isocaloric high-fat/high-cholesterol diet (black bars) or a low-fat/no-cholesterol diet (white bars) for 4 months were sacrificed, and L4 was removed, cleaned, and analyzed using MicroCT. Measurements included bone mineral content (BMC) (A), bone mineral density (BMD) (B), tissue mineral content (TMC) (C), and tissue mineral density (TMD) (D). Data are given as mean ± SD versus group. In all cases, the Student's t-test was used to calculate significance. Data were considered significant at P < 0.05. *P ^ 0.05, ***P ^ 0.005.

analysis revealed a significant increase in the number of osteoclasts present in the bones of the hypercholesterolemic versus the normocholesterolemic groups (Figure 8, B and C).

Discussion

In the present study, we demonstrated that diet-induced hypercholesterolemia in mice is associated with a reduction in bone quality measures that resemble the bone quality in humans with osteoporosis.

We concluded that there is likely a direct influence of hypercholesterolemia on bone health due to the following: i) mice were fed isocaloric diets, and, thus, there is no effect of additional calories in this model, even though the diets differed in their levels of fat and cholesterol; ii) the mice did not gain weight with the HFHC diet versus the LFNC diet, and, thus, any changes that occurred in bone are not reflective of differences in animal weight (Figure 1A); iii) multiple parameters of bone quality were altered by diet, and, thus, the changes detected are likely to originate in a systematic loss of bone homeostasis and are not due to an isolated event (Figures 2 to 6); iv) femors, calvariae, and vertebrae were all affected, indicating that the source of altered bone quality is systemic (Figures 2 to 6); v) the C57BL/6 HFHC cohort exhibited reduced mechanical properties (Figure 7); and vi) diet did not influence androgen levels (Figure 1D).

Mice received the HFHC or LFNC diets for 4 months, which is approximately 16% to 25% of their lifespan. This should be sufficient time to affect cortical and trabecular bone, should either diet have any effect.

Testing of the femors of the C57BL/6 mice demonstrated significant loss of their mechanical properties, including substantial decreases in stiffness, failure load, ultimate load, and energy to failure in the HFHC cohort,

which suggests that the loss of bone mineral and thinning of the trabeculae were substantial enough to significantly weaken the bone. The loss of both physical and mechanical properties implies a resemblance to osteoporosis in humans. The similarities to osteoporosis in humans include thinning of the diaphyseal cortical bone, reduced BVF, thinning trabeculae, loss of trabeculae, increased osteoclast numbers, and reductions in failure load and energy to failure. Given the loss of femoral stiffness, which is likely a reflection of the thinner bone, it is expected that the bones would be somewhat less brittle; rather, we found that the bones were more brittle. Thus, the loss of energy to failure induced by the HFHC diet likely reflects a change in the material properties of the bone.

Although in many respects the physical properties noted in the C57BL/6 mice fed the HFHC versus the LFNC diet were also present in the SCID mice, for the most part, they were less pronounced, and most mechanical measures were not affected by diet in the SCID strain.

Although not all of the factors that contribute to native BMD are known, a combination of environmental and genetic factors is most probably involved. The effect of serum cholesterol concentrations on development of an osteoporotic phenotype is likely, to some degree, also dependent on native BMD (ie, the BMD before cholesterol manipulation). On the basis of previous analy-ses,54-56 and our own data (Figure 2E) for the native femoral BMD of these strains, C57BL/6 mice have a lower BMD than does the SCID strain, which shares a background with Balb/C mice. This result suggests that the native BMD may influence the degree to which hypercholesterolemia affects bone. A study by Ishimori et al57 suggested that the differential effect of atherogenic diets in C57BL/6 and 129 strains, which displayed different

Figure 7. Mechanical properties of femors. Mice (14 to 16 per group) fed an isocaloric high-fat/high-cholesterol diet (black bars) or a low-fat/no-cholesterol diet (white bars) for 4 months were sacrificed, and their femors were removed, cleaned, and analyzed via mechanical testing. Whole-bone mechanical properties of the femors to failure were performed in 4-point bending using a servohydraulic testing machine. Load-displacement curves were then analyzed for whole-bone yield load, ultimate load, stiffness, failure energy, and displacement ratio (ultimate displacement:yield displacement), and other measures using MATLAB software. The difference between "start" (centroid coordinate position) and "end" positions was used to calculate stiffness (A), ultimate load (B), failure load (C), and energy (D). Data are given as mean ± SD. In all cases, the Student's t-test was used to calculate significance. Data were considered significant at P < 0.05. *P ^ 0.05, ***P ^ 0.005.

Figure 8. Bone loss in mice with hypercholesterolemia is linked to increased osteoclastic activity and osteoclasts. Mice fed an isocaloric high-fat/high-cholesterol diet (black bars) or a low-fat/no-cholesterol diet (white bars) for 4 months were sacrificed, and serum pyridinoline (PYD) concentration in the murine blood was determined via enzyme-linked immunosorbent assay (A), the femors were analyzed at histology, and the osteoclasts were counted (B). A: Data are plotted as mean ± SD PYD concentration versus group (n = 8 or 9 mice per group). B: Five-micrometer sections of femors were cut and stained using the Acid Phosphatase, Leukocyte (TRAP) kit, and the total number of osteoclasts in 5 X20 fields were counted. Data are given as mean ± SD number of osteoclasts versus group (n = 7 bones per group). In all cases, the Student's t-test was used to calculate significance. Data were considered significant at P < 0.05. C: Histomorphometric images demonstrate osteoclasts (arrows) in trabecular regions of the femors in C57BL/6 mice. These are representative images of the X20 fields used in the osteoclast quantification at X40 magnification. Left: Most panels are enlargements of original X60 images demonstrating osteoclasts used for quantification with characteristic TRAP staining (light pink) and multiple nuclei (purple). *P < 0.05.

total and vertebral BMD after being fed an atherogenic diet, are linked to four quantitative trait loci: Bmd9, Bmd20, Bmd21, and Bmd22, which collectively accounted for 21.6% of total BMD and 17.3% of vertebral BMD of the F2 population variances. It is likely that genetic variations, which influence BMD, may have a role in the effect of diet on the development of osteopenia in mice.

Another potential reason for the disparate response between C57BL/6 and SCID mice is that SCID mice are profoundly immunodeficient owing to the inability of these mice to perform VDJ recombination, rendering them without functional T and B cells. Although certain inflammatory responses from natural killer cells and macrophages remain, most inflammation is impaired in these animals. In a separate study (accepted for publication, PLOS One), we measured particular inflammatory responses in the animals as related to diet, and found that the HFHC diet did contribute to inflammation in the C57BL/6 mice, which was absent in the SCID mice, consistent with the known effects of hypercholesterolemia on inflammation.

Proinflammatory cytokines have a role in osteoclasto-genesis and bone resorption, and are increased during menopause, which suggests that inflammation and osteoporosis are linked.58 Our results are consistent with a role for inflammation in response to hypercholesterolemia as a mediator of bone loss, but also suggest that noninflammatory pathways likely mediate some of the effect of diet on development of osteoporosis because the SCID mice did lose bone.

Our data (Figure 8) indicate that hypercholesterolemia increases the number and activity of osteoclasts within bone. Although increased osteoclast activity could explain the reduced BMD in both strains, it is notable that statistically significant effects of hypercholesterolemia were observed only for serum pyridinoline concentration (a measure of osteoclast activity) in SCID mice, whereas more extensive effects of diet on bone were evident in C57BL/6 mice. This is likely a reflection of the timing of the assay, which was necessarily performed as an experimental end point (ie, after 4 months on the diet) at which the loss of bone in C57BL/6 mice had already peaked.

Our data suggest that hypercholesterolemia promotes osteoclastogenesis, consistent with unpublished obser-

vations from our laboratory that increasing the exogenous cholesterol available to macrophages increases the number and size of the polykaryon (data not shown). It is possible that hypercholesterolemia affects the membrane composition of both osteoblasts and osteoclasts, which we hypothesized would in turn affect their ability to respond to environmental cues. We have published59,60 and unpublished evidence that altering membrane cholesterol levels in osteoblasts affects their responses to growth factors, but we have no evidence that this is occurring in vivo.

A number of studies have suggested that high serum cholesterol concentrations, per se, are associated with low BMD and osteoporosis in humans.5,14-21 For example, Broulik and Kapitola14 measured higher cholesterol concentrations in patients with osteoporosis with and without vertebral fractures, respectively, compared with age-matched patients without osteoporosis. Poli et al21 reported that women with plasma low-density lipoprotein (LDL) cholesterol concentrations >160 mg/dL had a more than twofold probability of having osteopenia than did women with lower LDL cholesterol concentrations (47.9% versus 21.2%, respectively). Of note, some studies have not found an association between high cholesterol concentration (or LDL concentration) and osteopo-rosis.61,62 For example, Samelson et al,61 using data from the Framingham Study, examined the relationship between BMD (femoral head) and total cholesterol concentration, but found no correlation between high cholesterol concentrations and osteoporosis. Of importance, this study did not use vertebral measures, LDL cholesterol concentrations were unknown, and, perhaps most critical, the subjects were enrolled in 1948, yet their BMD was not measured until 1988-1989,61 thereby effectively skewing the data in favor of the null hypothesis because many subjects with high cholesterol concentrations and low BMD were likely deceased by 1989. Assessing the complex relationships between diet, cholesterol concentrations, and osteoporosis in a human population is problematic because all important variables cannot be controlled and data are difficult to obtain.

In the present study, we created a hypercholester-olemic animal model with which we can determine how hypercholesterolemia affects bone health. By controlling calories, animal weight, and other variables, we were

able to accurately measure the effect of hypercholesterolemia In mice. It is noteworthy that a previous analysis of diet and bone in mice determined that an atherogenic diet that raised cholesterol concentrations was correlated with loss of BMD. The diet we used in our study also raised serum cholesterol concentrations significantly, but was not atherogenic. Mice do not naturally develop atherosclerotic lesions in the aorta similar to those found in humans.63 An atherogenic diet in mice requires the addition of sodium cholate, a bile acid.63 Sodium cholate is not required to raise cholesterol concentrations, but does cause liver toxicity. Thus, hypercholesterolemic non-atherogenic diets, such as those used in the present study, are preferred when determining the relationship between hypercholesterolemia and disease.

Although the type of analysis performed in the present study will be helpful in hypothesis testing and in discovery, this model is not without caveats; measuring osteoporosis in mice will not accurately predict which bones and bone regions will be most affected by hypercholes-terolemia in humans because weight distribution in mice is not the same.

In conclusion, a hypercholesterolemic non-athero-genic diet contributes to development of an osteoporotic bone phenotype in mice, including an increase in oste-oclasts, loss of trabeculae, thinning of trabeculae and cortex, and reductions in failure load and energy to failure. This model of osteoporosis may provide an opportunity to critically evaluate various interventional strategies to counteract the changes induced by a hypercho-lesterolemic diet. Our data suggest that the correlation between hypercholesterolemia and osteoporosis may be a manifestation of failure to maintain cholesterol homeostasis and an increase in osteoclast activity. Given the increase in hyperlipidemia, metabolic syndrome, and obesity in the United States and around the world, we anticipate an increase in diet-induced osteoporosis, making our novel model particularly timely.

References

1. Raisz LG: Clinical practice. Screening for osteoporosis. N Engl J Med 2005, 353:164-171

2. Rachner TD, Khosla S, Hofbauer LC: Osteoporosis: now and the future. Lancet 2011, 377:1276-1287

3. Migliaccio S, Greco EA, Fornari R, Donini LM, Lenzi A: Is obesity in women protective against osteoporosis? Diabetes Metab Syndr Obes 2011, 4:273-282

4. Yesil Y, Ulger Z, Halil M, Halacli B, Yavuz BB, Yesil NK, Kuyumcu ME, Cankurtaran M, Ariogul S: Coexistence of osteoporosis (OP) and coronary artery disease (CAD) in the elderly: it is not just a by chance event. Arch Gerontol Geriatr 2011, 54:473-476

5. Burnett JR, Vasikaran SD: Cardiovascular disease and osteoporosis: is there a link between lipids and bone? Ann Clin Biochem 2002, 39:203-210

6. Masse PG, Tranchant CC, Dosy J, Donovan SM: Coexistence of osteoporosis and cardiovascular disease risk factors in apparently healthy, untreated postmenopausal women. Int J Vitam Nutr Res 2005, 75:97-106

7. McFarlane SI, Muniyappa R, Shin JJ, Bahtiyar G, Sowers JR: Osteoporosis and cardiovascular disease: brittle bones and boned arteries, is there a link? Endocrine 2004, 23:1-10

8. Lidfeldt J, Holmdahl L, Samsioe G, Nerbrand C, Nyberg P, Schersten B, Agardh CD: The influence of hormonal status and features of the

metabolic syndrome on bone density: a population-based study of Swedish women aged 50 to 59 years. The women's health in the Lund area study. Metabolism 2002, 51:267-270 9. von Muhlen D, Safii S, Jassal SK, Svartberg J, Barrett-Connor E: Associations between the metabolic syndrome and bone health in older men and women: the Rancho Bernardo Study. Osteoporos Int 2007, 18:1337-1344

10. Yaturu S, Humphrey S, Landry C, Jain SK: Decreased bone mineral density in men with metabolic syndrome alone and with type 2 diabetes. Med Sci Monit 2009, 15:CR5-9

11. Maxfield FR, Wustner D: Intracellular cholesterol transport. J Clin Invest 2002, 110:891-898

12. SoccioRE, BreslowJL: Intracellular cholesterol transport. Arterioscler Thromb Vasc Biol 2004, 24:1150-1160

13. Weber LW, Boll M, Stampfl A: Maintaining cholesterol homeostasis: sterol regulatory element-binding proteins. World J Gastroenterol 2004, 10:3081-3087

14. Broulik PD, Kapitola J: Interrelations between body weight, cigarette smoking and spine mineral density in osteoporotic Czech women. Endocr Regul 1993, 27:57-60

15. Orozco P: Atherogenic lipid profile and elevated lipoprotein(a) are associated with lower bone mineral density in early postmenopausal overweight women. Eur J Epidemiol 2004, 19:1105-1112

16. Parhami F: Possible role of oxidized lipids in osteoporosis: could hyperlipidemia be a risk factor? Prostaglandins Leukot Essent Fatty Acids 2003, 68:373-378

17. Parhami F, Garfinkel A, Demer LL: Role of lipids in osteoporosis. Arterioscler Thromb Vasc Biol 2000, 20:2346-2348

18. Yamaguchi T, Sugimoto T, Yano S, Yamauchi M, Sowa H, Chen Q, Chihara K: Plasma lipids and osteoporosis in postmenopausal women. Endocr J 2002, 49:211-217

19. Nuzzo V, de Milita AM, Ferraro T, Monaco A, Florio E, Miano P, Montemarano E, Zuccoli A, de Terlizzi F: Analysis of skeletal status by quantitative ultrasonometry in a cohort of postmenopausal women with high blood cholesterol without documented osteoporosis. Ultrasound Med Biol 2009, 35:717-722

20. Tarakida A, Iino K, Abe K, Taniguchi R, Higuchi T, Mizunuma H, Nakaji S: Hypercholesterolemia accelerates bone loss in postmenopausal women. Climacteric 2011, 14:105-111

21. Poli A, Bruschi F, Cesana B, Rossi M, Paoletti R, Crosignani PG: Plasma low-density lipoprotein cholesterol and bone mass densitometry in postmenopausal women. Obstet Gynecol 2003, 102:922-926

22. Desager JP, Horsmans Y: Clinical pharmacokinetics of 3-hydroxy-3-methylglutaryl-coenzyme A reductase inhibitors. Clin Pharmacokinet 1996, 31:348-371

23. Sirtori CR: Tissue selectivity of hydroxymethylglutaryl coenzyme A (HMG CoA) reductase inhibitors. Pharmacol Ther 1993, 60:431-459

24. Solomon KR, Freeman MR: Do the cholesterol-lowering properties of statins affect cancer risk? Trends Endocrinol Metab 2008, 19:113121

25. Bauer DC, Mundy GR, Jamal SA, Black DM, Cauley JA, Harris F, Duong T, Cummings, S.R.: Statin Use, Bone Mass and Fracture: An Analysis of Two Protective Studies. Edited by Drezner MK. St. Louis, MO: Blackwell Science, Inc.; 1999:S179

26. Bocan TM: Pleiotropic effects of HMG-CoA reductase inhibitors. Curr Opin Investig Drugs 2002, 3:1312-1317

27. Chan KA, Andrade SE, Boles M, Buist DS, Chase GA, Donahue JG, Goodman MJ, Gurwitz JH, LaCroix AZ, Platt R: Inhibitors of hy-droxymethylglutaryl-coenzyme A reductase and risk of fracture among older women. Lancet 2000, 355:2185-2188

28. Lupattelli G, Scarponi AM, Vaudo G, Siepi D, Roscini AR, Gemelli F, Pirro M, Latini RA, Sinzinger H, Marchesi S, Mannarino E: Simvastatin increases bone mineral density in hypercholesterolemic postmenopausal women. Metabolism 2004, 53:744-748

29. Meier CR, Schlienger RG, Kraenzlin ME, Schlegel B, Jick H: HMG-CoA reductase inhibitors and the risk of fractures [see comments]. JAMA 2000, 283:3205-3210

30. Mundy G, Garrett R, Harris S, Chan J, Chen D, Rossini G, Boyce B, Zhao M, Gutierrez G: Stimulation of bone formation in vitro and in rodents by statins. Science 1999, 286:1946-1949

31. Mundy GR: Statins and their potential for osteoporosis. Bone 2001, 29:495-497

32. Rogers MJ: Statins: lower lipids and better bones? Nat Med 2000, 6:21-23

33. Schoofs MW, Sturkenboom MC, van der Klift M, Hofman A, Pols HA, Stricker BH: HMG-CoA reductase inhibitors and the risk of vertebral fracture. J Bone Miner Res 2004, 19:1525-1530

34. Waldman A, Kritharides L: The pleiotropic effects of HMG-CoA reductase inhibitors: their role in osteoporosis and dementia. Drugs 2003, 63:139-152

35. Wang PS, Solomon DH, Mogun H, Avorn J: HMG-CoA reductase inhibitors and the risk of hip fractures in elderly patients [see comments]. JAMA 2000, 283:3211-3216

36. Werner N, Nickenig G, Laufs U: Pleiotropic effects of HMG-CoA reductase inhibitors. Basic Res Cardiol 2002, 97:105-116

37. Whitfield JF: Statins: new drugs for treating osteoporosis? Expert Opin Investig Drugs 2001, 10:409-415

38. BakhirevaLN, Shainline MR, Carter S, Robinson S, Beaton SJ, Nawar-skas JJ, Gunter MJ: Synergistic effect of statins and postmenopausal hormone therapy in the prevention of skeletal fractures in elderly women. Pharmacotherapy 2010, 30:879-887

39. Cruz AC, Gruber BL: Statins and osteoporosis: can these lipid-low-ering drugs also bolster bones? Cleve Clin J Med 2002, 69:277-278, 280-282, 287-288

40. Bocan TM, Ferguson E, McNally W, Uhlendorf PD, Bak Mueller S, Dehart P, Sliskovic DR, Roth BD, Krause BR, Newton RS: Hepatic and nonhepatic sterol synthesis and tissue distribution following administration of a liver selective HMG-CoA reductase inhibitor: CI-981: comparison with selected HMG-CoA reductase inhibitors. Biochim Biophys Acta 1992, 1123:133-144

41. Corsini A, Maggi FM, Catapano AL: Pharmacology of competitive inhibitors of HMG-CoA reductase. Pharmacol Res 1995, 31:9-27

42. Duggan DE, Chen IW, Bayne WF, Halpin RA, Duncan CA, Schwartz MS, Stubbs RJ, Vickers S: The physiological disposition of lovastatin. Drug Metab Dispos 1989, 17:166-173

43. Germershausen JI, Hunt VM, Bostedor RG, Bailey PJ, Karkas JD, Alberts AW: Tissue selectivity of the cholesterol-lowering agents lo-vastatin, simvastatin and pravastatin in rats in vivo. Biochem Biophys Res Commun 1989, 158:667-675

44. Gerson RJ, MacDonald JS, Alberts AW, Kornbrust DJ, Majka JA, Stubbs RJ, Bokelman DL: Animal safety and toxicology of simvastatin and related hydroxy- methylglutaryl-coenzyme A reductase inhibitors. Am J Med 1989, 87:28S-38S

45. Koga T, Shimada Y, Kuroda M, Tsujita Y, Hasegawa K, Yamazaki M: Tissue-selective inhibition of cholesterol synthesis in vivo by pravastatin sodium, a 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor. Biochim Biophys Acta 1990, 1045:115-120

46. Komai T, Kawai K, Tokui T, Tokui Y, Kuroiwa C, Shigehara E, Tanaka M: Disposition and metabolism of pravastatin sodium in rats, dogs and monkeys. Eur J Drug Metab Pharmacokinet 1992, 17:103-113

47. Parker RA, Clark RW, Sit SY, Lanier TL, Grosso RA, Wright JJ: Selective inhibition of cholesterol synthesis in liver versus extrahepatic tissues by HMG-CoA reductase inhibitors. J Lipid Res 1990, 31: 1271-1282

48. Tse FL, Smith HT, Ballard FH, Nicoletti J: Disposition of fluvastatin, an inhibitor of HMG-COA reductase, in mouse, rat, dog, and monkey. Biopharm Drug Dispos 1990, 11:519-531

49. Vickers S, Duncan CA, Chen IW, Rosegay A, Duggan DE: Metabolic disposition studies on simvastatin, a cholesterol-lowering prodrug. Drug Metab Dispos 1990, 18:138-145

50. White CM: A review of the pharmacologic and pharmacokinetic aspects of rosuvastatin. J Clin Pharmacol 2002, 42:963-970

51. Parhami F, Tintut Y, Beamer WG, Gharavi N, Goodman W, Demer LL: Atherogenic high-fat diet reduces bone mineralization in mice. J Bone Miner Res 2001, 16:182-188

52. You L, Sheng ZY, Tang CL, Chen L, Pan L, Chen JY: High cholesterol diet increases osteoporosis risk via inhibiting bone formation in rats. Acta Pharmacol Sin 32:1498-1504

53. Solomon KR, Pelton K, Boucher K, Joo J, Tully C, Zurakowski D, Schaffner CP, Kim J, Freeman MR: Ezetimibe is an inhibitor of tumor angiogenesis. Am J Pathol 2009, 174:1017-1026

54. The Jackson Laboratory: Densitometric survey of 11 inbred strains of mice. MPD:224, Jaxpheno1. Mouse Phenome Database; 2006.

55. Beamer WG, Donahue LR, Rosen CJ, Baylink DJ: Genetic variability in adult bone density among inbred strains of mice. Bone 1996, 18:397-403

56. Wergedal JE, Sheng MH, Ackert-Bicknell CL, Beamer WG, Baylink DJ: Genetic variation in femur extrinsic strength in 29 different inbred strains of mice is dependent on variations in femur cross-sectional geometry and bone density. Bone 2005, 36:111-122

57. Ishimori N, Li R, Walsh KA, Korstanje R, Rollins JA, Petkov P, Pletcher MT, Wiltshire T, Donahue LR, Rosen CJ, Beamer WG, Churchill GA, Paigen B: Quantitative trait loci that determine BMD in C57BL/6J and 129S1/SvImJ inbred mice. J Bone Miner Res 2006, 21:105-112

58. Mundy GR: Osteoporosis and inflammation. Nutr Rev 2007, 65:S147-151

59. Solomon KR, Adolphson LD, Wank DA, McHugh KP, Hauschka PV: Caveolae in human and murine osteoblasts. J Bone Miner Res 2000, 15:2391-2401

60. Solomon KR, Danciu TE, Adolphson LD, Hecht LE, Hauschka PV: Caveolin-enriched membrane signaling complexes in human and murine osteoblasts. J Bone Miner Res 2000, 15:2380-2390

61. Samelson EJ, Cupples LA, Hannan MT, Wilson PW, Williams SA, Vaccarino V, Zhang Y, Kiel DP: Long-term effects of serum cholesterol on bone mineral density in women and men: the Framingham Osteoporosis Study. Bone 2004, 34:557-561

62. Wu LY, Yang TC, Kuo SW, Hsiao CF, Hung YJ, Hsieh CH, Tseng HC, Hsieh AT, Chen TW, Chang JB, Pei D: Correlation between bone mineral density and plasma lipids in Taiwan. Endocr Res 2003, 29: 317-325

63. Lichtman AH, Clinton SK, Iiyama K, Connelly PW, Libby P, Cybulsky MI: Hyperlipidemia and atherosclerotic lesion development in LDL receptor-deficient mice fed defined semipurified diets with and without cholate. Arterioscler Thromb Vasc Biol 1999, 19:1938-1944