Osteocalcin is necessary for the alignment of apatite crystallites, but not glucose metabolism, testosterone synthesis, or muscle mass

Abstract

The strength of bone depends on bone quantity and quality. Osteocalcin (Ocn) is the most abundant noncollagenous protein in bone and is produced by osteoblasts. It has been previously claimed that Ocn inhibits bone formation and also functions as a hormone to regulate insulin secretion in the pancreas, testosterone synthesis in the testes, and muscle mass. We generated Ocn-deficient (Ocn-/-) mice by deleting Bglap and Bglap2. Analysis of Ocn-/-mice revealed that Ocn is not involved in the regulation of bone quantity, glucose metabolism, testosterone synthesis, or muscle mass. The orientation degree of collagen fibrils and size of biological apatite (BAp) crystallites in the c-axis were normal in the Ocn-/-bone. However, the crystallographic orientation of the BAp c-axis, which is normally parallel to collagen fibrils, was severely disrupted, resulting in reduced bone strength. These results demonstrate that Ocn is required for bone quality and strength by adjusting the alignment of BAp crystallites parallel to collagen fibrils; but it does not function as a hormone.

Fig 1. Generation of Ocn^–/–mice and a μ-CT analysis of femurs.

(A) Schematic presentation of Bglap and Bglap2 gene loci, targeting vectors, and mutated alleles. (B) Southern blot analysis. DNA from tails was digested with SacI and hybridized with the probe shown in A. Bands are indicated corresponding to wild-type (12.8 kb) and mutated (11.7 kb) alleles. (C) Genotyping by PCR. PCR was performed using primers 1, 2, and 3 shown in A. Primers 1 and 3 amplify 706 bp DNA of the wild-type allele, and primers 2 and 3 amplify 980 bp DNA of the mutated allele. (D) Real-time RT-PCR analysis of Bglap and Bglap2 expression using RNA from the osteoblast fractions of femurs in wild-type (wt) and Ocn^–/–mice at 9 months of age. Primers were designed to detect the expression of both Bglap and Bglap2. wt: n = 3, Ocn^–/–: n = 7. E, The levels of serum Ocn in female mice at 11 weeks of age. wt: n = 5, Ocn^–/–: n = 7. ***P<0.001. (F-I) μ-CT analyses of femurs in male wild-type and Ocn^–/–mice at 14 weeks, 6 months, and 9 months of age. μ-CT images of femoral distal metaphyses (F) and mid-diaphyses (G) are shown. Bars = 500 μm. H, Trabecular bone parameters, including the trabecular bone volume (BV/TV), trabecular thickness (Tb.Th), and trabecular number (Tb.N). I, Cortical bone parameters including cortical thickness (Ct. Th), the periosteal perimeter (Ps.Pm), and endocortical perimeter (Ec.Pm). wt (n = 12), Ocn^–/–(n = 6) at 14w; wt (n = 14), Ocn^–/–(n = 24) at 6m; wt (n = 7), Ocn^–/–(n = 9) at 9m. *vs. wild-type mice. ***P<0.001. X symbols in box plots show outliers.

Fig 2. Bone histomorphometric analysis.

(A) Bone histomorphometric analysis of trabecular bone in femurs at 6 months of age. The trabecular bone volume (bone volume/tissue volume, BV/TV), osteoid surface (OS/BS), osteoid thickness (O.Th), osteoblast surface (Ob.S/BS), number of osteoblasts (N.Ob/B.Pm), osteoclast surface (Oc.S/BS), number of osteoclasts (N.Oc/B.Pm), eroded surface (ES/BS), number of osteocytes (N.Oct/BV), mineral apposition rate (MAR), mineralizing surface (MS/BS), and bone formation rate (BFR/BS) were compared between male wild-type (blue column, n = 7) and Ocn^–/– (red column, n = 7) mice. B.Pm, bone perimeter; BS, bone surface. (B and C) Dynamic histomorphometric analysis of the periosteum (B) and endosteum (C) in the mid-diaphyses of femoral cortical bones in male wild-type (n = 10) and Ocn^–/–(n = 12) mice at 6 months of age. * vs. wild-type mice. **P<0. 01. X symbols in box plots show outliers.

Fig 3. Serum markers and real-time RT-PCR.

(A) The levels of serum P1NP, TRAP5b, and CTX1 in wild-type (wt) and Ocn^–/–mice at 11 weeks, 6 months, and 9 months of age. Male 11 weeks, wt (n = 5) and Ocn^–/–(n = 6); male 6 months, wt (n = 10) and Ocn^–/–(n = 10); female 6 months, wt (n = 6) and Ocn^–/–(n = 8); male 9 months, wt (n = 5) and Ocn^–/–(n = 6); female 9 months, wt (n = 4) and Ocn^–/–(n = 7). (B) Real-time RT-PCR analysis. RNA was extracted from osteoblast fractions from femurs in male wild-type and Ocn^–/–mice at 6 months of age. The values in wild-type mice were defined as 1, and relative levels are shown. wt: n = 7, Ocn^–/–: n = 6. **: P<0.01. X symbols in box plots show outliers.

Fig 4. TEM analyses of cortical bone.

(A-F) Immunoelectron microscopic analysis. Ultrathin sections from a wild-type femur at 10 days of age (A, D) and humerus at 4 months of age (B, E), and Ocn^–/–humerus at 4 months of age (C, F) were immunolabeled with an anti-osteocalcin antibody and analyzed by TEM. Ocn molecules immunolabeled with 20-nm gold particles were detected in wild-type (A, B, D, E), but not Ocn^–/–(C, F) bone. Boxed regions in A, B, and C are magnified in D, E, and F, respectively. The region about 8 μm from osteoid is shown in A, and those about 2 μm from osteoid are shown in B and C. (G-J) TEM analysis of non-decalcified sections. The osteoid regions of the metaphyses of tibiae in wild-type (G, H) and Ocn^–/–(I, J) mice at 4 weeks of age. The boxed regions in G and I are magnified in H and J, respectively. ob: osteoblast, mbm: mineralized bone matrix. H and J show mineralized nodules in osteoids. Bars: 500 nm (A-C, H, J), 100 nm (D-F), and 5 μm (G, I).

Fig 5. Raman microspectroscopic analysis of femurs in male wild-type and Ocn^–/–mice at 14 weeks of age.

(A) A bright field image of the section at mid-diaphysis. The analyzed region is boxed (arrow). (B) Raman spectrum showing the assignments for PO₄^3– (959cm^–1), CO₃^2– (1070 cm^–1), Amide III (1243–1320 cm^–1), and Amide I (1616–1720 cm^–1). (C) Parameters including mineralization, crystallinity, CO₃^2–/PO₄^3–, collagen maturity, and remodeling were calculated as shown in Table 1. n = 6.

Fig 6. BMD and orientations of collagen fibers and the BAp c-axis in the femoral cortical bone of female wild-type and Ocn^–/–mice at 9 months of age.

(A) Schematic presentation of analyzed positions and the correlation of angles and colors used for collagen orientation in B. (B) The orientation of collagen fibers shown in color. (C) BMD. (D) Collagen orientation degree. If the orientation of collagen fibers is completely parallel to the longitudinal direction of bone, the degree is one. (E) BAp c-axis orientation degree. The preferential alignment of the BAp c-axis along the bone longitudinal direction was the intensity ratio of the (002) diffraction peak to the (310) peak. Higher values indicate a more preferential alignment to the longitudinal direction. *: P<0.05, **: P<0.01. (F) Single regression analysis of the orientation of collagen fibers and the BAp c-axis. wt: blue dots, Ocn^–/–: red dots. wt: n = 5, Ocn^–/–: n = 6 in C-F.

Fig 7. Crystallite size in the BAp c-axis and Young’s modulus in femurs of female wild-type and Ocn^–/–mice at 9 months of age.

(A) The crystallite size in the BAp c-axis in the longitudinal and vertical directions at position 5. (B) Young’s modulus along the bone longitudinal axis at position 5 in Fig 6A as measured by nanoindentation. **: P<0.01. X symbols in box plots show outliers. (C) Single regression analysis of Young’s modulus to BMD and each degree of the preferential alignment of collagen fibers and the BAp c-axis in the bone longitudinal direction at position 5. wt: blue dots, Ocn^–/–: red dots. wt: n = 5, Ocn^–/–: n = 6. (D) Schematic presentation of orientations of collagen fibrils and BAp c-axis. In wild-type mice, the orientations of collagen fibrils and BAp c-axis are parallel to the bone longitudinal axis. In Ocn^–/–mice, the orientation of collagen fibrils is parallel to the bone longitudinal axis, whereas that of BAp c-axis is severely disrupted from the longitudinal axis. The disruption of the BAp c-axis orientation reduces Young’s modulus in the longitudinal axis.

Fig 8. Body weight, blood glucose, HbA1c, and adipose tissues in Ocn^–/–mice.

(A)Body weights of male (m) and female (f) wild-type and Ocn^–/–mice. Male (wt: n = 5, Ocn^–/–: n = 6), female (n = 9, 6) at 11w; male (n = 25, 18) at 14w; male (n = 14, 23), female (n = 6, 8) at 6m; male (n = 7, 9), female (n = 8, 12) at 9m; female (n = 8, 6) at 12m; male (n = 7, 4) at 18m. (B) Blood glucose levels. Male (wt: n = 5, Ocn^–/–: n = 6) at 14w; male (n = 16, 16), female (n = 8, 13) at 9m; male (n = 7, 5) at 18m. (C) HbA1c levels. Male (wt: n = 5, Ocn^–/–: n = 6), female (n = 9, 10) at 11w; male (n = 14, 18), female (n = 6, 7) at 6m; male (n = 9, 7), female (n = 9, 13) at 9m; male (n = 7, 5) at 18m. (D-H) Measurement of adipose tissues. D-G, Sagittal sections at the midline (D, E) and cross sections at the level of L5 (F, G) of μ-CT in male mice at 14 weeks of age. Subcutaneous adipose tissues are shown orange and visceral adipose tissues are shown yellow in the range of the first to fifth vertebra. H, Volumes of subcutaneous and visceral adipose tissues in wild-type (n = 10) and Ocn^–/–(n = 6) male mice at 14 weeks of age and wild-type (n = 8) and Ocn^–/–(n = 5) female mice at 9 months of age. The percentages of adipose tissue volumes against abdominal volumes in the range of the first to fifth vertebra are shown. X symbols in box plots show outliers.

Fig 9. GTTs and ITTs in Ocn^–/–mice.

(A) GTT by glucose (2 g/kg body weight) injection. Glucose levels in male mice (wt: n = 9, Ocn^–/–: n = 12) at 5 months of age fed a normal diet. (B and C) GTT by glucose (1 g/kg body weight) injection. Glucose (B) and insulin (C) levels in male mice at 9 months of age fed a normal diet. wt: n = 9, Ocn^–/–: n = 7. (D and E) GTT by glucose (2 g/kg body weight) injection. Glucose (D) and insulin (E) levels in male mice (wt: n = 8, Ocn^–/–: n = 12) at 6 months of age fed a high-fat diet for 1 month. (F) ITT. Glucose levels in male mice (wt: n = 20, Ocn^–/–: n = 12) at 4 months of age fed a normal diet. (G) ITT. Glucose levels in male mice (wt: n = 7, Ocn^–/–: n = 12) at 8 months of age fed a high-fat diet for 3 months.

Fig 10. Relationships between exercise, glucose metabolism, and bone formation.

(A-D) Body weights (A), HbA1c (B), serum levels of carboxyleted Ocn (Gla-OC), uncarboxyleted Ocn (Glu-OC), P1NP, and CTX-1 (C), and μ-CT analysis of trabecular bone (D) in male wild-type mice fed a normal diet with or without exercise on a treadmill for 7 weeks. Mice were analyzed at 4 months of age. n = 13. (E-H) Body weights (E), HbA1c (F), serum levels of carboxyleted Ocn, uncarboxyleted Ocn, P1NP, and CTX-1 (G), and μ-CT analysis of trabecular bone (H) in male KK/TaJcl mice with or without exercise on a treadmill for 7 weeks. Mice were fed a high-fat diet and analyzed at 4 months of age. Control: n = 15, treadmill: n = 14. *: P<0.05, **: P<0.01, ***: P<0.001. X symbols in box plots show outliers.

Fig 11. Analyses of testosterone synthesis and spermatogenesis.

(A and B) Appearance of testes (A) and their weights (B) in wild-type (n = 5 at 11w, n = 7 at 6m, n = 7 at 8m) and Ocn^–/–(n = 6 at 11w, n = 7 at 6m, n = 11 at 8m) mice. Bars: 2 mm. (C) Serum testosterone levels in wild-type (n = 4 at 11w, n = 12 at 6m, n = 6 at 18m) and Ocn^–/–(n = 3 at 11w, n = 15 at 6m, n = 4 at 18m) mice. (D) H-E stained sections of the cauda epididymides at 4 months of age. Bars: 10 μm. (E) The number of spermatozoa. The number of spermatozoa in 1 mg of seminal fluid of each mouse at 8 months of age are shown. wt: n = 7, Ocn^–/–: n = 11. (F) Morphology of sperms. Sperms were stained with Wright-Giemsa or analyzed with fluorescence microscope after the treatment with Hoechst 33342. Bars: 2 μm. (G) The frequencies of the sperms with acrosomal defects. More than 300 sperms in each mouse at 8 months of age were evaluated using the smears stained with Wright-Giemsa. wt: n = 7, Ocn^–/–: n = 11. (H) TUNEL staining. The sections were counterstained with hematoxylin. Bars: 50 μm. (I) The number of TUNEL-positive cells in seminiferous tubules. All tubules (more than 200 tubules) in one section were evaluated in each mouse at 11 weeks of age (wt: n = 5, Ocn^–/–: n = 6) and 6 months of age (wt: n = 5, Ocn^–/–: n = 4), and the means of the number of TUNEL-positive cells in one tubule are shown. (J) Real-time RT-PCR analysis using RNA from testes at 8 months of age. The values in wild-type mice were defined as 1, and relative levels are shown. wt: n = 5, Ocn^–/–: n = 8. X symbols in box plots show outliers.

Fig 12. Analyses of muscles.

(A and B) Body weights (A) and muscle weights (B) in male mice at 4 months (wt: n = 5, Ocn^–/–: n = 6) and 8 months (wt: n = 7, Ocn^–/–: n = 11) of age. (C and D) H-E stained sections of the quadriceps (C), and the average area of myofibers in the quadriceps (D) of male wild-type and Ocn^–/–mice at 4 months of age. The areas of sixty myofibers in the cross sections at mid-diaphyses of femurs were measured. wt: n = 7, Ocn^–/–: n = 7. EDL: extensor digitorum longus. Bars: 50 μm. X symbols in box plots show outliers.