Runx2 is essential for the transdifferentiation of chondrocytes into osteoblasts

Abstract

Chondrocytes proliferate and mature into hypertrophic chondrocytes. Vascular invasion into the cartilage occurs in the terminal hypertrophic chondrocyte layer, and terminal hypertrophic chondrocytes die by apoptosis or transdifferentiate into osteoblasts. Runx2 is essential for osteoblast differentiation and chondrocyte maturation. Runx2-deficient mice are composed of cartilaginous skeletons and lack the vascular invasion into the cartilage. However, the requirement of Runx2 in the vascular invasion into the cartilage, mechanism of chondrocyte transdifferentiation to osteoblasts, and its significance in bone development remain to be elucidated. To investigate these points, we generated Runx2fl/flCre mice, in which Runx2 was deleted in hypertrophic chondrocytes using Col10a1 Cre. Vascular invasion into the cartilage was similarly observed in Runx2fl/fl and Runx2fl/flCre mice. Vegfa expression was reduced in the terminal hypertrophic chondrocytes in Runx2fl/flCre mice, but Vegfa was strongly expressed in osteoblasts in the bone collar, suggesting that Vegfa expression in bone collar osteoblasts is sufficient for vascular invasion into the cartilage. The apoptosis of terminal hypertrophic chondrocytes was increased and their transdifferentiation was interrupted in Runx2fl/flCre mice, leading to lack of primary spongiosa and osteoblasts in the region at E16.5. The osteoblasts appeared in this region at E17.5 in the absence of transdifferentiation, and the number of osteoblasts and the formation of primary spongiosa, but not secondary spongiosa, reached to levels similar those in Runx2fl/fl mice at birth. The bone structure and volume and all bone histomophometric parameters were similar between Runx2fl/fl and Runx2fl/flCre mice after 6 weeks of age. These findings indicate that Runx2 expression in terminal hypertrophic chondrocytes is not required for vascular invasion into the cartilage, but is for their survival and transdifferentiation into osteoblasts, and that the transdifferentiation is necessary for trabecular bone formation in embryonic and neonatal stages, but not for acquiring normal bone structure and volume in young and adult mice.

Fig 1. Skeletal system and histological analyses in Runx2^fl/fl and Runx2^fl/fl/Cre embryos at E15.5.

(A and B) Immunohistochemical analysis in tibia from Runx2^fl/fl (A) and Runx2^fl/fl/Cre (B) mice using anti-Runx2 antibody. The boxed regions in A and B are magnified in A’ and B’, respectively. (C) The number of Runx2-positive hypertrophic chondrocytes was counted and shown. (D and E) Lateral view of the whole skeletons of Runx2^fl/fl (D) and Runx2^fl/fl/Cre (E) mice. Hind limbs were magnified in D’ and E’. (F-P) Histological analyses using femoral sections from Runx2^fl/fl (F, H, J, L, N) and Runx2^fl/fl/Cre (G, I, K, M, O) mice. (F and G) H-E staining. The boxed regions in F, G, F’, and G’ are magnified in F’, G’, F”, and G”, respectively. (H and I) von Kossa staining. (J and K) ALP staining. (L and M) Safranin O staining. (N and O) TUNEL staining. (P) The number of TUNEL-positive hypertrophic chondrocytes was counted. Scale bars: 100 μm (A-B’, F-O), 1 mm (D-E’). The number of mice analyzed: Runx2^fl/fl: 4–6, Runx2^fl/fl/Cre: 3–5. (Q and R) The expression of the genes related to chondrocyte and osteoclast differentiation and apoptosis in hypertrophic and terminal hypertrophic layers. (Q) The hypertrophic and terminal hypertrophic layers in femurs and tibiae at E15.5 were isolated using laser capture microdissection system and RNA was prepared. (R) Real-time RT-PCR analysis. The values in Runx2^fl/fl mice were defined as 1, and relative levels are shown. The number of mice analyzed: Runx2^fl/fl: 7, Runx2^fl/fl/Cre: 7. Data are the mean ± SD. *Versus Runx2^fl/fl, *p<0.05, **p<0.01, ***p<0.001.

Fig 2. In situ hybridization at E16.5.

The femoral sections from Runx2^fl/fl (A, D, G, J, M, P, S), Runx2^fl/+/Cre (B, E, H, K, N, Q, T), and Runx2^fl/fl/Cre (C, F, I, L, O, R, U) mice were used. (A-C) H–E staining. (D-U) In situ hybridization using Col2a1 (D-F), Col10a1 (G-I), Mmp13 (J-L), Ibsp (M-O), Spp1 (P-R), and Col1a1 (S-U) probes. The boxed regions in A-U are magnified in A’-U’, respectively. (V) Intensity on in situ hybridization. The gray values in the red boxes in G’-U’ were measured. The gray values in Runx2^fl/fl mice were set as 1, and the relative levels are shown. Scale bars: 100 μm. The number of mice analyzed: Runx2^fl/fl: 3–6, Runx2^fl/+/Cre: 3, Runx2^fl/fl/Cre: 3–5. Data are the mean ± SD. *Versus Runx2^fl/fl, #Versus Runx2^fl/+/Cre, *^,#p<0.05, **^,##p<0.01.

Fig 3. Histological analyses of Runx2^fl/fl and Runx2^fl/fl/Cre embryos at E16.5.

The femoral sections from Runx2^fl/fl (A, C, E, I, L, P, R, T) and Runx2^fl/fl/Cre (B, D, F, J, M, Q, S, U) mice were used. (A and B) ALP staining. (C-F) Immunohistochemical analysis using anti-Runx2 (C, D) and anti-Sp7 (E, F) antibodies. The boxed regions in A-F were magnified in A’-F’, respectively. (G and H) The numbers of Runx2 (G)- and Sp7 (H)-positive cells in the red boxes of C’-F’ were counted. (I-K) Immunohistochemical analysis using anti-CD34 antibody. The boxed regions in I and J were magnified in I’ and J’, respectively. The densities of stained endothelial cells were measured in the boxed region in I and J, the value in Runx2^fl/fl mice was set as 1, and the relative levels are shown (K). The number of mice analyzed: Runx2^fl/fl: 3, Runx2^{fl/fl Cre}: 3. (L and M) BrdU labeling. The boxed regions in L were magnified in L’ and L”, and the boxed regions in M were magnified in M’ and M”. (N and O) Frequencies of BrdU-positive cells. The frequencies of BrdU-positive chondrocytes in the proliferating layers (L’, M’, N) and those of BrdU-positive cells in the primary spongiosa regions (red boxes in L” and M”, O) were counted. (P and Q) Safranin O staining. The boxed regions in P and Q were magnified in P’ and Q’, respectively. (R and S) TRAP staining. (T and U) TUNEL staining. Serial sections were used in P and T, and in Q and U. (V) The percentage of the safranin O-positive area in bone marrow. (W) TRAP-positive cell number in bone marrow. (X) The number of TUNEL-positive cells. TUNEL-positive cells were counted in the regions that correspond to the boxed regions in P and Q using the serial sections. (Y) Real-time RT-PCR analysis of the genes related to osteoclastogenesis. RNA was extracted from the diaphysis of femurs and tibiae in Runx2^fl/fl and Runx2^fl/fl/Cre mice using razor blade and stereoscopic microscope at E16.5. The values in Runx2^fl/fl mice were defined as 1, and the relative levels are shown. Scale bars: 100 μm. The number of mice analyzed: Runx2^fl/fl: 4–5, Runx2^fl/fl/Cre: 4 in ALP staining, immunohistochemical analysis, and TUNEL staining. Runx2^fl/fl: 8, Runx2^fl/fl/Cre: 5 in BrdU labeling. Runx2^fl/fl: 17, Runx2^fl/fl/Cre: 15 in safranin O staining. Runx2^fl/fl: 9, Runx2^fl/fl/Cre: 6 in TRAP staining. Runx2^fl/fl: 8, Runx2^{fl/fl Cre}: 6 in real-time RT-PCR analysis. Data are the mean ± SD. *Versus Runx2^fl/fl, *p<0.05, **p<0.01, ***p<0.001.

Fig 4. Vegfa expression in Runx2^fl/fl and Runx2^fl/fl/Cre embryos at E16.5.

H–E staining (A, D, F, I, K, N, P, S) and immunohistochemical analysis using anti-Vegfa antibody (C, E, H, J, M, O, R, T) or same working concentration of normal mouse IgG (B, G, L, Q) of femoral sections from Runx2^fl/fl (A-C, F-H, K-M, P-R) and Runx2^fl/fl/Cre (D, E, I, J, N, O, S, T) mice. The boxed regions in A-E were magnified in F-J, respectively. The boxed regions in F-J were magnified in K and P, L and Q, M and R, and N and S, and O and T, respectively. Scale bars: 100 μm. The number of mice analyzed: Runx2^fl/fl: 2, Runx2^fl/fl/Cre: 2.

Fig 5. β-galactosidase staining of femoral sections from Runx2^{fl/+/Cre LacZ} and Runx2^{fl/fl/Cre LacZ} mice at E15.5, E16.5, and E17.5.

β-galactosidase staining of femoral sections from Runx2^{fl/+/Cre LacZ} (A, I, K) and Runx2^{fl/fl/Cre LacZ} (C, J, L) mice at E15.5 (A, C), E16.5 (I, J), and E17.5 (K, L). The boxed regions in A, E, B, C, G, D, I, M, J, N, K, O, L, and P were magnified in E, B, F, G, D, H, M, Q, N, R, O, S, P, and T, respectively. The cells in the boxes in Q, R, S, and T were magnified in the windows. (U) The number of β-gal positive cells in Q and R. Scale bars: 100 μm. The number of mice analyzed: Runx2^{fl/+/Cre LacZ}: 2, Runx2^{fl/fl/Cre LacZ}: 3 at E15.5. Runx2^{fl/+/Cre LacZ}: 5, Runx2^{fl/fl/Cre LacZ}: 8 at E16.5. Runx2^{fl/+/Cre LacZ}: 4, Runx2^{fl/fl/Cre LacZ}: 2 at E17.5. Data are the mean ± SD. *Versus Runx2^{fl/+/Cre LacZ}, **p<0.01.

Fig 6. Tracing of the transdifferentiated osteoblasts in trabecular bone at E17.5, newborn, and one week of age.

Frozen femoral sections from Runx2^{fl/+Cre ROSA tomato} (A, C, E, G, I, K, M, O, Q) and Runx2^{fl/flCre ROSA tomato} (B, D, F, H, J, L, N, P, R) mice at E17.5 (A-F), newborn (G-L), and 1 week of age (M-R) were observed by confocal microscopy. Osteoblasts express tdTomato (red), the cells derived from hypertrophic chondrocytes express mTFP (blue-green), and the osteoblasts derived from hypertrophic chondrocytes exhibit their merged whitish color. The boxed regions in A-R and A’-R’ are magnified in A’-R’ and A”-R”, respectively. Arrows show the osteoblasts derived from hypertrophic chondrocytes. Scale bars: 500 μm (A-R), 100μm (A’-R’), and 50μm (A”-R”). (S-U) The number of tdTomato-positive (t⁺) cells and tdTomoto- and mTFP-double positive (t⁺m⁺) cells were counted in trabecular bone, and the percentages of t⁺m⁺ cells in t⁺ cells were calculated at E17.5 (S), newborn (T), and 1 week of age (U). Arrows show t⁺m⁺ cells. The number of mice analyzed: Runx2^{fl/+Cre Rosa tomato}: 3–4, Runx2^{fl/flCre Rosa tomato}: 3–4. Data are the mean ± SD. *Versus Runx2^{fl/+ Cre Rosa tomato}, *p<0.05, **p<0.01, ***p<0.001.

Fig 7. Tracing of the transdifferentiated osteoblasts in cortical bone at 1 week of age, and trabecular and cortical bone at 3 weeks of age.

(A-G) Cortical bone in frozen femoral sections in Runx2^{fl/+Cre Rosa tomato} (A, C, E) and Runx2^{fl/flCre Rosa tomato} (B, D, F) mice at 1 week of age. The boxed regions in A-F and A’-F’ were magnified in A’-F’ and A”-F”, respectively. The number of tdTomato-positive (t⁺) cells and tdTomoto- and mTFP-double positive (t⁺m⁺) cells were counted in the endosteum of cortical bone, and the percentages of t⁺m⁺ cells in t⁺ cells were calculated (G). (H-P) Trabecular (H, K, N) and cortical (J, M, P) bone in the frozen femoral sections in Runx2^{fl/+Cre Rosa tomato} mice at 3 weeks of age. The boxed region in H, I, I’, J, J’, K, L, L’, M, M’, N, O, O’, P, and P’ were magnified in I, I’, I”, J’, J”, L, L’, L”, M’, M”, O, O’, O”, P’, and P”, respectively. The number of tdTomato-positive (t⁺) cells and tdTomoto- and mTFP-double positive (t⁺m⁺) cells were counted in the trabecular bone and in the endosteum of cortical bone, and the percentages of t⁺m⁺ cells in t⁺ cells were calculated (Q). Arrows show t⁺m⁺ cells. Scale bars: 1mm (H, K, N), 500 μm (A-F, I, J, L, M, O, P), 100 μm (A’-F’, I’, J’, L’, M’, O’, P’), and 50 μm (A”-F”, I”, J”, L”, M”, O”, P”). The number of mice analyzed: Runx2^{fl/+Cre Rosa tomoto}: 3–4, Runx2^{fl/flCre Rosa tomato}: 3–4. *Versus Runx2^{fl/+Cre Rosa tomato}, *p<0.05, **p<0.01.

Fig 8. Histological analyses of Runx2^fl/fl and Runx2^fl/fl/Cre embryos at E17.5.

Femoral (A-J, R-U) and tibial (K, L) sections from Runx2^fl/fl (A, C, E, G, I, K, R, T) and Runx2^fl/fl/Cre (B, D, F, H, J, L, S, U) mice were used for the analyses. (A and B) H–E staining. (C and D) Safranin O staining. (E–L) In situ hybridization using Ibsp (E, F), Spp1 (G, H), Col1a1 (I, J), and Bglap2 (K, L) probes. (M) The percentage of the safranin O-positive area in red boxes of C’ and D’. (N–Q) Intensity of in situ hybridization using Ibsp, Spp1, Col1a1, and Bglap2 probes. The gray values in the red boxes in E’–L’ were measured. The gray values in Runx2^fl/fl mice were set as 1, and the relative levels are shown. (R-U) Immunohistochemical analysis using anti-Runx2 (R, S) and anti-Sp7 (T, U) antibodies. The boxed regions in A-L, R-U, and R’-U’ were magnified in A’-L’, R’-U’, and R”-U”, respectively. Scale bars: 100 μm. (V, W) The number of Runx2-positive cells in R” and S” (V) and Sp7-positive cells in T” and U” (W). The number of mice analyzed: Runx2^fl/fl: 4–7, Runx2^fl/fl/Cre: 4–7. Data are the mean ± SD. *Versus Runx2^fl/fl, *p<0.05.

Fig 9. Histological and micro-CT analyses of Runx2^fl/fl and Runx2^fl/flCre newborn mice.

The femoral sections from Runx2^fl/fl (A, C, F, H) and Runx2^fl/fl/Cre (B, D, G, I) mice were used for histological analyses. (A and B) H–E staining. (C and D) Immunohistochemical analysis using anti-Runx2 antibody. (E) The number of Runx2-positive cells in C” and D”. (F-I) In situ hybridization using Col1a1 and Bglap2 probes. (J and K) Intensity of in situ hybridization using the Col1a1 probe in F’ and G’(J) and Bglap2 probe in H’ and I’ (K). The values in Runx2^fl/fl mice were set as 1, and the relative levels are shown. (L, M) Physical appearance of Runx2^fl/fl and Runx2^fl/flCre mice. (N-Q) Micro-CT images of whole femurs (N, O) and the trabecular bone (P, Q) in Runx2^fl/fl (N, P) and Runx2^fl/flCre (O, Q) mice. (R) Body weight of Runx2^fl/fl and Runx2^fl/flCre mice. (S-U) The percentage of the total bone volume/total tissue volume in N and O (S), that of total trabecular bone volume/total tissue volume in P and Q (T), and that of secondary spongiosa volume/tissue volume in the red boxes in P and Q (U). The boxed regions in A, B, C, D, C’, D’, F, G, H, and I are magnified in A’, B’, C’, D’, C”, D”, F’, G’, H’, and I’, respectively. Scale bars: 100 μm (A-D”, F-I’), 0.5 cm (L, M), 0.5 mm (N-Q). The number of mice analyzed: Runx2^fl/fl: 4–7, Runx2^fl/fl/Cre: 3–4. Data are the mean ± SD. *Versus Runx2^fl/fl, *p<0.05.

Fig 10. Micro-CT and bone histomorphometric analyses of Runx2^fl/fl and Runx2^fl/flCre female mice at 6 weeks of age.

(A, B) Physical appearance of Runx2^fl/fl (A) and Runx2^fl/flCre (B) mice. (C) Body weight. (D-K) Three-dimensional images of trabecular bone in the primary ossification center (D, E), trabecular bone in the secondary ossification center (F, G), and cortical bone at mid-diaphysis (H, I) in femurs, and those of trabecular bone in first lumbar vertebrae (J, K) in Runx2^fl/fl (D, F, H, J) and Runx2^fl/flCre (E, G, I, K) mice. The views from the top are shown for the trabecular bone in the secondary ossification center (F, G). (L-O) Trabecular bone volume (BV/TV), trabecular thickness (Tb.Th), trabecular number (Tb.N), and trabecular bone mineral density (BMD) in the primary ossification center (L), secondary ossification center (M), and first lumber vertebrae (O); and cortical area (CtAr/TtAr), cortical thickness (Ct.Th), and cortical bone mineral density (BMD) at mid-diaphysis of femurs (N). (P-W) Bone histomorphometric analyses of cortical bone. Cross-sections from the mid-diaphyses of femurs in Runx2^fl/fl (P, R, S) and Runx2^{fl/fl Cre} mice (Q, T, U), in which calcein had been injected twice, were analyzed. R and T show periosteum (per) and S and U show endosteum (end). (V, W) Mineral apposition rate (MAR), mineralizing surface (MS/BS), and bone formation rate (BFR/BS) in the periosteum (V) and endosteum (W). BS, bone surface. Scale bars: 1 cm (A, B), 1 mm (D-K), 200 μm (P, Q) and 100 μm (R–U). The number of mice analyzed: Runx2^fl/fl: 12, Runx2^fl/fl/Cre: 10. Data are the mean ± SD.

Fig 11. Micro-CT and bone histomorphometric analyses of Runx2^fl/fl and Runx2^fl/flCre male mice at 20 weeks of age.

(A, B) Physical appearance of Runx2^fl/fl (A) and Runx2^fl/flCre (B) mice. (C) Body weight. (D-K) Three-dimensional images of trabecular bone in the primary ossification center (D, E), trabecular bone in the secondary ossification center (F, G), and cortical bone at mid-diaphysis (H, I) in femurs, and those of trabecular bone in first lumbar vertebrae (J, K) in Runx2^fl/fl (D, F, H, J) and Runx2^fl/flCre (E, G, I, K) mice. The views from the top are shown for the trabecular bone in the secondary ossification center (F, G). (L-O) Trabecular bone volume (BV/TV), trabecular thickness (Tb.Th), trabecular number (Tb.N), and trabecular bone mineral density (BMD) in the primary ossification center (L), secondary ossification center (M), and first lumber vertebrae (O); and cortical area (CtAr/TtAr), cortical thickness (Ct.Th), and cortical bone mineral density (BMD) at mid-diaphysis of femurs (N). (P-W) Bone histomorphometric analyses of cortical bone. Cross-sections from the mid-diaphyses of femurs in Runx2^fl/fl (P, R, S) and Runx2^{fl/fl Cre} mice (Q, T, U) were analyzed. R and T show endosteum (end) and S and U show periosteum (per). (V, W) Mineral apposition rate (MAR), mineralizing surface (MS/BS), and bone formation rate (BFR/BS) in the endosteum (V) and periosteum (W). BS, bone surface. Scale bars: 1 cm (A, B), 1 mm (D-K), 200 μm (P, Q) and 100 μm (R–U). The number of mice analyzed: Runx2^fl/fl: 10, Runx2^fl/fl/Cre: 9 in femurs and Runx2^fl/fl: 6, Runx2^fl/fl/Cre: 6 in vertebrae. Data are the mean ± SD.

Fig 12. Bone histomorphometic analyses of vertebrae and tibiae in females at 6 and 12 weeks of age.

(A-D) H-E and von Kossa staining of lumbar vertebra of Runx2^fl/fl (A, C) and Runx2^{fl/fl Cre} (B, D) mice at 6 weeks (A, B) and 12 weeks (B, D) of age. (E and F) The trabecular bone volume (bone volume/tissue volume, BV/TV), osteoblast surface (Ob.S/BS), osteoblast number (Ob.N/BS), osteoid surface (OS/BS), osteoid thickness (O.Th), osteoclast surface (Oc.S/BS), osteoclast number (Oc.N/BS), eroded surface (ES/BS), mineral apposition rate (MAR), mineralizing surface (MS/BS), bone formation rate (BFR/BS) at 6 weeks (E) and 12 weeks (F) of age. BS, bone surface. (G and H) H-E and von Kossa staining of tibiae in Runx2^fl/fl (G) and Runx2^{fl/fl Cre} (H) mice at 12 weeks of age. (I) Bone histomorphometric analysis of tibiae. Scale bars: 200 μm. The number of mice analyzed: Runx2^fl/fl: 12, Runx2^{fl/fl Cre}: 10 at 6 weeks of age; Runx2^fl/fl: 7, Runx2^{fl/fl Cre}: 6 at 12 weeks of age.

Fig 13. The three-point bending test.

(A and B) Representative load-displacement curves for female Runx2^fl/fl (A) and Runx2^fl/flCre (B) mice at 12 weeks of age. (C-G) Body weight (C), maximum load (D), displacement (E), stiffness (the slope of the linear part of the load) (F), and energy to failure (shaded area under the curve) (G) in female Runx2^fl/fl and Runx2^fl/flCre mice. The number of mice analyzed: Runx2^fl/fl: 6, Runx2^fl/flCre: 6.