Osx-CKO and littermate control (FIP200F/F) mice were born with similar body weight (Supplemental Fig. S2). However, by as early as 1 week, both male and female Osx-CKO mice weighed less than controls. This weight differential was extended over the 6 to 12 months of observation. To investigate the effects of FIP200 deficiency on bone development, we first evaluated deletion of FIP200 by Western blotting of lysates from tissues of Osx-CKO and control neonatal mice. As shown in Supplemental Fig. S3, a significantly reduced level of FIP200 was found in the calvaria but not heart and liver samples from Osx-CKO mice compared with control mice, indicating efficient and specific deletion of FIP200 as expected. We then performed whole mount skeleton staining of newborn mice. Compared with control bones, Osx-CKO mice had no obvious developmental defects in major skeletal elements including humerus, radius, and ulna (Supplemental Fig. S4A), femur, tibia, and fibula (Supplemental Fig. S4B), ribs and vertebrae (Supplemental Fig. S4C), suggesting that bone formation was not affected at this stage of development. Next, we used micro-CT to evaluate the effects of FIP200 deletion on postnatal bone morphometry. Micro-CT analysis revealed an osteopenic phenotype in Osx-CKO mice. In femur trabecular bone of Osx-CKO female mice, we observed a decrease (59%, 37%, and 53% at 1, 2, and 6 months, respectively) in trabecular bone volume (BV/TV) (Fig. 3A), a decrease (52%, 33%, and 52% at 1, 2, and 6 months, respectively) in trabecular number (TbN) (Fig. 3B), a slight decrease (14% and 9% at 1 and 2 months, respectively) in trabecular thickness (TbTh) (Fig. 3C), and an increase (172%, 57%, and 129% at 1, 2, and 6 months, respectively) in trabecular spacing (TbSp) (Fig. 3D). In femur cortical bone of Osx-CKO mice, we observed a decrease (17%, 23%, and 23% at 1, 2, and 6 months, respectively) in cortical thickness (Fig. 3E), a decrease (16%, 9%, and 11% at 1, 2, and 6 months, respectively) in outer cortical bone perimeter (Fig. 3F), and a decrease (16%, 6%, and 6% at 1, 2, and 6 months, respectively) in inner cortical bone perimeter (Fig. 3G). Besides the femurs, we performed micro-CT analysis on 1-month-old L3 vertebrae and observed similar phenotype. There was a 34%, 21%, and 17% decrease in trabecular bone volume, trabecular number, and trabecular thickness, respectively, in Osx-CKO mice (Supplemental Fig. S5). In addition to the endochondrally formed bones, we also examined calvaria (intramembranous bone). The calvarial thickness of CKO mice was significantly decreased compared with controls (Fig. 3H). Similar osteopenic phenotype was observed in male mutant mice (Supplemental Fig. S6). To determine to what extent the low bone mass phenotype was owing to the Osx-Cre transgene itself, we analyzed the femur and vertebrae L3 of Osx-Cre mice and their littermate control mice. Except the slightly decreased outer and inner femur cortical perimeters, Osx-Cre mice had comparable cortical and trabecular bone parameters compared with the control mice (Supplemental Fig. S7), suggesting that the phenotype observed in Osx-CKO mice was largely because of the deletion of FIP200. To further confirm that the observed phenotype was indeed resulting from the specific deletion of FIP200 in osteoblasts, we used other lines of osteoblast-targeting Cre transgenic mice, including Col3.6-Cre and Col2.3-Cre to delete FIP200 in osteoblasts and observed a similar osteopenic phenotype in FIP200F/F;Col3.6-Cre (Col3.6-CKO) mice (data not shown) and FIP200F/F;Col2.3-Cre (Col2.3-CKO) mice (Supplemental Fig. S8A–D), highlighting the role of FIP200 in osteoblast functions. As expected, significantly reduced level of FIP200 was found in femur but not heart and liver samples from Col2.3-CKO mice compared with control mice (Supplemental Fig. S8E). In addition, FIP200-null osteoblasts isolated from Col2.3-CKO mice demonstrated similar autophagy deficiency (Supplemental Fig. S8F, G) as the cells isolated from Osx-CKO mice. Taken together, our data demonstrated that FIP200 deletion in osteoblasts leads to compromised bone development and decreased bone mass in mice.
To determine the extent to which FIP200 deletion would affect the mechanical properties of bone, we tested the femurs of Osx-CKO and control mice by four-point bending. Femurs of female mutant mice had decreased yield load (Fig. 3I), decreased ultimate load (Fig. 3J), decreased stiffness (Fig. 3K), and decreased elastic energy (Fig. 3L). In contrast, increased plastic energy and total failure energy were noted in the 6-month-old mutant group (Fig. 3M, N), which may be an adaptive response or just a reflection of a change in the properties of the extracellular matrix. Similar mechanical property changes were also found in femurs of mutant male mice (Supplemental Fig. S3I–L). Interestingly, the bones of the 6-month-old mutant male mice had comparable yield load (Supplemental Fig. S3I) and ultimate load (Supplemental Fig. S3J), which is consistent with the comparable cortical bone thickness at this stage (Supplemental Fig. S3E). Collectively, these data demonstrated that bones in FIP200 mutant mice had significantly altered mechanical properties with decreased strength to resist fracture compared with that of control mice.