ABSTRACT
PURPOSE: Mitofusin 2 (Mfn2) is one of two mitofusins involved in regulating mitochondrial size, shape and function, including mitophagy, an important cellular mechanism to limit oxidative stress. Reduced expression of Mfn2 has been associated with impaired osteoblast differentiation and function and a reduction in the number of viable osteocytes in bone. We hypothesized that the genetic absence of Mfn2 in these cells would increase their susceptibility to aging-associated metabolic stress, leading to a progressive impairment in skeletal homeostasis over time. METHODS: Mfn2 was selectively deleted in vivo at three different stages of osteoblast lineage commitment by crossing mice in which the Mfn2 gene was floxed with transgenic mice expressing Cre under the control of the promoter for Osterix (OSX), collagen1a1, or DMP1 (Dentin Matrix Acidic Phosphoprotein 1). RESULTS: Mice in which Mfn2 was deleted using DMP1-cre demonstrated a progressive and dramatic decline in bone mineral density (BMD) beginning at 10 weeks of age (n = 5 for each sex and each genotype from age 10 to 20 weeks). By 15 weeks, there was evidence for a functional decline in muscle performance as assessed using a rotarod apparatus (n = 3; 2 males/ 1 female for each genotype), accompanied by a decline in lean body mass. A marked reduction in trabecular bone mass was evident on bone histomorphometry, and biomechanical testing at 25 weeks (k/o: 2 male/1 female, control 2 male/2 female) revealed severely impaired femur strength. Extensive regional myofiber atrophy and degeneration was observed on skeletal muscle histology. Electron microscopy showed progressive disruption of cellular architecture, with disorganized sarcomeres and a bloated mitochondrial reticulum. There was also evidence of neurodegeneration within the ventral horn and roots of the lumbar spinal cord, which was accompanied by myelin loss and myofiber atrophy. Deletion of Mfn2 using OSX-cre or Col1a1-cre did not result in a musculoskeletal phenotype. Where possible, male and female animals were analyzed separately, but small numbers of animals in each group limited statistical power. For other outcomes, where sex was not considered, small sample sizes might still limit the strength of the observation. CONCLUSION: Despite known functional overlap of Mfn1 and Mfn2 in some tissues, and their co-expression in bone, muscle and spinal cord, deletion of Mfn2 using the 8 kB DMP1 promoter uncovered an important non-redundant role for Mfn2 in maintaining the neuromuscular/bone axis.