Sox9 restrains chondrocyte to osteoblast lineage progression-implications for the pathogenesis of Campomelic Dysplasia

Abstract

In endochondral bone development bi-potential chondroosteoprogenitors form mesenchymal condensates followed by a cascade of chondrocyte differentiation steps to form cartilage. Osteoblasts, originating from the perichondrium, accompany vascular invasion and lay down endochondral bone to replace cartilage. It has been generally accepted that chondrocytes and osteoblasts are separate lineages derived from a common osteo-chondroprogenitor and in the terminal step, differentiated hypertrophic chondrocytes (HCs) are thought to undergo apoptosis. But whether hypertrophy is the terminal differentiation stage of chondrocytes, with apoptosis as the ultimate fate, or marks a transition preceding osteogenesis has been debated. Using a genetic recombination lineage tracing approach in mice, we showed that HCs directly contribute to the osteoblast lineage and can become osteogenic cells in fetal and postnatal endochondral bones, persist into adulthood. Therefore, hypertrophy is not the terminal differentiation state for chondrocytes, but is part of a continuum in which HCs directly contribute to the full osteoblast lineage. SOX9 is the master transcriptional regulator of chondrogenesis and mutations in the gene cause campomelic dysplasia (CD) characterized by cleft palate, short stature, bowed limbs (campomelia), XY sex reversal. The human CD SOX9Y440X nonsense mutation leads to truncation of the transactivation domain and is suggested to be hypomorphic. We generated a conditional mouse equivalent of the SOX9Y440X mutation. All the mutants recapitulated the human CD skeletal phenotypes. But heterozygous SOX9Y440X + mice, display a more severe skeletal phenotype especially with respect to campomelia than in Sox9-/+, suggestive of a dominant negative or neomorphic pathogenesis. We hypothesized that the campomelia is caused by the impact of the SOX9Y440Xmutation on the chondrocyte to osteoblast lineage transition. We found dysregulated hedgehog signaling in Sox9Y440X/Y440X growth plates. Lineage tracing showed an increased number of HC descendants suggesting an accelerated HC to osteoblast transition. We propose that SOX9 normally restrains the transition of HCs to osteoblasts. The campomelia in CD is a consequence of loss of control of the lineage progression from chondrocyte to osteoblast.