A murine model of neurofibromatosis type 1 tibial pseudarthrosis featuring proliferative fibrous tissue and osteoclast-like cells

Abstract

Neurofibromatosis type 1 (NF1) is a common genetic condition caused by mutations in the NF1 gene. Patients often suffer from tissue-specific lesions associated with local double-inactivation of NF1. In this study, we generated a novel fracture model to investigate the mechanism underlying congenital pseudarthrosis of the tibia (CPT) associated with NF1. We used a Cre-expressing adenovirus (AdCre) to inactivate Nf1 in vitro in cultured osteoprogenitors and osteoblasts, and in vivo in the fracture callus of Nf1(flox/flox) and Nf1(flox/-) mice. The effects of the presence of Nf1(null) cells were extensively examined. Cultured Nf1(null)-committed osteoprogenitors from neonatal calvaria failed to differentiate and express mature osteoblastic markers, even with recombinant bone morphogenetic protein-2 (rhBMP-2) treatment. Similarly, Nf1(null)-inducible osteoprogenitors obtained from Nf1 MyoDnull mouse muscle were also unresponsive to rhBMP-2. In both closed and open fracture models in Nf1(flox/flox) and Nf1(flox/-) mice, local AdCre injection significantly impaired bone healing, with fracture union being <50% that of wild type controls. No significant difference was seen between Nf1(flox/flox) and Nf1(flox/-) mice. Histological analyses showed invasion of the Nf1(null) fractures by fibrous and highly proliferative tissue. Mean amounts of fibrous tissue were increased upward of 10-fold in Nf1(null) fractures and bromodeoxyuridine (BrdU) staining in closed fractures showed increased numbers of proliferating cells. In Nf1(null) fractures, tartrate-resistant acid phosphatase-positive (TRAP+) cells were frequently observed within the fibrous tissue, not lining a bone surface. In summary, we report that local Nf1 deletion in a fracture callus is sufficient to impair bony union and recapitulate histological features of clinical CPT. Cell culture findings support the concept that Nf1 double inactivation impairs early osteoblastic differentiation. This model provides valuable insight into the pathobiology of the disease, and will be helpful for trialing therapeutic compounds.