Abstract

Neurofibromatosis type I (NF1) is a congenital disorder resulting from loss-of-function of the tumor suppressor gene, NF1. 50% of NF1 patients have osseous manifestations, including short stature, scoliosis, and reduced bone mineral density. Osteoclasts are hematopoietic stem cell-derived cells that function to resorb bone. We recently reported that osteoclasts derived from NF1 patients and Nf1 heterozygous (Nf1+/−) mice have elevated migration, adhesion, and bone resorption and our studies indicate that the gain-of-function of the Nf1+/− osteoclasts is, at least in part, caused by hyperactivation of macrophage colony-stimulating factor (M-CSF)-stimulated Ras, phosphoinositol-3-kinase (PI3K), and Erk. Rho GTPases function downstream of Ras and PI3K and act as binary switches, cycling between an inactive (GDP-bound) and active (GTP-bound) state, to regulate osteoclast actin ring formation, bone resorption, and development of filamentous actin structures associated with migration and adhesion. We hypothesized that hyperactivation of Rac1, Rac2, or both Rac1 and Rac2 contribute to the increased osteoclast function observed in Nf1+/− mice and NF1 patients. To examine this hypothesis, we intercrossed Nf1+/− mice with conditional Rac1flox/floxMxcre+ mice or with Rac2/− mice to generate WT, Nf1+/, Rac1/, Nf1+/;Rac1/, Rac2/, and Nf1+/;Rac2/− mice. Genetic disruption of Rac1, but not of Rac2, restored the increased colony forming unit-macrophage (CFU-M), tatrate resistant acid phosphate+ (TRAP+) CFU-M, osteoclast migration, and bone resorption observed in Nf1+/− cultures. Osteoclast bone resorbing capacity is dependent on the organization of the actin cytoskeleton into a large f-actin-rich structure referred to as the sealing zone. The podosome belt evolves into the sealing zone in actively resorbing osteoclasts. A significantly higher level of belt formation, seen in mature osteoclasts, was observed in Nf1+/− cultures as compared to WT. Upon genetic deletion of Rac1, the Nf1+/;Rac1/− osteoclasts demonstrated belt formation at a similar level to that of WT osteoclasts. These data indicate that Rac1 plays an essential role in functional f-actin organization and suggest that inhibition of Rac1 in the setting of Nf1 haploinsufficiency is able to normalize osteoclast hyperactivity by correcting the cytoskeletal organization of f-actin-based structures. Mechanistically, Rac1 deficiency normalized M-CSF-stimulated phospho-Erk and phospho-Akt and pharmacologic inhibition of MEK and PI3K using PD98059 or Ly294002, respectively, normalized Nf1+/− osteoclast development and maturation. The critical role of Rac1, but not of Rac2, in osteoclast function is significant as it suggests that the Rac GTPases contribute non-redundant functions in various myeloid cell types and imply that blocking Rac1 function, while sparing that of Rac2, may provide a level of specificity to therapeutics for skeletal diseases. Collectively, these data demonstrate that Rac1 critically contributes to increased osteoclast function induced by haploinsufficiency of Nf1 and imply that Rac1 may be a rational therapeutic target for dysplastic and erosive bone diseases.

Author notes

Disclosure: No relevant conflicts of interest to declare.