RANK-dependent autosomal recessive osteopetrosis: characterization of five new cases with novel mutations.

Autosomal recessive osteopetrosis (ARO) is a genetically heterogeneous disorder attributed to reduced bone resorption by osteoclasts. Most human AROs are classified as osteoclast rich, but recently two subsets of osteoclast-poor ARO have been recognized as caused by defects in either TNFSF11 or TNFRSF11A genes, coding the RANKL and RANK proteins, respectively. The RANKL/RANK axis drives osteoclast differentiation and also plays a role in the immune system. In fact, we have recently reported that mutations in the TNFRSF11A gene lead to osteoclast-poor osteopetrosis associated with hypogammaglobulinemia. Here we present the characterization of five additional unpublished patients from four unrelated families in which we found five novel mutations in the TNFRSF11A gene, including two missense and two nonsense mutations and a single-nucleotide insertion. Immunological investigation in three of them showed that the previously described defect in the B cell compartment was present only in some patients and that its severity seemed to increase with age and the progression of the disease. HSCT performed in all five patients almost completely cured the disease even when carried out in late infancy. Hypercalcemia was the most important posttransplant complication. Overall, our results further underline the heterogeneity of human ARO also deriving from the interplay between bone and the immune system, and highlight the prognostic and therapeutic implications of the molecular diagnosis.

Signal peptide mutations in RANK prevent downstream activation of NF-κB.

Familial expansile osteolysis and related disorders are caused by heterozygous tandem duplication mutations in the signal peptide region of the gene encoding receptor activator of NF-κB (RANK), a receptor critical for osteoclast formation and function. Previous studies have shown that overexpression of these mutant proteins causes constitutive activation of NF-κB signaling in vitro, and it has been assumed that this accounts for the focal osteolytic lesions that are seen in vivo. We show here that constitutive activation of NF-κB occurred in HEK293 cells overexpressing wild-type or mutant RANK but not in stably transfected cell lines expressing low levels of each RANK gene. Importantly, only cells expressing wild-type RANK demonstrated ligand-dependent activation of NF-κB. When overexpressed, mutant RANK did not localize to the plasma membrane but localized to extensive areas of organized smooth endoplasmic reticulum, whereas, as expected, wild-type RANK was detected at the plasma membrane and in the Golgi apparatus. This intracellular accumulation of the mutant proteins is probably the result of lack of signal peptide cleavage because, using two in vitro translation systems, we demonstrate that the mutations in RANK prevent cleavage of the signal peptide. In conclusion, signal peptide mutations lead to accumulation of RANK in the endoplasmic reticulum and prevent direct activation by RANK ligand. These results strongly suggest that the increased osteoclast formation/activity caused by these mutations cannot be explained by studying the homozygous phenotype alone but requires further detailed investigation of the heterozygous expression of the mutant RANK proteins.

The matricellular protein CYR61 inhibits osteoclastogenesis by a mechanism independent of alphavbeta3 and alphavbeta5.

Cysteine-rich protein 61 (CYR61/CCN1) belongs to the family of CCN matricellular proteins. Most of the known effects of CCN proteins appear to be due to binding to extracellular growth factors or integrins, including alpha(v)beta(3) and alpha(v)beta(5). Although CYR61 can stimulate osteoblast differentiation, until now the effect of CYR61 on osteoclasts was unknown. We demonstrate that recombinant human CYR61 inhibits the formation of multinucleated, alpha(v)beta(3)-positive, or tartrate-resistant acid phosphatase-positive human, mouse, and rabbit osteoclasts in vitro. CYR61 markedly reduced the expression of the osteoclast phenotypic markers tartrate-resistant acid phosphatase, matrix metalloproteinase-9, calcitonin receptor, and cathepsin K. However, CYR61 did not affect the formation of multinucleated osteoclasts when added to osteoclast precursors prior to fusion or affect the number or resorptive activity of osteoclasts cultured on dentine discs, indicating that CYR61 affects early osteoclast precursors but not mature osteoclasts. CYR61 did not affect receptor activator of nuclear factor-kappaB (RANK) ligand-induced phosphorylation of p38 or ERK1/2 in human macrophages and did not affect RANK ligand-induced activation of nuclear factor-kappaB, indicating that CYR61 does not appear to inhibit osteoclastogenesis by affecting RANK signaling. Furthermore, a mutant form of CYR61 defective in binding to alpha(v)beta(3) also inhibited osteoclastogenesis, and CYR61 inhibited osteoclastogenesis similarly in cultures of mouse wild-type or beta(5)(-/-) macrophages. Thus, CYR61 does not appear to inhibit osteoclast formation by interacting with alpha(v)beta(3) or alpha(v)beta(5). These observations demonstrate that CYR61 is a hitherto unrecognized inhibitor of osteoclast formation, although the exact mechanism of inhibition remains to be determined. Given that CYR61 also stimulates osteoblasts, CYR61 could represent an important bifunctional local regulator of bone remodeling.