Cathepsin K (CatK) inhibition in preclinical models results not only in lower bone resorption but also in higher bone formation (BF) on both remodeling and modeling surfaces. The mechanisms for greater BF at modeling surfaces remain unexplained. In vitro data suggest that periostin is proteolytically degraded by CatK. Periostin (Postn) is a matricellular protein expressed in the periosteum and osteocytes that mediates β-catenin signalling and skeletal response to loading. We hypothesised that higher BF in Ctsk−/− mice might be related to increased periostin expression. For this purpose, BMD, microstructure and histomorphometry were evaluated in the progeny of a Postn+/− X Ctsk+/− mouse cross. Postn immunostaining was more intense in Ctsk−/− osteocytes and periosteum surfaces vs WT. Postn+/+;Ctsk−/− have higher BMD, BV/TV, CtBV, CtTh, whereas Postn−/−;Ctsk+/+ have lower femoral BMD, CtBV, and CtTh compared to WT. Postn+/+;Ctsk−/− have higher Ps-MAR and -BFR (+58% and +137% vs WT, P<0.05), as well as Ec-MAR (+74% vs WT, P<0.05) but lower Ec-MPm/BPm (−40% vs WT, P<0.05), indicating that BF is increased on both cortical envelopes and independently of remodeling surfaces. Removing both Postn alleles in Ctsk−/− mice, i.e. Postn−/−;Ctsk−/−, prevented the increase in CtBV and Ps-BFR, but had no effect on Ec-MPm/BPm which remained low (51% vs WT, P<0.05), indicating that Postn mediates cortical bone formation but has no influence on CatK dependent bone remodeling. In contrast, trabecular microarchitecture at the distal femur and vertebrae was not affected by Postn deletion in Ctsk−/− mice, indicating that Postn-dependent bone formation in these mice occurs specifically at cortical surfaces. In conclusions, periostin expression is increased in Cstk−/− mice and is responsible for the increased modelling-based bone formation observed on cortical bone in the absence of cathepsin K.
Disclosure: This work was supported by MSD.