We therefore hypothesized that the balance between the rate of collagenolysis and demineralization might serve as a mechanism determining the duration of a resorption event, and thereby also the excavation geometry. A definitive demonstration of this hypothesis requires testing the effect of direct and specific inhibitors of either mineral solubilization or collagen degradation, on the resorption pattern of OCs. We used inhibitors of CatK to slow down the relative rate of collagen degradation compared to the rate of mineral solubilization

[18], [19] and [20], and we used low concentrations of a carbonic anhydrase inhibitor to increase the relative rate of collagen degradation compared to mineral solubilization [21]. Thus, as illustrated in Fig. 1, 3-Methyladenine price according to our hypothesis, CatK inhibitors should accelerate the accumulation of collagen in the resorption pit thereby leading to early termination of the local resorption event and a shallower pit. In contrast, mild inhibition of carbonic anhydrase should allow collagenolysis to proceed as fast as demineralization, thereby ensuring continuation of the local resorption event, thus promoting the formation of trenches at

the expense of round pits. The following inhibitors of OC resorption were used: 6-ethoxyzolamide (Sigma-Aldrich, Broendby, Denmark), specific inhibitor of carbonic anhydrase, 20 mM stock in DMSO, stored at − 20 °C; E64 (Sigma-Aldrich), cysteine-protease Doxorubicin clinical trial inhibitor, 1 mM stock in H2O, stored at − 20 °C; L873724, an inhibitor specific of CatK [20], [22] and [23] (a generous gift from MSD, Rahway, USA), 10 mM stock in DMSO (Sigma-Aldrich) stored at − 20 °C. Human CD14+ cells were isolated from buffy coats of healthy volunteers (approved by the local ethics committee, 2007-0019) and differentiated into multinucleated OCs through the use of 25 ng/ml M-CSF and 25 ng/ml RANKL (R&D, Abingdon, England, UK) as described previously [17]. Differentiated

OCs were re-seeded on bovine cortical bone slices adapted for 96-well plates (IDS Nordic, Herlev, Denmark) Niclosamide at a density of 50,000 to 100,000 cells per bone slice, and cultured for 72 h in the presence or not of various resorption inhibitors at the indicated concentrations. DMSO was added at a final concentration of 0.2% to controls when relevant. The resorption features (i.e. cavitations as well as superficial demineralization patches) were stained with toluidine blue as described previously [17] and analyzed through light microscopy. Resorbed bone surface area, number of resorption cavities and maximal erosion depth measurements were measured as previously described [17]. A resorption feature with a continuous and distinct perimeter at the surface was counted as one.