Introduction:​ Osteocyte apoptosis has been linked to bone resorption resulting from estrogen depletion and other resorptive stimuli;​ however, precise spatial and temporal relationships between the two events have not been clearly established. The purpose of this study was to characterize the patterns of osteocyte apoptosis in relation to bone resorption following ovariectomy to test whether osteocyte apoptosis occurs preferentially in areas known to activate resorption. Moreover, we report that osteocyte apoptosis is necessary to initiate endocortical remodeling in response to estrogen withdrawal.

Materials and methods:​ Adult female C57BL/6J mice (17 weeks old) underwent either bilateral ovariectomy (OVX), or sham surgery (SHAM) and were euthanized on days 3, 7, 14, or 21 days after OVX. Diaphyseal cross-sections were stained by immunohistochemistry for activated caspase-3 as a marker of apoptosis. The percentages of caspase-positive stained osteocytes (Casp + Ot.) were measured along major and minor anatomical axes around the femoral diaphysis to evaluate the distribution of osteocyte apoptosis after estrogen loss;​ resorption surface was measured at the adjacent endocortical regions. In a second study to test whether osteocyte apoptosis plays a regulatory role in the initiation of bone resorption, a group of OVX mice received the pan-caspase inhibitor, QVDOPh, to inhibit osteocyte apoptosis. Remaining experimental and sham groups received either QVD or Vehicle.

Results:​ OVX increased osteocyte apoptosis in a non-uniform distribution throughout the femoral diaphyses. Increases in Casp + osteocytes were predominately located in the posterior diaphyseal cortex. Here, the number of apoptotic osteocytes 4- to 7-fold higher than sham controls (p < 0.005) by day 3 post-OVX and remained elevated. Increases in resorption post-OVX also occurred along the posterior endocortical surface overlying the region of osteocyte apoptosis, but these increases occurred only at 14 and 21 days post-OVX (p < 0.002) well after the increases in osteocyte apoptosis. Treatment with QVD in OVX animals suppressed osteocyte apoptosis, with levels in QVD-treated samples equivalent to baseline. Moreover, the increases in osteoclastic resorption normally observed after estrogen loss did not occur in OVX mice treated with QVD.

Conclusions:​ The results of this study demonstrate that osteocyte apoptosis following estrogen loss occur regionally, rather than uniformly throughout the cortex. We also showed that estrogen loss increased osteocyte apoptosis. Apoptotic osteocytes were overwhelmingly localized within the posterior cortical region, the location where endocortical resorption was subsequently activated in ovariectomized mice. Finally, the increases in osteoclastic resorption normally observed after estrogen withdrawal did not occur in the absence of osteocyte apoptosis indicating that this apoptosis is necessary to activate endocortical remodeling following estrogen loss.