The bone resorption inhibitors odanacatib and alendronate affect post-osteoclastic events differently in ovariectomized rabbits

Abstract

Odanacatib (ODN) is a bone resorption inhibitor which differs from standard antiresorptives by its ability to reduce bone resorption without decreasing bone formation. What is the reason for this difference? In contrast with other antiresorptives, such as alendronate (ALN), ODN targets only the very last step of the resorption process. We hypothesize that ODN may therefore modify the remodeling events immediately following osteoclastic resorption. These events belong to the reversal phase and include recruitment of osteoblasts, which is critical for connecting bone resorption to formation. We performed a histomorphometric study of trabecular remodeling in vertebrae of estrogen-deficient rabbits treated or not with ODN or ALN, a model where ODN, but not ALN, was previously shown to preserve bone formation. In line with our hypothesis, we found that ODN treatment compared to ALN results in a shorter reversal phase, faster initiation of osteoid deposition on the eroded surfaces, and higher osteoblast recruitment. The latter is reflected by higher densities of mature bone forming osteoblasts and an increased subpopulation of cuboidal osteoblasts. Furthermore, we found an increase in the interface between osteoclasts and surrounding osteoblast-lineage cells. This increase is expected to favor the osteoclast-osteoblast interactions required for bone formation. Regarding bone resorption itself, we show that ODN, but not ALN, treatment results in shallower resorption lacunae, a geometry favoring bone stiffness. We conclude that, compared to standard antiresorptives, ODN shows distinctive effects on resorption geometry and on reversal phase activities which positively affect osteoblast recruitment and may therefore favor bone formation.

Fig. 1

Effect of odanacatib (ODN) and alendronate (ALN) treatment on bone resorption parameters in ovariectomized (OVX) rabbits. The extent of eroded surface (ES/BS) and osteoclast surface (Oc.S/BS) was assessed in the vertebral trabecular bone of sham and OVX rabbits treated with vehicle (Veh), ALN, or ODN. The results are shown as means with SD a or as medians with upper and lower quartiles b. Differences between the groups were analyzed by the Mann–Whitney test b or the t-test a, applying Welch correction in the presence of unequal group variances. *p < 0.05, **p < 0.01, ***p < 0.001

Fig. 2

Histological appearance of osteoclasts in odanacatib (ODN)– and alendronate (ALN)–treated ovariectomized (OVX) rabbits. a Masson-Goldner trichrome staining shows the appearance of osteoclast profiles (asterisks) attached to the vertebral trabecular bone surface from a vehicle-treated (left panel) and an ALN-treated (middle panel) OVX rabbit. Osteoclast profiles frequently appeared away from the bone surface in ALN-treated OVX rabbits as shown by the presence of a giant osteoclast (asterisk) above bone forming osteoblasts (right panel). b Examples of osteoclast profiles (asterisks) detached from the bone surface by mononucleated reversal cells (Rv.C, arrows) in an ODN-treated OVX rabbit as they appear using Masson-Goldner trichrome staining (left panel), TRAP activity staining (middle panel), and electron microscopy (right panel). c TRAP-positive osteoclast profiles (asterisks) attached to a bone surface without visible broken lamellae. The framed area in the left panel is enlarged in the right panel. d Example of an osteoclast profile (asterisk) with demineralized collagen below (arrowheads), visualized by Masson-Goldner trichrome staining. e Toluidine blue staining showing the absence and presence of intracellular vesicles in osteoclast profiles (asterisks) in the vertebral trabecular bone from an ALN-treated (left panel) and an ODN-treated (right panel) OVX rabbit. f Immunohistochemical staining showing elevated intracellular levels of cathepsin K in trabecular osteoclasts (asterisks) of ODN-treated OVX rabbits (right panel) compared to ALN-treated (middle panel) and vehicle-treated (left panel) OVX rabbits. Scale bar = 50 μm, except for b, right panel = 10 μm and c, left panel = 200 μm

Fig. 3

Effects of odanacatib (ODN) and alendronate (ALN) treatment on osteoclast features. a The extent of osteoclast profiles detached from the bone surface was evaluated in the vertebral trabecular bone of sham and ovariectomized (OVX) rabbits treated with vehicle (Veh), ALN, or ODN. Data are expressed as the percentage of detached osteoclast profiles per total number of osteoclast profiles. b Among the osteoclast profiles detached from the bone surface, some covered tightly reversal cells (Rv.Cs). The fraction of this population of osteoclasts is expressed as percentage of the total detached osteoclasts. c Percentage of osteoclast profiles attached onto bone surfaces with an erosion depth of ≤2 lamellae. d Percentage of osteoclast profiles with demineralized collagen below. e Percentage of osteoclast profiles with intracellular vesicles. Results are shown as medians with upper and lower quartiles a or as means with SD b–e. Rabbits with fewer than 20 osteoclast profiles were excluded from the analyses. Differences between groups were analyzed by the Mann–Whitney test a or the t-test b–e, applying the Welch correction in the presence of unequal group variances. *p < 0.05, **p < 0.01, ***p < 0.001

Fig. 4

Effect of odanacatib (ODN) and alendronate (ALN) treatment on the reversal phase. a The extent of reversal surface (Rv.S/BS) was assessed in the trabecular bone of a vertebra of sham and OVX rabbits treated with vehicle (Veh), ALN, or ODN. b The Rv.S shown in a was plotted against the Oc.S/BS shown in Fig. 1 in order to evaluate the relationship between osteoclast and reversal phase prevalence for each rabbit. Each dot represents the measurement obtained in a bone specimen isolated from one rabbit (sham: black circles, black dotted line; OVX + Veh gray open squares, gray dotted line; OVX + ALN gray diamonds, gray line; OVX + ODN black triangles, black line). c The proximity of the reversal surfaces to formative surfaces is shown for each group as the percentage of Rv.S with OS in the vicinity as defined in “Materials and Methods.” d The density of mononucleated cells on the reversal surfaces, i.e., reversal cells (Rv.C), is shown for each of the four groups of rabbits in cell profiles per millimeter. Results are shown as means with SD a, d or medians with upper and lower quartiles c. The Mann–Whitney c or the t-test a, d was used to analyze whether the parameters were significantly different in the four groups, and correlations between the extent of Oc.S/BS and Rv.S/BS b were analyzed using the Spearman rank correlation test, *p < 0.05, **p < 0.01, ***p < 0.001. The lines in b were made by linear regressions. Comparison of the slopes of the regression lines for sham, OVX + Veh, and OVX + ALN revealed no differences. Therefore, the slope of the best-fitted regression line for the pooled groups (line not shown, Spearman rank correlation for pooled groups r _s = 0.80, p < 0.0001) was compared with the slope of the regression line for OVX + ODN and proved to be significantly different (p = 0.0012)

Fig. 5

Effect of odanacatib (ODN) and alendronate (ALN) treatment on bone formation parameters. The extent of osteoid surface (OS/BS) a, total osteoblast surface (Ob.S/BS) b, and cuboidal osteoblast surface (C.Ob.S/BS) c was evaluated in a vertebra from the sham and OVX rabbits treated with vehicle (Veh), ALN, or ODN. d The density of osteoblasts (cell profiles per millimeter) is shown for each of the four groups. Results are shown as medians with upper and lower quartiles a–c or means with SD d. The Mann–Whitney a–c or the t-test d was used to analyze whether the parameters were significantly different in the four groups. *p < 0.05, **p < 0.01