Zoledronic Acid Is Not Equally Potent on Osteoclasts Generated From Different Individuals

Abstract

Zoledronic acid is a bisphosphonate commonly used to treat bone diseases such as osteoporosis and cancer-induced bone disease. Patients exhibit a variable sensitivity to zoledronic acid; the underlying explanation for this remains unclear. The objective of this study was to obtain more knowledge in this regard. We hypothesized that osteoclasts generated from different individuals would show a variable sensitivity to zoledronic acid in vitro. Osteoclasts were generated using monocytes from 46 healthy female blood donors (40 to 66 years). Matured osteoclasts were reseeded onto bone slices precoated with different concentrations of zoledronic acid. IC50 values were determined based on total eroded bone surface after 3 days of resorption. The IC50 for inhibition of osteoclastic bone resorption varied from 0.06 to 12.57μM zoledronic acid; thus, a more than 200-fold difference in sensitivity to zoledronic acid among osteoclasts from different individuals was observed. Multiple linear regression analyses showed that the determined IC50 correlated with smoking status, and the average number of nuclei per osteoclast in vitro. Further analyses showed that: (i) increasing protein levels of mature cathepsin K in osteoclast cultures rendered the osteoclasts less sensitive to zoledronic acid; (ii) surprisingly, neither the gene nor the protein expression of farnesyl diphosphate synthase was found to correlate with the IC50; and (iii) trench-forming osteoclasts were found to be more sensitive to zoledronic acid than pit-forming osteoclasts within the same cell culture. Thus, we conclude that there indeed is a high degree of variation in the potency of zoledronic acid on osteoclasts when generated from different individuals. We propose that our findings can explain some of the varying clinical efficacy of zoledronic acid therapy observed in patients, and may therefore be of clinical importance, which should be investigated in a clinical trial combining in vitro and in vivo investigations. © 2020 The Authors. JBMR Plus published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research.

Keywords: ANTIRESORPTIVES; OSTEOCLASTS; OSTEOPOROSIS; TUMOR‐INDUCED BONE DISEASE.

Fig 1

Examples of responses to Zol for OCs generated from three different donors. Responses to Zol were determined by: (A) The IC50 value, based on total eroded bone surface after 3 days of resorption (n = 5). Each data point represents the mean ± SD; baseline is indicated by a dotted line. (B) The IC50 value, based on the concentration of CTx in the conditioned media, after 3 days of resorption (n = 3). Each data point represents the median ± range; baseline is indicated by the dotted line. (C) Tartrate‐resistant acid phosphatase (TRACP) activity (arbitrary units; n = 3); statistics: Kruskal‐Wallis test. Each data point represents the median + range. (D) Metabolic activity (arbitrary units; n = 3) statistics: Kruskal‐Wallis test. Each data point represents the median and range.

Fig 2

The sensitivity to Zol varies significantly among OC preparations (based on total eroded surface after 3 days of resorption). (A) Variation in IC50. (B) Variation in IC100. (C) Variation in maximum effect of Zol. How much (in percent) Zol could reduce baseline resorption levels. (D) Correlation between IC50 and IC100. (E) Correlation between IC50 and maximum effect. Statistical correlation analyses were performed using either Pearson's correlation (r ²) or Spearman's rank correlation (r _s). Each data point represents the results obtained from osteoclasts generated from an individual donor (n = 46); the red line represents the median.

Fig 3

Zol treatment decreases the metabolic activity of OCs in vitro, but not for all donors. (A) Metabolic activity of OCs at baseline compared to the metabolic activity of OCs at IC50 (based on total eroded surface). (B) Metabolic activity of OCs at baseline compared to the metabolic activity of OCs at IC100. (C) The percent change in metabolic activity from baseline to IC50 (based on total eroded surface). (D) The percent change in metabolic activity from baseline to IC100. ( E) Comparison of the IC50 value with the metabolic activity at IC50 (normalized to baseline; based on total eroded surface). (F) Comparison of the IC50 value with the metabolic activity at IC100 (normalized to baseline). Baseline is indicated by the dotted line. Statistical tests: (A) Wilcoxon matched‐pairs signed rank test, (B) paired t test, (C + D) One sample Wilcoxon test, and (E + F) Spearman's rank correlation (r _s). For all graphs, each data point represents the results obtained from OCs generated from an individual donor (n = 46).

Fig 4

OC activity and resorption mode (pits and trenches) affect the sensitivity to Zol. (A) Examples of pits (stipulated black box) and trenches (stipulated white box) on a cortical bovine bone slice stained with toluidine blue. Enlargement of these can be seen on the right. The black scale bar corresponds to 50 μm. (B) Correlation analysis between the total eroded surface at baseline and the IC50 value (based on total eroded surface) for each OC preparation. (C) Correlation analysis between the extent of pit surface/bone surface (BS) at baseline and the IC50 value (based on total eroded surface) for each OC preparation. (D) Correlation analysis between the amount of trench surface/BS at baseline and the IC50 value (based on total eroded surface) for each OC preparation. (E) Matching the IC50 value for pit and trench surfaces in the same culture (medians are indicated by horizontal lines). Statistical correlation analyses were performed using (B–D) Spearman's rank correlation (r _s) and (E) Wilcoxon matched‐pairs signed rank test. Each data point represents the results obtained from OCs generated from an individual donor (n = 46).

Fig 5

The sensitivity to Zol correlates with the amount of mature CatK in the cell, but not the amount of FDPS. Correlation analyses between the IC50 value (based on total eroded bone surface) for each OC preparation and: (A) CTSK gene expression, (B) FDPS gene expression, (C) the amount of mature CatK protein, and (D) the total amount of FDPS protein. Statistical correlation analyses were performed using Spearman's rank correlation (r _s). Each data point represents the results obtained from OCs generated from an individual donor (n = 45).