Alendronate Alters the Release of EVs by Raw 264.7 Osteoclasts

Abstract

Alendronate (ALN), a nitrogen-containing bisphosphonate, is widely used to treat bone disorders. While its inhibitory effect on osteoclast activity is well-established, its impact on the release of extracellular vesicles (EVs) is less understood. This study investigated the effect of ALN on the quantity and size distribution of EVs released by osteoclasts cultured on bovine bone slices pretreated with 10-µM ALN, 100-µM ALN, or vehicle. Raw 264.7 cells were differentiated into osteoclasts using RANK-ligand, and EVs were isolated from conditioned media. Tartrate-resistant acid phosphatase (TRAP) staining, phalloidin staining for actin rings, and nanoparticle tracking analysis (NTA) were performed. TRAP staining showed a significant reduction in the number of TRAP-positive multinucleated cells in the 100-µM ALN group, confirming that high-concentration ALN also impairs osteoclast formation. Phalloidin staining showed a significant decrease in actin ring formation in the 100-µM ALN group, confirming ALN’s inhibitory effect on osteoclast activity. NTA revealed a lower total EV concentration in the 100-µM ALN group, with a distinct peak of smaller EVs (<100 nm), suggestive of exosomes. These findings indicate that ALN, especially at higher concentrations, alters the release profile of osteoclast-derived EVs, potentially affecting intercellular communication and bone remodeling beyond its direct inhibition of resorption.

Introduction

Bisphosphonates (BPs) are drugs widely used to treat bone disorders, including osteoporosis, osteopenia, Paget’s disease, and others. One of the most prescribed BP is alendronate (ALN), a nitrogen-containing type of BP marketed as Fosamax.^1,2 ALN acts by inhibiting farnesyl pyrophosphate synthase (FPPS), an important metabolite at mevalonate pathway. The inhibitory effects of this pathway change the prenylation of small GTPases, which are essential to osteoclast activation to resorb bone. Key elements of this are rearrangement of the cytoskeleton into actin rings and the formation of ruffled membranes, which are the primary machinery of the resorption apparatus. ³ This resorption apparatus is required for bone resorption. Therefore, BP promotes inactivation of the clastic cells. ⁴

The function of ALN and other BP as therapeutics is also aided because the pyrophosphate component that is characteristic of BPs quickly targets these agents to areas of bone that are undergoing resorption. It remains in the bone until osteoclasts activate to resorb bone, at which point the BP is released so that it can inhibit the osteoclast. ⁵ BPs remain in bone for extended periods of time. Besides their beneficial effect in reducing bone resorption by osteoclasts, BPs have also been linked to rare but severe off-target effects. Medicine-related osteonecrosis of the jaw ⁶ has been linked to the use of ALN and other BPs, as has atypical femoral fractures. ⁷

Although osteoclasts’ primary function is to resorb bone, they also contribute to intercellular signaling, particularly in regulating osteoblasts, the bone-forming cells. ⁸ Various signaling agents that released from osteoclasts have been identified, and some have been identified as coupling factors, regulatory molecules that promote the balance between bone resorption by osteoclasts and bone formation by osteoblasts.

Extracellular vesicles (EVs) that contain receptor activator of nuclear factor kappa B (RANK) are one type of coupling factor released by osteoclasts.^9,10 EVs are release by most or all cells of the body and some serve as intracellular regulatory molecules. There are three types of EVs: microvesicles, formed by budding directly from the plasma membrane; exosomes, an intraluminal type of vesicle derived from multivesicular bodies; and apoptotic bodies. ¹¹ Several functions have been related to EV-mediated intercellular communication, both in health and disease. ¹² RANK-EVs may competitively inhibit the interaction between RANK-ligand (RANKL) and RANK of the osteoclast’s surface, which is required for osteoclast formation and activity. It has also been shown that RANK-containing EVs bind to RANKL on the surface of osteoblasts and trigger their differentiation toward bone formation. ⁹

Other molecules in EVs reported to act on bone physiology are microRNAs (miRNAs). ¹³ It was shown that osteoclast-derived EVs enriched with miRNA-324 were able to stimulate osteogenic differentiation. ¹⁴ Also, studies have shown that EVs derived from osteoclasts containing miRNA-23a-5p92 ¹⁵ and miRNA-214-3p ¹⁶ inhibited osteoblast activity.

While the impact of BPs on the osteoclast cytoskeleton is well-established, their effects on the release of signaling molecules, including EVs, are less understood. Recent data show that the composition of EVs changes significantly depending on whether osteoclasts are on plastic, bone, or dentine. It is reasonable to expect that BPs would trigger alterations in signaling molecules released by osteoclasts and perhaps other cell types in the bone microenvironment. Our study investigated the effect of ALN on the quantity of EVs released by osteoclasts in vitro. We hypothesized that ALN treatment would alter EV release, potentially contributing to both its therapeutic and adverse effects.

Methods

Preparation of Bone Substrates

Bovine cortical bone was sectioned into slices with a thickness of 200 μm and a surface area of approximately 2 cm² using a low-speed diamond saw (Buehler, Lake Bluff, IL). To better simulate the in vivo environment, these bone slices were used as substrates for cell culture. First, the bone slices were decontaminated by immersion in α-MEM media containing 1% penicillin-streptomycin. Subsequently, the decontaminated slices were incubated overnight in solutions containing 10-µM ALN, 100-µM ALN, or vehicle (PBS). These concentrations were previously determined considering the total amount of ALN that the bone substrate was able to absorb. ¹⁷

Generation of Osteoclasts

Raw 264.7 cells (American Type Culture Collection, Manassas, VA), were cultivated in six-well plate at 1.8 × 10⁵ cells per well over pretreated bone substrate. The cells were stimulated to differentiate into osteoclast-like cells with 5 ng/ml of RANKL ¹⁸ in an α-MEM media (Sigma-Aldrich, St. Louis, MO) plus 10% exosome-free fetal bovine serum (Systems Biosciences, Palo Alto, CA), 1% of glutamine (Thermo Fisher Scientific, Waltham, MA), and 1% penicillin/streptomycin. The cells were cultured for 5 days in a CO₂ incubator at 37C and medium was replaced on the third day. All tests were performed in triplicate.

TRAP Histochemistry

After 5 days of culture, the remaining cells around the bone substrate were fixed in 2% formaldehyde, washed with PBS, permeabilized in 1% Triton X-100 and stained for TRAP assay (Sigma-Aldrich, St. Louis, MO) following the manufacturer’s instructions. Osteoclasts were identified as TRAP-positive cells and counted in four random locations per well (1000 µm × 1000 µm square) by a blinded examiner. Cells were counted and then classified according to the number of nuclei: multinucleated (2–5 nuclei) and giant (>6 nuclei).

Phalloidin Staining for Actin Rings Detection

After 5 days of culture, cells adhered to bone substrates were fixed in 2% formaldehyde, washed with PBS, permeabilized in 1% Triton X-100 and stained with 10 μg/ml of phalloidin-FITC (Sigma-Aldrich, St. Louis, MO) for 30 minutes. The bone slices were washed with PBS and examined with Zeiss Axioplan II microscope (Zeiss, Oberkochen, Germany). Actin rings were counted in five different areas in each bone substrate.

EV Isolation

EVs were isolated as previously described.^10,19 The conditioned media from day 3 to day 5 was removed, dispensed in a 15-ml sterile conical tube, and centrifuged for 30 minutes at 3000 × g to exclude cells and cell debris. The supernatant was carefully transferred to new 15 ml sterile tubes. The samples were incubated overnight with 1:5 ratio of ExoQuick TC solution (System Biosciences, Palo Alto, CA) at 4C with gentle agitation, following manufacturer’s protocol. The next day the tubes were centrifuged for 30 minutes at 3000 ×g, the supernatant discarded, and the pellet was resuspended in 50 µl of sterile PBS. These isolated EVs were stored in a −80C freezer.

Nanoparticle Tracking Analysis

Previously isolated EVs were analyzed with nanoparticle tracking analysis (NTA) method using a NanoSight NS-300 (Malvern Panalytical, Malvern, UK). This method allows a quantitative analysis of concentration and diameter size of the EVs. The samples were diluted in PBS in 1:20 ratio and the following parameters were established in the device: image capture at 25 frames per second, temperature between 21C and 25C and video recording time for 60 s. The videos were analyzed using NanoSight NTA software (Malvern Panalytical) in raw format in triplicate.

Statistical Analysis

All quantitative data are presented as mean ± standard deviation (SD) or standard error of the mean (SEM), as indicated in the figure legends. Statistical analyses were performed using Minitab 19 (Minitab LLC, State College, PA). Comparisons between the experimental groups were performed using a one-way analysis of variance (ANOVA) followed by a Tukey’s post hoc test. A p value of <0.05 was considered statistically significant.

Results

TRAP Histochemistry

After 5 days of culture, cells were stained for TRAP to identify and quantify multinucleated and giant osteoclasts (Fig. 1). Quantitative analysis revealed a dose-dependent reduction in the number of TRAP-positive cells. This reduction was statistically significant in the 100-µM ALN group when compared with the control group (p<0.05). No significant difference was observed between the control and 10-µM ALN groups, nor between the 10-µM and 100-µM ALN groups.

Figure 1.

TRAP staining of osteoclasts differentiated from Raw 264.7 cells. Representative images show TRAP-positive multinucleated and giant cells in (A) control, (B) 10-µM alendronate, and (C) 100-µM alendronate group. (D) Quantification of TRAP-positive cells per area. A significant reduction in the number of osteoclasts was observed in the 100-µM ALN group compared with the control group (p<0.05). Data are presented as mean ± SD. Scale bar = 100 µm.

Phalloidin Staining for Actin Rings Detection

Clastic cells that adhered to the bone substrate were stained with phalloidin-FITC. Actin rings were identified and counted (Fig. 2). The data were submitted to ANOVA test and Tukey post-test with 5% significance. We found a decrease of actin rings formation (p<0.05) in the 100-µM ALN group, when compared with control and 10-µM ALN group.

Figure 2.

Actin rings identification of Raw 264.7 cells stimulated with RANKL over bone substrate. In all groups were observed actin rings after phalloidin staining (A = control; B = alendronate 10 µM; C = alendronate 100 µM). In (D), graphs show the number of actin rings per bone area. A significant decrease of actin rings was observed in the alendronate 100-µM group (p<0.05). Scale bar = 100 µm.

NTA

EVs isolated from conditioned media were analyzed by NTA (Fig. 3). The size distribution profiles showed that while all groups had a predominant population of EVs between 100 and 200 nm, the 100-µM ALN group displayed an additional, distinct peak of EVs smaller than 100 nm, consistent with the size of exosomes (Fig. 3B). This shift in distribution resulted in a dose-dependent decrease in the mean particle size, measuring 227.9 ± 4.9 nm for the control group, 183.6 ± 1.8 nm for the 10-µM ALN group, and 173.4 ± 2.2 nm for the 100-µM ALN group. Furthermore, the lowest total concentration of particles was also observed in the 100-µM ALN group.

Figure 3.

Size and concentration of osteoclast-derived EVs. Representative nanoparticle tracking analysis (NTA) profiles showing particle size distribution and concentration for each experimental group (A = control; B = alendronate 10 µM; C = alendronate 100 µM). Alendronate treatment caused a dose-dependent decrease in the mean particle size (control: 227.9 ± 4.9 nm; 10-µM ALN: 183.6 ± 1.8 nm; 100-µM ALN: 173.4 ± 2.2 nm). Note the emergence of a sub-100-nm particle population in the 100-µM ALN group and the different scales on the y-axis, indicating a lower total particle concentration in this group. Data for mean size are presented as mean ± standard error of the mean (SEM).

Discussion

This study showed that ALN treatment alters the release of EVs by osteoclasts, particularly at higher concentrations. Our findings confirm the inhibitory effects of ALN on osteoclasts in a dose-dependent manner. First, TRAP staining revealed a significant reduction in the number of differentiated, multinucleated, and giant osteoclasts in the 100-µM ALN group, indicating that ALN impairs osteoclast formation or survival. Furthermore, for the cells that did form, phalloidin staining showed a significant decrease in actin ring formation, consistent with the established mechanism of ALN, which inhibits osteoclasts by disrupting cytoskeletal rearrangements. ⁴ Together, these results demonstrate that ALN released from the bone substrate reduces both the number and the functional activity of osteoclasts. It is important to remember that the effects of ALN are the result of the amounts locally released from the bone slices. For example, 100-µM free ALN would quickly kill the osteoclasts. However, we had showed previously ¹⁷ that at 100 µM there was no detectable leakage of ALN from coated bone slices. The amount released from the bone was sufficient to perturb actin ring formation, but living cells were present.

Actin rings are formed through the polymerization of microfilaments, creating filamentous actin (F-actin). ²⁰ The formation of actin rings reflects osteoclast activation. Treatment with nitrogen-containing BP, among other actions, affects the prenylation of small GTPases such as Rho, Rac, and Rab. These GTPases play a role in osteoclast cytoskeletal organization, thus impairing actin ring formation. BPs also impact cell adhesion, migration, polarization, and intracellular vesicular trafficking. ⁴

It is well known that ALN disturbed intracellular vesicle transportation. ²¹ However, NTA analysis revealed alterations in the EV profile released by osteoclasts treated with ALN. While all groups showed a predominant population of EVs within the 100–200 nm range, the 100-µM ALN group displayed a unique peak of smaller EVs (<100 nm), consistent with the size of exosomes (ranging between 30 and 150 nm) ²² and a lower overall concentration of EVs.

This suggests that ALN, at higher concentrations, may not only inhibit osteoclast activity but also modulate the biogenesis and/or release of EVs. The reduced overall EV concentration in this group could be attributed to impaired vesicle trafficking or release mechanisms within the osteoclast. More studies will be necessary to better understand whether the drug directly influences EV release or whether the observed changes are secondary to its effects on cytoskeletal organization and intracellular vesicular trafficking.

The observed changes in EV release profiles are particularly relevant considering the role of osteoclast-derived EVs, especially RANK-containing EVs, in intercellular communication and bone remodeling. RANK-EVs have been shown to influence both osteoclast and osteoblast activity, acting as a coupling factor between bone resorption and formation. Therefore, the alteration in EV release by ALN could have indirect effects on bone remodeling beyond the direct inhibition of osteoclast resorption.

The culture system described here, where osteoclasts are plated on ALN-coated bone slices, provides a model for studying BP’s effects on osteoclasts and other bone cells in a setting that is similar to the in vivo situation. BP pharmaceuticals like ALN quickly absorb to exposed bone in sites of active remodeling when given to patients. The BP is then mobilized as osteoclasts begin to resorb the bone releasing BP to act on the resorbing osteoclast. This may not only inhibit the osteoclast, for example, by disrupting actin rings and thus disrupting resorption, but also may alter signaling molecules released by the osteoclasts, including regulatory EVs. Such secondary effects of BP-treated bone have not been extensively studied. It is known that EV composition is altered based on the substrate they are plated on. Significant differences exist in the composition of EVs released by osteoclasts resorbing bone, dentine, or setting passively on plastic.^19,23

We acknowledge that a limitation of our study is the use of a single cell line, Raw 264.7. Although this is a well-established model, future research should aim to corroborate these findings in other osteoclast models, such as primary bone marrow macrophages or human CD14+-derived osteoclasts, to confirm the broader applicability of these mechanisms. Future research should also characterize the specific EV populations affected by BPs, as ALN, especially the levels of RANK-containing EVs. Given the role of osteoclast-derived EVs, particularly RANK-containing EVs, in intercellular communication and bone remodeling, these changes in EV release profile could have significant implications. RANK-EVs influence both osteoclast and osteoblast activity, acting as a crucial coupling factor between bone resorption and formation. Therefore, the alteration in EV release by ALN could have indirect effects on bone remodeling beyond its direct inhibition of osteoclast resorption. Further investigation into the functional consequences of these alterations on bone remodeling and their potential clinical implications is warranted. In conclusion, this study demonstrates that bone-bound ALN not only inhibits osteoclast activation and reduces the number of differentiated osteoclasts in a dose-dependent manner, but also significantly modulates EV release by osteoclasts. Specifically, high-concentration ALN treatment decreased the total concentration of EVs. These findings suggest that the biological effects of ALN extend beyond direct cellular inhibition to include the modulation of intercellular signaling, a mechanism that may contribute to both its therapeutic and potential adverse effects.