Effects of Alendronate and Dexamethasone on Osteoclast Gene Expression and Bone Resorption in Mouse Marrow Cultures

Abstract

Osteoclasts are cells whose main function is the resorption of bone matrix. However, several factors, including medications, can interfere with the resorption process. Alendronate (ALN), a nitrogen-containing type of bisphosphonate, and dexamethasone (DEX), a glucocorticoid, are drugs that may affect the resorption activity. The aim of this study is to investigate the effects of ALN, and/or DEX on osteoclast gene expression and resorption activity in primary mouse marrow cultures stimulated with 1,25-dihydroxyvitamin D3, a model for the bone microenvironment. Cultures were treated only with ALN (10⁻⁵ M), DEX (10⁻⁶ M), and with a combination of both agents. Viability assays performed at days 5, 7, and 9 showed the highest number of viable cells at day 7. All the following assays were then performed at day 7 of cell culture: tartrate resistant acid phosphatase (TRAP) histochemistry, receptor activator of nuclear factor kappa B ligand (RANKL) immunofluorescence, osteoprotegerin (OPG), and RANKL gene expression by qPCR and resorption analysis by scanning electron microscopy. Treatment with ALN, DEX, and the combination of both did not promote significant changes in the number of TRAP+ cells, although larger giant cells were detected in groups treated with DEX. DEX treatment increased the gene expression of RANKL and reduced OPG. The treatment with ALN reduced the depth of the resorption pits, but their inhibitory effect was less effective when administered with DEX:

Introduction

Osteoclasts are highly specialized cells, formed by the fusion of mononuclear cells from hematopoietic origin. Control of osteoclast differentiation is regulated mainly by macrophage-colony stimulating factor and receptor activator of nuclear factor kappa B ligand (RANKL).^1,2 The primary function of osteoclasts is the resorption of bone matrix, which is composed of organic molecules and hydroxyapatite. Resorption of bone by osteoclasts and formation of new bone matrix by osteoblasts to replace the bone removed occurs throughout life and is known as bone remodeling. ³

Several factors can influence the resorptive processes, including medications. As bisphosphonates (BP) are drugs capable of inhibiting bone resorption, they are used to prevent and treat diseases that are associated with an increase in osteoclast activity, such as osteoporosis. Their structures are similar to those of inorganic pyrophosphate, a molecule that regulates bone biomineralization. ⁴ Alendronate (ALN) is a second-generation nitrogen-containing BP, being one of the most widely used drugs currently for the treatment of osteoporosis; its mechanism of action occurs through the inhibition of the mevalonate pathway, inactivating osteoclast differentiation and activity. ⁵ It has also been indicated that the administration of small doses of ALN can inhibit the release of cytokines such as RANKL, interfering with the recruitment, differentiation, and activity of clastic cells. ⁶

Glucocorticoids (GC) are drugs that have several effects on the immune response and are also anti-inflammatory and immunosuppressive. ⁷ Studies have shown that chronic treatment with these drugs can ultimately stimulate bone loss and cause osteoporosis, a phenomenon known as glucocorticoid-induced osteoporosis (GIOP).^8,9 The mechanism of action of GC on bone loss occurs both in the processes of neoformation and resorption. While bone formation is affected by the decrease in the number and activity of osteoblasts, ¹⁰ in resorption, it can promote the increase of proliferation and differentiation of osteoclastic precursors. ¹¹

Clinically, there are two situations that can lead to combined treatment with ALN and DEX. The first one is the use of BP as a treatment option for GIOP,^12,13 and the second, patients undergoing treatment for osteoporosis using BP, who for some period (either brief or chronic), need to use a GC for the purpose of controlling inflammatory or immune reactions. Considering that the mechanism of action of GC on osteoclasts and the combined effects of ALN and DEX are not fully established, the aim of this study was to elucidate the effects of ALN, and/or DEX on osteoclastogenesis and bone resorption activity in primary mouse marrow cultures.

Materials and Methods

Animal care and experimental procedures were performed following the Guidelines for Animal Experimentation of the University of São Paulo School of Dentistry, with approval from the Ethics Committee for Animal Research.

Preparation of Bone Substrates

Bone substrates were obtained from the same bovine femur cortical bone. Slices, 100 µm in thickness, were prepared using an IsoMet 1000 Precision Cutter (Buehler; Rockville, IN) and trimmed to produce a surface area of 1 cm². To better simulate the in vivo condition, the bone slices had the smear layer (resulting from the cutting process) removed. For this, the substrates were initially immersed in a 1% sodium hypochlorite solution for 30 min, followed by rinsing with ultrapure water (MilliQ) with ultrasonic agitation. Then, they were treated with 4.13% EDTA for 5 min, followed by another ultrasonic washing. Finally, they were immersed in a 1% sodium hypochlorite solution for 5 min, washed with ultrapure water with ultrasonic agitation, followed by an overnight period with ultrapure water and left to dry. These disks were sterilized with 1% penicillin/streptomycin solution and kept until use. The bone disks used in ALN groups were previously incubated overnight in ALN solutions (see below).

Saturation Analysis of Bone Substrates With ALN

To confirm the concentration of ALN to be used in the experiment, tests were carried out to detect how much BP was necessary to be fully incorporated into the bone substrate. For this, we detected ALN using a solution of O-phthalaldehyde (OPA) and mercaptoethanol (ME), which reacts with ALN, becoming detectable in a spectrophotometer at 333 nm. ¹⁴ ALN solutions were prepared with 1 mM, 100 µM, 10 µM, 1 µM, and 100 nM and then, the spectrophotometer reading was performed at 333 nm. The total volume used in each well was 1 ml (100 µl of ALN solution and 900 µl of OPA-ME solution). After reading at zero-time, bovine bone slices were incubated in these solutions so that ALN could be incorporated into the bone matrix. The solution reacted overnight, and then, the bone substrates were removed, and a new reading was performed to detect how much ALN was incorporated into the bone and how much remained in the solution.

Primary Cultures

For the primary cultures, 30-day-old Balb-c mice of both sexes were used as a source of cells. They were anesthetized with isoflurane (BioChimico; Itatiaia, RJ, Brazil) and euthanized by cervical dislocation. The tibia and femur were dissected, and the knee side epiphysis removed. The cells were obtained from the bone marrow, which was expelled from the spinal canal through centrifugation. Briefly, a small hole was made at a 0.5-ml microcentrifugation tube containing the bones; these tubes were placed inside a 1.5-ml tube and then, centrifuged for 30 sec at 800 × g. The cells were washed twice with α-MEM medium and plated at 1 × 10⁶ density in 24- or 96-well cell culture plates, over bone slices previously treated as described above. α-MEM medium was added with 10% fetal bovine serum, 1% glutamine, and 1% penicillin/streptomycin and supplemented with 1,25-dihydroxyvitamin D3 (Biovision; Milpitas, CA) at 10⁻⁸ M. ¹⁵ In DEX groups, the medium was added with dexamethasone (Sigma-Aldrich; St. Louis, MO) at 1 µM and the medium replacements were made every 2 days.

Cell Viability Assay

To determine cell viability, after 5, 7, and 9 days in a 96-well cell culture plate with bone substrate, the cells of the control (C) and treated [A (ALN), D6 (DEX), AD6 (ALN+DEX)] groups had the medium removed and were then washed twice with sterile PBS. The cells were incubated in 100 μl of MTT (thiazolyl blue tetrazolium bromide) solution (Sigma-Aldrich), which was added at a concentration of 0.5 mg/ml, at 37⁰C for 3 hr. After that, the MTT solution was removed and 100 μl of DMSO was added. The plate was shaken for 5 min, followed by an additional 5 min without shaking for color stabilization. Afterwards, spectrophotometer readings were made at 540 nm.

TRAP Histochemistry

After 7 days of culture in a 24-well cell culture plate with a bone substrate, the cells of the control (C) and treated groups (A, D6, AD6) were washed with PBS and fixed in 2% formaldehyde, the plate being positioned on ice during the 20 min of fixation. Then, they were washed again with PBS, permeabilized with 1% Triton X-100, and stained with leukocyte acid phosphatase kit (Sigma-Aldrich) following the manufacturer’s instructions. After fixing and staining, 3 non-overlapping and representative photos for each of triplicate culture wells were obtained at 100× magnification in a ZeissAxiovert 40 CFL microscope. The total number of multinucleated (with 2–9 nuclei) and giant (with 10 or more nuclei) TRAP positive cells were counted.

RANKL Immunofluorescence

After 7 days of culture in a 24-well cell culture plate over glass coverslip, the cells of the control (C) and treated groups (A, D6, AD6) were fixed in methanol at −20⁰C for 10 min, washed with PBS, and permeabilized with 5% Tween. Then, the specimens were blocked with milk for 1 hr and incubated with primary antibody for RANKL at 4⁰C (Santa Cruz Biotechnology Inc.). After washing with PBS, the specimens were incubated with anti-rabbit secondary antibody (Vector Laboratories; UK), incubated in Fit-C for 30 min, and counterstained with DAPI, 4′,6-diamidino-2-phenylindole (Vector Laboratories). Cells were examined under 200× magnification using an Olympus BX60 microscope equipped with an Olympus DP72 CCD camera.

Real Time Quantitative Polymerase Chain Reaction (RT-qPCR) Analysis

After 7 days, the cells were removed from the substrate and the plate by scraping and the RNA was isolated with ChargeSwitch Total RNA Cell Kit (Thermo Fischer Scientific, Waltham, MA USA). RNA was converted into cDNA using SuperScript First Strand kit (Thermo Fisher Scientific), following the manufacturer’s instructions. The gene expression analysis of the osteoclastogenesis regulators osteoprotegerin (OPG) and RANKL was performed with Taqman probes in a StepOnePlus device (Thermo Fisher Scientific). The housekeeping gene HMBS was used as a reference gene. All reactions were performed in triplicate.

Bone Resorption Assays

After 7 days of culture on the cortical bone substrates, previously used in cell culture, resorption assays were performed by scanning electron microscopy (SEM). Briefly, the bones were rinsed with 2.5% sodium hypochlorite for 15 min, washed with ultrapure water (MilliQ), dehydrated in crescent ethanol concentrations of ethanol, and left to dry in a hood. The specimens were coated with gold in a Balzers SDC 050 device and then examined using a LEO 430 scanning electron microscope operating at 10–15 kV. To quantify the resorption area, grids were overlayed on electron micrographs using ImageJ program and the area per grid point was 500 pixels. ² Resorption areas were considered at the intersections of the grid, and the total number of pits per bone sample and the percentage of the total resorption area were calculated (for this, the number of pits was considered by the total number of intersections).

Statistical Analysis

The data of MTT, TRAP+ cell count, qPCR, and resorption evaluated by SEM are presented as mean and standard deviation. These data were subjected to ANOVA followed by post hoc Tukey using Minitab 19 program. Differences were considered to be significant at the level of p<0.05.

Results

Treatment of Bone Substrates

To confirm the effectiveness of the smear layer removal protocol, bovine bone disks previously treated or not were cultured with osteoclasts for 7 days. Samples were evaluated using a scanning electron microscope. The smear layer removal protocol applied proved to be effective. It is possible to observe in more detail the structures of the mineralized bone matrix, such as the Haversian channels and Haversian systems, in the bone disks without the presence of a smear layer (Fig. 1).

Figure 1.

Scanning electron micrographs of the bone substrate. (A) Bone substrate not subjected to the smear layer removal process, observe the presence of smear layer (white arrow). (B) Bone substrate after the removal protocol, without the smear layer, exhibits a Haversian canal (asterisk) and Haversian system (between white triangles). Scale bar, 20 µm.

Saturation Analysis of Bone Substrates With ALN

After spectrophotometer analysis, it was possible to calculate how much ALN was actually incorporated into the bone matrix. At concentrations of 100 nM and 1, 10, and 100 µM, the ALN was completely incorporated into the bone matrix and no ALN was detected in the solution. At a concentration of 1 mM, ALN was detected in the solution, an indication of saturation of the bone substrate. It is important to note that saturation depends both on the concentration and volume used. Bone substrates of ALN-treated groups were incubated overnight in solutions of 10 µM of ALN and non-treated bones were incubated in PBS (Fig. 2).

Figure 2.

Saturation curve of bovine bone substrates incubated with ALN. At concentrations below 1 mM, all ALN was fully absorbed into the bone matrix. At concentrations above 1 mM, it was detected in the solution after an overnight incubation period, indicating that the substrate has been saturated. Abbreviation: ALN, alendronate.

Cell Viability

The cell viability was tested after 5, 7, and 9 days. The doses of ALN and DEX used did not cause any significant differences in cytotoxic effects between the treated and control groups at each experimental time. However, when comparing the experimental groups over the days (5, 7, and 9), an increase in viable cells was observed from days 5–7, and a decrease in the number of viable cells on day 9. The number of viable cells was significantly higher at the seventh day of culture (p<0.05%), in all groups (Fig. 3).

Figure 3.

Cell viability assay (thiazolyl blue tetrazolium bromide - MTT). The highest cell viability was detected on the seventh day of culture in all groups.^a,b Values without a common superscript differed (p<0.05).

TRAP Histochemistry

After 7 days of culture, the bone substrate was removed and the cells remaining on the plate were stained by TRAP histochemistry for osteoclast detection (Fig. 4). The highest number of total TRAP+ cells (multinucleated and giant) was detected in the AD6 group and this number was statistically significant compared with group A (p<0.05). No significant differences were detected between the other groups. Giant cells (10 or more nuclei) were found in greater number on DEX-treated groups (p<0.05).

Figure 4.

Quantification of TRAP+ cells after 7 days of culture. Representative micrographs showing the resorption pattern in each treatment. (A) Group C, (B) group A, (C) group D6, and (D) group AD6. (E) Greater amount of total TRAP+ cells (multinucleated and giant) was observed in the AD6 group. (F) Greater amount of giant TRAP+ cells (more than 10 nuclei) was found in the D6 and AD6 group.^a,b Values without a common superscript differed (p<0.05). Scale bar, 100 µm.

RANKL Immunofluorescence

After 7 days of culture, cells cultured over a glass coverslip were processed for RANKL immunofluorescence protocol. RANKL was homogeneously detected in the cells (Fig. 5). There was no significant distribution between groups.

Figure 5.

RANKL immunofluorescence. In green, immunostaining for RANKL and blue for DAPI. A homogeneous distribution of RANKL is observed between the groups. Scale bar, 50 µm. Abbreviations: RANKL, receptor activator of nuclear factor kappa B ligand; DAPI, 4′,6-diamidino-2-phenylindole.

Gene Expression of Osteoclastogenesis Regulators

The gene expression of osteoclastogenesis regulating cytokines, RANKL and OPG, was quantified on the seventh day of culture. RANKL was significantly more expressed in both groups treated with dexamethasone (D6 and AD6). The OPG gene was less expressed in these same groups (p<0.05). There were no significant differences between groups C and A, in both genes (Fig. 6).

Figure 6.

Gene expression of (A) OPG and (B) RANKL markers in bone marrow cells cultured for 7 days.^a,b Values without a common superscript differed (p<0.05). Abbreviations: RANKL, receptor activator of nuclear factor kappa B ligand; OPG, osteoprotegerin.

Bone Resorption

After 7 days of culture, the examination of bone slices by SEM revealed pits of several sizes on the bone surface in all groups. The lowest amount of resorption pits was observed in group A, with significant difference (p<0.05) between groups A and D6 (Fig. 7E). When quantifying resorption by area, the same pattern between groups was observed, with a statistically significant difference (p<0.05) only between groups A and D6 (Fig. 7F). The resorption areas were also assessed quantitatively. When only the pits per bone sample were counted, a very similar number was seen between groups C and A. However, when they were evaluated individually, a change in the pattern was detected. While in group C, deeper areas of resorption were observed, in group A, the areas were more superficial. Deep gaps were also detected in the D6 and AD6 groups. In all groups, shallow gaps were observed, mainly at the edges of the bone and in regions between the Havers systems (Fig. 7).

Figure 7.

Quantification of bone resorption after 7 days of culture. Representative scanning electron micrographs showing the resorption pattern in each treatment. (A) Group C, (B) group A, (C) group D6, and (D) group AD6. In (E), the number of resorption pits and in (F), the percentage of the resorbed area.^a,b Values without a common superscript differed (p<0.05). Scale bar, 10 µm.

Discussion

The main objective of this in vitro study was to investigate the effects of BP (ALN) and/or GC (DEX) therapy on osteoclastogenesis and osteoclast activity. Our results showed that the administration of DEX on primary mouse marrow cultures increased the expression of RANKL and decreased OPG, stimulating osteoclastogenesis and activation of osteoclasts, and, consequently, bone resorption. The resorption stimulus, however, was not reversed by the treatment with ALN.

To closely simulate the in vivo environment, a protocol for chemically removing the smear layer generated during the process of cutting bovine bone disks was applied; the complete removal was confirmed by SEM examination. Indeed, smear layer that contains remains of organic and inorganic tissue materials ¹⁶ is not present on bone surfaces in vivo, when osteoclasts adhere on mineralized matrix to start the resorptive activity.

Another clinical condition simulated was to determine how much of the ALN added to the medium was actually incorporated into the bone disks. BP strongly adheres to the bone matrix; in patients using BP such as ALN, BP is absorbed by osteoclasts during their resorption activity. ¹⁷ By using a spectrophotometer analysis, it was observed that all ALN at a concentration of 10 μM was fully incorporated into the bone matrix. As it was the same ALN concentration administered in previous studies of osteoclast cultivation,^18–20 it was the ALN concentration used in this study. On the other hand, DEX concentration was determined from previous dose-dependent study, in which the peak of DEX promoting effect on osteoclastogenesis was reached with 1 μM. ²¹ In addition, the viability of the cultured osteoclasts was analyzed by carrying out an MTT assay at days 5, 7, and 9. As the highest number of viable cells was detected on day 7, with a decrease on day 9, the 7th day was the period of choice for TRAP histochemistry, gene expression, immunofluorescence, and bone resorption experiments.

In general, treatment with DEX, associated or not with ALN, yielded an increase in RANKL and a decrease in OPG gene expression, suggesting its promoting effect on osteoclastogenesis. The distribution of the produced RANKL was also evaluated by immunofluorescence, which showed its labeling in all groups with a wide distribution throughout the cell, especially in association with the plasma membrane. Changes in these cytokines represent an indirect action on osteoclasts because the cells that control their release are osteoblasts and undifferentiated mesenchymal cells. In this context, indirect effects on osteoclastogenesis have been associated with GC.^12,22–24 In this study, as we used cells from the bone marrow, in which both osteoblasts and osteoclast precursors are present, it must be considered that part of the effects found on osteoclasts were indirect and related to the other cells of bone marrow.

In addition, an increase in the total number of TRAP+ cells in DEX-treated groups, even when it was administered with ALN, was observed, especially when just giant TRAP+ cells were considered. Significantly a higher number of giant TRAP+ cells were found in DEX groups (with or without ALN). Studies have indicated that GC promote the initial increase in osteoclast number.^7,21,25 Recently, it was reported that the combination of ALN and DEX worsens the healing of soft and hard tissue wounds around implants in rats. ²⁶ They also detected few TRAP+ cells in treated groups. However, the ALN dose used was about 50× greater than the previously reported amount needed to prevent ovariectomy-induced bone loss in rats. ²⁷ These effects are possibly dose dependent and should be investigated in more detail. To test whether the increase in the number of TRAP+ cells and osteoclastic activity markers correlated with an increase in the activity of these cells, SEM resorption assays were performed on the bone substrate. The resorbed area was higher when the cells were treated exclusively with DEX; this increase did not occur with combined treatment of DEX and ALN. This finding suggests that ALN bound to the bone disks was able to minimize the resorptive effects of dexamethasone.

It is important to note that BP bound to the bone matrix does not prevent the adhesion of the osteoclast to the bone. The activated osteoclast adheres to the matrix, reorganizes its cytoskeleton, and starts pumping acid. At that moment, with the acidification of the lacunae, the BP ions dissociate from the matrix, and are incorporated by osteoclasts. These molecules, now inside the cytoplasm, initiate their inhibitory effects on osteoclast activity. ¹⁷ It indicates that even in bone treated with ALN, areas of initial resorption are present. In this study, although the amount of resorption pits between the control group and the one treated with ALN were similar, the depth of the resorption pits was quite different. While the concavities in the bone treated with ALN were shallow, the areas of resorption of the control were deeper. This was also observed in the groups treated with DEX, especially in the one submitted only to treatment with this GC, where the largest number of pits, and also the greatest depth of the resorption lacunae was observed.

Thus, it was found that treatment with ALN, DEX, and the combination of both did not promote significant changes in the number of TRAP+ cells, compared with the control group. DEX promoted an increase in expression of osteoclastogenesis and osteoclastic activity marker genes, and the treatment with ALN reduced the depth of the resorption pits; however, its inhibitory effect was minimized when administered with DEX.