Morphological changes of osteoclasts by a MEK1 inhibitor, PD98059, were investigated to clarify a role of ERK. PD98059 promoted apoptosis of osteoclasts and the loss of ruffled borders. This study supports the importance of ERK in survival and polarity of osteoclasts.
Introduction: Extracellular signal-regulated kinase (ERK) is a mitogen activated protein kinase (MAPK) that has been reported to play a role in the survival and apoptosis of osteoclasts. However, the precise signal transduction mechanism is not fully understood. The aim of this study was to clarify the role of ERK in osteoclasts by histological analysis.
Materials and Methods: Using a rat calvarial organ culture system, the inhibition of ERK phosphorylation by PD98059, a MAPK/ERK kinase 1 (MEK1) inhibitor, was assayed by immunoblotting. Morphological changes in osteoclasts induced by PD98059 were elucidated by light and electron microscopy. The cellular localization of ERK was also determined by immunoelectron microscopy.
Results: PD98059 inhibited phosphorylated ERK after a 1-h incubation. Ultrastructural study demonstrated that PD98059 induced the accumulation of vesicles and vacuoles in osteoclasts and the loss of ruffled border at 1 h. At 3 h, some osteoclasts showed apoptosis with nuclear condensation, and at 6 h after PD98059 treatment, many osteoclasts were detached from the bone surface and had lost their cell polarity. Electron microscopic immunohistochemistry revealed that ERK was mainly localized in the cytoplasm of clear zones in control osteoclasts, but apoptotic osteoclasts also showed immunoreactivity in clear zone-like structures in contact with osteoblast-lineage cells.
Conclusion: These findings indicate that ERK in osteoclasts is involved in their survival and may be involved in the formation of a ruffled border and the maintenance of cell polarity.
Bone volume is maintained through a balance of bone formation and resorption. Bone resorption is regulated in part by osteoclasts, which are multinuclear giant cells arising from the monocyte-macrophage lineage of hematopoietic precursors.(1) Their differentiation is controlled by cell-cell interactions between osteoblast-lineage cells and osteoclast progenitors,(2–5) as well as by osteotropic hormones and cytokines. Results of recent studies have demonstrated that RANK, expressed in osteoclast-lineage cells, and RANKL, produced by osteoblast-lineage and stromal cells, participate in the differentiation of osteoclasts.(6–9)
For survival of mature osteoclasts, cytokines, such as macrophage colony-stimulating factor (M-CSF),(10) or cell-cell interactions though RANK(11) are required. Extracellular signal-regulated kinase (ERK) is a member of the mitogen-activated protein kinases (MAPKs), which are important in cell growth, differentiation, and apoptosis.(12–14) More specifically, they are necessary for the survival of hematopoietic cells.(15) In the MAPK cascade, the Ras-Raf interaction induces membrane translocation and activation of Raf, which then stimulates MAPK/ERK kinases (MEKs). These MEKs, in turn, phosphorylate ERK1 and ERK2, and this ERK activation is followed by phosphorylation of transcription factors and induction of immediate early gene expression. While this cascade may be stimulated by both cytokines, such as epidermal growth factor (EGF) and fibroblast growth factor (FGF), and serum,(16) ERK activation is also induced by cell-matrix interaction through integrins.(17) In osteotropic hormones, M-CSF,(18) interleukin-1 (IL-1),(19) and TNF-α(20) have been demonstrated to induce phosphorylation of ERK and prolong the survival of osteoclasts,(10,11) suggesting that they promote osteoclast survival through the ERK pathway. Recently, Miyazaki et al.(21) reported that ERK is involved in survival of osteoclasts by showing that inhibition of ERK activity induced the apoptosis of osteoclast-like cells while ERK activation remarkably lengthened their survival by preventing spontaneous apoptosis. Moreover, ERK has also been shown to play a role in osteoclast differentiation by participating in RANKL signaling(22,23) and in osteoclast activation through FGF signaling.(24) However, the localization of ERK in osteoclasts and the ultrastructural characteristics in the process of osteoclastic apoptosis induced by ERK inhibition remains to be determined.
This study describes the morphological changes in osteoclasts induced by PD98059, a MEK1 inhibitor,(25) in a rat calvarial organ culture system, using light and electron microscopy. Electron microscopic immunohistochemistry was also used to clarify the role of ERK in osteoclasts.
MATERIALS AND METHODS
All animal procedures were in accordance with The Guidelines for Animal Experiments, Okayama University Graduate School of Medicine and Dentistry.
Forty 5-day-old Wistar rats were used in this study. Calvariae were removed, cut through the sagittal suture, and incubated, dura mater side up, in BGJb medium (Gibco BRL, Life Technologies Inc., Rockville, MD, USA), supplemented with 10% FBS (PAA Laboratories GmbH, Linz, Austria), penicillin G (100 U/ml; Gibco BRL), and streptomycin (100 μg/ml; Gibco BRL). Samples were incubated with or without PD98059 (Calbiochem, La Jalla, CA, USA), at 37°C in a humidified incubator with an atmosphere of 5% CO2. After incubation for 15 minutes, 30 minutes, 1 h, 3 h, and 6 h, samples were subjected to biochemical and histological analyses.
Samples were dissolved, placed in 100 μl of sample buffer containing 4% SDS, 20% glycerol, and 12% mercaptoethanol in 100 mM Tris-HCl (pH 6.8), and heated at 100°C. SDS-PAGE was carried out using a 12% polyacrylamide gel. Samples were electrophoresed at 150 V for 60 minutes and transferred to a nitrocellulose membrane using 192 mM glycine and 20% methanol in 25 mM Tris-HCl (pH 8.3) at a constant amperage of 50 mA for 60 minutes. The membrane was then immersed in 10% skim milk in 10 mM PBS for 1 h to block nonspecific binding and washed with PBS containing 0.05% Tween 20. The membrane was probed with 1:1000 dilution of anti-phospho-ERK polyclonal antibody (New England Biolabs Inc., Beverly, MA, USA) at 4°C for 12 h, followed by incubation with horseradish peroxidase (HRP)-conjugated anti-rabbit Ig G (Sigma, St Louis, MO, USA) at room temperature for 1 h. Immunoreactivity was visualized using ECL Western blotting detection reagents (Amersham Pharmacia Biotech UK Ltd., Bucks, UK) according to the manufacturer's instructions. The same membrane was stripped and reprobed with 1:5000 dilution of anti-ERK1 monoclonal antibody (Transduction Laboratories, Lexington, KY, USA) followed by HRP-conjugated anti-mouse Ig G (Sigma).
Tissue preparation for histology
Samples were fixed with 4% paraformaldehyde and 0.1% glutaraldehyde in 0.05 M phosphate buffer (pH 7.3) for enzyme histochemistry and immunohistochemistry, or with 2% paraformaldehyde and 2.5% glutaraldehyde for ultrastructural study. All samples were decalcified in 5% EDTA (pH 7.3) at 4°C for 2 days.
Specimens were dehydrated in graded N,N-dimethylformamide and embedded in glycol methacrylate (GMA). Polymerization was performed under ultraviolet irradiation at 4°C for 24 h, and 2-μm-thick sections were cut and mounted on silane-coated glass slides. For TRACP detection, samples were incubated in a mixture of 0.01% naphthol AS-BI phosphate (Sigma), 0.06% fast red violet LB salt (Sigma), and 50 mM l(+)-tartaric acid (Wako, Osaka, Japan) in 0.1 M acetate buffer (pH 5.0) at 37°C for 3 h. Samples were then stained with 0.1% toluidine blue and observed under a Leitz Vario Orthomat (Leitz Wetzlar Gmbh, Wetzlar, Germany). The apoptosis rate was evaluated as the percentage of TRACP+ osteoclasts with condensed nuclei at 1, 3, and 6 h. The results were expressed as the mean ± SD of 10 samples.
Specimens were dehydrated in graded acetone and embedded in Epon 812 (TAAB, Berks, UK). Ultrathin sections were cut with an Ultracut E ultramicrotome (Reichert-Jung, Vienna, Austria) and mounted on copper grids. The samples were stained with uranyl acetate and lead citrate and observed under a Hitachi H-800 transmission electron microscope (TEM; Hitachi, Tokyo, Japan) at an accelerating voltage of 100 kV.
Light microscopic and electron microscopic immunohistochemistry
For light microscopic immunohistochemistry, 1-μm-thick GMA sections were mounted on silane-coated glass slides, immersed for 15 minutes in PBS containing 10% bovine serum albumin (BSA), incubated in anti-pan ERK monoclonal antibody (Transduction Laboratories) diluted to 1:500 at 4°C for 12 h, and incubated at room temperature for 1 h in 10 nm gold-conjugated goat anti-mouse IgG (British Biocell International, Cardiff, UK) diluted to 1:50. Finally, samples were washed with PBS and distilled water and air-dried. Silver intensification was carried out according to the method by Uchida et al.(26) Briefly, sections were immersed in 50 mM citrate buffer (pH 3.5), containing 5 mM maleic acid, 0.85% hydroquinone, 0.11% silver nitrate, and 2% gum arabic, at room temperature for 10 minutes, stained with 0.1% toluidine blue, and observed by light microscopy.
For electron microscopic immunohistochemistry, ultrathin GMA sections were mounted on nickel grids. The same procedures were performed as outlined for light microscopic immunohistochemistry except that the silver intensification step was omitted. Specimens were stained with 1% tannic acid and uranyl acetate and observed by TEM.
Inhibition of ERK phosphorylation by PD98059
Activated (phosphorylated) ERK were assayed by immunoblotting with an anti-phospho-ERK antibody. Preliminary experiments revealed that phosphorylation of ERK was transiently stimulated by the incubation with BGJb containing 10% FCS. Slight increase of phosphorylated ERK was also seen in the organ culture without FCS (data not shown). The inhibition of ERK phosphorylation by PD98059 at various concentrations was determined. After a 1-h incubation, phosphorylation of ERK was inhibited by PD98059 in a dose-dependent manner. This inhibition was clearly observed at a concentration of more than 10 μM PD98059 (Fig. 1A). A dose of 50 μM was chosen to investigate the effect of culture time period on phosphorylation of ERK. Phosphorylated ERK was not detected in samples without incubation. In the control samples, phosphorylated ERK was detected at 15 minutes and reached a maximum at 1 h. After this time, levels of phosphorylated ERK gradually declined and were only weakly detectable at 6 h. In samples treated with PD98059, phosphorylated ERK was detected at 15 and 30 minutes, but levels at 1 h were less than those in the control samples. Phosphorylated ERK was barely detectable at 3 and 6 h in PD98059-treated samples (Fig. 1B).
Apoptosis of osteoclasts stimulated by PD98059
Osteoclasts on the bone surface of the dura matter, rather than the periosteal side, of 5-day-old rat calvariae were examined, because this region is continuously resorbed in accordance with the growth of brain, and thus, contains more osteoclasts.
With regards osteoclasts, numerous vesicles and vacuoles were visible after a 1-h incubation with 50 μM PD98059 (Fig. 2A). After 3- and 6-h incubation with PD98059, some osteoclasts exhibited an apoptotic appearance with condensed nuclei, and after the 6-h incubation, apoptotic osteoclasts tended to be detached from the bone surface (Fig. 2B). Several apoptotic osteoclasts were also seen in control samples at 3 and 6 h. On the other hand, osteoblasts showed minimal apoptotic appearance at any time in both control and PD98059-treated samples.
The percentage of TRACP+ osteoclasts that appeared apoptotic at 1, 3, and 6 h was calculated. At 1 h, no apoptosis was detected in either control or PD98059-treated samples. Apoptotic rates at 3 and 6 h after PD98059 treatment were 23.9% and 38.5%, respectively, but only 5.4% at 3 h and 19.9% at 6 h in the control samples (Fig. 3).
Electron microscopic observations
After 1-h incubation with PD98059, numerous vesicles and vacuoles were visible in the osteoclast cytoplasm. Although osteoclasts were attached to the bone surface through clear zones, their ruffled borders appeared poorly developed (Fig. 4). At 3 h, several osteoclasts exhibited an apoptotic appearance with nuclear condensation. These osteoclasts contacted the bone surface through clear zones but did not possess ruffled borders (Fig. 5A). Many vesicles and vacuoles were seen in their cytoplasm. At 6 h, some apoptotic osteoclasts were detached from bone surface and fragmented. They contained fragmented nuclei and numerous vesicles and mitochondria in their cytoplasm (Fig. 5B).
Immunolocalization of ERK in osteoclasts and the effect of PD98059
The anti-ERK1 antibody used in Western blotting did not show a specific reactivity in immunohistochemistry. We used an anti-panERK antibody to detect the localization of ERK. Although this antibody reacts with ERK1 and ERK2 in Western blotting, its reactivity to ERK2 is stronger than that to ERK1 (data not shown).
In the control osteoclasts, immunoreactivity was detected by silver intensification on GMA sections in the region of attachment to the bone surface (Fig. 6A). Under electron microscopy, numerous gold particles were visible in the clear zones of osteoclasts, indicating the localization of ERK to this region (Fig. 7A). A few gold particles were also seen in their ruffled borders (Fig. 7B), but the labeling here was minimal compared with that in the clear zones. A few gold particles were also apparent near the basolateral plasma membrane of the control osteoclasts. In the PD98059-treated samples, ERK reactivity was seen in the clear zones of osteoclasts contacting bone surfaces as in control. In addition, moderate reactivity was also detected by light microscopy in the region between the osteoclasts and osteoblast-lineage cells (Fig. 6B). Under electron microscopy, clear zone-like structures were occasionally observed in apoptotic osteoclasts at the region of contact with osteoblast-lineage cells (Fig. 8A), and gold particles were visible in this area as well as in the clear zones (Fig. 8B).
The results of this study suggest that PD98059 inhibits ERK activation and accelerates osteoclastic apoptosis in calvarial organ culture. Phosphorylated ERK was not detected in calvariae immediately after dissection, and this is most likely because of a balance between the activities of MEK and phosphatases, such as PP2A, in the cytoplasm.(27) Because phosphorylation and de-phosphorylation of ERK may occur rapidly, this is likely to hinder detection of phosphorylated ERK under physiological conditions. Transient upregulation of ERK phosphorylation was detected in our organ culture system, and this is attributed to BGJb medium and/or unknown growth factors in FCS, because in vitro studies have shown phosphorylation of ERK to be induced by cytokines, such as EGF and FGF, and serum.(16) Inhibition of ERK phosphorylation after a 1-h incubation with PD98059 suggests that PD98059 prevents phosphorylation of ERK through inactivation of MEK1.
We have confirmed that ERK plays a role in survival of osteoclasts, as reported by Miyazaki et al.(21) Moreover, in the current study, the results obtained after PD98059 treatment were also seen after incubation with 10 μM U0126, a MEK1 and MEK2 inhibitor (data not shown). Previously, the TUNEL method has been used for the morphological detection of apoptosis, although this method sometimes causes false positive reactivity.(28) Indeed, we experienced similar shortcomings of this method and thus defined osteoclasts possessing TRACP+ reactivity and condensed nuclei as apoptotic in the present study. PD98059-induced apoptosis of osteoclasts was dependent on incubation time. Morphological changes induced by PD98059 were observed until 24 h, and fragmented apoptotic osteoclasts showing weak TRACP reactivity were more readily apparent after 6 h. Because it was difficult to distinguish fragmented osteoclasts from other cell types, such as osteoblasts, fibroblasts, and endothelial cells, evaluation of apoptotic cells after incubation periods greater than 6 h was not performed. In the present study, 50 μM of PD98059 was used for incubation experiments. Although we cannot exclude the possibility that the inhibition of ERK phosphorylation was caused by cytotoxicity of the relatively high dose of PD98059, this concentration did not seem to cause any morphological abnormality or apoptosis in osteoblasts. Therefore, apoptosis and other morphological changes in osteoclasts seem to reflect the prevention of phosphorylation of ERK by PD98059 through the inhibition of MEK1.
Apoptotic osteoclasts were also seen at 3 and 6 h in control samples. This may reflect the short life of mature osteoclasts without cytokines, such as M-CSF(10) and RANKL,(11) which were not added to our culture system. Moreover, it may be reflective of the difference between in vivo and in vitro microenvironments. In vivo, the bone surface of the dura matter side of calvariae receives continuous compressive forces from the growing brain. As a result, osteoclasts differentiate on this surface and resorb bone matrices constantly. Compressive force is reported to be important for the regulation of osteoclastic activity.(29,30) Because this mechanical force is removed in a in vitro culture system, this may have detrimental affects on the survival of osteoclasts. Phosphorylated ERK translocates to the nucleus and regulates expression of genes important for cell proliferation and differentiation.(31–33) Results of our study demonstrated that osteoclastic apoptosis was induced by PD98059 in a relatively short time period, indicating that new gene expression may not be required for apoptosis. Indeed, the ERK pathway also seems to promote cell survival by the inactivation of a component of the apoptotic process. ERK has been shown to be involved in the Ser112 phosphorylation of Bad, a pro-apoptotic member of the Bcl-2 family.(34,35) Non-phosphorylated Bad translocates from the cytosol into mitochondria and heterodimerizes with Bcl-xl to promote apoptosis.(36) Although ERK cannot directly phosphorylate Bad at Ser112, because of a lack of consensus phosphorylation site surrounding Ser112, phosphorylation of Bad is suggested to occur through a downstream kinase of ERK, such as Rsk.(37) In osteoclasts, activation of ERK by M-CSF, IL-1, TNF-α, and RANKL is consistent with promotion of cell survival.(10,11,18–20) Taken collectively, the ERK pathway is an important component of the survival signal in osteoclasts. The PI3-K-Akt pathway and the cross-talk with the Raf-MEK-ERK are also suggested to participate in osteoclastic apoptosis.(19) The downstream targets that regulate apoptosis of osteoclast remain to be determined. On the other hand, some bisphosphonates(38) and osteoprotegerin(39) have been reported to induce apoptosis of osteoclasts. The effects of these pro-apoptotic reagents on ERK also need to be addressed in future research.
In this study, PD98059 also seemed to inhibit the formation of the ruffled border in osteoclasts. A similar phenomenon has been reported after treatment with wortmannin, a PI-3 kinase inhibitor.(40,41) Ruffled borders are essential for bone resorption by osteoclasts and believed to be formed by the fusion of lysosomes.(42) In this study, vacuoles were visible in osteoclasts after a 1-h incubation with PD98059, and this may be because of the prevention of transport of lysosomes. Microtubules play an important role in cell-organelle transport and accumulation of vesicles, and stathmin, a microtubules-associated protein that regulates the dynamics of microtubules through its phosphorylation/de-phosphorylation, may be a substrate for ERK.(43,44) Thus, the loss of osteoclast ruffled borders and the accumulation of vesicles seen in this study may be attributed to the inhibition of stathmin, or other microtubules-associated proteins, and the subsequent prevention of lysosomal transport.
Previously, Miyazaki et al.(21) reported that ERK did not affect bone resorbing activity. Our finding of the loss of ruffled borders is not in agreement with their results, and this discrepancy may be the result of methodological differences. In their study, resorbing activity was quantified by the pit forming activity per osteoclast,(21) and there is a possibility that this results in exclusion of apoptotic osteoclasts detached from the bone surface.
In an attempt to further characterize the role of ERK in the osteoclast, we performed electron microscopic immunohistochemistry using an anti-pan ERK antibody. Our results show that ERK accumulates in clear zones of osteoclasts and occasionally localizes near the basolateral plasma membrane. In addition, ERK is also concentrated in clear zone-like structures, which are observed in the contact area between apoptotic osteoclasts and osteoblast-lineage cells. These findings lead us to conclude that ERK participates in the maintenance of cell polarity in osteoclasts, because clear zones are thought to be important for cell-matrix interactions and contain several signal transduction molecules, including Src. This area is also rich in cytoskeletal proteins such as actin filaments. While previous reports have indicated that the activation of ERK is linked to nuclear translocation, ERK can also be targeted to cellular attachment sites. Klemke et al.(45) have demonstrated that myosin light chain kinase (MLCK) is a downstream component of ERK, and activation of ERK with subsequent phosphorylation of MLCK leads to enhanced cell migration.(46,47) Moreover, both growth factor receptors and integrins promote signaling events leading to MAPK activity and the immediate induction of cell migration.(17,48) Phosphorylation of myosin light chains by MLCK is a critical regulatory step in myosin function, leading to a functional actin-myosin interaction involved in generating contractile force necessary for cell movement.(49) Taken together, these results show that ERK localized to the clear zones may be involved in osteoclast movement. This is supported by the finding that M-CSF, which facilitates the spreading of osteoclasts, also upregulates phosphorylated ERK.(18)
In summary, the results of this study show that ERK is localized in the clear zones of osteoclasts. From the results of the PD98059-treated samples, it seems that both the prevention of ruffled border formation and the acceleration of apoptosis in osteoclasts are associated with a decrease in phosphorylated ERK. Our results suggest that ERK in osteoclasts may be involved in their survival, motility, and cell polarity.
We thank Dr Noriyuki Nagaoka and Tomoko Yamamoto for their technical support. This study was supported in part by a grant (13671903) for scientific research from the Ministry of Education, Culture, Sports, Science and Technology of Japan.