Recent studies have demonstrated that bone is highly innervated and contains neuromediators that have functional receptors on bone cells. However, no data exist concerning the quantitative changes of innervation during bone loss associated with estrogen withdrawal. To study the involvement of nerve fibers in the regulation of bone remodeling, we have evaluated the modifications of innervation in a classical in vivo model of osteopenia in rats, ovariectomy (OVX). Skeletal innervation was studied by immunocytochemistry using antibodies directed against specific neuronal markers, neurofilament 200 and synaptophysin, and the neuromediator glutamate. Sciatic neurectomy, another model of bone loss due to limb denervation and paralysis, was used to validate our quantitative image analysis technique of immunostaining for nerve markers. Female Wistar rats at 12 wk of age were sham-operated (SHAM) or ovariectomized (OVX). Bone mineral density measurement and bone histomorphometry analysis of tibiae 14 d after surgery demonstrated a significant bone loss in OVX compared with SHAM. We observed an important reduction of nerve profile density in tibiae of OVX animals compared with SHAM animals, whereas innervation density in skin and muscles was similar for OVX and control rats. Quantitative image analysis of immunostainings demonstrated a significant decrease of the percentage of immunolabeling per total bone volume of neurofilament 200, synaptophysin, and glutamate in both the primary and secondary spongiosa of OVX rats compared with SHAM. These data indicate for the first time that OVX-induced bone loss in rat tibiae is associated with a reduction in nerve profile density, suggesting a functional link between the nervous system and the bone loss after ovariectomy.

We previously identified functional N‐methyl‐d‐aspartate (NMDA) glutamate receptors in mature osteoclasts and demonstrated that they are involved in bone resorption in vitro. In the present work, we studied the expression of NMDA receptors (NMDAR) by osteoclast precursors and their role in osteoclastogenesis using two in vitro models, the murine myelomonocytic RAW 264.7 cell line and mouse bone marrow cells, both of which differentiate into osteoclasts in the presence of macrophage colony‐stimulating factor (M‐CSF) and Rank ligand (RankL). Using RT‐PCR analysis with specific probes, we showed that RAW 264.7 cells and mouse bone marrow cells express mRNA of NMDAR subunits NMDA receptor 1 (NR1) and NMDA receptor 2 (NR2) A, B, and D. These subunits are expressed all along the differentiation sequence from undifferentiated precursors to mature resorbing osteoclasts. Semi‐quantitative PCR analysis showed no regulation of the expression of these subunits during the differentiation process. Two specific non competitive antagonists of NMDAR, MK801 and DEP, dose‐dependently inhibited osteoclast formation in both models, indicating that osteoclastogenesis requires the activation of NMDAR expressed by osteoclast precursors. MK801 had no effect when added only during the first 2 days of culture, suggesting that NMDAR are rather involved in the late stages of osteoclast formation. Finally, we demonstrated using Western‐blotting and immunofluorescence that activation of NMDAR in RAW 264.7 cells by specific agonists induces nuclear translocation of NF‐kappa B, a factor required for osteoclast formation. Altogether, our results indicate that osteoclast precursors express NMDAR that are involved in the osteoclast differentiation process through activation of the NF‐kappa B pathway. J. Cell. Biochem. 90: 424–436, 2003. © 2003 Wiley‐Liss, Inc.

To the Editor: In the April 2001 issue of the Journal, Gray et al.1 reported that glutamate does not play a major role in controlling bone growth, a conclusion that is at variance with multiple data published in recent years by independent groups that show evidence for a role of glutamate in bone remodeling.2-9 To sustain their argument, the authors first claim that N-methyl-D-aspartate (NMDA) and the NMDA glutamate receptor antagonist D(−)-2-amino-5-phosphonopentanoic acid (D-AP5) have no effect on in vitro bone formation. An unusual model was used to evaluate bone formation, consisting of primary osteoblasts cultured on slices of dentine, that may be complicated by the fact that dentin contains many factors that may stimulate osteoblast differentiation and interfere with the regulation pathway analyzed. The authors' data derived from this model are difficult to compare with a characterized and validated model of osteogenesis, such as the bone nodule formation assay in which osteoblast colonies can be counted and give direct quantitative assessment of osteoprogenitor differentiation.10 It is in such an assay that Dobson and Skerry showed an inhibitory effect on bone formation of MK801, another antagonist of the NMDA receptor acting on the channel site and not tested by Gray et al.11 Moreover, studies using agonists and antagonists of the NMDA receptor ligand site such as NMDA itself and D-AP5 need to take account of the fact that culture medium and fetal calf serum contain glutamate and glycine in sufficient concentrations to induce NMDA receptor activation; this may well have been the case in the authors' study. Specific antagonists targeting other sites of the NMDA receptor should also have been tested to establish that glutamate does not affect bone formation. Gray et al. conclude from their in vitro bone resorption assays that NMDA glutamate receptor agonists and antagonists have no effect on bone resorption, a conclusion that does not seem to reflect their own results. They show that MK801 significantly inhibits bone resorption evaluated by the number of pits, as previously reported by our group.3 Because of the small, nonsignificant inhibitory effect in the same experiments of the NMDA receptor antagonist D-AP5, they attribute the effect of MK801 on bone resorption to an action independent of its binding to the NMDA receptor. The lack of significant effect of D-AP5 in this culture model is again unsurprising considering that all experiments were performed with culture medium and 10% fetal calf serum containing glutamate that may compete with D-AP5 for the NMDA receptor ligand site. To validate the authors' conclusions, other noncompetitive antagonists of NMDA receptor, such as 1-(1,2-diphenylethyl)piperidine (DEP), should have been used. Our group has recently reported that several specific antagonists of NMDA receptors that bind to different sites of the receptor inhibit bone resorption, and that the specific channel blockers are the most potent inhibitors.8 This study was unfortunately not mentioned by Gray et al. Finally, the fact that the glutamate aspartate transporter (GLAST) knockout mice do not exhibit an abnormal bone phenotype is not necessarily an indication that glutamate signaling does not play a role in bone remodeling. The expression of different glutamate receptors and transporters in bone suggests several pathways for glutamatergic signaling in this tissue that may compensate for each other.2, 9, 12 Furthermore, there are multiple examples of gene knockouts that do not exhibit apparent morphological or physiological abnormalities under basal conditions but in which the essential role of the gene was shown in functional studies. In the bone field, this was recently illustrated by the osteopontin gene knockout mice that do not exhibit a spontaneous bone phenotype but show anomalies of bone cell function when skeletal homeostasis is challenged by ovariectomy.13 In the case of the GLAST knockout mice, ovariectomy or denervation experiments will need to be performed before firm conclusions can be drawn. We think therefore that the present study by Gray et al. may have been compromised by these potential technical pitfalls together with an incomplete consideration of the previous work in this interesting field.

The N‐methyl‐D‐aspartate (NMDA) subtype of the glutamate receptor has recently been identified in bone, but the molecular composition of this receptor expressed by bone cells is unknown. NMDA receptor (NMDAR) is a hetero‐oligomeric protein composed of two classes of subunits, the essential subunit NR1 and NR2A to D subunits that do not by themselves produce functional channels but potentiate NR1 activity and confer functional variability to the receptor. These subunits coassemble in different combinations to form functionally distinct NMDAR. In this study, we have investigated the molecular composition of NMDAR expressed by osteoblasts and osteoclasts in culture, using RT‐PCR analysis, in situ hybridization and immunocytochemistry. Specific probes were designed for the different subunits of the NMDAR, and we showed by RT‐PCR analysis that mammalian osteoclasts expressed NR2B and NR2D subunits mRNAs but not NR2A and NR2C mRNAs. Rat calvaria and MG63 osteoblastic cells also expressed several NR2 subunits mRNAs, namely NR2A, NR2B, and NR2D. In situ hybridization on isolated rabbit osteoclasts and MG63 cells has confirmed the localization of NR1, NR2B, and NR2D transcripts in osteoclasts and NR1, NR2A, NR2B, and NR2D transcripts in MG63 cells. The expression of NR2D protein by bone cells was shown by immunofluorescence. These results demonstrate for the first time that osteoblasts and osteoclasts express several NR2 subunits, suggesting a molecular diversity of NMDAR channels similar to what was shown for brain. The presence of distinct functional NMDAR on bone cells may be associated with various states of bone cell differentiation and function. J. Cell. Biochem. 82: 134–144, 2001. © 2001 Wiley‐Liss, Inc.

N-Methyl-d-aspartate (NMDA) glutamate receptors, widely distributed in the nervous system, have recently been identified in bone. They are expressed and are functional in osteoclasts. In the present work, we have studied the effects of specific antagonists of NMDA receptors on osteoclast activation and bone resorption. Using an in vitro assay of bone resorption, we showed that several antagonists of NMDA receptors binding to different sites of the receptor inhibit bone resorption. Osteoclast activation requires adhesion to the bone surface, cytoskeletal reorganization and survival. We demonstrated by autoradiography that the specific NMDA receptor channel blocker, MK 801, binds to osteoclasts. This antagonist had no effect on osteoclast attachment to bone and did not induce osteoclast apoptosis. In contrast, MK 801 rapidly decreased the percentage of osteoclasts with actin ring structures that are associated with actively resorbing osteoclasts. These results suggest that NMDA receptors expressed by osteoclasts may be involved in adhesion-induced formation of the sealing zone required for bone resorption.

1 The N-methyl-D-aspartate (NMDA) glutamate receptor, widely distributed in the mammalian nervous system, has recently been identified in bone. In this study, we have investigated whether NMDA receptors expressed by osteoclasts have an electrophysiological activity. 2 Using the patch clamp technique two agonists of the NMDA receptor, L-glutamate (Glu) and NMDA, were shown to activate whole-cell currents recorded in isolated rabbit osteoclasts. 3 The current-voltage (I-V) relationships of the currents induced by Glu (IGlu) and NMDA (INMDA) were studied using Mg2+-free solutions. The agonist-induced currents had a linear I-V relationship with a reversal potential near 0 mV, as expected for a voltage independent and non-selective cationic current. 4 I Glu and INMDA were sensitive to specific blockers of NMDA subtype glutamate receptors, such as magnesium ions, (5R, 10S)-(+)-5-methyl-10,11-dihydro-5H-dibenzo[a, d]cyclohepten -5,10-imine (MK-801) and 1-(1,2-diphenylethyl) piperidine (DEP). The block of IGlu and INMDA by these specific antagonists was voltage dependent, strong for negative potentials (inward current) and absent for positive potentials (outward current). 5 These results demonstrate that NMDA receptors are functional in rabbit osteoclasts, and that their electrophysiological and pharmacological properties in these cells are similar to those documented for neuronal cells. Active NMDA receptors expressed by osteoclasts may represent a new target for regulating bone resorption.