Cannabis Research A-Z

Endocrinology, doi:10.1210/en.2008-0150

Endocrinology Vol. 149, No. 11 5619-5626

Copyright © 2008 by The Endocrine Society

Aymen I. Idris, Antonia Sophocleous, Euphemie Landao-Bassonga, Robert J. van't Hof and Stuart H. Ralston

Rheumatic Diseases Unit, Institute of Genetics and Molecular Medicine, University of Edinburgh, Western General Hospital, Edinburgh EH4 2XU, United Kingdom

Address all correspondence and requests for reprints to: Stuart H. Ralston, Rheumatic Diseases Unit, University of Edinburgh, Western General Hospital, Edinburgh EH4 2XU, United Kingdom. E-mail: [email protected].

 Abstract

The endocannabinoid system has recently been shown to play a role in the regulation of bone metabolism. The type 2 cannabinoid receptor (CB2) has been reported to regulate bone mass, but conflicting results have been reported with regard to its effects on bone resorption and osteoclast function. Here we investigated the role that CB2 plays in regulating bone mass and osteoclast function using a combination of pharmacological and genetic approaches.

The CB2-selective antagonist/inverse agonist AM630 inhibited osteoclast formation and activity in vitro, whereas the CB2-selective agonists JWH133 and HU308 stimulated osteoclast formation. Osteoclasts generated from CB2 knockout mice (CB2) were resistant to the inhibitory effects of AM630 in vitro, consistent with a CB2-mediated effect.

There was no significant difference in peak bone mass between CB2 mice and wild-type littermates, but after ovariectomy, bone was lost to a greater extent in wild-type compared with CB2^–/– mice. Furthermore, AM630 protected against bone loss in wild-type mice, but the effect was blunted in CB2 mice.

We conclude that CB2 regulates osteoclast formation and bone resorption in vitro and that under conditions of increased bone turnover, such as after ovariectomy, CB2 regulates bone loss. These observations indicate that CB2 regulates osteoclast formation and contributes to ovariectomy-induced bone loss and demonstrate that cannabinoid receptor antagonists/inverse agonists may be of value in the treatment of bone diseases characterized by increased osteoclast activity.

Introduction

 BONE REMODELING IS a complex process that is regulated by an interplay between circulating hormones and locally produced factors that act in a concerted manner to regulate osteoblast and osteoclast activity. Over recent years, there has been increasing interest in the role that the nervous system and neurotransmitters play in the regulation of bone remodeling. Reflecting this fact, the endocannabinoid pathway has recently been implicated as are important regulator of bone turnover and bone mass.

The endocannabinoid system comprises two known receptors [the type 1 (CB1) and type 2 (CB2) cannabinoid receptors], a family of endogenous ligands, and a molecular machinery for ligand synthesis, transport, and inactivation. Cannabinoid receptors belong to the G protein-coupled receptor superfamily and are highly expressed in the brain (CB1), immune system (CB2), and a number of other tissues.

A variety of natural and synthetic ligands have been identified that bind to these receptors. Cannabinoid receptor agonists such as 9-tetrahydrocannabinol, anandamide, 2 arachinodyl glycerol, CP55,490, and JWH133 bind to the receptors causing inhibition of adenylyl cyclase and activation of ERK kinases and other signaling pathways in target cells. Another class of cannabinoid receptor ligands have been synthesized including AM251, AM630, and SR141716A (rimonabant), which block the effects of cannabinoid receptor agonists.

These compounds were originally referred to as cannabinoid receptor antagonists, but it has now become clear that these compounds cause activation of the receptors in the absence of agonist binding, with opposite effects on downstream signaling cascades as those of the agonists. Accordingly, these compounds are now referred to as cannabinoid receptor antagonists/inverse agonists.

We previously reported that osteoblasts and osteoclasts express cannabinoid receptors and that mice with inactivation of CB1 have increased bone mass and are protected from ovariectomy-induced bone loss. We also found that the CB1-selective antagonist/inverse agonist AM251 inhibited osteoclast differentiation and prevented ovariectomy-induced bone loss in vivo, confirming the importance of CB1 as a regulator of bone resorption. In another study, Tam and colleagues reported that mice with CB1 deficiency on a CD1 background had high bone mass but surprisingly found that mice with CB1 deficiency on a C57BL/6 background had low bone mass.

There is evidence that CB2 also plays a role in regulating bone metabolism. Mice with deficiency of CB2 (CB2^–/–) have been reported to develop age-related osteoporosis in association with increased bone turnover.

The CB2-selective agonist HU308 has also been reported to promote osteoblast differentiation, to inhibit osteoclast differentiation, and to protect against ovariectomy-induced bone loss.

On the basis of these observations, it has been suggested that an important function of CB2 is to suppress bone turnover and regulate osteoblast-osteoclast coupling. Contrasting with these findings, however, we and others have reported that pharmacological agonists of cannabinoid receptors stimulate osteoclast formation and bone resorption in vitro in both mouse and human cells. In these studies, however, it was not possible to determine whether the effects were mediated by CB1, CB2, or both receptors.

In view of this, the aim of the present study was to clarify the role that CB2 plays in osteoclast function and ovariectomy-induced bone loss using a combination of pharmacological and genetic approaches.

Materials and Methods

 Materials
Reagents were obtained from Sigma (Dorset, UK) unless otherwise indicated. The cannabinoid receptor ligands JWH133, AM251, and AM630 were purchased from Tocris Bioscience (Bristol, UK) and HU308 was a kind gift from Dr. R. J. Arends (Organon). The pharmacological properties of the cannabinoid receptor ligands used in the study are summarized in Table 1.

Tissue culture medium was obtained from Invitrogen (Paisley, UK), and primary antibodies were purchased from Cell Signaling Technology (Beverly, MA), unless otherwise stated.

Human IL-1β was obtained from Roche (London, UK); human macrophage colony-stimulating factor (M-CSF) was obtained from R&D Systems (Abingdon, UK); 1,25-(OH)₂-vitamin D₃ was obtained from Alexis Biochemicals (Nottingham, UK); and mouse receptor activator of nuclear factor-B ligand (RANKL) was a gift from Dr. F. P. Ross (Washington University, St. Louis, MO).

 Mouse osteoblast-bone marrow cocultures
Osteoblast-bone marrow cocultures were performed essentially as previously described. Briefly, osteoblasts were isolated from the calvarial bones of 2-d-old mice by sequential collagenase digestion (type I collagenase; Sigma) and cultured in -MEM supplemented with 10% fetal calf serum (FCS) and penicillin and streptomycin at 37 C in 5% CO₂.

Bone marrow cell populations containing osteoclast precursors were isolated from the long bones of 3- to 5-month-old mice, and erythrocytes were removed by Ficoll Hypaque density gradient centrifugation. The cells were washed with PBS and resuspended in culture medium. Osteoblasts and bone marrow cells were plated together at 10 x 10³ cells per well and 2 x 10⁵ cells per well, respectively, on dentine slices in 96-well plates in 150 µl -MEM supplemented with 10% FCS, antibiotics, and 10 nM 1,25-(OH)₂-vitamin D3 for 7 d. The medium was replaced with the addition of IL-1β (10 U/ml), test compounds were added, and the cultures were terminated on d 8 for actin-ring identification or d 9 for quantification of bone resorption. At the end of the culture period, dentine slices with adherent cells were fixed in 4% paraformaldehyde, washed with PBS, and incubated with phalloidin at 37 C for 30 min to identify actin rings.

The cultures were counterstained for tartrate-resistant acid phosphatase (TRAcP) as previously described, and TRAcP-positive cells with three or more nuclei bearing an actin ring were considered to be active osteoclasts. Cells were then removed from the dentine slices by wiping with tissue paper. Resorption pits were visualized by reflected light microscopy, and the area resorbed was quantified by Image Analysis using custom software developed using Aphelion ActiveX objects (ADCIS, Gif sur Yvette, France) as previously described.

Osteoclast cultures
Osteoclast formation and survival were studied using RANKL- and M-CSF-generated osteoclasts from bone marrow macrophages as described. Briefly, bone marrow cells were flushed from the long bones of 3- to 5-month-old mice, plated onto petri dishes, and incubated for 48 h in the presence of M-CSF (100 ng/ml). Nonadherent erythrocytes were removed, and the adherent cells were washed with PBS and resuspended in culture medium.

The resulting M-CSF-dependent bone marrow macrophages were plated onto either 96-well plates (10⁴ cells per well) or 24-well plates (2 x 10⁵ cells per well) in 150 or 400 µl -MEM supplemented with 10% FCS, antibiotics, M-CSF (25 ng/ml), and RANKL (100 ng/ml). Test agents were added on d 5 and the cultures terminated on d 6. The cells were fixed in 4% paraformaldehyde, washed with PBS and stained for TRAcP as described above. TRAcP-positive cells with more than three nuclei were counted as osteoclasts.

Apoptosis and caspase assays
Adherent and nonadherent cells from the osteoclast cultures were collected, fixed with 4% paraformaldehyde, and cytospun onto glass slides. Apoptosis was identified on the basis of characteristic changes in nuclear morphology using 4',6-diamidino-2-phenylindole (DAPI) staining and by terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick end labeling (TUNEL) staining. An average of six microscopic fields per treatment group was analyzed at x200 magnification, and the number of apoptotic cells was quantitated in relation to total cell number.

Levels of inactive (p-30) and activated (p-19) caspase-3 expression were evaluated by Western blotting. Briefly, both adherent and nonadherent cells were washed in PBS and homogenized in lysis buffer (0.1% wt/vol sodium dodecyl sulfate, 0.5% wt/vol sodium deoxycholate in PBS) supplemented with 2% vol/vol protease inhibitor cocktail (Sigma).

The samples were cleared by centrifugation at 12,000 rpm for 10 min at 4 C, and protein concentration was determined using a Bio-Rad (Hercules, CA) protein assay kit. Cell extracts (20 µg/lane) were run on a 10% acrylamide gel and blotted onto a nylon membrane. Activated caspase-3 was detected using a rabbit polyclonal anti-cleaved caspase-3 antibody (1:1000 dilution; Cell Signaling Technology), and inactive caspase-3 was detected using a goat polyclonal anti-caspase-3 antibody (1:100 dilution; Santa Cruz Biotechnology, Inc., Santa Cruz, CA). β-Actin was detected using a mouse monoclonal antibody (1:1000 dilution; Santa Cruz Biotechnology).

Subsequently, the membranes were washed in Tris-buffered saline, incubated with the appropriate secondary antibodies coupled to horseradish peroxidase, and washed in Tris-buffered saline again, and bands were visualized using chemiluminescence (Amersham, Little Chalfont, UK) on a Bio-Rad Fluomax gel imaging system.

Animals and ovariectomy-induced bone loss

Mice with CB2 deficiency were obtained from Dr. Susana Winfield at the National Institutes of Health and were generated as previously described.

These mice have previously been crossed with wild-type C57BL/6 mice for at least 10 generations to create a congenic strain on a C57BL/6 background. To minimize the chance of genetic drift, CB2^–/– mice used in this study were generated by mating of heterozygote breeding pairs, and all comparisons between CB2^–/– and wild-type mice were performed on littermates.

Animal experiments were approved by the ethical review board of the University of Edinburgh and were conducted in accordance with United Kingdom Home Office regulations. Ovariectomy or sham ovariectomy was performed in 9-wk-old adult female mice. Treatment with test agents was commenced 2 d after ovariectomy or sham ovariectomy by ip administration of the drug in corn oil. Controls received vehicle in corn oil.

The treatment was continued for 19 d and the experiment terminated on d 21. Bone mineral density was measured at the tibial metaphysis by microcomputed tomography (micro-CT), using a Skyscan 1172 scanner. After bone mineral density scanning, the limbs were embedded and processed for bone histomorphometry as previously described.

We also performed some in vitro experiments on osteoclasts generated from mice with CB1 deficiency on an ABH background that were a kind gift from Dr. David Baker (University of London).

Statistical analysis
Comparison between groups was by ANOVA followed by Dunnett’s post test using SPSS for Windows version 11. A P value of 0.05 was considered statistically significant.

 Results

 Cannabinoid receptor antagonists/inverse agonists inhibit osteoclast formation and bone resorption in osteoblast-bone marrow cocultures
Treatment of mouse osteoblast-bone marrow cocultures for 48 h with the CB2-selective antagonist/inverse agonist AM630 and the CB1-selective antagonist/inverse agonist AM251 inhibited osteoclast formation and bone resorption in a concentration-dependent manner (Fig. 1).

The mean ± SEM concentration of AM630 that half-maximally inhibited osteoclast formation and bone resorption (IC₅₀) was 6.3 ± 0.3 and 7.2 ± 0.1 µM, respectively. Corresponding values for AM251 were 5.9 ± 1.3 and 6.6 ± 0.4 µM (Fig. 1, A–D). Exposure of the cultures to these agents for 24 h also inhibited actin ring formation by about 50% (Fig. 1, E and F). Next, we investigated the effects of AM630, AM251, and the CB2-selective agonists JWH133 and HU308 on RANKL- and M-CSF-induced osteoclast formation in vitro (Fig. 2). We found that AM630 and AM251 both significantly inhibited RANKL-induced osteoclast formation in a concentration-dependent manner with 50% inhibition at values of 330 ± 40 nM for AM630 and 870 ± 100 nM for AM251.

Conversely, JWH133 and HU308 both stimulated osteoclast formation from about 10 nM with maximal stimulation at about 300 nM. At higher concentrations, HU308 inhibited osteoclast formation, whereas JWH133 stimulated osteoclast formation at all concentrations tested (Fig. 2, A and B). We also studied the effect of JWH133 and HU308 on osteoclast fusion by counting the number of osteoclasts with 20 or more nuclei.

As shown in Fig. 2, JWH133 and HU308 significantly increased osteoclast size and nuclearity such that the proportion of cells with more than 20 nuclei rose from about 15% in vehicle-treated cells to 20–25% in JWH133- and HU308-treated cells (Fig. 2B). None of the ligands that we tested significantly affected macrophage growth or viability at concentrations up to 10 µM, excluding a nonspecific toxic effect of the compounds on cell viability (data not shown).

 FIG. 1. Cannabinoid receptor inverse agonists inhibit osteoclast formation and bone resorption in vitro. A, Osteoclast number in osteoblast bone marrow cocultures exposed to AM630 or AM251 for 48 h; B, representative photomicrographs from the cultures shown in A at low and high power; C, resorbed area in osteoblast-bone marrow cocultures exposed to AM630 or AM251 for 48 h as assessed by reflected light microscopy; D, representative photomicrographs at low and high power from the cultures shown in C; E, osteoclasts with intact actin rings after 24 h exposure to the test substances as visualized by phalloidin staining counted and expressed as a percentage of total osteoclast numbers; F, representative photomicrographs from the cultures are shown in E at low and high power.

The areas of the photomicrographs in the high-power images in B, D, and F are indicated by boxes in the lower-power images. Values in the graphs are means ± SEM and were obtained from three to five independent experiments. *, P < 0.05; **, P < 0.01 vs. vehicle.

 FIG. 2. Regulation of RANKL-induced osteoclast formation by cannabinoid receptor ligands. A, Mouse bone marrow macrophages were cultured in M-CSF (25 ng/ml) and RANKL (100 ng/ml) for 5 d and then exposed to vehicle, JWH133, AM630, AM251, or HU308 at the concentrations indicated for 24 h. Osteoclast numbers were assessed by counting multinucleated TRAcP-positive cells with three or more nuclei and expressed as a percentage of those in vehicle-treated cultures. B, Number of osteoclasts per well with 20 or more nuclei in the cultures from A. C, Representative photomicrographs from the cultures in A.

Values in the graphs are means ± SEM and were obtained from three to five independent experiments. *, P < 0.01 vs. control cultures; +, P < 0.05 vs. HU308-treated cultures; $, P < 0.05 vs. AM251- and AM630-treated cultures.

 Cannabinoid receptor antagonists/inverse agonists induce osteoclast apoptosis in vitro
To investigate the mechanism of osteoclast inhibition, we studied the effect of AM630 and AM251 on osteoclast apoptosis. These studies showed that both agents induced apoptosis in mature osteoclast at concentrations greater than 1 µM (Fig. 3A) and activated caspase-3 as early as 12 h (Fig. 3C). After 24 h, both agents had induced apoptosis in about 90% of cells, whereas the CB2-selective agonist JWH133 had no significant effect on apoptosis. Representative photomicrographs of osteoclast cultures treated with the cannabinoid receptor ligands JWH133, AM251, and AM630 are shown in Fig. 3B.

FIG. 3. Cannabinoid receptor inverse agonists induce apoptosis in mouse osteoclast cultures. A, Mouse bone marrow macrophages were cultured in the presence of M-CSF (25 ng/ml) and RANKL (100 ng/ml) for 6 d and then exposed to vehicle, JWH133, AM630, or AM251 for 24 h at the concentrations indicated. Apoptotic cells were visualized by DAPI and TUNEL staining, and cells with fragmented DNA were counted and expressed as a percentage of the total cell number. B, Representative photomicrographs of apoptotic osteoclasts visualized by DAPI and TUNEL staining in the cultures from A. Nuclei from apoptotic cells are stained bright green with the TUNEL assay, and condensed chromatin in the DAPI assay is seen as a bright blue stain (arrows). C, Total cellular protein (50 µg/lane) from the cultures in A was analyzed for expression of β-actin, intact caspase-3 (C3i). and cleaved (active) caspase-3 (C3a) by Western blot analysis. Values in the graphs are means ± SD and were obtained from three independent experiments. **, P < 0.01 vs. control.

FIG. 4. Inhibitory effects of AM630 on osteoclast formation are mediated through CB2. A, Mouse bone marrow macrophages from wild-type mice were cultured in the presence of M-CSF (25 ng/ml) and RANKL (100 ng/ml) for 6 d and then exposed to AM630 in the presence of vehicle or JWH133 at a concentration of 1 µM for 24 h at the indicated concentrations. B, Osteoclasts were generated from CB1^–/–, CB2^–/–, and wild-type littermates as described above and cultured in the presence of AM630 at the concentrations indicated for 24 h. C, Representative photomicrographs from the cultures in B. Values in the graphs are means ± SEM and were obtained from three independent experiments. #, P < 0.05, JWH133 vs. vehicle (A) and wild-type vs. CB2^–/– (B); *, P < 0.05 vs. vehicle; **, P < 0.01 vs. vehicle.

 Cannabinoid receptor antagonists/inverse agonists prevents ovariectomy-induced bone loss
To determine whether cannabinoid receptor antagonists/inverse agonists prevented ovariectomy-induced bone loss in vivo, we studied the effects of AM630 on ovariectomy-induced bone loss in wild-type and CB2^–/– mice.

Before ovariectomy, there was no significant difference in trabecular bone volume at the tibial metaphysis as assessed by micro-CT in wild-type and CB2^–/– mice (Fig. 5, A and B). Analysis of cortical bone and total bone volume similarly showed no difference between the genotypes (data not shown). After ovariectomy, loss of trabecular bone volume (Fig. 5E), trabecular thickness (Fig. 5D), and trabecular number (Fig. 5E) were slightly but significantly less in CB2^–/– mice as compared with wild-type littermates.

Treatment of wild-type mice with AM630 at a dose of 0.1 mg/kg·d and 1 mg/kg·d completely prevented bone loss, whereas the CB2^–/– mice were resistant to the protective effects of AM630 at 0.1 mg/kg but not at 1.0 mg/kg (Fig. 5F). Bone histomorphometry showed that osteoclast numbers and active resorption surfaces were increased after ovariectomy in vehicle-treated wild-type mice, whereas this was prevented in AM630-treated mice.

Osteoblast numbers increased after ovariectomy as expected but were unaffected by AM630 treatment (Table 2). These data indicate that AM630 inhibits ovariectomy-induced bone loss by interacting with the CB2 receptor at low concentration, but at higher concentrations, selectivity is lost and prevention of bone loss occurs probably by an effect on CB1.

FIG. 5. Regulation of ovariectomy-induced bone loss by CB2. A, Trabecular bone volume (BV/TV) in wild-type and CB2 mice at age 3 months assessed by micro-CT of the tibia; B, representative micro-CT images from the tibial metaphysis; C, trabecular bone volume (BV/TV) in CB2^–/– mice and wild-type littermates subjected to ovariectomy (Ovx) or sham operation at age 9 wk, with changes in BV/TV normalized to those in wild-type sham-operated animals and expressed as percent change; D, trabecular thickness (Tb.Th) from the same experiment, expressed in the same way; E, trabecular number (Tb.N) from the same experiment expressed in the same way; F, changes in BV/TV in CB2^–/– and wild-type littermates subjected to ovariectomy or sham ovariectomy and injected daily with AM630 or vehicle at the dose indicated. Values in each panel are means ± SEM from seven to eight mice per group. *, P < 0.05; **, P < 0.01 CB2^–/– vs. wild-type; #, P < 0.01, AM630 vs. vehicle.

 Discussion

We have previously reported that CB1 is required for ovariectomy-induced bone loss in mice and that AM251, an antagonist/inverse agonist with selectivity toward CB1, causes osteoclast inhibition and prevents ovariectomy-induced bone loss in mice. We and others also reported that cannabinoid receptor agonists stimulate osteoclast formation and bone resorption in vitro . In contrast, other workers have reported that the CB2-selective agonist HU308 inhibits osteoclast formation in vitro and partially prevents ovariectomy-induced bone loss in vivo. In view of this, the aim of the present study was to further investigate the effects of pharmacological activation and blockade of CB2 on osteoclast differentiation and function in vitro and in vivo in wild-type and CB2^–/– mice.

We found that the CB2-selective antagonist/inverse agonist AM630 inhibited osteoclast formation and bone resorption in osteoblast cocultures, with similar potency to the CB1-selective antagonist/inverse agonist AM251. The inhibitory effect in bone resorption may have been due in part to disruption of the actin ring formation, but both agents were also found to inhibit osteoclast differentiation and to cause osteoclast apoptosis in RANKL- and M-CSF-generated osteoclast cultures.

These data therefore suggest that cannabinoid receptor antagonists/inverse agonists inhibit bone resorption mainly by reducing osteoclast formation but indicate that there may also be an inhibitory effect on the ability of mature osteoclasts to form actin rings and adhere to the bone surface.

There was a striking difference in potency in the ability of AM251 and AM630 to cause osteoclast inhibition in RANKL- and M-CSF- generated isolated osteoclast cultures where they were effective in the nanomolar range as opposed to bone marrow osteoblast cocultures where micromolar concentrations were required for osteoclast inhibition. This raises the possibility that cannabinoid receptor antagonists/inverse agonists may influence osteoblast-osteoclast cross talk in a manner that interferes with the direct inhibitory effects of these agents on osteoclasts, but further work will be required to explore the mechanisms responsible.

In keeping with the fact that cannabinoid receptor antagonists/inverse agonists caused osteoclast inhibition, the CB2-selective agonists HU308 and JWH133 significantly increased osteoclast formation at concentrations in the nanomolar range. Although HU308 cause osteoclast inhibition at 10 µM, this is approximately 2000 times greater than the concentration of HU308 required for CB2-mediated adenylyl cyclase inhibition.

This suggests that the inhibitory effects of HU308 at these concentrations may have been nonspecific and mediated by an interaction with pathways other than CB2.

The stimulatory effects of HU308 and JWH133 on osteoclast formation are consistent with previous work that has shown that nonselective cannabinoid receptor agonists including anandamide, 2-arachinodyl glycerol, and CP55,940 stimulate osteoclast formation and bone resorption in vitro at nanomolar concentrations.

The observations reported here differ from those of Ofek and colleagues, who found that HU308 caused osteoclast inhibition at concentrations in the nanomolar range. We cannot readily explain this difference except to note that the observations of Ofek were based in part on studies of RAW 264.7 cells rather than authentic osteoclasts. Although Ofek also studied M-CSF- and RANKL-stimulated bone marrow cultures, the numbers of osteoclasts generated were very low (an average of 15 osteoclasts per culture), and this might also have contributed to the differences observed.

Because the pharmacological ligands currently available are not completely specific and can interact with both CB1 and CB2, further studies were conducted to determine whether the inhibitory effects of AM630 on osteoclast activity were truly mediated by CB2.

Studies in wild-type mice showed that at low concentrations, the osteoclast-inhibitory effect of AM630 was reversed by the CB2-selective agonist JWH133 consistent with a CB2-mediated effect.

At higher concentrations of AM630, however, JWH133 did not reverse the osteoclast inhibition, probably because AM630 is known to act as an inverse agonist of CB1 at high concentrations.

In keeping with this, osteoclasts generated from CB2^–/– mice were resistant to the inhibitory effects of AM630 at low concentrations but inhibited osteoclast formation in CB2^–/– mice at higher concentrations. In contrast, osteoclasts generated from CB1^–/– mice were equally sensitive as wild type to the inhibitory effects of AM630.

To determine whether pharmacological blockade of CB2 can also prevent the bone loss that results from estrogen deficiency, we studied the effects of the CB2-selective antagonist AM630 on ovariectomy-induced bone loss in wild-type and CB2^–/– mice.

Administration of AM630 to ovariectomized wild-type mice in a dose of 0.1 and 1 mg/kg·d prevented ovariectomy-induced bone loss, with results virtually identical to those observed with the CB1-selective inverse agonist AM251 as previously reported. Analysis of bone histomorphometry showed that AM630 blocked the increase in osteoclast numbers and active resorption surfaces that followed ovariectomy, demonstrating that prevention of bone loss with AM630 was due to an inhibitory effect on osteoclasts and bone resorption rather than an effect on osteoblasts and bone formation. We found that CB2^–/– mice lost slightly less bone than wild-type mice after ovariectomy, indicating that CB2 deficiency may result in a subtle defect in osteoclastic bone resorption.

Treatment with AM630 at a low dose had no significant protective effect on ovariectomy-induced bone loss in CB2^–/– mice, but at a higher dose of 1.0 mg/kg·d, AM630 was equally effective at preventing ovariectomy-induced bone loss in CB2^–/– mice and wild-type littermates, probably due to an effect on CB1.

In summary, the results reported here indicate that CB2 regulates osteoclast differentiation and function in vitro and ovariectomy-induced bone loss in vivo. We have confirmed that pharmacological inverse agonists of cannabinoid receptors inhibit osteoclast differentiation and bone resorption by a CB2-mediated pathway as well as by interacting with CB1 as previously reported.

Conversely, it appears that the stimulatory effects of cannabinoid receptor agonists on osteoclast formation and bone resorption that we and others have previously observed can also be mediated by an interaction with either or both receptors.

These data suggest that cannabinoid receptor inverse agonists have potential value as antiresorptive drugs that might be of clinical value in the prevention and treatment of diseases such as osteoporosis that are characterized by increased osteoclastic bone resorption.

They also raise the possibility that pharmacological or recreational use of cannabinoid receptor agonists might increase the risk of osteoporosis, by causing increased bone resorption.

	Footnotes

Disclosure Statement: S.H.R., R.V.H., and A.I.I. are inventors on patent applications to protect the use of cannabinoid receptor ligands as treatments for bone disease. A.S. and E.B.-L. have no interests to declare.

First Published Online July 17, 2008

 References

Khosla S 2001 The OPG/RANKL/RANK system. Endocrinology 142:5050–5055

Idris AI, Van't Hof RJ, Greig IR, Ridge SA, Baker D, Ross RA, Ralston SH 2005 Regulation of bone mass, bone loss and osteoclast activity by cannabinoid receptors. Nat Med 11:774–779

Ofek O, Karsak M, Leclerc N, Fogel M, Frenkel B, Wright K, Tam J, Attar-Namdar M, Kram V, Shohami E, Mechoulam R, Zimmer A, Bab I 2006 Peripheral cannabinoid receptor, CB2, regulates bone mass. Proc Natl Acad Sci USA 103:696–7

Tam J, Ofek O, Fride E, Ledent C, Gabet Y, Muller R, Zimmer A, Mackie K, Mechoulam R, Shohami E, Bab I 2006 Involvement of neuronal cannabinoid receptor, CB1, in regulation of bone mass and bone remodeling. Mol Pharmacol 70:786–792

Tam J, Trembovler V, Di M, V, Petrosino S, Leo G, Alexandrovich A, Regev E, Casap N, Shteyer A, Ledent C, Karsak M, Zimmer A, Mechoulam R, Yirmiya R, Shohami E, Bab I 2008 The cannabinoid CB1 receptor regulates bone formation by modulating adrenergic signaling. FASEB J 22:285–294[Abstract/Free Full Text]

 Top

 Home

PMID:

18571910

[PubMed - indexed for MEDLINE]

 

 

 Top

The cannabis plant has been a source of medicinal preparations since the earliest written records on pharmacobotany. A major obstacle to acceptance of the drug has been its potent psychoactivity.

This problem has been studied in recent years in attempts to discover synthetic analogs that would retain medicinal properties without the psychotropic effects.

A class of cannabinoid, the carboxyl tetrahydrocannabinols, shows potential as therapy that is free from cannabimimetic central nervous system activity.

These substances, which are metabolites of tetrahydrocannabinol (THC), the psychoactive principle of Cannabis, do not produce behavioral changes in humans at doses several times greater than THC doses given to the same volunteers.

The parent compound in this series, the THC metabolite THC-11-oic acid (Fig. 1 ), is effective in animal models of inflammation and pain at oral doses of 20–40 mg/kg. However, more potent activity is needed for clinical use.

Figure 1.

Cannabinoid structures. A) Δ9-THC (tetrahydrocannabinol) is the component of Cannabis responsible for psychoactivity. B) THC-11-oic acid is a nonpsychoactive metabolite of Δ9 THC. C) AjA (1′,1′-dimethylheptyl THC-11-oic acid) is a nonpsychoactive synthetic analog of THC-11-oic acid.

It has been known for some time that modifications of the pentyl side chain of THC increase its potency.

In particular, extending the chain length to 7 carbons and introducing branching close to the ring leads to compounds with potencies that are 50–100 times greater than that of THC.

This strategy was employed in designing the structure of 1′,1′-dimethylheptyl-THC-11-oic-acid [trivial name ajulemic acid (AjA)] (5). This synthetic analog of THC-11-oic acid is a potent anti-inflammatory and analgesic agent in several animal models .

We have shown (8)that oral administration of AjA at a dose of 0.1 mg/kg 3×/wk reduces significantly the severity of adjuvant-induced olyarthritis in rats.

Histomorphological evaluation of the joints suggested that although synovial inflammation occurred in AjA-treated animals, it did not progress to cartilage degradation, bone erosion, and distortion of joint architecture as observed in rats given placebo. Thus, AjA treatment appears to facilitate resolution of inflammation in this animal model of joint tissue injury.

Lipoxins are a new class of eicosanoids that arise from the sequential actions of lipoxygenases. Because these compounds are generated through an interaction between lipoxygenase pathways, the term lipoxins (lipoxygenase interaction products) was introduced.

In humans, lipoxins are formed in vivo during multicellular responses such as inflammation. They serve as stop signals in that they prevent leukocyte-mediated tissue injury and stimulate the uptake of apoptotic polymorphonuclear leukocytes (PMNs) at sites of inflammation, thereby facilitating resolution of inflammation. A major problem in joints of patients with rheumatoid arthritis is that inflammation—designed to be protective—often does not resolve. Stable analogs of lipoxin A₄ (LXA₄) block chemotaxis of PMNs and reduce inflammation in animal models. As noted, AjA treatment of rats with adjuvant arthritis appears to promote resolution of synovial inflammation. Therefore, we designed studies to determine whether AjA—added to cells in vitro or administered to mice in vivo—is associated with an increase in LXA₄.

We present results here that indicate that AjA does stimulate production of LXA₄, a process associated with counterregulation of an inflammatory response in a murine model of peritoneal inflammation.

RESULTS

AjA increased LXA₄ production by TNFα stimulated PMNs in a dose-dependent manner (Fig. 2A ). In a series of 4 experiments, 30 μM AjA increased LXA₄ release from TNFα-stimulated PMNs 2.60 ± 0.35-fold (mean±sd); P = 0.04 vs. untreated control cells.

Figure 2.

AjA stimulates LXA₄ production by human cells. A) PMNs (3×10⁶) incubated with vehicle alone (0 AjA) or treated 1 h with AjA and stimulated 4 h with 10 ng/ml rhTNFα. Supernatant LXA₄ determined by ELISA. Values are means of 6 replicates; error bars represent sd. *P < 0.04 vs. 0 AjA; Student’s t test (n=6). B) Peripheral blood incubated without or with AjA (3, 10, 30 μM) for 5 h. Supernatant LXA₄ concentration measured by ELISA. Values are means of 5 experiments with blood from 5 separate donors; error bars are sd. *P < 0.05 vs. 0 AjA; ANCOVA (3 and 10 μM, n=4; baseline and 30 μM, n=5).

Optimal generation of LXA₄ requires transcellular biosynthesis. The addition of AjA to whole blood in vitro also enhanced production of LXA₄ in a dose-dependent manner (Fig. 2B ). In cells in which LXA₄ was increased maximally by zymosan, LXA₄ was not increased substantially more by AjA (not shown). The observations indicate that AjA is by itself an agonist for LXA₄. Although the ratio of AjA-induced LXA₄ to baseline LXA₄ was similar across experiments, the absolute amounts of LXA₄ induced by stimulation of cells by zymosan and/or by AjA varied widely from experiment to experiment. Therefore, results from 3 experiments presented in Fig. 2B were assessed by ANCOVA.

Although joint tissue injury in patients with rheumatoid arthritis is likely due to a multicellular assault on cartilage and bone, studies in animals and humans suggest that joint damage can proceed with participation of synovial cells alone.

Therefore, we examined the influence of AjA on LXA₄ production by 4 different cultures of human FLSs. We have not observed differences in responses of FLSs to AjA whether cells were derived from patients with inflammatory polyarthritis (rheumatoid, psoriatic) or osteoarthritis. Addition of AjA to FLSs increased LXA₄ release from unstimulated and TNFα (1 ng/ml)-stimulated cells in a concentration-dependent manner (Fig. 3 ). In these experiments, TNFα did not stimulate LXA₄ maximally (not shown). Thus, AjA did stimulate further LXA₄ production in TNFα-stimulated cells.

Figure 3.

AjA stimulates LXA₄ production by human FLSs. FLSs were incubated with vehicle (0 AjA) or AjA (10, 20, 30 μM) for 1 h, then stimulated with 1 ng/ml TNFα for 18 h. Supernatant LXA₄ determined by ELISA. Values are means of 4 experiments; error bars are sd. *P < 0.05 vs. 0 AjA; ANCOVA (n=4).

We next assessed the actions of AjA in an in vivo murine model of acute inflammation. Results from these experiments in which 6 CD-1 mice were administered 10 mg/kg/day AjA by mouth for 3 days (total daily dose ∼0.3 mg) before intraperitoneal injection of zymosan are shown in Fig. 4 . AjA treatment reduced the total number of cells invading the peritoneum by 69%.

Figure 4.

Treatment of murine peritonitis with oral administration of AjA. Male CD-1 mice were treated by mouth with 10 mg/kg/day AjA or with safflower oil (vehicle control) for 3 days before i.p. injection of 1 mg zymosan A.

Peritoneal fluid obtained 3 h after zymosan installation. Values are means of total cell counts determined by light microscopy; error bars are sd. *P = 0.03 vs. untreated controls; ANCOVA, two separate experiments, 3 mice/group (n=6). NS, no stimulation.

Having shown an induction of LXA₄ by AjA in vitro, we next examined whether LXA₄ generation was also increased in vivo. To this end, we analyzed LXA₄ in peritoneal lavages from 4 AjA-treated animals by LC/MS/MS and compared these with lavages from 5 mice that received vehicle. Complete MS/MS analysis of the fraction at 7.2 min resulted in a fragmentation pattern consistent with that for LXA (Fig. 5A ).

The prominent MS/MS product ions that are diagnostic for LXA₄ were noted at m/z 351 ([M-H]) (Fig. 5A, a ), m/z 333 ([M-H]–H₂O), m/z 315 ([M-H]–2H₂O), m/z 307 ([M-H]–CO₂ (Fig. 5A, b ), m/z 289 ([M-H]–H₂O–CO₂), m/z 271 ([M-H]–2H₂O–CO₂), m/z 251 ([M-H]–CHO(CH2)4CH₃) (Fig. 5A, d ), m/z 233 ([M-H]–CHO(CH₂)4CH₃–H₂O), m/z 207 ([M-H]–CHO(CH₂)4CH₃–CO₂), m/z 189 ([M-H]–CHO(CH₂)4CH₃–H₂O–CO₂), m/z 235 ([M-H]–CHO(CH₂)3COOH) (Fig. 5A, c ), and m/z 115 (CHO(CH2)3COOH) (Fig. 5A, c′ ).

Having confirmed the identity of the material eluted at 7.2 min as LXA₄ via MS/MS identification of diagnostic ions, it was then determined that this material represented ∼0.5 ng of LXA₄ (Fig. 5A ). When compared to baseline levels of 0.06 ng/mouse obtained in the mice who received vehicle only, this represented a 7.33-fold increase in LXA₄ generation (Fig. 5B ).

AjA treatment reduced total cell counts in peritoneal lavages but did not reduce numbers of mononuclear cells significantly (not shown). The reduced total cell count was due to a reduction in the PMN counts in peritoneal lavages from AjA-treated mice (Fig. 5C ). Taken together, these results demonstrate that LXA₄ generation is associated with reduction of PMN infiltration by AjA in vivo.

Figure 5.

AjA stimulation of endogenous LXA₄ generation and reduction of peritoneal infiltration of PMNs in zymosan-A-induced peritonitis. AjA (1.5 mg/kg/100 μl sterile saline) was injected into tail vein of 6- to 8-wk-old male FVB mice, followed by i.p. injection of 1 mg zymosan A.

Peritoneal lavages were collected, volumes were measured, and 2 vol of methanol was added. Following solid-phase extraction, samples were then analyzed using LC/MS/MS. Aliquots of lavages were assessed for PMN concentration as described in Materials and Methods. A) MS/MS spectrum of AjA-treated samples.

Chromatogram spectra are representative of n = 4; a–d and c′ denote prominent diagnostic ions used for identification (see inset). a, ([M-H]) = m/z 351; b, ([M-H]–CO₂) = m/z 307; c, ([M-H]–CHO(CH₂)3COOH) = m/z 235; c′, (CHO(CH2)3COOH) = m/z 115; d, ([M-H]—CHO(CH2)4CH3) = m/z 215. B) LXA₄ generation quantified in AjA-treated samples vs. control samples. Values represent LXA₄ generation per mouse. Samples were combined before quantitation to allow for adequate detection. C) PMN concentrations quantified in peritoneal lavages from mice treated with AjA vs. untreated controls. Values are means; error bars are sd. *P < 0.05 vs. untreated controls; Student’s t test (n=4).

In an effort to determine which biosynthetic pathway might be influenced by AjA, we treated human whole blood with inhibitors of cytoplasmic phospholipase A₂ (cPLA₂), and 12/15 lipoxygenase (12/15 LOX). Results suggest that most of the increase in LXA₄ initiated by AjA resulted from the conversion of endogenous arachidonic acid to LXA₄ by the 15 LOX-initiated route. The addition to cells of the 12/15 LOX inhibitor baicalein blocked the AjA-induced increase in LXA₄ substantially (Fig. 6 ). It is likely that AjA stimulated the release of arachidonic acid from phospholipid precursors in both murine exudate cells and human whole blood. However, the addition to mouse peritoneal exudate cells or human whole blood of the cPLA₂ inhibitor methylarachidonylfluorophosphate did not alter the AjA-induced increase in LXA₄ (data not shown), thus precluding our assessment of specific phospholipases in the process.

Figure 6.

Blockade of AjA enhancement of LXA_4. Human blood was incubated with AjA for 1 h, and LXA₄ was assessed as described in Materials and Methods. AjA (25 μM)-induced production of LXA₄ was reduced significantly when the 12/15 LOX inhibitor baicalein was added to cells before incubation with AjA. Values are means; error bars represent sd. *P < 0.05 vs. AjA alone; Student’s t test (n=3).

• Received July 24, 2008.
• Accepted November 26, 2008.

• FASEB

 

 Top

Researchers of the University of Wroclaw, Poland, investigated the effects of an ointment with structured physiological lipids and endocannabinoids in 21 patients with pruritus due to end-stage failure of kidney function. So-called uremic pruritus is still a common symptom in patients with end-stage renal failure.

However, there is no effective treatment for this condition. All patients applied the tested cream twice daily for a period of three weeks. Global pruritus and dry skin were examined before the trial, on study visits at weekly intervals, and two weeks after completion of the study.

After 3-week therapy pruritus was completely eliminated in 8 patients. Dry skin was significantly improved. Researchers noted that "it is very probable that the observed decrease of pruritus with the test product therapy was not only the result of dry skin improvement but that the addition of endocannabinoids may have also played a role."

(Source: Szepietowski JC, Szepietowski T, Reich A. Efficacy and tolerance of the cream containing structured physiological lipids with endocannabinoids in the treatment of uremic pruritus: a preliminary study. Acta Dermatovenerol Croat 2005;13(2):97-103.)

 Top