ACCELERATED COMMUNICATION Involvement of Neuronal Cannabinoid Receptor CB1 in Regulation of Bone Mass and Bone Remodeling

The CB1 cannabinoid receptor has been implicated in the regulation of bone remodeling and bone mass. A high bone mass (HBM) phenotype was reported in CB1-null mice generated on a CD1 background (CD1 / mice). By contrast, our preliminary studies in cb1 / mice, backcrossed to C57BL/6J mice (C57 / mice), revealed low bone mass (LBM). We therefore analyzed CB1 expression in bone and compared the skeletons of sexually mature C57 / and CD1 / mice in the same experimental setting. CB1 mRNA is weakly expressed in osteoclasts and immunoreactive CB1 is present in sympathetic neurons, close to osteoblasts. In addition to their LBM, male and female C57 / mice exhibit decreased bone formation rate and increased osteoclast number. The skeletal phenotype of the CD1 / mice shows a gender disparity. Female mice have normal trabecular bone with a slight cortical expansion, whereas male CD1 / animals display an HBM phenotype. We were surprised to find that bone formation and resorption are within normal limits. These findings, at least the consistent set of data obtained in the C57 / line, suggest an important role for CB1 signaling in the regulation of bone remodeling and bone mass. Because sympathetic CB1 signaling inhibits norepinephrine (NE) release in peripheral tissues, part of the endocannabinoid activity in bone may be attributed to the regulation of NE release from sympathetic nerve fibers. Several phenotypic discrepancies have been reported between C57 / and CD1 / mice that could result from genetic differences between the background strains. Unraveling these differences can provide useful information on the physiologic functional milieu of CB1 in bone. The endogenous cannabinoids bind to and activate the CB1 and CB2 cannabinoid receptors. Both are seven-transmembrane domain receptors and they share 44% identity. They are coupled to the Gi/o subclass of G-proteins and inhibit stimulated adenylyl cyclase activity (Rhee et al., 1997). CB1 is present in the brain and in peripheral neurons and accounts for most of the central nervous system actions of cannabinoid drugs and endocannabinoids (Herkenham et al., 1990; Zimmer et al., 1999). CB2 is found mainly in the immune system (Munro et al., 1993). In vertebrates, bone mass and shape are determined by continuous remodeling consisting of the concerted and balanced action of osteoclasts, cells that resorb bone, and This work was supported by National Institute on Drug Abuse grants DA9789 (to R.M.) and DA00286, DA11322 (to K.M.), and Israel Science Fondation (ISF) grants 482/01 and 4007/02-Bikura (to I.B. and E.S.). Purchase of the CT system was supported in part by ISF grant 9007/01 (to I.B.). J.T. and O.O. contributed equally to this article. Article, publication date, and citation information can be found at http://molpharm.aspetjournals.org. doi:10.1124/mol.106.026435. ABBREVIATIONS: CD1 / mice, CB1-null mice generated on a CD1 background; C57 / mice, CB1-null mice generated on a C57BL/6J background; M-CSF, macrophage colony-stimulating factor; BV/TV, trabecular bone volume density. 0026-895X/06/7003-786–792$20.00 MOLECULAR PHARMACOLOGY Vol. 70, No. 3 Copyright © 2006 The American Society for Pharmacology and Experimental Therapeutics 26435/3134748 Mol Pharmacol 70:786–792, 2006 Printed in U.S.A. 786 at A PE T Jornals on Jne 1, 2017 m oharm .aspeurnals.org D ow nladed from osteoblasts, cells that form bone. Osteoporosis, the most prevalent degenerative disease in developed countries, results from impaired remodeling balance, which leads to bone loss and increased fracture risk. Bone remodeling is subject to central control through pathways that involve signaling by the hypothalamic receptors for leptin and neuropeptide Y (Ducy et al., 2000; Baldock et al., 2002), which are also associated with the regulation of endocannabinoid brain levels (Di Marzo et al., 2001). Along these lines, we have recently reported that 1) the peripheral CB2 cannabinoid receptor is normally expressed in osteoblasts, osteoclasts, and in their precursors; 2) mice deficient for CB2 have a low bone mass (LBM) phenotype; and 3) specific activation of CB2 attenuates ovariectomy-induced bone loss by restraining osteoclastogenesis and stimulating bone formation (Ofek et al., 2006). Two mutant mouse lines with deficiency in the CB1 gene have been generated. In one line, backcrossed to C57BL/6J mice (C57 / ), almost the entire protein-encoding sequence was removed (Zimmer et al., 1999). In the other line, backcrossed to CD1 mice (CD1 / ), the N-terminal 233 codons of cb1 were ablated (Ledent et al., 1999). Although both lines demonstrate a null mutation missing all the CB1 responsiveness to cannabinoid ligands, they display significant phenotypic discrepancies (Lutz, 2002; Hoffman et al., 2005). Regarding the skeleton, CD1 / mice have a high bone mass (HBM) phenotype, suggesting that activation of CB1 down-regulates bone mass (Idris et al., 2005). By contrast, our preliminary studies in C57 / mice pointed to a LBM phenotype occurring in the absence of functional CB1 receptors. These studies used different methods to characterize the skeletal phenotype. Hence, in an attempt to solve this critical discrepancy, we analyzed the expression and distribution of CB1 in bone and compared the skeletons of the C57 / and CD1 / mouse lines using identical methods, equipment, and expertise. Fig. 1. CB1 is expressed in sympathetic nerve fibers in trabecular bone. A and B, RT-PCR analysis. A, bone marrow stromal cells undergoing osteoblastic differentiation in “osteogenic medium”; note absence of CB1positive bands. PTHRc1, PTH receptor 1; TNSALP, tissue nonspecific alkaline phosphatase. NOM, cells grown for 20 days in nonosteogenic medium (negative control). Lane C, cerebellar mRNA (positive control). B, bone marrow-derived monocytic cells undergoing osteoclastogenesis in medium containing M-CSF and RANKL; Ocl, osteoclast-like tartrate-resistant acid phosphatase–positive multinucleated cells. Mon, undifferentiated monocytes. Lane C, cerebellar mRNA (positive control). Cells were derived from WT C57BL/6J mice. C–H, immunohistochemical staining in secondary spongiosa in distal femoral metaphysis from C57BL/6J (C–F) and CD1BL/6J (G and H) mice. C, D and G, anti-tyrosine hydroxylase antibodies; E, F, and H, consecutive sections stained with anti CB1 antibodies. T, bone trabeculae; arrows, osteoblasts; arrowheads, tyrosine hydroxylaseCB1 positive fibers. Hematoxylin counterstaining. CB1 Cannabinoid Receptor and Bone 787 at A PE T Jornals on Jne 1, 2017 m oharm .aspeurnals.org D ow nladed from Materials and Methods Animals. All animals in the study were 9to 12-week-old mice. C57 / mice were generated as reported previously (Zimmer et al., 1999). We have crossed heterozygous animals of this line for at least 10 generations to WT C57BL/6J mice. Heterozygous animals from the last generation were then intercrossed to obtain congenic C57BL/6J mice that are homozygous for the respective mutation. CD1 / mice were generated by homologous recombination as described previously (Ledent et al., 1999). Heterozygous mice were bred for 17 generations on a CD1 background before generating the WT and cb1-null littermates used in this study. To study bone formation, newly formed bone was vitally labeled in mice intended for microcomputed tomographic ( CT)-histomorphometric analysis by the fluorochrome calcein (Sigma, St. Louis, MO), injected i.p. (15 mg/kg) 4 days and 1 day before sacrifice. The use of animals was approved by the Institutional Animal Care Committee of the Hebrew University of Jerusalem. Immunohistochemistry. Mice were killed by transcardial perfusion of phosphate-buffered saline followed by 4% paraformaldehyde. The femora were dissected and further fixed with paraformaldehyde for 2 h at 4°C. The specimens were decalcified in 0.5 M EDTA, pH 7.4, and embedded in paraffin. For immunohistochemical analysis, serial 5m frontal sections were reacted with anti-tyrosine hydroxylase (TH) (Chemicon, Temecula, CA) or anti-CB1 antibodies (Nyı́ri et al., 2005). Further processing was carried out using the SuperPicture polymer detection Kit (Zymed Laboratories, South San Francisco, CA) according to the manufacturer instructions. Cell Cultures and mRNA Analysis. Primary bone marrow stromal cell cultures from WT adult femoral and tibial diaphyseal bone marrow were established as described previously. For testing CB1 expression, the cells were grown in osteogenic medium (Ofek et al., 2006). Bone marrow-derived osteoclastogenic cultures were established from Ficoll-separated monocytic precursors and grown for 5 to 6 days in medium containing macrophage colony-stimulating factor (M-CSF) and RANK ligand (RANKL; R&D Systems, Minneapolis, MN) (Ofek et al., 2006). Total RNA was extracted from the cells, purified, and reverse-transcribed using routine procedures. The following primers were used for PCR: CB1, sense: 5 -TGGTGTATGATGTCTTTGGG-3 , antisense: 5 -ATGCTGGCTGTGTTATTGGC-3 ; tissue nonspecific alkaline phosphatase, sense: 5 -GACACAAGCATTCCCACTAT-3 , antisense: 5 -ATCAG-CAGTAACCACAGTCA-3 ; parathyroid hormone receptor I, sense: 5 -CAAGAAGTGGATCATCCAGGT-3 , antisense: 5 -GCTGCTACTCCCACTTCGTGCTTT-3 ; and -actin, sense, 5 -GAGACCTTCAACACCCCAGCC-3 ; antisense, 5 -GGCCATCTCTTGCTCGAAGTC-3 . CT Analysis. Whole femora were examined by a CT system ( CT 40; SCANCO Medical, Bassersdorf, Switzerland) as reported recently (Bajayo et al., 2005; Ofek et al., 2006). Scans were performed at a resolution of 20 m in all three spatial dimensions. Morphometric parameters were determined as reported previously (Kram et al., 2006). Trabecular and cortical bone parameters were measured in metaphyseal and mid-diaphyseal segments, respectively. Histomorphometry. After CT image acquisition, the specimens were embedded undecalcified in Technovit 9100 (Heraeus). LongituFig. 2. Bone mass phenotype in CB1-null mice. Tri-dimensional CT images of distal femoral metaphysial trabecular bone from mice with median BV/TV values. Quantitative data are mean S.E.M. obtained in 16 C57 / and 16 WT C57BL/6J mice per gender, six CD1 / and six WT CD1 male mice and 20 CD1 / and 20 WT CD1 female mice. , P 0.05. 788 Tam et al. at A PE T Jornals on Jne 1, 2017 m oharm .aspeurnals.org D ow nladed from dinal sections through the midfrontal plane were left unstained for dynamic histomorphometry, based on the vital calcein double labeling. To identify osteoclasts, consecutive sections were stained for tartrate-resistant acid phosphatase. Parameters were determined according to a standardized nomenclature (Parfitt et al., 1987). Statistical Analysis. Differences between cb1 / and WT mice were analyzed with the use of the Student’s t test. Results and Discussion Expression of CB1 in Bone. mRNA analyses were carried out in cells derived from WT C57BL/6J mice. Unlike the expression of CB2, which is absent in undifferentiated bone marrow stromal cells but increases progressively when these cells undergo osteoblastic differentiation (Ofek et al., 2006), we were unable to identify CB1 mRNA transcripts in either undifferentiated or differentiated stromal cells, even after 40 PCR cycles (Fig. 1A). Monocytic osteoclast precursors from these mice also did not show CB1 expression. However, a weak signal was present when these cells underwent osteoclastogenesis with M-CSF and RANKL (Fig. 1B). Bone, especially trabecular bone, is densely innervated by sympathetic fibers (Serre et al., 1999; Mach et al., 2002). These fibers release norepinephrine, thus potently mediating central signals that restrain bone formation and stimulate bone resorption (Elefteriou et al., 2005). Because CB1 is expressed in such nerve fibers elsewhere (Schlicker and Kathmann, 2001), we further explored its presence in bone sympathetic nerve fibers. Indeed, immunohistochemical analysis using the sympathetic marker TH (Bjurholm et al., 1988) confirmed the occurrence of a network of TH-positive fibers in the intertrabecular spaces of cancellous bone in both C57BL/6J and CD1 mice (Fig. 1, C and G). The fibers were close to the bone trabeculae with terminal nerve processes penetrating the osteoblast palisades, thus being in intimate proximity to these cells (Fig. 1, D and G). Consecutive histological sections show CB1 immunoreactivity of the same nerve fibers (Fig. 1, E, F, and H), indicating the presence of CB1 receptors in sympathetic fibers that innervate the trabecular bone. This CB1 immunoreactivity was missing in the CB1-null mice (data not shown). Fig. 3. CT-based structural morphometric parameters in secondary spongiosa of distal femoral metaphysis of CB1-null mice. A–C, C57 / mice and their WT C57BL/6J control mice; D–F, CD1 / mice and their WT CD1 control mice. A and D, trabecular number. B and E, trabecular thickness. C and F, trabecular connectivity. Data are mean S.E.M. obtained in 16 C57 / and 16 WT C57BL/6J mice per gender, six CD1 / and six WT CD1 male mice and 20 CD1 / and 20 WT CD1 female mice. , P 0.05. CB1 Cannabinoid Receptor and Bone 789 at A PE T Jornals on Jne 1, 2017 m oharm .aspeurnals.org D ow nladed from Skeletal Phenotype of CB1-Null Mice. Our results demonstrate that the background WT strains, in which the C57 / and CD1 / mouse lines had been established, display vast differences in both trabecular and cortical bone mass. More importantly, cb1 inactivation in these lines resulted in opposing skeletal effects (Figs. 2–4). Compared with their WT control mice, both male and female C57 / mice exhibited low bone mass (LBM) phenotype characterized by a lower density of their trabecular network. The trabecular bone volume density (BV/TV) in female and male null mice was 20 and 15% lower than that of WT C57BL/6J control mice, respectively (Fig. 2). Apparently, the lower BV/TV in the C57 / mice resulted from decreases in the trabecular number (Fig. 3A) without changes in the trabecular thickness (Fig. 3B). The trabecular connectivity density, a parameter measuring the structural integrity of the trabecular network (Stampa et al., 2002), was also decreased in these animals (Fig. 3C) but did not reach statistical difference. In addition, both the diaphyseal shaft diameter and medullary cavity diameter were narrower in the C57 / mice (Fig. 4, A and B), with unchanged cortical thickness (Fig. 4C). By contrast, the CD1 / skeletal phenotype showed a marked gender bias. The trabecular bone, the main skeletal compartment affected in osteoporosis, appeared normal in female CD1 / mice (Figs. 2 and 3, D–F). Male CD1 / mice had a pronounced HBM phenotype demonstrating 27.5% increase in trabecular BV/TV (Fig. 2) accompanied by increased trabecular thickness (Fig. 3E) and slightly decreased connectivity density (Fig. 3F). The female CD1 / diaphysis was mildly abnormal, exhibiting cortical expansion portrayed as increases in both diaphyseal shaft diameter and medullary cavity diameter (Fig. 4, D and E). The male CD1 / diaphysis appeared normal (Fig. 4, D–F). To gain further insight into the processes leading to the LBM phenotype in C57 / mice, we analyzed their bone remodeling. Consistent with the results of the structural CT parameters, the histomorphometric analysis demonstrated that the LBM in these mice is associated with unbalanced bone remodeling. The bone formation rate was markedly decreased in both female and male mice (Fig. 5A), mainly because of a decrease in mineral appositional rate, a surrogate of osteoblast activity (Fig. 5B), inasmuch Fig. 4. CT-based measurements of diaphyseal dimensions in CB1-null mice. A–C, C57 / mice and their WT C57BL/6J control mice; D–F, CD1 / mice and their WT CD1 control mice. A and D, overall mid-diaphyseal diameter. B and E, mid-diaphyseal medullary cavity diameter. C and F, cortical thickness. Data are mean S.E.M. obtained in 16 C57 / and 16 WT C57BL/6J mice per gender, six CD1 / and six WT CD1 male mice and 20 CD1 / and 20 WT CD1 female mice. , P 0.05. 790 Tam et al. at A PE T Jornals on Jne 1, 2017 m oharm .aspeurnals.org D ow nladed from as the mineralizing perimeter, a surrogate of osteoblast number, remains unchanged (Fig. 5C). The osteoclast number was increased, significantly in female mice and insignificantly in male mice (Fig. 5D). We were surprised to find no significant differences in bone remodelling parameters between the CD1 / mice and their WT control mice, even not in male mice (Table 1). Together, these results suggest that in the C57BL/6J mice, CB1 signaling positively regulates trabecular bone mass and radial diaphyseal growth by up-regulating bone formation and down-regulating bone resorption. The absence of significant changes in bone remodeling parameters of the male CD1 / mice suggests that CB1 in these animals is associated only with the accrual of peak bone mass, which occurs at a younger age than that studied here. Apparently, the LBM phenotype is exhibited in the C57 / mice consequent to a decrease in bone formation and increase in bone resorption attributable to the absence of sympathetic CB1, which normally inhibits norepinephrine release (Ishac et al., 1996). Although either genetic modification leads to a null mutation missing all CB1 responsiveness to its ligands, the occurrence of phenotypic differences is not entirely surprising, inasmuch as these mouse lines exhibit other substantial discrepancies ranging from nociceptive perception to locomotor activity, life expectancy, and embryo implantation (Lutz, 2002). Furthermore, at least to some extent, skeletal dissimilarity between the C57 / and CD1 / mice could be expected from the differences in bone mass and structure observed between the WT CD1 and C57BL/6J background strains. More surprising is the gender bias portrayed by the