Bones and Joints: The Effects of Cannabinoids on the Skeleton

Abstract

The endocannabinoid system, consisting of cannabinoid ligands and the enzymes involved in their biosynthesis and degradation as well as the cannabinoid receptors, can be found in almost every cell and tissue type (Fig. 1). In addition to Δ⁹-tetrahydrocannabinol (THC), the predominant phytocannabinoid present in Cannabis species (i.e., marijuana), many other exogenous and endogenous substances can activate endocannabinoid receptors throughout the body. In addition to the well-known effect of these agents on the central nervous system (CNS), growing evidence indicates that the endocannabinoid system can influence osteoblast and osteoclast activity, bone mass, and bone remodeling. The expanding legalization of cannabis throughout the world and the growing use of cannabinoids by older individuals at increased risk for osteoporosis (1) have increased attention on the pharmacologic effects of cannabinoids on bone metabolism.

Figure 1.

A schematic overview of endocannabinoid involvement in bone metabolism is shown. The illustration to the right of the long bone shows endocannabinoids, including anandamide, 2-arachidonylglycerol (2-AG), and lipoamines (e.g., oleoyl serine), which are found in the bone microenvironment. Endocannabinoid receptors CB1, CB2, TRPV1, and GPR55 are located on osteoblast and osteoclast cell membranes. Sympathetic innervation of bone is shown in the two illustrations to the left of the long bone. Axons from dorsal root ganglion neurons, depicted in the top left drawing, travel to periosteum and cortical bone. Postganglionic synaptic structures, shown in the bottom left illustration, synthesize endocannabinoids on demand and release these lipophilic compounds into the synaptic cleft, where they travel in a retrograde direction to bind to membrane receptors found on osteoblasts and osteoclasts. Osteoblasts and osteoclasts contain enzymes for endocannabinoid synthesis [N-acyl-phosphatidylethanolamine (NAPE), N-acyl-phosphatidylethanolamine-specific phospholipase (NAPE-PLD), diacylglycerol lipase (DAGL)] and degradation (fatty acid amide hydrolase (FAAH) and monacylglycerol lipase (MAGL). Anandamide preferentially binds to CB2, the predominant endocannabinoid receptor found in bone, and is degraded by FAAH into arachidonate and ethanolamine. 2-AG is metabolized by MAGL into arachidonate and glycerol. DAG, diacylglycerol. Used with permission from Intermountain Healthcare. © 2019 Intermountain Healthcare. All rights reserved.

Open in new tabDownload slide

In this review, we describe the role of the endocannabinoid system in modulating bone and skeletal metabolism and review how specific ligands, through activation of diverse receptor types, contribute to bone remodeling and bone mass. The review is based on a comprehensive search of PubMed and Google Scholar databases as of June 2019 for all English-language publications that relate to cannabinoids and bone. Retrieved articles were evaluated for relevance, experimental design, data acquisition, statistical analysis, and conclusions. Non‒peer-reviewed publications and publications on non‒bone-related aspects of endocannabinoids were excluded from this literature review.

Biology of the Endocannabinoid System

The endocannabinoid system utilizes four classes of organ-specific cellular receptors (2): G protein‒coupled endocannabinoid receptors type 1 (CB1) and type 2 (CB2) (3); transient receptor potential vanilloid (TRPV) proteins (4); orphan G protein‒coupled receptors (5); and peroxisome proliferator-activated receptor (PPAR) signaling pathways (6). The system also comprises the endogenous ligands that bind to these receptors (7, 8) as well as the enzymes that are required for synthesis and degradation of the endocannabinoids (9, 10) (Table 1).

Endocannabinoids are derived from cell membrane lipids; because they are highly lipophilic, they can readily cross the blood-brain barrier. Their cellular effects, though tissue specific, are principally dependent on widely expressed CB1 and CB2 cannabinoid receptors that are coupled to intracellular effector pathways via the G proteins G_i and G_o (11). Activation of these receptors leads to inhibition of adenylyl cyclase, which leads to a decrease in intracellular cyclic AMP and activation of phospholipase C, which in turn leads to increases in intracellular calcium and activation of protein kinase C (12, 13). These intracellular events collectively trigger an integrated array of genomic and nongenomic effects (14). The endocannabinoid system modulates synaptic transmission, immune responses, inflammation, pain, carbohydrate metabolism, and thermoregulation (15).

THC and cannabadiol (CBD), two of nearly 500 compounds found in Cannabis species, including at least 120 other cannabinoids, are the most studied and best-known phytocannabinoids. THC acts as an agonist for CB1 and CB2 (16). CBD is a weak CB1 and CB2 antagonist with effects on TRPV1 and TRPV2, the orphan G protein‒coupled receptor GPR55, 5-hydroxytryptamine, and dopamine receptors (17, 18); however, the mechanism of action for CBD is not well understood (19). Genetic and pharmacologic manipulation of the endocannabinoid system as well clinical observations (20) have yielded intriguing and often unexpected insights into the role of this evolutionary ancient cellular signaling pathway on neural, immune, and metabolic functions (21).

The Medical History of Endocannabinoids: Bedside to Bench to Bedside

Therapeutic use of cannabis for treatment of neuropsychiatric, gastrointestinal, ophthalmic, and obstetric conditions dates back more than 3000 years (22). O’Shaughnessy, an Irish physician working for the British East India Company in Calcutta, introduced the therapeutic use of marijuana to Western medicine in 1833 (23). On the basis of Indian folk uses of cannabis, he conducted a series of animal experiments and human studies. These studies showed that hashish, the cannabis inflorescence exudate consumed by humans, had analgesic, anxiolytic, and muscle relaxant properties (24). A 2017 National Academy of Medicine review of the medical cannabis literature confirmed O’Shaughnessy’s observations (25).

The discovery of the endocannabinoid system arose from a series of studies designed to identify the mechanism of action of cannabis. Using synthetic cannabinoids that were produced in the 1940s to understand the analgesic properties of phytocannabinoids (26), investigators in the 1980s performed a series of structure-activity experiments that demonstrated cannabinoid’s stereospecificity, which in turn suggested the existence of distinctive cannabinoid receptors (27). On the basis of evidence for stereospecific cannabinoid receptors, coupled with the knowledge that phytocannabinoids reduce concentrations of intracellular cyclic AMP in the rat brain as well as the availability of radiolabeled cannabinoid analogues (28), investigators were able to identify, characterize, and clone the G protein‒coupled cannabinoid receptor CB1 (29), encoded by the Cnr1 gene located at 6q15 (30). CB1 is found primarily on the presynaptic terminal boutons of central and peripheral neurons (31), and variant transcripts encode three CB1 isoforms that exhibit differential ligand specificities (32). Endocannabinoids in the CNS have both genomic and nongenomic effects. Nongenomic effects are mediated by CB1-dependent decreases in cell membrane potassium gradients and protein kinase C activity. Release of presynaptic glutamate and γ−aminobutyric acid is decreased and is thought to account for the effects of endocannabinoids on the pituitary-adrenal stress response (33).

In addition to CB1, a second endocannabinoid receptor, CB2, was subsequently identified through studies of G protein‒coupled receptors as a mediator of cannabinoid action outside the CNS (34). CB2 is encoded by the Cnr2 gene located at 1p36.11 (35). CB2 mRNA expression is greatest in the lymph nodes, spleen, tonsils, and appendix (36). The CB2 receptor appears to function in the inflammatory response, nociceptive neural transmission, and bone homeostasis (37). CB2A and CB2B, two distinct transcripts that differ only in the 5′ untranslated region, are produced by tissue-specific use of alternative promoters (35). CB2 protein is increased in the spinal cord tissue of patients with multiple sclerosis and amyotrophic lateral sclerosis (38), in an animal model of peripheral neuropathy (36), and in hepatic fibrosis (39). CB2-selective agonists regulate cell proliferation, differentiation, transformation, and death by stimulating major components of the MAPK pathway, including ERK1/2, p38, and c-Jun N-terminal kinases (40). CB2 also induces apoptosis, necrosis, and autophagy through modulations of the Akt-phosphoinositide 3-kinase pathways, as well as modulating arrestin activity and ceramide production (41‒44). Structurally, CB2 shares approximately 50% amino acid homology with CB1 in the transmembrane region; however, these two receptors are functionally distinct, as they exhibit differences in binding affinity for cannabinoid agonists and antagonists (45).

The endocannabinoid system is present in all chordates. According to phylogenic comparisons and the limited sequence homology between the Cnr1 and Cnr2 genes, it is thought that the gene duplication event that gave rise to these two genes must have occurred before teleosts and tetrapods diverged from a common ancestor some 350 to 400 million years ago (46). In addition to CB1 and CB2, additional endocannabinoid receptors, such as TRPV1 and orphan G protein receptors (GPRs) 18, 19, 55, and 119, were identified as a result of observations that cannabinoids such as cannabinol and CBD, which do not bind to CB1 or CB2, have pharmacologic effects in Cnr1 and Cnr2 knockout mice (47–50). Lastly, both endocannabinoids and phytocannabinoids have agonist activity on PPARα and PPARγ (6). Cannabinoid signaling via PPARα may play a role in the pathogenesis of diabetic ketoacidosis and could account for the increased incidence of diabetic ketoacidosis in self-reported cannabis users (51, 52).

The pharmaceutical industry has developed synthetic cannabinoids to treat pain, epilepsy, chemotherapy-induced nausea and vomiting, metabolic syndrome, asthma, and glaucoma (22) (53‒55). Dronabinol is an US Food and Drug Administration (FDA)‒approved synthetic form of THC used in the treatment of chemotherapy-related nausea and vomiting (56). CBD was recently approved by the US FDA for the treatment of seizures in patients with Lennox-Gastaut and Dravet syndromes (57, 58). An oromucosal spray formulation of THC and CBD, is approved in the United Kingdom to treat spasticity, neuropathic pain, and symptoms of multiple sclerosis; it has not been approved in the United States (59, 60). Moreover, several other drugs that target the endocannabinoid system are in phase 2 and 3 clinical trials, including lipoamines for the treatment of metabolic bone disease. Rimonabant, a synthetic endocannabinoid antagonist, was briefly marketed in Europe and South America for weight loss in patients with metabolic syndrome, but it was withdrawn from the market because of major untoward psychiatric effects (61).

Endocannabinoid Involvement in Bone Homeostasis

Demonstration that the endocannabinoid system is involved in bone metabolism emerged from experimental observations that linked the CNS via endocannabinoids to regulation of bone growth (62, 63). Mouse models in which Cnr1, Cnr2, or both genes were deleted experimentally provided insights into the roles of the CB1 and CB2 endocannabinoid receptors in the regulation of bone mass. Table 2 summarizes the results of experiments in Cnr1 and Cnr2 knockout mice. Although both receptors have substantial effects on the skeleton, CB2 is more highly expressed than CB1 in bone cells, including osteoblasts, osteocytes, and osteoclasts (64). Additional evidence for a role of the endocannabinoid system in the skeleton comes from observations that the endocannabinoids anandamide and 2-arachidonoylglycerol (2-AG), as well as enzymes for their synthesis and degradation, are present in the bone microenvironment (65, 66).

Pharmacological and genetic studies have highlighted the effect of genetic background on the influence of the endocannabinoid system on skeletal homeostasis. Idris et al. (67) showed that female adult Cnr1 knockout mice on inbred (ABH) and outbred (CD1) backgrounds have high peak bone mass and are protected against ovariectomy-induced loss of trabecular bone due to an osteoclast defect. Bone mineral density was greater in the spine and femur of knockout mice than in wild-type littermates. Bone histomorphometry showed that Cnr1 knockout mice had significantly increased trabecular bone volume at the tibial metaphysis compared with wild-type mice. Moreover, no differences were observed between Cnr1 knockout mice and their wild-type littermates in terms of osteoclast numbers, eroded surfaces, or osteoblast numbers. These mice lose bone mass as they age owing to a defect in bone formation.

In contrast to these findings, Tam et al. (68) showed that female mice with a Cnr1 deletion had low peak bone mass when inbred onto a C57BL/6 background. Idris et al. (69) confirmed these findings in vitro using primary mouse osteoblast cultures and receptor activator of nuclear factor κB ligand (RANKL)–generated mouse osteoclast cultures. Cannabinoid ligands did not affect osteoblast growth or viability, but the CB1-selective antagonist AM251and the CB2-selective antagonists SR144528 and AM630 significantly inhibited osteoclast formation in RANKL- and macrophage colony-stimulating factor‒stimulated mouse bone marrow cultures in a concentration-dependent fashion. In contrast, cannabinoid receptor agonists stimulated osteoclast formation in a concentration-dependent manner. Moreover, pharmacologic inhibition of CB1 signaling reproduced the effects of Cnr1 gene deletion with increased osteoclast apoptosis, decreased RANKL-stimulated ERK phosphorylation, and nuclear translocation of nuclear factor of activated T-cells, cytoplasmic 1, c-jun, and c-fos (69). These results indicate that CB1 has an important role in the regulation of bone mass and bone loss resulting from estrogen deficiency (70). In addition to direct effects on bone, the loss of CB1 function on bone may reflect an indirect mechanism of action resulting from decreased sympathetic nervous system stimulation of bone, which in turn results in decreased osteoblast activity (71, 72).

Studies on the role of CB2 in the regulation of skeletal homeostasis have demonstrated important, but inconsistent effects that reflect the confounding influences of genetic background, age, and sex on CB2 effects on bone (73). Peak bone mass in the trabecular compartment at the tibial and femoral metaphyses of 3-month-old female Cnr2 knockout mice was significantly greater, with a significantly reduced trabecular patterning factor, than in wild-type littermates. Remarkably, these effects were not observed in the lumbar spine or in male Cnr2 knockout mice (74). In contrast, cortical bone volume at the tibial diaphysis was slightly greater in male Cnr2 knockout mice than in wild-type mice, whereas in female Cnr2 knockout mice, the cortical diameter was slightly but significantly lower than in wild-type littermates (74). Finally, there were no changes in biomechanical characteristics of femoral bones of 3-month-old female Cnr2 knockout mice compared with wild-type mice (75). The findings in Cnr2 knockout mice on a CD1 genetic background contrast with findings of earlier studies of C57BL/6 mice with Cnr2 ablation, in which trabecular bone volume / total volume ratio and bone turnover were reported to be normal in one study and reduced in association with high bone turnover in another study (65).

Differences have also been observed between mouse strains with regard to the effects of CB2 deficiency on the aging skeleton of 12-month-old mice. Aged Cnr2 knockout mice on a C57BL/6 genetic background had significantly greater trabecular bone loss and cortical bone expansion than their wild-type littermates as a result of excessive bone resorption (76). By contrast, Sophocleous et al. (77) found that aged Cnr2 knockout mice on a CD1 background had greater bone mass than aged wild-type mice, but this effect was due to the higher peak bone mass of Cnr2 knockout mice, as no differences were observed in age-related bone loss between wild-type and knockout mice. These seemingly discrepant observations may indicate that sex and genetic background, as well as other unidentified factors, can influence the effects of cannabinoid receptor signaling on specific skeletal sites (78, 79). In addition, the phenotypic differences observed with various knockout models may reflect off-target effects of gene deletion effects (73).

Pharmacological studies demonstrate that activation of CB2 receptors in osteoblasts by endocannabinoids and synthetic agonists can alter levels of RANKL and osteoprotegerin, which suggests that CB2 may participate in the regulation of osteoblast-osteoclast crosstalk. Moreover, activation of CB2 with the CB2-selective nonpsychotropic agonist HU308 protected against bone loss due to estrogen deficiency by stimulating bone formation rather than inhibiting osteoclast number and bone resorption (67, 74). In addition, pharmacological studies have shown that selective blockade of CB2 using the antagonist/inverse agonist AM630 prevented ovariectomy-induced bone loss (80). Comprehensive in vitro and in vivo analyses of bone cell differentiation and activity in adult Cnr2 knockout mice and wild-type littermates after treatment with the CB2-specific inverse agonist AM630 revealed that these effects were likely due to a reduction in osteoclast number and activity rather than an increase in bone formation. Pharmacologic stimulation of CB2 does not suppress ex vivo osteoclast generation in Cnr2 knockout mice, as occurs in similar cultures from wild-type mice (80). In wild-type mice, CB2 agonists blunted bone loss after oophorectomy in association with increased endocortical bone formation. By decreasing skeletal oleoyl serine, a CB2 mimetic that regulates bone turnover, ovariectomy may account for the protective effects of Cnr2 deletion (81). Contrary results were observed with administration of CB2 receptor antagonists, as these agents are reported to inhibit osteoclastogenesis (80). The mitogenic effects of CB2 activation result in phosphorylation of ERK1/2 and increases in MAPK and cyclic AMP response element-binding protein transcription (82). CB2 receptor agonists may also act indirectly to decrease bone resorption via anti-inflammatory effects, including decreased expression of cytokines, which promote bone resorption, TNF, and IL-1 and increased expression of IL-1 receptor antagonist, which is present in bone and suppresses osteoclast formation (74). The interpretation of pharmacologic manipulation of the endocannabinoid system assumes that the effects of agonist, antagonists, and reverse agonists are target specific. This is not the case, however, as these agents do not appear to be entirely specific (48).

Because of the phenotypic differences in bone mass found with deletion of each type of endocannabinoid receptor, studies of the effects of deletion of both the Cnr1 and Cnr2 genes on bone development were performed. These experiments provide additional insight into the cannabinoid regulation of bone growth and turnover (83). Double-knockout mice showed increased trabecular bone mass at 3 months of age compared with controls. Bone in mice lacking both endocannabinoid receptors contained fewer osteoclasts than bone in controls. Osteoclast deficiency was responsible for protecting against bone loss due to estrogen deficiency and aging. Although double-knockout mice had diminished bone formation and reciprocal increased amounts of bone marrow fat, the loss of osteoclasts had a greater effect on bone mineral density than the reduced bone formation observed. These findings are consistent with prior observations that Cnr1 gene deletion results in a phenotype that is distinct from that found with Cnr2 gene deletion.

Observations in zebrafish provide additional support for the role of CB2 in bone metabolism (84). Cannabinoids increased markers of both osteoblast differentiation and synthetic activity without effects on markers of osteoclast differentiation or metabolism in zebrafish scales. In addition, in the zebrafish model, the anabolic effects of cannabinoids on bone mitigated bone loss due to glucocorticoid administration (85). The antiabsorptive effects of cannabinoids in the setting of glucocorticoid-induced osteoporosis are mediated by both CB2 and TRPV1 (85).

It has been suggested that GPR55 is a novel endocannabinoid receptor. GPR55, which is encoded by the GPR55 gene, is only 13.5% identical to CB1 and 14.4% identical to CB2, and its mRNA is expressed widely, including in both mouse and human osteoclasts. Studies suggest that GPR55 is involved in osteoclast bone resorption (86). Male but not female Gpr55 knockout mice have increased bone mass and osteoclast number, but diminished osteoclast function. In vitro, CBD, which is a GPR55 antagonist, increased osteoclast number in GPR55 wild-type, Cnr1 knockout, and Cnr2 knockout mice, but not Gpr55 knockout mice. O-1602, a GPR55 agonist, increased GTP-Rho; this is blocked by CBD in both mouse and human osteoclasts. O-1602 also increased in vitro mouse, but not human, osteoclast numbers.

In a cell culture of human periodontal ligament cells, Cnr2 activity was associated with increased osteogenic gene transcription and decreased expression of RANKL (87). The differentiation of cultured human monocytes, which express CB2, into osteoclasts is regulated by alterations in CB2 receptor signaling associated with modulation of endocannabinoid synthesis (88). Decreasing CB2 expression using small interference RNA reduces mesenchymal cell differentiation into osteogenic cells. Conversely, CB2 activation increases osteogenic differentiation (89). These effects are indirect and require an accessory bone marrow cell, which communicates with osteoclast precursors (90). In vitro treatment of human osteoblasts with anandamide increases differentiation, whereas 2-AG incubation has a biphasic effect to initially produce osteoblast differentiation and then decrease markers of osteoblastic differentiation (91).

In addition to the roles of the CB1, CB2, and GPR55 endocannabinoid receptors in regulation of bone, a third endocannabinoid receptor, TRPV1, also participates in bone homeostasis. Human osteoblasts express TRPV1 receptors and the enzymatic machinery for their synthesis and degradation (92). Activation of TRPV1 in human osteoclasts in vitro by anandamide stimulates osteoclast activity and bone turnover and increases CB2 expression in osteoclasts 10-fold (93). TRPV1 is also involved in glucocorticoid-related bone loss, as in vitro exposure of human osteoclasts to methylprednisolone increases TRPV1 signaling; this results in increased osteoclast number and activity (94).

These preclinical findings demonstrate that CB2 influences skeletal homeostasis in many experimental models and raise the possibility that therapeutic targeting of skeletal CB2 may represent a novel mechanism for reducing loss of bone mass in a variety of bone disorders. Whether additional therapeutic effects on age-related bone loss can be obtained by targeting CB1 receptors in bone, as suggested by the findings in double-knockout mice, is not known.

Clinical observations of intrauterine growth retardation and decreased birth weight, length, and head circumference in babies born to mothers who used cannabis during pregnancy implicate endocannabinoids in human fetal skeletal development (95). CB2 is expressed in the fetal human brain as early as 9 weeks’ gestation (96). First-trimester human placental tissue synthesizes fatty acid amino hydrolase, an enzyme responsible for digesting endocannabinoids, and functional CB1 and CB2 receptors are found in first-trimester human syncytiotrophoblast cells (97). The administration of THC to gravid mice and rats impairs both intrauterine and femoral and vertebral growth in 5- to 11-week-old wild-type and Cnr2 knockout, but not Cnr1 knockout, mice (98). However, THC decreases in vitro epiphyseal growth plate chondrocyte hypertrophy only in Cnr1 knockout mice (64).

Another developmental pathway in bone cell differentiation and growth affected by endocannabinoids is the differentiation of mesenchymal stem cells into fibroblasts. In a cell culture model, phytocannabinoids increased fibroblast colony-forming units, alkaline phosphatase expression, and collagen synthesis. These effects are mediated by CB2 but do not involve direct action on mesenchymal stem cells (99). In addition to their role in bone growth and remodeling, cannabinoids are present in human fascia (100) and affect connective tissue elements found in intervertebral disk tissue and human cartilage (101). In a rat model of intervertebral disk degeneration, the phytocannabinoid CBD attenuated the inflammatory response to disk injury (102). This is consistent with action of CBD in reducing inflammation in inflammatory peripheral neuropathy, periodontitis, experimental arthritis, and hepatic ischemia (19). Human chondrocytes obtained from osteoarthritic joints express a range of cannabinoid receptors, including CB1 and CB2, TRPV1, GPR18, and GPR55. Receptor concentration does not correlate with amount of joint degeneration (103).

A number of human and animal studies provide evidence for the involvement of the endocannabinoid system in musculoskeletal disorders. In a mouse model of osteoarthritis, Cnr2 knockout mice exhibited increased age- and trauma-related osteoarthritis of the knee. Administration of a CB2-selective agonist blunted the extent of joint damage. Furthermore, articular chondrocytes obtained from Cnr2 knockout mice produced less proteoglycans in vitro than wild-type chondrocytes (104). Administration of a monoacylglycerol lipase (MAGL) inhibitor, the enzyme responsible for the inactivation of 2-AG, to mice with tumor-associated osteolysis decreased tumor growth, bone metastases, ectopic bone formation, and cachexia and prolonged survival. In contrast, MAGL inhibition in control animals decreased bone volume due to CB1- and CB2-mediated increased osteoclast activity (105). Ligature-induced periodontitis in rats treated with electroacupuncture increased CB2 expression in periodontal tissue and increased CB1 expression in spinal cord afferent nocioceptive pathways (105). Two studies (89, 106) have reported a qualitative decrease in human bone marrow CB2 expression in marrow biopsies obtained from patients with osteoporosis. The expression of CB1, TRPV1, triacylglycerol lipase α, and MAGL in osteocytes generated from human peripheral blood monocytes is inversely proportional to the amount of age-related bone loss (93). Synovial biopsy samples from patients with osteoarthritis or rheumatoid arthritis contain CB1 and CB2 RNA and protein (103). These are not found in normal control synovial biopsy samples. In addition, synovial biopsy tissue shows upregulation of some endocannabinoids compared with controls.

Human gene linkage studies revealed an association between Cnr2 polymorphisms and both osteoporosis and decreased bone density in some, but not all, populations (107). Genetic studies of families with decreased bone density identified 1p36, the location of the Cnr2 gene, as a site involved in the determination of bone density (108‒110). A retrospective study of recreational cannabis users in the United Kingdom showed decreased lumbar spine and hip bone mineral density, increased fracture rate, and modest elevations in type 1 C-terminal telopeptide and N-terminal type 1 procollagen as well as decreased 25-hydroxyvitamin D levels in heavy cannabis users compared with moderate users. Multivariate analysis modeling found that the decreased bone mineral density was indirect and secondary to decreased body mass index (111). In contrast to these findings, a retrospective study of self-reported cannabis use and bone mineral density in normal subjects aged 20 to 59 years obtained from 2007 to 2010 National Health and Nutrition Examination Survey (NHANES) data showed no association between self-reported cannabis use and bone mineral density (112). However, this study did confirm the finding of decreased body mass index in self-reported heavy cannabis users.

Drug targeting of the endocannabinoid system represents an attractive approach for the treatment of bone and joint diseases (93, 113). In vitro and in vivo animal studies showed that endocannabinoids, phytocannabinoids, and synthetic cannabinoids affect bone metabolism and can protect against osteoporosis due to ovariectomy and glucocorticoid administration (81) (114). In animal models of inflammatory arthritis, CBD and synthetic CBD analogues had anti-inflammatory effects (115‒117). These effects were mediated by a number of cannabinoid receptors and may also involve displacement of endocannabinoids from intracellular fatty acid‒binding proteins (118). CBD enhanced fracture healing in rats (119) and reduced bone loss in rats after thoracic spinal cord transection (120). There are no published studies on the treatment of metabolic bone disease or inflammatory arthritis in humans with cannabinoids. However, the absence of human studies is remarkable and may reflect the legislative restrictions surrounding cannabis research (121). The preclinical findings demonstrate that CB2 influences skeletal homeostasis in many experimental models and raise the possibility that therapeutic targeting of skeletal CB2 may represent a novel mechanism for reducing loss of bone mass in a variety of bone disorders. On the basis of findings in double-knockout mice, targeting CB1 in bone represents an additional direction for endocannabinoid-based drug development to prevent and treat osteopenia and osteoporosis.

As an alternative to phytocannabinoids found in Cannabis sativa and Cannabis indica, synthetic cannabinoids as well as natural products obtained from non-Cannabis plant species, which target the endocannabinoid system, are widely available and provide lead compounds for cannabinoid-based drugs for the treatment of bone loss (122, 123). Lipoamines found in olive oil, such as oleoyl serine, are in clinical trials for the treatment of metabolic bone disease (124). In addition to phytocannabinoid agonists and antagonists, synthetic agonists and inverse agonists (i.e., antagonists) of CB2 have in vitro effects on bone (78). High throughput screening has identified selective estrogen receptor modifiers as CB2 receptor reverse agonists (125). Selective estrogen receptor modifiers are FDA-approved drugs for the treatment of postmenopausal osteoporosis. Their mechanism of action is thought to result from mimicking the effects of estrogenic compounds in inhibiting osteoclast activity (126). However, the finding that this class of compounds is an inverse antagonist of CB2 suggests an alternative and additional mechanism of action that accounts for the effects of these drugs on bone resorption.

In summary, evidence from animal and human studies demonstrate a role for the endocannabinoid system in bone development, bone remodeling, and metabolic bone disease. The results of animal studies are complex, inconsistent, and dependent on the species, strain, and sex of the animal model. The applicability of results obtained in animal and in vitro studies to humans is not clear. In humans, CB2 is expressed by osteoblasts, osteoclasts, fibroblasts, and synovial tissue, and their expression is altered in arthritides (127). Retrospective human studies implicate Cnr2 gene polymorphisms and recreational cannabis use as possible risk factors for decreased bone density. Population genetic studies have localized gene polymorphisms associated with altered bone mineral density to the chromosome locus that includes Cnr2.

Metabolic bone disease represents a common medical problem and substantial public health concern. The long-term use of medical cannabis, especially among older populations, may affect the burden of age-related bone loss. Despite the increased availability of cannabis and the identification of noncannabis sources of natural products that affect the endocannabinoid system, the clinical relevance of the effects of cannabis on bone is unknown. Cannabis is safe and effective in the treatment of chronic pain, chemotherapy-related nausea and vomiting, neuromuscular complications of multiple sclerosis, sleep disorders, HIV and AIDS cachexia, Tourette syndrome, anxiety, posttraumatic stress disorders, and glaucoma (128). These findings confirm the traditional uses of cannabis in the management of inflammatory, neurologic, psychiatric, and ophthalmic diseases (129‒131). With the widespread availability and long-term use of medical cannabis (132), the effects of cannabinoids and cannabinoid congeners on bone deserves attention. Prospective studies on bone metabolism in cannabis users are needed. The efficacy of cannabinoids and cannabinoid mimetics in accelerating fracture healing and preventing and treating bone loss due to estrogen deficiency, glucocorticoid use, or aging requires clinical investigation. Studies of the physiology, pathology, and pharmacology of skeletal endocannabinoids merit the attention of endocrinologists.

Acknowledgments

Disclosure Summary: The authors have nothing to disclose.

Abbreviations:

Abbreviations:
 
• 2-AG

  2-arachidonoylglycerol

 
• CBD

  cannabadiol

 
• CNS

  central nervous system

 
• FDA

  Food and Drug Administration

 
• GPR

  G-protein receptor

 
• MAGL

  monoacylglycero lipase

 
• PPAR

  peroxisome proliferator-activated receptor

 
• RANKL

  receptor activator of nuclear factor κB ligand

 
• THC, Δ⁹-tetrahydrocannabinol

  TRPV, transient receptor potential vanilloid

References and Notes