The dynamic nature of type 1 cannabinoid receptor (CB1) gene transcription

Abstract

The type 1 cannabinoid receptor (CB₁) is an integral component of the endocannabinoid system that modulates several functions in the CNS and periphery. The majority of our knowledge of the endocannabinoid system involves ligand–receptor binding, mechanisms of signal transduction, and protein–protein interactions. In contrast, comparatively little is known about regulation of CB₁ gene expression. The levels and anatomical distribution of CB₁ mRNA and protein are developmental stage-specific and are dysregulated in several pathological conditions. Moreover, exposure to a variety of drugs, including cannabinoids themselves, alters CB₁ gene expression and mRNA levels. As such, alterations in CB₁ gene expression are likely to affect the optimal response to cannabinoid-based therapies, which are being developed to treat a growing number of conditions. Here, we will examine the regulation of CB₁ mRNA levels and the therapeutic potential inherent in manipulating expression of this gene.

Linked Articles

This article is part of a themed section on Cannabinoids. To view the other articles in this section visit http://dx.doi.org/10.1111/bph.2012.167.issue-8

Introduction

In the past decade, evidence has accumulated indicating that the endocannabinoid system (ECS) plays a critical role in the regulation of numerous biological processes including embryonic development, metabolism and neurotransmission (Mechoulam and Hanu, 2001; Howlett et al., 2002; Pertwee et al., 2010). The ECS consists of endogenously synthesized endocannabinoids [eCBs, anandamide (AEA) and 2-arachidonoylglycerol (2-AG)], their receptors (the type 1 and type 2 cannabinoid receptors) and their anabolic and catabolic enzymes (Figure 1; Matsuda et al., 1990; Munro et al., 1993; Di Marzo et al., 1994; Cravatt et al., 1996; Martin et al., 1999). In addition to eCBs, phytocannabinoids and synthetic cannabinoids act as cannabinoid receptor ligands. The type 1 cannabinoid receptor (CB₁) mediates cannabinoid-dependent signal transduction in the CNS and periphery (Howlett et al., 2002; Basavarajappa et al., 2009), while the type 2 cannabinoid receptor (CB₂) is localized to, and highly inducible in, peripheral haemopoietic cells and glial cells in specific areas of the CNS during the inflammatory response (Basavarajappa et al., 2009; Atwood et al., ). To date, the majority of CB₁ research has focused on ligand–receptor binding, signal transduction and protein–protein interactions. In contrast, knowledge of CB₁ gene regulation is limited. CB₁ receptor abundance and the function of the ECS may change in response to altered CB₁ gene expression in different developmental or disease conditions or in response to drug exposure. While other reviews have explored the general factors that regulate CB₁ and CB₂ levels during disease pathogenesis (Miller and Devi, 2011), this review will focus on regulation of CB₁ mRNA expression during development and in several pathological conditions where specific pharmacologically tractable regulators of CB₁ transcription have been elucidated. The data reviewed here demonstrate that CB₁ mRNA transcription is malleable and may be exploited for therapeutic benefit.

Figure 1.

The ECS and CB₁. The GPCR CB₁ is activated by endocannabinoid ligands such as AEA and exogenous ligands such as THC. In the CNS, activation of CB₁, which is typically coupled to G_i/o-proteins, inhibits AC and causes changes in gene expression (e.g. TH, L1 adhesion molecule, PPE and the Bcl-2/Bax regulators of apoptosis). Activation of CB₁ also causes inhibition of L-, N- and P/Q-type Ca²⁺ channels. CB₁ can couple to several other GPCRs, which influences receptor trafficking and ligand affinity. Other components of the ECS include CB₂, the putative cannabinoid receptors, GPR55 and GPR18, and the catabolic enzymes of cannabinoids fatty acid amide hydrolase (FAAH) and monoacylglycerol lipase (MAGL).

Architecture, splice variants and isoforms of the CNR1 gene

The human CB₁ gene (CNR1) spans 26.1 kb of chromosome 6 (6q14–q15). CNR1 contains four exons (Figure 2A), and the protein coding region of CB₁ is contained entirely within exon 4 (Zhang et al., 2004). Outside of the coding region, alternative splicing of CB₁ mRNA produces six 5′ untranslated region (5′-UTR) splice variants. The precise transcription start sites within exon 1 included in 5′ UTR variants 1, 3, 4 and 5 have not been defined; although it appears that multiple transcription start sites may exist within the first 60 bp of exon 1 (Shire et al., 1996). Transcription of variant 6 begins within intron 2, and thus the 5′ most exon of variant 6 has been redefined as exon 3a. Transcription of variant 2 begins at the 5′ end of exon 4. Transcript variants 1, 3, 4, 5 and 6 encode full-length CB₁ that is 472 amino acids in length encoded without interruption by a single region in exon 4. Exon 4, however, can be differentially spliced to remove 102 nts separating the 5′ end of exon 4 and a new exon identified as exon 4a. This splicing occurs in transcript variant 2 that encodes the truncated, 439 amino acid and CB_1b protein (Figure 2D). In CB_1a, different intra-exon 4 splice sites result in the loss of 167 nts. Furthermore, two translation start sites are present at the 5′ end of exon 4. Translation from the first produces CB₁ and CB_1b. Translation from the second is thought to produce the amino-terminal variant CB_1a, also known as CB_1short (Ryberg et al., 2012). The macaque monkey (Macaca mulatta) CB₁ gene is located on chromosome 4. Although the number of exons is not known, the protein coding region of the gene is contained entirely within one contiguous coding region (Figure 2B) (National Centre for Biotechnology Information (NCBI), 2011). The mouse and rat CB₁ genes are located on chromosomes 4 and 5, respectively; both genes contain 2 exons with the protein coding regions existing entirely within the second exon in both species (Figure 2C; Miller and Devi, 2011).

Figure 2.

(A) The human CB₁ gene, CNR1, spans 26.1 kb on chromosome 6. Six splice variants of the 5′ UTR have been identified by sequencing cDNA ESTs. Splice variants are illustrated in blue and numbered on the left. Exons are numbered within the blue boxes. Each splice variant is aligned with respect to its nucleotide sequence in the CNR1 gene (at top). The scale bar represents 5 kb of nucleotides. (B) The non-human primate (M. mulatta) CB₁ gene is poorly characterized, yet it is known that the entire protein coding region is contained within 1 exon (NCBI, 2011). The protein isoforms CB1a and CB1b have been described in non-human primates (Gustafsson et al., 2008). (C) The rat and mouse CB₁ genes span approximately 20 kb on chromosome 4 and contain two exons. The second exon contains the entire protein coding region. (D) Three CNR1 coding region variants for protein isoforms of CB₁ have been described in humans and non-human primates: the 472 amino acid, intron-less CB₁, the 439 amino acid CB_1b and the 411 amino acid CB_1a. In this figure, position 1 is 300 bp downstream of the 5′ end of exon 4 in CNR1. Translation of CB₁ and CB_1b begins at the same ATG codon located 309 bp downstream of the first nucleotide in exon 4. Translation of CB_1a begins 326 bp downstream of the first nucleotide in exon 4. Fifty-nine basepairs downstream of the CB_1b translation start site, CB_1b contains a 102 bp intron that is spliced from the pre-mRNA at an atypical intron–exon splice junction (s, CT/cc and ag/GG). Eighty-eight basepairs downstream of the CB_1a translation start site, CB_1a contains a 167 bp intron that is spliced from the pre-mRNA at a typical 5′ intron-exon boundary (AG/gt) and an atypical ag/GA 3′ splice junction. Downstream of the CB_1a intron–exon junction the coding sequences of the three CB₁ isoforms are identical.

To date, CB_1a and CB_1b isoforms have only been identified in humans and higher primates (Ryberg et al., 2012; Gustafsson et al., 2008; Palermo et al., 2009), and some evidence suggests CB_1a may be expressed in the rat (Shire et al., 1996). Several authors have demonstrated that CB₁, CB_1a and CB_1b receptors signal via G_i/o-type G-proteins and that the amino-terminal variants CB_1a and CB_1b have reduced affinity for cannabinoid agonists and antagonists (Rinaldi-Carmona et al., 1996; Ryberg et al., 2012). However, Xiao et al. (2008) did not observe differences in the ligand affinity or localization of the three CB₁ protein isoforms. Moreover, the signalling properties of CB₁ receptor variants may be altered depending on the model system they are being studied in (Straiker et al., 2012), which complicates our ability to understand receptor differences. In the majority of reports, steady-state CB₁ mRNA levels were measured via amplification of the 3′ end of the CB₁ coding region outside of the 5′ region in exon 4 involved in differential splicing. The cell-specific relative abundance of CB₁ versus CB_1a or CB_1b is, therefore, poorly characterized (Gustafsson et al., 2008). Early research suggested that CB_1a mRNA accounted for approximately 20% of the CB₁ transcript population (Shire et al., 1996), yet more recent evidence suggests that less than 5% of the total population of CB₁ transcripts obtained from human fetal and adult brain tissue are CB_1a or CB_1b (Xiao et al., 2008). Studies to define the relative abundance and distribution of the 5′ UTR variants 1–6 have measured the levels of expressed sequence tags. The 5′ UTR transcript variants 1 (5732 bp), 3 (5863 bp), 4 (5901 bp) and 5 (5776 bp) are most abundant in the brain, lymphocytes, testes and liver, relative to other tissues (NCBI, 2011). Transcript variant 2 (5387 bp mRNA) is expressed at highest levels in the brain and testes (NCBI, 2011). Transcript variant 6 (8974 bp mRNA) has only been isolated from brain tissue (NCBI, 2011). Regulation of the transcription of 5′ UTR variants and how 5′ UTR differences relate to CB₁ mRNA stability and translation to different CB₁ isoforms has not been characterized. The abundance and activity of the different amino-terminal CB₁ isoforms may be regulated by different physiological conditions, isoform-specific ligand–receptor affinity and the CB₁ isoform complement expressed in a given cell type (Ryberg et al., 2012).

The ‘when and where’ of CB₁ mRNA expression

In mammals, steady-state levels of CB₁ mRNA vary in different tissues and during different developmental periods. In humans, CB₁ is detected in neocortical progenitor cells and in the subventricular zone during the early cortical plate stages of development (9 to 17 weeks gestation; Zurolo et al., 2010). CB₁ mRNA is also abundant at 19 weeks gestation in humans in white matter, which is nearly devoid of CB₁ expression in adulthood. In the human visual cortex, CB₁ mRNA levels rise during early development and plateau approximately 1 year after birth (Romero et al., 1997; Pinto et al., 2010). Following the steady-state CB₁ mRNA plateau achieved 1 year after birth, CB₁ mRNA levels increase further in the visual cortex to reach a new steady-state level during adolescence, after which CB₁ mRNA abundance declines throughout adulthood (Pinto et al., 2010). In the non-human primate, M. mulatta, high levels of CB₁ mRNA have been observed in the prefrontal cortex during neonatal development (Eggan et al., 2010); CB₁ mRNA abundance increases in the prefrontal cortex until reaching a steady-state at postnatal day 5 (Eggan et al., 2010). In the same manner as is observed in the human visual cortex, a higher steady-state level of CB₁ mRNA is observed in the M. mulatta prefrontal cortex during adolescence, and steady-state CB₁ mRNA levels decline in the prefrontal cortex following adolescence (Eggan et al., 2010). In mice, CB₁ mRNA is detectable during embryonic development as early as four-cell and eight-cell/morula stages (Paria et al., 1995), and can still be detected at embryonic day 12 in glutamatergic neurons of the cerebral cortex and hippocampus (Vitalis et al., 2008). CB₁ mRNA is abundant in the adult mouse thalamus, amygdala, dorso-lateral prefrontal cortex, hypothalamus and pituitary (NCBI, 2011). Furthermore, CB₁ expression is enriched in the striatum, relative to other brain regions, within the adult mouse CNS (Fernandez-Ruiz et al., 2012; McCaw et al., 2004). It is within the striatum that a high steady-state level of CB₁ expression are dysregulated in Parkinson's and Huntington's diseases (Zeng et al., 1999; Denovan-Wright and Robertson, 2000). The temporal and anatomical distribution of CB₁ expression during early development is similar in mice and rats (NCBI, 2011). Six to 8-week-old rats, which are sexually mature, have lower levels of CB₁ mRNA in the limbic/associative brain areas compared with adolescents (Heng et al., 2011). Following periods of peak neurodevelopment associated with high CB₁ levels, CB₁ mRNA abundance declines in these brain regions (Heng et al., 2011). Taken together, these data demonstrate that, in mammals, CB₁ mRNA levels peak during adolescence within the prefrontal cortex, limbic/associative areas and visual cortex and subsequently with age. Early development and adolescence represent critical developmental windows where the regulation of CB₁ expression changes in order for higher levels of expression to be achieved. It is likely that developmental stage-specific transcription factors or modifiers regulate the different steady states of CB₁ expression. A representative illustration of the temporal–spatial expression of CB₁ mRNA, in the CNS, based on data obtained from mouse, rat, monkey and human is presented in Figure 3.

Figure 3.

CB₁ mRNA abundance and distribution shift throughout development in the CNS. This simplified schematic illustrates the areas of the brain where CB₁ mRNA levels are moderate and high relative to other regions of the CNS during mid – late embryonic development, adolescence and adulthood. The images were created using data obtained in humans, non-human primates, mice and rats. cb, cerebellum; cc, cerebral cortex; hp, hippocampus; hyp, hypothalamus; ls, limbic system; mc, motor cortex; pfc, prefrontal cortex; svz, sub-ventricular zone; th, thalamus; str, striatum; vc, visual cortex; wm, white matter (Denovan-Wright and Robertson, 2000; Vitalis et al., 2008; Eggan et al., 2010; Gensat, 2010; Zurolo et al., 2010; Heng et al., 2011; NCBI, 2011).

High levels of CB₁ expression are related to the establishment of neuronal circuitry. During critical development periods, such as late gestation and early postnatal life, areas associated with neurogenesis and synapse formation, such as the subventricular zone and white matter, are transiently enriched for CB₁ and subsequently depleted of CB₁ expression in adulthood (Romero et al., 1997). The activity or abundance of the factors that enabled high steady-state CB₁ levels during development and adolescence may decrease in concentration or activity as part of the aging process (Eggan et al., 2010; Heng et al., 2011). Greater expression and subsequent activation of CB₁ receptors facilitates higher expression of several genes required for brain development, including tyrosine hydroxylase (TH), preproenkephalin (PPE), the neural adhesion molecule L1 and Bcl-2/Bax genes involved in apoptotic regulation of development (reviewed in Fernandez-Ruiz et al., 2012). Mice lacking CB₁ exhibit transcriptional dysregulation of PPE and substance P (Steiner et al., 1999), altered dendritic morphology and lower synapse density in the prefrontal cortex (Fitzgerald et al., 2012), impaired locomotor activity (Zimmer et al., 1999) and increased anxiety (Hill et al., 2011) compared to wild-type littermates. Thus, the developmental stage-specific expression of CB₁ facilitates the proper establishment of neuronal circuitry and the consequent normalization of behaviour (Fernandez-Ruiz et al., 2012).

Expression of CB₁ is cell-specific within the CNS. Striatal medium spiny projection neurons and interneurons are enriched for CB₁ mRNA expression, relative to other cell populations, within the basal ganglia (Marsicano and Lutz, 1999; Fernandez-Ruiz et al., 2012). Consistent with CNS anatomical distribution, CB₁ appears to be involved with aspects of motor coordination, mechanisms of reward and motivation, emotion and central endocrine regulation during adulthood (Fernandez-Ruiz et al., 2012). CB₁ is co-localized, co-expressed and can dimerize with the pre-synaptic type 2 dopamine short (D₂Sh), μ-and δ-opioid, adenosine A_2a, and orexin, receptors (Navarro et al., 2008; Pacheco et al., 2009; Uriguen et al., 2009; Bortolato et al., 2010; Rozenfeld et al., ). Thus, in addition to alterations in CB₁-mediated signalling, changing CB₁ mRNA and protein abundance could cause dysregulated signalling via other GPCRs, such as D₂Sh, opioid and orexin receptors; by affecting dimer formation, receptor trafficking and localization, and signal transduction (Hudson et al., 2010; reviewed in Smith et al., 2010).

In non-neuronal tissue, CB₁ expression is associated with immune and endocrine homeostasis as well as reproductive system development and maturation. CB₁ mRNA is abundant in helper (CD4) and cytotoxic (CD8) T cells, hepatocytes, beta-islet cells and adipose tissue, during adulthood where these receptors regulate inflammatory and metabolic processes through autocrine and paracrine signalling mechanisms (Borner et al., 2008; Mukhopadhyay et al., 2010). CB₁ mRNA levels increase in primary cultured rat leydig cells (testosterone-producing testis cells) from postnatal day 14 onwards, spermatids from postnatal days 31 through 61 and sertoli cells from postnatal day 41 onwards (Cacciola et al., 2008). Cells within the testes exhibit a biphasic pattern of CB₁ expression, in which CB₁ levels are elevated at 1 week post partum, decline to a minimum by 2 weeks post partum and rise again to highest levels as mice reach sexual maturity at 4 weeks post partum (Cacciola et al., 2008). Certain fish species, for example the gilthead seabream (Sparus aurata) and the puffer fish (Fugu rubripes), are capable of undergoing sexual reversal, which is the process of shifting the reproductive organs from being functionally male to functionally female, or vice versa. Sexual reversal, therefore, represents a period of altered gene expression leading to changes in cellular phenotype within teleost reproductive organs. CB₁ mRNA expression increases in the testes of these fish during the process of sexual reversal (Cottone et al., 2008). Increased CB₁ expression could contribute to sexual reversal by altering the complement of genes expressed and thus the phenotype of the testes (Cottone et al., 2008). In the developing CNS, changes in CB₁ expression facilitate downstream changes in the expression of many other genes (Fernandez-Ruiz et al., 2012). Therefore, up-regulation of CB₁ expression may also sufficiently alter the gene expression profile in reproductive tissue to produce changes in phenotype and facilitate developmental processes. In reproductive systems, as in the CNS, CB₁ expression appears to be coordinated in order to facilitate development and maturation during early development and adolescence.

Understanding changes in CB₁ mRNA expression that occur in diverse pathological conditions

CB₁ mRNA expression is induced by inflammation in non-neuronal tissue

Although CB₂ receptors are considered the major eCB receptor in the periphery, particularly as regulators of inflammation (Rajesh et al., 2008; reviewed in Atwood and Mackie, 2010), CB₁ receptors also contribute to regulation of the inflammatory response. Pro-inflammatory molecules induce CB₁ and CB₂ mRNA expression in cells that mediate the inflammatory responses (Gutierrez et al., 2006; Borner et al., 2008). The involvement of CB₁ in the inflammatory response was first examined in rat dorsal root ganglia (DRG), where complete Freund's adjuvant increased CB₁ mRNA abundance in glial cells of the DRG 4 h post-treatment, relative to untreated controls (Amaya et al., 2006). Freund's adjuvant produces an inflammatory response and activates such transcription factors as nuclear factor of activated T cells (NFAT) and NF-κB in glial cells (Amaya et al., 2006; Borner et al., 2007a). Activation of NFAT and NF-κB is dependent on the endogenous pro-inflammatory cytokines CD3/28 and IL-4 (Borner et al., 2007a). CD3/28 and IL-4 induce CB₁ mRNA expression in human peripheral T cells and immortalized Jurkat cells (Borner et al., 2007a; 2008). Borner et al. (2007a) examined CD3/28- or IL-4-mediated induction of CB₁ via a promoter–reporter plasmid in which chloramphenicol acetyl transferase activity was driven by a 3 kb fragment of the CNR1 promoter. Short, double-stranded, decoy oligonucleotides containing the consensus sequences normally bound by NFAT or NF-κB were used to titrate NFAT or NF-κB enhancers of transcription away from their endogenous promoters (Borner et al., 2007a). NFAT and NF-κB facilitate a CD3/28- or IL-4-dependent increase in CB₁ expression (Borner et al., 2007a). Using the same techniques, it was found that activator protein 1 (AP-1) and the signal transducers and activators of transcription 5 and 6 (STAT5 and STAT6) are also recruited to the CNR1 promoter to mediate increased mRNA expression in Jurkat cells (Borner et al., 2007a,2007b; 2008). Together, these data demonstrate that pro-inflammatory cues mediate an increase in CB₁ mRNA levels from an initial steady-state to a second, higher state through common mechanisms.

CB₁ mRNA expression is changed in various cancers

The ECS regulates cell fate and division during oncogenesis (Malfitano et al., 2012). For example, in breast cancer, CB₂ agonism inhibits cell cycle progression via down-regulation of Cdc2 (Caffarel et al., 2006), whereas in fibrosarcoma cells, CB₁ antagonism up-regulates the cell cycle inhibitor p21^WAF1 and down-regulates cyclins E and D (Malfitano et al., 2012). Up- and down-regulation of CB₁ levels influences cell growth, just as cannabinoid treatment affects cell growth. Treatment of transformed cells with anti-neoplastic agents has been shown to both increase and decrease CB₁ mRNA expression depending on drug and cell line (Larrinaga et al., 2010b; Proto et al., 2011). CB₁ mRNA levels are reduced in the DLD-1 and SW620 cell culture models of colorectal cancer, relative to primary colorectal cell cultures (Proto et al., 2011). Furthermore, the CB₁ promoter is highly methylated in human colorectal carcinoma cells, compared with healthy tissue (Wang et al., 2008), suggesting that elevated CB₁ expression consequently alters the capacity for cell division in colorectal carcinoma cells. This may be the result of CB₁-mediated up-regulation of p21^WAF1 or down-regulation of Cdc2 and cyclins, or both (Caffarel et al., 2006; Malfitano et al., 2012). Treatment of DLD-1 and SW620 cells with 17β-estradiol or the synthetic cannabinoid meth-anandamide (mAEA) increases CB₁ mRNA levels (Proto et al., 2011). Following treatment with 17β-estradiol or mAEA, the rate of division of these colorectal cancer cells is significantly reduced (Proto et al., 2011). CB₁ mRNA levels are lower in primary adrenocarcinoma tumour cells than in healthy tissue (Larrinaga et al., 2010b). In healthy tissue, CB₁ activation decreases cell division and proliferation (Larrinaga et al., 2010a,b). The chemotherapeutic agent gemcitabine arrests adrenocarcinoma tumour growth (Larrinaga et al., 2010a). Gemcitabine has also been reported to induce CB₁ mRNA expression via NF-κB, as demonstrated by chromatin immunoprecipitation of the CNR1 promoter (Larrinaga et al., 2010a). These data demonstrate that CB₁ mRNA expression can be manipulated pharmacologically, and that the level of CB₁ expression and activity negatively correlates with cell division. Consequently, pharmacological manipulation of CB₁ expression may represent a therapeutic option for the treatment of adrenocarcinomas (Larrinaga et al., 2010a,b).

In contrast to observations of decreased CB₁ mRNA levels in adrenocarcinoma, CB₁ mRNA abundance is increased in biopsied human tissue taken from patients with prostate cancer or benign prostate hyperplasia relative to healthy tissue (reviewed in Gustafsson et al., 2008). Similarly, CB₁ mRNA levels are increased in non-Hodgkin lymphoma tissues, relative to healthy tissues (Gustafsson et al., 2008). Despite the increase in CB₁ mRNA levels, Gustafsson et al. (2008) found that the relative proportions of CB_1a and CB_1b mRNA levels were lower in biopsied lymphoma tissue compared with normal lymphocytes (Gustafsson et al., 2008). If CB₁, CB_1a and CB_1b are not equally abundant, then this suggests that expression of each isoform is differentially regulated (Gustafsson et al., 2008), which is likely the result of differences in mRNA processing and stability (Shire et al., 1996; Ryberg et al., 2012). Conversely, the hypothesis that each CB₁ isoform is regulated by different promoter elements and upstream differences in cell signalling (Gustafsson et al., 2008) is unlikely because all three isoforms share a common, highly active promoter (Borner et al., 2007a,2007b; 2008; reviewed in Miller and Devi, 2011). CB₁ levels are also elevated in alveolar rhabdosarcoma, which is caused by expression of a chimeric PAX3/7-FOX01 transcription factor (Marshall et al., 2011). Up-regulation of CB₁ in alveolar rhabdosarcoma does not affect the rate or capacity of cells to divide but rather increases the metastatic ability of the cells (Marshall et al., 2011). Consequently, elevated CB₁ expression in certain cancers, such as prostate cancer, lymphoma or rhabdosarcoma may impact metastasis, but not cell division.

CB₁ mRNA abundance fluctuates in obesity and diabetes

The hormone 17β-estradiol and mediators of the inflammatory response, such as CD3/28, increase CB₁ mRNA levels. It is not surprising, therefore, that obesity and diabetes – two pathologies associated with hormonal dysregulation and inflammation – are also associated with changes in CB₁ mRNA levels (Howlett et al., 2002; Kempf et al., 2007). Whether obesity correlates with higher or lower CB₁ mRNA levels remains controversial. CB₁ mRNA abundance has been measured in primary cultured adipocytes of lean and obese individuals. In one study, CB₁ mRNA was shown to be less abundant in primary cultured adipocytes derived from white adipose tissue of obese children compared with that from lean children (Karvela et al., 2010). Similarly, CB₁ levels were lower in adipocytes derived from the visceral adipose tissue of obese adults compared with healthy adults (Kempf et al., 2007). In another study, CB₁ mRNA was more abundant in adipocytes derived from the visceral adipose tissue of obese adults, relative to non-obese individuals (Sarzani et al., 2009). Different cell culture conditions may account for the discrepancies in CB₁ mRNA levels reported by these groups. Karvela et al. (2010) and Kempf et al. (2007) cultured adipocytes in the presence of adiponectin, while Sarzani et al. (2009) measured CB₁ mRNA abundance in tissue samples without culturing the adipocytes. Therefore, the primary culturing and treatment of adipocytes may alter CB₁ expression. However, the functional consequence of altered CB₁ levels in adipocytes may be altered by cell survival and proliferation because CB₁ activation is often associated with pro-survival signalling (Kempf et al., 2007; Sarzani et al., 2009; Karvela et al., 2010). In obese individuals, enhanced survival of adipocytes may exacerbate their condition (Sarzani et al., 2009).

Diabetes is associated with a decreased production of, or response to, insulin. Insulin can penetrate the blood–brain barrier, act on insulin receptors and stimulate glucose uptake in the central nervous system (Bingham et al., 2002). Streptozotocin-treated rats lack insulin-producing beta-islet cells and are used to model diabetes (Zhang et al., 2007). CB₁ levels are increased in the striatum and hypothalamus of streptozotocin-treated rats compared to untreated controls (Diaz-Asensio et al., 2008). In the rat pancreas, β-, α, and δ-islet cells express CB₁ mRNA (Zhang et al., 2007). Treatment of rats with glucose (20–50 mM) is associated with an increase in plasma insulin concentration and a decrease in CB₁ mRNA levels in the pancreas, white adipose tissue and DRG, relative to untreated rats (Zhang et al., 2007). Thus, increased glucose, leading to increased plasma insulin is associated with CB₁ down-regulation. Therefore, insulin appears to inhibit CB₁ expression in the CNS. Conversely, up- or down-regulation of CB₁ may lead to altered insulin receptor expression (Zhang et al., 2007), glucose uptake (Diaz-Asensio et al., 2008) or both.

Susceptibility to schizophrenia is associated with changes in CB₁ mRNA expression

During adolescence, CB₁ mRNA and eCB levels peak in the dorso-lateral prefrontal cortex in mammals (Eggan et al., 2010). Mice exposed to Δ⁹-tetrahydrocannabinol (THC) during adolescence (postnatal day 40) are more likely to develop schizophrenia modelling behaviours, namely pre-attentional sensorimotor and executive function deficits, than mice exposed to THC later in adulthood (Heng et al., 2011). Eggan et al. (2010) also found that CB₁ mRNA abundance peaks in the dorso-lateral prefrontal cortex of macaque monkeys 2 months after puberty. These data suggest that adolescence may be a period of hypersensitivity to cannabinoids (Caspi et al., 2005). Polymorphisms that relate cannabis use to the CNR1 gene have been described. One study reported that a TAG allele in the 5′ region of CNR1 exon 3 and a polymorphic AAT repeat in the 3′ region of exon 4 were more prevalent among European, African American and Japanese substance abusers than individuals that did not have a history of substance abuse from the same region (Zhang et al., 2004). Intriguingly, the presence of the ‘TAG’ allele was associated with less CB₁ mRNA compared to other alleles (Zhang et al., 2004). Subsequent analyses support that the length of the polymorphic AAT repeat region may be correlated with substance abuse (Benyamina et al., 2011). Therefore, a strong connection exists between the allelic variability of CNR1 and adolescent exposure cannabis use.

CB₁ mRNA abundance differs in individuals with schizophrenia compared with healthy controls. CB₁ mRNA levels were higher in postmortem tissue from the dorso-lateral prefrontal cortex of individuals with schizophrenia compared with age-matched healthy subjects (Uriguen et al., 2009). Immunohistochemical analyses of postmortem brains isolated from individuals with schizophrenia revealed that individuals treated with the atypical antipsychotics olanzapine or clozapine expressed significantly less CB₁ protein in the dorso-lateral prefrontal cortex than age-matched individuals with schizophrenia who did not receive atypical antipsychotics (Uriguen et al., 2009). Olanzapine and clozapine treatment were not associated with a change in CB₁ mRNA levels (Uriguen et al., 2009). Therefore, these atypical antipsychotics that are, among other activities, D₂ receptor antagonists, decrease CB₁ protein, but not mRNA, levels. The functional implication of this observation is that pharmacological manipulation of the dopaminergic system can impact the cannabinergic system and possibly vice versa (El Khoury et al., 2012). Thus, therapeutics aimed at reducing cannabinergic or dopaminergic tone, such as antipsychotics, may impact both systems simultaneously. Conversely, certain compounds may increase the tone of both systems, which could have undesirable effects on cognition, behaviour and motor control if the ECS is over-activated and beneficial effects on cognition, behaviour and motor control in pathologies where CB₁ levels are reduced.

A striatal cell-specific decrease in CB₁ mRNA is observed in Parkinson's disease

CB₁ mRNA is highly expressed in the caudate and putamen, globus pallidus and substantia nigra of healthy individuals (Fernandez-Ruiz et al., 2012). Hurley et al. (2003) examined CB₁ mRNA levels in postmortem tissue from normal controls and individuals with Parkinson's disease. CB₁ levels were reduced in the caudate nucleus, anterior dorsal putamen and external segment of the globus pallidus, relative to controls or other brain regions of Parkinson's patients (Hurley et al., 2003). A key feature of Parkinson's disease is decreased dopamine levels. Parkinson's disease can be modelled by selective lesioning of the nigrostriatal pathway by 6-hydroxydopamine or administration of 6-hydroxydopamine to the medial forebrain bundle causes cell loss in the substantia nigra, which in turn depletes the striatum of dopamine (Zeng et al., 1999). Consequently, CB₁ mRNA levels are reduced within the dopamine-depleted rat striatum (Zeng et al., 1999). CB₁ mRNA expression can be increased in the striatum of 6-hydroxydopamine-lesioned rats by subsequent, chronic, treatment with L-DOPA, which increases dopaminergic signalling (Zeng et al., 1999; Garcia-Arencibia et al., 2009). Reserpine depletes catecholamines, including dopamine, in the CNS (Thrash et al., 2009). Reserpine-treated rats have been used to model Parkinson's disease and depression (Silverdale et al., 2001; Thrash et al., 2009). Similarly to 6-hydroxydopamine-lesioned rats, CB₁ mRNA levels were reduced in the caudate and putamen of reserpine-treated rats relative to age-matched, control rats (Silverdale et al., 2001). Overall, CB₁ mRNA levels appear to be regulated by dopamine in Parkinson's disease, which provides more evidence for a link between cannabinergic and dopaminergic signalling.

A cell-specific decrease in CB₁ mRNA is observed in Huntington's disease and contributes to disease pathogenesis

CB₁ mRNA levels are reduced in the caudate and putamen of human subjects with Huntington's disease and the striatum of all Huntington's disease mouse models tested to date relative to age-matched controls (Pazos et al., 2008). Expression and nuclear localization of the amino-terminus of mutant huntingtin protein reduces transcription of CB₁ in striatal medium spiny projection neurons early in disease progression (Gafni and Ellerby, 2002; McCaw et al., 2004). CB₁ expression is not altered in the presence of mutant huntingtin protein in the hippocampus, or prior to adulthood in mice (Denovan-Wright and Robertson, 2000; McCaw et al., 2004). Also, CB₁ mRNA levels are lower in cultured neuronal cell models of Huntington's disease, which lack inter-cellular signalling, compared to cells that do not express mutant huntingtin (Blázquez et al., 2011). Wild-type huntingtin binds to the repressive element 1 silencing transcription factor (REST), which can interact with repressor elements at the CB₁ promoter (Blázquez et al., 2011). Blázquez et al. (2011) utilized a CB₁ promoter–reporter and decoy oligonucleotide-based assay to demonstrate that REST inhibits transcription of CB₁ in cells expressing mutant huntingtin. Therefore, tissue-, cell- and developmental stage-specific factors that normally accommodate high-level CB₁ mRNA expression in the adult striatum are affected by the cell-autonomous overexpression of amino-terminal mutant huntingtin. In contrast, CB₁ mRNA expression does not appear to change in Alzheimer's disease, a neurodegenerative disease, like Huntington's and Parkinson's diseases, characterized by protein misfolding and aggregation (Kalifa et al., 2011). While CB₁ protein levels (Kalifa et al., 2011) and receptor binding (Westlake et al., 1994) may be decreased, in the hippocampus, neocortex and basal ganglia during Alzheimer's disease progression (Westlake et al., 1994; Lee et al., 2010), mRNA abundance is unaffected in humans and mouse models.

Decreased CB₁ receptor function may contribute to progressive decline in Huntington's disease. Separate research groups bred two different mouse models of Huntington's disease with homozygous CB₁ knock-out mice (CB₁^−/−; Blázquez et al., 2011; Mievis et al., 2011). Both research groups found that mice over-expressing amino-terminal mutant huntingtin and having reduced CB₁ exhibited an earlier Huntington's disease symptom onset, a more rapid disease progression and a greater degree of medium spiny projection neuron degeneration than wild-type mice or mice over-expressing amino-terminal mutant huntingtin with a full complement of CB₁ (Blázquez et al., 2011; Mievis et al., 2011). Their findings suggest CB₁ normally performs a neuroprotective role in the striatum and loss of this receptor correlates with Huntington's disease pathogenesis. Thus, therapeutic strategies capable of elevating CB₁ mRNA abundance may restore or enhance the neuroprotective role of CB₁ where decreased expression of this receptor may contribute to disease pathology.

Pharmacological manipulation of CB₁ mRNA abundance

CB₁ mRNA levels can be modulated by methamphetamine and alcohols

Methamphetamine use is associated with region-specific changes in CB₁ mRNA levels in the CNS. Acute treatment of rats with methamphetamine is associated with increases in steady-state CB₁ mRNA levels in the prefrontal cortex, caudate and putamen, basolateral amygdala, CA1 hippocampal region and perirhinal cortex, relative to other brain regions and untreated controls (Bortolato et al., 2010). These region-specific increases are detectable up to 3 weeks after a single post-acute treatment (Bortolato et al., 2010). Acute methamphetamine use is associated with increased dopamine neurotransmission. These findings align with findings in schizophrenia and Parkinson's disease that suggest CB₁ levels are influenced by dopamine acting on pre- and post-synaptic receptors.

Ethanol use is also associated with region-specific changes in CB₁ mRNA levels in the CNS and in cell culture. In humans, CB₁ protein levels have been compared in the ventral striatum of individuals that were alcohol-dependent to age-matched, non-alcoholic individuals (Vinod et al., 2005). CB₁ protein levels were lower in alcohol-dependent individuals relative to non-alcoholic controls (Vinod et al., 2005). Barbier et al. (2012) found that mice exposed to ethanol in utero have significantly lower levels of CB₁ mRNA in the cortex, striatum and hippocampus from postnatal days 14 through 90 relative to age-matched controls. Consequently, exogenous ethanol exposure appears to alter CB₁ expression during early development and adulthood and may lead to chronic alterations in neurotransmission and gene expression that are normally facilitated by CB₁.

CB₁ mRNA levels can be modulated by estradiol and retinoic acid

Estradiol and retinoic acid alter CB₁ mRNA levels. 17β-estradiol and retinoic acid act upon their respective nuclear receptors to increase CB₁ mRNA levels in cell culture (Mukhopadhyay et al., 2010; Proto et al., 2011). Administration of 17β-estradiol increases CB₁ mRNA levels in DLD-1 and SW620 colon cancer cells (Nortarnicola et al., 2008; Proto et al., 2011). The effect of 17β-estradiol requires the oestrogen receptor, retinoic acid receptor (RAR)α and PPARγ to co-localize within the promoter region 1 kb from the human CNR1 transcription start site (Proto et al., 2011). Retinoic acid also increases CB₁ mRNA levels in mouse primary hepatic stellate cells (Mukhopadhyay et al., 2010). This induction requires retinaldehyde dehydrogenase and RARγ (Mukhopadhyay et al., 2010). RARγ interacts with a retinoic acid response element approximately 350 bp upstream of the mouse Cnr1 transcription start site. Together, these data demonstrate 17β-estradiol and retinoic acid induce CB₁ transcription via oestrogen receptor, RAR- and PPAR-dependent mechanisms from a basal state to a higher steady state. Therefore, CB₁ levels can be manipulated by the activation or inhibition of well-known, pharmacologically tractable regulators of transcription.

CB₁ mRNA level is modulated by cannabinoids

Cannabinoids modulate steady-state CB₁ mRNA abundance. Chronic treatment with THC has been shown to decrease CB₁ mRNA levels in the CNS of rodents. Repeated exposure to THC, once daily for 14 days by i.p. injection, decreases CB₁ mRNA levels in the caudate and putamen of adult male rats (Corchero et al., 1999). The extent of CB₁ mRNA decrease correlates to the number of repeated exposures. In another study, THC treatment increased CB₁ mRNA levels in the rat cerebellum and hippocampus over a 3 day period, while simultaneously decreasing CB₁ mRNA levels in the rat striatum (Zhuang et al., 1998). Cannabinoids have also been shown to increase CB₁ mRNA levels in primary and immortalized cell culture systems. Treatment of primary mouse hepatic, stellate cells with 2-AG induces CB₁ mRNA, up to 30-fold relative to basal expression in untreated cells (Mukhopadhyay et al., 2010). 2-AG-mediated CB₁ induction is RARγ- and CB₁ receptor-dependent in this model system (Mukhopadhyay et al., 2010). AEA has been reported to increase CB₁ mRNA levels in DLD-1 and SW620 cells (Proto et al., 2011). This effect was oestrogen receptor- and RARα-dependent. Finally, THC, methanandamide and the CB₂-selective agonist JWH-015 induce CB₁ mRNA expression in Jurkat cells in a CB₂-dependent manner (Borner et al., 2007b). Borner et al. (2007b) observed that CB₂ activation leads to phosphorylation of STAT5, transactivation of IL-4 and activation of STAT6, thereby inducing CB₁ promoter activity (Borner et al., 2007b). Thus, in some systems, cannabinoid-dependent activation of CB₁ and CB₂ receptors stimulates the activity of specific transcription factors, such as the oestrogen receptor, RARα and STAT6, and augments steady-state CB₁ mRNA levels above basal levels. In other systems, cannabinoid exposure down-regulates CB₁ mRNA levels (Corchero et al., 1999). Cannabinoid treatment therefore, as in various pathological conditions, is associated with malleable context-specific regulation of CB₁ expression. In vivo, repeated exposure to cannabinoid agonists is associated with tachyphylaxis (Corchero et al., 1999), whereas in cell culture, single acute doses of cannabinoid agonists induce CB₁ mRNA expression (Borner et al., 2007b; Mukhopadhyay et al., 2010). From these observations, it is clear that the response of the CB₁ mRNA levels to cannabinoid treatment depends on the nature of treatment, chronic versus acute as well as the potency and efficacy of the ligand. For example, CB1 mRNA expression may be inducible in in vivo studies examining acute doses of cannabinoids, indirect cannabinoid agonism via fatty acid amide hydrolase inhibitors (Kim and Alger, 2010), or allosteric modulation of CB₁ receptor activity (Navarro et al., 2009; Ahn et al., 2012).

CB₁ protein levels are also increased following acute cannabinoid-dependent induction of CB₁ mRNA levels (Mukhopadhyay et al., 2010; Proto et al., 2011). This increase is modest (four- to fivefold) compared with the increased mRNA expression observed (29- to 30-fold, Mukhopadhyay et al., 2010; Proto et al., 2011) yet represents an increase in the pool of CB₁ receptors. In these studies, CB₁ protein abundance was quantified via western blot. Therefore, it is not known whether cannabinoid-mediated CB₁ induction affects the localization or functionality of CB₁ receptors.

Conclusions

Manipulation of CB₁ expression may have wide ranging affects on physiological processes such as embryogenesis (Paria et al., 1995) and neural development (Fitzgerald et al., 2012). Given that CB₁ gene expression is highest in many regions of the brain during early development and adolescence, and it is these areas of high expression where expression is often altered during disease progression, the effect of CB₁ therapeutics on CB₁ expression will depend on the existing level of CB₁ expression. CB₁ levels can be modulated pharmacologically by pro-inflammatory peptides, oestrogen, insulin, atypical antipsychotics, methamphetamine, ethanol, retinoic acid and, importantly, endogenous and exogenous cannabinoids (Figure 4). The observation that cannabinoids can induce CB₁ mRNA expression in cell culture suggests that cannabinoid ligands could regulate CB₁ levels. Although long-term treatment with THC in vivo is associated with tachyphylaxis (Corchero et al., 1999), other cannabinoids have not been examined; the extent of tachyphylaxis may vary in a ligand-specific manner (Hudson et al., 2010), as well as with the duration of exposure (Zhuang et al., 1998), half-life, efficacy, and potency of the cannabinoid. For instance, acute doses of less-potent agonists of CB₁ (e.g. AEA relative to THC; Pertwee et al., 2010), or allosteric modulators of CB₁ (Ahn et al., 2012) may induce CB₁ expression in vivo, whereas chronic treatment with more-potent agonists may cause receptor desensitization. If this is the case, then cannabinoid-dependent manipulation of CB₁ levels may represent a useful therapeutic strategy for diseases where reduced or elevated CB₁ levels correlate with disease progression. Overall, the affect of cannabinoid-based therapies may depend upon their modulation of CB₁ gene expression.

Figure 4.

CB₁ mRNA expression is regulated by several co-activators of transcription. This schematic illustrates regulation of CB1 transcription at several functional transcription factor binding sites known to enhance CB₁ expression within 3 kb of the CNR1 transcription start site at exon 1. STAT5, STAT6 (approximately −2769 bp, Borner et al., 2007a,b), NFAT, NF-κB, AP-1 (within −2490 bp, Borner et al., 2008), REST (approximately −898 bp, Blázquez et al., 2011), RARα/γ (−350 bp, McCaw et al., 2004; Mukhopadhyay et al., 2010; Proto et al., 2011), ERE (−1073 bp and −366 bp, Proto et al., 2011), TAG and AAT polymorphisms 5′ of exon 1 and 3′ of exon 4, respectively (Zhang et al., 2004), Me, DNA methylation leading to promoter repression (Miller and Devi, 2011).

Glossary

2-AG

2-arachidonyl glycerol

AEA

anandamide

AP-1

activator protein 1

CB₁

type 1 cannabinoid receptor

CB₂

type 2 cannabinoid receptor

CNR1

type 1 cannabinoid receptor gene

D₂Sh

type 2 dopamine receptor – short variant

DRG

dorsal root ganglion

eCBs

endocannabinoids

ECS

endocannabinoid system

mAEA

meth-anadamide

NFAT

nuclear factor of activated T cells

PPE

preproenkephalin

RAR

retinoic acid receptor

REST

repressive element 1 silencing transcription factor

STAT

signal transducers and activators of transcription

TH

tyrosine hydroxylase

THC

Δ⁹-tetrahydrocannabinol

UTR

untranslated region

Conflict of interest

None declared.