Rapid advances in designing genetically engineered laboratory animals are producing not only research models but also models that are more effective research tools. As a result, the need for precision genetic characterization and denition of laboratory animals is of primary concern.
However, the use of transgenic mouse models for the over-expression and other forms of modi-cation of CBR genes (except for the CBR gene inactivation described above) to study the regulation and site-specic mechanisms of the action of cannabinoids is currently being explored. For example, cannabinoid transgenes, in which genomic regulatory sequences of interest are coupled to a reporter gene, can be used to probe further the mechanism of regulation of the CBR genes. But the current lack of information about the 5$ and 3$ untranslated regions and other regulatory elements of CBR genes makes it difcult to make necessary CBR gene construct modications for generating such rodent CBR transgenic models. Obviously, the use of CBR transgenic animals will provide new in vivo systems for studying genetic regulation, development, and normal function and dysfunction associated with EPCS. Our current investigation and those of others indicate that a number of lines of evidence make the gene that encodes this G-protein-coupled cannabinoid receptors a strong candidate to harbor variants that might contribute to individual differences in human addiction vulnerability.
FATTY ACID AMIDE HYDROLASE KNOCKOUT MICE
The currently known endocannabinoids, derived from membrane phospholipids that contain arachi-donate (Mecholam et al., 1998) are metabolized by FAAH (Deutsch and Chin, 1993), which is a membrane-associated serine hydrolase enriched in brain and liver. There is an overlap of the distribution of FAAH and its activity in the rat brain with the expression of CB1, which has led to the suggestion that FAAH is probably the major enzyme in the brain responsible for inactivation of fatty acid amides (Elphick and Egertova, 2001). Thus, in the human brain the distribution of CB1R and the FAAH enzyme frequently overlap in many structures. As the identity of endocan-nabinoid transporters is still unknown, the generation of FAAH-mutant mice has shed some light into the activity of FAAH. Using homologous recombination, the role that FAAH plays in control-ling fatty acid amide levels and activity in vivo was evaluated in mice that possess a targeted disruption of the FAAH gene (Cravatt et al., 2001). It was demonstrated that mice lacking FAAH were supersensitive to anandamide with an enhanced endogenous cannabinoid signaling (Cravatt et al., 2001). Thus mice lacking FAAH were severely impaired in their ability to degrade anandamide
and when treated with this compound, the mutant animals exhibited an array of intense CB1 R-dependent behavioral responses. In addition, these workers reported that FAAH-mutant mice pos-sess 15-fold augmented endogenous brain levels of anandamide and display reduced pain sensation, which was reversed by the CB1 CB1R antagonist rimonabant. The data from Cravatt et al. (2001) support an important role for FAAH in the metabolism of the amide series and perhaps also for other endocannabinoids. The mechanisms involved in the inactivation of endocannabinoids in vivo are not completely understood. However, functional studies indicate that the biological actions of endocannabinoids are probably terminated by a two-step inactivation process consisting of a carrier-mediated uptake and intracellular hydrolysis by FAAH (Piomelli et al., 1999; Di Marzo et al., 1999;
Hillard, 2000). FAAH has been puried, cloned, sequenced from mouse, rat, and human, and thus fairly well characterized. It is a single-copy gene with 579 amino acids and a highly conserved primary structure, homologous in mouse, rat, and human species (Giang and Cravatt, 1997). This membrane-associated enzyme is 63 kDa and possesses the ability to hydrolyze a range of fatty acid amides including anandamide, 2-AG, and oleamide. The distribution of FAAH and CB1R in rat brain is similar, with FAAH often occurring in neuronal somata that are postsynaptic CB1 R-expressing axons and therefore consistent with a potential role in the regulation of endocannabinoids (Egertova et al., 2000). Whereas FAAH has gene sequence homologous to FAAH enzymes in other species, it is the rst mammalian member of this enzyme family. So far, this enzyme has been identied also in rat and mouse with more than >90% sequence homology to humans, and this homology indicates a general role for the fatty acid amides in mammalian neurobiology. In addition to PMSF, numerous compounds have been identied that block FAAH reversibly and irreversibly.
Among these is ibuprofen, which is an active inhibitor, but not other nonsteroidal anti-inammatory agents (NSAID) such as naproxen. The ability of neuronal tissue to synthesize and rapidly metab-olize anandamide with the aid of a specic transport carrier mechanism suggests either a role for anandamide as an new member of fatty acid-derived neuromodulators or that it could act as a specic neurotransmitter. Enhanced NAE biosynthesis and turnover have been demonstrated in peritoneal macrophages from mice treated with a calcium ionophore (Kuwae et al., 1999). It was suggested that arachidonic acid mobilization induced by ionophore treatment of macrophages could result in the selective generation of anandamide. The biosynthesis and inactivation of endocannab-inoids and other cannabimimetic fatty acid derivatives have been extensively reviewed by Di Marzo et al. (1998b). An acid amidase hydrolyzing anandamide and other N-acylethanolamines distinct from FAAH that can hydrolyze N-acylethanolamines have been reported (Ueda et al., 2000). There is evidence that the synthesis of 2-AG and anandamide can be independently regulated, even though the plasma membranes contain precursor molecules for both anandamide and 2-AG (Piomelli, 2003). Whereas the existence of cannabinoid transporters continues to be controversial (see text following), the differential intracellular hydrolysis of 2-AG and anandamide has been reported (Piomelli, 2003). Apparently, anandamide and 2-AG can be hydrolyzed by distinct serine hydrolases with anandamide predominantly hydrolyzed by FAAH, as discussed, and 2-AG by monoglyceride lipase (MGL), to yield inactive breakdown products (Piomelli, 2003). It appears that there is partial overlap in the distribution of FAAH and MGL in the CNS, with FAAH predominantly localized in postsynaptic structures and MGL mostly associated with nerve endings (Piomelli, 2003). Further research will undoubtedly continue to unravel the biochemical pathways associated with deactiva-tion of endocannabinoids and the role of these processes in marijuana-smoking dependence.
GENES ENCODING ENDOCANNABINOID TRANSPORTERS
Although there is evidence from functional studies for the existence of some form of cannabinoid transporter(s), their identity, sequence information, and biological characteristics at the molecular level are unknown. There is functional evidence that the transport of endocannabinoids such as anandamide and 2-AG across a biological membrane is accomplished via a protein carrier (Piomelli et al., 1999;
Di Marzo et al., 1999; Hillard, 2000). Further evidence has been shown for this carrier-mediated, transmembrane transport of anandamide in human neuroblastoma and lymphoma cells (Maccarone et al., 1998), in mouse macrophages and RBL-2H3 cells (Bisogno et al., 1998), and in neurons (Di Marzo et al., 1994; Hilliard et al., 1997). This transport process fullls several criteria of a carrier-mediated process, including saturability, temperature dependence, high afnity, substrate selectivity, facilitated diffusion, and Na+-independence (Piomelli et al., 1999; Di Marzo et al., 1999; Hillard, 2000). Some of these features make this process fundamentally different from other known transport carriers such as catecholamine and amino acid transporters. Using a relatively potent uptake inhibitor N-(4hydroxyphenyl) arachidonylamide, AM404, these investigators have demonstrated that a high-afnity transport system present in neurons and astrocytes has a role in anandamide uptake and subsequent inactivation by FAAH (Piomelli et al., 1999; Di Marzo et al., 1999; Hillard, 2000). AM404 has been reported to inhibit anandamide uptake by rat-cultured cortical neurons and astrocytes and to potentiate anandamide both in vitro and in vivo (Piomelli et al., 1999; Di Marzo et al, 2000c). When administered to rats by itself, AM404 increases plasma levels of anandamide and shares the ability of this endocannabinoid to decrease locomotor activity, depress plasma levels of prolactin, and alter tyrosine hydroxylase activity in different brain regions. The inhibitory effect of AM404 on locomotor activity has been found to be susceptible to antagonism by rimonabant.
However, AM404 does not elicit typical cannabinoid responses of catalepsy in the Pertwee ring test, nor does it produce signs of analgesia in the hot plate test. While there is ample scientic evidence to support the concept that anandamide transport across membranes is protein-mediated, denitive evidence awaits its molecular characterization. Other studies have suggested that passive diffusion alone is sufcient for anandamide uptake, with intracellular hydrolysis by FAAH main-taining the concentration gradient (Glaser et al., 2003). But recent work using FAAH knockout mice show that anandamide uptake is the same in FAAH-mutant and wild-type mice and is blocked by anandamide transport inhibitors, suggesting that anandamide transport is independent of FAAH (Fegley et al., 2004). Although this work and the report of the blockade of endocannabinoid long-term depression (LTD) induction by anandamide uptake inhibitors (Ronesi et al., 2004) functionally support the existence of an anandamide transporter, the molecular identity of the anandamide transporter is still elusive. Furthermore, the differential uptake of the different endocannabinoids (for example, anandamide and 2-AG in different cell types) may indicate the possibility of different cannabinoid transporters for the different endocannabinoids (Di Marzo et al., 1999; Hillard, 2000).
Further research will show whether the additional endocannabinoids being discovered, such as noladin ether, will be metabolized and taken up by similar metabolic and uptake inhibitors. This will not be unprecedented because the monoamines have different transporters for dopamine, serotonin, and norepinephrine.
POLYMORPHIC STRUCTURE OF CANNABINOID RECEPTOR GENES
New information on the CBR gene and its allelic variants in humans and rodents can add to our understanding of vulnerabilities to addictions and other neuropsychiatric disorders and the genetic basis of marijuana use and addiction in vulnerable individuals. Little information is, however, available at the molecular level about CBR gene structure, regulation, and polymorphisms. Different human CBR gene polymorphisms have been reported. A silent mutation of a substitution from G to A at nucleotide position 1359 in codon 453 (Thr), which turned out to be a common polymorphism in the German population, was reported (Gadzicki et al., 1999). In this study, allelic frequencies of 1359(G/A) in genomic DNA samples from German GTS patients and controls were determined by screening the coding exon of the CB1R gene using PCR single-stranded conformation polymor-phism (SSCP) analysis (Gadzicki et al., 1999). This was accomplished by the use of a PCR-based assay by articial creation of a MSP1 restriction site in amplied wild-type DNA (G-allele), which is destroyed by A-allele (Gadzicki et al., 1999). They found no signicant differences of the allelic distributions between GTS patients and controls within the coding region of the CB1R gene.
In our studies, the frequencies of this polymorphism are signicantly different between the Caucasian, African-American, and Japanese population (Ishiguro et al., unpublished observation).
There is a HindIII restriction fragment length polymorphism (RFLP) located in an intron approxi-mately 14 kb in 5$ region of the initiation codon of the CB1R gene. Caenazzo et al. (1991) genotyped 96 unrelated Caucasians, using hybridization of human DNA digested with HindIII, and identied two allele with bands at 5.5 (A1) and 3.3 kb (A2). The frequencies of these alleles were 0.23 and 0.77, respectively. Another polymorphism is a triplet repeat marker for the CB1R gene. This is a simple sequence-repeat polymorphism (SSRP) consisting of nine alleles containing (AAT) 12-20- repeat sequences that was identied by Dawson, (1995). This polymorphism has been used in linkage and association studies of the CB1R gene with mental illness and drug abuse in a different population. This CB1R gene triplet repeat marker was used to test for a linkage with schizophrenia using 23 multiplex schizophrenia pedigrees (Dawson, 1996), and associations with heroin abuse in a Chinese population (Li et al., 2000) and intravenous (IV) drug use in Caucasians (Comings, 1997). There was no linkage and association of the marker with schizophrenia, indicating that the CB1R gene is not a polymorphism of major etiological effect for schizophrenia, but a CB1R gene might be a susceptibility locus in certain individuals with schizophrenia, particularly those whose symptoms are apparently precipitated or exacerbated by cannabis use (Dawson, 1996). Comings et al.
(1997) hypothesized that genetic variants of the CB1R gene might be associated with susceptibility to alcohol or drug dependence and analyzed the triplet repeat marker in the CB1R gene. They found a signicant association of the CB1R gene with a number of different types of drug dependence (cocaine, amphe-tamine, cannabis) and IV drug use but no signicant association with variables related to alcohol abuse and dependence in nonHispanic Caucasians. In addition, this group also reported that a signicant association of the triplet repeat marker in the CB1R gene alleles with the P300-event-related potential that has been implicated in substance abuse (Johnson et al., 1997). Li et al. (2000) attempted to replicate the nding of Comings et al. (1997) in a sample of Chinese heroin addicts and did not nd any evidence that CB1R gene AAT repeat polymorphism confers susceptibility to heroin abuse. CB1R gene is located in human chromosome 6q14-q15, and it is interesting that previous reports showed evidence for suggestive linkage to schizophrenia with chromosome 6q markers (Martinez et al., 1999). Suggestive evidence also exists for a schizophrenia susceptibility locus on chromosome 6q (Cao et al., 1997). Although there was no linkage and association of the CB1R gene triplet marker with schizophrenia, it remains to be determined if linkage and association to schizophrenia might exist with other unknown polymorphisms that might exist in the CB1R gene structure, which is currently poorly characterized. Three other variants have been reported in the CB1R gene of an epilepsy patient (Kathmann et al., 2000). This was obtained from PCR assay with cDNA from hippocampal tissue taken from patients undergoing neurosurgery for intractable epilepsy. They detected four mutations in the coding region of the CB1R gene, with the rst three mutations yielding amino acid substitutions.
We have initiated a series of studies to analyze CB1R gene structure, regulation, and expression in the mouse and human models to determine genotypic and haplotypic associations of CB1R gene with addictions and other neuropsychiatric disturbances. Genotypes at markers near the mouse Chr 4, 13.9 cM CB1/CNR locus in 9 mouse strains reveal apparent haplotypes that extend from at least D4Mit213 to D4Mit90. These haplotypes can be correlated with strain differences in cannabinoid effects.
The human Chr 6, 91.8~96.1 cM CB1R gene locus encodes at least four exons which account for 24 to 28 kb of sequence Figure 3.3. Examination of CB1R gene sequence variations in distinct populations has revealed a G/A single nucleotide polymorphism (SNP) in CB1 5$ anking sequences.
The initial values for linkage disequilibrium between these markers and genotypic frequencies of the markers in drug abusing and control populations were calculated. The A-allele of the SNP polymor-phism was present in fewer African-Americans and Asians than in Caucasians. However, in the Caucasian and African-American samples used, no association between drug abuse and the 1359(G/A) polymorphism could be found (Ishiguro et al., unpublished observation). Furthermore, 1359(G/A)
polymorphism has been determined in healthy control male Japanese subjects (aged 20 to 30) for whom personality traits were measured with a temperament and character inventory (TCI). Although the statistical power is weak because of fewer frequency of allelic distribution, no association between TCI scores and this polymorphism could be found (Ishiguro et al., unpublished observa-tion). While these studies continue, these ndings add to the characterization of these CB1R genes in species in which they can be tested for impact on substance abuse and other neuropsychiatric disorders. The amino acid sequence alignments and phylogenetic tree of known CBRs indicate some similarities and signicant divergence between CB1 and CB2 CBRs. Polymorphisms at the CB1R gene may be associated with the diverse actions of marijuana use. They may be a major factor which, when triggered by environmental, age, and metabolic factors, could lead to the continuous but mild circle of marijuana dependence and use.
CHROMOSOMAL MAPPING OF THE CBR GENES
With in situ hybridization using a biotinylated cosmid probe, the CB1R gene in humans was localized at 6q14-q15, thus conrming the linkage analysis and dening a precise alignment of the genetic and cytogenetic maps (Hoehe et al., 1991). These investigators found that the location of the human CB1R gene is very near the gene encoding the alpha subunit of chorionic gonadotropin (CGA). It has been determined that the mouse CB1 and CB2 CBR genes are located in proximal chromosome 4 (Stubbs et al., 1996). This location is within a region in which other homologs of human 6q genes are located. In order to localize the mouse CB1 and CB2 CBR genes in the mouse genome, Stubbs et al. (1996) traced the inheritance of species-specic variants of the gene in 160 progeny of an interspecic backcross. Therefore, using the interspecic and four DNA probes we mapped the mouse CB1 and CB2 CBR genes to chromosome 4 with map positions calculated for Mos, Cntfr, Pax5, and Cd72, in excellent agreement with previously published results that clearly established linkage between CB1R gene and other genes known to be located on mouse chromosome 4. The CB1R gene, GABRR1, GABRR2, and Cga are linked together both in the mouse and on human chromosome 6q. The genes encoding the peripheral CB2R and "-L-fucosidase have been shown to be located near a common virus integration site, Ev11 (Valk et al., 1997). They showed that Ev11 is located at the distal end of mouse chromosome 4 in a region that is synthenic with human 1p36, in agreement with our report (Onaivi et al., 2002), that the mouse CB2R gene is also located at the distal end of mouse chromosome 4. The results of the chromosomal location of the human CB1R gene (Hoehe et al., 1991) and the mouse (Stubbs et al., 1996; Onaivi et al., 2002) CBR genes add a new marker to this region of the mouse-human homology, and conrm the close linkage of FIGURE 3.3 Part of human chromosome 6 genomic sequences (complete or draft) of a part of chromosome 6p. The completed sequence covers more than 200kb around the CB1/CNR gene. The computer program GENSCAN (http://CCR-081.mit.edu/GENCAN.html) estimated four fairly possible gene sequences (including CB1/CNR) in this region. These estimated sequences showed similarity to known protein sequences, revealed by computer program BLAST (http://www.ncbi.nlm.nih.gov/BLAST/).
AL 121835
ACTB T8 ( b actin pseudogene 8 )
AL 136096
AL 121942
SPACA 1
Cannabinoid receptor type 1
10 kb
CBR genes in both species. The location of the rat CB1R gene in the rat genome has been determined and localized at 5q13-q24 and ts the rodent-human homology as the CB1R genes are highly conserved in the mammalian species. The physical and genetic localization of the bovine CB1R gene has also been mapped to chromosome 9q22 using uorescence in situ hybridization (FISH) and R-banding to identify the chromosome (Pster-Genskow et al., 1997). The genetic mapping of the CB1R gene on bovine chromosome 9q22 by in situ localization and linkage mapping of a dinucleotide repeat, D9S32, also adds to coverage of the bovine genome map and contributes to the mammalian comparative gene map (Pster-Genskow et al., 1997). The location of the human CB1 genomic sequences (using complete or draft) of a part of chromosome 6p is shown in Figure 3.3.
The completed sequences cover more than 200kb around the CB1R gene. The program GENESCAN was used to estimate four fairly possible gene sequences including CB1R in this region. These estimated sequences showed similarity to known protein sequences, revealed by BLAST. As the neurobiological effects of marijuana and other cannabinoids suggest the involvement of the CBR genes in mental and neurological disturbances, the mapping of the genes will undoubtedly enhance our understanding of the linkage and possible cannabinoid genetic abnormalities. In the case of the cattle, research into the role of CBRs in mediating responses to natural and production-induced stressors could lead to improvement of production inefciencies that exists in meat and milk animal systems (Pster-Genskow et al., 1997). The chromosomal location and genomic structure of human and mouse FAAH genes have been mapped to chromosomes 1p34-p35 and 4, respectively (Wan
The completed sequences cover more than 200kb around the CB1R gene. The program GENESCAN was used to estimate four fairly possible gene sequences including CB1R in this region. These estimated sequences showed similarity to known protein sequences, revealed by BLAST. As the neurobiological effects of marijuana and other cannabinoids suggest the involvement of the CBR genes in mental and neurological disturbances, the mapping of the genes will undoubtedly enhance our understanding of the linkage and possible cannabinoid genetic abnormalities. In the case of the cattle, research into the role of CBRs in mediating responses to natural and production-induced stressors could lead to improvement of production inefciencies that exists in meat and milk animal systems (Pster-Genskow et al., 1997). The chromosomal location and genomic structure of human and mouse FAAH genes have been mapped to chromosomes 1p34-p35 and 4, respectively (Wan