Conserved Chromosomal Location and Genomic Structure of Human

GENOMICS 54, 408–414 (1998) ARTICLE NO. GE985597

Conserved Chromosomal Location and Genomic Structure of Human and Mouse Fatty-Acid Amide Hydrolase Genes and Evaluation of clasper as a Candidate Neurological Mutation Minghong Wan,* Benjamin F. Cravatt,†,‡ Huijun Z. Ring,* Xianyu Zhang,§ and Uta Francke*,§,1 *Department of Genetics, Stanford University School of Medicine, and §Howard Hughes Medical Institute, Stanford, California 94305; †The Skaggs Institute for Chemical Biology and ‡Department of Cell Biology, The Scripps Research Institute, La Jolla, California 92307

Received July 31, 1998; accepted September 22, 1998

affect serotonergic systems by potentiating the re- Fatty-acid amide hydrolase (FAAH) is a membrane- sponse of 5-HT receptors to serotonin (Boger et al., bound enzyme that degrades neuromodulatory fatty 1998; Huidobro-Toro and Harris, 1996; Thomas et al., acid amides, such as oleamide and anandamide, and is 1997a). Anandamide was originally isolated from brain expressed in the mammalian central nervous system. tissue as an endogenous ligand for the CB1 cannabi- To evaluate FAAH genes as candidates for neuroge- noid receptor (Devane et al., 1992) and has been shown netic diseases in humans and mice, we have mapped the loci in both species and have determined their to induce cannabinoid-like effects in mice (Smith et al., intron–exon structures. The human FAAH gene was 1994). Furthermore, both oleamide and anandamide mapped to region 1p34–p35, closely linked to D1S197 have been found to block gap junction communication and D1S443, by using PCR analysis of somatic cell in glial cells (Guan et al., 1997). hybrid (SCH) and radiation hybrid mapping panels. FAAH cDNAs have recently been cloned and char- Analysis of an SCH mapping panel and a mouse inter- acterized from rat, mouse, and human liver cDNA specific backcross panel has localized the Faah gene libraries (Cravatt et al., 1996; Giang and Cravatt, to the conserved syntenic region on mouse chromo- 1997). Transient expression of the FAAH enzyme in some 4, close to the neurological mutation clasper. COS-7 cells demonstrated that the enzyme was able Faah gene rearrangements were excluded by South- to hydrolyze both oleamide and anandamide, as well ern blot analysis of clasper DNA. No sequence abnor- as several other fatty acid amides (Cravatt et al., mality was detected in PCR products containing the 15 exons and splice junctions of the mouse Faah gene. 1996). In addition, in situ hybridization has revealed FAAH protein levels were normal in clasper mouse widespread distribution of rat FAAH mRNA in neu- tissues as determined by enzyme activity assays and ronal cells in the central nervous system (CNS) Western blotting. © 1998 Academic Press (Thomas et al., 1997b). The primary structure of the enzyme is well conserved in mammals, with mouse and rat FAAHs sharing 91% of their amino acids and INTRODUCTION human FAAH being 82 and 84% identical to rat and mouse FAAH, respectively. As a metabolic enzyme Fatty-acid-amide hydrolase (FAAH) is a membrane- coupled with endogenous neuromodulatory mole- bound enzyme involved in the degradation of bioactive cules, FAAH may play important roles in the CNS by fatty acid amides. FAAH’s substrates include the neu- ensuring rapid termination of signaling. To evaluate romodulatory signaling molecules oleamide and anan- FAAH genes as candidates for neurogenetic diseases damide. Oleamide (cis-9-octadecenamide) is an endog- in humans and mice, we mapped the chromosomal enous sleep-inducing lipid originally isolated from the locations of the human and mouse FAAH genes by cerebrospinal fluid of sleep-deprived cats (Lerner et al., 1994; Cravatt et al., 1995). In addition, oleamide can using PCR analysis of two panels of somatic cell hybrid (SCH) lines and one radiation hybrid (RH) Sequence data from this article have been deposited with the panel as well as by linkage analysis of a mouse GenBank Data Library under Accession Nos. AF097997–AF098019. interspecific backcross panel. We also determined 1 To whom correspondence should be addressed at the Howard the genomic structures of both human and mouse Hughes Medical Institute, Stanford University School of Medicine, Stanford, CA 94305-5323. Telephone: (650) 725-8089. Fax: (650) FAAH genes and evaluated the mouse FAAH gene as 725-8112. E-mail: [email protected]. a candidate gene for the clasper mouse mutant.

MATERIALS AND METHODS forward and reverse primers were designed about 20 bp away from putative splice sites and used for sequencing the BAC inserts into the Somatic cell hybrids and PCR amplifications. The human FAAH putative introns. Direct sequencing of BAC DNA was performed on gene was localized to a human chromosome by using a panel of 17 an ABI Prism 377 DNA sequencer using the Big Dye Terminator human ϫ Chinese hamster hybrid cell lines derived from several Cycle Sequencing Kit (Perkin–Elmer ABI), following the protocol independent fusion experiments (Francke et al., 1986). PCR primers recommended by the manufacturer. Briefly, 5 ␮g of BAC genomic (hFAAH-F 5Ј-GAGGCTTCCGTGTCCTCTC-3Ј and hFAAH-R 5Ј- DNA was used for each 20-␮l reaction containing 8 ␮l Big Dye CCTATGTCATACCCATGGGC-3Ј) were designed to specifically am- Terminator Mix and 15 pmol primers. The cycle conditions were plify a sequence of 138 bp from the 3Ј untranslated region (UTR) of 95°C for 5 min and then 35 cycles of 95°C for 30 s, 55°C for 20 s, and the human FAAH gene. Genomic DNAs from the mapping panel 60°C for 4 min. The intron–exon boundaries were determined by were analyzed by PCR. Similarly, the mouse Faah gene was assigned comparing the cDNA sequence and the corresponding genomic se- to a mouse chromosome by PCR analysis of DNA from a mapping quences using Sequencher 3.0 (GeneCodes Corporation, Ann Arbor, panel consisting of 19 mouse ϫ Chinese hamster somatic cell hybrid MI). The introns were PCR amplified using exonic primers, and the lines as described previously (Li et al., 1993). Primers (sense, mFaah- intron sizes were estimated by comparing the PCR products with 2F, 5Ј-GGCCATCTTGAGGGTCATTC-3Ј; antisense, mFaah-1R, 5Ј- 1-kb and 100-bp DNA size markers (GIBCO BRL, Gaithersburg, GGGAGCTGGTCCAGACTTG-3Ј) were designed to amplify a 197-bp MD) as well as a 20-bp DNA ladder (McFrugal’s Lab Depot) running fragment from the 3Ј UTR of the mouse Faah gene. PCR amplifica- on 3% agarose gels or by sequencing the PCR products directly using tions were carried out using 50 ng genomic DNA in 50-␮l final the same exonic PCR primers. volume of 10 mM Tris–HCl (pH 8.3), 50 mM KCI, 1.25 mM MgCI2, Mutation analysis of the mouse Faah gene. Genomic DNAs from 250 ␮M each dNTP, and 2.5 U Taq DNA polymerase (Perkin–Elmer a homozygous mutant mouse B6 ϫ BALB/cBy–cla/cla and a possibly ABI, Foster City, CA). The PCR cycle conditions for both human and heterozygous BALB/cBy–cla/ϩ mouse were purchased from The mouse FAAH gene amplifications were 94°C for 3 min and then 35 Jackson Laboratory. Southern blot analysis was carried out follow- cycles of 94°C for 30 s, 55°C for 45 s, and 72°C for 1 min followed by ing a standard protocol (Sambrook et al., 1989). All 15 exons of the 72°C for 5 min. The PCR products were sequenced to confirm their mouse Faah gene were PCR amplified by using the intronic primers identities. listed in Table 3 using 50 ng of genomic DNA in 50 ␮l final volume of

Human radiation hybrid mapping. A human RH mapping panel, 10 mM Tris–HCl (pH 8.3), 50 mM KCI, 1.25 mM MgCI2, 250 ␮M each GeneBridge 4, was used to confirm and define further the localization dNTP, and 2.5 U Taq DNA polymerase. The PCR products were of the human FAAH gene (Walter et al., 1994). Typing of the map- sequenced on the ABI Prism 377 DNA sequencer. The sequences ping panel was carried out by using the same 3Ј UTR human FAAH were compared with published sequences (Giang and Cravatt, 1997) primers and PCR conditions described above. The maximum likeli- using Sequencher 3.0 to search for potential mutations. hood analysis (Boehnke et al., 1991) results were obtained by sub- FAAH enzyme assay and Western blot analysis. BALB/c, Black mitting the raw scores to http://www-genome.wi.mit.edu/cgi-bin/ Swiss, and B6 ϫ BALB/cBy-cla/ϩ mice were obtained from The contig/rhmapper.pl. Jackson Laboratory. Adult mice were sacrificed by cervical disloca- tion, and heart, liver, and brain tissues were isolated and Dounce- Linkage mapping with interspecific backcross mice and single- homogenized in phosphate-buffered saline, and the protein was con- strand conformation analysis (SSCA). Fine localization of the centrated (D Protein Assay kit; Bio-Rad). Protein from each tissue mouse Faah gene was done by genotyping the interspecific backcross c was analyzed by Western blotting (50 ␮g protein) and by FAAH (C57BL/6JEi ϫ SPRET/Ei)F ϫ SPRET/Ei (BSS) panel (The Jackson 1 enzymatic assays (100 ␮g protein) as described previously (Cravatt Laboratory DNA Mapping Resource, Bar Harbor, ME). To find a et al., 1995, 1996), with the following exceptions: (1) affinity-purified polymorphism, a 94-bp fragment was PCR amplified from DNA anti-FAAH polyclonal antibodies raised against a FAAH–GST fusion samples of the parental strains by using primer pair mFaah-1F protein (containing amino acids 30–579 of FAAH; Patricelli et al., (5Ј-CATGAGGAGATGCCCAGC-3Ј) and mFaah-1R (described 1998) were used for Western blotting and (2) in the FAAH enzymatic above). Five microliters of PCR products were denatured and loaded assay, the conversion of [14C]oleamide to oleic acid was quantified by onto a 12% nondenaturing polyacrylamide minigel for SSCA as de- phosphorimager analysis of thin-layer chromatography plates (Pack- scribed previously (Sugano et al., 1993). All 94 DNA samples from ard Phosphorimager). the BSS panel were typed by PCR-SSCA and scored. The typing results were sent to The Jackson Laboratory backcross service for comparison with previous typing data of markers on mouse chromosome 4. The mouse Faah gene mapping figures were prepared by RESULTS Lucy Rowe (The Jackson Laboratory). Chromosomal Localization of the Human FAAH Gene Characterization of exon–intron boundaries of human and mouse FAAH genes. The genomic structures of the FAAH genes were The human FAAH gene was first mapped to a human determined by direct sequencing of bacterial artificial chromosome chromosome by PCR typing of an SCH mapping panel. (BAC) genomic clones and comparing the sequence with the cDNA A 138-bp PCR product was generated from DNA ex- sequence. The human FAAH genomic BAC clone was isolated by tracted from human cells and from five hybrid cell lines PCR screening of human BAC DNA pools release III (Research Genetics, Inc., Huntsville, AL), with the same 3Ј UTR primers de- that contained human chromosome 1, but not from signed for mapping. The mouse FAAH genomic BAC clone was hamster DNA. The human-specific PCR product was isolated by Genome Systems by hybridization with a 400-bp cDNA not obtained from a hybrid cell line (XV-18A-10b) that fragment from the 5Ј region as a probe. Both human and mouse contained only chromosome region 1p32–qter (data not FAAH BAC clones were grown in LB media containing 12.5 ␮g/ml chloramphenicol. The BAC DNA was purified using the QIAGEN shown). By concordant segregation and by exclusion of plasmid kit/QIAGEN-Tip 500 (Qiagen, Inc., Valencia, CA), following all other chromosomes, the human FAAH gene was the manufacturer’s protocol for purification of low-copy-number plas- assigned to chromosome region 1pter–p3200 (Fig. 1A). mids. RH mapping confirmed the chromosomal location of The putative splice sites in the human and mouse FAAH cDNA the human FAAH gene and also identified the position sequences (GenBank Accession Nos: U82535 for human and U82536 for mouse) were predicted using the Baylor College of Medicine Gene of this gene relative to polymorphic markers on the Finder Program at http://dot.imgen.bcm.tmc.edu:9331/gene-finder/ linkage map. Seventeen of the 93 RH cell lines were gf.html by choosing “HSPL/Search For Potential Splice Sites.” Then positive for the human-specific FAAH gene signal (re- 410 WAN ET AL.

marker D1S200 were found on the Whitehead Institute integrated map, and their physical map positions cor- responded to 1p34–p35 (Bray-Ward et al., 1996). Therefore, the FAAH gene is most likely located at 1p34–p35 in a region that also contains the MPL gene, the human homolog of the mouse myeloproliferative leukemia virus gene v-mpl, a member of the cytokine receptor superfamily (Skoda et al., 1993) (Fig. 1).

Chromosomal Mapping of the Mouse Faah Gene The mouse Faah gene was ﬁrst mapped to a mouse chromosome by PCR analysis of an SCH mapping panel as described under Materials and Methods. A 197-bp PCR product was ampliﬁed from genomic DNA extracted from mouse cells and from mouse ϫ Chinese hamster hybrid lines containing mouse chromosome 4 when primers mFaah-2F and mFaah-1R were used. The mouse Faah gene was assigned to chromosome 4

FIG. 1. Chromosomal localization of the human FAAH gene. (A) Idiogram of the short arm of human chromosome 1 (Francke, 1994) illustrating the localization of the FAAH gene by somatic cell hybrid (SCH) and radiation hybrid (RH) mapping. (B) A partial linkage map FIG. 2. Chromosomal localization of the mouse Faah gene. (A) of human chromosome 1 showing the FAAH locus and flanking SSCA of Faah-specific PCR products from parental C57BL/6JEi (B) markers. Distances between the STS markers (in centirays) were and SPRET/Ei (S) mice as well as from a (C57BL/6JEi ϫ SPRET/Ei) calculated using a two-point linkage program. ϫ SPRET/Ei (BS) backcross animal. The two arrowheads a and b point to two single conformers present in either the B or the S strain. MW indicates the 100-bp DNA size marker. (B) Partial linkage map tention frequency 18.3%). By maximum likelihood of the distal region of mouse chromosome 4 with markers typed on analysis, the human FAAH gene was placed 3.36 cR The Jackson Laboratory BSS backcross. The centromere is toward from chromosome 1 marker CHLC.GATA72H07 (lod Ͼ the top. A 3-cM scale bar is shown on the left. Loci mapping to the 3.0) proximal to D1S443 and distal to D1S197 (Fig. same position are listed in alphabetical order. Raw typing data were obtained from the website at http://www.jax.org/resources/documents/ 1B). To get cytogenetic information about these two cmdata. (C) Haplotypes from The Jackson Laboratory BSS backcross STS markers and the human FAAH gene, we looked for showing part of chromosome 4 with loci linked to Faah. Loci are two more distantly located flanking markers and ex- listed in order with the most proximal at the top. The black boxes amined their chromosomal locations by searching represent the C57BL/6J (B) allele and the white boxes represent the SPRET/Ei (S) allele. The numbers of offspring inheriting each type of available integrated mapping information in the liter- chromosome are listed at the bottom of each column. The percentage ature and in the on-line genome center databases. The of recombination (R) between adjacent loci is given to the right, with more distal marker D1S441 and the more proximal the standard error (SE) for each R. MAPPING AND STRUCTURE OF HUMAN AND MOUSE FAAH GENES 411

TABLE 1 Intron–Exon Boundaries of the Human FAAH Gene

Exon Length (bp) 5Ј splice site 3Ј splice site

1 195a ATGGT TCCAG gtgactgccggagcgtattg 2 114 tccgagtttgttccccacag AACCC GAAAG gtaaggccagtcaaggccag 3 135 gtctacttccccttcctcag GCCTG ACAAG gtatgctctgcctcagcgcc 4 134 gtgcccatccctcctcccag GGCCA TTCAG gttgggtcttgggttggggc 5 207 cccattcttggctcctccag CTATG CTCAG gtaaggtgggtggagggcgct 641atcttatgtttcttatccag CAAGA GGCAG gtgaggtccgtggtgctctc 7 125 ctcacctctctgccccacag TGCGT AAGAG gtgagcagggctgggtggg 8 126 cccctgtgttttccctccag GTCTA ACACG gtatgactgcagggtcctgg 998acacccctcaatccctgcag CTGGT AACTT gtgagtgatagtgggctttg 10 100 ccgagcattttgtttcccag CAAAG CTCTG gtgagggcacaaggagtgga 11 41 ccctgcgtctgtctgtgcag CTGCC TCTCG gtaaggttcttctgtgtcta 12 40 ccacactccttctgcccaag TTCGG TCGAG gtgaggcaagagcctctgga 13 109 gtgtgttgtgtcctccgcag GTGTA CACAG gtgaggcccgacaccctgcc 14 146 ctgcctgtaatgtgttccag GGGCC AGAAG gtgaggactgacctgccctt 15 129b gatgcctgtatcccctatag GGCAT CCTGA

a Length from initiator codon (underlined). b Length to the stop codon (underlined) of the open reading frame. by complete concordance of the 197 bp PCR product cation of the mouse Faah gene is consistent with as- with mouse chromosome 4 in these hybrid lines. All signment of the human FAAH gene to 1p34–p35. other mouse chromosomes were excluded by at least three discordant hybrids. Characterization of Exon–Intron Boundaries of the We also performed linkage mapping of the mouse Human and Mouse FAAH Genes Faah gene by PCR-SSCA genotyping of The Jackson Laboratory backcross BSS panel. A polymorphism was To facilitate screening for mutations in human ge- identified when primer pair mFaah-1F/mFaah-1R was netic disorders or mouse mutants mapped close to the used for PCR-SSCA. As shown in Fig. 2A, several FAAH locus, we determined the genomic structure of bands were detected by SSCA; conformer a was present the human and mouse FAAH genes. Two BAC genomic only in C57BL/6J and not in SPRET/Ei, while con- clones, each containing either the human or the mouse former b was specific for SPRET/Ei. The mouse Faah gene, were isolated by screening human and mouse gene was mapped to a distal region of mouse chromo- BAC libraries, respectively. The intron–exon bound- some 4, cosegregating with several other loci, including aries and flanking intron sequences were determined Mpl (Skoda et al., 1993) (Fig. 2B). Since this part of by using primers in the FAAH coding sequence for mouse chromosome 4 contains a large conserved syn- direct sequencing of BAC genomic DNA as templates. tenic region with human chromosome 1p, the map lo- We found that both human and mouse FAAH genes

TABLE 2 Intron–Exon Boundaries of the Mouse Faah Gene

Exon Length (bp) 5Ј splice site 3Ј splice site

1 195a ATGGT TGCAG gtgaaggcgggcgccggggc 2 114 acgtaatgtgttacccgcag AATCC GAAAG gtaagccctaagttagtcc 3 135 ccttgtatccctttccctag GCCTG ACAAG gtacaccctgcttgtcctgc 4 134 cgctgtcttctttctctaag GGCCA CTAAG gttggtcccccggagctggt 5 207 ccttctcctgcttcctccag CTATG CTCAG gtacgttggggacagggagg 641atcttgtgtttctgacccag CAAGA AGCAG gtgagtgaggtctatggtac 7 125 tctcccttcccttcctacag TGCAG AGGAG gtgagtgaggataggaatga 8 126 ttgcctacgtttgcttctag ATCTA ACACG gtatggcatctcctttccct 998acccccttcaatctccgcag CTGGT AACTT gtgagtgtgatgggcttgag 10 100 ctaagccttttgtcttccag CAAAG CTCTG gtgagagcaccaggcatgca 11 41 ctgtgtctgttctggtgcag TTTCC CCTCG gtgagaacttcaatgcctgc 12 40 cttcctctctctctgctcag GTCAG TTGAG gtaaggtctgggccctggag 13 109 atgcgcttgtgtctctgcag ATGTA CACAG gtgaggcctattgcccctct 14 146 ccagtgtgttcttccccccag GGGCT AGAAG gtaaagcctgcctgtccccc 15 129b gatgtctgtatcccatgtag GGCAT CTTGA

a Length from initiator codon (underlined). b Length to the stop codon (underlined) of the open reading frame. 412 WAN ET AL.

TABLE 4 Activity of FAAH Enzyme in Tissues Isolated from Normal and clasper Mice

BALB/c FAAH Black Swiss FAAH clasper FAAH activities activities activities (nmol/min mg) (nmol/min mg) (nmol/min mg)

Liver 3.9 Ϯ 0.5 3.4 Ϯ 0.4 3.4 Ϯ 0.5 Brain 2.3 Ϯ 0.5 2.2 Ϯ 0.4 2.4 Ϯ 0.5

FIG. 3. Comparison of the genomic organization of the human and mouse FAAH genes. The coding regions are represented by filled erozygous clasper mouse were searched for by South- boxes and untranslated regions by open boxes. The exon and intron ern blot analysis and by direct sequencing of PCR sizes are drawn to scale. The intron sizes of the human gene are 1, amplimers. Ten micrograms of DNA samples from both ϳ330 bp; 2, ϳ5 kb; 3, 222 bp; 4, 81 bp; 5, 239 bp; 6, 160 bp; 7, ϳ2 kb; 8, ϳ550 bp; 9, ϳ1 kb; 10, 312 bp; 11, ϳ780 bp; 12, ϳ0.5 kb; 13, ϳ0.8 clasper mutant mice and from a normal C57BL/6J kb; 14, 235 bp. The intron sizes of the mouse gene are 1, ϳ450; 2, control mouse were digested with EcoRI and probed ϳ200 bp; 3, 158; 4, 93 bp; 5, ϳ270; 6, ϳ230 bp; 7, ϳ1 kb; 8, ϳ450 bp; with a PCR-amplified mouse 1.9-kb cDNA probe. Two 9, ϳ1 kb; 10, ϳ320 bp; 11, ϳ780 bp; 12, ϳ0.5 kb; 13, ϳ0.8 kb; 14, distinct bands, between 11 and 13 kb, were detected in 224 bp. both clasper and normal control DNA samples on Southern blots (data not shown). These results exclude contain 15 exons. The positions of all exon–intron any large-scale deletions, duplications, or rearrange- boundaries are conserved. Exon 1 contains the consen- ments of the Faah gene in clasper mice. The intron- sus ATG translation initiation codon, and exon 15 con- based primers listed in Table 3 were used to amplify tains the TGA translation stop codon and the polyad- and to sequence directly the PCR products. No muta- enylation consensus sequence for both human and tion was found in the coding regions and splice junc- mouse FAAH genes. The intron–exon boundaries are tions. Since mutations in promoter or regulatory ele- summarized in Table 1 and Table 2. As shown in the ments located outside of the amplified regions would tables, most splice sites match the AG/GT rule. Even not have been detected, we compared the FAAH en- the sizes of introns were strikingly similar (Fig. 3), zyme activities in tissues from normal and clasper with the exceptions of intron 2, which spans ϳ5kbin mice. the human gene compared to ϳ0.2 kb in the mouse Comparisons of FAAH enzyme activity in liver and gene, and intron 7, which is ϳ2 kb in human and ϳ1kb brain tissues from adult normal (BALB/c and Black in mouse. The genomic structures of the human and Swiss) and clasper mice revealed no significant differ- mouse FAAH genes are compared in Fig. 3. ences in relative FAAH activities (Table 4). Addition- ally, Western blotting showed equivalent FAAH pro- Evaluation of the Mouse Faah Gene as a Candidate tein levels in normal and clasper mice (Fig. 4). Finally, Gene for the clasper Mutation no FAAH enzyme activity or protein was observed in Potential mutations in the mouse Faah gene in DNA heart tissue isolated from clasper mice, indicating that samples from one homozygous and one possibly het- an upregulation of FAAH levels did not occur in these

TABLE 3 Primer Sequences for PCR Ampliﬁcation of Coding Regions of the Mouse Faah Gene

Anneal. Amplimer temp. length Exon Forward primer (5Ј–3Ј) Reverse primer (5Ј–3Ј) (°C) (bp)

1 GCTGTTGGTGTGCGCGTGCCGAGTCCTCTC GCCCTCTGGCGTCCGGGCAGTTGAAGTCGCA 68 340 2 AGATTGGAATGTAAAGTGAATAATTAATG TCAGCTTGTAAAGAACCAGGAC 55 234 3 TCTGGGCCATGCTTCTGGTTAC ACCAAGGCATCTTGGACCCAAAGC 60 294 4 TGCATCTAGCTTGGGTCCAGATGC TCACCTCATCAACTGCCTCTGTC 61 321 5 AGCTGGTCTAACCAATGAGGTG CAAGGTTGTGCTTTCCTTAGGTC 59 340 6 CTAGGCTGCAGGAGGGATTCACTAACCAGC CCATACACACCCTTCAGGGTACCA 63 264 7 TGCCAAGCCACTGTAGGTGTCAATC GATCATTCCTATCCTCACTCAC 58 274 8 TCCTGCTAAGTTGCCAGGAAGAC ACCAACATTTGCCCCAATACTCTCAC 61 481 9 TGTGATTATGTGGCGCTGATATCCTCG GATCATCTGGAAGAGCATCCCGTGCTC 63 301 10 AGGTGGACAGAACTGTTGAGTTACGG GCCAGAGACCACAAGGTAGTGTGAG 63 279 11 AAGGGTCATGACCTGGACCTGGAAGGT ACGTGAACTCAGACTAGACCAGTTGA 61 329 12 ATCCTCTCTTGTAGCTCAAGAGTCAC CCACTCACTAACACAGACATAGAC 60 537 13 CTATTGGTGGCTCAGCTACAACAGATGC ATAGCTCCACAGGGCACAGGGCTGAACTC 64 317 14 ATGTGTGGACAAGGGTGACTTAGGC GTCCCAGGAATACCTTGTGCTTCTGC 63 401 15 GTATCCTGGACCTATGGTGTGGT TGGTCCAGACTTGGTTATGGA 58 373 MAPPING AND STRUCTURE OF HUMAN AND MOUSE FAAH GENES 413

have identified this transmembrane region of FAAH as an autonomously functioning homo-oligomerization domain (Patricelli et al., 1998). Degradation of neuromodulatory fatty acid amides by FAAH may have distinct physiological effects in the CNS, and mutations in the FAAH gene could poten- tially cause inheritable neurological disorders in mammals. The 1p36–p34 region of human chromosome 1p, where the FAAH gene was mapped, carries genes for human neurogenetic diseases, such as Schwartz– Jampel syndrome (Fontaine et al., 1996; Singh et al., 1997; Timmerman et al., 1996), hereditary congenital ptosis 1 (Engle et al., 1997), and Charcot–Marie–Tooth neuropathy type 2A (Saito et al., 1997). Several neurological mouse mutants, audiogenic seizure prone-2 (asp2), cerebellar folial pattern (Cfp1), cribriform de- FIG. 4. Western blot of total liver extract (50 ␮g protein) from generation (cri), meander tail (mea), and clasper (cla), normal (BALB/c and Black Swiss) and clasper mice probed with have been mapped to chromosome 4 (Mouse Genome anti-FAAH antibodies. A single 65-kDa protein band was identified Database, Mouse Genome Informatics Project, The in each tissue sample. An equivalent profile was also observed in Jackson Laboratory: http://www.informatics.jax.org). brain extracts from these animals (data not shown). Molecular mass markers are given in kDa. Among the mouse mutants, cla is located most closely to the Faah gene (Mock et al., 1996). Cla is a recessive neurological mutation that arose in BALB/cByJ mice. mutant mice (normal mice do not express FAAH in the The clasper mice exhibit a whole-body tremor and heart). Taken together, these results support the idea hobbling gait with clasping of both fore- and hindfeet that FAAH levels are normal in adult clasper mice. when picked up by the tail (Sweet, 1985). Given that fatty acid amides have been shown to induce hypother- DISCUSSION mia and motor defects in mice (Mechoulam et al., 1997), it was tempting to speculate that the clasper Comparative mapping of homologous genes on hu- phenotype might represent an exaggerated reflex man and mouse chromosomes is a powerful means to caused by failure of FAAH to downregulate certain study the evolution of genome organization and to fatty acid amide-mediated activities in the CNS. How- identify candidate genes for genetic disorders. In this ever, our thorough mutation analysis and enzymatic study, we mapped the human FAAH gene to chromo- assays strongly suggest that the Faah gene is an un- some region 1p34–p35 and the mouse Faah gene to the likely candidate gene for the clasper mutation. Still, we distal portion of mouse chromosome 4, which contains cannot at this time exclude the possibility that tran- a large region of conserved synteny with human chro- sient changes in the developmental expression of mosome 1p. The genomic structures of the human and FAAH could account for the clasper phenotype. mouse FAAH genes are highly conserved, with the It is possible that mutations in the FAAH gene are same splice sites and similar sizes for both introns and involved in the development of other neurological dis- exons. The conserved chromosomal locations and orders the genes for which will be mapped to mouse genomic structures strongly suggest that the FAAH chromosome 4 or human chromosome 1p in the future. enzyme plays a similar role in different mammals. The best approach may be to create Faah mutant mice Indeed, the FAAH enzymes are well conserved in pri- by gene targeting. The phenotypes of such FAAH-defi- mary structure from rat and mouse to human (Giang cient mice may help to identify any existing neurolog- and Cravatt, 1997). Transiently expressed rat and hu- ical disorders as candidates for spontaneous FAAH man FAAHs were found to share many biochemical mutations that can be tested directly for allelism with and enzymological features (Giang and Cravatt, 1997). Faah Ϫ/Ϫ mice. The complete genomic characteriza- In addition, in situ hybridization demonstrated wide- tion of the human and mouse FAAH genes reported spread distribution of FAAH mRNA in neuronal cells here will facilitate the screening for mutations in these in the rat CNS (Thomas et al., 1997b). Interestingly, and other genetic disorders that become linked to this the present study revealed that the FAAH genes show region in the future. an exon–intron structure compatible with certain func- tional protein domains being encoded by individual ACKNOWLEDGMENTS exon units. For example, FAAH’s amidase signature sequence, which defines this family of enzymes, was We thank D. K. Giang for technical assistance and N. B. Gilula for found entirely within exon 5, the SH3-binding domain ongoing encouragement and support. This work was aided by NIH within exon 7, and the N-terminal transmembrane P01 Grants HD24234 (P. I. Daniel Glaze) and MH58542 (P. I. B.F.C.) domain within exon 1. In this regard, recent studies and NIH Postdoctoral Fellowship NS10447 (H.Z.R.) as well as by the 414 WAN ET AL.

Howard Hughes Medical Institute in which U.F. is an Investigator. served mouse/human synteny on mouse chromosome 2 and 18. The authors are grateful to Lucy Rowe for figure preparation and to Genomics 18: 667–672. Kathy Redman for expert administrative assistance. Mechoulam, R., Fride, E., Hanus, L., Sheskin, T., Bisogno, T., Di Marzo, V., Bayewitch, M., and Vogel, Z. (1997). Anandamide may REFERENCES mediate sleep induction. Nature 389: 25–26. Boehnke, M., Lange, K., and Cox, D. R. (1991). Statistical methods Mock, B. A., Stoye, J., Spence, J., Jackson, I., Eppig, J. T., Fiedorek, for multipoint radiation hybrid mapping. Am. J. Hum. Genet. 49: F. T., Jr., and Neumann, P. E. (1996). Mouse chromosome 4. 1174–1188. Mamm. Genome 6: S73–S96. Boger, D. L., Patterson, J. E., and Jin, Q. (1998). Structural require- Patricelli, M. P., Lashuel, H. A., Giang, D. K., Kelly, J. W., and ments for 5-HT2A and 5-HT1A serotonin receptor potentiation by Cravatt, B. F. (1998). Comparative characterization of a wild type the biologically active lipid oleamide. Proc. Natl. Acad. Sci. USA and transmembrane domain-deleted fatty acid amide hydrolase: 95: 4102–4107. Identification of the transmembrane domain as a site for oligomer- Bray-Ward, P., Menninger, J., Lieman, J., Desai, T., Mokady, N., ization. Biochemistry, in press. Banks, A., and Ward, D. C. (1996). Integration of the cytogenetic, Rowe, L. B., Nadeau, J. H., Turner, R., Frankel, W. N., Letts, V. A., genetic, and physical maps of the human genome by FISH map- Eppig, J. T., Ko, M. S. H., Thurston, S. J., and Birkenmeier, E. H. ping of CEPH YAC clones. Genomics 32: 1–14. (1994). Maps from two interspecific backcross DNA panels avail- Cravatt, B. F., Prospero-Garcia, O., Siuzdak, G., Gilula, N. B., Hen- able as a community genetic mapping resource. Mamm. Genome 5: rikse, S. J., Boger, D. L., and Lerner, R. A. (1995). Chemical 253–274. characterization of a family of brain lipids that induce sleep. Sci- Saito, M., Hayashi, Y., Suzuki, T., Tanaka, H., Hozumi, I., and Tsuji, ence 268: 1506–1509. S. (1997). Linkage mapping of the gene for Charcot–Marie–Tooth Cravatt, B. F., Giang, D. K., Mayfield, S. P., Boger, D. L., Lerner, disease type 2 to chromosome 1p (CMT2A) and the clinical features R. A., and Gilula, N. B. (1996). Molecular characterization of an of CMT2A. Neurology 49: 1630–1635. enzyme that degrades neuromodulatory fatty-acid amides. Nature Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989). “Molecular 384: 83–87. Cloning. A Laboratory Manual,” 2nd ed., Cold Spring Harbor Lab- Devane, W. A., Hanus, L., Breuer, A., Pertwee, R. G., Stevenson, L. A., oratory Press, Cold Spring Harbor, NY. Griffin, G., Gibson, D., Mandelbaum, A., Etinger, A., and Mechoulam, Singh, B., Biary, N., Jamil, A. A., and al-Shahwan, S. A. (1997). R. (1992). Isolation and structure of a brain constituent that binds to Schwartz–Jampel syndrome: Evidence of central nervous system the cannabinoid receptor. Science 258: 1946–1949. dysfunction. J. Child Neurol. 12: 214–217. Engle, E. C., Castro, A. E., Macy, M. E., Knoll, J. H. M., and Beggs, Skoda, R. C., Seldin, D. C., Chiang, M. K., Peichel, C. L., Vogt, T. F., A. H. (1997). A gene for isolated congenital ptosis maps to a 3-cM and Leder, P. (1993). Murine c-mpl: A member of the hematopoi- region within 1p32–p34.1. Am. J. Hum. Genet. 60: 1150–1157. etic growth factor receptor superfamily that transduces a prolifer- Fontaine, B., Nicole, S., Topaloglu, H., Ben Hamida, C., Beighton, P., ative signal. EMBO J. 12: 2645–2653. Spaans, F., Cantu, J. M., Bakouri, S., Romero, N., Ricker, K., Smith, P. B., Compton, D. R., Welch, S. P., Razdan, R. K., Mechou- Barros-Nunez, P., Ponsot, G., Ben Hamida, M., Weissenbach, J., lam, R., and Martin, B. R. (1994). The pharmacological activity of Hentati, F., and Lehmann-Horn, F. (1996). Recessive Schwartz– anandamide, a putative endogenous cannabinoid, in mice. J. Phar- Jampel syndrome (SJS): confirmation of linkage to chromosome macol. Exp. Ther. 270: 219–227. 1p, evidence of genetic homogeneity and reduction of the SJS locus to a 3-cM interval. Hum. Genet. 98: 380–385. Sugano, K., Kyogoku, A., Fukayama, N., Ohkura, H., Shimosato, Y., Francke, U. (1994). Digitized and differentially shaded human chro- Sekiya, T., and Hayashi, K. (1993). Rapid and simple detection of mosome ideograms for genomic applications. Cytogenet. Cell Genet. c-Ki-ras2 gene codon 12 mutation by nonradioisotopic single- 65: 206–219. strand conformation polymorphism analysis. Methods Lab. Invest. 65: 361–366. Francke, U., Yang, F. T., Brissenden, J. E., and Ullrich, A. (1986). Chromosomal mapping of genes involved in growth control. Cold Sweet, H. O. (1985). Clasper (cla). Mouse News Lett. 73: 18. Spring Harbor Symp. Quant. Biol. 51: 855–866. Timmerman, V., De Jonghe, P., Spoelders, P., Simokovic, S., Lofgren, Giang, D. K., and Cravatt, B. F. (1997). Molecular characterization of A., Nelis, E., Vance, J., Martin, J. J., and Van Broeckhoven, C. human and mouse fatty acid amide hydrolases. Proc. Natl. Acad. (1996). Linkage and mutation analysis of Charcot–Marie–Tooth Sci. USA 94: 2238–2242. neuropathy type 2 families with chromosomes 1p35–p36 and Xq13. Guan, X., Cravatt, B. F., Ehring, G. R., Hall, J. E., Boger, D. L., Neurology 46: 1311–1318. Lerner, R. A., and Gilula, N. B. (1997). The sleep-inducing lipid Thomas, E. A., Carson, M. J., Neal, M. J., and Sutcliffe, J. G. (1997a). oleamide deconvolutes gap junction communication and calcium Unique allosteric regulation of 5-hydroxytryptamine receptor-me- wave transmission in glial cells. J. Cell Biol. 139: 1785–1792. diated signal transduction by oleamide. Proc. Natl. Acad. Sci. USA Huidobro-Toro, J. P., and Harris, R. A. (1996). Brain lipids that 94: 14115–14119. induce sleep are novel modulators of 5-hydroxytrypamine recep- Thomas, E. A., Cravatt, B. F., Danielson, P. E., Gilula, N. B., and tors. Proc. Natl. Acad. Sci. USA 93: 8078–8082. Sutcliffe, J. G. (1997b). Fatty acid amide hydrolase, the degrada- Lerner, R. A., Siuzdak, G., Prospero-Garcia, O., Henriksen S. J., tive enzyme for anandamide and oleamide, has selective distribu- Boger, D. L., and Cravatt, B. F. (1994). Cerebrodiene: A brain lipid tion in neurons within the rat central nervous system. J. Neurosci. isolated from sleep-deprived cats. Proc. Natl. Acad. Sci. USA 91: Res. 50: 1047–1052. 9505–9508. Walter, M. A., Spillet, D. J., Thomas, P., Weissenbach, J., and Good- Li, X., Pereira, L., Zhang, H., Sanguineti, C., Ramirez, F., Bonadio, fellow, P.N. (1994). A method for constructing radiation hybrid J., and Francke, U. (1993). Fibrillin genes map to regions of con- maps of whole genomes. Nature Genet. 7: 22–28.