Sequence and Genomic Organization of the Human G-Protein

Home , Genomic organization, GNAL

Molecular Psychiatry (2000) 5, 495–501  2000 Macmillan Publishers Ltd All rights reserved 1359-4184/00 $15.00 www.nature.com/mp ORIGINAL RESEARCH ARTICLE Sequence and genomic organization of the human G-protein Golf␣ gene (GNAL) on chromosome 18p11, a susceptibility region for bipolar disorder and schizophrenia JT Vuoristo1, WH Berrettini2, J Overhauser3, DJ Prockop4, TN Ferraro2, L Ala-Kokko1,4

1Collagen Research Unit, Biocenter Oulu and Department of Medical Biochemistry, Aapistie 7, FIN-90014 University of Oulu, Finland; 2Department of Psychiatry, University of Pennsylvania, Philadelphia, Pennsylvania 19104, USA; 3Department of Biochemistry and Molecular Pharmacology, Thomas Jefferson University, Philadelphia, Pennsylvania 19107, USA; 4Center for Gene Therapy, MCP Hahnemann University, Philadelphia, Pennsylvania 19102, USA

The sequence and genomic organization of the human Golf␣ (GNAL) gene were determined. The human GNAL gene was found to contain 12 coding exons, and it spans over 80 kb on chromosome 18p11. 5Ј RACE analysis suggested an additional transcription initiation start site. Sequence analysis of the putative promoter region revealed conserved binding sites for several transcription factors. Sequence analysis of the 3Ј-untranslated region revealed the presence of two Alu sequences and two polyadenylation signals. 3Ј RACE analysis conﬁrmed the functionality of the most downstream poly-a signal. The human GNAL was found to be expressed as a single transcript of about 5.9 kb in the brain. One highly informative dinucleotide repeat was found in intron 5. Additionally, a processed pseudogene for asparagine synthetase was found about 6 kb upstream of the GNAL gene. Knowledge of the sequence and structure of the human GNAL gene provides essential information for further analysis of the GNAL locus at chromosome 18p11 which has been linked to bipolar disorder and schizophrenia. Molecular Psychiatry (2000) 5, 495–501. Keywords: G-protein; schizophrenia; bipolar; signal transduction; gene structure; polyadenylation; expression

Introduction effectors. Such effectors include ion channels and enzymes that generate regulatory molecules or second G proteins are a large family of cytosolic proteins that messengers. Second messengers such as cAMP, in turn, bind guanine nucleotides and are coupled to trans- generate intracellular changes including protein phos- membrane receptors. They play a crucial role in the phorylation, gene transcription, cytoskeleton reorgani- transduction of many extracellular signals.1–6 The pro- zation, secretion and membrane depolarization. Ter- teins are heterotrimeric and comprised of ␣ subunits mination of the signal occurs when the intrinsic of 39–52 kDa, ␤-subunits of 35–36 kDa, and ␥-subunits GTPase activity of the ␣ subunit hydrolyses the bound of 8–10 kDa. Each of the G proteins are coupled to GTP to GDP. The ␣-subunit then re-associates with the receptors that have seven membrane spanning domains ␤␥ complex. and show considerable amino acid sequence similarity. Mammals have over 20 different ␣ subunits of G pro- Signal transduction by the receptors is initiated by tein.2 The ␣ subunit of the G protein Golf was orig- ligand binding which stabilizes an alternate confor- inally isolated during a screen of a cDNA library from mational form of the receptor and thus transmits infor- olfactory tissue of the rat.7 mRNA from the gene was mation across the cell membrane by activation of the found in olfactory neuroepithelium but not in six other heterotrimeric G protein. Interaction of the G protein tissues examined. Also, antisera to the protein was with the activated receptor promotes the exchange of bound primarily to the sensory apparatus and neurons. GDP for GTP on the ␣ subunit. The GDP/GTP exchange ␣ ␤␥ The protein was found to share extensive amino acid dissociates the -GTP complex from the hetero- ␣ ␣ ␤␥ identity with the subunit of stimulatory G protein dimer. The free and subunits each activate target 7 GS␣. In addition, the protein was shown to stimulate adenyl cyclase in a heterologous system. Subsequently, Golf␣ was found to be expressed in the basal ganglia, Correspondence: L Ala-Kokko, Dept of Medical Biochemistry, University of Oulu, Aapistie 7, FIN-90014 University of Oulu, where it apparently couples dopamine D1 receptors to 8 ␣ Finland. E-mail: leena.ala-kokkoȰoulu.fi adenyl cyclase. In further work, Golf was shown to Received 16 December 1999; revised and accepted 10 April 2000 be expressed in insulinomas as well as in testis, retina, Human GNAL gene JT Vuoristo et al 496 liver, lung and spleen.9–12 The results, therefore, sug- Version 8.0-UNIX (Genetics Computer Group) and gested that the expression was not specific for olfactory MacDNASIS program (Hitachi, Yokohama, Japan). neuroepithelium. Instead, it suggested a role for Golf␣ in a diverse population of cells in many different 3Ј RACE tissues. The human GNAL gene is located on the short 3Ј RACE analysis was used to define the functional arm of chromosome 1813 in a region that has been polyadenylation signals in the gene. About 1 ␮g of total linked to bipolar disorder and/or schizophrenia.14–18 human brain RNA, provided by Brain Bank at St Eliza- Here we have isolated 15 cosmid clones for the beth Hospital, NIMH, was reverse transcribed human GNAL gene, determined the structure of the (GeneAmp RNA PCR kit, Perkin Elmer Corp, Foster gene, its expression in human brain, and defined the City, CA, USA) using random hexamers. In the first transcription initiation site and functional polyadenyl- PCR reactions the single-stranded cDNA was amplified ation site by 5Ј RACE and 3Ј RACE, respectively. The with an oligo(dT) primer linked to a random sequence Ј results indicate the gene has 12 exons and is over 80 kb 5 -TAC TGA GCA GCG AAT TCT ACG TCG C(T)20 and in length primarily because of the large sizes of introns forward primers designed to hybridize to the 3Ј-end of 4 and 5. These results provide essential information for the gene, ie JE5-F (5Ј-CGA TGA AGG CGT GAA GGC further studies of the GNAL locus and its putative role ATG, in exon 5), JE9-F (GCT ACA ACA TGG TGA TTC in the etiology of bipolar disorder and schizophrenia. G, in exon 9), LN6 (5Ј-GCA GTA TGA GCT CTT GT, in exon 12), 14-F (5Ј-TGC AAA ATG TGT CCT GTC Ј Materials and methods TA, nt 605–624 beyond the stop codon), 14-F2 (5 -AGT TAT ACA CGT TAA TCA TC, nt 959–978), 14-F3 (5Ј- Isolation and nucleotide sequencing of the gene ACT TTG GAC ACA GGC TAC GT, nt 1213–1232), 14- The human GNAL gene was isolated from a chromo- F5 (5Ј-ACG TGT CAA TGT TTG TGT C, nt 1575–1593), some 18-specific cosmid library (Lawrence Livermore 14-F8 (5Ј-CCC TTG CAC TGT CAA AGT, nt 2160– National Laboratory, Livermore, CA, USA) with cDNA 2177), 14-F12 (5Ј-TTA CAT ACT GTA CAG ATG, nt probes covering the entire coding sequence. The cDNA 3043–3060, 14-F13 (5Ј-TCC ACT TGA TTC TAA CAT, clones were kindly provided by Zigman et al.9 Nucleo- nt 3247–3264), 14-F14 (5Ј-TGA ACA CCG CAG TCT tide sequencing was carried out either with the cosmid TAG, nt 3552–3569), 14-F15 (5Ј-TCC ATA ACA GAG clones or with subclones in a plasmid vector (pT7 ATT CAG, nt 3808–3825), 14-F16 (5Ј-AAT GTT AGG Blue, Novagen Inc, Madison, WI, USA) as templates or GAT TAC TCC, nt 4025–4042), and 14-F19 (5Ј-AAC directly from PCR products (Sequenase PCR Product CAG GCG CGG CTG GAG, nt 4658–4674). The second Sequencing Kit, US Biochemical Corp, Cleveland, OH, round amplifications were performed with a reverse USA). The nucleotide sequencing was performed with primer that was identical to the random sequence at the dideoxynucleotide method19 and modified T7 DNA the 5Ј-end of the oligo(dT) primer and to the above polymerase (Sequenase 2.0, US Biochemical Corp). nested forward primers, and the additional primer 14- Additional nucleotide sequencing was carried out by F19C (5Ј-TCT GAG ATA CTG TGA AAG CC, nt 5202– cycle sequencing of cosmid clones (dsDNA Cycle 5221). The first PCR was in a volume of 50 ␮l with Sequencing System, Gibco BRL, Rockville, MD, USA). 10 pmol of reverse primer and 10 pmol of forward Analysis of the nucleotide sequences was carried out primer at 94.5°C for 60 s, 53°C for 60 s, and 72°C for with the Wisconsin Sequence Analysis Package (GCG) 90 s for 40 cycles. About 1 ␮lof50␮l was taken for a

Figure 1 Diagram of the GNAL gene and cosmid clones analyzed here. As indicated in the text, the gene has 12 exons. A pseudogene (ASN) corresponding to a cDNA for asparagine synthetase was about 6 kb upstream of the GNAL gene. Regions from the GNAL gene that were sequenced are presented as solid lines. Cosmid clones are identiﬁed by capital letters. Sites for two rare cutting enzymes, Kpnl and Sphl, are indicated.

Molecular Psychiatry Human GNAL gene JT Vuoristo et al 497 second amplification with 10 pmol of reverse primer gene were isolated. In a second series of experiments, and 10 pmol of nested forward primer in a final volume a cDNA probe for more 5Ј sequences was employed. of 50 ␮l. The conditions were 94.5°C for 60 s, 53°C for Eight cosmids were isolated containing the 5Ј end of 60 s, and 72°C for 60 s for 35 cycles. The products of the gene. The middle region of the gene was obtained the second PCR were analyzed on agarose gels and by as a cosmid that gave weak hybridization signals with direct sequencing (Sequenase PCR Product Sequencing the two cDNA probes. The clones spanned over 100 kb Kit, US Biochemical Corp). and included all coding sequences of the gene (Figure 1). Over 32 kb of the gene were sequenced. The 5Ј RACE sequences defined twelve exons of the gene and the For 5Ј RACE analysis, pooled adapter-ligated cDNAs exon boundaries (Figure 2). The relatively large size of from human brain were used as a template (Marathon- the gene was explained by the large sizes of introns 4 Ready cDNA Kit, Clontech, Palo Alto, CA, USA). The and 5. Because of their large sizes, these two introns first PCR was performed with a forward primer AP-1 were not completely sequenced here. The exon sizes

(Marathon-Ready cDNA Kit, Clontech) designed to of the gene were very similar to the related gene GS␣ hybridize to the adapter sequence and a reverse primer with the exception that exon 1 was 6 bp longer. Also, JE4-R (5Ј-GTA TTA TAG TAC TCA TTG CTG, in exon GNAL did not contain an exon corresponding to the 21,22 4) complementary to part of exon 4. For the second alternatively spliced exon 3 of the Gs␣ gene. The round amplification another nested forward primer intron sequences analyzed included 22 Alu repeat AP-2 (Marathon-Ready cDNA Kit, Clontech) and JE1-R sequences that were widely distributed in the gene. (5Ј-TGC TTC TCG ATC TTT TTG TTG G, in exon 1) Two of the Alu repeats were in the 3Ј-untranslated primer were used. The conditions for the PCRs were region of the gene. essentially the same as in the 3Ј RACE analysis. The About 6 kb upstream of the GNAL, a processed products of the second PCR were analyzed by agarose pseudogene was found. It had a high degree of identity gels and direct sequencing. with a cDNA for the asparagine synthetase gene.23–25 About 1 kb from the 5Ј-end of the K clone and about Northern analysis 0.4 kb from the 5Ј-end of the O clone were sequenced To assay expression of GNAL, commercially supplied (Figure 1). The sequences consisted of 600 bp that filter containing mRNA from different regions of extended from nt 167 to nt 740 of the asparagine syn- human brain was used (Human Brain Multiple Tissue thetase cDNA23 followed by an Alu repeat sequence of Northern Blot, Clontech). Each lane on the filter con- about 300 bp and then an additional 250 bp extending tained about 2 ␮g of polyA-enriched RNA. A PCR-pro- to nt 989 of the cDNA. If the Alu sequences were duct from the 3Ј-untranslated region (nt 1689–2457 excluded, there was 83% identity with the reported downstream of the translation termination codon) of sequences of the cDNA for asparagine synthetase. the GNAL gene was labeled by nick translation (Nick Several stop codons were found in all three reading Translation System, Gibco BRL) with ␣32PdCTP. frames. Hybridization and washing of the filter were performed according to the manufacturer’s protocol (Clontech). Analysis of the 3Ј-untranslated region Over 5 kb for the 3Ј-untranslated region of the mRNAs Dinucleotide repeat were sequenced from the genomic clones A and B To assay for the intragenic dinucleotide repeat (gnal)20 (Figure 1). The region included two Alu sequences in intron 5, a PCR was carried out with the forward (Figure 3). They also defined two polyadenylation primer JVA (5Ј-CTC AGT CGG CTC TTA CAC ATA, nt sequences of -AATAAA- at about 4.2 and 4.5 kb −467–−446 from 3Ј-end of the intron) and the reverse beyond the termination codon for translation. Also, a primer JVB (5Ј-CAT GGA AGG TGA CAG ATG TC, nt third polyadenylation sequence of -ATTAAA- was −375–−356). The forward primer was labeled with T4 found at about 2.4 kb beyond the termination codon polynucleotide kinase (US Biochemical Corp) and [␥- and located between the two Alu repeats. In addition, 32P]ATP (DuPont NEN, Wilmington, DE, USA). The over 10 variants of the consensus polyadenylation sig- PCR was carried out in 15 ␮l with 1 pmol of each nal26 were present (not shown). To confirm that the primer. The conditions were 94.5°C for 40 s, 52°C for polyadenylation signals were used, 3ЈRACE assay was 60 s, and 72°C for 40 s with 27 cycles. performed. The total human RNA isolated from caudate nucleus was reverse transcribed using random Genbank accession numbers hexamers. The cDNA was then amplified with a series U55180, U55181, U55182, U55183, U55184. of forward primers specific for different regions of the 3Ј-untranslated region (Figure 3) and an oligo(dT) primer that contained a predefined random sequence Results at its 5Ј-end. The cDNA amplified in the first step was Structure of the GNAL gene then re-amplified with a nested forward primer and a To isolate the GNAL gene, a cosmid library specific for reverse primer with the random sequence found in the chromosome 18 was screened with cDNA probes.9 In oligo(dT) primer. Apparently because of the poly-A the initial screen with a cDNA for the 3Ј-coding sequences within the Alu repeats of the 3Ј-untranslated sequences, six cosmids containing the 3Ј end of the regions, most of the PCR products were anomalous.

Molecular Psychiatry Human GNAL gene JT Vuoristo et al 498

Figure 2 Summary of the exon sizes, intron sizes and exon/intron boundaries of the GNAL gene, in comparison with the

related Gs␣ gene are indicated. Exon sequences are presented in capital letters and intron sequences in lower case.

Molecular Psychiatry Human GNAL gene JT Vuoristo et al 499

Figure 3 3Ј RACE analysis and sequencing of the 3Ј-untranslated region of the gene. Symbols: ATG, codon for initiation of translation of the GNAL gene; TGA, signal for termination of translation of the GNAL gene; open circles, Alu repeat sequences; arrow heads, direction of the Alu repeats; vertical lines, polyadenylation signals (A, B, C) deﬁned as described in text; solid horizontal arrows, primers used for successful 3Ј RACE assay; shaded horizontal arrows, primers that did not provide products in the 3Ј RACE assay.

Figure 4 Analysis of the 5Ј-end of the GNAL gene. Symbols: Top line, sequences of the human GNAL gene defined here; second line, sequences of the rat Gnal gene; algebraic numbers above sequences, conserved consensus sites for binding of transcription factors: H1S4 US (1), ZESTEFCS (2), S1 HS (3), gammaFIREFCS (4), GCFFCS (5), AP2 CS6 (6), CAPFSITE (7), SP1 CS2 (8), JCV repeated seq (9); A, published 5Ј-end of the human GNAL cDNA; B, start of transcription as defined by 5Ј RACE carried out here; C, start of transcription as defined by primer extension for the rat mRNA (see Wang et al);27 upper case letters, coding sequences of the gene.

However, a product of the expected size was obtained Analysis of the promoter sequence with the two primers that were closest to the most dis- To identify sequences in the GNAL promoter region tal polyadenylation signal. Sequencing of the products that might be involved in regulation of transcription, demonstrated that transcription of GNAL is terminated we sequenced 3 kb DNA upstream from the ATG trans- at the most 3Ј polyadenylation signal in human brain. lation start codon and analyzed the data for conserved Thus, the predicted total length of the message is sites for the binding transcription factors (MacDNASIS about 5.9 kb. program, Hitachi) at 100% homology between the

Molecular Psychiatry Human GNAL gene JT Vuoristo et al 500 human sequences and previously published rat sequences.27 One or more conserved sites were found for nine different transcription factors (Figure 4). As was previously reported for the promoter of the rat Gnal gene,27 the promoter for the human gene did not contain a consensus CCAAT box or a consensus TATA box. The 800 bp of sequences from the human promoter region had about a 75% degree of identity with the rat promoter. Further analyses suggested that the gene had multiple start sites for transcription. A 5Ј RACE assay indicated that the major transcription initiation site in human brain was 100 nt upstream of the translation initiation codon. Assays by RT-PCR suggested that some transcripts extended even further upstream, since a product was obtained with a 5Ј- primer that was located 216 nt upstream of the start of translation. A cDNA previously isolated from a human insulinoma extended 255 nt upstream of the start of translation.9 Additional analysis of the GNAL promoter sequence was performed by Recognition of Human Poll Promoter Region and Start of Transcription28 through the WWW (http://dot.imgen.bcm.tmc.edu:9331/seq- search/gene-search.html). The sequence analysis predicted an additional transcription initiation site at 318 nt upstream of the start of translation.

Northern analysis Northern blot analysis of the gene demonstrated that GNAL is expressed as a single mRNA of about 6 kb in size in human brain. Highest level of expression was detected in caudate nucleus and amygdala (Figure 5).

Intragenic dinucleotide repeat Analysis of the genomic sequences revealed dinucleo- Figure 5 Northern analysis. Poly(A) RNA isolated from dif- tide repeat located in intron 5). The CA-repeat (gnal) ferent regions of human brain were hybridized with a GNAL in intron 5 (401 nt from the 3Ј end of the intron) was cDNA probe. The positions of size markers are indicated. found to be useful for linkage analysis because at least 11 alleles were present. The heterozygosity was 74%, and the allele frequencies were 106 nt 3%, 108 nt 17%, 110 nt 1%, 112 nt 5%, 114 nt 17%, 116 nt 40%, 118 nt untranslated region. Alu sequences are found in about 5%, 120 nt 2%, 122 nt 1%, 124 nt 17%, and 126 nt 3% 5% of expressed sequences.35 in 212 chromosomes. Together with previously published transcription initiation sites of human9 and rat Gnal,27 our 5Ј RACE results suggest that in human brain GNAL has one Discussion major transcription initiation site located at −100 nt We have isolated and determined the sequence and upstream of the translation initiation codon. As a more genomic structure of the human GNAL gene. The exon sensitive method, RT-PCR revealed that some tran- structure is very similar to the gene for the analogous scripts extend even further upstream. To identify the 21,22 G protein GS␣ except that exon 1 is 6 bp longer. cis-acting regulatory sequences that direct GNAL tran- Also, the alternatively spliced exon 3 of the GS␣ gene scription, we compared human and rat promoter was not found in the GNAL gene by complete sequen- sequences. Comparison revealed several highly con- cing of the corresponding introns. In addition, no alter- served transcription factor binding sites. natively spliced exons were found by RT-PCR (results GNAL is expressed in a wide variety of tissues not shown). The GNAL gene contained a relatively including various areas of brain. In addition to being large number of Alu repeat sequences in the introns. anosmic, mice homozygous for a null mutation in Gnal The frequency of Alu repeats was 1 per 1.45 kb which exhibit hyperactivity and inadequate maternal may increase chances for genomic rearrangements behaviors.36 These results suggest that G protein Golf␣ within the GNAL locus. The average frequency of Alu may, in addition to olfactory functions, have more roles repeats in the human genome is 1 per 4 kb,29 but some in a diverse population of cells in a variety of tissues. regions have frequencies of as high as 1 per 1 kb.30–34 Together with the northern analysis the 3Ј RACE In addition, two Alu sequences were found in the 3Ј- results of mRNA from human caudate nucleus con-

Molecular Psychiatry Human GNAL gene JT Vuoristo et al 501 firmed that GNAL is expressed in brain outside the 16 Schwab SG, Hallmayer J, Lerer B, Albus M, Borrmann M, Honig S areas responsible for olfaction. et al. Support for a chromosome 18p locus conferring susceptibility to functional psychoses in families with schizophrenia, by associ- Gnal dinucleotide repeat in intron 5 of the GNAL ation and linkage analysis. Am J Hum Genet 1998; 63: 1139–1152. gene on chromosome 18p11 has been linked to schizo- 17 Detera-Wadleigh SD, Badner JA, Berrettini WH, Yoshikawa T, Gol- phrenia in some families.16 Interestingly, deficient din LR, Turner G et al. A high-density genome scan detects evi- olfactory identification performance has been fre- dence for a bipolar-disorder susceptibility locus on 13q32 and 37 other potential loci on 1q32 and 18p11.2. Proc Natl Acad Sci USA quently reported in patients with schizophrenia. Our 1999; 96: 5604–5609. data provide a fundamental basis for functional analy- 18 Nothen MM, Cichon S, Rohleder H, Hemmer S, Franzek E, Fritze sis of the GNAL promoter and search for mutations in J et al. Evaluation of linkage of bipolar affective disorder to chromo- the human GNAL gene in patients with schizophrenia some 18 in a sample of 57 German families. Mol Psychiatry 1999; or bipolar disorder. 4: 76–84. 19 Sanger F, Nicklen S, Coulson AR. DNA sequencing with chain- terminating inhibitors. Proc Natl Acad Sci USA 1977; 74: 5463– Acknowledgements 5467. 20 Berrettini WH, Vuoristo J, Ferraro TN, Buono RJ, Wildenauer D, We wish to thank Ms Rohini Dhulipala and Ms Aila Ala-Kokko L. Human Golf gene polymorphisms and vulnerability Jokinen for excellent technical assistance. This work to bipolar disorder. Psychiatr Genet 1998; 8: 235–238. 21 Bray P, Carter A, Simons C, Guo V, Puckett C, Kamholz J et al. was supported in part by a grant from the Lucille P Human cDNA clones for four species of G␣s signal transduction Markey Charitable Trust. protein. Proc Natl Acad Sci USA 1986; 83: 8893–8897. 22 Kozasa T, Itoh H, Tsukamoto T, Kaziro Y. Isolation and characteriz- ation of the human Gs␣ gene. Proc Natl Acad Sci USA 1988; 85: References 2081–2085. 23 Andrulis IL, Chen J, Ray PN. Isolation of human cDNAs for aspara- 1 Gilman AG. G proteins: transducers of receptor-generated signals. gine synthetase and expression in Jensen rat sarcoma cells. Mol Annu Rev Biochem 1987; 56: 615–649. Cell Biol 1987; 7: 2435–2443. 2 Simon MI, Strathmann MP, Gautam N. Diversity of G proteins in 24 Greco A, Ittmann M, Basilico C. Molecular cloning of a gene that signal transduction. Science 1991; 252: 802–808. is necessary for G1 progression in mammalian cells. Proc Natl Acad 3 Spiegel AM. G proteins in cellular control. Curr Opin Cell Biol Sci USA 1987; 84: 1565–1569. 1992; 4: 203–211. 25 Zhang YP, Lambert MA, Cairney EL, Wills D, Ray PN, Andrulis 4 Neer EJ. Heterotrimeric G proteins: organizers of transmembrane IL. Molecular structure of the human asparagine synthetase gene. signals. Cell 1995; 80: 249–257. Genomics 1989; 4: 259–265. 5 Sprang SR. G protein mechanisms: insights from structural analy- 26 Sheets MD, Ogg SC, Wickens MP. Point mutations in AAUAAA sis. Annu Rev Biochem 1997; 66: 639–678. and the poly(A) addition site: effects on the accuracy and efficiency 6 Hamm HE. The many faces of G protein signaling. J Biol Chem 1998; 273: 669–672. of cleavage and polyadenylation in vitro. Nucleic Acids Res 1990; 7 Jones DT, Reed RR. Golf: an olfactory neuron specific-G protein 18: 5799–5805. involved in odorant signal transduction. Science 1989; 244: 790– 27 Wang MM, Tsai RYL, Schrader KA, Reed RR. Genes encoding 795. components of the olfactory signal transduction cascade contain a 8 Herve D, Levi-Strauss M, Marey-Semper I, Verney C, Tassin J-P, DNA binding site that may direct neuronal expression. Mol Cell Biol 1993; 13: 5805–5813. Glowinski J et al.Golf and Gs in rat basal ganglia: possible involve- 28 Soloyev V, Salamov A. The gene-finder computer tools for analysis ment of Golf in the coupling of dopamine D1 receptor with adenylyl cyclase. J Neurosci 1993; 13: 2237–2248. of human and model organisms genome sequences. Ismb 1997; 5: 9 Zigman JM, Westermark GT, LaMendola J, Boel E, Steiner DF. 294–302. ␣ 29 Moyzis RK, Torney DC, Meyne J, Buckingham JM, Wu J-R, Burks Human Golf : complementary deoxyribonucleic acid structure and expression in pancreatic islets and other tissues outside the olfac- C et al. The distribution of interspersed repetitive DNA sequences tory neuroepithelium and central nervous system. Endocrinology in the human genome. Genomics 1989; 4: 273–289. 1993; 133: 2508–2514. 30 Chen EY, Liao Y-C, Smith DH, Barrera-Saldana HA, Gelinas RE, 10 Herve D, Rogard M, Levi-Strauss M. Molecular analysis of the mul- Seeburg PH. The human growth hormone locus: nucleotide tiple Golf ␣ subunit mRNAs in the rat brain. Mol Brain Res 1995; sequence, biology, and evolution. Genomics 1989; 4: 479–497. 32: 125–134. 31 Edwards A, Voss H, Rice P, Civitello A, Stegemann J, Schwager C 11 Frayon S, Pessah M, Giroix MH, Mercan D, Boissard C, Malaisse et al. Automated DNA sequencing of the human HPRT locus. Gen- WJ et al. Galphaolf identification by RT-PCR in purified normal omics 1990; 6: 593–608. pancreatic B cells and in islets from rat models of non-insulin- 32 Carter PE, Duponchel C, Tosi M, Fothergill JE. Complete nucleotide dependent diabetes. Biochem Biophys Res Commun 1999; 254: sequence of the gene for human C1 inhibitor with an unusually 269–272. high density of Alu elements. Eur J Biochem 1991; 197: 301–308. 12 Ferrand N, Pessah M, Frayon S, Marais J, Garel JM. Olfactory recep- 33 Martin-Gallardo A, McCombie WR, Gocayne JD, FitzGerald MG, tors, Golf alpha and adenylyl cyclase mRNA expressions in the rat Wallace S, Lee BM et al. Automated DNA sequencing and analysis heart during ontogenic development. J Mol Cell Cardiol 1999; 31: of 106 kilobases from human chromosome 19q13.3. Nature Genet 1137–1142. 1992; 1: 34–39. 13 Overhauser J, Mewar R, Rojas K, Lia K, Kline AD, Silverman GA. 34 McCombie WR, Martin-Gallardo A, Gocayne JD, Fitzgerald M, Dub- STS map of genes and anonymous DNA fragments on human chro- nick M, Kelley JM et al. Expressed genes, Alu repeats and polymor- mosome 18 using a panel of somatic cell hybrids. Genomics 1993; phisms in cosmids sequenced from chromosome 4p16.3. Nature 15: 387–391. Genet 1992; 1: 348–353. 14 Berrettini WH, Ferraro TN, Goldin LR, Weeks DE, Detera-Wadleigh 35 Yulug IG, Yulug A, Fisher EMC. The frequency and position of Alu S, Nurnberger JI Jr et al. Chromosome 18 DNA markers and manic- repeats in cDNAs, as determined by database searching. Genomics depressive illness: evidence for a susceptibility gene. Proc Natl 1995; 27: 544–548. Acad Sci USA 1994; 92: 5918–5921. 36 Belluscio L, Gold GH, Nemes A, Axel R. Mice deficient in G(olf) 15 Stine OC, Xu J, Koskela R, McMahon FJ, Gschwend M, Friddle C are anosmic. Neuron 1998; 20: 69–81. et al. Evidence for linkage of bipolar disorder to chromosome 18 37 Moberg PJ, Agrin R, Gur RE, Gur RC, Turetsky BI, Doty RL. Olfac- with a parent-of-origin effect. Am J Hum Genet 1995; 57: 1384– tory dysfunction in schizophrenia: a qualitative and quantitative 1394. review. Neuropsychopharmacology 1999; 21: 325–340.

Molecular Psychiatry