American Journal of Medical Genetics Part B (Neuropsychiatric Genetics) 144B:37 –40 (2007) Further Evidence of MAO-A Gene Variants Associated With Bipolar Disorder Daniel J. Müller,1,3 Alessandro Serretti,2* Tricia Sicard,3 Subi Tharmalingam,3 Nicole King,3 Paola Artioli,4 Laura Mandelli,2 Cristina Lorenzi,4 and James L. Kennedy3** 1 Department of Psychiatry, Charite´ University Medicine Berlin, PUK der Charite´ im SHK, Charite´ Campus Mitte, Berlin, Germany Institute of Psychiatry, University of Bologna, Bologna, Italy 3 Neurogenetics Section, Centre for Addiction and Mental Health, Department of Psychiatry, University of Toronto, Toronto, ON, Canada 4 Department of Psychiatry, San Raffaele Scientific Institute, Milan, Italy 2 The aim of this study was to investigate MAOA gene variants in bipolar disorder by using a family-based association approach. The first sample included 331 nuclear families from Western and Central Canada with at least 1 offspring affected with bipolar disorder comprising a total of 1,044 individuals. All subjects were genotyped for MAOA–941T > G and uVNTR gene variants using PCR techniques. Haplotype TDT was statistically significant (LRS ¼ 12.17; df ¼ 3; P ¼ 0.0068; permutation global significance ¼ 0.00098), with the T-4 haplotype significantly associated with bipolar disorder (OR ¼ 1.63, 95% CI ¼ 1.11–2.37). Single marker analysis evidenced a borderline association for MAOA–941T > G (P ¼ 0.04), but not for the uVNTR. Pooling the Canadian sample with a second previously reported Italian sample genotyped for the uVNTR variant, negative results were obtained as well. No different results were detected when analyzing female subjects separately. In conclusion, our family-based association study gives mild but further support of the involvement of MAOA variants in bipolar disorder. ß 2006 Wiley-Liss, Inc. KEY WORDS: bipolar disorder; affective disorder; MAO-A; genetics Please cite this article as follows: Müller DJ, Serretti A, Sicard T, Tharmalingam S, King N, Artioli P, Mandelli L, Lorenzi C, Kennedy JL. 2007. Further Evidence of MAOA Gene Variants Associated With Bipolar Disorder. Am J Med Genet Part B 144:37–40. Grant sponsor: Canadian Institutes of Health Research (CIHR); Grant sponsor: CIHR postdoctoral fellowship award. *Correspondence to: Alessandro Serretti, M.D., Institute of Psychiatry, University of Bologna, Viale Carlo Pepoli 5, 40123 Bologna, Italy. E-mail: email@example.com **Correspondence to: James L. Kennedy, M.D., Neurogenetics Section, Centre for Addiction and Mental Health, Department of Psychiatry, University of Toronto, 250 College Street R30, Toronto, ON, M5T 1R8, Canada. E-mail: James_Kennedy@camh.net Received 21 March 2006; Accepted 23 May 2006 DOI 10.1002/ajmg.b.30380 ß 2006 Wiley-Liss, Inc. INTRODUCTION The gene coding for the monoamine oxidase A, the principal enzyme for the degradation of biogenic amines, is of particular interest for mood disorders [Preisig et al., 2005]. Pharmacological and genetic observations suggest that variations in either the structural or in the regulatory sequences of the MAOA gene could be associated with behavioral or physiological variability in humans [Shih et al., 1999; Shih, 2004]. MAOA-deficient mice have elevated brain levels of serotonin (5-HT), norepinephrine (NE), and dopamine (DA), which are MAO-preferred substrates, and manifest aggressive behavior [Cases et al., 1995; Shih and Chen, 1999]. The MAOA gene is located on the short arm of the X chromosome (Xp11.23) [Sabol et al., 1998] and several different polymorphisms in the MAOA gene have been reported. Two of them gained interest for their functional relevance. First, a G to T substitution at position 941 has been reported [Hotamisligil and Breakefield, 1991], with the 941 T allele associated with lower MAOA activity in human male fibroblast lines. Second, a polymorphism located 1.2 kb upstream of the MAOA coding sequence that affects the transcriptional activity of the MAOA gene promoter has been described [Sabol et al., 1998]. This gene polymorphism consists of 30 bp repeated sequence present in 3, 3.5, 4, or 5 copies. In luciferase reporter gene assays, allele copies 3.5 and 4 proved to be 2–10 times more active than the shorter one (3), while opposite findings were reported for the rarer allele copy 5 [Sabol et al., 1998; Deckert et al., 1999]. Previous reports of MAOA studies in bipolar disorder were mixed but with some evidence of association. Despite negative findings [Sasaki et al., 1998], meta analyses reported mild but significant associations for the 941 T variant [Rubinsztein et al., 1996; Furlong et al., 1999] in bipolar disorder. As for the uVNTR variant, evidence was less unequivocal with negative findings [Rubinsztein et al., 1996; Furlong et al., 1999; Kirov et al., 1999; Kunugi et al., 1999; Syagailo et al., 2001; Serretti et al., 2002], though significant sex-specific effects were observed [Deckert et al., 1999; Lin et al., 2000; Preisig et al., 2000; Schulze et al., 2000; Gutierrez et al., 2004]. The aim of this study was to investigate the possible association of the two MAOA gene variants with bipolar disorder in a large sample of nuclear families recruited in the Toronto area and Central Canada. The Canadian sample has subsequently been pooled with an Italian sample where results for the VNTR polymorphism were previously reported yielding no significant results to detect possible false negative findings due to the sample size [Serretti et al., 2002]. MATERIALS AND METHODS Sample The Canadian sample was composed of 336 patients (212 females and 124 males), affected with bipolar disorder 38 Müller et al. (n ¼ 319) and schizoaffective disorder, bipolar type (n ¼ 17), from 331 nuclear families recruited in the Toronto area and Central Canada. Patients mean age was 35.36 10.37 years with a mean age at the onset of 20.00 7.58 years. Diagnostic assessment procedures for this sample have been published in detail elsewhere [Carter et al., 2003]. Briefly, a semi-structured clinical interview (SCID-I) [First et al., 1995] was performed to assign best estimate consensus DSM-IV diagnoses. A second sample was composed by Italian patients, with probands affected by Bipolar disorder (n ¼ 309), from 99 nuclear families recruited in the area of Milan (Northern Italy). The sample was composed of 154 females (49.8%) and 155 males (50.2%), a mean age of 31.80 7.70 years and a mean age at the onset of 23.13 5.41 years. Findings on the uVNTR polymorphism of this sample were reported in the context of a previous study [Serretti et al., 2002]. Lifetime diagnoses were assigned according to DSM-IV criteria on the basis of the Structured Clinical Interview for DSM-IV, axis I disorders (SCID-I) and all available sources of information. Informed consent was obtained from all subjects after the aim of the study had been fully explained. DNA ANALYSIS Genomic DNA was obtained from peripheral leukocytes, using high salt extraction methods [Lahiri and Nurnberger, 1991]. Genotypes of the MAOA-VNTR polymorphism was determined by Polymerase Chain Reaction (PCR). PCR was performed with primers: 50 -GGACCTGGGCAGTTGTGC-30 , and 50 -CCCAGGCTGCTCCAGAAA-30 . Two and a half microliters of 20 ng/ml stock genomic DNA was used in a 20 ml reaction volume in the Canadian sample while 100 ng DNA was diluted to 5 ml in the Italian sample. Amplification was performed with an annealing temperature of 598C (Canadian sample) and 618C (Italian sample). PCR products were resolved using 3.0% high resolution plus agarose gel for 2.5 hr at 100 volts [Sabol et al., 1998]. The MAOA941T > G polymorphism was analyzed by Applied Biosystems (ABI) Taqman assay only in the Canadian sample [Hotamisligil and Breakefield, 1991]. An ABI Taqman assay was specially designed (Assay-by-design) using the following primers; forward ¼ CTTCCAGCAGAGAGAAACCAGTTAA and reverse ¼ GGCCTCCTTGTAATACATCATGCA. The assay was analyzed on the ABI7000 sequence detection instrument (Applied Biosystems, Inc., Foster City, CA). For each 10 ml reaction, we used 1 ml of 20 ng/ml DNA, 5.0 ml of Taqman reaction mix, 0.25 ml of ABI assay, and 3.75 ml of water. Statistical Analysis The transmission disequilibrium test (TDT) was applied for single marker and haplotype family-based association analysis [Spielman et al., 1993]. ETDT22 (extended TDT analysis) [Sham and Curtis, 1995] was used for single allele significance in single marker analysis and applied for X-linked analyses. Haplotype analyses were performed with the program TDTPHASE 2.403 [Dudbridge, 2003] that also includes an option for X-linked analyses. Haplotypes-analyses were performed with the COCAPHASE 2.35 program [Dudbridge, 2003]. Using a standard unconditional logistic regression, this program performs likelihood ratio tests (LRT) under a log-linear model of the probability that an haplotype belongs to the case rather than the control group; the expectation–maximization (EM) algorithm is used to resolve uncertain haplotypes and provides maximumlikelihood estimates of frequencies. Permutation analyses were used to determine empiric P levels, by using 10,000 replications. TABLE I. Haplotype Frequencies for MAOA 941T > G and VNTR Polymorphisms, in the 331 Families of Bipolar Patients (Canadian Sample) MAOA 941T > G MAOA VNTR Haplotype frequency T-4 T-3 G-4 G-3 T-5 T-3.5 G-5 G-3.5 T-2 G-2 0.4802 0.2403 0.1708 0.0856 0.0080 0.0067 0.0030 0.0024 0.0022 0.0008 We calculated the power of our sample (Canadian) to detect transmission differences for TDT analysis with an alpha level of 0.05. We had a sufficient power (0.80) to detect a genotypic relative risk of 1.81 (Aa) and 3.08 (AA) with a minimum detectable difference of frequency of 0.17 between the two groups (considering a frequency of the disease allele of 0.4, a disease prevalence of 0.01, a phenocopy rate of 0.1, a penetrance of 0.8, a codominant transmission model, and a strong LD (D-prime ¼ 0.9) between disease and marker) [Purcell and Sham, 2001; Purcell et al., 2003]. RESULTS In the Canadian sample, MAOA–941T > G and VNTR were in strong linkage disequilibrium (Chi-sq ¼ 863.85 df ¼ 4 P ¼ 0.0001; D0 ¼ 0.84), with high frequency rate of the T-4 haplotype in the sample of 331 nuclear families (Table I). Single marker TDT, performed with ETDT22, showed a positive, though marginal, association between the T allele of the 941T > G polymorphism and bipolar disorder (Chisquared for allele-wise TDT ¼ 4.11; df ¼ 1; P ¼ 0.04), while no association was detected for the VNTR polymorphism (Chisquared for allele-wise TDT ¼ 1.70; df ¼ 2; P ¼ 0.43) (Table II). Similar results were obtained using COCHAPHASE single marker analysis (941T > G: LRS ¼ 4.58; df ¼ 1; P ¼ 0.032; VNTR: LRS ¼ 2.21; df ¼ 3; P ¼ 0.53) (see Table II). Considering males and females separately, neither the VNTR nor the 941T > G polymorphism were associated to bipolar disorder (in males, respectively: P > 0.11, in females, respectively: P > 0.14), probably because of decrease in sample sizes. By the haplotype TDT analysis, the T-4 haplotype was found significantly associated with bipolar disorder (LRS ¼ 14.94; TABLE II. Individual Transmissions for MAOA 941T > G (P ¼ 0.04) and VNTR Alleles (P ¼ 0.43) in the Canadian Sample MAOA markers Ta f(T) NTb F(NT) 941T > G T G 59 54 0.52 0.48 39 75 0.34 0.66 (Chi-sq, P) (4.11, 0.04) VNTR 2 3 3.5 4 5 0 51 0 66 2 — 0.43 — 0.55 0.02 0 65 0 52 2 0.55 0.44 0.02 (1.70, 0.43) a b Transmitted allele count. Untrasmitted allele count. Bipolar Disorder and MAO-A 39 TABLE III. Transmissions of the 941T > G and VNTR Haplotype in Bipolar Patients (P ¼ 0.021) in the Canadian Sample MAOA markers 941T > G a b VNTR T-4 G-3 T-3 G-4 T-5 T-3.5 G-5 Ta f(T) NTb F(NT) Chi-sq P 165 44 22 2 1 1 0 0.702 0.187 0.094 0.009 0.004 0.004 — 139 58 24 12 0 1 1 0.591 0.247 0.102 0.051 — 0.004 0.004 6.32 2.46 0.10 8.14 1.39 0.00 1.39 0.012 0.117 0.756 0.004 0.239 1.000 0.240 Transmitted haplotype count. Untrasmitted haplotype count. df ¼ 6; P ¼ 0.021; permutation analysis ¼ 0.02) (Table III). We repeated the calculations dropping rare haplotypes (frequency <0.01) and the result was even more significant (LRS ¼ 12.17; df ¼ 3; P ¼ 0.0068; permutation analysis ¼ 0.00098). Carriers of the T-4 haplotype had a significant risk effect (OR ¼ 1.63; 95%CI ¼ 1.11–2.37). Considering exclusively females did not change the observed haplotype frequencies. The Italian sample was only genotyped for the VNTR polymorphism. According to the data found in the Canadian sample, single marker TDT did not reveal a significant association between the VNTR polymorphism and bipolar disorder (Chi-squared for allele-wise TDT ¼ 1.31 df ¼ 1 P ¼ 0.29; LRS 1.09 df ¼ 1 P ¼ 0.30, permutation analysis ¼ 0.21). Finally, merging the Italian sample with the Canadian one, no significant TDT was still observed for the VNTR polymorphism (Chi-squared for allele-wise TDT ¼ 0.43 df ¼ 2 P ¼ 0.80; LRS ¼ 0.82 df ¼ 3 P ¼ 0.84, permutation analysis ¼ 0.85) (data not shown). DISCUSSION In the present sample, we observed a marginal association between the MAOA gene and bipolar disorder with one single nucleotide polymorphism (941T > G) and with haplotype analyses using a family-based association strategy. The use of those strategies has the purpose of circumvent biases linked to the traditional case-control association studies, for example population stratification (or admixture) due to ethnic variation [Falk and Rubinstein, 1987; Spielman et al., 1993]. Consistent with functional analyses [Hotamisligil and Breakefield, 1991], our data suggest an association of increased bipolar affective disorder risk in particular with the MAOA 941 T which would be expected to correlate with relatively low levels of MAOA activity. This finding is in line and in the same direction of the one observed in previous reports in Caucasians [Rubinsztein et al., 1996; Furlong et al., 1999; Preisig et al., 2000]. Interestingly, the other polymorphism we investigated, the MAOA VNTR, presented a moderate overtransmission of the allele 4 which was more present also in panic disorder subjects [Deckert et al., 1999]. However, a number of negative articles were also reported, as stated in the introduction section as well as discrepancies in the risk haplotype. The two markers used in the present study could be considered quite informative of the MAOA gene, firstly because it has been reported that all the 90 kb length of the MAOA gene can be represented by a single block with strong LD within it [Jansson et al., 2005], then because the two polymorphism we analyzed are in strong LD and they located at approximately 70 kb from each other spanning a large part of the gene (Regarding haplotype structure, the reader is referred to www.hapmap. org or http://www.ncbi.nlm.nih.gov/SNP/GeneGt.cgi?geneID¼ 4128 in order to retrieve the latest version of haplotype interpretation of the MAOA gene). Our sample size allowed us to observe effect sizes in the medium range and this means that smaller differences were not observable. In detail, we could reliably observe a difference of about 0.17 between the two groups; therefore, the observed value of 0.11 (OR ¼ 1.63) should be considered with caution. However, single gene variants are not expected to have much smaller effects than this on complex traits; therefore, the risk of having an underpowered sample should not be considered high [Kendler, 2005]. A possible explanation of our borderline findings is that genetic influence is limited to other features that are independent from both diagnoses and symptomatology factors [Kendler, 2005]. This view is recently gaining interest following, as example, some reports of associations with antidepressant response (for a review see Serretti et al., 2005), or with neuropsychology traits [Egan et al., 2001]. Such strategies try to circumvent the complexities of psychiatric disturbances following more linear hypotheses of functional polymorphisms associated with simple traits. A more comprehensive analysis of those features should, therefore, complement future studies. Finally, family-based association studies are not free from biases. In most instances case-control studies are more powerful compared to them. For instance, if we consider that for a TDT analysis one needs to recruit not only cases but also both parents, the samples appear generally undersized. Moreover, a bias toward a selection of younger probands is implicit in the method [Khoury and Yang, 1998; Schulze et al., 2001]. In conclusion, our family-based association study gives mild but further support of the involvement of MAOA variants in mood disorders. ACKNOWLEDGMENTS This work was supported by a Canadian Institutes of Health Research (CIHR) operating grant to J.L.K. and a CIHR postdoctoral fellowship award to D.J.M. REFERENCES Carter TD, Mundo E, Parikh SV, Kennedy JL. 2003. Early age at onset as a risk factor for poor outcome of bipolar disorder. J Psychiatr Res 37: 297–303. Cases O, Seif I, Grimsby J, Gaspar P, Chen K, Pournin S, Muller U, Aguet M, Babinet C, Shih JC, et al. 1995. Aggressive behavior and altered amounts of brain serotonin and norepinephrine in mice lacking MAOA. Science 268:1763–1766. Deckert J, Catalano M, Syagailo YV, Bosi M, Okladnova O, Di Bella D, Nothen MM, Maffei P, Franke P, Fritze J, Maier W, Propping P, Beckmann H, Bellodi L, Lesch KP. 1999. Excess of high activity monoamine oxidase A gene promoter alleles in female patients with panic disorder. Hum Mol Genet 8:621–624. 40 Müller et al. Dudbridge F. 2003. Pedigree disequilibrium tests for multilocus haplotypes. Genet Epidemiol 25:115–121. Egan MF, Goldberg TE, Kolachana BS, Callicott JH, Mazzanti CM, Straub RE, Goldman D, Weinberger DR. 2001. Effect of COMT Val108/158 Met genotype on frontal lobe function and risk for schizophrenia. Proc Natl Acad Sci USA 98:6917–6922. Falk CT, Rubinstein P. 1987. Haplotype relative risks: An easy reliable way to construct a proper control sample for risk calculations. Ann Hum Genet 51:227–233. First MB, Spitzer RL, Gibbon M, Williams JB. 1995. Structured Clinical Interview for DSM-IV Axis I Disorders- Patient Edition (SCID- I/P, Version 2.0). New York: Biometrics Research Department, New York State Psychiatric Institute. Furlong RA, Ho L, Rubinsztein JS, Walsh C, Paykel ES, Rubinsztein DC. 1999. Analysis of the monoamine oxidase A (MAOA) gene in bipolar affective disorder by association studies, meta-analyses, and sequencing of the promotor. Am J Med Genet 88:398–406. Gutierrez B, Arias B, Gasto C, Catalan R, Papiol S, Pintor L, Fananas L. 2004. Association analysis between a functional polymorphism in the monoamine oxidase A gene promoter and severe mood disorders. Psychiatr Genet 14:203–208. Hotamisligil GS, Breakefield XO. 1991. Human monoamine oxidase A gene determines levels of enzyme activity. Am J Hum Genet 49:383–392. Jansson M, McCarthy S, Sullivan P, Dickman P, Andersson B, Oreland L, Schalling M, Pedersen N. 2005. MAOA haplotypes associated with thrombocyte-MAO activity. BMC Genet 6:46. Kendler KS. 2005. ‘‘A gene for. . .’’: The nature of gene action in psychiatric disorders. Am J Psychiatry 162:1243–1252. Preisig M, Ferrero F, Malafosse A. 2005. Monoamine oxidase a and tryptophan hydroxylase gene polymorphisms: Are they associated with bipolar disorder? Am J Pharmacogenomics 5:45–52. Purcell S, Sham P. 2001. Genetic Power Calculator. London: Social, Genetic & Developmental Research Centre. Purcell S, Cherny SS, Sham P. 2003. Genetic Power Calculator: Design of linkage and association genetic mapping studies of complex traits. Bioinformatics 19:149–150. Rubinsztein DC, Leggo J, Goodburn S, Walsh C, Jain S, Paykel ES. 1996. Genetic association between monoamine oxidase A microsatellite and RFLP alleles and bipolar affective disorder: Analysis and meta-analysis. Hum Mol Genet 5:779–782. Sabol SZ, Hu S, Hamer D. 1998. A functional polymorphism in the monoamine oxidase A gene promoter. Hum Genet 103:273–279. Sasaki T, Hattori M, Sakai T, Kato T, Kunugi H, Hirose T, Nanko S. 1998. The monoamine oxidase-A gene and major psychosis in Japanese subjects. Biol Psychiatry 44:922–924. Schulze T, Muller DJ, Krauss H, Scherk H, Ohlraun S, Syagailo YV, Windemuth C, Neidt H, Grassle M, Papassotiropoulos A, Heun R, Nothen MM, Maier W, Lesch KP, Rietschel M. 2000. Association between a functional polymorphism in the monoamine oxidase A gene promoter and Major Depressive Disorder. Am J Med Genet 96:801–803. Schulze TG, Muller DJ, Krauss H, Gross M, Bauer I, Fangerau-Lefevre H, Illes F, Ohlraun S, Fimmers R, Cichon S, Held T, Propping P, Nothen MM, Maier W, Rietschel M. 2001. Caught in the trio trap? Potential selection bias inherent to association studies usings parent-offspring trios. Am J Med Genet 105:351–353. Khoury MJ, Yang Q. 1998. The future of genetic studies of complex human diseases: An epidemiologic perspective. Epidemiology 9:350–354. Serretti A, Cristina S, Lilli R, Cusin C, Lattuada E, Lorenzi C, Corradi B, Grieco G, Costa A, Santorelli F, Barale F, Nappi G, Smeraldi E. 2002. Family-based association study of 5-HTTLPR, TPH, MAO-A, and DRD4 polymorphisms in mood disorders. Am J Med Genet 114:361–369. Kirov G, Norton N, Jones I, McCandless F, Craddock N, Owen MJ. 1999. A functional polymorphism in the promoter of monoamine oxidase A gene and bipolar affective disorder. Int J Neuropsychopharmcol 2:293–298. Serretti A, Artioli P, Quartesan R. 2005. Pharmacogenetics in the treatment of depression: Pharmacodynamic studies. Pharmacogenet Genomics 15:61–67. Kunugi H, Ishida S, Kato T, Tatsumi M, Sakai T, Hattori M, Hirose T, Nanko S. 1999. A functional polymorphism in the promoter region of monoamine oxidase—A gene and mood disorders. Mol Psychiatry 4: 393–395. Sham PC, Curtis D. 1995. An extended transmission/disequilibrium test (TDT) for multi-allele marker loci. Ann Hum Genet 59:323–336. Lahiri DK, Nurnberger JIJ. 1991. A rapid non-enzimatic method for the preparation of HMW DNA from blood for RFLP studies. Nucleic Acid Res 19:5444. Lin SC, Jiang SD, Wu XD, Qian YP, Wang DX, Tang GM, Gu NF. 2000. Association analysis between mood disorder and monoamine oxidase gene. Am J Med Genet 96:12–14. Preisig M, Bellivier F, Fenton BT, Baud P, Berney A, Courtet P, Hardy P, Golaz J, Leboyer M, Mallet J, Matthey ML, Mouthon D, Neidhart E, Nosten-Bertrand M, Stadelmann-Dubuis E, Guimon J, Ferrero F, Buresi C, Malafosse A. 2000. Association between bipolar disorder and monoamine oxidase A gene polymorphisms: Results of a multicenter study. Am J Psychiatry 157:948–955. Shih JC. 2004. Cloning, after cloning, knock-out mice, and physiological functions of MAO A and B. Neurotoxicology 25:21–30. Shih JC, Chen K. 1999. MAO-A and -B gene knock-out mice exhibit distinctly different behavior. Neurobiology (Bp) 7:235–246. Shih JC, Chen K, Ridd MJ. 1999. Role of MAO A and B in neurotransmitter metabolism and behavior. Pol J Pharmacol 51:25–29. Spielman RS, McGinnis RE, Ewens WJ. 1993. Transmission test for linkage disequilibrium: the insulin gene region and insulin-dependent diabetes mellitus (IDDM). Am J Hum Genet 52:506–516. Syagailo YV, Stober G, Grassle M, Reimer E, Knapp M, Jungkunz G, Okladnova O, Meyer J, Lesch KP. 2001. Association analysis of the functional monoamine oxidase A gene promoter polymorphism in psychiatric disorders. Am J Med Genet 105:168–171.