Human Molecular Genetics

East Anglian Medical Genetics Service Molecular Genetics Laboratory, Box 158, Addenbrooke's NHS Trust, Hills Road, Cambridge, CB2 2QQ, UK and Departments of ¹Medical Genetics and ²Psychiatry, Cambridge University, Addenbrooke's Hospital, Hills Road, Cambridge, CB2 2QQ, UK

Received January 6, 1996; Revised and Accepted March 13, 1996

The monoamine oxidase A locus (MAOA) at Xp11 was considered a good candidate to investigate in bipolar affective disorder since this enzyme plays an important role in the degradation of various neurotransmitters and a mutation in this gene has been associated with borderline mental retardation and a behavioural phenotype that has some resemblance to the manic syndrome. Previous association studies comparing allele frequencies of a microsatellite and RFLP at the monoamine oxidase A locus in bipolar affective disorder cases and controls in the UK have yielded conflicting results: Lim and colleagues reported a positive association, while no evidence for allelic association was obtained by Cradock and co-workers. A significant allelic association was observed between Japanese bipolar cases and controls at the MAOA microsatellite but different alleles seemed to be overrepresented in the bipolar cases in this population compared to the UK. In order to resolve these differences, we have examined this locus in our series of unrelated bipolar cases and age- and sex-matched controls and found significantly different MAOA microsatellite allele frequencies. In addition, we have pooled the data from the two previous UK studies with ours to create a total data set including 67 males and 113 females with bipolar affective disorder and a similar number of matched controls. No evidence for heterogeneity was observed for the control MAOA microsatellite or RFLP allele frequencies in these three studies. However, we found a significant difference between the pooled normal and bipolar allele frequencies both for the microsatellite and the RFLP at MAOA.

INTRODUCTION

Bipolar affective disorder is a psychiatric condition characterised by episodes of depression and mania which has a prevalence of 1-2%. Although much data support a genetic component in the etiology of bipolar affective disorder, the mode(s) of inheritance are unclear and the specific genes associated with predisposition to this condition have not been characterised (1 ).

The monoamine oxidase A locus (MAOA) at Xp11 was considered a good candidate to investigate in bipolar affective disorder, since this enzyme plays an important role in the degradation of the neurotransmitters norepinephrine, epinephrine, dopamine and serotonin (2 ). Furthermore, a nonsense mutation in this gene has been associated with borderline mental retardation and a behavioural phenotype that has some resemblance to the manic syndrome (3 ).

Association studies comparing allele frequencies of a microsatellite and RFLP at the the monoamine oxidase A locus in bipolar affective disorder cases and controls in the UK have yielded conflicting results: Lim and colleagues (2 ) reported a positive association, while no evidence for allelic association was obtained by Cradock and co-workers (4 ). A significant allelic association was observed between Japanese bipolar cases and controls at the MAOA microsatellite but different alleles seemed to be overrepresented in the bipolar cases (5 ) in this population compared to the UK (2 ).

In order to resolve these differences, we have examined this locus in our series of unrelated bipolar cases and age- and sex-matched controls and found significantly different MAOA microsatellite allele frequencies at this locus. In addition, we have pooled the data from the two previous UK studies (2 ,4 ) with ours. No evidence for heterogeneity was observed for the control MAOA microsatellite or RFLP allele frequencies from these three studies. However, we found a significant difference between the pooled control and bipolar allele frequencies both at the microsatellite and the RFLP loci.

RESULTS

Allele frequencies of the microsatellite in intron 2 of MAOA (2 ) were initially compared in 39 bipolar cases from the Cambridge region and age and sex-matched controls (Table 1 ). There are at least nine alleles at this locus. All but three alleles were rare and resulted in expected values of below 5 for their respective cells in a [chi]² contingency table. Accordingly, alleles a₂, a₅ and a₆ were analysed in separate cells, while the fourth cell contained the remaining rare alleles from the distribution. Thus, we were left with a 4 * 2 contingency table. A significant deviation from the normal allele distribution was observed for the bipolar patients compared to the controls ([chi]² = 9.1, 3 d.f., p = 0.028). However, no significant differences were observed when the sexes were examined separately. Our data provide strong support for the original suggestion of Lim and colleagues (2 ), since related alleles (a₅ and a₆) showed a trend towards over-representation in the bipolar cases, while allele a₂ tended to be over-represented in the controls.

In order to resolve the previously described discrepancies between association studies of bipolar affective disorder with this locus, we pooled the data from the two previous UK studies (2 ,4 ) with ours. The descriptions of these populations are summarised in the Methods section. The Japanese data were excluded for three reasons. First, Japanese allele frequencies are likely to differ from those in the UK, as suggested by the data of Kawada et al. (5 ). Second, the Japanese observed a different pattern of allelic association with bipolar affective disorder at this locus (5 ). Third, the Japanese study did not report allele distributions by sex (5 ).

The control allele distributions at this microsatellite locus were compared across the UK populations in order to confirm the rationale for pooling the data of the three studies [the data can be reconstructed by comparing our control data to those of Cradock et al. (4 ) and Lim et al. (2 )]. For the reasons outlined above, alleles a₂, a₅ and a₆ were analysed in separate cells, while the fourth cell contained the remaining rare alleles from the distribution. Although the studies from Cardiff (4 ) and London (2 ) did not indicate the microsatellite allele sizes but only the allele names (a₁, a₂ etc.), we were able to confidently identify which alleles corresponded to a₂, a₅ and a₆ and pool microsatellite allele data of the controls and cases from our study with those previously published data, for the following reasons. First, since the allele distribution at this locus is asymmetrical and bimodal (see Table 1 and refs 2 and 4 ), it was apparent that all three laboratories were using the same naming convention by calling the largest alleles a₀. Second, the control allele distributions at this microsatellite locus were compared across the three UK populations in order to confirm the rationale for pooling the data of the three studies [the data can be reconstructed by comparing our control data to those of Cradock et al. (4 ) and Lim et al. (2 )]. Again, for the reasons outlined above, alleles a₂, a₅ and a₆ were analysed in separate cells, while the fourth cell contained the remaining rare alleles from the distribution. No significant difference was observed for the allele distributions of the total normal populations ([chi]² = 3.2, 6 d.f., p = 0.8). In addition, no differences were detected when these data were decomposed by sex (males: [chi]² = 0.87, 6 d.f., p = 0.99; females: [chi]² = 6.3, 6 d.f., p = 0.39). Inspection of the allele distributions at this locus (Table 1 and refs 2 and 4 ) will make it obvious that any misclassification of alleles a₂, a₅ and a₆ would have been revealed by significant differences in these control allele frequencies from London, Cambridge and Cardiff. These data also suggest that the three populations that were pooled were relatively homogeneous.

Significant allelic association with bipolar affective disorder was observed at the MAOA microsatellite for the total combined sample ([chi]² = 13.8, 3 d.f., p = 0.003) (Table 2 ). A similar effect was also observed for the females ([chi]² = 16.4, 3 d.f., p = 0.0009) but not for the males ([chi]² = 0.93, 3 d.f., p = 0.8).

Table 1 Allele distributions at the MAOA microsatellite and RFLP loci in bipolar affective disorder patients and controls from Cambridge (18 males and 21 females age- and sex-matched)

Microsatellite	Bipolar	Controls	Bipolar	Controls	Bipolar	Controls
Allele size (bp)	(total)	(total)	(males)	(males)	(females)	(females)
126	1 (0.02)	3 (0.05)		2 (0.11)	1 (0.02)	1 (0.02)
124	1 (0.02)	1 (0.02)			1 (0.02)	1 (0.02)
122 (a2)	6 (0.10)	12 (0.20)	1 (0.06)	2 (0.11)	5 (0.12)	10 (0.24)
120
118		2 (0.04)				2 (0.05)
116 (a5)	10 (0.17)	9 (0.15)	3 (0.17)	2 (0.11)	7 (0.17)	7 (0.17)
114 (a6)	41 (0.68)	28 (0.47)	14 (0.78)	10 (0.56)	27 (0.64)	18 (0.43)
112		5 (0.08)		2 (0.11)		3 (0.07)
106	1 (0.02)				1 (0.02)
RFLP allele
1	47 (0.79)	40 (0.67)	17 (0.94)	15 (0.83)	30 (0.71)	25 (0.60)
2	13 (0.21)	20 (0.33)	1 (0.06)	3 (0.17)	12 (0.29)	17 (0.40)

For [chi]² tests microsatellite alleles a₂, a₅ and a₆ [using nomenclature of Lim et al.(2) and Cradock et al. (4)] were examined in independent cells and the remaining rare alleles were pooled into one cell. The [chi]² value of the microsatellite alleles for the total sample was 9.1, 3 d.f., p = 0.028. Males and females alone were non-significant. Proportions of alleles in each cell are shown in parentheses.The Fnu4H1 RFLP allele frequencies in the Cambridge sample are also shown. These allele frequencies did not show significant differences between the cases and controls.

Table 2 . Allele distributions at the MAOA microsatellite and Fnu4H1 RFLP loci in bipolar affective disorder patients and controls using data from Table 1 combined with results reported by Cradock et al. (4) and Lim et al. (2)

Microsatellite	bipolar	controls	bipolar	controls	bipolar	controls
allele class	(total)	(total)	(males)	(males)	(females)	(females)
a6	177 (0.61)	144 (0.51)	41 (0.62)	44 (0.57)	136 (0.60)	102 (0.49)
a5	43 (0.15)	30 (0.11)	8 (0.12)	11 (0.14)	35 (0.15)	19 (0.09)
a2	39 (0.13)	52 (0.18)	9 (0.14)	9 (0.12)	30 (0.13)	43 (0.21)
rest	33 (0.11)	57 (0.20)	8 (0.12)	13 (0.17)	25 (0.11)	44 (0.21)
RFLP allele
1	223 (0.79)	174 (0.68)	54 (0.84)	48 (0.74)	169 (0.77)	126 (0.65)
2	61 (0.21)	83 (0.32)	10 (0.16)	17 (0.26)	51 (0.23)	66 (0.34)

For the microsatellite [chi]² tests, alleles a₂, a₅ and a₆ [using nomenclature of Lim et al. (2) and Cradock et al. (4)] were examined in independent cells and the remaining rare alleles were pooled into one cell. The [chi]² value for the total sample was 13.8, 3 d.f., p = 0.003; males [chi]² = 0.93, 3 d.f., p = 0.8; females [chi]² = 16.4, 3 d.f., p = 0.0009. The proportions of alleles in each cell are shown in parentheses.The RFLP allele frequencies were significantly different in the total sample ([chi]² = 8.08, 1 d.f., p = 0.0045) and in the females ([chi]² = 6.31, 1 d.f., p = 0.012), but no significant difference was detected in the males.

A G to T substitution at the third base of codon 941 of MAOA results in a Fnu4H1 RFLP site (6 ). This substitution does not result in a change in amino acid at this position. RFLP genotypes were determined at the MAOA locus for our sample. When the allele frequencies at this RFLP were examined in the Cambridge sample no significant differences were observed between the cases and controls (Table 1 ), although allele 1 showed a trend towards being more common in the bipolar cases, as was found by Lim et al. (2 ).

The RFLP allele frequencies in our control population were compared to the controls reported by Lim et al. (2 ) and Cradock et al. (4 ). Again, no significant differences were observed for either the total populations or males or females ([chi]² = 1.0, 1.0 and 1.6 respectively, 2 d.f., p for all cases > 0.45). We then pooled the allele frequencies from the two previous UK studies with our data (Table 2 ). [The Japanese study did not examine the MAOA RFLP (5 ).] The RFLP allele frequencies were significantly different in the total sample ([chi]² = 8.1, 1 d.f., p = 0.0045) and in the females ([chi]² = 6.31, 1 d.f., p = 0.012), but no significant difference was detected in the males.

DISCUSSION

The MAOA locus was examined in our series of bipolar affective disorder cases and controls to try to resolve the conflicting results of the two previous UK studies (2 ,4 ). Our sample showed significant allelic association for this condition at the MAOA microsatellite. When our data were combined with those from the two previous conflicting studies (2 ,4 ), significant allelic association was observed both at the MAOA microsatellite and RFLP loci. The combined data incorporate the results which argued against an allelic association at this locus (4 ) and therefore also effectively account for the only study which has doubted this possibility.

The strength of the association is weak (see Tables 1 and 2 ) and our individual data set was too small to allow confident relative risk estimates. However, Lim et al. estimated that the relative risk at the Fnu4H1 locus was 2.71 (2 ). There are at least three explanations for the low relative risk. First, a putative variant in linkage disequilibrium with the MAOA locus may be rare and only account for a small proportion of bipolar affective disorder. Second, this putative variant may only exert a small effect and have a low penetrance. Third, the variant may exert a major effect but the initial disequilibrium between it and the markers at MAOA may have been eroded over time. All of these scenarios include the possibility that the variant(s) predisposing to bipolar affective disorder may not be in the MAOA gene but closely linked. Linkage experiments would provide a complementary approach to studying this locus and may be able to resolve some of the above scenarios.

The MAOA RFLP has been associated with high MAOA activity in vitro when the FnuH1 site is present (6 ). The RFLP does not change the amino acid sequence at codon 941. Thus, it is probably in linkage disequilibrium with another functional variant. Since a significantly smaller number of bipolar patients compared to controls have this site, Lim et al. suggested that these patients may have lower MAOA levels (2 ). This possibility is tempting to consider in the context of the family with the MAOA nonsense mutation where the males showed a phenotype which had some overlap with the manic syndrome (2 ,3 ).

It is interesting that a significant effect was seen for females but not for males. Lim et al. have suggested that such a discrepancy may result if the putative mutation close to this locus results in a different phenotype from bipolar affective disorder in males (who only have one X chromosome) compared to females (2 ). Such a model can account for the observed excess of inheritance of bipolar affective disorder from females (7 ), since male carriers of this putative mutation would not be recognised as having bipolar affective disorder.

These data suggest that sequence variation in the MAOA or closely neighbouring genes contributes to susceptibility to bipolar affective disorder and serve as a stimulus to replicate these findings in other populations. Our results reflect the same general trends observed previously by Lim et al. (2 ) and a significant difference between MAOA allele frequencies remains when the results of three independent data sets from the UK are combined. Further confirmation of these data will rapidly advance searches for susceptibility loci for bipolar affective disorder.

MATERIALS AND METHODS

The bipolar affective disorder patients that were analysed in Cambridge were recruited from in-patient and out-patient clinics at a hospital in East Anglia. The sample met research diagnostic criteria (8 ) for bipolar affective disorder I. All were of English Caucasian origin. Lifetime psychopathology was assessed by trained clinicians using the SADS-L interview (9 ) supplemented by case note review. All cases were unrelated. Eighteen males and 21 females were investigated. These were age and sex-matched (within 2 years) with East Anglian individuals who were referred for the diagnosis of Huntington's disease and spinocerebellar ataxia. Ethical approval has been granted by the Addenbrooke's Hospital Ethics committee for genetic analyses of bipolar affective disorder patients.

The complete material in the meta analysis comprised our data set, 15 male and 40 females bipolar cases from the London study (2 ), 34 males and 50 females from the Cardiff study (4 ) and matched controls. All cases and controls were of Caucasian origin (2 ,4 ). The cases in the London study were recruited in London hospitals and diagnosed using the Schedule for Affective Disorders and Schizophrenia (Lifetime version) and satisfied Research Diagnostic criteria for bipolar affective disorder (2 ). These cases were matched for ethnic origin, age and sex with individuals attending a general hospital hematology clinic. The patients in the Cardiff study were recruited in the West Midlands of England and South Wales and met DSM III-R criteria for bipolar disorder. Seventy of the 84 probands also met Research Diagnostic Criteria for bipolar I disorder (4 ). Eighty four British Caucasian controls for these cases were recruited from married-in members of families seeking non-psychiatric care at the University of Wales College of Medicine Department of Medical Genetics, who were matched for socioeconomic background.

PCR analysis of the MAOA microsatellite and RFLP was performed as described (2 ). The map positions of these polymorphisms at MAOA are shown in ref. 2 . Allele sizes were compared to a sequence ladder [pGEM 3Z(+) (Promega)]. A reference individual was also run on each gel to allow additional gel-to-gel standardisation.

Statistics were performed using the SPSS/PC+ software package (SSPC, Chicago, Ill.)

ACKNOWLEDGEMENTS

S.J. was supported by a Commonwealth Fellowship.

REFERENCES

2 Lim, L.C.C., Powell, J., Sham, P., Castle, D., Hunt, N., Murray, R. and Gill, M. (1995) Evidence for a genetic association between alleles of monoamine oxidase A gene and bipolar affective disorder. Am. J. Med. Genet. 60, 325-331.

3 Brunner, H.G., Nelen, M., Breakfield, X.O., Ropers, H.H. and van Oost, B.A. (1993) Abnormal behaviour associated with a point mutation in the structural gene for monoamine oxidase A. Science 262, 578-580. MEDLINE Abstract

4 Cradock, N., Daniels, J., Roberts, E., Rees, M., McGuffin, P. and Owen, M.J. (1995) No evidence for allelic association between bipolar disorder and monoamine oxidase A gene polymorphisms. Am. J. Med. Genet. 60, 322-324.

5 Kawada, Y., Hattori, M., Dai, X.Y. & Nanko, S. (1995) Possible association between monoamine oxidase gene and bipolar affective disorder. Am J. Hum. Genet. 56, 335-336. MEDLINE Abstract

6 Hotamisligil G.S. and Breakefield X.O. (1991) Human monoamine oxidase A gene determines levels of enzyme activity. Am. J. Hum. Genet. 49, 383-392.

7 McMahon, F.J., Stine, O.C., Meyers, D.A., Simpson, S.G. and DePaulo, J.R. (1995) Patterns of maternal transmission in bipolar affective disorder. Am. J. Hum. Genet. 56, 1277-1286. MEDLINE Abstract

8 Spitzer, R.L., Endicott, J. and Robins, E. (1978) Research diagnostic criteria: rationale and reliability. Arch. Gen. Psychiatry 35, 773-782. MEDLINE Abstract

9 Endicott, J., Spitzer, R.L. (1978) A diagnostic interview: The schedule for affective disorders and schizophrenia. Arch. Gen. Psychiatry 35, 837-862. MEDLINE Abstract

*To whom correspondence should be addressed⁺ Present address: National Institute of Mental Health and Neurosciences, Bangalore, 560011, India

CiteULike

Connotea

Del.icio.us

This article has been cited by other articles:

				J. K. Pinsonneault, A. C. Papp, and W. Sadee Allelic mRNA expression of X-linked monoamine oxidase a (MAOA) in human brain: dissection of epigenetic and genetic factors Hum. Mol. Genet., September 1, 2006; 15(17): 2636 - 2649. [Abstract] [Full Text] [PDF]

				M. Preisig, F. Bellivier, B. T. Fenton, P. Baud, A. Berney, P. Courtet, P. Hardy, J. Golaz, M. Leboyer, J. Mallet, et al. Association Between Bipolar Disorder and Monoamine Oxidase A Gene Polymorphisms: Results of a Multicenter Study Am J Psychiatry, June 1, 2000; 157(6): 948 - 955. [Abstract] [Full Text]

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (42) Request Permissions Google Scholar Articles by Rubinsztein, D. C. Articles by Paykel, E. S. Search for Related Content PubMed PubMed Citation Articles by Rubinsztein, D. C. Articles by Paykel, E. S. Social Bookmarking          What's this?