Letters to the Editor
226
homozygotes. In this case, the heterozygous TG genotype shows a greater effect for a normal mentally healthy trait or to put it differently, a weaker effect for the ADHD trait. One proposed mechanism for this phenomenon is that homozygotes produce too much or too less amount of protein for optimal biological activity, whereas heterozygotes produce just the right amount of protein.10 Previous studies of SNAP-25 also indicate that underexpression or overexpression of SNAP-25 can cause disruption of normal neural functioning.11,12 This phenomenon is reported to be widespread in human genetic associations and failure to consider such effects in family-based association studies may result in false-negative findings.10 In summary, we detected a significant association of the MnlI polymorphism in our ADHD sample. Owing to our modest sample size, further investigation by other groups is required to confirm our findings. Especially, whether the ADHD patients with no co-morbid disorders are specifically associated with this SNP warrants further replication in independent samples. This study is also, to our knowledge, the first report on the association of SNAP-25 with ADHD in the East Asian population.
Acknowledgments We thank Dr Dae Hyun Cho for the genotyping and genotype analysis. We also thank Drs Chan Hyung Kim and Se Joo Kim for helpful discussions of the manuscript. This study was supported by NARSAD Young Investigator Award to JWK. JWK is a NARSAD Sidney R Baer, Jr Foundation Investigator. TK Choi1, HS Lee2, JW Kim3, TW Park4, DH Song2, KW Yook1, SH Lee1, JI Kim1 and SY Suh1 1 Department of Psychiatry, College of Medicine, Pochon CHA University, Seongnam, Korea; 2 Department of Psychiatry, College of Medicine, Yonsei University, Seoul, Korea; 3 Psychiatric & Neurodevelopmental Genetics Unit, Massachusetts General Hospital, Boston, MA, USA and 4Department of Psychiatry, College of Medicine, Chonbuk National University, Jeonju, Korea E-mail:
[email protected]
References 1 Barr CL, Feng Y, Wigg K, Bloom S, Roberts W, Malone M et al. Identification of DNA variants in the SNAP-25 gene and linkage study of these polymorphisms and attention-deficit hyperactivity disorder. Mol Psychiatry 2000; 5: 405–409. 2 Brophy K, Hawi Z, Kirley A, Fitzgerald M, Gill M. Synaptosomalassociated protein 25 (SNAP-25) and attention deficit hyperactivity disorder (ADHD): evidence of linkage and association in the Irish population. Mol Psychiatry 2002; 7: 913–917. 3 Kustanovich V, Merriman B, McGough J, McCracken JT, Smalley SL, Nelson SF. Biased paternal transmission of SNAP-25 risk alleles in attention-deficit hyperactivity disorder. Mol Psychiatry 2003; 8: 309–315. 4 Mill J, Richards S, Knight J, Curran S, Taylor E, Asherson P. Haplotype analysis of SNAP-25 suggests a role in the aetiology of ADHD. Mol Psychiatry 2004; 9: 801–810. Molecular Psychiatry
5 Hughes TA. Regulation of gene expression by alternative untranslated regions. Trends Genet 2006; 22: 119–122. 6 APA. Diagnostic and Statistical Manual of Mental Disorders (DSMIV-TR), 4th edn (text revision). American Psychiatric Association: Washington, DC, 2000. 7 Jensen PS, Hinshaw SP, Kraemer HC, Lenora N, Newcorn JH, Abikoff HB et al. ADHD comorbidity findings from the MTA study: comparing comorbid subgroups. J Am Acad Child Adolesc Psychiatry 2001; 40: 147–158. 8 Faraone SV, Biederman J, Monuteaux MC. Toward guidelines for pedigree selection in genetic studies of attention deficit hyperactivity disorder. Genet Epidemiol 2000; 18: 1–16. 9 Faraone SV, Perlis RH, Doyle AE, Smoller JW, Goralnick JJ, Holmgren MA et al. Molecular genetics of attention-deficit/ hyperactivity disorder. Biol Psychiatry 2005; 57: 1313–1323. 10 Comings DE, MacMurray JP. Molecular heterosis: a review. Mol Genet Metab 2000; 71: 19–31. 11 Raber J, Mehta PP, Kreifeldt M, Parsons LH, Weiss F, Bloom FE et al. Coloboma hyperactive mutant mice exhibit regional and transmitter-specific deficits in neurotransmission. J Neurochem 1997; 68: 176–186. 12 Owe-Larsson B, Berglund M, Kristensson K, Garoff H, Larhammar D, Brodin L et al. Perturbation of the synaptic release machinery in hippocampal neurons by overexpression of SNAP-25 with the Semliki Forest virus vector. Eur J Neurosci 1999; 11: 1981–1987.
No preferential transmission of paternal alleles at risk genes in attention-deficit hyperactivity disorder Molecular Psychiatry (2007) 12, 226–229. doi:10.1038/sj.mp.4001936
Attention-deficit hyperactivity disorder (ADHD) is a heterogeneous neurodevelopmental condition for which a significant genetic component has been suggested, although the mode of inheritance appears to be complex. Recently, Hawi et al.1 reported data supporting the presence of parent-of-origin effects for the transmission of risk alleles for nine dopaminergic and serotoninergic genes showing association with ADHD. After conducting a transmission/disequilibrium test (TDT) analysis, Hawi and colleagues observed an enhancement of the association findings when parental transmissions were analyzed separately, with association due principally to paternal overtransmission of risk alleles. To verify this hypothesis, we undertook the investigation of parent-of-origin effects in an independent sample of nuclear families from the Toronto area. The characteristics of the sample, methods of assessment and inclusion/exclusion criteria have been described previously.2,3 The number of families genotyped for each gene varied between 142 and 266 families for the different markers, dependent on the
Letters to the Editor
for paternal transmission (DRD1-D1P.6), the association results were enhanced for maternal transmission for others (SNAP25-9806 and GNAL-7595). The reasons for the discrepancies between the two studies are unclear. In most cases, the sample size and the number of informative transmissions for each marker allele tested both in the Irish and Toronto samples are similar, as well as the total number of maternal and paternal transmissions. Both studies used Diagnostical and Statistical Manual of Mental Disorders (DSM)-IV-based diagnosis procedure and the distribution of families between the DSM-IV ADHD combined, predominantly inattentive or predominantly hyperactive/impulsive subtypes was not strikingly different (Irish sample: 77–15–8%, respectively, vs Toronto sample: 62–24–14%, respectively). The boys:girls ratio is also similar (Toronto: 19% girls vs Irish: B12% girls12). Ethnic differences cannot be excluded despite both samples being mostly composed of Caucasians. The sample in Hawi et al. is almost exclusively composed of Irish families while our sample is composed of a majority of families of European descent with a small proportion of ‘other’ or ‘mixed’ backgrounds, including Native-American, African, Chinese and Indian. Although differences in sample characteristics can explain the lack of replication between samples due to different linkage disequilibrium between markers, it is hard to explain how ethnicity could explain preferential parental transmission. Beyond sample difference considerations, we have no explanation for the observed global overtransmission of paternal risk alleles in the study by Hawi et al. One point for consideration for these analyses is the use of a single marker for each gene and the choice of the ‘risk’ alleles. The identification of the true risk
stage of family collection at the time of genotyping. We included in our analysis the genes for DAT1, DRD4, DRD5, SNAP25 and 5HT1B, since they are the most replicated genes associated with ADHD and were also tested in Hawi et al.’s study.1 For DAT1, SNAP25 and DRD5, the most significant variant in our sample was different from those previously reported by others; thus, we also included the marker/allele yielding the best P-value in our sample for these genes. Association results for these five genes using the Toronto sample have been reported fully or partially previously.4–8 Note that the DRD4616 genotyping assay was redesigned and the marker reanalyzed since our original study to avoid interference in the interpretation of the results from a second polymorphism at 615. Similar to Hawi et al., we also show the results from analysis of five additional genes that demonstrated significant evidence of association in the ADHD sample from Toronto. These were the genes for dopamine receptor D1 (DRD1),9 D1-interacting protein calcyon (DRD1IP),2 a-subunit of the Gs-like protein Golf (GNAL) (Laurin et al.10), glutamate receptor 2B (GRIN2B) (Dorval et al.11) and serotonin receptor 2A (HTR2A).3 We performed a TDT analysis and analyzed transmission from mothers and fathers separately to detect any bias in parental transmission. These analyses were carried out using the extended TDT program. Table 1 shows the TDT analysis results for the total transmissions and for paternal and maternal transmissions separately. In contrast to Hawi et al.’s report,1 no clear pattern of transmission of risk alleles dependent on the sex of the transmitting parent was observed across the markers tested in our sample. Although some DNA variants showed stronger results
Table 1
227
Paternal vs maternal transmission of ADHD-associated gene alleles in a sample from Toronto
Marker
ID number
Allelea
All transmissions T
NT
Paternal transmissions
Maternal transmissions
w2
P
T
NT
w2
P
T
NT
w2
P
DRD4 (616) DRD5 (CA)n DRD5 (CA)nb DAT1 (VNTR) DAT1 (MspI)b SNAP25 (MnlII) SNAP25 (9806)b 5HTR1B (861G/C)
rs12720383 C 85 80 GDB:270167 148 bp 82 71 146 bp 16 27 rs28363170 480 bp 76 74 rs27072 G 61 36 rs3746544 T 78 78 rs6039806 C 108 70 rs6296 G 81 71
0.152 0.791 2.814 0.027 6.443 0.000 8.112 0.658
0.697 0.374 0.094 0.870 0.011 1.000 0.004 0.417
35 32 5 31 26 24 37 34
28 35 11 33 15 28 23 25
0.780 0.134 2.250 0.063 2.951 0.308 3.267 1.373
0.377 0.714 0.134 0.803 0.086 0.579 0.070 0.240
34 45 9 29 27 32 47 29
36 31 14 25 13 28 23 28
0.057 2.579 1.087 0.296 4.900 0.267 8.229 0.018
0.811 0.108 0.297 0.586 0.027 0.606 0.004 0.895
DRD1 (D1P.6) CALCYON (8721) GNAL (1961) GRIN2B (2144) HTR2A (His452Tyr)
rs265981 rs4838721 rs2161961 rs2284411 rs6314
1.939 4.846 6.444 8.308 4.414
0.164 0.027 0.011 0.004 0.036
30 25 38 38 19
15 15 29 21 10
5.000 2.500 1.209 4.898 2.793
0.024 0.112 0.272 0.027 0.095
27 28 49 43 17
26 17 25 24 10
0.019 2.689 7.784 5.388 1.815
0.891 0.101 0.005 0.020 0.178
T A A A T
74 56 101 96 37
58 35 68 60 21
Abbreviation: VNTR, variable number tandem repeat. According to the gene orientation. b Best marker or allele for this gene in our sample. a
Molecular Psychiatry
Letters to the Editor
228
variant(s) has not been determined for most of the genes tested and the functionality of those ‘risk’ alleles has not yet been thoroughly demonstrated in ADHD. As significance levels (P-values) are not necessarily good predictors of biological effects, the selection of the most significant polymorphism for a given gene may not be an optimum strategy as different markers and alleles are identified depending on the sample tested. Also, the choice of the marker within a gene could lead to different results when looking at parental transmission. For example, a careful examination of the data previously published by the same group on tryptophan hydroxylase 2 (TPH2),12 which was included in Hawi et al.’s report,1 showed that some TPH2 markers displayed preferential paternal transmission whereas other markers showed a preferential transmission of the maternal alleles. Such observations are difficult to explain because a consistent pattern of parental transmission for multiple markers within a gene would generally be expected, although complex epigenetic regulation has been demonstrated for a few genes. The protein Gsa subunit gene (GNAS), for example, uses four alternatively first exons with distinct tissue-specific promoters, two of which are expressed from the maternal chromosome and two from the paternal chromosome. Mutations inherited from the mother or the father give rise to overlapping but different phenotypes. A possible concern for segregating based on parental gender is chance findings resulting from a low number of informative transmissions and, therefore, cautious interpretation is required. For several genes in Hawi et al.’s study, the overtransmission of the risk allele was also observed for the mothers, although to a lesser extent, and it is not clear if the pattern of preferential paternal transmission would stand in a larger sample. The examination of the 861G marker allele for the 5HTR1B gene clearly illustrates this point. With a subset of our sample from Toronto (115 families), we have previously reported evidence of biased transmission for paternal alleles (TDT Ppat = 0.03; Pmat = 0.49). With an updated sample (209 families), we no longer observe this (Table 1). Similarly, according to a previous report13 on a subset of the sample from Ireland, the 5HTR1B861G allele showed stronger results for maternal transmissions (TDT: Ppat = 0.052 vs Pmat = 0.014; and HHRR: Ppat = 0.01 vs Pmat = 0.0035). Current pooled data analysis from six independent study samples supports an excess of paternal transmission for 5HTR1B-861G.14 Variation in the sample characteristics could potentially explain the discrepancy between earlier and more recent results; however, chance findings observed in smaller samples are more likely. Although a mechanism for the parent-of-origin effect observed by Hawi and colleagues has not been determined, one can hypothesize the possibility of a genomic imprinting mechanism, the silencing of one allele of a gene according to its parental origin.
Molecular Psychiatry
In the majority of documented cases, imprinted genes exist in clusters in the genome; however, as noted by Hawi et al., none of the ADHD-associated genes of their study maps to any known imprinted genomic clusters. In addition, no biological evidence exists so far to support a parent-of-origin effect for these genes (e.g., expression level differences between paternal and maternal alleles). Furthermore, no molecular evidence exists to support the involvement of genomic imprinting in ADHD. That does not exclude the possibility that epigenetic factors are important in the regulation of isolated genes, which could also be temporal and/or tissue specific. A neurodevelopmental etiology may be possible for ADHD, with imprinting effects affecting brain development leading to alterations in brain function influencing cognition and behavior. However, we believe it is premature to postulate a general pattern of preferential transmission of the paternal risk alleles of ADHD-associated genes, as this does not replicate in this independent sample and biological evidence for such a regulatory mechanism is not present at this point. N Laurin1, Y Feng1, A Ickowicz2, T Pathare2, M Malone3, R Tannock2,4, R Schachar2, JL Kennedy5 and CL Barr1,2 1 Cell and Molecular Biology Division, Toronto Western Research Institute, University Health Network, Toronto, ON, Canada; 2Department of Psychiatry, Brain and Behaviour Programme, The Hospital for Sick Children, Toronto, ON, Canada; 3 Division of Neurology, Brain and Behaviour Programme, The Hospital for Sick Children, Toronto, ON, Canada; 4Ontario Institute for Studies in Education at the University of Toronto, Toronto, ON, Canada and 5Neurogenetics Section, Centre for Addiction and Mental Health, Department of Psychiatry, University of Toronto, Toronto, ON, Canada E-mail:
[email protected]
References 1 Hawi Z, Segurado R, Conroy J, Sheehan K, Lowe N, Kirley A et al. Am J Hum Genet 2005; 77: 958–965. 2 Laurin N, Misener VL, Crosbie J, Ickowicz A, Pathare T, Roberts W et al. Mol Psychiatry 2005; 10: 1117–1125. 3 Quist JF, Barr CL, Schachar R, Roberts W, Malone M, Tannock R et al. Mol Psychiatry 2000; 5: 537–541. 4 Feng Y, Crosbie J, Wigg K, Pathare T, Ickowicz A, Schachar R et al. Mol Psychiatry 2005; 9: 9. 5 Feng Y, Wigg KG, Makkar R, Ickowicz A, Pathare T, Tannock R et al. Am J Med Genet B Neuropsychiatr Genet 2005; 4: 4. 6 Barr CL, Feng Y, Wigg KG, Schachar R, Tannock R, Roberts W et al. Am J Med Genet B Neuropsychiatr Genet 2001; 105: 84–90. 7 Barr CL, Wigg KG, Feng Y, Zai G, Malone M, Roberts W et al. Mol Psychiatry 2000; 5: 548–551. 8 Ickowicz A, Feng Y, Wigg K, Quist J, Pathare T, Roberts W et al. Am J Med Genet B Neuropsychiatr Genet 2006; 51: 325–328. 9 Misener V, Luca P, Azeke O, Crosbie J, Waldman I, Tannock R et al. Mol Psychiatry 2004; 9: 500–509.
Letters to the Editor 10 Laurin N, Ickowicz A, Pathare T, Malone M, Tannock R, Schachar R et al. J Psychiatric Res 2006; doi:10.1016/j.jpsychires. 2006.10.010. 11 Dorval KM, Wigg KG, Crosbie J, Tannock R, Kennedy JL, Ickowicz A et al. Genes Brain Behav 2006; doi:10.1111/j.1601-183X.2006. 00273.x. 12 Sheehan K, Lowe N, Kirley A, Mullins C, Fitzgerald M, Gill M et al. Mol Psychiatry 2005; 10: 944–949. 13 Hawi Z, Dring M, Kirley A, Foley D, Kent L, Craddock N et al. Mol Psychiatry 2002; 7: 718–725. 14 Smoller JW, Biederman J, Arbeitman L, Doyle AE, Fagerness J, Perlis RH et al. Biol Psychiatry 2006; 59: 460–467.
Global DNA hypomethylation and DNA hypermethylation of the alpha synuclein promoter in females with anorexia nervosa Molecular Psychiatry (2007) 12, 229–230. doi:10.1038/sj.mp.4001931
Genetic variants contribute to the pathophysiology of eating disorders such as anorexia nervosa (AN) or bulimia nervosa (BN). However, the complex interplay between genetic and environmental factors in eating disorders is only incompletely understood. The regulation of gene expression is tightly controlled and well balanced in the organism by different mechanisms such as DNA methylation and histone modifications, which are known to be disturbed under several disease conditions.1 Recent years have witnessed growing interest in epigenetic contributions to psychiatric disorders2 and evidence for epigenetic alterations has been found in different psychiatric conditions, among them schizophrenia,2 impaired stress-response3 and depression.4 Recently, we have reported genomic and gene-specific DNA hypermethylation in patients with alcohol dependence that was associated with hyperhomocysteinemia.5 Hyperhomocysteinemia is known to be present also in patients suffering from AN.6 The aim of this study was therefore to investigate possible alterations of genomic and gene-specific DNA methylation in the peripheral blood of a well-defined sample of females with eating disorders (HEaD-study).6 This being a pilot-study, we examined the promoter methylation of two genes previously described to be regulated by DNA methylation, the alpha synuclein gene (SNCA) and the homocysteine-induced endoplasmatic reticulum protein (HERP) gene.5,7
We included 46 in-patients of a specialized hospital into the study, among them 22 women with AN (mean age 26.5710.3 years; body mass index 15.972.0 kg/m2) and 24 women with BN (25.877.7 years; 22.672.6 kg/m2). Measurements were controlled against 30 age-matched healthy females (22.074.8 years; 21.773.7 kg/m2). All patients met Diagnostic and Statistical Manual of Mental Disorder (DSM) IV criteria for AN or BN. Diagnoses were established using the German version of the Structured Clinical Interview for DSM IV diagnoses. Written informed consent was obtained from all patients after the procedures had been fully explained to them. Fasting blood samples were collected on admission in ethylenediaminetetraacetic acidcontaining tubes. Homocysteine was determined by high-performance liquid chromatography using a Bio-Rad kit (BioRad, Hercules, CA, USA). Analysis of global and gene-specific DNA methylation and expression of the two genes investigated was performed as was recently described by us.5,7 Betweengroup comparisons were performed by t-tests or one-way analysis of variance (ANOVA), followed by all pairwise comparisons according to Bonferroni using SPSS for Windows 14.0 (SPSS Inc., Chicago, IL, USA). We found a global DNA hypomethylation in patients with AN (47.7719.6%; ANOVA: F = 6.67; P = 0.002) when compared to controls (65.2710.5%; Bonferroni: P = 0.001). In addition, patients with AN showed a trend towards a lower global methylation compared to patients with BN (60.4722.3%; Bonferroni: P = 0.058). There was no difference between patients with BN and controls. After dividing the participants into those with normal homocysteine and those with hyperhomocysteinemia (X11.7 mmol/l), we found a significantly lower DNA methylation in the hyperhomocysteinemia group (50.8721.0% vs 62.2718.8%; t-test: T = 2.25; df = 33.4; P = 0.031). Patients suffering from both, AN and BN had significantly lower levels of SNCA mRNA (AN: DCT 2.2072.2; BN: DCT 3.2473.2; ANOVA: F = 11.7; P < 0.001) than controls (DCT 0.1271.2). Expression of HERP did not differ between groups. We found a significantly higher DNA methylation of the SNCA gene promoter in patients with AN (86.34711.4%; ANOVA: F = 3.38; P = 0.04) compared to controls (74.06723.8%; Bonferroni: P = 0.048). Patients with BN (82.83710.0%) did not differ from patients with AN and controls. No significant differences were found in the amount of DNA methylation in the HERP promoter. To our knowledge, this is the first study reporting alterations of global and gene-specific DNA methylation in a sample of patients with eating disorders. We found a global DNA hypomethylation in females with AN, but not in those with BN. Elevated plasma homocysteine was associated with a reduced DNA methylation in peripheral blood cells, which is in line with several investigations in healthy probands.8 DNA methylation occurs by transfer of a methyl
229
Molecular Psychiatry
