Attention Deficit Hyperactivity Disorder and the Gene for Dopamine Beta-Hydroxylase

Recent molecular genetic studies have reported linkage of several genes to attention deficit hyperactivity disorder (ADHD), including the gene for dopamine beta hydroxylase (DBH). Daly et al. (1) observed an association of allele 2 (TaqI site present, allele T) of the TaqI polymorphism located in intron 5 in a sample that consisted of 105 probands and 191 parents collected in Ireland (haplotype-based haplotype relative risk χ²=9.0, df=1, p=0.003; risk ratio=1.31, 95% confidence interval [CI]=1.09–1.56).

The level of DBH in the plasma and CSF is a stable, heritable trait, and the DBH gene has been shown to be the major locus influencing this level (2). The alleles of several polymorphisms identified for the DBH locus have been found to be associated with serum DBH levels, suggesting that these alleles are in linkage disequilibrium with a DNA variant controlling the function or expression of this gene. Studies have found that the alleles of the (CA)n repeat polymorphism and the 19 base-pair insertion/deletion polymorphism located in the region 5′ to the transcription start site are associated with either high or low levels of DBH in the serum (2, 3). For the 19 base-pair insertion/deletion polymorphism, the insertion allele has been reported to be associated with high serum levels and the deletion with low serum levels (2). Individuals with a heterozygous genotype have been found to have an intermediate level. For the (CA)n repeat, there are two common alleles, A3 and A4. The A4 allele has been found to be associated with high levels of serum DBH, the A3 allele with low, and the heterozygous genotype with intermediate levels (3).

In this study, we sought to replicate the work of Daly et al. (1) by studying 117 families with a proband who had ADHD. The study group consisted of 92 families with both parents genotyped, 25 families with a single parent genotyped, and 33 affected siblings of the probands for a total of 150 children with ADHD. We tested for linkage to the DBH gene using the same TaqI polymorphism as Daly et al. In addition, we tested the (CA)n repeat and 19 base-pair polymorphisms associated with DBH serum levels as discussed. Both polymorphisms are located approximately 4.7 kilobases (kb) 5′ to the transcriptional start site (∼6 kb from the intron 5 polymorphism).

Method

The assessment and characteristics of the subjects for this study have been described previously (4). This protocol was approved by the Hospital for Sick Children’s Research Ethics Board, and written informed consent was obtained for all participants.

The intron 5 TaqI polymorphism was genotyped as previously described (1). The second marker we used contains two polymorphisms—a (CA)n repeat and a 19-base-pair insertion/deletion—and was genotyped as previously described (5). For the analysis the two polymorphisms were considered separately. We classified the alleles for the (CA)n repeat according to the designation of Wei et al. (3).

The transmission/disequilibrium test (TDT) statistic was calculated with the extended TDT program (6). Because we used the genotypes from affected siblings in the analysis, the test is more accurately described as a test for linkage as opposed to a test for association, and we use the term “linkage” throughout the manuscript to describe our results. The TRANSMIT program (7) was used to analyze the transmission of the haplotypes. Association between the markers was estimated with the EH program (8).

Results

The allele frequencies in the parental chromosomes were 0.480 for allele 1 (C) and 0.520 for allele 2 (T) for the TaqI polymorphism and 0.565 for the 19 base-pair insertion and 0.435 for the deletion. For the DBH (CA)n repeat the allele frequencies were 0.012, 0.070, 0.340, 0.571, and 0.007 for alleles A1 through A5, respectively. The haplotype frequencies are shown in Table 1. We observed significant association of the genotypes of the three polymorphisms (χ²=72.33, df=8, p=2 ×10^–12).

The transmission/disequilibrium test results were not significant for the biased transmission of the alleles of the TaqI polymorphism; however, the biased transmission of allele 2 (transmitted 68 times compared with 52 times not transmitted) was also nonsignificant (TDT χ²=2.13, df=1, p=0.07, one-sided). We used the one-sided test because there was an a priori hypothesis for the transmission of this allele. Biased transmission was not observed for either allele of the 19 base-pair polymorphism (TDT χ²=0.29, df=1, p=0.59) or for either of the two common alleles, A3 (TDT χ²=0.09, df=1, p=0.77) or A4 (TDT χ²=0.03, df=1, p=0.85), of the (CA)n repeat. We also found no significant evidence for biased transmission of the haplotypes (Table 1).

Discussion

The biased transmission of allele 2 of the TaqI polymorphism, as previously reported, failed to reach statistical significance. We studied a slightly larger group of families than Daly et al. (1), and therefore we should presumably have had sufficient power to detect linkage if this gene has the same effect size in our study group. However, differences in the composition of the group because of variation in ascertainment, in diagnostic assessment, and particularly in ethnicity could influence power to detect linkage.

Although our results for the TaqI polymorphism support the previous finding, it is puzzling that we did not find any evidence for linkage with the other two polymorphisms. Because the alleles of these polymorphisms have been shown to be associated with serum DBH levels, it would be expected that the phenotype would link to either the high or the low alleles or to either of the two common haplotypes containing these alleles. One explanation is that the phenotype is not related to DBH activity but, instead, that the TaqI*2 allele is in linkage disequilibrium with another functional change contributing to the phenotype, such as a null allele or an allele that changes the expression of the gene. Alternatively, the finding for the biased transmission of the TaqI*2 allele may be a chance finding from multiple tests. Further experiments would be necessary to determine an explanation.

TABLE 1

Enlarge table