Linkage study of two polymorphisms at the dopamine D3 receptor gene and attention-deficit hyperactivity disorder
Abstract
Data from animal studies suggest that the dopamine D3 receptor gene may have a role in locomotion and behavioral regulation. Therefore, this gene has been suggested as a candidate for attention-deficit hyperactivity disorder (ADHD). The dopamine D3 receptor gene (DRD3) has two common polymorphisms, one in exon I that changes a Serine to Glycine (Ser9Gly) and alters the recognition site for the restriction enzyme MscI [Lannfelt et al., 1992]. The other common polymorphism is located in intron 5 and results in the change of a restriction site for MspI [Griffon et al., 1996]. We investigated the possibility of linkage of the dopamine D3 receptor gene in 100 small, nuclear families consisting of a proband with ADHD, their parents, and affected siblings. We examined the transmission of the alleles of each of these polymorphisms and the haplotypes of both polymorphisms using the transmission disequilibrium test [Spielman et al., 1993]. We did not observe biased transmission of the alleles at either polymorphism or any haplotype. Our findings using this particular sample do not support the role of the dopamine D3 gene in ADHD. Am. J. Med. Genet. (Neuropsychiatr. Genet.) 96:114–117, 2000. © 2000 Wiley-Liss, Inc.
INTRODUCTION
The dopamine D3 receptor is an interesting candidate as a genetic susceptibility factor for ADHD. Experimental manipulation of this gene and the receptor in animals suggests that it plays a role in locomotion in animals [see, for example, Accili et al., 1996; Ekman et al., 1998]. The study of Ekman et al. [1998] used a modified oligodeoxynucleotide targeted against rat dopamine D3 receptor mRNA to decrease the levels of D3 mRNA and observed a significant increase in the spontaneous locomotor activity of rats. The creation of a mouse lacking functional D3 receptors further supports the inhibitory role of this receptor in behavior [Accili et al., 1996]. Homozygous mice lacking D3 receptors displayed increased locomotor activity in an exploratory test and increased rearing behavior. Mice heterozygous for the mutation showed similar but less-pronounced behavior [Accili et al., 1996].
It is not known if hyperlocomotion in experimental animals represents a true model of hyperactivity in children. Clearly, ADHD is more complex than simple motor behavior but other aspects of ADHD cannot be easily measured in an animal model. The molecular genetic test for linkage or association of ADHD and genes contributing to hyperlocomotion in animals is straightforward and therefore worthwhile to test specific hypotheses concerning the role of these genes in ADHD as they may contribute to the hyperlocomotor aspects of the behavior in humans.
The D3 receptor is primarily located in the nucleus accumbens and ventral striatum [Meador-Woodruff et al., 1996; Sokoloff et al., 1990], providing anatomical support for a role for D3 in motivation and motor behavior. Of further interest, abnormal involuntary movement in humans has also been associated with the D3 receptor gene in three studies [Basile et al., 1999; Segman, 1999; Steen et al., 1997]. The pharmacological investigation of D3 has been limited due to the relative paucity of D3-specific ligands. One D3 preferring agonist, pramipexole (Boehringer Ingleheim, Germany), has been reported to induce a decrease of locomotor activity in rats [Lagos et al., 1998]. In terms of ADHD specificity, there is a preliminary report of an association of the DRD3 locus and ADHD [Asherson et al., 1998].
The dopamine 3 receptor gene is composed of six exons spanning 48 kilobases (kb) and has been localized to chromosome 3q13.3 [Giros et al., 1990; Le Coniat et al., 1991; Sokoloff et al., 1990]. The Ser9Gly polymorphism appears to have functional variation [Lundstrom and Turpin, 1996]. The glycine allele has been reported to have significantly higher affinity for dopamine than the serine allele when expressed in a Chinese hamster ovary (CHO) cell line [Lundstrom and Turpin, 1996]. The D3 receptor gene has received considerable attention as a candidate for disorders thought to be the result of dysregulation of the dopamine system, in particular schizophrenia [see, for example, Asherson et al., 1996; Kennedy et al., 1995; Shaikh et al., 1996], Parkinson disease [see for example, Nanko et al., 1994] and Tourette Syndrome [see, for example, Brett et al., 1993; Comings et al., 1993; Devor et al., 1998; Hebebrand et al., 1993]. Several meta-analyses have suggested that homozygosity for the alleles at the Ser9Gly polymorphism confers some risk for schizophrenia [Dubertret et al., 1998; Shaikh et al., 1996; Williams et al., 1998] despite a number of reports not supporting this finding. Increased risk of substance abuse has also been reported to be linked to homozygosity of the alleles of this polymorphism [Krebs et al., 1998]. This holds some interest to ADHD as several studies have reported individuals with ADHD and their family members to be at an increased risk for substance abuse [Biedermann et al., 1992; Mannuzza et al., 1993].
In this study, we tested two common polymorphisms at the dopamine D3 receptor locus using the transmission disequilibrium test [Spielman et al., 1993]. We tested for biased transmission from parent to child in our sample of 100 families with an ADHD proband.
MATERIALS AND METHODS
Diagnostic Criteria
The assessment and characteristics of the subjects for this study have been previously described [Barr et al., in press]. Briefly, subjects were included if they met the following criteria for ADHD: (a) six symptoms of inattention and/or hyperactivity-impulsiveness either in the home or school setting as determined by clinical interview; (b) evidence of pervasiveness defined as a minimum of four symptoms in the non-criterion setting; (c) diagnostic decision in borderline cases included information from parent and teacher questionnaires; (d) not meeting exclusion criteria (discussed next) based on parent or teacher interview or on child assessment of anxiety or depression; (e) onset before 7 years of age.
Subjects were excluded if they scored below 80 on both the Performance and Verbal Scales of the WISC-III [Wechsler, 1991], had evidence of neurological or chronic medical illness, bipolar affective disorder, psychotic symptoms, Tourette Syndrome, chronic multiple tics, or had a comorbid anxiety, depressive, or developmental disorder that could better account for the behaviors (as specified by DSM-IV). Children were free of medication for a minimum of 24 hr before their assessment. This protocol was approved by The Hospital for Sick Children (Toronto, Ontario) Research Ethics Board and informed consent was obtained for all participants.
Isolation of DNA and Marker Typing
DNA was extracted from blood lymphocytes using a high salt extraction method [Miller et al., 1988]. The MscI (isoschizomer for BalI, New England Biolabs) polymorphism located in the first exon of DRD3 was genotyped according to Lannfelt et al. [1992] and the MspI (New England Biolabs) polymorphism in the fifth intron was genotyped as described [Griffon et al., 1996]. The restriction enzyme fragments were electrophoresed on 6% polyacrylamide mini gels (Novex) followed by silver staining to detect the alleles.
Statistical Analysis
Statistical analysis was performed using the ETDT program [Sham and Curtis, 1995]. Linkage disequilibrium between the two markers was estimated with the EH program [Ott, 1991].
RESULTS
We used for this study 100 families, consisting of 84 families with a proband and both parental DNAs genotyped and 16 families with a proband with DNA available and genotyped for a single parent. We also used genotypes from 28 siblings of the probands who also met our criteria for ADHD. Haplotypes could be determined unambiguously in 79 of these families—62 were probands with genotypes available from both parents, 16 of these were parent/child pairs, and there were 15 affected siblings.
The allele frequencies in the parental chromosomes for the MscI (Ser9Gly) polymorphism were 0.636 for the serine allele (absence of the restriction site) and 0.364 for the glycine allele (presence of the MscI restriction site). The frequency of the serine allele has been reported to vary according to the population sampled [For a summary of 10 published studies, see Shaikh et al., 1996] from between 0.12 (seen in a sample from a Congolese population) [Crocq et al., 1996] and 0.73 (seen in a sample from a Japanese population) [Nanko et al., 1993].
For the MspI polymorphism, the allele frequencies in our parental sample were 0.479 for the allele without the restriction site and 0.521 for the allele with the restriction site. These allele frequencies were similar to the published frequency (0.52 for the allele with the absence of the MspI site) for a sample of French and Alsatian ancestry populations [Crocq et al., 1996]. However, the allele frequencies observed in our sample were substantially different (0.24) from a sample taken from a Congolese population [Crocq et al., 1996].
The haplotype frequencies for the two polymorphisms were as follows: for the absence of the restriction sites at both polymorphisms, 0.266; for the absence of the restriction site at the MscI site (serine) and presence of the restriction site at MspI, 0.370; for the presence of the restriction site at the MscI site (glycine) and absence of the restriction site at MspI, 0.213, and for the presence of the restriction sites at both the MscI and MspI sites, 0.151.
The possibility of linkage disequilibrium between the genotypes of the two polymorphisms was tested using the EH program [Ott, 1991]. We observed significant disequilibrium for the distribution of the genotypes, χ2 = 3.81, df = 1, P = 0.05. This agrees with previous studies of these two polymorphisms. For example, Devor et al. [1998] reported significant evidence for linkage disequilibrium, χ2 = 20.80, df = 4, P < 0.0001), in a sample of 77 individuals. Griffon et al. [1996] also observed that the distribution of the MspI and MscI genotypes were not independent in 297 individuals, χ2 = 10.5, df = 4, P = 0.03.
We tested for biased transmission of the alleles at each polymorphism and the haplotype of the polymorphisms using the transmission disequilibrium test (TDT). The TDT test was not significant for either allele of the MscI (Table I) or MspI (Table II) polymorphisms or for the haplotypes of these two polymorphisms (Table III).
| Allele | Ser | Gly |
|---|---|---|
| Transmitted | 56 | 46 |
| Not transmitted | 46 | 56 |
| chi-squared | 0.980 | 0.980 |
| P values | 0.322 | 0.322 |
| Allele | Absence | Presence |
|---|---|---|
| Transmitted | 54 | 51 |
| Not transmitted | 51 | 54 |
| chi-squared | 0.086 | 0.086 |
| P values | 0.769 | 0.769 |
| MscI MspI | Ser absence | Ser presence | Gly absence | Gly presence |
|---|---|---|---|---|
| Transmitted | 28 | 36 | 26 | 20 |
| Not transmitted | 21 | 36 | 31 | 22 |
| chi-squared | 1.000 | 0.000 | 0.439 | 0.095 |
| P value | 0.317 | 1.000 | 0.508 | 0.758 |
DISCUSSION
Despite the possibility that the dopamine D3 receptor appears to be an interesting candidate from animal models as a genetic susceptibility factor for ADHD, our preliminary study does not suggest that this is the case. The anatomical distribution of the D3 receptor in the ventral striatum suggests that it would be more likely to mediate processes of locomotion than attention. In general, D3 receptors have a limbic distribution (e.g., nucleus accumbens) and therefore may also have a role in motivation and regulation of emotion.
We cannot rule out the possibility that the DRD3 locus is contributing to the susceptibility to ADHD in a proportion of the sample as we do not have sufficient power to detect this. According to the power calculations for TDT [Speer, 1998], given the presence of an associated allele, our sample of 100 families would have 95% power to detect a gene with relative risk of between 10.3 and 2.9 (depending on the mode of inheritance).
We have previously examined a number of candidate genes in this sample including the dopamine receptor D4 [Barr et al., in press], catechol-O-methyltransferase [Barr et al., in press], the dopamine transporter [Barr et al., personal communication], and the gene for SNAP-25 [Barr et al., personal communication]. Two of these genes have shown some evidence for linkage and are being further pursued (dopamine receptor D4 and SNAP-25). We will continue to test DRD3 as a candidate gene in our sample as the size increases during the continuation of our research.
