Association studies of −3081(A/T) polymorphism of norepinephrine transporter gene with attention deficit/hyperactivity disorder in Korean population^†‡

Attention-deficit/hyperactivity disorder (ADHD) is a highly heritable disorder including symptoms of inattention, hyperactivity, and impulsivity. Several lines of pharmacologic studies have suggested that its pathophysiology is involved in the alteration of the dopaminergic and noradrenergic pathways [Mikkelsen et al., 1981; Zametkin et al., 1985a,b,c; Scahill et al., 2001; Hechtman, 2005; Pliszka, 2005; Wilens, 2006]. Recent studies using the highly selective noradrenergic reuptake inhibitor, atomoxetine, demonstrate the role played by the noradrenergic system in this disorder [Michelson et al., 2001, 2002, 2003; Kratochvil et al., 2002; Thomason and Michelson, 2004]. Moreover, the latest studies reported the association with genetic markers in SLC6A2 and ADHD [Bobb et al., 2005; Kim et al., 2008; Xu et al., 2008]. The reports provide the evidence that SLC6A2 is involved in the genetic cause of ADHD. Recently, the structural–functional relationship of SLC6A2 promoter has been characterized [Kim et al., 1999, 2001, 2002]. The distal part at −4.0 kb to −3.1 kb contains a noradrenergic-specific enhancer and the proximal region at −133 bp to −75 bp appears to be critical for the noradrenergic-specific transcriptional activity of SLC6A2. We hypothesized that a polymorphism in SLC6A2 promoter region might be involved in the genetic susceptibility to the development of ADHD. We performed an association study to contrast allele and genotype frequencies for the −3081(A/T) polymorphism (rs28386840) in ADHD cases and normal controls. The frequency of the −3081(T) allele in ADHD patients was significantly higher than in controls [Kim et al., 2006]. Given that the sample size was relatively small, more extensive investigation would be warranted to test an association between ADHD and −3081(A/T) polymorphism. The objective of this study was to examine an association between ADHD and a novel polymorphism located in the promoter region of SLC6A2 [Kim et al., 2006] in Asian ethnic group including lager sample size. The association of SLC6A2 allelic variants in ADHD diagnostic subtypes was also examined to investigate whether the evidences of the association between the allelic variants of the SLC6A2 and ADHD subtype, clinical severity, or dimensional measures might be found.

A total of 103 unrelated males with ADHD and 103 normal male controls were included in this study, all of whom were Korean. They were selected from among the patients treated at the Samsung Medical Center ADHD Clinic between March 2004 and February 2007. The patients were children and adolescents, with range from 6 to 14 years, who met the DSM-IV-TR [American Psychiatric Association, 2000] criteria for ADHD, were determined by clinical assessment and semi-structured interview of Korean version of K-SADS-PL [Kaufman et al., 1997; Kim et al., 2004]. ADHD diagnostic subtypes were determined according to the DSM-IV diagnostic classification. The symptom severity and the symptom dimension were evaluated using the ADHD Rating Scale-IV—Parent version-scored scale (ADHDRS-IV-Parent) [DuPaul et al., 1998; Kim et al., 2003]. The patients were required to have at least borderline intelligence (IQ > 70) as assessed by the full KEDI-WISC-III. The exclusion criteria included any serious medical illnesses, psychosis, bipolar disorder, a history of a seizure disorder, learning disorder, and autism spectrum disorder including Asperger's disorder. The study protocol was approved by the institutional review board of Samsung Medical Center, and appropriate written informed consent was obtained from the parents of all subjects (ADHD patients and normal controls). The normal controls were volunteers randomly recruited from a middle school in Seoul. All of the control subjects were screened for apparent medical illness, ADHD, learning disorder, or mental retardation by means of a teacher's and parent's informal report, the ADHDRS-IV-teacher scale (teacher's rating score <18), and a school record. None of the control had any history of medical or psychiatric illness.

The analysis of genomic DNA samples was carried out under the conditions used in our previous study [Kim et al., 2006]. Chi-square analysis with a two-tailed P-value was used to compare the allele and genotype frequencies between the patients and controls. Fisher's exact test with the permutation method was used for multiple testings to compare the frequencies of the three types of genotypes. Allele frequencies were calculated and analyzed using the SPSS statistical package version 17.0 (Cornell University mainframe computer). The tests for Hardy–Weinberg equilibrium (HWE) using R genetic package are conducted in patient and control groups separately. The genotype and allele odds ratios were estimated by logistic regression using additive model. The genotype distribution and allele frequencies in the three subgroups of ADHD were analyzed by using chi-square test. The ANOVA was used to compare the symptom severity among the subjects with three genotypes.

The demographic characteristics of the ADHD patients are listed in Table I. Testing for deviation from HWE was performed using a Fisher's exact test (R Genetics Package) in the full sample (χ² = 0.59, P = 0.57), as well as separately in cases (χ² = 0.35, P = 0.71) and controls (χ² = 0.001, P = 1.000) indicating that there is no possibility of a deletion polymorphism or a segmental duplication caused by a mutant PCR primer site or due to a tendency to miscall heterozygotes as homozygotes. We first evaluated allele frequency between groups using a Fisher's exact test. Significant differences were found in the allele frequencies (χ² = 5.60, P = 0.02). Cochran–Armitage test was performed to apply more conservative/robust analysis and not to rely on a HWE assumption. The findings confirm the results from Fisher's exact test (χ² = 5.60, P = 0.02). The recurrence of the −3,081 T allele is higher in ADHD patients compared to controls (Table II). To further investigate that T allele is critical to make group difference, we performed another analysis using logistic regression model for the linear trend contrasting the number of T alleles. It is revealed that the frequency of T alleles in ADHD patients is significantly higher than that from control group [χ² = 5.60, P = 0.02, log odds ratio = 1.59, 95% confidence interval (CI) = 1.08–2.34] (Table II). Differences in the genotypes of ADHD cases and controls were further examined. Significant differences were found in the AA, AT, and TT genotypes (χ² = 7.1, P = 0.02). Similar to the allele-wise results, the TT genotypes were more presented in ADHD cases whereas the AT genotypes were slightly overrepresented in ADHD cases compared to control cases (Table II). Logistic analyses with T-dominant model confirmed the results that TT genotypes were overrepresented in ADHD [χ² = 26.41, P < 0.0001, log odds ratio = 4.57, 95% CI = 2.56–8.15] (Table II). We also examined the significance by using permutation tests given in our sample size. The permutation-based significance levels were comparable with standard P-values further confirming our conclusions regarding the associations between ADHD and the −3,081(A/T) SNP. From the total ADHD subjects of 103, 81 children were a combined type. Sixteen children were an inattentive type. Six children were a not otherwise specified (NOS) type (Table I). There was no significant association between genotypes and allele variants and subtypes of ADHD. Among ADHD subgroups with the three genotypes (AA, AT, and TT), there were no significant differences of total, hyperactive–impulsive, and inattentive symptom severity scores.

In this study, we examined the novel polymorphism, which was recently discovered in the promoter region of SLC6A2 [Kim et al., 2006]. Significant differences in the allele and genotype frequencies of this variant were found between the patients and controls. In particular, the odds ratio for ADHD with homozygous TT was significant. Our data provided further evidence for the existence of an association between ADHD and the A/T variants in the 5′ untranslated region.

Polymorphisms in the promoter region of the SLC6A2 may be functional or nonfunctional. It is important to note that the A/T polymorphism is a functional one. We found that the −3081T alleles significantly decreases promoter function compared with the A allele [Kim et al., 2006]. The presence of a functional polymorphism in the promoter region means that a genetic variation may cause the alteration of protein functions or the level of gene expression. Down-regulated promoter function of SLC6A2 results in decreased transcriptional activities, which may give rise to reduced level of norepinephrine transporter. The study that found a significantly reduced level of norepinephrine transporter in the locus coeruleus of major depressive subjects compared with age-matched normal control subjects has suggested that the transcriptional activities of the SLC6A2 may play an important role in the development of psychiatric illnesses [Klimek et al., 1997], though the findings have not made mention of ADHD.

Because ADHD affects multiple domains and heterogenetic populations, it is very important for the sample group to have highly homogenous characteristics in an association study. All of the subjects were of Korean descent, which is generally assumed to constitute a very homogenous ethnic group.

Although the current findings represent a significant step forward in the quest to identify the specific genetic influences on ADHD, there are limitations to this study. There was a significant difference in the mean ages of the ADHD and normal control groups. Nevertheless, the research findings reported the continuity of diagnosis or impairment of ADHD between childhood and adolescence impaired by their symptoms till adolescence [Gittelman et al., 1985; Lambert et al., 1987; Barkely, 1996; Biederman et al., 1998]. These data suggest that adolescents with ADHD from childhood may not miss the diagnosis of ADHD when they evaluate first in later stage. Therefore, the probability of such a problem that normal control group may include ADHD patients arising is very low. To our knowledge, as the first report of association between promoter region of SLC6A2 and ADHD in Asian population with homogenous ethnic group, the current study provides further insight into the role of SLC6A2 in ADHD.