Prefrontal executive function and D₁, D₃, 5-HT_2A and 5-HT₆ receptor gene variations in healthy adults

Abstract

Objective

The Val158Met polymorphism of the catechol-O-methyltransferase gene has been demonstrated to be associated with prefrontal executive function explaining 4% of variance in perseverative errors on the Wisconsin Card Sorting Test (WCST). Studies suggest that dopamine D₁ and D₃ and serotonin 5-HT_2A and 5-HT₆ receptors may also be involved in prefrontal cognitive function and that genetic polymorphisms (D₁ A-48G, D₃ Ser9Gly, 5-HT_2A T102C, and 5-HT₆ T267C) of these receptors may be associated with brain glucose metabolism or neurophysiological function. The current study's objective was to investigate whether executive function varies with these genetic variations.

Methods

A sample of 216 healthy Han Chinese adults were measured with the WCST and genotyped for the 4 genetic polymorphisms.

Results

Kruskal–Wallis tests showed a significant difference in WCST perseverative errors among the genotypes D₃ Ser9Gly (p = 0.009), 5-HT_2A T102C (p = 0.038) and 5-HT₆ T267C (p = 0.010), but not in the genotype D₁ A-48G. Multiple regression analysis for the WCST natural logarithm values (i.e., for fulfilling the normal distribution requirement) showed that subjects' perseverative errors were significantly influenced by D₁ A-48G, D₃ Ser9Gly, 5-HT_2A T102C and 5-HT₆ T267C polymorphisms after adjustment of other variables.

Conclusion

The preliminary data suggest that D₁, D₃, 5-HT_2A and 5-HT₆ genetic mutations may influence prefrontal executive cognition in healthy adults. Further studies in larger samples with other ethnicities or in mentally ill patients are warranted.

Introduction

Genetic factors can influence frontal structures and related test performances.1,2 Converging evidence suggests that dopaminergic neurotransmission affects prefrontal functioning3 and other cognitive domains.4–6 Catechol-O-methyltransferase (COMT) is the key enzyme regulating dopamine metabolism within the prefrontal cortex.7 It has been demonstrated that Val158Met, a functional polymorphism in the COMT gene, is associated with prefrontal cognitive function, explaining 4% of variance in perseverative errors on the Wisconsin Card Sorting Test (WCST) in schizophrenia patients, unaffected siblings and healthy individuals.8 The WCST, a widely used measure of prefrontal cognitive functions (including executive function, abstraction and working memory), is sensitive to a person's ability to generate hypotheses, establish response sets and fluently shift sets.9,10 A recent meta-analysis11 suggests that there is a small but significant relation between Val158Met genotype and WCST performance in healthy people but not in those with schizophrenia.

Because the COMT Val158Met polymorphism accounts for a small portion of the variability in prefrontal cognition, it is warranted to explore whether other genetic mutations have potential effects. Of dopamine receptors, D₁ and D₃ may deserve foremost attention. D₁ receptors play a critical role in the activation of prefrontal cognition such as working memory.12 In the the rat medial prefrontal cortex, the D₁ receptor is also essential for selecting response in skilled nonautomatic tasks.3 In addition, D₁ activation modulates rodents' social cognition,13 which depends substantially on prefrontal function.14 Among genetic variants of D₁ receptors, the A-48G polymorphism is associated with glucose metabolic rates in human brain regions, including the prefrontal cortex.15

The D₃ receptor, although structurally highly homologous to other D₂-like dopamine receptors, differs from them in regulating neuropsychological performance.16,17 In healthy people, D₂ receptor occupancy is correlated with motor tasks rather than with domains involving prefrontal cognitive functions.6 D₂ may also regulate spatial working memory, but the latter is principally attributed to D₁ receptor–mediated mechanisms.16 In comparison, D₃ receptors mediate dopamine-related prefrontal neurocognition18 independently with D₁ activation.16 Of D₃ genetic variances, a functional polymorphism, Ser9Gly, determines eye movement disturbances,19 which are related to deficient working memory20,21 and attention.22

By interacting with the dopaminergic system, serotonergic transmission also plays a significant role in prefrontal cognitive function.23 Its action is mediated via specific receptors, possibly serotonin 5-HT_2A and 5-HT₆, located in crucial brain structures, primarily the nucleus basalis magnocellularis–frontal cortex. In rats, 5-HT_2A receptors regulate prefrontal cortex–related execution of primed responses.3 In healthy volunteers, 5-HT_2A agonists impair the continuous performance task,24 which is mainly determined by prefrontal function.25 Among 5-HT_2A receptor polymorphisms, T102C has received considerable attention. For instance, this genetic variance affects treatment response to antipsychotics.26 It also affects N100 amplitude of event-related potentials,27 which are mediated by both frontal and temporal networks.28 Although the T102C substitution does not alter amino acid sequence, recent evidence29 indicates that total levels of 5-HT_2A receptor messenger ribonucleic acid (mRNA) and protein in healthy people with the C/C genotype are lower than those in healthy people with the T/T genotype.

The 5-HT₆ receptor is responsible for endogenous 5-HT–mediated facilitation of dopamine release in the rat prefrontal cortex30 and regulates brain cholinergic neurotransmission.31 This receptor is thus involved in rodents' learning and memory function.31,32 It has been indicated that, among 5-HT₆ receptor polymorphisms,33 T267C is associated with Alzheimer disease.34 Whether this polymorphism also contributes to variances in prefrontal function requires investigation. Although this is a silent mutation, it has been suggested that this variant may affect translation through the secondary structure and stability of mRNA or that it may be in linkage disequilibrium with a functional polymorphism.35

The current study aims to explore the effects of the aforementioned genetic variances of D₁, D₃, 5-HT_2A and 5-HT₆ on WCST performance, which reflects prefrontal executive function, in a healthy population.

Methods

Participants

This study was approved by the institutional review board of China Medical University Hospital and carried out in accordance with the Declaration of Helsinki (www.wma.net /e/policy/b3.htm). The participants were 216 unrelated healthy volunteers (81 men and 135 women) with a mean age of 48.6 (standard deviation [SD] 9.4) years (range 20–65 y) and a mean education level of 10.9 (SD 3.7) years (range 0–20 y). All were Han Chinese living in Taiwan. The subjects in this study gave their consent to participate after procedures were explained to them. All subjects were free of any axis I or II psychiatric disorders, as determined by a research psychiatrist using the Structured Clinical Interview for DSM-IV.36 All participants were also in good physical health, as determined by physical examination, electrocardiogram and laboratory tests including liver, renal and thyroid function tests and urinalysis.

WCST assessments

An experienced psychologist (Y-LC) administered the WCST.9,10 In this test, subjects are required to sort stimulus cards on the basis of perceptual attributes (colour, form, number). The only feedback provided by the administrator is to indicate whether each response is correct or incorrect. The sorting rule is changed after 10 consecutive correct responses. Testing is discontinued when the subject has reached 128 trials.37 The a priori primary outcome measure was the percentage of perseverative errors because it is sensitive to a person's ability to fluently shift cognitive sets and is thought to best reflect the prefrontal cognitive function.8 The accessory outcome variable was the number of categories completed, as used by other investigators.34 To minimize multiple comparisons, we determined a priori not to analyze other WCST parameters.

Genotyping

Genomic DNA was extracted from the subjects' white blood cells. The 4 polymorphisms were genotyped according to the polymerase chain reaction-restriction fragment length polymorphism technique.

Forward and reverse primers and restriction enzymes for each polymorphism were as follows: 5′-ACT GAC CCC TAT TCC CTG CT-3′, 5′-AGC ACA GAC CAG CGT GTT C-3′, DdeI for A-48G polymorphisms in the promoter region of the dopamine D₁ receptor gene;38 5′-GCT CTA TCT CCA ACT CTC ACA-3′, 5′-AAG TCT ACT CAC CTC CAG GTA-3′, MscΙ for the D₃ Ser9Gly polymorphism;39,40 5′-TCT GCT ACA AGT TCT GGC TT-3′, 5′-CTG CAG CTT TTT CTC TAG GG-3′, MspI for 5-HT_2A T102C;41,42 and 5′-AAC TTC TTC CTG GTG TCG CTC TTC-3′, 5′-ATG AGC AGG TAG CGG TCC AGG-3′, RsaI for T267C polymorphisms in the 5-HT₆ receptor gene.37,43

DNA fragments were visualized with 2% agarose gel electrophoresis and staining with ethidium bromide. Duplicate genotyping was conducted for all 4 polymorphisms, with 100% concordance. All assays were performed blind to the subjects' cognitive measurements.

Data analyses

We used a 2-tailed Mann–Whitney U test for WCST performance comparisons between male and female subjects. Kruskal–Wallis tests were carried out with each genetic polymorphism as the independent variable and percentage of perseverative errors and number of categories as the dependent measures. These nonparametric statistical methods were used because the distributions of perseverative errors and categories numbers were skewed to the right (data not shown). We also compared the subjects' demographic characteristics among genotypes of each genetic polymorphism, using Kruskal–Wallis tests for continuous variables and χ² tests for categorical variables. Statistical significance was defined as p < 0.05 or corrected for multiple comparisons as needed (p < 0.0125 = 0.05/4 [genetic polymorphisms]).

To control for the effects of other confounding factors, we used multiple linear regression in a single analysis to evaluate impacts of sex, age, sex × age interaction, education level and the 4 genetic polymorphisms simultaneously on WCST performance. Because all these comparisons were performed in a single multiple regression analysis, no correction for multiple comparisons was required. However, the distributions of the WCST data were skewed to the right and unsuitable for regression analyses. To convert the skewed distribution to a normal one, we measured each perseverative error value in its natural logarithmic scale before the regression analysis.44,45

Results

In all 216 subjects, the mean percentage of WCST perseverative errors was 20.1% (SD 13.1%) and the mean number of WCST categories completed was 4.18 (SD 3.08). Men and women had comparable mean perseverative errors at 18.7 (SD 12.0) and 20.9 (SD 13.8) (p = 0.46), respectively, and comparable mean numbers of categories completed at 4.4 (SD 3.2) and 4.1 (SD 3.0) (p = 0.55).

Allele frequencies of the genetic polymorphisms in our Han Chinese sample are similar to those in other Han Chinese populations but may be different from those in white populations (Table 1). The genotype distribution of each of the 4 genetic polymorphisms did not deviate significantly from the Hardy–Weinberg equilibrium: D₁ A-48G (χ² = 0.839, p = 0.657), D₃ S er9G1y (χ² = 1.873, p = 0.392), 5-HT_2A T102C (χ² = 0.76, p = 0.683) and 5-HT₆ T267C (χ² = 1.17, p = 0.558).

Table 1

WCST perseverative errors and genotypes

Table 2 shows differences in WCST perseverative errors among genotypes of each of the 4 genetic variances. D₃ Ser9Gly, 5-HT_2A T102C and 5-HT₆ T267C, but not D₁ A-48G, significantly influenced perseverative errors. If corrected for multiple comparisons of the 4 genetic polymorphisms, D₃ Ser9Gly and 5-HT₆ T267C remained significant in modulating the functional outcome measures (α = 0.05/4 = 0.013). Sex, age and education level did not significantly differ among genotypes of each genetic polymorphism (all p values > 0.05, data not shown).

Table 2

Because various genetic factors may simultaneously determine executive function, we then conducted multiple linear regression analysis to evaluate effects of the 4 genetic polymorphisms on WCST perseverative errors after controlling for sex, age, gender × age interaction and education level. Prior to multiple regression, each WCST value was transformed to its natural logarithm to obtain normal distributions (data not shown). As revealed in Table 3, D₁ A-48G, D₃ Ser9Gly, 5-HT_2A T102C and 5-HT₆ T267C polymorphisms significantly influence perseverative errors after adjustment for sex, age, sex × age interaction, education level and other genetic variants. Compared with subjects having the D₁ A-48G A/A genotype, those with A/G had a 74% more perseverative errors (e0.5561 = 1.7438, 1.7438 — 1 = 0.7438), and those with G/G had a 101% more perseverative errors (e^0.6968 = 1.9447, 2.0073 – 1 = 1.0073). Compared with subjects with the D₃ Ser9Gly Ser/Ser genotype, those with Ser/Gly had 23% more perseverative errors (e^0.2103 = 1.2340). However, the difference in perseverative errors between those with Gly/Gly and those with Ser/Ser did not achieve statistical significance. Compared with subjects having the 5-HT_2A T102C T/T genotype, those with T/C committed 25% more perseverative errors (e^0.2257 = 1.2532). The 2 homozygote groups, however, did not have significant differences in perseverative errors. Compared with subjects with the 5-HT₆ T267C T/T genotype, those with T/C had 42% fewer perseverative errors (e^–0.5468 = 0.5788, 1 – 0.5788 = 0.4212) and those with C/C had 32% fewer errors (e^–0.3920 = 0.6757, 1 – 0.6757 = 0.3243). Because all these comparisons were performed in a single multiple regression analysis (Table 3), no correction for multiple comparisons was needed. There was a sex × age interaction effect on perseverative errors. Compared with men, perseverative errors made by women increased less rapidly with age. Education level did not significantly influence the cognitive measures.

Table 3

Genotypes and WCST categories completed

Regarding the number of categories completed, we found that only the D₃ Ser9Gly polymorphism had significant influence. The mean number of categories completed for the subjects with the Ser/Ser genotype was 4.7 (SD 3.1); for those with Ser/Gly, it was 3.7 (SD 3.1); and for those with Gly/Gly, it was 3.5 (SD 2.9) (p = 0.029). However, if corrected for the 4 multiple comparisons (see the above section), the difference did not achieve statistical significance. Multiple linear regression analysis also did not reveal statistical significance for any genetic variance. These findings support the notion that the percentage of perseverative errors is the most sensitive of the WCST parameters in terms of reflecting prefrontal function.8

Discussion

The main finding is that D₁ A-48G, D₃ Ser9Gly, 5-HT_2A T102C and 5-HT₆ T267C variants can significantly influence executive function in healthy adults, supporting the hypothesis that D₁, D₃, 5-HT_2A and 5-HT₆ receptors are involved in prefrontal executive cognition. Certainly, chance false-positive findings might have happened. Although we included a total of 216 subjects, cell sizes for some genotypes that were used as the basis of between-genotype comparisons were small (n = 5 and n = 12 for the D₁ A-48G and 5-HT₆ T267C polymorphisms, respectively). Larger samples are needed to substantiate the results.

Interestingly, while subjects homozygous for the D₃ Ser/Ser genotype performed better in terms of perseverative errors than those heterozygous for Ser/Gly, subjects homozygous for Gly/Gly tended to have intermediate performances (although this was nonsignificant). This phenomenon is not unique for functional studies of D₃ Ser/Gly polymorphisms. For instance, people homozygous for Gly/Gly demonstrate intermediate levels of eye movement disturbances whether they are schizophrenia patients or healthy subjects.19 Ser/Gly heterozygotic subjects have lower 5-hydroxyindoleacetic acid levels in their cerebral spinal fluid than Ser/Ser and Gly/Gly homozygotic subjects.46 Excess homozygosity (both the wild type and the variant) can also be a feature of certain subgroups of schizophrenia patients, such as those who respond to neuroleptic treatment.47 The heterozygote of the 5-HT_2A T102C polymorphism also displayed unusual functional consequences. Whereas subjects with the 5-HT_2A T/T genotype had the fewest perseverative errors, those with T/C (rather than those with C/C) had the poorest performances (Table 2, Table 3). In accordance, heterozygous patients with T/C have the lowest N100 amplitude when compared with the other 2 Han Chinese genetic groups.27 Last, 5-HT₆ T267C polymorphisms may have similar effects. The T/T genotype was associated with the highest perecentage of perseverative errors, and the C/C genotype tended to have intermediate performance (Table 2, Table 3). Likewise, schizophrenia patients with the 5-HT₆ C/C genotype show intermediate treatment response to clozapine, compared with those with T/T or T/C.35 Such findings on “positive or negative heterosis (the highest or lowest phenotypic expression in heterozygotes)”48 have been reported in the neuroscience49 and other50,51 literature.

An advantage of this study is the apparent ethnic homogeneity of Han Chinese people. In general, the Han Chinese population is genetically homogeneous. However, given the large population and the wide distribution of the Han Chinese, we cannot exclude the potential population admixture effect on the study results. Of note, cognitive capacity, as measured by performance on neuropsychological tests, is related to many aspects of function in schizophrenia patients.52,53 Whether our findings could be extrapolated to other ethnic populations or to patients with psychiatric disorders remains unclear. In another study,54 which recruited antipsychotics-treated schizophrenia patients, D₁ A-48G, but not D₃ Ser9Gly or D₄ C-521T, influenced WCST performance. In female patients, a relation was also found between D₂-141C Ins/Del polymorphism and perseverative errors. Patients with the G/G genotype of the D₁ receptor gene tended to obtain poorer WCST results,54 similar to our finding in healthy people. The findings for the D₃ variances differ from ours. Because both first-and second-generation antipsychotics act on D₃ receptors,17,39 the potential confounding influences from drugs need consideration. Differences in subject ethnicity may also contribute to the discrepancy.

The WCST perseverative error performance in this study is similar to that in another healthy Chinese population55 but inferior to that in a healthy white population.56 Further, the percentage of perseverative errors and the number of completed categories in the healthy population in the current study are similar to the results obtained in schizophrenia patients in studies performed in the United States and Europe.8,11 In conclusion, although our results are of interest, the definitive answers and the possible applications to other ethnic groups or mentally ill patients require future, larger studies. Besides, the question of whether polymorphisms of other serotonin or dopamine receptors or serotonin or norepinephrine transporters (the latter of which is responsible for dopamine reuptake in the prefrontal cortex) are also associated with WCST performance deserves more study. Finally, glutamate neurotransmission57–59 and other pathways2 are also crucial for high cortical function in the prefrontal areas. Therefore, possible effects of other genetic variances than dopamine-or serotonin-related ones on executive function or other neuroncognitions60 also require elucidation.