There is strong evidence for genetic influences on personality traits. Interest in one such gene, the dopamine D4 receptor (DRD4) grew after an exon III polymorphism was associated with Novelty Seeking and related measures of Extraversion. However, the findings were not confirmed in later studies. Recently, a −521C/T single nucleotide polymorphism (SNP) within the promoter region of the DRD4 gene was found to be related to Novelty Seeking scores in populations from Japan and Hungary. Since little is known about the role DRD4 plays in personality in other populations we evaluated if two DRD4 promoter SNPs, −521C/T and −616C/G, were related to personality traits in African Americans. Personality traits were measured by the NEO-FFI in 71 unrelated African Americans. Genotyping was performed using PCR-RFLP. Multivariate analyses of variance (MANOVA) were performed to evaluate the effects of gender and −616 and −521 genotypes on personality traits. A significant three-way interaction effect from gender, −616 genotype, and −521 genotype was observed for Extraversion scores (F(1,54) 5.86, P < 0.02). Subsequent analyses revealed that the association was mainly due to −521C/T genotype among females (P = 0.01). This study provides further evidence that genetic variation within the DRD4 promoter and gender differences contribute to variation in Novelty Seeking behaviors such as Extraversion.

Human personality is influenced by inherited components.¹ However, no clear genetic candidates have consistently been shown to be associated with human personality traits. Various questionnaires have been developed to measure personality such as the Tridimensional Personality Questionnaire (TPQ, now called TCI),² and the NEO Five Factor Inventory (NEO-FFI). The NEO-FFI is a hierarchical organization of personality traits portrayed in five basic dimensions: Neuroticism, Extraversion, Openness to Experience, Agreeableness, and Conscientiousness.³ All five dimensions are thought to be independent and linked to specific neurogenetic pathways.⁴ Exploratory, impulsive, extravagant and sensation seeking behavior has been linked to dopaminergic systems.⁵ Dopamine is a monoamine neurotransmitter found in parts of the brain that regulate movement, mood, and the rewarding effects of drugs.⁶ The D4 receptor is expressed in limbic areas of the brain involved in cognition and emotion⁷ and has been shown to mediate exploratory behavior in experimental animals.⁸ Variation within the DRD4 gene has been implicated for involvement with the personality trait of Novelty Seeking^9,10,11 which is related to Extraversion and is measured using the Tridimensional Personality Questionnaire (TPQ). The reported associations were between the 48-base pair repeat in exon III of DRD4 and Novelty Seeking in an American population consisting mainly of European Americans⁹ and Ashkenazi Jews from Israel.¹⁰ Subsequent research using different populations has been controversial and has yielded no consensus on the involvement of DRD4 with Novelty Seeking behavior.^{12,13,14,15,16,17}

Recently several polymorphisms in the 5′ promoter region of DRD4 have been identified.^16,18 One of these polymorphisms, a C to T single nucleotide polymorphism (SNP) upstream of the translation start site at position −521 has been associated with Novelty Seeking behavior in separate studies of Japanese and Hungarians.^16,17 This polymorphism may have potential functional importance because it occurs in a region regulating transcriptional activity. In a transient expression system performed by Okuyama et al, the −521 C allele was shown to be 40% more active than the T allele.¹⁶ Another interesting polymorphism in the promoter region of DRD4 is the −616C/G SNP that may result in a gain of an AP-2 binding site.¹⁹

The close proximity, possible regulatory significance, and the ability to construct haplotypes using the −616 and −521 SNPs increases statistical power in order to evaluate the role DRD4 plays in personality. In addition, complex phenotypes such as behavior are the result of strong interactions between genetic loci and the environment. Thus, it is important to perform these studies in different populations since each may represent distinct environments that could potentially interact with genetic variants. Presently, few data exist on the genetic involvement in human temperament among African Americans. In this study we evaluated if two promoter SNPs, at positions −616 and −521 relative to the start site of the gene, contribute to human temperament as measured by the NEO-FFI among 71 African Americans from Washington, DC, USA. The NEO-FFI questionnaire (Form S) is a 60-item scale that measures Neuroticism, Extraversion, Openness to Experience, Agreeableness, and Conscientiousness. No significant differences in personality scores were observed between African American males and females (Table 1). Allele frequencies for the −616 C/G (0.33/0.67) and the −521 C/T (0.55/0.45) SNPs were consistent with observed frequencies in Japanese and Hungarians^16,17 and were not significantly different from frequencies observed in Nigerians, Scandinavians, and Han Chinese (Bookman et al, unpublished data). However, a Canadian study of mixed Europeans did observe quite different frequencies for the −616 SNP.¹⁹ Genotypes for both markers (−621 CC: 0.13; CG: 0.31; GG: 0.56; and −521 CC: 0.32; CT: 0.42; TT: 0.26) were in Hardy–Weinberg equilibrium (P > 0.05). Haplotypes were estimated from the marker genotypes by a maximum likelihood method.²⁰ All four possible DRD4 −616 and −521 haplotypes were observed within the African American sample. Interestingly, no significant linkage disequilibrium (LD) was detected between the −616C/G and −521C/T SNPs (D′ = 0.09; P = 0.46). The lack of LD between markers separated by only 95 nucleotides was quite striking but not surprising since the ability to detect LD is influenced by marker allele frequencies, population history, selection, and sample size. The lack of LD between the −521 and −616 SNPs in African Americans is in contrast to the strong LD recently observed between the −521 SNP and the exon III 48-bp repeat in the Finnish population.²¹

We found no significant differences between the CG and GG genotypes for the C −616G polymorphism and between the CT and TT genotypes for the C −521T polymorphism among any of the personality dimensions. Similar to previous studies we assumed a dominant model and combined the genotype categories for the analyses. Three-way ANOVA (MANOVA) was performed to test the effects of the −616C/G genotype, −521C/T genotype, and gender on NEO-FFI scores. We observed no significant effect of gender alone on NEO-FFI scores, however there was a significant main effect observed for Extraversion scores with the −521C/T genotype (F(1,54) = 3.99; P = 0.05) and nominal significance between Extraversion scores and the −616C/G genotype (F(1,54) = 3.56; P = 0.06). The MANOVA also revealed a strong three-way interaction effect on Extraversion from the two promoter SNPs and gender (F(1,54) = 5.86, P = 0.02). There was no DRD4 genotype effects observed for any of the other personality dimensions. Subsequent analyses of the interaction effects with Extraversion are presented in Table 2. No association was observed in the follow-up analyses of C−616G genotype and Extraversion scores in any of the separate groups. The analysis of C−521T genotypes revealed an association between genotype and Extraversion among females (Table 2). Specifically, females with at least one copy of the −521 T allele scored significantly lower in mean extraversion scores (CC: 32.1 ± 1.03 vs CT/TT: 28.2 ± 1.03, F (1, 35) = 6.78, P = 0.01).

The two promoter SNPs analyzed here and other polymorphisms in exons I, II and III of the DRD4 gene have been implicated in an array of personality dimensions and psychiatric disorders such as Novelty Seeking, schizophrenia, bipolar disorder, and anorexia nervosa.^{9,10,11,16,17,22,23,24,25} While we did not find a significant difference in mean personality scores among −616 DRD4 marker genotypes, a strong association with the Novelty Seeking related behavior of Extraversion was observed for the −521 marker. Consistent with previous findings related to Novelty Seeking behaviors our study revealed significantly lower Extraversion scores among females with at least one T allele at the −521 DRD4 site. Various scenarios may explain the strong effect observed among the female subjects. There may be interactions between the DRD4 −521 SNP and environmental influences that differ between males and females. For example, environmental experiences that affect Extraversion when the −521 T allele is present may be more common among females (or inversely among males). If this is true, then the genetic effect of the −521C/T SNP may be small compared to environmental effects. This is important because several sociocultural models posit that social and cultural factors directly influence gender differences in personality.^26,27 Another potential contributor to the sex differences may be gender-related epistatic effects (interaction among different genes).

Our findings here provide support to earlier research on the role the DRD4 gene plays in Extraversion and Novelty Seeking behavior. Previous studies found that the VNTR polymorphism in exon III of the DRD4 gene was associated with Extraversion and Novelty Seeking behavior.^9,10,11 However, some researchers have not confirmed these findings.^12,13,14,15 It is likely that the exon III 48-base pair polymorphism does not actually contribute to Extraversion but is linked to some other DRD4 variant that does. Events such as recombination could place the causal allele on different DRD4 haplotypic backgrounds. If this were true then single marker association studies would produce conflicting results. In addition there appears to be some sample size issues and gender-related effects that were neglected or poorly examined in previous studies.^13,14,21

We note that allele frequency differences between ethnic populations can be a confounder in association studies if not controlled for,^28,29 especially genetic association studies on the African American population, which is highly heterogeneous due to its African ancestry and recent admixture with European Americans. However the likelihood that our finding is spurious due to population stratification is very low because the allele frequencies for the two SNPs were consistent with frequencies previously described in two other ethnic groups.^16,17 In fact, the allele frequencies observed in our sample of African Americans are quite similar to frequencies observed in divergent populations.³⁰ Another reason why the association is likely is because it has consistently been shown in independent studies across ethnically divergent populations.

In summary, this study replicates the results of two other studies using different populations, which show a relationship between DRD4 −521 genotype and Novelty Seeking related behavior and further provides evidence of a sex-specific relationship between DRD4 −521 genotype and personality. This study is the first to examine the role DRD4 haplotypes play in personality within healthy African Americans. For complex phenotypes such as personality it is likely that multiple polymorphisms within different genes contribute a range of effects on the trait. Since allele frequencies may vary significantly among different populations, as is the case for several polymorphisms with the DRD4 gene,^30,31 it is extremely important that different ethnic groups, such as African Americans, are included in association studies of complex phenotypes. While the findings are intriguing, they suggest that more research on the functional consequence of the DRD4 gene promoter variants is warranted. Future studies may increase our knowledge of the role the DRD4 gene plays in Extraversion and Novelty Seeking types of behavior. In addition it is likely that other genes or genomic regions will be identified that contribute to variation in human personality traits of Extraversion, Novelty Seeking, and positive emotional experiences.

Subjects and personality inventory

Healthy, unrelated African American male (n = 33; mean age = 35.1 years) and female (n = 38; mean age = 31.9 years) volunteers were recruited from Howard University and the Washington, DC metropolitan area. Ethnicity was self-reported and none of the volunteers have ever been diagnosed with a psychiatric disorder. All volunteers completed the NEO-FFI questionnaire (Form S). Cronbach alpha values for internal consistency of each scale were 0.88 for Neuroticism; 0.80 for Extraversion; 0.79 for Openness to Experiences; 0.76 for Agreeableness; and 0.83 for Conscientiousness. The Howard University Institutional Review Board approved this study and written consent was obtained from all subjects.

Genotyping

DNA was extracted from whole blood using the PureGene kit (Gentra, Inc, Plymouth, MN, USA). The PCR was performed in a total volume of 25 μl containing 50 mM KCl, 10 mM Tris-HCl (pH 8.3), 1.5 mM MgCl₂, 200 mM each of dATP, dCTP and dTTP, 50 mM of dGTP and 150 mM of 7-deaza-dGTP, 10% DMSO, 0.6 units of TaqGold polymerase, 20 pmol of each primer, and 50 ng of genomic DNA. A single PCR reaction was used to produce an amplicon containing both the −616 and −521 nucleotide sites. Primer sequences were forward, 5′- TCA ACT GTG CAA CGG GTG −3′ and reverse, 5′-GAG AAA CCG ACA AGG ATG GAG-3′. Amplification conditions include 95°C for 1 min, 58.0°C for 1 min and 72°C for 1 min for 40 cycles followed by 10 min at 72°C in a PE 9700 thermal cycler. Restriction enzyme digestion was performed on the PCR fragment in separate reactions using 10 units of AvaII for the −616C/G SNP and FspI for the −521C/T SNP. Fragments were visualized using a 4% agarose gel. Genotyping was performed blind to phenotypic information. All samples were genotyped twice directly from genomic DNA along with control samples of known genotype. Due to limited genomic DNA the −616C/G SNP was not genotyped in three individuals reducing the sample size for the −616C/G analyses to 68 individuals.

Statistical analyses

Genotype and allele frequencies were calculated for the −616 and −521 markers by gene counting. Hardy–Weinberg equilibrium tests were performed by contingency table analysis. Haplotypes were estimated from the marker genotypes by a maximum likelihood method using the computer program 3LOCUS (written by J Long), which implements the EM algorithm.^20,32 The program also calculated the total ‘normalized’ disequilibrium value,³³ which is an extension of Lewontin's D′,³⁴ for the marker pair. The SAS Version 6.12 computer program (SAS Institute, Inc, Cary, NC, USA) was used to perform the Multivariate Analyses of Variance (MANOVA) to test for associations of NEO-FFI personality scores with DRD4 promoter SNPs. Interaction effects were then subsequently evaluated using one-way ANOVAs.

We would like to express our appreciation to Dr W Chen, Dr K Panguluri, N Sylvester, G Vosganian, T Vaughn, and A Massac for recruitment and technical assistance. This work was supported by grant 5U24AA11898–03 from NIAAA and the Office of Research on Minority Health, NIH and funds from the National Human Genome Center at Howard University.

Affiliations

1. Department of Genetics and Human Genetics, Howard University, Washington, DC, USA

   • E B Bookman

2. Department of Pharmacology, Howard University, Washington, DC, USA

   • R E Taylor

3. Howard University Cancer Center, Howard University, Washington, DC, USA

   • L Adams-Campbell
   •  & R A Kittles

4. National Human Genome Center at Howard University, Howard University, Washington, DC, USA

   • E B Bookman
   •  & R A Kittles

Corresponding author

Correspondence to R A Kittles.

To obtain permission to re-use content from this article visit RightsLink.

Publication history

DOI

https://doi.org/10.1038/sj.mp.4001075