CYP2D6 polymorphism in a Gabonese population: contribution of the CYP2D6*2and CYP2D6*17alleles to the high prevalence of the intermediate metabolic phenotype

Abstract

Aims

To determine the molecular basis of the intermediate extensive metaboliser (EM) CYP2D6 phenotype in healthy Gabonese subjects.

Methods

The CYP2D6 phenotype of 154 healthy Gabonese subjects was assessed by giving the subject a single dose of 30 mg dextromethorphan, and collecting their urine for the next 8 h. The CYP2D6 genotype was determined for 50 individuals of the EM phenotypic group by Southern blotting and various PCR-based procedures aimed at identifying different CYP2D6 alleles.

Results

We found that in the studied Gabonese population, as compared with a French population, there is significantly higher frequency of intermediate EM phenotype having lower frequency of CYP2D6 PM alleles. To clarify this discrepancy phenotype-genotype relationship was studied. We found that the CYP2D6*17 and CYP2D6*2 alleles, prevalent in this black population, are characterised by their low capacity for dextromethorphan demethylation. Our data also show that the CYP2D6*1 allele is associated with the highest in vivo activity followed by the CYP2D6*2 allele and then the CYP2D6*17 allele.

Conclusions

The higher frequencies of the CYP2D6*2 and CYP2D6*17 alleles than the CYP2D6*1 allele account for the high frequency of the intermediate EM phenotype in this black population. The polymorphism of the CYP2D6 enzyme activity in African populations could have important implications for use of drugs that are substrates for CYP2D6 and have a narrow therapeutic window.

Introduction

The debrisoquine/sparteine type polymorphism is a well-characterised genetic variation in drug metabolism associated with cytochrome P450 CYP2D6 [1]. People with active CYP2D6 are termed extensive metaboliser (EM). The enzyme activity in this group shows wide variation and individuals can be classified as ultrarapid, intermediate and slow EMs. Between 1 and 7% of Caucasians and 29% of Ethiopians are ultrarapid metabolisers with multiple copies of functional CYP2D6 genes [2–4]. The genetic basis of the poor metaboliser (PM) group is now well understood, but much less is known about the molecular basis of the heterogeneity within the EM group. EMs with intermediate activity are either heterozygotes for one of the poor metaboliser alleles or carry at least one of the functionally active but less efficient CYP2D6 alleles (CYP2D6*2 for Caucasians, CYP2D6*10 for Chinese, CYP2D*17 for Zimbabweans [5, 6, 7, 8, 9, 10]).

We investigated the relationship between metabolic phenotype and CYP2D6 genotype in a Bantu-speaking black population in Gabon. We also explored the molecular basis of interethnic differences in CYP2D6 activity by comparing data obtained from the Gabonese population with those previously obtained from a French population [11].

Methods

One hundred and fifty-four unrelated, healthy Gabonese volunteers (102 men and 52 women) gave their informed consent to participate in the phenotype studies performed at the ‘Centre International de Recherches Médicales de Franceville’ (Gabon). CYP2D6 phenotype was assessed by giving the subject a single dose of 30 mg dextromethorphan, and collecting their urine for the next 8 h. The amount of dextromethorphan (DEM) and dextrorphan (DOR) in the urine was determined by high performance liquid chromatography (h.p.l.c.) [11] and the urinary metabolic ratio (MR) DEM/DOR was calculated. Subjects with a log DEM/DOR ratio higher than −0.5 were assigned to the PM phenotype.

Blood samples for genotyping were obtained from 50 individuals of the EM group. These DNA samples were analysed by Southern blotting (XbaI polymorphism) and using various published PCR-based procedures aimed at identifying the following CYP2D6 alleles: CYP2D6*1, *2, *3, *4, *10 and*17 [7, 9, 12, 13] except for the CYP2D6*2 allele. The CYP2D*2 allele frequency was obtained by substracting the frequency of the CYP2D6*17 allele (exon 2 variation [9]) from the overall frequency of the exon 6 mutation [7].

The analysis of variance test was used for statistical evaluation of the differences between mean MR in individual genotypes. Where statistically significant differences were found groups were compared using the Student’s t-test with a Bonferroni correction. Chi-square analysis was used to test the differences in genotype and allelic variations between different ethnic groups.

Results

The distribution profile of the log MR was bimodal (Figure 1a) with a PM phenotype frequency of 0.02 (n =3 subjects). The distribution of the data for this black population was right-skewed as compared with that obtained for a French Caucasian population (Figure 1b) [11], suggesting a higher frequency of intermediate EM phenotype.

Figure 1.

a) Frequency distribution of the log dextromethorphan (DEM) and dextrorphan (DOR) urinary metabolic ratio (MR) in 154 healthy Gabonese individuals. EM and PM individuals were defined as those with a log DEM/DOR MR lower (EM) and higher (PM) than—0.5. b) Frequency distribution of the log DEM/DOR MR in 422 healthy French Caucasians. The data are from [11].

The CYP2D6 genotype and phenotype observed in the Gabonese subjects are shown in Table 1a. Analysis of allele segregation in family members indicated that the 16+9 kb-XbaI allele was associated with the CYP2D6*4 poor metaboliser (PM) allele whereas the 42 kb-XbaI fragment was associated with the CYP2D6*2X2 extensive metaboliser (EM) allele. As with other black populations [14–16], the frequency of the CYP2D6 PM alleles in the Gabonese EM group was significantly lower (0.09) than that in Caucasians (0.21) (P<0.02). The most prevalent mutant allele in the black population was the CYP2D6*5 allele (frequency 0.07). The heterozygotes (EM/PM) had a lower mean CYP2D6 activity, as shown by their low MR values (−1.4±0.9 (n =9) vs−2.1±0.9 for the homozygotes XbaI 29 kb/29 kb EM/EM (n =39)): (P<0.05). Within the homozygous EM XbaI 29/29kb group (n =39 individuals), the observed frequencies were 0.32 for CYP2D6*1, 0.44 for CYP2D6*2, and 0.24 for CYP2D6*17.

Table 1.

CYP2D genotype and phenotype in the Gabonese.

The relationship between MR values and six genotypes defined by the CYP2D6*1, CYP2D6*2, CYP2D6*17 alleles was analysed in the Gabonese EM group (Table 1b)(only subjects homozygous for 29 kb fragments were analysed to avoid a bias due to differences in CYP2D gene copy numbers). These results show a non random distribution of the CYP2D genotypes within the EM phenotypes. Finally, there was no significant difference in mean MR values between the French Caucasian [11] and Gabonese individuals with CYP2D6*1/CYP2D6*1, CYP2D6*2/CYP2D6*1 and CYP2D6*2/CYP2D6*2 genotypes (data not shown).

Discussion

Our results show that the CYP2D6*17 allele was associated with the lowest capacity for dextromethorphan demethylation in the Gabonese population (Table 1b). Although the CYP2D6*2 allele has been reported to have similar catalytic activity (for debrisoquine) to that of CYP2D6*1 [2, 9], we found that the CYP2D6*2 allele was associated with lower levels of CYP2D6 activity, consistent with previous in vivo observations in Caucasians [7, 10]. We do not know the reason for this difference but the association of the CYP2D6*2 allele with lower CYP2D6 activity might apply only to dextromethorphan and not to debrisoquine, as these substrates belong to two different chemical classes [17]. Recently, Oscarson et al. [18] reported in vitro studies showing that the CYP2D6*2 protein exhibits lower activity with bufuralol than CYP2D6*1, which is in agreement with our data. Further studies using different substrates are required to determine the molecular basis for the lower CYP2D6 activity of the CYP2D6*2 enzyme. Nevertheless, the CYP2D6*2 allele in vivo exhibited higher CYP2D6 activity than the CYP2D6*17 allele (Table 1b) (also in agreement with in vitro studies [18]). The CYP2D6*17 allele was relatively frequent in the Gabonese population studied consistent with results for another black population [9]. However, the frequency of the CYP2D6*2 allele was lower in Zimbabwean [9] than in Gabonese (0.13 vs 0.40). This difference may be due to i) genuine ethnic differences between these two population groups, or ii) a difference in the CYP2D6*2 allele detection procedure. Concerning the latter possibility, the CYP2D6*2 allele was detected by analysis of sequence variation in intron 1 [9] rather than by detection of CYP2D6 coding sequence variations (in exons 3, 6 and 9) (see Method section).

This analysis in a Bantu population from Gabon shows that the CYP2D6*1 allele is associated with the highest activity in vivo followed by the CYP2D6*2 allele and then the CYP2D6*17 allele. In conclusion, the higher frequencies of the CYP2D6*2 and CYP2D6*17 alleles than the CYP2D6*1 allele may account for the high incidence of the intermediate EM phenotype in this black population. The polymorphism of the CYP2D6 enzyme activity in African populations could have important implications for use of drugs that are substrates for CYP2D6 and have a narrow therapeutic window [19, 20].