Genetic polymorphisms of CYP2C9 and CYP2C19 in the Beninese and Belgian populations


  • Aurel C. Allabi,

    1. Medical Faculty of National University of Benin (UNB),
    2. Applied Molecular Technologies, Centre for Human Genetics, Université Catholique de Louvain (UCL), Louvain, Belgium,
    3. Clinical Pharmacology Unit, St Luc University Hospital (UCL), Brussels, Belgium, and
    Search for more papers by this author
  • Jean-Luc Gala,

    Corresponding author
    1. Applied Molecular Technologies, Centre for Human Genetics, Université Catholique de Louvain (UCL), Louvain, Belgium,
    2. Defence Laboratories Department (DG MR), Belgian Armed Forces, Brussels, Belgium
    Search for more papers by this author
  • Jean-Pierre Desager,

    1. Clinical Pharmacology Unit, St Luc University Hospital (UCL), Brussels, Belgium, and
    Search for more papers by this author
  • Michel Heusterspreute,

    1. Applied Molecular Technologies, Centre for Human Genetics, Université Catholique de Louvain (UCL), Louvain, Belgium,
    Search for more papers by this author
  • Yves Horsmans

    1. Clinical Pharmacology Unit, St Luc University Hospital (UCL), Brussels, Belgium, and
    Search for more papers by this author

J.-L. Gala MD, PhD, Laboratory of Applied Molecular Technology, Centre for Human Genetics–UCL, Clos Chapelle-aux-Champs, 30-UCL/30.46, B-1200, Brussels, Belgium. Tel.: + 32 2 764 3165; Fax: + 32 2 764 3166; E-mail:


Aims  To investigate the distribution of cytochrome P450 2C9 (CYP2C9) and 2C19 (CYP2C19) genotype frequencies in the Beninese and Belgian Caucasian populations.

Methods  Beninese (n = 111) and Belgian (n = 121) were genotyped for CYP2C9*2, *3, *4, *5, and *11 as well as for CYP2C19*2 and*3.

Results  The distribution of alleles was: CYP2C9*1: 95.5 vs. 82.2% (P < 0.001); CYP2C9*2: 0 vs. 10% (P < 0.001); CYP2C9*3: 0 vs. 7.4% (P < 0.01); CYP2C9*4: both 0%; CYP2C9*5: 1.8 vs. 0% (P = 0.05); and CYP2C9*11: 2.7 vs. 0.4% (P < 0.05). The frequencies of the CYP2C19*2 allele were 13 vs. 9.1%, respectively. CYP2C19*3 was not detected in either population. The 95% confidence intervals for the differences of frequencies of CYP2C9*1, CYP2C9*2, CYP2C9*3, CYP2C9*4, CYP2C9*5, CYP2C9*11, CYP2C19*1, CYP2C19*2 and CYP2C19*3 between  Belgian  and  Beninese  were  7%,  19%;  − 14%,  − 6%;  − 11%, − 4%; − 1%, 1%; 0%, 4%; 0%, 5%; − 10%, 2%; − 2%, 10%; − 1%; respectively. The distributions of CYP2C9 genotypes in the Beninese and Belgian individuals were: CYP2C9*1/*1: 91 vs. 67% (P < 0.00001); CYP2C9*1/*2: 0 vs. 18.2% (P < 0.0001); CYP2C9*1/*3: 0 vs. 11.6% (P < 0.001); CYP2C9*1/*5: 3.6 vs. 0% (P = 0.05); CYP2C9*1/*11: 5.4 vs. 0.8% (P = 0.05); CYP2C9*2/*3: 0 vs. 1.6% (NS); CYP2C9*3/*3: 0 vs. 0.8% (NS). The distributions of CYP2C19 genotypes between these ethnic groups were: CYP2C19*1/*1: 73.9 vs. 83.5% (NS); CYP2C19*1/*2: 26.1 vs. 14.9% (P < 0.05); CYP2C9*2/*2: 0 vs. 1.6% (NS).

Conclusions  Differences of allele frequencies between Beninese and Belgian populations were statistically significant for CYP2C9*2, *3, *5 and *11, but not for CYP2C9*4 or for CYP2C19*2 and *3.


The cytochrome P450 2C (CYP2C) subfamily of enzymes metabolizes approximately 20% of drugs commonly used in clinical practice [1]. It comprises four members (CYP2C8, 2C9, 2C18 and 2C19), which exhibit a range of genetic polymorphisms. Those affecting CYP2C9 and CYP2C19 result in an impaired capacity to metabolize a range of drugs, the consequences of which may be clinically significant [1, 2]. The CYP2C9 and CYP2C19 polymorphisms have both been extensively characterized in Caucasians and Orientals, and to a lesser extent in Black populations. Whereas data are available for African-American and East-African populations, West African populations have not been studied in this context.

CYP2C9 is involved in the oxidation of a wide range of drugs, including S-warfarin, phenytoin, tolbutamide, losartan and torasemide. Several nonsteroidal anti-inflammatory drugs, including diclofenac, naproxen, piroxicam and ibuprofen, as well as the selective COX-2 inhibitor celecoxib, are also mainly metabolized by CYP2C9 [3].

The CYP2C9 gene displays functional genetic polymorphisms unambiguously associated with impaired CYP2C9-mediated metabolism. The products of the CYP2C9*2 (Arg144Cys) and CYP2C9*3 (Ile359Leu) alleles (see are enzymes with impaired activity towards a number of substrates, both in vivo and in vitro[1, 3–5]. The CYP2C9*4 (Ile359Thr) polymorphism was first identified in a Japanese patient presenting with epilepsy [6], but its functional influence remains so far unclear. The CYP2C9*5 variant is derived from a C1080G transversion in exon 7 that leads to Asp360Glu amino acid substitution. In vitro data suggest that CYP2C9*5 carriers eliminate CYP2C9 substrates at a slower rate compared with those expressing the wild-type protein [7]. The CYP2C9*11 genotype is derived from a C1003T mutation in exon 7, resulting in a Arg335Thr substitution ( Distribution of the CYP2C9*11 polymorphism in Caucasian or Black populations is currently unknown, as is the distribution of the CYP2C9*2, *3, *4 and *5 genotypes in the West Black African population.

Cytochrome P450 2C19 (CYP2C19) metabolizes a number of drugs, such as S-mephenytoin, omeprazole, propranolol, and imipramine. Several polymorphisms of the CYP2C19 gene have been identified and produce an inactive enzyme. Two variant alleles account for the majority of the poor metabolizer (PM) phenotypes, namely CYP2C19*2 which carryies a G→A nucleotide substitution in exon 5 resulting in an aberrant splice site, and CYP2C19*3 which carries a G→A nucleotide substitution at position 636 in exon 4 and produces a premature stop codon. Whereas CYP2C19*2 appears to be the most prevalent allele associated with the PM phenotype [8], CYP2C19*3 is mainly found in Orientals [9].

The aim of this study was to assess the distribution of CYP2C9 and selected CYP2C19 genotypic variants in the Beninese population and to compare these data with the distribution observed in the Belgian Caucasian population.



One hundred and eleven unrelated indigenous Beninese healthy adults were recruited. They all lived in the Zou department area of the country. Written informed consent was obtained from all participants. The study protocol was approved by the Ministry of Public Health of Benin and by Ethic Review Committee of Medical Faculty of National University of Benin (UNB).

Blood samples from 121 white Belgian Caucasian subjects were collected in Saint-Luc Hospital (Brussels, Belgium), in accordance with a protocol approved by the local ethics committee.

Blood collection and DNA isolation

For each subject, venous blood samples (5 ml) were collected in EDTA tubes. Genomic DNA was extracted using the Qiagen Qiamp DNA Blood kit (Qiagen, Leusden, the Netherlands), according to the manufacturer's protocol.


CYP2C19*2 and CYP2C19*3 genotyping was performed according to the single-tube tetra-primer polymerase chain reaction (PCR) assay method as described previously [10]. A new single-tube tetra-primer PCR assay was developed to detect the CYP2C9*2 allele (Table 1). Primers 3S and 3AS complementary to exon 3 and allele-specific primers matching the wild-type (3IS) or the *2 allele (3IAS) were mixed together. The following reaction mixture was used: 5 µl of PCR buffer (10 mmol l−1 Tris hydrochloride, pH 8.3 and 50 mmol l−1 potassium hydrochloride), 2.25 mm of MgCl2, 1 U of Gold Taq, 200 µm of each deoxyribonucleoside triphosphate (dNTPs), 10 µm of each primer except for primer 3S (3 µm), and 250 ng of genomic DNA in a final volume of 50 µl. Cycling conditions were as follows: 5 min at 95 °C; 35 cycles of 95 °C for 30 s, 55 °C for 30 s, and 72 °C for 60 s; and a final extension of 7 min at 72 °C. The PCR products were separated by 2% agarose gel electrophoresis. The expected size of 3S-3AS amplicon, used as internal control for the quality of the PCR amplification and as template for the allele-specific amplification (ASA), was 367 bp. The expected sizes of the ASA amplicon for CYP2C9*2 and wild-type alleles were 243 and 156 bp, respectively.

Table 1. CYP2C9 and CYP2C19 genotyping: primer selection.
  • *

    Nucleotide positions according to GenBank Accession no. L32982 (exon 4), L31506 (exon 5), L16878 (exon 3) and L16881 (exon 7).

  • From Xiao et al.[22].

  • From Hersberger et al.[10].

  • §

    §Nucleotides that correspond to the target point mutation.

 Exon 33S 3AS 31S 31AS5′-GGAGGATGGAAAACAGAGAC-3′ (337–356)
5′-AGGAGCATTGAGGACC-3′§-3′ (548–563)
5′-GCTTCCTCTTGAACACA-3′§-3′ (579–563)
5ISc5′-ATCATTGATTATTTCCCA §-3′ (100–117)
4IS5′-GTAAGGACCCCCTGA §-3′ (220–234)
4IAS5′-GGCCTTACCTGGATC §-3′ (248–234)

Genotypes identified using the new single-tube tetra-primer assays were confirmed by sequence analysis. Identification of the CYP2C9*3, *4, *5 and *11 alleles was performed by sequence analysis. The primers used for amplification of CYP2C9*5 are shown in Table 1. For this allele, 250 ng genomic DNA was amplified in a final volume of 50 µl using 2.25 mm of MgCl2, 1 U of Gold Taq, 5 µl of PCR buffer, 200 µm of each dNTPs, 10 µm of 7S, 10 µm of 7AS. PCR conditions for CYP2C9*5 consisted of an initial denaturation of 5 min at 95 °C, followed by 40 cycles (40 s at 95 °C, 40 s at 60 °C and 1 min 20 s at 72 °C) and a final elongation of 7 min at 72 °C. The amplicons were first purified with Microcon (Amicon Millipore Corporation, Bedford, MA 01730, USA) columns. Sequence analysis was performed in both orientations with Big Dye terminators (PRISM; Applied Biosystems, Nieuwerkerk, the Netherlands), using an automated ABI 3100 capillary sequencer (Applied Biosystems).

Statistical analysis

Allele distributions were compared using χ2 and Fisher's exact tests. P ≤ 0.05 was considered significant. These analyses were performed with SPSS (SPSS Inc. Chicago, USA) for Windows (version 10.00).


CYP2C9 and CYP2C19 allele and genotype frequencies in both populations are summarized in Table 2. Expression of the genotypes relies upon the assumption that the different mutations are not linked on the same allele [4]. Accordingly, we observed five different non homozygous wild-type CYP2C9 genotypes in the 121 Belgian individuals: CYP2C9*2/*3 (n = 2); CYP2C9*3/*3 (n = 1). CYP2C9*1/*2 (n = 22), CYP2C9*1*3 (n = 14) or CYP2C9*1*11 (n = 1). The remainder were homozygous wild-type CYP2C9*1*1 (n = 81). In the 119 Beninese, four were heterozygous for the CYP2C9*5 allele and six for the CYP2C9*11 allele.

Table 2.  Allele frequencies (a) and genotype prevalences (b) for CYP2C9 and CYP2C19 in a Beninese and Belgian Caucasian population. :
 Frequency in Belgian population % (95% CI)Frequency in Beninese population % (95 CI)95% CI on the differenceP-value
  1. CI, Confidence interval; NS, no significant differences (P > 0.05). CYP2C9*1 and CYP2C19*1 wild-type alleles frequencies have been inferred from the observed frequencies of mutated alleles, as assessed in the study. Calculation of allele frequencies was made on the assumption of compound heterozygosity when two polymorphisms were observed.

a Variant allele CYP2C9 allele
CYP2C9*182.2 (76.7, 86.7)95.5 (91, 97.3) 0.07, 0.19<0.001
CYP2C9*210.0 (6.6, 14.5) 0−0.14, − 0.06<0.001
CYP2C9*3 7.4 (4.6, 11.7) 0−0.11, − 0.04<0.01
CYP2C9*4 0 0−0.01, 0.01NS
CYP2C9*5 0 1.8 (0.5, 4.8) 0.00, 0.04  0.05
CYP2C9*11 0.4 (0.0, 2.6) 2.7 (1.1, 6) 0.00, 0.05<0.05
CYP2C19 allele
CYP2C19*190.9 (86.4, 94.1)87   (81.6, 91)−0.10, 0.02NS
CYP2C19*2 9.1 (6, 13.6)13   (9, 18.4)−0.02, 0.10NS
CYP2C19*3 0 0−0.01, 0.01NS
b Genotype
CYP2C9*1/*167   (57.7, 75.0)91   (83.6, 95.3) 0.13, 0.34<0.00001
CYP2C9*1/*218.2 (12, 26.4) 0−0.26, − 0.12  0.0001
CYP2C9*1/*311.6 (6.7, 19) 0−0.18, − 0.06  0.001
CYP2C9*1/*5 0   (0, 3.8) 3.6 (1.1, 9.5) 0.00, 0.09  0.05
CYP2C9*1/*11 0.8 (0.04, 5.2) 5.4 (2.2, 11.8) 0.00, 0.10  0.05
CYP2C9*2/*3 1.6 (0.3, 6.4) 0−0.06, 0.02NS
CYP2C9*3/*3 0.8 (0.04, 5.2) 0−0.04, 0.02NS
CYP2C19*1/*183.5 (75.4, 89.4)73.9 (64.5, 81.5)−0.20, 0.01NS
CYP2C19*1/*214.9 (9.3, 22.7)26.1 (18.4, 35.4) 0.00, 0.22<0.05
CYP2C19*1/*3 0 0−0.03, 0.03NS
CYP2C19*2/*2 1.6 (0.3, 6.4) 0−0.06, 0.02NS
CYP2C19*2/*3 0 0−0.03, 0.03NS
CYP2C19*3/*3 0 0−0.03, 0.03NS

There was a statistically significant difference in the allelic distribution of CYP2C9*2 and CYP2C9*3 between Beninese Africans and Belgian Caucasians. The CYP2C9*5 and CYP2C9*11 alleles were statistically more common among Beninese compared with Belgians.

Eighteen of the 121 Belgians had the CYP2C19*2/CYP2C19*1 genotype, and two were homozygous for CYP2C19*2, and would be predicted to be PMs. Allelic distributions were similar in the Beninese and Belgian subjects.


Whereas CYP2C genetic polymorphisms have previously been assessed in Caucasians and Orientals, data in West African populations are lacking. As expected, the frequencies of the CYP2C9*2, *3, *4 and *5 alleles in the Belgian Caucasian subjects are similar to those of other Caucasian populations [5, 11, 12]. The absence of a CYP2C9*2 allele in the Beninese group differs significantly from its frequency of 2.5% (P = 0.03) reported in 120 African-Americans by Dickmann et al.[7], whereas the frequency of CYP2C9*3 was similar in the two groups (P = 0.2). The frequencies of CYP2C9*2 and *3 differ significantly from those in 150 Ethiopians [4% (P = 0.001) and 2% (P = 0.02), respectively][13], and from those in 247 Egyptians [12% (P < 0.001) and 6% (P < 0.001), respectively][14].

These discrepant data suggest genetic heterogeneity among African populations. The absence of CYP2C9*2 was reported previously in Oriental and Inuit populations [11, 15, 16], and the latter did not also possess the CYP2C9*3 allele [11]. In contrast, current and previous data suggest that CYP2C9*2 and *3 alleles are found more frequently in Caucasian (6–13%) than in Black populations [4, 5, 7, 11, 12, 17].

The frequency of CYP2C9*5 allele in Beninese (1.8%) appears slightly higher than in Tanzanians (0.82%) [12], but is similar to the frequency reported in African Americans (1.7%) [8]. CYP2C9*5 is not found in the Belgian or other Caucasian populations [12].

Whereas the functional importance of CYP2C9*11 has not been assessed, it is thought that the activity of CYP2C9 in heterozygous genotypes is less than in homozygous wild-type genotypes [17]. Individuals homozygous for the CYP2C9*3 allele show markedly decreased metabolic capacities for most CYP2C9 substrates [1, 3, 5]. The frequency of these PM subsets (0.8%) in the Belgian is similar to that in other Caucasian populations [4, 5, 13, 17].

Our data on CYP2C19*2 and *3 allele frequencies are also in agreement with previous data obtained from Zimbabweans and Tanzanians [18], as well as in Dutch [19], French [20], German [21] or Egyptian populations [14]. In the Venda (South Africa), CYP2C19*2 was found at a high frequency (21.7%) [18], but CYP2C19*3 was also absent. The frequency of CYP2C19*3 was reported to be higher (between 5 and 11%) in Oriental populations [2, 14, 22]. Considering that CYP2C19*2 and *3 account for approximately 83% of all PM alleles in Caucasians [8], we estimate a PM phenotype frequency of approximately 2% in the Belgian population. In contrast, we found no homozygous or compound heterozygous CYP2C19 mutations in the Beninese population. These findings suggest that the prevalence of PM phenotype with respect to CYP2C19 in the Beninese could be lower than in Belgians.

In conclusion, frequencies of CYP2C9*2 and *3 alleles appear higher in Caucasian than in Black populations. Conversely, the frequency of CYP2C9*5 allele is higher in the Beninese population than in Belgians or other Caucasian populations. Likewise, the frequency of the CYP2C9*11 allele is significantly higher in Beninese than in Belgian Caucasians. The allelic distributions of CYP2C19*2 and *3 are comparable in both ethnic groups.

This study was supported by the Walloon Region (Waleo 2002, Grant no. 215131/0938840). We also thank the Belgian Ministry of Defence and The Belgian Military Air Forces for their help in transporting samples from Benin to Belgium for analysis.