Distribution of <i>CYP2C</i> Polymorphisms in an Amerindian Population of Brazil

The human cytochrome P450 (CYP450) superfamily members CYP2C8, CYP2C9 and CYP2C19 are drug-metabolizing enzymes that provide the main pathway for the biotransformation of various therapeutic classes, including taxanes and antimalarials (CYP2C8) [1,2] anticoagulants, anticonvulsants, non-steroidal anti-inflammatory drugs (CYP2C9) [3–5], proton pump inhibitors and thienopyridine antiplatelet medicines (CYP2C19) [6,7]. The encoding CYP2C genes, located in a cluster on chromosome 10q23–24, are highly polymorphic (http://www.cypalleles.ki.se), and several functional variants have been shown to affect the clinical efficacy, toxicity and/or dose requirement of CYP2C substrates, notably clopidogrel and warfarin [8]. The frequency of distinct CYP2C polymorphisms varies significantly across and within continental populations [9,10] (http://hapmap.ncbi.nlm.nih.gov/cgi-perl/gbrowse/hapmap3r3_B36/) and also within different strata of admixed populations, such as Brazilians [11]. For example, the reduced activity of CYP2C8*2 allele [12] is absent in Japanese [13], rare (0–1%) in European populations [14], ranges in frequency from 11 to 17% in sub-Saharan Africans [15] and occurs in 2% and 12% of Brazilians, self-identified as White or Black, respectively [11]. A PubMed search on 31 August, 2011 retrieved no frequency data on CYP2C8 polymorphisms and a paucity of information on CYP2C9 and CYP2C19 genetic variation in native American (Amerindian) populations [16–21]. This prompted us to examine the distribution of twelve functional CYP2C polymorphisms in Guarani, an endogamous Amerindian group belonging to the Tupi-Guarani linguistic family living in the south, south–east and centre-west regions of Brazil and in neighbouring regions of Paraguay, Argentina and Bolivia [22]. Study population. Three Guarani groups, namely Guarani-Kaiowá, Guarani-Ñandeva and Guarani-M′byá, were contacted in the indigenous reservation areas named Amambai (23°06′S, 55°12′W), Limão Verde (23°12′S, 55°06′W) and Porto Lindo (23°48′S, 54°30′W), located in Mato Grosso do Sul State, centre-west region and in Rio das Cobras (25°18′S, 52°32′W), located in Paraná State, South region, Brazil. Ninety healthy, unrelated individuals (54 women/36 men), aged 27.5 ± 7.2 years, 30 from each Guarani group, were recruited in the context of a population genetics study of Amerindians living in Brazil, approved by the Brazilian National Ethics Committee [23]. The study cohort had been previously genotyped for polymorphisms in VKORC1 [24]. CYP2C genotyping. DNA was extracted from peripheral blood by standard salting out and phenol/chloroform/isoamyl alcohol methods. Table 1 lists the 12 SNPs in CYP2C8, CYP2C9 and CYP2C19, genotyped in this study. Allele discrimination at each locus was performed on a Fast 7500 Real-Time System (Applied Biosystems, Foster City, CA, USA) using TaqMan assays purchased from Applied Biosystems, following the manufacturer’s protocols, with reaction volumes reduced to 7 μl. Allele frequency of the CYP2C polymorphisms was derived by gene counting. Linkage disequilibrium and haplotype analyses. The ARLEQUIN software version 3.1 (available at http://lgb.unige.ch.arlequin/) [25] was used to estimate haplotype frequencies and linkage disequilibrium (LD) between all pairs of loci. Maximum-likelihood haplotype frequencies were estimated by the expectation–maximization (EM) algorithm. Haplotype frequencies were used to compute the LD coefficient standardized by the maximum value it can take (D’), given the allele frequencies [26] and r2, the square of the correlation coefficient between allele frequencies, expressed as a function of D’. Table 1 shows the observed frequencies of the CYP2C variant alleles genotyped in the Guarani. Seven alleles, namely CYP2C19*3, CYP2C9*6, CYP2C9*11, CYP2C9*3, CYP2C9*5, CYP2C8*4 and CYP2C8*3, were not detected in the study cohort. CYP2C8*2 occurred in heterozygosis in one individual (frequency 0.6%), whereas two individuals were heterozygous for CYP2C19*17 and two other were heterozygous for CYP2C9*2 (frequency 1.1%). The remaining alleles, CYP2C19*2 and CYP2C9*8, displayed frequencies of 11.1% and 4.4%, respectively. Each of these alleles occurred in homozygosis in one distinct individual, and no significant deviations from HWE were observed at these loci in the overall cohort. Five haplotypes were inferred for the polymorphic loci with minor allele frequency of more than 1%: two haplotypes, namely GUA01 (wild type) and GUA02 (including allele CYP2C19*2), accounted for 95% of the genetic diversity in the study cohort. The D’ and r2 for LD between pairs of polymorphic loci were 1 and <0.004, respectively, for all pairs. It is interesting that the two Guarani having the CYP2C9*2 allele were not carriers of CYP2C8*3. This contrasts with the strong LD of these two alleles in Europeans [10,27], African Americans [28] and non-Amerindian Brazilians [11,29]. However, CYP2C9*2 and CYP2C8*3 were reported to be in low LD in South Indians [30]. The extent of LD between SNPs in the CYP2C cluster is known to vary across populations [31], as illustrated recently for CYP2C19*17 and CYP2C8*2 in populations of European versus African ancestry [32,33]. Comparison of our data for CYP2C8 in Guarani with other Amerindian populations is not possible, as this is the first study to genotype CYP2C8 polymorphisms in Amerindians. Regarding CYP2C9 and CYP2C19 polymorphisms, the observed frequency in our Guarani cohort agrees with some extent with published data for other Amerindian populations, namely Inuit and Canadian Native Indians (CNI) from Canada, Tepehuanos from Mexico [20] and combined Tupinikin plus Guarani from Brazil’s south–east region [21]. As shown in table 2, CYP2C19*3, CYP2C9*5 and CYP2C9*6 were absent, CYP2C9*2 was either absent or rare (<3%), and CYP2C19*2 was common (frequency range, 10.4–19.2%) in all groups investigated. CYP2C9*3 was absent in Guarani and Inuit but was detected in Tepehuano (1.5%) and CNI (6%). A large discrepancy, however, was observed with respect to the frequency of the gain-of-function CYP2C19*17 allele between our cohort (1.1%) and the combined Tupinikin plus Guarani sample (15.8%) studied by Santos et al. [21]. This discordance may reflect differences in composition and geographical origin of the study cohorts: we enrolled Guarani-Kaiowá, Guarani-Ñandeva and Guarani-M′byá from the centre-west and south regions, whereas Santos et al. [21] reported combined data for Guarani and Tupinikin groups living in the south–east Brazilian Coast. Genetic distance estimates have shown considerable divergence among extant Amerindian populations [23], and to our knowledge, there are no population genetic studies of the Guarani and Tupinikin groups studied by Santos et al. [21] which is in contrast to our cohort [23]. A greater extent of European and sub-Saharan admixture and consequently greater gene flow, in the combined Tupinikin and Guarani cohort compared to our study population, would contribute to the higher frequency of CYP2C19*17 in the former, as this allele is common in Europeans and sub-Saharan Africans [34]. Genetic drift is another possible factor for the difference in frequency of CYP2C19*17 in the two studies. Gene flow and genetic drift have been previously proposed to explain the large differences in frequency of other pharmacogenetic polymorphisms among Amerindians, notably in CYP1A1 and CYP2E1 (reviewed in [35]). The distinct CYP2C19*17 genotyping procedures used in our study (validated TaqMan® allele discrimination assay) and by Santos et al. [21] (high-resolution melting analysis) may also have contributed to the discordant CYP2C19*17 allele frequencies. In this regard, it is noteworthy that the CYP2C19*17 genotype distribution in both the Amerindian and ‘European-descent’ groups studies by Santos et al. [21] deviate significantly from HWE, whereas no significant deviation was observed in our study. A possibly related observation is that the CYP2C19*17 allele frequency in the ‘Mulatto’ and ‘African-descent’ groups of Santos et al. [21] is considerably higher than in self-identified brown and black individuals from a large, representative Brazilian cohort analysed by the Brazilian Pharmacogenetics Network [11]. Admixture of Amerindians with Europeans and sub-Saharan Africans has been observed for centuries, the extent of admixture varying considerably across the American continent [36,37]. The data shown in table 2 offer insights into the impact of admixture on the distribution of CYP2C polymorphisms in groups of predominant or extensive Amerindian ancestry such as ‘Mestizos’ from Bolivia [38] and Colombia [39], Mexicans (http://hapmap.ncbi.nlm.nih.gov/cgi-perl/gbrowse/hapmap3r3_B36/) and a selected group of Brazilians with estimated Amerindian ancestry >50% [40]. It can be seen that the frequency of CYP2C19*2 and, to a lesser extent, CYP2C9*3 in Amerindians overlaps those reported for the Amerindian-admixed populations listed in table 2. For other polymorphisms, notably CYP2C9*2, CYP2C8*3 and CYP2C19*17, the range of frequencies in the admixed groups is considerably higher than in Amerindians. Significantly, these three polymorphisms are relatively common in Europeans (22.0%, 15.3% and 11.9, respectively; http://hapmap.ncbi.nlm.nih.gov/cgi-perl/gbrowse/hapmap3r3_B36/). The historical evidence of extensive and gender biased (European men and Amerindian women) intermarriage between Amerindians and Europeans [36] provides a reasonable explanation for the higher frequencies of CYP2C19*17, CYP2C9*2 and CYP2C8*3 in the admixed groups of table 2. The notion that Asia is the likely origin of the first migrants into the American continent 25 000–15 000 years ago [36] led us to compare the frequencies of the CYP2C9 polymorphisms in Guarani with those reported for the Chinese (CHB) sample in HapMap3 (http://hapmap.ncbi.nlm.nih.gov/cgi-perl/gbrowse/hapmap3r3_B36/). Guarani and CHB differed by <1.2% in frequency of CYP 2C19*17, CYP2C9*2, *11 and *5, and CYP2C8*2, *3 and *4, but differed markedly in frequency of CYP2C19*2 (11%versus 33%), whereas CYP2C19*3 and CYP2C9*3 were absent in Guarani but detected in 4 and 5% of CHB individuals, respectively. We have previously reported significant differences in frequency of polymorphisms in other pharmacogenes (e.g. VKORC1 and CYP4F2) between Guarani and Asian [24,41] and ascribed these divergences to random genetic drift and/or a founder effect, i.e., a consequence of a small number of migrants colonizing new lands. A similar explanation may explain the discrepancies in the distribution of CYP2C polymorphisms. In conclusion, the present study provides the first analysis of CYP2C haplotypes in a Native American population, namely Guarani living in Brazil. From a pharmacogenetic perspective, these data suggest that CYP2C-poor metabolizers, i.e., individuals with two defective alleles of a given gene in the CYP2C cluster, are quite rare (<1%) or absent in Guarani, whereas intermediate metabolizers (one defective allele) of CYP2C19, CYP2C9 and CYP2C8 substrates account for 11%, 5.5% and 0.6%, respectively, of the study cohort. CYP2C19 ultrarapid metabolizers, i.e., individuals homozygous for the gain-of-function CYP2C19*17 allele [34], were not detected among the Guarani. Collectively, these genotype–phenotype inferences may be useful for pharmacogenetic-informed prescription of CYP2C19 (e.g. clopidogrel) and, to a lesser extent, CYP2C9 substrates (e.g. warfarin) in Guarani. For example, the relatively low frequency (4.4%) of CYP2C9 alleles with reduced- or null-activity, combined with the high frequency (59%) of the VKORC1 3673A allele in Guarani [24], suggests that this Amerindian group comprises high proportions of individuals requiring reduced warfarin doses. However, in the absence of phenotypic data on the warfarin dosage in Amerindians, this suggestion remains tentative. As a final caveat, the present data should not be interpreted as representative of extant Amerindian populations, in view of their remarkable diversity [36,37]. The research was supported by grants from Financiadora de Estudos e Projetos (FINEP 01.08.01230.00), Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) and Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ).

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.