Genotyping of CYP2C9 and VKORC1 in the Arabic Population of Al-Ahsa, Saudi Arabia

Polymorphisms in the genes encoding CYP2C9 enzyme and VKORC1 reductase significantly influence the dose variability of coumarinic oral anticoagulants (COAs). Substantial inter- and intraethnic variability exists in the frequencies of CYP2C9 ∗2 and ∗3 and VKORC1 –1639A alleles. However, the prevalence of CYP2C9 and VKORC1 genetic variants is less characterized in Arab populations. A total of 131 healthy adult subjects from the Al-Ahsa region of Saudi Arabia were genotyped for the CYP2C9 ∗2 and ∗3 and VKORC1 –1639G>A polymorphisms by PCR-RFLP method. The frequencies of the CYP2C9 ∗2 and ∗3 and VKORC1 –1639A alleles were 13.3%, 2.3%, and 42.4%, respectively, with no subjects carrying 2 defective alleles. The frequencies of the CYP2C9 ∗3 and VKORC1 –1639A alleles were significantly lower than those reported in different Arabian populations. None of the subjects with the VKORC1 –1639AA genotype were carriers of CYP2C9 ∗1/∗3 genotypes that lead to sensitivity to COAs therapy. The low frequency of the CYP2C9 ∗3 allele combined with the absence of subjects carrying 2 defective CYP2C9 alleles suggests that, in this specific population, pharmacogenetic COAs dosing may mostly rely upon VKORC1 genotyping.


Introduction
Pharmacogenomics is the first step towards personalized medicine and is a promising field of investigation that may explain some of the interindividual variations in responses to various classes of drugs [1,2]. Pharmacogenomics can be applied to all fields of medicine. In particular, recent advances in genotype-phenotype associations in cardiology point towards the application of pharmacogenomics to oral coumarinic anticoagulants (COAs), including warfarin, acenocoumarol, and phenprocoumon, in routine clinical practice [3].
Among other factors, interindividual COAs dose variability is significantly influenced by variations in the genes encoding two enzymes: cytochrome P450 2C9 (CYP2C9), the enzyme that metabolizes COAs, and vitamin K epoxide reductase (VKORC1), the pharmacologic target of these drugs [3].
Polymorphisms in the CYP2C9 gene seriously affect the enzymatic activity of the encoded CYP2C9 protein. Based on phenotype, populations can be divided into extensive (EM), intermediate (IM), and poor metabolizers (PM), and more than 35 different allelic variants have been identified in the CYP2C9 gene [4]. Among these alleles, the 2 9 * 2 (rs1799853) and * 3 (rs1057910) variants, which reduce CYP2C9 enzymatic activity, allow for the prediction of more than 85% of PMs [5]. In addition, it has recently been shown that VKORC1 gene polymorphisms also affect COAs dosing requirements [6]. The VKORC1 -1639G>A polymorphism (rs9923231) is located in the promoter of the VKORC1 gene and results in reduced promoter activity and lower mRNA levels, which lead to lower levels of synthesized protein and eventually the reduced production of active clotting factors in subjects with the AA genotype [6].
Whereas the 2 9 * 2 and * 3 alleles affect coumarin pharmacokinetics, the VKORC1 -1639G>A polymorphism affects the pharmacodynamic response to coumarins [3]. It has been reported that polymorphisms in the CYP2C9 and VKORC1 genes together account for 35%-50% of the variability in COAs dose requirements for initiation and maintenance [7]. Moreover, carriers of the 2 9 * 2 or * 3 alleles and the VKORC1 -1639G>A polymorphism are at higher risk for bleeding and require lower mean daily doses [3]. These associations between genotype and COAs response led the U.S. Food and Drug Administration (FDA) to release a warning in the warfarin insert to indicate the range of expected therapeutic warfarin dosages based on CYP2C9 and VKORC1 genotypes [8]. Furthermore, the clinical feasibility of incorporating CYP2C9 and VKORC1 genotyping-based COAs dosing regimens into routine clinical practice is being tested in large prospective clinical trials [9][10][11].
Few data are available concerning the prevalence of the 2 9 * 2 and * 3 alleles and the VKORC1 -1639G>A polymorphism in distant populations of Saudi Arabia where social and in some areas religious beliefs favor consanguineous marriages and therefore limit genetic flow. In the Al-Ahsa region, which is part of the eastern province of Saudi Arabia, the rate of consanguineous marriage has been reported to be 59.1%, with marriages between first-degree relatives at 40% [22]. The reported inter-and intraethnic variability in the frequencies of the 2 9 * 2 and * 3 and VKORC1 -1639A alleles was our rationale for studying the incidence of these variant alleles in the Al-Ahsa population. The results of this investigation will be critical for the coming era, in which genotype-guided dosing algorithms will be increasingly utilized to guide the prescription of COAs [9][10][11].
The aim of the present study was to investigate the frequency of the 2 9 * 2 and * 3 alleles and the VKORC1 -1639G>A polymorphism as well as the number and percentages of individuals with genotypes predictive of COAs response in a representative sample of the population of Al-Ahsa, Saudi Arabia. We also sought to compare the data obtained with existing published data for other populations residing in a wider area of the Middle East.

Subjects.
All participants were of Arabian origin and residents of the Al-Ahsa urban area, which is located on the east coast of Saudi Arabia. The study protocol was approved by the Ethics Committee of King Faisal University, Hofouf, Saudi Arabia. All of the study participants were nonrelative volunteers and provided informed consent. A total of 131 healthy adult subjects (70 males and 61 females) were genotyped to determine the frequencies of the 2 9 * 2 and * 3 alleles and the VKORC1 -1639G>A polymorphism. The mean (±SD) age of the subjects was 25 (±7) years (range: 19-52 years). The majority of the subjects (57%) reported to originate from a consanguineous marriage.

Genotyping.
Genomic DNA was extracted from peripheral blood leukocytes using the QIAamp DNA Blood Mini Kit (Qiagen, Germany) according to the manufacturer's instructions. All subjects were genotyped for the 2 9 * 2 and * 3 alleles and the VKORC1 -1639G>A polymorphism using PCR-restriction fragment length polymorphism (RFLP) protocols, as previously described [13,20]. PCR amplifications were performed in duplicate by two independent researchers in an MJ Research PTC-200 thermocycler (Watertown, MA, USA). To ensure the accuracy of the results, an internal positive control was utilized for each polymorphism (rare allele) in each PCR-RFLP run.

Statistics.
Data were analyzed using the Statistical Package for Social Sciences (SPSS) version 17.0 and are presented as the medians with 95% confidence intervals. Departure from the Hardy-Weinberg equilibrium was estimated using an exact 2-sided probability test using the formula provided by Weir [23]. Allele frequencies were compared to other ethnic population utilizing the two-tailed Fisher's exact test [24].

Results
The distributions of genotypes and alleles of 2 9 * 2 and * 3 and VKORC1 -1639G>A polymorphisms in the studied population are shown in Table 1.
We also analyzed the data for potential gender differences. However, there were no significant differences in the genotype frequencies of the 2 9 * 2 and * 3 alleles and the VKORC1 -1639G>A polymorphism in our study group of 131 Saudi Arabian subjects (70 male and 61 female) (data not shown).
Finally, to investigate possible differences between individuals who originated from consanguineous marriages and those who did not, we analyzed the distribution of the 2 9 * 2 and * 3 alleles and the VKORC1 -1639G>A polymorphism according to this factor. However, the frequencies of the genotypes and alleles studied did not differ with respect to origin from a consanguineous marriage (data not shown).

Discussion
The pharmacogenomics of COAs is one field that is most ready to apply genotype-guided dosing in clinical practice. It has been estimated that polymorphisms in the CYP2C9 and VKORC1 genes together account for 35%-50% of the variability in COAs initiation and maintenance dosage requirements [20,40]. Thus, efforts are focused on incorporating this knowledge into the dosing regimens currently used, and genotype-based algorithms are currently being tested in large randomized trials to validate the accuracy, safety, and cost effectiveness of incorporating CYP2C9 and VKORC1 genotypes into the optimization of anticoagulant therapy [3,9].
In the era of developing and testing genotype-guided COAs dosing algorithms, there remain populations in which the frequency of the major CYP2C9 and VKORC1 polymorphisms has not been assessed. Thus, potential differences in the prevalence of CYP2C9 and VKORC1 genetic variants in different populations may lead to adjustments of genotypebased COAs dosing algorithms or may serve as motivation to identify novel genetic variants that influence COAs therapeutic responses. Towards this goal, the current study reported the frequency of the 2 9 * 2 and * 3 alleles and VKORC1 -1639G>A polymorphism in a distant population residing on the east coast of Saudi Arabia. Some of the CYP2C9 variants that lead to decreased enzymatic activity, such as 2 9 * 5, * 6, and * 11, were not included in this study because they have not been observed in Caucasian or Middle Eastern populations, in contrast to their higher frequency in African populations. Although consanguineous marriage is traditionally favored among Saudis, genetic inflow in Al-Ahsa is further limited due to social and religious beliefs, which could have led to the differences in the prevalence of the CYP2C9 and VKORC1 genotypes and alleles studied.
In general, the genetic characteristics of Arabian populations are not well characterized. In the case of CYP2C9,  [39] there are scattered reports on CYP2C9 frequencies in Arabs in general and Saudis in particular, whereas to the best of our knowledge, this is the first report of the frequency of the VKORC1 -1639G>A polymorphism in a population in Saudi Arabia. Among Arabs in general, there exists great intraethnic variability in the frequencies of the 2 9 * 2 and * 3 alleles. In Saudi Arabia, Mirghani et al. [14] reported that the frequencies of the 2 9 * 2 and * 3 alleles among Saudis residing in Riyadh were similar to those in Caucasian populations (11.7% and 9.1%, resp.) [14]. Unpredictably, the frequency of 2 9 * 3 was significantly different from the frequency of * 3 in our study group (Table 3). In the Omani population, 2 9 * 2 and * 3 allele frequencies are markedly lower and have been estimated at 7.4% and 2.9%, respectively [17]. The later was the closest Among Arab populations to the frequency of 2 9 * 3 in our study subjects (2.3%). In the Egyptian population, the frequencies of the 2 9 * 2 and * 3 alleles have been estimated to be 12% and 6%, respectively [16]. In Lebanese individuals, the frequencies of 2 9 * 2 and * 3 have been reported to be 11.2% and 9.6%, respectively [18].
In our study population, the 2 9 * 2 allele frequency (13.4%) was similar to that reported for other Arabian and Caucasian populations (Table 3). However, we found a significantly reduced frequency of the 2 9 * 3 allele (2.3%). In addition, genotypes predicting the CYP2C9 PM phenotype (i.e., 2 9 * 2/ * 2, 2 9 * 2/ * 3, and 2 9 * 3/ * 3) were absent in the subjects studied. One possible explanation for this finding is that our study population comprised residents of Al-Ahsa, where the population can be divided based on religious beliefs into two main populations (Sunni and Shiaah) between which intermarriage rarely occurs. Moreover, within each of the two populations, social customs may further limit intermarriage between nonrelatives, further resulting in decreased genetic inflow.
Although we did not find statistically significant differences in the distribution of 2 9 * 2 and * 3 alleles and VKORC1 -1639G>A polymorphisms in our study group according to presence or absence of consanguineous marriage, different frequencies in variants predicting low CYP2C9 enzymatic activity should be expected in similar populations elsewhere in Saudi Arabia.
Regarding the VKORC1 -1639G>A polymorphism, we found that the frequencies of genotypes and alleles were similar to those reported in Caucasian populations. The VKORC1 -1639AA genotype and VKORC1 -1639A allele frequencies