Design and implementation of a novel pharmacogenetic assay for the identification of the CYP2D6*10 genetic variant

Tamoxifen is considered to be the most widely used adjuvant therapy for hormone receptor positive breast cancer in premenopausal women. However, it is reported that nearly 30% of patients receiving tamoxifen therapy have shown reduced or no benefits. This may be due to the high inter-individual variations in the CYP2D6 gene that is involved in tamoxifen metabolism. The CYP2D6*10 gene variant (rs1065852C>T) is reported to be commonly found in Asian and South Asian populations. The present study was undertaken to design a novel pharmacogenetic assay (Single step-Tetra Arms Polymerase Chain Reaction) for the identification of the CYP2D6*10 variant and implement the designed assay by genotyping a cohort of breast cancer patients. The novel assay was successfully designed, optimized and validated using Sanger sequencing. Blood samples from 70 patients were genotyped. The following bands were observed in the gel image: Control band at 454 bp; band for C allele at 195 bp; band for T allele at 300 bp. The genotype frequencies for the CYP2D6*10 (rs1065852C>T) variant were: CC-24.28% (17/70), CT-75.71% (53/70), TT-0% (0/70). The allele frequencies were: T-allele-37.86% and C-allele-62.14%.


Introduction
Breast cancer is the most common cancer among women and overall, it is the second most common cancer in the world [1]. In Sri Lanka, it is the most common cancer among females with a prevalence rate of 27% [2]. Although tamoxifen has been the drug of choice to treat hormone receptor positive breast cancer in premenopausal women, nearly a one third of patients treated with adjuvant tamoxifen therapy do not obtain the desired benefits. It is reported that there is a substantial interindividual variability in the response to this drug and the beneficial and adverse effects also appear to be variable and unpredictable for individual patients [3]. Variants in the CYP2D6 gene have been reported to affect the metabolism of tamoxifen in breast cancer patients resulting in variable responses to the drug due to reduced enzyme activity. Tamoxifen is converted to its more active form endoxifen by the CYP2D6 enzyme [4]. However, due to marked variability in the CYP2D6 genotypes among population groups, it has been difficult to formulate a common guideline for using the genotype as a determinant factor in prescribing tamoxifen. Although many pharmacogenomic studies have been conducted to evaluate the possibility of incorporating CYP2D6 genotyping into the therapeutic management of hormone receptor positive breast cancer, conflicting results have posed as a barrier to achieving this. In order to support the establishment of pharmacogenomic testing for tamoxifen Open Access BMC Research Notes *Correspondence: nadeeka@kiu.ac.lk Department of Anatomy, Genetics and Biomedical Informatics, Faculty of Medicine, University of Colombo, Colombo 08, Sri Lanka therapy in breast cancer patients, it is fundamental to establish prevalence data of the most common variants in under-represented populations. Existing data reports CYP2D6*10:rs1065852C > T as one of the most common variants in the South Asian and Sri Lankan populations that is known to affect tamoxifen metabolism. Knowledge of the prevalence of this variant would be of importance to gain the maximum benefit associated with the use of this drug [5]. The current study was undertaken to design a novel pharmacogenetic assay for easy detection of the CYP2D6*10:rs1065852C>T variant and implement the designed assay by genotyping a cohort of Sri Lankan breast cancer patients.

Methods
The study was conducted at the Human Genetics Unit, Faculty of Medicine, University of Colombo. It was an experimental study, where a novel assay was designed for the targeted variant, and a cohort of hormone receptor positive breast cancer patients were genotyped for the CYP2D6*10 variant using the optimized and validated assay. An existing resource of stored venous blood samples obtained from 70 hormone receptor positive breast cancer patients which were collected into EDTA tubes and stored at − 20 °C were used for the present study.
The primers for the CYP2D6*10 rs1065852 variant were newly designed using web-based tools. CYP2D6 gene sequence was obtained from the NCBI (CYP2D6 ID: 1565). The Primer1 is a special software designed for obtaining primer tetrads that are specific for T-ARMS PCR technique [6]. The Primer1 software was operated through the web interface where an input of a target DNA sequence is enabled. The maximum size of the DNA sequence is up to 1000 bases in the 5′ to 3′ direction. Serial cloner enabled cutting a 1000 bp fragment (which included the variant in the middle) that was inserted to Primer1. Other parameters were also considered for primer design: the desired minimum, maximum and average melting temperatures (°C) for a primer oligo; the millimolar concentration of salt and the nanomolar concentration of annealing oligos in the PCR (this is used in the calculation of T m -melting temperature); the % GC content (the minimum and maximum allowable for any primer); the primer lengths and the minimum and maximum complementarity and product sizes [7]. In par with the selected parameters, the program generated primer sequence sets. To select the optimum primer sets the melting temperatures were considered for each primer set (inner and outer). NCBI primer blast was used to check the specificity of the primers. The details of the designed set of primers, and their product sizes are given in Table 1.
Genomic DNA was extracted using QIAamp ® DNA mini kits (Qiagen Ltd., UK) according to the manufacturer's protocol. The Thermal Cycler (Bio-Rad) was used to conduct gradient PCR for the targeted variant CYP2D6*10. The primer annealing temperatures mentioned in the primer vials by the manufacturer were considered for the annealing temperature gradient in the first attempt at PCR optimization. After a series of optimization attempts, the final optimized cyclic conditions for this variant consisted of an initial denaturation of 3 min at 94 °C (initial denaturation), followed by 35 PCR cycles of 1 min at 94 °C (denaturation), 30 s at 65 °C (annealing temperature), 1 min at 72 °C and the final extension of 5 min at 72 °C. The optimized PCR master mixture con-  (homozygous wild type) and Figure S1D (homozygous wild type) show the Sanger sequencing electropherogram images of the variant]. The bands for each genotype were observed in the gel images: Control band at 454 bp; band for C allele at 195 bp; band for T allele at 300 bp. The genotype frequencies were determined using the data (allele counts) observed from the gel images. They were systematically recorded, and genotype frequencies for homozygous wild-type (CC), heterozygous variant (CT), and homozygous variant (TT) genotypes and the allele frequencies for C (wild type) and T (variant) alleles were calculated using an online calculator (available from: https:// wpcalc. com/ en/ equil ibrium-hardy-weinb erg/) and tested for Hardy-Weinberg equilibrium. Standard descriptive statistics were used to analyze the demographic data.

Results
All patients in the sample cohort that was genotyped were females with hormone receptor positive breast cancer. Their mean age was 60.12 ± 6.56 years.
T-ARMS PCR protocol was successfully designed and implemented for the identification of the CYP2D6*10 variant. The expected gel band patterns were observed in the gel image results (Control band: 454 bp; band for C allele: 195 bp; band for T allele: 300 bp). Under Additional file information, Additional file 2: Figure S2 shows the expected gel band patterns for the targeted variant. Figure 1 shows an actual gel image of the genotype results.
The gel images of all 70 patient samples (with the sample numbers indicated in table) are given in Additional file 3: Figure S3.
The genotype and allele frequencies are shown in Table 2.
Out of the 70 samples which were genotyped, homozygous wild type (CC) and heterozygous (CT) samples were detected but no homozygous samples were detected for the variant T allele ( Table 2). The predominant genotype was CT (75.7%) and the minor allele frequency of the CYP2D6*10 variant in this cohort was 37.9%. The results of all the genotyped samples are shown in Additional file 4: Table S1.

Discussion
The T-ARMS PCR is a technique which many researchers are beginning to adopt as it is a cost-effective method compared to other techniques that are used to detect SNVs [8,9]. However, the optimization step can be timeconsuming and laborious [9]. The presence of GC rich regions near the SNV of interest might be a limiting factor for the use of this technique [10]. One of the common  problems faced with T-ARMS PCR is the specificity, where the MgCl 2 concentration, primer ratio, and dNTP ratio can play a vital part in the optimization process [10]. Different gradients of MgCl 2 , dNTP and annealing temperatures were used during the optimization process in this study. The use of Betaine [10] and Dimethyl sulfoxide (DMSO) [11] is also commonly reported across studies that have used this technique. In this study, Betaine was used in a lower concentration and volume during the optimization process. The CYP2D6*10 variant has been studied in different populations and the present study is an addition to this data from the South Asian region. The genotyping results obtained were comparable to data available from a previous Sri Lankan study conducted by Tharanga et al. in 2013, where the CYP2D6*10 minor allele frequency in a Sri Lankan population (n = 75) was reported to be 39%, similar to the 37.9% obtained in the cohort of 70 hormone receptor positive breast cancer patients genotyped in this study. The study by Tharanga et al. further identified that the minor allele frequency was highest among the Moors and 50% of the participants in that study were males [12]. The current study did not enumerate the ethnic specific variations since the ethnicity of the breast cancer patients was not available for subgroup analysis.

Conclusions
A novel T-ARMS PCR assay was successfully designed, optimized and validated to genotype the CYP2D6*10 variant and could be implemented as a cost-effective technique. Homozygous wild type (CC) and heterozygous variant (CT) alleles in CYP2D6*10 were found in the genotyped samples. Homozygotes for the variant allele were not identified in this cohort.

Limitations
• The T-ARMS PCR assay development is challenging specially when the target region has a high GC percentage which could lead to prolongation of the optimization process. • With the use of Betaine, the intensity of the control and other bands may reduce, this could cause difficulty in detecting the relevant bands on the gel image. • Primer dimer formation. • The small sample size used for genotyping in this study. A larger study would be ideal for frequency estimation of the targeted variant in the population.