Special RepoRt

951ISSN 1462-241610.2217/PGS.12.56 © 2012 Future Medicine Ltd Pharmacogenomics (2012) 13(8), 951–954

Special RepoRt

CYP2D6*11 and challenges in clinical genotyping of the highly polymorphic CYP2D6 gene

CYP2D6 is a clinically important gene encoding a phase I drug-metabolizing enzyme. CYP2D6 is a member of a superfamily of heme-dependent monooxygenases that use oxygen and NADPH to catalyze conversion of hydrogen (-H) with oxy-gen and reduced flavoprotein to -OH, water and oxidized flavoprotein. The enzyme is localized to the endoplasmic reticulum in the liver although some expression occurs in other tissues to a lesser degree. The majority of activity of the enzyme is in the liver [1]. The CYP2D6 gene is located on chromosome 22q13.1 and it is one of the most polymorphic CYP isoforms with nomenclature ranging from CYP2D6*1A to CYP2D6*105 [101]. CYP2D6 may be duplicated and multiplications of up to 13 copies on a single chromosome have been described [2]. A CYP2D7 pseudogene is located 5́ to CYP2D6 and it is highly homolo-gous to CYP2D6. Recombinant events between CYP2D6 and CYP2D7 have given rise to a variety of hybrid alleles (e.g., CYP2D6*4N, *13, *36, *68 and others). The gene can also be deleted as in the case of CYP2D6*5. Allelic frequencies vary widely among ethnic groups. In addition, enzyme func-tion is largely driven by the allelic variant present; however, there is also variability in function for the same apparent allele in different ethnic groups. The degree to which this variation is caused by undetected polymorphisms that cluster in one ethnic group versus another for a given allele is unknown. The availability of accurate haplotypic data is of critical relevance in paving the way to better evaluation and interpretation of genotype–phenotype correlations regarding CYP2D6.

CYP2D6 is involved in the phase I drug metabolism of many medications. It is involved in the bioactivation of tamoxifen and codeine and it is involved in the inactivation of several anti-depressants including desipramine, fluoxetine, mirtazapine, nortriptyline, paroxetine and ven-lafaxine. CYP2D6 is a prototypic example of a pharmacogenomic target and was listed as one of only two ‘valid pharmacogenomic biomark-ers’ in the 2003 US FDA Draft Guidance for Pharmacogenomic data [102].

CYP2D6 is the focus of much ongoing research. As new information on this gene is con-stantly evolving, the CYP Allele Nomenclature Committee, in vitro diagnostic (IVD) manufac-turers, researchers and clinical laboratories strug-gle to resolve the new and previously published information. There is no database containing complete DNA sequence data from the promoter region through the 3´ untranslated region for all of the CYP2D6 alleles. In fact, many of the alleles listed on the nomenclature page [101] have only had the exonic regions sequenced [3] and others have apparently not had the promoter region evaluated [4]. For this reason, data listed for these alleles may not include the information needed to accurately infer the alleles present and the associated pheno-type for a given sample. This occurred with the xTAG® CYP2D6 kit v3 IVD during verification testing in our laboratory where a CYP2D6*1/*11 was assigned a CYP2D6*2/*11 genotype on the basis of heterozygous calls for -1584C>G, 883G>C, 1661G>C, 2850C>T and 4180G>C polymorphisms, and a CYP2D6*4/*11 allele

CYP2D6 is genotyped clinically for prediction of response to tamoxifen, psychotropic drugs and other medications. Phenotype prediction is dependent upon accurate genotyping. The CYP Allele Nomenclature Committee maintains the allelic nomenclature for CYP2D6; however, in some cases, the list of polymorphisms associated with a given allele is incomplete. Clinical laboratories and in vitro diagnostic manufacturers rely upon this nomenclature, in addition to the literature, to infer allelic function and haplotypes and when they design CYP2D6-testing platforms. This article provides more complete sequencing data for the CYP2D6*11 allele and describes the difficulties encountered in genotyping CYP2D6 when incomplete data are available. The CYP Allele Nomenclature Committee should provide clear information about the completeness of the original data used to define each allele.

KEYWORDS: CYP2D6 n nomenclature n pharmacogenomic Jennifer M Skierka1, Denise L Walker2, Sandra E Peterson1, Dennis J O’Kane1 & John Logan Black III*1,2

1Nucleotide Polymorphism Laboratory, Department of Laboratory Medicine & Pathology, Mayo Clinic, 200 1st Street SW, Rochester, MN 55902, USA 2Functional Neurogenomics Laboratory, Department of Psychiatry & Psychology, Mayo Clinic & Mayo Medical School, Rochester, MN 55902, USA *Author for correspondence: Tel.: +1 507 284 3374 Fax: +1 507 266 4176 black.john@mayo.edu

part of

For reprint orders, please contact: reprints@futuremedicine.com

Special RepoRt Skierka, Walker, Peterson, O’Kane & BlackSpecial RepoRt Skierka, Walker, Peterson, O’Kane & Black

Pharmacogenomics (2012) 13(8)952 future science group

was assigned a ‘no call’ (e.g., failed to generate a genotype call) on the kit due to the presence of the -1584C>G polymorphism. We subsequently resequenced (see below) all available CYP2D6*11 samples obtained from previous testing which had been de-identified and stored with Mayo Clinic Institutional Review Board approval.

The CYP2D6*11 allele is a null allele by vir-tue of a polymorphism at 883G>C, which causes a splice defect. In the original description of the CYP2D6*11 allele, Marez et al. identified the allele’s sequence as containing 883G>C, 1661G>C, 2850C>T and 4180G>C [4]; how-ever, it appears that the -1584C>G polymor-phism was not evaluated in this study nor was the intron 1 CYP2D7 conversion evaluated, which is commonly associated with alleles apparently derived from further variation of the CYP2D6*2 alleles. It should be noted that the nucleotide numbering has been updated since the publica-tion of the Marez et al. article and is currently based on GenBank sequence M33388 [4]. We have subsequently found that the -1584C>G polymor-phism and several other polymorphisms, includ-ing the CYP2D7 gene conversion in intron 1, are found with the defining CYP2D6*11 polymor-phism, 883G>C, in all six samples we examined. This information is critical to IVDs manufac-turers, clinical laboratories and researchers as the -1584C>G polymorphism is used to infer both genotype and phenotype, and improper assignment to one allele or another has clinical consequences.

Methods & resultsGenomic DNA was isolated from ethylenediami-netetraacetic acid whole blood using the Qiagen EZ1 BioRobot™ according to product protocol (Qiagen, CA, USA). Original CYP2D6 geno-typing was performed using xTAG® CYP2D6 kit v1 and v2 and the Luminex 100 xMAP™ IS system, software versions 2.3 (Luminex, TX, USA) because these versions were used clinically. Variations detected by this method include: gene duplication, gene deletion, -1584 C>G, 100C>T, 124G>A, 138insT, 883G>C, 1023C>T, 1661G>C, 1707T>del, 1758G>T/A, 1846G>A, 2549A>del, 2613-2615 delAGA, 2850C>T, 2935A>C, 2988G>A and 4180G>C. When the xTAG® CYP2D6 kit v3 IVD became available, we began an IVD verification process which led to this study. Each version of the kit has added additional polymorphisms to the assay to aid in accuracy of allelic determination. For example, v2 added 138insT, 1661G>C, 2988G>A and 4180G>C. The v3 IVD kit

has added an additional three polymorphism (31G>A, 1659G>A and 3183G>A).

A 7-kb CYP2D6-specific fragment encompass-ing the entire gene (i.e., promoter, exons 1–9 and 3 -́UTR) was amplified similar to the method described previously [5,6]. Two allele-specific PCR amplicons were generated using the same forward primer as the long-PCR amplification and an allele-specific reverse primer where the 3´ base was 883G for the wild-type, and 883C for the variant copy of the gene, respectively. Using a PCR master mix similar to that previously described 1 µl of a 1:100 diluted product from the long-PCR amplification was added as a tem-plate to each nested PCR reaction. Thermocycler parameters with an initial denaturation step of 94°C for 1 min, followed by 18 cycles of 98°C for 10 s, 73°C for 3 min and a final elongation step of 72°C for 10 min were utilized. The wild-type and variant nested PCR products (~2800 bp) were bidirectionally sequenced to verify the -1584C>G and 883G>C polymorphisms.

Six heterozygous CYP2D6*11 (883G>C) samples were studied (Table 1). Samples 1–3, had an original genotype of CYP2D6*2A/*11, with a homozygous -1584C>G and all other *2 classical variations, but a heterozygous 883G>C using the xTAG® CYP2D6 kit. Allele-specific PCR product verified the presence of -1584C>G on both copies of the gene. Sample 4 (tested with CYP2D6 kit v1) and sample 5 (tested with CYP2D6 kit v2) were originally genotyped as being heterozygous for -1584C>G, 100C>T, 883G>C, 1846C>T and 2850C>T. Because sample 5 was testing using the CYP2D6 v2 kit, it was also found to be homozy-gous for 1661G>C and 4180G>C. These findings resulted in a CYP2D6*4/*11 genotype. Sample 5 yielded a ‘no call’ on the xTAG(R) CYP2D6 kit v3 as a result of the presence of the -1584G poly-morphism. However, it yielded a ‘no call’ on the xTAG® CYP2D6 kit v3 as a result of the presence of the -1584G polymorphism. The allele-specific PCR verified the presence of the -1584G variation in cis with 883C (*11). Sample 6 was originally studied as part of a verification process for the xTAG® CYP2D6 kit v3 and was heterozygous for -1584C>G, 883G>C and the classical *2 variations (1661G>C, 2850C>T and 4180G>C), resulting in this being called a CYP2D6*2/*11 genotype with that kit but when run with a previous ver-sion kit, it was genotyped as a CYP2D6*1/*11. The allele-specific PCR verified the presence of the -1584G variation in cis with 883C; thus, the actual genotype was CYP2D6*1/*11.

The CYP2D6*11 allele of sample 6 was sequenced using the allele-specific nested PCR

Special RepoRt Skierka, Walker, Peterson, O’Kane & BlackSpecial RepoRt Skierka, Walker, Peterson, O’Kane & Black

www.futuremedicine.com 953future science group

CYP2D6*11 & challenges in clinical genotyping of the polymorphic CYP2D6 gene Special RepoRt

method as described above and also in another allele-specific nested PCR. This additional allele-specific nested PCR used the 883G>C SNP as the forward allele-specific primer and ampli-fied both copies of the gene separately through the remainder (exon 2–9) of the gene with the reverse primer from the long-PCR amplification of CYP2D6.

DiscussionSequencing of the CYP2D6*11 allele revealed the following polymorphisms: -1584C>G (rs1080985), -1235A>G (rs28735595), -740C>T (rs28624811), -678G>A (rs28633410), 214G>C (rs1080995), 221C>A (rs1080996), 223C>G (rs1080997), 227T>C (rs1080998), 232-233GA>CC (rs1080999), 245A>G (rs1081000), 310G>T (rs28371699), 746C>G (rs28371701), 843T>G (rs28371702), 883G>C (rs5030863), 1661G>C (rs1058164), 2850C>T (rs16947), 3384A>C (rs1985842), 3584G>A (rs28371730), 3790C>T (rs28371731), 4180G>C (rs1135840) and 4481G>A (rs116390392). The complete CYP2D6 *11 allele sequence is available in NCBI (accession number JQ421296) [103].

We report that in all six samples with a CYP2D6*11 allele, a -1584G polymorphism was found in cis with the 883C allele indicating that -1584G is a part of the CYP2D6*11 allele in addi-tion to the 883G>C, 1661G>C, 2850C>T and 4180G>C polymorphisms already associated with this allele. In addition, 15 other polymorphisms, including the CYP2D7 gene conversion in intron 1, typical of alleles in the CYP2D6*2 family, were found with this allele.

Clarifying the haplotype for this allele is important because at least two FDA-approved CYP2D6 genotyping platforms (Luminex xTAG CYP2D6® kit and the Roche Amplitaq® CYP test, Roche Diagnostics Corp., IN, USA) evaluate the

-1584C>G polymorphism and use it to assign genotype. Correctly assigning this polymorphism to the proper allele has implications on predicting phenotype for a clinical sample. For example, the -1584G polymorphism appears to increase tran-scription of CYP2D6 [7] and has been associated with a trend towards ultrarapid metabolism [8] or extensive metabolism even in the presence of other polymorphisms that result in a deficiency allele such as CYP2D6*2 [9].

We have recently published research involv-ing CYP2D6 hybrid alleles, which showed that CYP2D6-2D7 hybrid tandems are found in 18.7% of CYP2D6*4 and 14.5% of CYP2D6*10 alleles and CYP2D7-2D6 hybrid tandems are seen in 7.3% of samples yielding duplication signals using a xTAG® CYP2D6 kit [5,6]. This collective body of work underscores the importance of understand-ing this gene’s molecular diversity as well as cis and trans relationships of CYP2D6 polymorphisms when developing in vitro diagnostic tools.

Financial & competing interests disclosureThe Mayo Clinic and JL Black and DJ O’Kane have licensed intellectual property, which aids in the selection of psychotropic drugs based upon genotyping to AssureRx. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

Ethical conduct of research The authors state that they have obtained appropriate insti-tutional review board approval or have followed the princi-ples outlined in the Declaration of Helsinki for all human or animal experimental investigations. In addition, for investi gations involving human subjects, informed consent has been obtained from the participants involved.

Table 1. SNP results for six samples containing 883C.

Sample Original genotype -1584C>G call from kit§

-1584G cis/trans relationship

CYP2D6 v1 kit† CYP2D6 v2 kit† CYP2D6 v3 kit‡

1 *2A/*11 *2/*11 G/G (MUT) -1584G on both copies

2 *2A/*11 *2/*11 G/G (MUT) -1584G on both copies

3 *2A/*11 DUP *2/*11 DUP G/G (MUT) -1584G on both copies

4 *4/*11 C/G (HET) -1584G cis with 883C

5 *4/*11 No call C/G (HET) -1584G cis with 883C

6 *1/*11 *2/*11 C/G (HET) -1584G cis with 883C

The original genotype by xTAG® CYP2D6 kit v1, v2 and v3. Note that when using the xTAG® CYP2D6 v3 in vitro diagnostic, sample five yielded a ‘no calls’ for the genotype and sample six yielded a CYP2D6*2/*11 caused by the presence of the -1584G polymorphism. The -1584G cis/trans relationship column shows the association between this polymorphism and the 883C (CYP2D6*11) polymorphism. †Genotype call assigned by the clinical laboratory. ‡Genotype call assigned by ana lysis software provided from manufacturer. §Note that the -1584 variation call was concordant from kit to kit.

Special RepoRt Skierka, Walker, Peterson, O’Kane & Black

Pharmacogenomics (2012) 13(8)954 future science group

Executive summary

Background � CYP2D6 is a clinically relevant gene that is tested often for pharmacogenomic purposes.

� CYP2D6 is a very complex gene with over 105 reported variations including duplications, multiplications, deletions, hybrid alleles and tandem hybrid alleles.

� Existing databases for CYP2D6 allele sequences are incomplete.

� In this article, we updated the literature on CYP2D6*11 sequence.

Method � With Mayo Clinic IRB approval, six de-identified samples known to have the CYP2D6*11-defining polymorphism, 883G>C, were evaluated by allele specific nested PCR and DNA sequencing to determine the cis and trans relationship of the 883G>C and -1584C>G polymorphisms. In all cases, these two polymorphisms were found to be in a cis relationship.

Results � Sequencing of the CYP2D6*11 allele revealed the following polymorphisms: -1584C>G (rs1080985), -1235A>G (rs28735595), -740C>T (rs28624811), -678G>A (rs28633410), 214G>C (rs1080995), 221C>A (rs1080996), 223C>G (rs1080997), 227T>C (rs1080998), 232–233GA>CC (rs1080999), 245A>G (rs1081000), 310G>T (rs28371699), 746C>G (rs28371701), 843T>G (rs28371702), 883G>C (rs5030863), 1661G>C (rs1058164), 2850C>T (rs16947), 3384A>C (rs1985842), 3584G>A (rs28371730), 3790C>T (rs28371731), 4180G>C (rs1135840) and 4481G>A (rs116390392).

Conclusion � All six CYP2D6*11 alleles contained the -1584G polymorphism.

� The CYP2D6*11 allele appears to be in the CYP2D6*2 subfamily of CYP2D6 alleles since it has the polymorphisms commonly associated with CYP2D6*2.

� Genotyping errors that may impact patient-management decisions by leading to erroneous assignment of phenotype, will occur unless there is a complete understanding of the DNA sequence of CYP2D6 alleles.

� The CYP Allele Nomenclature Committee should provide clear information about the completeness of the original data used to define each allele including if sequencing of exonic, intronic and promoter regions was done and if gene conversion and copy number variation were assessed.

ReferencesPapers of special note have been highlighted as:nn of considerable interest

1 Purich D, Allison R. The Enzyme Reference: a Comprehensive Guide to Enzyme Nomenclature, Reactions, and Methods. Academic Press, Elsevier Science, CA, USA (2002).

2 Lundqvist E, Johansson I, Ingelman-Sundberg M. Genetic mechanisms for duplication and multiduplication of the CYP2D6 gene and methods for detection of duplicated CYP2D6 genes. Gene 226, 327–338 (1999).

3 Marez D, Legrand M, Sabbagh N et al. Polymorphism of the cytochrome P450 CYP2D6 gene in a European population: characterization of 48 mutations and 53 alleles, their frequencies and evolution. Pharmacogenetics 7, 193–202 (1997).

4 Marez D, Sabbagh N, Legrand M, Lo-Guidice J, Boone P, Broly F. A novel CYP2D6 allele with an abolished splice recognition site associated with the poor metabolizer phenotype. Pharmacogenetics 5(5), 305–311 (1995).

5 Kramer W, Walker D, O’Kane D et al. CYP2D6: novel genomic structures and alleles. Pharmacogenet. Genomics 19, 813–822 (2009).

nn Describes novel arrangements of the CYP2D locus and novel alleles.

6 Black J, Walker D, O’Kane D, Harmandayan M. Frequency of undetected CYP2D6 hybrid genes in clinical samples: impact on phenotype prediction. Drug Metabol. Disposition 40, 111–119 (2012).

nn Describes novel arrangements of the CYP2D locus and novel hybrid genes as well as frequency of CYP2D6 hybrid genes in clinical samples.

7 Zanger U, Fischer J, Raimundo S et al. Comprehensive ana lysis of the genetic factors determining expression and function of hepatic CYP2D6. Pharmacogenetics 11, 573–585 (2001).

8 Lovlie R, Daly AK, Matre GE, Molven A, Steen VM. Polymorphisms in CYP2D6 duplication-negative individuals with the ultrarapid metabolizer phenotype: a role for the CYP2D6*35 allele in ultrarapid metabolism? Pharmacogenetics 11, 45–55 (2001).

9 Raimundo S, Fischer J, Eichelbaum M, Griese EU, Schwab M, Zanger U. Elucidation of the genetic basis of the common ‘intermediate metabolizer’ phenotype for drug oxidation by CYP2D6. Pharmacogenetics 10, 577–581 (2000).

n Websites101 CYP2D6 Allele Nomenclature.

www.cypalleles.ki.se/cyp2d6.htm (Accessed 7 May 2012)

102 US FDA Guidance for Industry: Pharmacogenomic Data Submissions. www.fda.gov/downloads/RegulatoryInformation/Guidances/UCM126957.pdf?utm_campaign=Google2&utm_source=fdaSearch&utm_medium=website&utm_term=pharmacogenomic AND 2002&utm_content=3

103 NCBI website. www.ncbi.nlm.nih.gov