The drug metabolizer phenotype of the major P450 enzyme, CYP2D6, can be predicted by genetic analysis. The combination of full, reduced, or no function alleles, in addition to gene copy number, is used to determine ultrarapid, extensive, intermediate, or poor metabolizer status. Data from quantitative PCR (qPCR) experiments using TaqMan™ SNP Genotyping assays and TaqMan Copy Number assays can be translated to star allele diplotypes associated with metabolizer phenotypes. While qPCR can decipher most of the information necessary, phenotypes cannot be unequivocally assigned for samples that contain a CYP2D6 duplication or multiduplications and are heterozygous for alleles that are known to be duplicated and that have different functional levels. For these samples, allele-specific copy number analysis by digital PCR (dPCR) can be done in order to identify the duplicated allele. Digital PCR enables single sample analysis with high precision and sensitivity by partitioning target molecules into 20,000 individual reactions, thus elucidating the ratio of the heterozygous alleles. For the allele-specific copy number variation (ASCNV) application, TaqMan SNP assays to CYP2D6 variants that are associated with specific duplicated alleles were run in dPCR on samples of known SNP genotype and CNV status. Initial validation was conducted on Coriell cell line gDNA samples. Here we describe work done on buccal swab samples collected from two different sites. Representative samples carrying CYP2D6 duplications that were heterozygous were selected to run with ASCNV dPCR.

1. Allele specific digital PCR to differentiate
CYP2D6 heterozygous duplication events
1Patricia Hegerich, 1Sunali Patel, 2Jessica Forcellini, 2Mark Borgman, 1Toinette Hartshorne,
1Thermo Fisher Scientific, South San Francisco, CA; 2PGXL Laboratories, Louisville, KY
ABSTRACT
The drug metabolizer phenotype of the major P450 enzyme,
CYP2D6, can be predicted by genetic analysis. The
combination of full, reduced, or no function alleles, in
addition to gene copy number, is used to determine ultra-
rapid, extensive, intermediate, or poor metabolizer status.
Data from quantitative PCR (qPCR) experiments using
TaqMan™ SNP Genotyping assays and TaqMan Copy
Number assays can be translated to star allele diplotypes
associated with metabolizer phenotypes. While qPCR can
decipher most of the information necessary, phenotypes
cannot be unequivocally assigned for samples that contain a
CYP2D6 duplication or multiduplications and are
heterozygous for alleles that are known to be duplicated and
that have different functional levels.
For these samples, allele-specific copy number analysis by
digital PCR (dPCR) can be done in order to identify the
duplicated allele. Digital PCR enables single sample
analysis with high precision and sensitivity by partitioning
target molecules into 20,000 individual reactions, thus
elucidating the ratio of the heterozygous alleles. For the
allele-specific copy number variation (ASCNV) application,
TaqMan SNP assays to CYP2D6 variants that are
associated with specific duplicated alleles were run in dPCR
on samples of known SNP genotype and CNV status. Initial
validation was conducted on Coriell cell line gDNA samples.
Here we describe work done on buccal swab samples
collected from two different sites. Representative samples
carrying CYP2D6 duplications that were heterozygous were
selected to run with ASCNV dPCR.
INTRODUCTION
Figure 1. The CYP2D6 gene is highly polymorphic. Genotype analysis
requires both SNP genotyping and CNV analysis. (figure adapted from ref. 1)
haplotype Major SNPs
Enzyme
Function
Activity score
*1 reference Full 1.0
*2 2850C>T; 4180G>C Full 1.0
*2A -1584C>G; 2850C>T; 4180G>C Full 1.0
*4 100C>T; 1846G>A; 4180G>C None 0
*9 2615_2617delAAG Reduced 0.5
*10 100C>T; 4180G>C Reduced 0.5
*17 1023C>T; 2850C>T; 4180G>C Reduced 0.5
*35 -1584C>G; 31G>A; 2850C>T; 4180G>C Full >=1.0
Table 1. CYP2D6 star allele haplotypes that can be duplicated2. Star allele
haplotypes are associated with DME phenotypes.
Figure 2. DME metabolizer phenotype can vary depending on which allele has been
duplicated in individuals heterozygous for alleles in different functional classes
MATERIALS AND METHODS
Initial CYP2D6 genotyping was performed on genomic DNA
isolated from buccal swabs using TaqMan® DME assays in the
OpenArray™ format on the QuantStudio™ 12K Flex instrument.
CYP2D6 copy number (CN) was determined using qPCR assays
targeting exon 9 and additionally intron 2 where applicable.
Samples were run at two sites. The first set of samples 4 of the 8
assays listed in the workflow were run on each sample. At the
second site, representative samples with duplication positive
genotypes were selected to be run with one defining assay.
Allele-specific copy number variation (ASCNV) was done by first
digesting samples with a restriction enzyme to separate tandem
duplicated CYP2D6 alleles then loading onto QuantStudio 3D
Digital PCR 20K chips with CYP2D6 TaqMan DME assays
targeting the -1584C>G, 100C>T, 1023C>T, 1846G>A, 2850C>T,
or 4180G>C variants, depending on genotype in question.
Reactions positive for each allele, detected by allele-specific
VIC™ or FAM™ dye-labeled probes, were visualized, and allele
ratios determined using data from the QuantStudio 3D
AnalysisSuite™ software on the Thermo Fisher Cloud. For
samples heterozygous for target SNPs, 3-copy samples gave
close to 1:2 ratios and the duplicated allele was readily identified.
Where there was a discrepancy between the ASCNV results and
the qPCR results, additional Copy Number assays were run on
digital PCR for further clarification.
4. Analyze data
o Calculate VIC and FAM™ dye ratios
using from data from QuantStudio 3D
AnalysisSuite™ software on the Thermo
Fisher Cloud.
3. Run digested gDNA with TaqMan
SNP Assays on QuantStudio 3D
Digital PCR system
Figure 3. CYP2D6 allele-specific copy number dPCR workflow using the QuantStudio 3D Digital PCR System
QuantStudio® 3D
AnalysisSuite™
Software
Likely phenotype Frequency
Activity
score
Genotypes
Examples of
diplotypes
Ultrarapid metabolizer 1-2% >2.0 more than two copies of functional alleles
*1/*1x2,
*1/*2x2
Extensive metabolizer 77-92% 1.0-2.0
two full or reduced function alleles or
one full function allele plus either one
nonfunctional or one reduced function allele
*1/*1, *1/*2,
*10/*10,
*1/*4, *10/*5
Intermediate metabolizer 2-11% 0.5 one reduced and one nonfunctional allele
*4/*10,
*5/*17
Poor metabolizer 5-10% 0 no functional alleles
*4/*4, *4/*5,
*5/*5, *4/*6
*1 / *10 x 2
Extensive
metabolizer
*1 x 2 / *10
Ultrarapid
metabolizer
*10 reduced function allele
*1 full function allele
Example1
*4 / *17 x 2
Extensive
metabolizer
*4 x 2 / *17
Intermediate
metabolizer
*17 reduced function allele
*4 nonfunctional allele
Example2
Table 2. CYP2D6 diplotype predicts the drug metabolism phenotype, which is used to aide
drug selection and starting dosages. (table adapted from ref. 3)
E.g., with respect to the metabolism of codeine to morphine by CYP2D6:
• Ultrarapid metabolizers - avoid due to potentially toxic morphine levels.
• Poor metabolizers - avoid due to lack of efficacy.
• Extensive & Intermediate metabolizers use age- & weight-specific dosing
• Intermediate metabolizers may not respond as well as extensive metabolizers
Star Allele Results
2. Select appropriate TaqMan Drug
Metabolism Genotyping Assays
1. Identify samples with CYP2D6
duplications that are
heterozygous for functionally
different alleles (~1-2% of
samples)
RESULTS
Copy Number = 4
*2x3/*4
Copy Number = 3
*2/*4x2
Copy Number = 3
*2Ax2/*4
A
A
G
G
G
A
AG
PG # Ethnicity
Predicted
Genotype
Diplotype Assay
Variant
%*
Copy
Number
4 Caucasian *1/*10 *1x2/*10 *10 100C>T 36% 3
10 Caucasian *1/*10 *1x2/*10 *10 100C>T 35% 3
A unknown *1/*17 *1x2/*17 *17 1023C>T 34% 3
B unknown *1/*17 *1x2/*17 *17 1023C>T 35% 3
C unknown *1/*17 *1x2/*17 *17 1023C>T 30% 3
20 Ashkenazi *1/*4 *1/*4x2 *4 1846G>A 66% 3
19 Hispanic *1/*41 *1x2/*41 4180G>C 32% 3
17 Af Amer *2/*10 *2/*10x2 2850C>T 58% 3
7 Hispanic *2/*10 *2x2/*10 2988G>A 66% 3
18 Af Amer *2/*4 *2/*4x2 *4 1846G>A 66% 3
5 Caucasian *2/*4 *2x3/*4 *4 1846G>A 21% 4
8 Hispanic *2/*9 *2x2/*9 2850C>T 66% 3
2 Hispanic *2A/*10 *2A/*10x2 *2A-1584C>G 35% 3
15 Hispanic *2A/*17 *2Ax2/*17 *17 1023C>T 38% 3
9 Af Amer *2A/*29 *2Ax3/*29 *2A-1584C>G 70% 4
13 Hispanic *2A/*4 *2Ax2/*4 *4 1846G>A 33% 3
6 Caucasian *2A/*6 *2Ax2/*6 2850C>T 66% 3
1 Unknown *2A/*9 *2A/*9 4180G>C 51% 3
Table 4. Table 4. Accurate calling of copy number ratios from buccal swab samples run at
Site 2. The % FAM allele within each sample run with the indicated assay is shown: the
average value for 2 experiments is compared to the expected value for each 3 copy
sample.
* Note: the FAM % were calculated in Excel using data from AnalysisSuite™
Figure 4. Examples of dPCR clusters produced with a
CYP2D6 SNP assay, *4 1846G>A, run on one 4 copy
and two 3 copy samples. Data shown was generated by
Site 2.
For research use only. Not for use in diagnostic procedures. Thermo Fisher Scientific • 5791 Van Allen Way • Carlsbad, CA 92008 • www.thermofisher.com
REFERENCES
1. Nature Reviews Drug Discovery (2004)
2. The Human Cytochrome P450 (CYP) Allele Nomenclature Database. www.cypalleles.ki.se.
3. Pharmacogenomics Knowledge Database. www.PharmGKB.org.
TRADEMARKS/LICENSING
© 2015 Thermo Fisher Scientific Inc. All rights reserved. All trademarks are the property of
Thermo Fisher Scientific and its subsidiaries unless otherwise specified. TaqMan is a registered
trademark of Roche Molecular Systems, Inc., used under permission and license.
CYP2D6
exon 9
CYP2D6
intron2
2D6 g.-
1584C>G
2D6
g.100C>T
2D6
g.1023C>
T
2D6
g.1846G>
A
2D6
g.2850C>
T
2D6
g.4180G>
C
Expected
Hs00010
001_cn
Hs04083
572_cn
C__3240
7252_30
C__1148
4460_40
C___222
2771_A0
C__2710
2431_D0
C__2710
2425_10
C__2710
2414_10
VIC:FAM
C__11484460_40 47 1:1
C__27102431_D0 48 1:1
C__27102425_10 97 0:1
C__27102414_10 50 1:1
C__32407252_30 34 1:2 or 2:1
C__11484460_40 34 1:2 or 2:1
C__27102425_10 69 2:1 or 1:2
C__27102414_10 99 1:0 or 0:1
C__32407252_30 66 2:1 or 1:2
C___2222771_A0 1 0:1 or 1:0
C__27102425_10 35 1:2 or 2:1
C__27102414_10 66 2:1 or 1:2
C__32407252_30 51 1:1
C__11484460_40 50 1:1
C__27102431_D0 49 1:1
C__27102425_10 49 1:1
C__32407252_30 35 1:2or 2:1
C__11484460_40 36 1:2 or 2:1
C__27102431_D0 68 2:1 or 1:2
C__27102425_10 67 2:1 or 1:2
C__11484460_40 51 1:1
C___2222771_A0 50 1:1
C__27102431_D0 48 1:1
C__27102425_10 50 1:1
C__11484460_40 34 1:2 or 2:1
C__27102431_D0 65 2:1 or 1:2
C__27102425_10 100 1:0 or 0:1
C__27102414_10 49 1:1
32
Sample
ID
Predicted
Genotype
Diplotype Assay
38 *4/*17 *4/*17 2 2
A/G
Variant
%*
1 *1/*4 *1/*4 2 2 G/G A/G G/G C/T G/G C/G
15
*2Ax2/*36
or
*10x2/*63
*2A/*36-
*10
G/G
28
*1/*2Ax2
or
*1x2/*2A
*1/*2Ax2 3 3 G/C G/G G/G C/C A/G C/G
G/C A/G G/G C/C
C/T A/G33 *2A/*4 *2A/*4 2 2
A/G
G/G
36
*2A/*4x2
or
*2Ax2/*4
*2A/*4x2 3 3 G/C A/G G/G C/T A/G G/G
G/C A/G G/G
G/G
43 *3/*4 *3/*4N-*4 2 3 G/G A/G G/G C/T G/G C/G
G/G A/G G/A C/T
Table 3. Accurate calling of copy number ratios from assay run on buccal swabs gDNA from site 1 with
either 2 or 3 copies of CYP2D6. A representative set of samples were selected to run with 4 assays per
sample. Results were consistent across all assays run for each sample shown.
* Note: the FAM % were calculated in Excel using data from AnalysisSuite™
CONCLUSIONS
We have demonstrated that allele-specific copy number analysis using dPCR and TaqMan SNP assays is a simple and effective method for identifying specific duplicated alleles in heterozygous samples.
This method facilitates accurate CYP2D6 allele genotyping and better prediction of drug metabolizer phenotype.
ASCNV was able to discern the genotype for 100% of the samples tested at the first site. At the second site, 18 of 23 samples tested with a success rate of 78%. This is likely due to the condition of the
samples. Successful samples with CN=3 were evident by a %FAM/Total ratio of approximately 33% or 66% indicating which allele was present twice depending on the TaqMan assay. Equivocal samples
resulted from 50% FAM/Total in CN=3 samples. For samples with initial phenotype of EM/UM, 78% were identified as UM. For samples with initial phenotype of EM/IM, 44% were identified as IM.
Interestingly, one *2/*4 sample with CN=4 and phenotype of EM/IM returned a result of *2/*2/*2/*4 and thus a true phenotype of UM.