Hum Genet (1995) 95 : 337-341 © Springer-Verlag 1995

Ann K. Daly • Jul ian B. S. Leathar t - S tephanie J. London Jeffrey R. Idle

An inactive cytochrome P450 CYP2D6 allele containing a deletion and a base substitution

Received: 31 May 1994/Revised: 21 September 1994

A b s t r a c t The cy tochrome P450 CYP2D6 is a po lymor - phic enzyme, for which 5 % - 1 0 % of Caucas ians (poor metabol izers ) lack activity. The major i ty of mutat ions giv- ing rise to the def ic iency have now been ident i f ied but some individuals show anomalous pheno type-geno type re la t ionships when screened for the c o m m o n mutant alle- les. We have sequenced all nine exons and in t ron-exon boundar ies in a subject who was phenotyp ica l ly a poor metabol izer but genotypical ly heterozygous when screened for the c o m m o n alleles. A single base-pa i r dele t ion (T1795) was detected in exon 3 and a base substi tut ion (G2064A) resul t ing in an amino acid subst i tut ion (G212E) in exon 4. The dele t ion results in premature terminat ion of transla- t ion and a t runcated protein. In a group of 50 white A m e r - icans, the al lele f requency for the new mutant al lele was 0.01. The new al lele expla ins some cases of anomalous geno type /pheno type rela t ionships for CYP2D6.

Introduction

The cy tochrome P450 enzyme CYP2D6 metabol izes at least 30 therapeut ica l ly impor tant drugs and may also p lay a role in the act ivat ion of certain carc inogens (for a re- view, see Da ly et al. 1993). The enzyme is po lymorph ic and about 5 % - 1 0 % of individuals of var ious ethnic ori- gins lack this enzyme activity. A number of different mu- tat ions associa ted with the defect have now been de- scribed. Assays involv ing the po lymerase chain react ion (PCR) and res t r ic t ion f ragment length po lymorph i sm

(RFLP) analysis can now ident i fy over 90% of individuals who lack CYP2D6 act ivi ty and who are known as poor metabol izers (Broly et al. 1991; Daly et al. 1991; Evans and Rel l ing 1991; Dahl et al. 1992). However , several studies on pheno type /geno type rela t ionships in CYP2D6 have sugges ted that other inact ivat ing muta t ions occur (Broly et al. 1991; Daly et al. 1991; Evans and Rel l ing 1991). To detect muta t ions of this nature, we have se- quenced all the exons and in t ron-exon boundar ies of the CYP2D6 gene f rom a prev ious ly descr ibed indiv idual (Daly et al. 1991) who lacks CYP2D6 act ivi ty on the ba- sis of in vivo phenotyp ing with the probe drug debr iso- quine but who genotypes as he te rozygous on the basis of assays for all known CYP2D6 mutations.

Materials and methods

Subjects

The propositus was a 40-year-old white British male who was a poor metabolizer of debrisoquine (metabolic ratio of > 50 on two separate occasions) but who showed a CYP2D6B/wild-type geno- type. Fifty white American subjects selected at random were also studied, in addition to a group of 23 white American subjects who showed anomalous genotype/phenotype relationships that were de- fined either as being homozygous wild-type when genotyped for the CYP2D6A, CYP2D6B and CYP2D6D alleles and having a de- brisoquine metabolic ratio of greater than 1, or as being heterozy- gous for one of the above mutant alleles and with a metabolic ratio of greater than 10. All subjects gave informed consent and the study was carried out with the approval of the appropriate ethical committees.

A. K. Daly (IE~). J. B. S. Leathart - J. R. Idle Pharmacogenetics Research Unit, Department of Pharmacological Sciences, University of Newcastle upon Tyne, Medical School, Framlington Place, Newcastle upon Tyne NE2 4HH, UK

S. J. London Department of Preventive Medicine, Division of Occupational and Public Health, University of Southern California School of Medicine, Los Angeles, California, USA

CYP2D6 phenotyping and genotyping

CYP2D6 phenotyping with debrisoquine and genotyping for the known CYP2D6-inactivating mutations (CYP2D6A, CYP2D6B and CYP2D6D) was carried out as described previously (Daly et al. 1991).

Amplification of CYP2D6 by PCR for sequencing

The sequence of all exons and at least 20 bp of all adjoining introns of the apparent wild-type allele of the post metabolizer was deter-

338

4D 7E 5E 4C 1C 1G 2G 1A

PCR 1 PCR 2

6G

PCR 4

4B 7G 5B

2 I I 3 4 6

site

7'

PCR 3

1G 2G IE IA 4A

Fig. 1 PCR reactions and sequencing primers used for sequencing of CYP2D6 exons. The position and orientation of each oligonu- cleotide used as a PCR primer or internal sequencing primer is in- dicated

two clones negative for the CYP2D6B sequence from each PCR re- action were then sequenced using the primers shown in Fig. 1 and Table 1. Sequencing was carried out with a Sequenase II kit (USB) in accordance with the manufacturer 's recommendations.

mined by sequencing PCR products after cloning in Bluescript KS (Stratagene). As summarized in Fig. 1, four separate PCR products were amplified using the conditions and primers described in Table 1. General PCR conditions were as described previously (Daly et aI. 1991). The primers were designed so that only CYP2D6 and not the homologous pseudogenes CYP2D7P and CYP2D8P were amplified. PCR products were subcloned into Bluescript KS that had been digested with Sinai and treated with alkaline phosphatase (Sambrook et al. 1989). To avoid sequencing clones originating from the CYP2D6B allele, eight clones from each PCR reaction were screened by sequencing short stretches with appropriate primers (Table 1) to determine whether CYP2D6B- associated mutations were present. All exons present in at least

PCR assay for the exon 3 and exon 4 mutations

Allele-specific PCR assays were developed for the deletion in exon 3 and the base substitution in exon 4. For the exon 3 muta- tion, two PCR reactions were carried out in parallel using similar conditions to those described previously (Daly et al. 1991) with the primers 2G (CTCGGTCTCTCGCTCCGCAC; bp 2096 to 2115) and 9G (CAAGAAGTCGCTGGAGCTGT; bp 1776 to 1795), and 2G and 10G (CAAGAAGTCGCTGGAGCTGG; bp 1776 to 1795), respectively. The temperature conditions were 30 cycles of l m i n at 95°C, 1.5min at 54°C and 3min at 70°C in a Techne PHC-3 heating block. Primers 9G and 10G were allele-specific with an additional mismatch introduced in position 18 to give im-

Table 1 PCR primers and conditions for sequencing

Primer no. Position (bp) a Orientation Sequence Conditions

Primers and conditions for reactions 1 to 4 (see Fig. 1)

PCR1

4D - 100- -81 Forward 4C 1241-1260 Reverse

PCR2 1C 1394-1413 Forward 2G 2096-2115 Reverse

PCR3 1G 1824-1843 Forward 4A 3340-3359 Reverse

PCR4 1A 2623-2642 Forward 5B +411-+430 Reverse

Sequencing primers 7E 70 - 89 Forward 5E 965- 984 Forward 6G 3214-3233 Forward 4B 4207-4226 Forward 7G 4235-4254 Reverse 2L 1652-1671 Forward

CTGGCCTGACTCTGCCACTG CCACTCGCTGGCCTGTTTCA

CTGGAATCCGGTGTCGAAGT CTCGGTCTCTCGCTCCGCAC

TGCCGCCTTCGCCAACCACT CCAAAGCGCTGCACCTCATG

T G C T A A C T G A G C A C A G G A T G T G T C C A G T G G G C A C C G A G A A

A T T T G G T A G T G A G G C A G G T TCTGCAGTTGCGGCGCCGCT T G A C C C A T T G T G G G G A C G C A TTCTCGGTGCCCACTGGACA A A A G A C A C C A T G G T G G C T G G AATGCCTTCATGGCCACGCG

1 min 95°C, 1 min 65°C, 9.9 min 70°C (40 cycles)

1 min 95°C, 1 min 55°C, 9.9 min 70°C (40 cycles)

1 min 95°C, 1 min 65°C 9.9 min 70°C (30 cycles)

1 min 95°C, 1 min 65°C 9.9 min 70°C (30 cycles)

a Sequence positions are as described by Kimura et al. (1989)

Fig. 2a-c Detection of exon 3 and exon 4 mutations by DNA sequencing of a cloned PCR product, a Nucleotides 1394-2115 were amplified us- ing primers IC and 2G. After cloning into Bluescript KS, clones containing inserts were sequenced with primer 2L. The wild-type sequence is shown left and the mutant Hght. b Nucleotides 1824-3359 were amplified using primers 1G and 4A. After cloning into Bluescript KS, clones contain- ing inserts were sequenced with the primer 1G. The wild- type sequence is shown left and the mutant right, c Sum- mary of sequence differences between the wild-type and mu- tant alleles. Primers are given in Table 1

Wild-type

T C G A

Mutant

T C G A

339

T deleted

4 ~ i ; ~ i~ ~ ¸

T

Wild-type

C G A

Mutant

T C G A

- - G t o A

b

E x o n 3 fop 1788 - 1804)

W i l d - t y p e G G A G C A G T G G G T G A C C G M u t a n t G G A G C A G G G G T G A C C G

E x o n 4 fop 2 0 5 7 - 2069)

W i l d - t y p e A G G A G G C ~ C T G A M u t a n t AGGAGG/kACTGA

proved specificity. For the exon 4 mutation, two parallel PCR re- actions were carried out with the primers 1G (TGCCGCCTTCGC- CAACCACT; bp 1824-1843) and 1H (GCCCGACTCCTCCTTC- AGTC; bp 2064-2083), and 1G and 2H (GCCCGACTCCTCCTT- CAGTT; bp 2064-2083), respectively. The temperature conditions were 30 cycles of 1 min at 95°C, 1 min at 63°C and 2min at 70°C in a Techne PHC-3 heating block. Products from both PCR reac- tions were analysed on 1% agarose gels.

Results

Identif icat ion of a new mutat ion by sequencing of the apparent wild- type allele

The propositus was phenotypical ly a poor metabol izer on the basis of two separate rounds of debrisoquine pheno-

340

Wild-type

1 2 3 4 5 6 7 8

400bp

300bp

Exon3mutat ion

1 2 3 4 5 6 7 8

400 bp

300bp

Wild-type

1 2 3 4 5 6 7 8 9 10

- -300bp

- - 2 0 0 b p

Exon 4 mutation

1 2 3 4 5 6 7 8 9 10 300bp

13 200 bp

Fig. 3a,b Detection of the exon 3 and exon 4 mutations by allele- specific PCR. a Exon 3 assay. The upper panel shows eight DNA samples amplified using primers 9G and 2G and the lower panel the same samples amplified with 10G and 2G. Subjects 7 and 8 are heterozygous for the mutation, b Exon 4 assay. The upper panel shows ten DNA samples amplified using primers 1G and 1H and the lowerpanel the same samples amplified with 1G and 2H. Sub- jects 9 and 10 are heterozygous for the mutation

typing that gave identical results. He was in good health and not taking any interfering drugs. Genotyping for the CYP2D6A, CYP2D6B and CYP2D6D alleles by PCR and RFLP analysis gave a heterozygous CYP2D6B genotype with a 29-kb homozygous pattern on RFLP analysis with XbaI. It therefore appeared that this subject had an un- usual CYP2D6-inactivating mutation present in the appar- ently wild-type allele. Using a sequencing strategy that ensured that only the apparently wild-type allele was se- quenced, we found only two base pair differences be- tween this allele and the published genomic sequence for the nine CYP2D6 exons. There was a single base pair deletion of T1795 present in exon 3 and a base substitution of G2064A in exon 4 that resulted in the substitution of Glu for Gly at position 212 (Fig. 2).

Table 2 Subjects positive for the exon 3 and exon 4 mutations (nd not determined, wt wild type, Mut mutation

Code M e t a b o l i c CYP2D6A/B/D Exon 3 Exon 4 ratio genotype

N55 >50 B/wt Mut/wt Mut/wt LA16 13.6 wt/wt Mut/wt Mut/wt LA147 3.2 wt/wt Mut/wt Mut/wt LA428 nd wt/wt Mut/wt Mut/wt LA622 >50 B/wt Mut/wt Mut/wt LA678 1.74 wt/wt Mut/wt Mut/wt

Detection of the T1795 deletion and the G2064 A mutation by PCR

To enable population screening for the T1795 deletion and G2064 A mutation, allele-specific PCR assays were devel- oped. A typical result is shown in Fig. 3. A total of 50 ran- domly selected white American subjects and a separate group of 23 white Americans who might be more likely to have the mutations because of their anomalous pheno- type/genotype relationship were screened for both muta- tions. In the randomly selected group only one subject (LA428) was heterozygous for the mutations, thereby giv- ing an allele frequency of 0.01. When the individuals showing anomalous phenotype/genotype relationships were screened, an additional four subjects were heterozy- gous for both mutations making a total of six subjects. As summarized in Table 2, in addition to the propositus (sub- ject N55), two of these subjects were phenotypically poor metabolizers with debrisoquine metabolic ratios of 13.6 and >50, respectively. One of these individuals (meta- bolic ratio > 50) was also heterozygous for the CYP2D6B allele. Two of the other three subjects heterozygous for the new allele showed metabolic ratios of 1.74 and 3.19, which was in the range normally observed in those het- erozygous for other inactivating CYP2D6 mutations (Daly et al. 1991). In all subjects studied in the two groups, in- dividuals heterozygous for the exon 3 mutation were also heterozygous for the exon 4 mutation.

Discussion

This study has identified a new poor metabolizer-associ- ated CYP2D6 allele that we estimate occurs at a frequency of 0.01 in white Caucasians. Following discussions with the recently established ad hoc CYP2D6 allele nomencla- ture committee (J.R. Idle, M. Eichelbaum, M. Ingelman- Sundberg, C.R. Wolf, K. Brosen, L. Balant, U.A. Meyer, D.W. Nebert, FJ Gonzalez, manuscript in preparation), we propose that the allele should be termed CYP2D6*6. The frameshift in exon 3 appears to be an inactivating muta- tion since it predicts a stop codon at amino acid 153 just downstream of the deletion. The translation product is predicted to be a truncated protein of 152 amino acids,

341

which is shorter than the truncated products coded by the CYP2D6A and CYP2D6B alleles previously shown to lack activity in vitro (Kagimoto et al. 1990). We also have ev- idence that the amino acid substitution encoded by the exon 4 mutation may result in the loss of CYP2D6 activ- ity. Our data show that c D N A clones with the mutation encode a protein that lacks activity with bufuralol as a substrate in vitro. We have also recently found this substi- tution in a black American subject who is negative for the exon 3 mutation and the other known inactivating muta- tions but who has a metabolic ratio of 3.77 consistent with impaired CYP2D6 activity (A. K. Daly, J .B.S. Leathart, J.R. Idle, unpublished observations). The CYP2D6*6 al- lele is similar to the CYP2D6B allele where a total of six other base changes, including some that result in amino acid substitutions affecting enzyme activity, are normally observed in association with the inactivating mutation (Kagimoto et al. 1990). It has recently been observed that some individuals showing impaired, but not poor, metab- olism of debrisoquine are positive for certain of the asso- ciated base changes but lack the inactivating mutation; this suggests that some of the CYP2D6B-associated muta- tions predate the inactivating mutation (Wang et al. 1993; Yokota et al. 1993; Armstrong et al. 1994). This might also be the case for the new allele. Screening for the new allele in addition to the CYP2D6A, CYP2D6B and CYP2D6D alleles in a group of known phenotypically poor metabolizers o f British origin (Daly et al. 1991) in- creased the poor metabolizer detection rate to 100%. However, among the white American subjects who show discrepancies between phenotype and genotype the new allele completely accounted for the discrepancy in only three out of the 23 subjects (subjects LA147, LA622 and LA678). Therefore other as yet unknown inactivating mu- tations probably occur in CYP2D6.

Acknowledgements We are grateful to BAT Ltd, the Council for Tobacco Research, USA, and the State of California Tobacco Re- lated Disease Research Program (grant 1RT-0104) for financial support.

References

Armstrong M, Fairbrother K, Idle JR, Daly AK (1994) The cy- tochrome P450 CYP2D6 altelic variant CYP2D6J and related polymorphisms in a European population. Pharmacogenetics 4: 47-57

Broly F, Gaedigk A, Heim M, Eichelbaum M, Morike K, Meyer UA (1991) Debrisoquine/sparteine hydroxylation genotype and phenotype: analysis of common mutations and alleles of CYP2D6 in a European population. DNA Cell Biol 10: 545- 558

Dahl M-L, Johansson I, Palmertz MP, Ingelman-Sundberg M, Sjo- qvist F (1992) Analysis of the CYP2D6 gene in relation to de- brisoquin and desipramine hydroxylation in a Swedish popula- tion. Clin Pharmacol Ther 51:12-17

Daly AK, Armstrong M, Monkman SC, Idle ME, Idle JR (199l) The genetic and metabolic criteria for the assignment of de- brisoquine hydroxylation (cytochrome P450IID6) phenotypes. Pharmacogenetics 1:33-41

Daly AK, Cholerton S, Gregory W, Idle JR (1993) Metabolic poly- morphisms. Pharmacol Ther 57:129-160

Evans WE, Relling MV (1991) Concordance of P4502D6 (de- brisoquine hydroxylase) phenotype and genotype: inability of dextromethorphan metabolic ratio to discriminate reliably het- erozygous and homozygous extensive metabolizers. Pharmaco- genetics 1:143-148

Kagimoto M, Heim M, Kagimoto K, Zeugin T, Meyer UA (1990) Multiple mutations of the human cytochrome P450IID6 gene (CYP2D6) in poor metabolisers of debrisoquine. J Biol Chem 265:17209-17214

Kimura S, Umeno M, Skoda RC, Meyer UA, Gonzalez FJ (t989) The human debrisoquine 4-hydroxylase (CYP2D) locus: se- quence and identification of a polymorphic CYP2D6 gene, a related gene, and a pseudogene. Am J Hum Genet 45:889-905

Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: a laboratoo' manual, 2nd edn. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

Wang S-L, Huang J-D, Lai M-D, Liu B-H, Lai M-L (1993) Mole- cular basis of genetic variation in debrisoquin hydroxylation in Chinese subjects: polymorphism in RFLP and DNA sequence of CYP2D6. Clin Pharmacol Ther 53:410-418

Yokota H, Tamura S, Furuya H, Kimura S, Watanabe M, Kana- zawa I, Kondo I, Gonzalez FJ (1993). Evidence for a new vari- ant allele CYP2D6J in a Japanese population associated with lower in vivo rates of sparteine metabolism. Pharmacogenetics 3:256-263