CYP2D6 Genotype and Debrisoquine Hydroxylation Phenotype in Cubans and Nicaraguans

CYP2D6 genotype and debrisoquine hydroxylation phenotype in Cubans and Nicaraguans

A LLerena1, P Dorado1, CYP2D6 genotype and debrisoquine metabolic ratio (MR) were analyzed 2 ´ 3,5 in 133 Nicaraguan Mestizos (NMs) and 260 Cubans divided into Cuban R Ramı´rez , I Gonzalez , Mestizos (CMs) and White Cubans (WCs). The frequencies of poor 4 1 MA´lvarez , EM Penãs-LLedo´ , metabolizers (MRX12.6) were 6% in NMs, 3.9% in CMs and 5.3% in BPe´rez4 and LR Calzadilla3 WCs. The frequencies of ultrarapid metabolizers (MRp0.1) were 0% in NMs, 2.3% in CMs and 5.3% in WCs. Mean (±s.d.) MR among extensive 1CICAB-CAIBER Centro de Investigacioń Clıńica, metabolizers (MRo12.6) was higher in NMs (1.5±1.6; n ¼ 118) than in Hospital Universitario Servicio Extremenõde CMs (1.0±1.3; n ¼ 124; P 0.001) and WCs (0.7±1.0; n ¼ 124; P 0.001). Salud SES, Facultad de Medicina Universidad o o de Extremadura, Badajoz, Spain; 2Facultad MR correlated with the ‘activity score’ of CYP2D6 genotypes (Po0.05; de Medicina, UNAM Universidad Autońoma r ¼À0.55). Mean MR was higher among NMs than WCs and CMs for groups Nacional de Nicaragua, Leoń, Nicaragua; classified as 1 (P 0.05) or 2 (P 0.01) ‘activity score’. In addition, mean 3 o o Hospital Psiquia´trico de La Habana, La Habana, (±s.d.) MR was higher among subjects carrying CYP2D6*17 than in CYP2D6 Cuba and 4Facultad de Medicina Calixto Garcıá, Instituto Superior de Ciencias Me´dicas de La wt/wt (Po0.001). The CYP2D6*10 allele was higher in NMs (3.1%) than in Habana, La Habana, Cuba CMs (0.8%; Po0.05) and WCs (0.4%; Po0.05). CYP2D6*17 allele was higher in CMs (10.2%) than WC (2.7%; Po0.005) and NMs (0%). Thus, the Correspondence: variability in CYP2D6 phenotypes found may be related to differences in Dr A Llerena, CICAB Clinical Research Center-Fundesalud, Hospital Universitario allele frequency among groups (that is, CYP2D6*10 and *17 highest in NMs Infanta Cristina, Badajoz 06080, Spain. and CMs, respectively). However, the influence of environmental factors or E-mail: [email protected] alleles different than those studied here cannot be ruled out. The Pharmacogenomics Journal (2012) 12, 176–183; doi:10.1038/tpj.2010.85; published online 7 December 2010

Keywords: CYP2D6; debrisoquine hydroxylation; phenotype; Latinos; Cubans; Nicaraguans

Introduction

Interethnic differences in cytochrome P450 polymorphism might be partially responsible for the variations in drug disposition between populations. Cytochrome P450 2D6 (CYP2D6) is of great importance for the metabolism of B25% of all clinically used drugs. Phenotyping with specific probe drugs, such as debrisoquine, sparteine or dextromethorphan, allows classifying individuals into two hydroxylation phenotypes: Poor (PMs) and extensive metabolizers (EMs), including a group of individuals with very rapid enzyme activity or the so called ultrarapid metabolizers (UMs).1 The worldwide distribution of PMs varies markedly. In Europe, PMs are more common (5–10%) than in Asia (B1%), whereas in Africa PMs are reported to range from 0 to 19%.2–4 In Latino 5 Current address: Hospital de Llerena, Servicio populations, the frequency of PMs is also variable ranging from 3.2 to 10% Extremen˜o de Salud SES, Spain. (Table 1). In Amerindian groups, PMs range from 0% in Mexican Tepehuano and in Panama´ Cuna to 4.4% in Ngawbes living in Colombia and Panama (Table 1). Received 27 June 2010; revised 27 August 2010; accepted 30 September 2010; However, few studies have estimated the frequency of UMs in Latino or published online 7 December 2010 Amerindian populations, being the frequency in Spaniards of 5.2%.17 CYP2D6 among Cubans and Nicaraguans A LLerena et al 177

Table 1 Percentages of CYP2D6 variant alleles and poor metabolizers in Latino, Amerindian and Spanish populations

Populations N Defective alleles Reduced activity alleles Duplicationa % PMsb References

3a 4a 5a 6a 10a 17a

Brazilian–African 87 1.2 6.3 4.2 0 4.2 9.2 2.3+1.2 Kohlrausch et al.5 Brazilian–European 92 0 10.3 1.1 2.2 2.7 2.2 4.9+1.6 Kohlrausch et al.5 Colombian 121 1.2 19.4 0.8 0 1.6 1.2 Isaza et al.6 Cuna (Panama) 170 0 (spt) Arias et al.7 89 0 (dbq) Jorge et al.8 Embera (Panama, 153 0 14 0 1.1 6.9 Jorge et al.8 Colombia) 153 2.2 (spt) Jorge et al.8 Mapuche (Chile) 84 0 3.6 4.2 1.8 Munõz et al.9 Mexican–American 349 o1 10.3 2.3 7.4 o1 1 Mendoza et al.10 285 3.2 (dxt) Mendoza et al.10 Mexican–American 264 0.2 10 1.7 0.4 2.8 0.2 0.8 Luo et al.11 236 10 (dxt) Luo et al.11 Mexican–American 50 17 2 1 2 3 Casner12 50 6 (dxt) Casner12 Mexican–Mestizo 243 1.4 11.2 2.7 12.4 1.7 12.8 Lo´pez et al.13 100 10 (dxt) Lo´pez et al.13 Mexican–Mestizo 110 0.9 13.1 0 2.3 Sosa-Macıás et al.14 88 6.8 (dxt) Sosa-Macıás et al.14 Ngawbe (Panama, 344 0 17.1 0 0.5 17.5 Jorge et al.8 Colombia) 344 4.4 (spt) Jorge et al.8 Spanish 925 5.2 (dbq) Llerena et al.15 Spanish 327 0.5 18.5 1.4 3.1 2.0 0 4.1+0.3 —c Tepehuano (Mexico) 85 0 0.6 0 0 Sosa-Macıás et al.14 58 0 (dxt) Sosa-Macıás et al.14 Uruguayan 302 7.3 (dxt) Estevez et al.16

Abbreviations: dbq; debrisoquine; dxt, dextromethorphan; N, number of subjects; PMs, poor metabolizers; spt sparteine. aDuplications of functional plus non-functional CYP2D6 alleles. bCYP2D6 phenotyping test drugs: dbq, dxt and spt. cLLerena and Dorado, unpublished 2010.

The CYP2D6 is a highly polymorphic gene localized phenotype has been only observed in individuals with zero on chromosome 22q13.1. The CYP2D6 alleles are related active genes, whereas using the most common classification to absent, decreased, normal and increased catalytic activ- of 0, 1, 2 and 42 active genes. This correlation has been ity (http://www.cypalleles.ki.se/cyp2d6.htm). CYP2D6*3, largely improved since Gaedigk et al.22 proposed the ‘activity CYP2D6*4, CYP2D6*5, CYP2D6*6 are the most important score’ concept, which has been probed using dextromethor- variants related to absent enzyme activity and CYP2D6*10 phan as substrate test. Therefore, it first appears of scientific and CYP2D6*17 to reduced enzyme activity. The frequency interest to analyze it using debrisoquine, which is one of the of CYP2D6 alleles varies amongst ethnic groups. The most most accurate markers of CYP2D6 hydroxylation capacity. frequent non-functional alleles for Caucasians are in order Second, considering that United States Food and Drug CYP2D6*4, CYP2D6*5 and CYP2D6*3 (Table 1). CYP2D6*10 Association has already approved drugs with prescribing is more frequent in Asians,18 and CYP2D6*17 in Black guidelines targeted toward specific ethnic groups, it is also of Africans.19 Considering that Latinos have besides Caucasian, relevance to study groups that might be representative of Amerindian and African ancestors, a large representation of Latin American populations, including the largest minority non-functional but also of reduced function alleles is (15%) in United States of America. The studied populations expected to be found. The frequency of individuals with in this study resemble the constitution of Latinos; three duplicated CYP2D6 active alleles has been set around 5% in groups with an expected higher Caucasian (White Cubans; Spain,20 0.69% in Asia21 and 29% in Ethiopia.3 However, WC), African (Cuban Mestizos; CM) or Amerindian (Nicar- there are few reports in Latino populations, and it is not aguan Mestizos; NM) ancestry component. Therefore, this clearly shown whether the multiplication is of active alleles study was aimed to determine whether there were differ- (Table 1). ences in the frequency of CYP2D6 PMs and UMs, and the CYP2D6 genotyping is widely used to evaluate enzyme most common and relevant CYP2D6 alleles across the activity, but a perfect match between genotype and studied populations. The genotype–phenotype relationship

The Pharmacogenomics Journal CYP2D6 among Cubans and Nicaraguans A LLerena et al 178

using ‘activity score’ was also analyzed, as well as the drug and that of 4-hydroxydebrisoquine in the 0–8 h urine functional implication of the analyzed alleles (for example, output. The antimode for each histogram was calculated CYP2D6*17). according to the Probit plot. Individuals with an MR 412.6 (log10 MRX1.1) were considered as PMs, whereas the rest Materials and methods were classified as EMs. Among EMs, a cut-off point of MRp0.1 was used to classify the UMs.17,15,29 Subjects and procedure The Cuban population was divided into two groups: Genotyping procedure ‘Whites’ WCs that were those individuals with four Blood samples (10 ml) were collected in EDTA tubes, and Caucasian grandparents and ‘Mestizos’ CMs that repre- DNA was extracted using the QIAamp DNA blood kit sented the rest. Debrisoquine hydroxylation phenotype (QIAGEN, Hilden, Germany). The CYP2D6 genotype was was evaluated in 131 WCs, 129 CMs and 133 NMs. The analyzed by PCR and PCR–restriction fragment length CYP2D6 genotype was analyzed in 130 WCs, 126 CMs and polymorphism for the CYP2D6 *3,*4,*5,*6,*10,*17, and 30 98 NMs healthy volunteers. The group of WCs (45 males and multiplicated alleles. 86 females) had a mean (±s.d.) age of 24.8±7.2 (range The relationship between CYP2D6 genotype and pheno- 18–56) years and a body weight of 61.8±12.6 (43–136) kg. type was evaluated assigning a value relative to each variant 22 The group of 129 CM (20 males and 109 females) showed a CYP2D6 allele based on the ‘activity score’ system. The mean (±s.d.) age of 24.4±7.8 (range 18–65) years and a body value assigned to the reference allele CYP2D6 wt (*1 or *2) weight of 59.8±12.6 (40–120) kg. The NM group (54 males was 1, whereas to CYP2D6*3,*4,*4xN,*5,*6 variants were and 79 females) had a mean (±s.d.) age of 25±7(range18–61) assigned the value of 0, to CYP2D6*10 and *17 alleles were years and a body weight of 60.1±12.2 (42–98) kg. assigned 0.5, and to multiplications of CYP2D6wt (*1xN or The subjects were mostly Medical and Nursing school *2xN) allele with 2. students and staff from the Faculty of Medicine ‘Calixto Garcı´a’, Psychiatric Hospital of Havana (Cuba) and ‘Uni- Statistical analysis versidad Nacional Auto´noma de Nicaragua’, Leon (Nicar- Mean MR values of EMs in the three different populations agua). Both the Cuban and the Nicaraguan populations were were compared by using unpaired t-test. The correlations recruited following the same protocol previously used in between ‘activity scores’ and MR where analyzed by Spear- Spain,23 with a similar research team, in order to avoid man’s rank correlation coefficient. Mean MRs among potential selection bias because of the influence of CYP2D6 different ‘activity scores’ were compared with non-para- genotype on psychological features.24 A routine clinical metric Kruskal–Wallis test with Dunn’s multiple comparison examination was carried out and the medical history was test. The differences in CYP2D6 allele frequencies were 2 taken before the study. Volunteers with a previous history of compared by using the w -test and/or Fisher’s exact test. adverse drug effects and those with any drug intake in the 2 P-values o0.05 were regarded as statistically significant. weeks before the study were excluded. The subjects were Statistical analysis was performed using STATISTICA 4.3 informed about the aims of the study and gave their written (StatSoft, Tulsa, OK, USA) and GraphPad Prism 3.02 Informed Consent before their participation. The study was (GraphPad Software , San Diego, CA, USA) programs. conducted according to the Helsinki Declaration, and it was approved by the Local Ethics Committees of the institutions Results involved. Although some CYP2D6 allele frequencies have already been published for the Nicaraguan population,25 a Debrisoquine hydroxylation phenotype new genotyping analysis including CYP2D6*6, CYP2D6*17 The MR histograms revealed a bimodal distribution of and CYP2D6*4xN has been carried out in order to improve CYP2D6 activity in the three populations (Figure 1). The the accuracy of the geno–pheno relationship analysis. Some frequency of individuals classified as PMs was 5.3, 3.9 and data about Cubans’ debrisoquine hydroxylation phenotype 6.0% in WCs, CMs and NMs, respectively. The frequency of have already been published.26,27 UMs was 5.3% for WC, 2.3% for CM and 0% for NM. Mean (±s.d.) MR values among EMs (Figure 1) was higher among Phenotyping procedure NMs (1.5±1.6; n ¼ 118; Po0.001) than among CWs The subjects took a single oral dose of 10 mg of debrisoquine (0.7±1.0; n ¼ 124; Po0.001) and CMs (1.0±1.3; n ¼ 124; sulfate (Declinax; Hoffman-La Roche, Switzerland) and after Po0.001). No significant association was found with gender, an overnight fast all urine was collected over 8 h. The tobacco smoking or alcohol intake habits in Nicaraguans samples were kept frozen at –201C and transported from or Cubans. Havana (Cuba) and Leon (Nicaragua) to Badajoz (Spain) where analyses were carried out. The analytical determina- CYP2D6 genetic polymorphism tion was performed by high-performance liquid chromato- The frequencies of the CYP2D6 genotypes in all the studied graphy–ultraviolet according to our previously reported populations (Table 2) corresponded with those predicted by methodology.23,28 Debrisoquine hydroxylation phenotype the Hardy–Weinberg law. The percentage of individuals with was assigned by the debrisoquine metabolic ratio (MR), zero CYP2D6 active genes was 2.3% in WCs, 2.4% in CMs defined as the ratio of the molar concentration of the parent and 4.1% in NMs (Table 3).

The Pharmacogenomics Journal CYP2D6 among Cubans and Nicaraguans A LLerena et al 179

UMs White-Cubans PMs allele was higher in NMs (3.1%) than in WCs (0.4%; Po0.05) 18 100 and CMs (0.8%; Po0.05). With regard to duplications/ multiplications (CYP2D6*1xN or *2xN alleles), although in 16 NMs were observed to be lower (2%) than in WCs (3.8%), 80 14 and CMs (4.7%), there were no statistical differences. 12 60 CYP2D6 genetic polymorphism and debrisoquine hydroxylation 10 phenotype 8 A total of 123 WCs, 120 CMs and 93 NMs were phenotyped 40 and genotyped to CYP2D6. All individuals CYP2D6*4/*4 and 6 CYP2D6*4/*5 were also phenotypically PMs (n ¼ 10). The Number of subjetcs 4 Probit (percentage %) rest of phenotypically PMs (n ¼ 5) were individuals with 5.3% 5.3% 20 2 CYP2D6 wt/*10 (n ¼ 2), wt/*4,*5/*10 or wt/*4xN (n ¼ 3) genotypes. Among UMs, 40% carried multiplications of 0 0 -2 -1 0 1 2 CYP2D6 active alleles (42 active genes). In addition, these debrisoquine/4-OH debrisoquine (log10) individuals with more than two active genes versus those who carried one or two copies presented a lower MR UMs Cuban-Mestizos PMs (Po0.05). 18 100 MR correlated with the ‘activity score’ of CYP2D6 genotypes among all 336 healthy volunteers (Po0.05; 16 r ¼À0.55; Figure 2a). Moreover, MR varied across subjects 80 14 with different CYP2D6 ‘activity score’ (Po0.0001). Figure 2b 12 shows the relationship between ‘activity score’ and MR in 60 the three populations. Overall, NMs seemed to show a 10 higher MR than WCs or CMs (Figure 2b). In particular, MR 8 mean was higher in NMs than in WCs and CMs for the 40 6 groups classified as 1 (Po0.05) or 2 (Po0.01) ‘activity score’.

Number of subjetcs Mean (±s.d.) MR among subjects with CYP2D6wt/*17 4 Probit (percentage %) ± ± 2.3% 3.9% 20 (n ¼ 24; 1.42 2.05) and *4/*17 (n ¼ 6; 4.10 2.48) genotypes 2 was higher (Po0.001) than among those with CYP2D6wt/wt genotype (n ¼ 124; 0.59±0.77). 0 0 -2 -1 0 1 2 debrisoquine/4-OH debrisoquine (log10) Discussion

Nicaraguan-Mestizos UMs PMs This study demonstrates the existence of CYP2D6 hydro- 18 100 xylation polymorphism using debrisoquine as a test sub- 16 strate in Latino populations of WC, CM and NM (PMs were 5.3, 3.9, 6% and UMs were 5.3, 2.3, 0%, respectively). MR 14 80 among EMs was higher in NMs than in CMs and WCs. MR 12 was related to CYP2D6 ‘activity score’. MR was higher 60 10 among NMs than WCs and CMs for groups classified as 1 or 2 ‘activity score’. In addition, MR was higher among subjects 8 40 carrying CYP2D6*17 than in CYP2D6 wt/wt. Allelic dis- 6 tributions of CYP2D6 genes differed among the Latino Number of subjetcs

Probit (percentage %) groups studied here indicating ethnic variety within this 4 20 0.0% 6.0% large population (CYP2D6*10 and CYP2D6*17 were higher 2 in NMs and CMs, respectively). Thus, the variability in 0 0 CYP2D6 phenotypes found may be related to differences in -2 -1 0 1 2 allele frequency among groups. debrisoquine/4-OH debrisoquine (log10) Debrisoquine hydroxylation phenotypes Figure 1 Histograms of MRs of debrisoquine (log10) in 131 White Cuban, 129 Cuban Mestizos and 133 Nicaraguan Mestizo. Dotted lines: PMs PM phenotype frequencies were all similar to that found in 17 (log10MR41.1) and UMs (log10MRspÀ1). Continuous line shows the mean White Spaniards (4.9%), and higher than those reported in of MR among EMs. Dotted curves indicate the accumulative frequencies. Oriental31–33 and African populations.3 Studies in Ngawbe´ Guaymı´ Amerindians from Central America have reported a The percentage of CYP2D6*17 alleles among CMs (10.2%) similar frequency of PMs (around 5%). However, evidence was higher than among WC (2.7%; Po0.005) individuals about the absence of PMs in Cuna Amerindians also exists and NMs, in whom it was not detected. The CYP2D6*10 (Table 1).

The Pharmacogenomics Journal CYP2D6 among Cubans and Nicaraguans A LLerena et al 180

Table 2 Frequencies of CYP2D6 genotypes among White Cuban, Cuban Mestizo and Nicaraguan Mestizo populations

Genotype Active genes Active score White Cubans (n ¼ 130) Cuban Mestizos (n ¼ 126) Nicaraguan Mestizos (n ¼ 98)

N % 95% CI N % 95% CI N % 95% CI

*3/*4 00————————— *4/*4 0 0 3 2.31 0.49–6.86 2 1.57 0.08–5.92 2 2.04 0.11–7.59 *4/*5 0 0 — — — 1 0.79 0.00–4.76 2 2.04 0.11–7.59 *4/*6 00————————— *5/*6 00————————— *6/*6 00————————— *4/*10 1 0.5 — — — — — — — — — *4/*17 1 0.5 2 1.54 0.07–5.79 3 2.36 0.50–7.02 — — — *5/*10 1 0.5 — — — — — — 1 1.02 0.01–6.11 wt/*3 11————————— wt/*4 1 1 29 22.3 16.0–30.2 24 18.9 13.0–26.6 21 21.4 14.4–30.6 wt/*5 1 1 5 3.85 1.42–8.92 2 1.57 0.08–5.92 6 6.12 2.58–13.0 wt/*6 1 1 1 0.77 0.00–4.66 2 1.57 0.08–5.92 — — — wt/*4xN 1 1 1 0.77 0.00–4.66 — — — — — — *17/*17 2 1 — — — 1 0.79 0.00–4.76 — — — *10/*10 2 1 — — — 1 0.79 0.00–4.76 — — — wt/*10 2 1.5 1 0.77 0.00–4.66 — — — 5 5.10 1.91–11.7 wt/*17 2 1.5 4 3.08 0.94–7.91 21 16.5 11.0–24.0 — — — wt/wt 2 2 74 56.9 48.3–65.1 57 44.9 35.8–52.8 57 58.2 48.3–67.5 wtxN/*4 2 2 1 0.77 0.00–4.66 4 3.22 0.96–8.09 1 1.02 0.01–6.11 wtxN/*5 2 2 — — — 1 0.79 0.00–4.76 — — — wtxN/*6 2 2 1 0.77 0.00–4.66 1 0.79 0.00–4.76 — — — wt/wtxN 42 42 7 5.38 2.44–10.9 6 4.72 1.97–10.1 3 3.06 0.67–9.00 wtxN/*10 42 42—— — —— — —— — wtxN/*17 42 42 1 0.77 0.00–4.66 — — — — — —

Abbreviations: CI, confidence interval; N, number of subjects.

Table 3 Frequencies of CYP2D6 variant alleles among White Cuban, Cuban Mestizo and Nicaraguan Mestizo populations

CYP2D6 alleles White Cubans (n ¼ 260) Cuban Mestizos (n ¼ 252) Nicaraguan Mestizos (n ¼ 196)

Frequency 95% CI Frequency 95% CI Frequency 95% CI wt (*1 or *2) 0.754 0.698–0.802 0.663 0.606–0.722 0.761 0.696–0.815 *3 0— 0 — 0 — *4 0.146 0.108–0.195 0.143 0.108–0.196 0.142 0.100–0.199 *5 0.019 0.007–0.046 0.016 0.005–0.042 0.046 0.023–0.086 *6 0.008 0.000–0.030 0.012 0.002–0.036 0 — *10 0.004 0.000–0.024 0.008 0.000–0.030 0.031 0.013–0.067 *17 0.027 0.012–0.056 0.102 0.071–0.147 0 —

Multiplications wt (*1 or *2)xN 0.038 0.020–0.070 0.047 0.027–0.082 0.020 0.006–0.053 *4xN 0.004 0.000–0.024 0 — 0 —

Abbreviations: CI, confidence interval; N, number of alleles.

Among phenotypically EMs, MR was higher in NMs than CYP2D6 genetic polymorphism: frequency of defective (null or WCs or CMs, which may be probably because of the deficient activity) alleles differences in the presence of defective or null alleles The frequency of CYP2D6*4 in WCs was similar to other (CYP2D6*10 was highest in NMs), as well as to other American Caucasian populations.34–36 In CMs, it was environmental factors, such as food, lifestyle and so on. higher (14.6%) than in Black populations.3,34,35,37–41 In UM subjects were not found in NMs, whereas the NMs (14.2%) was similar to some Amerindians8 and frequency in WCs was similar to that found in White Spaniards (Table 1), and higher than in Mapuches and Spaniards (5.2%).17 Tepehuanos.9,14

The Pharmacogenomics Journal CYP2D6 among Cubans and Nicaraguans A LLerena et al 181

1000 reported in Central or South Americans,8,6 Caucasians from All subjects America,35,36 and other European populations.34,42 100 PMs The frequency of CYP2D6 duplication/multiplication in 10 WCs and CMs (4–5%) was similar to the one found in Spaniards17 (Tables 1 and 3). EMs

) DBQ/4DBQ 1 10 Effect of CYP2D6 genetic polymorphism on debrisoquine 0.1 hydroxylation

MR (log All individuals homozygous for the non-functional CYP2D6 0.01 alleles were also found to be phenotypically PMs. However, Number of subjects 10 6 86 28188 18 0.001 only 67% of PMs carried zero active genes. Therefore, 0 0.5 1 1.5 2 >2 reduced activity alleles need to be studied, similar to the Activity Score CYP2D6*29 reported in Africans from Tanzania,44 which is 1000 acknowledged as a limitation of this study. Moreover, only White-Cubans Cuban-Mestizos 23.5% of subjects with CYP2D6 gene multiplications were Nicaraguan-Mestizos PMs 100 phenotypically UMs, and only 40% of individuals with UM phenotype showed duplications. This result is consistent 10 with other previously observed in Caucasian populations.45 EMs It is also in agreement with previous findings showing that ) DBQ/4DBQ 1

10 genotyping for duplicated/multiplicated CYP2D6 active 0.1 alleles only explains a fraction (10–30%) of the ultrarapid 29,46,47

MR (log phenotype observed in Caucasian populations. 0.01 The results also show that CMs present a higher MR than Number of subjects 3/3/4 2/3/1 33/28/25 5/20/4 72/60/56 8/6/3 0.001 WCs, which is consistent with results from Black-African 0 0.5 1 1.5 2 >2 descendant studies.35,36 This could be explained by the Activity Score higher frequency of CYP2D6*17 defective alleles in CMs Figure 2 Relationship between debrisoquine MR and CYP2D6 ‘activity than in WCs, which is also shown by present results about score’ among all individuals (a) and among three healthy populations lower CYP2D6 activity in individuals carrying CYP2D6*17 of White Cuban, Cuban Mestizo, and Nicaraguan Mestizo (b). than in homozygous wild-type subjects (Figure 3). On the other hand, the MR of NMs was higher than the MR of p <0.05 Cubans, which may be related to the presence of CYP2D6*10 1.2 p <0.05 (Figure 2b). Moreover, differences between populations are 1.0 observed across activity score groups. Thus, these facts are 0.8 probably because of the environmental and genetic factors 0.6 that may influence variability in CYP2D6 activity. 0.4 This study has shown a relationship between CYP2D6

0.2 genotyping and debrisoquine phenotyping using ‘activity score’22 and variability across Latin groups. This is of clinical -0.0 relevance considering CYP2D6 involvement in the metabo- -0.2 lism of a large number of commonly prescribed drugs.1 -0.4 Thus, CYP2D6 genotyping may be a useful tool to predict debrisoquine/4-OH debrisoquine debrisoquine/4-OH ) 17 10 -0.6 drug related side effects, interactions or therapeutic failure -0.8 in these populations. Although population pharmacoge-

MR (log -1.0 netics could be of some use in terms of public health care

-1.2 policies, several studies have shown that genetic differences wt/wt wt/*17 wtxN/*17 *17/*17 *4/*17 between individuals are higher than between ethnic groups CYP2D6 genotype or races.1,2,17 Thus, in a heterogeneous population, such as Figure 3 Relationship between debrisoquine MR and CYP2D6*17 allele Latinos, it would be more adequate to consider the among Cuban healthy volunteers with CYP2D6 wt/wt genotypes and individual than an ethnic-based approach for personalized 48 CYP2D6*17 allele carriers (n ¼ 156). medicine. In the light of these findings, we could hypothesize that drugs metabolized by CYP2D6 may show The CYP2D6*10 defective allele occurred up to three times different profile of efficacy/adverse effects in NM, CM and more frequently in NMs than in WCs and CMs (Table 1). WC populations. Therefore, further research on different CYP2D6*17 allele with diminished function occurred up to ethnic groups in America is warranted. four times more frequently in CMs than in WCs, whereas it was not found in the NM population. CYP2D6*17 Conflict of interest frequency in CMs was similar to those observed in black populations,34,35,38,39,41–43 and higher than those The authors declared no conflict of interest.

The Pharmacogenomics Journal CYP2D6 among Cubans and Nicaraguans A LLerena et al 182

Acknowledgments 16 Estevez F, Giusti M, Parrillo S, Oxandabarat J. Dextromethorphan O-demethylation polymorphism in the Uruguayan population. Eur J This study was supported by Plan Nacional de Investigacioń Clin Pharmacol 1997; 52: 417–418. ˜ ´ Cientı´fica, Desarrollo e Innovacioń Tecnolo´gica (I þ D þ I) and 17 LLerena A, Dorado P, Penas-Lledo EM. Pharmacogenetics of debrisoquine and its use as a marker for CYP2D6 hydroxylation capacity. Fondo Social Europeo of the European Union (FEDER), Instituto Pharmacogenomics 2009; 10: 17–28. de Salud Carlos III-FIS CIBERSAM and CAIBER and Research Grants 18 Johansson I, Oscarson M, Yue QY, Bertilsson L, Sjo¨qvist F, Ingelman- (PI06/1681; CP06/0030 to PD), Ministerio de Educacioń y Ciencia Sundberg M. Genetic analysis of the Chinese cytochrome P4502D (SAF2006-13589), by AEXCID Cooperacioń Extremenã of the Junta locus: characterization of variant CYP2D6 genes present in subjects with de Extremadura (9IA006), and coordinated in the network Red diminished capacity for debrisoquine hydroxylation. Mol Pharmacol Iberoamericana de Farmacogene´tica y Farmacogeno´mica (CYTED 1994; 46: 452–459. 206RT0290). 19 Masimirembwa C, Persson I, Bertilsson L, Hasler J, Ingelman-Sundberg M. A novel mutant variant of the CYP2D6 gene (CYP2D6*17) common in a black African population: association with diminished debrisoquine References hydroxylase activity. Br J Clin Pharmacol 1996; 42: 713–719. 20 LLerena A, Dorado P, Penãs-Lledo´ EM, Caćeres MC, De la Rubia A. Low 1 Ingelman-Sundberg M. Genetic polymorphisms of cytochrome P450 frequency of CYP2D6 poor metabolizers among schizophrenia patients. 2D6 (CYP2D6): clinical consequences, evolutionary aspects and func- Pharmacogenomics J 2007; 7: 408–410. tional diversity. Pharmacogenomics J 2005; 5:6–13. 21 Sheng HH, Zeng AP, Zhu WX, Zhu RF, Li HM, Zhu ZD et al. Allelic 2 LLerena A, Cobaleda J, Martıńez C, Benı´tez J. Interethnic differences in distributions of CYP2D6 gene copy number variation in the Eastern Han drug metabolism: influence of genetic and environmental factors on Chinese population. Acta Pharmacol Sin 2007; 28: 279–286. debrisoquine hydroxylation phenotype. Eur J Drug Metab Pharmacokinet 22 Gaedigk A, Simon SD, Pearce RE, Bradford LD, Kennedy MJ, Leeder JS. 1996; 21: 129–138. The CYP2D6 activity score: translating genotype information into 3 Aklillu E, Persson I, Bertilsson L, Johansson I, Rodrigues F, Ingelman- a qualitative measure of phenotype. Clin Pharmacol Ther 2008; 83: Sundberg M. Frequent distribution of ultrarapid metabolizers of 234–242. debrisoquine in an ethiopian population carrying duplicated and 23 LLerena A, Herraı´z AG, Cobaleda J, Johansson I, Dahl ML. Debrisoquin multiduplicated functional CYP2D6 alleles. J Pharmacol Exp Ther 1996; and mephenytoin hydroxylation phenotypes and CYP2D6 genotype in 278: 441–446. patients treated with neuroleptic and antidepressant agents. Clin Pharmacol 4 Bernard S, Neville KA, Nguyen AT, Flockhart DA. Interethnic differences Ther 1993; 54: 606–611. in genetic polymorphisms of CYP2D6 in the U.S. population: clinical 24 Penãs-LLedo´ EM, Dorado P, Pacheco R, Gonza´lez I, LLerena A. Relation implications. Oncologist 2006; 11: 126–135. between CYP2D6 genotype, personality, neurocognition and overall 5 Kohlrausch FB, Gama CS, Lobato MI, Belmonte-de-Abreu P, Gesteira A, psychopathology in healthy volunteers. Pharmacogenomics 2009; 10: Barros F et al. Molecular diversity at the CYP2D6 locus in healthy 1111–1120. and schizophrenic southern Brazilians. Pharmacogenomics 2009; 10: 25 Aguńdez JA, Ramirez R, Hernandez M, Llerena A, Benı´tez J. Molecular 1457–1466. heterogeneity at the CYP2D gene locus in Nicaraguans: impact of gene- 6 Isaza CA, Henao J, Lo´pez AM, Cacabelos R. Isolation, sequence flow from Europe. Pharmacogenetics 1997; 7: 337–340. and genotyping of the drug metabolizer CYP2D6 gene in the 26 Gonza´lez I, Pe´rez B, Alvarez M, Dorado P, LLerena A. Study of debrisoquine Colombian population. Methods Find Exp Clin Pharmacol 2000; 22: hydroxylation polymorphism (CYP2D6)intheCubanpopulationcompared 695–705. to Spaniards. Med Clin (Barc) 2007; 128: 772–774. 7 Arias TD, Jorge LF, Lee D, Barrantes R, Inaba T. The oxidative 27 Gonza´lez I, Penãs-Lledo´ EM, Pe´rez B, Dorado P, Alvarez M, LLerena A. metabolism of sparteine in the Cuna Amerindians of Panama: absence Relation between CYP2D6 phenotype and genotype and personality in of evidence for deficient metabolizers. Clin Pharmacol Ther 1988; 43: healthy volunteers. Pharmacogenomics 2008; 9: 833–840. 456–465. 28 Dorado P, Berecz R, Caćeres MC, Gonza´lez I, Cobaleda J, Llerena A. 8 Jorge LF, Eichelbaum M, Griese EU, Inaba T, Arias TD. Comparative Determination of debrisoquine and 4-hydroxydebrisoquine by high- evolutionary pharmacogenetics of CYP2D6 in Ngawbe and Embera performance liquid chromatography: application to the evaluation Amerindians of Panama and Colombia: role of selection versus drift in of CYP2D6 genotype and debrisoquine metabolic ratio relationship. world populations. Pharmacogenetics 1999; 9: 217–228. Clin Chem Lab Med 2005; 43: 275–279. 9 Munõz S, Vollrath V, Vallejos MP, Miquel JF, Covarrubias C, Raddatz A 29 Dahl ML, Johansson I, Bertilsson L, Ingelman-Sundberg M, Sjo¨qvist F. et al. Genetic polymorphisms of CYP2D6, CYP1A1 and CYP2E1 in Ultrarapid hydroxylation of debrisoquine in a Swedish population. the South-Amerindian population of Chile. Pharmacogenetics 1998; 8: Analysis of the molecular genetic basis. J Pharmacol Exp Ther 1995; 274: 343–351. 516–520. 10 Mendoza R, Wan YJ, Poland RE, Smith M, Zheng Y, Berman N et al. 30 Dorado P, Caćeres MC, Pozo-Guisado E, Wong ML, Licinio J, Llerena A. CYP2D6 polymorphism in a Mexican American population. Clin Pharmacol Development of a PCR-based strategy for CYP2D6 genotyping Ther 2001; 70: 552–560. including gene multiplication of worldwide potential use. Biotechniques 11 Luo HR, Gaedigk A, Aloumanis V, Wan YJ. Identification of CYP2D6 2005; 39: 571–574. impaired functional alleles in Mexican Americans. Eur J Clin Pharmacol 31 Lou YC, Ying L, Bertilsson L, Sjo¨qvist F. Low frequency of slow 2005; 61: 797–802. debrisoquine hydroxylation in a native Chinese population. Lancet 12 Casner PR. The effect of CYP2D6 polymorphisms on dextro- 1987; 2: 852–853. methorphan metabolism in Mexican Americans. J Clin Pharmacol 32 Nakamura K, Goto F, Ray WA, McAllister CB, Jacqz E, Wilkinson GR et al. 2005; 45: 1230–1235. Interethnic differences in genetic polymorphism of debrisoquin and 13 Lo´pez M, Guerrero J, Jung-Cook H, Alonso ME. CYP2D6 genotype and mephenytoin hydroxylation between Japanese and Caucasian popula- phenotype determination in a Mexican Mestizo population. Eur J Clin tions. Clin Pharmacol Ther 1985; 38: 402–408. Pharmacol 2005; 61: 749–754. 33 Sohn DR, Kusaka M, Shin SG, Jang IJ, Chiba K, Ishizaki T. Utility of a one- 14 Sosa-Macıás M, Elizondo G, Flores-Pe´rez C, Flores-Pe´rez J, Bradley- point (3-hour postdose) plasma metabolic ratio as a phenotyping test Alvarez F, Alanis-Ban˜uelos RE et al. CYP2D6 genotype and phenotype using metoprolol in two east Asian populations. Ther Drug Monit 1992; in Amerindians of Tepehuano origin and Mestizos of Durango, Mexico. 14: 184–189. J Clin Pharmacol 2006; 46: 527–536. 34 Leathart JB, London SJ, Steward A, Adams JD, Idle JR, Daly AK. CYP2D6 15 LLerena A, Edman G, Cobaleda J, Benı´tez J, Schalling D, Bertilsson L. phenotype-genotype relationships in African-Americans and Caucasians Relationship between personality and debrisoquine hydroxylation in Los Angeles. Pharmacogenetics 1998; 8: 529–541. capacity. Suggestion of an endogenous neuroactive substrate or 35 Wan YJ, Poland RE, Han G, Konishi T, Zheng YP, Berman N et al. Analysis product of the cytochrome P4502D6. Acta Psychiatr Scand 1993; 87: of the CYP2D6 gene polymorphism and enzyme activity in African- 23–28. Americans in southern California. Pharmacogenetics 2001; 11: 489–499.

The Pharmacogenomics Journal CYP2D6 among Cubans and Nicaraguans A LLerena et al 183

36 Gaedigk A, Gotschall RR, Forbes NS, Simon SD, Kearns GL, Leeder JS. Tanzanians: analyses of the CYP2D6 genotype and phenotype. Optimization of cytochrome P4502D6 (CYP2D6) phenotype assign- Pharmacogenetics 1999; 9: 707–714. ment using a genotyping algorithm based on allele frequency data. 43 Panserat S, Sica L, Ge´rard N, Mathieu H, Jacqz-Aigrain E, Krishna- Pharmacogenetics 1999; 9: 669–682. moorthy R. CYP2D6 polymorphism in a Gabonese population: 37 Evans WE, Relling MV, Rahman A, McLeod HL, Scott EP, Lin JS. Genetic contribution of the CYP2D6*2 and CYP2D6*17 alleles to the high basis for a lower prevalence of deficient CYP2D6 oxidative drug prevalence of the intermediate metabolic phenotype. Br J Clin metabolism phenotypes in black Americans. J Clin Invest 1993; 91: Pharmacol 1999; 47: 121–124. 2150–2154. 44 Wennerholm A, Johansson I, Hidestrand M, Bertilsson L, Gustafsson LL, 38 Masimirembwa C, Hasler J, Bertilsson L, Johansson I, Ekberg O, Ingelman-Sundberg M. Characterization of the CYP2D6*29 allele Ingelman-Sundberg M. Phenotype and genotype analysis of debriso- commonly present in a black Tanzanian population causing reduced quine hydroxylase (CYP2D6) in a black Zimbabwean population. catalytic activity. Pharmacogenetics 2001; 11: 417–427. Reduced enzyme activity and evaluation of metabolic correlation of 45 Løvlie R, Daly AK, Matre GE, Molven A, Steen VM. Polymorphisms CYP2D6 probe drugs. Eur J Clin Pharmacol 1996; 51: 117–122. in CYP2D6 duplication-negative individuals with the ultrarapid meta- 39 Griese EU, Asante-Poku S, Ofori-Adjei D, Mikus G, Eichelbaum M. bolizer phenotype: a role for the CYP2D6*35 allele in ultrarapid Analysis of the CYP2D6 gene mutations and their consequences for metabolism? Pharmacogenetics 2001; 11: 45–55. enzyme function in a West African population. Pharmacogenetics 1999; 46 Sachse C, Brockmo¨ller J, Bauer S, Roots I. Cytochrome P450 2D6 9: 715–723. variants in a Caucasian population: allele frequencies and phenotypic 40 Dandara C, Masimirembwa CM, Magimba A, Sayi J, Kaaya S, Sommers consequences. Am J Hum Genet 1997; 60: 284–295. DK et al. Genetic polymorphism of CYP2D6 and CYP2C19 in east- and 47 Bathum L, Johansson I, Ingelman-Sundberg M, Hørder M, Brøsen K. southern African populations including psychiatric patients. Eur J Clin Ultrarapid metabolism of sparteine: frequency of alleles with duplicated Pharmacol 2001; 57: 11–17. CYP2D6 genes in a Danish population as determined by restriction 41 Gaedigk A, Bradford LD, Marcucci KA, Leeder JS. Unique CYP2D6 fragment length polymorphism and long polymerase chain reaction. activity distribution and genotype-phenotype discordance in black Pharmacogenetics 1998; 8: 119–123. Americans. Clin Pharmacol Ther 2002; 72: 76–89. 48 Suarez-Kurtz G, Pena SD. Pharmacogenomics in the Americas: 42 Wennerholm A, Johansson I, Massele AY, Lande M, Alm C, Aden-Abdi Y the impact of genetic admixture. Curr Drug Targets 2006; 7: et al. Decreased capacity for debrisoquine metabolism among black 1649–1658.

The Pharmacogenomics Journal