Research Article For reprint orders, please contact:
[email protected]
Flavin-containing monooxygenase 3 polymorphisms in 13 ethnic populations from Europe, East Asia and sub-Saharan Africa: frequency and linkage analysis Aims: To investigate intra- and inter-ethnic differences in three widespread (E158K, V257M and E308G) and two African-specific (D132H and L360P) flavin-containing monooxygenase 3 (FMO3) polymorphisms. Materials & methods: Allele frequencies were determined by TaqMan® allelic discrimination assay in 2152 healthy volunteers from Europe (Swedes, Italians and Turks), East Asia (Japanese) and sub-Saharan Africa (nine ethnic groups covering eastern, southern and western regions), followed by haplotype and linkage analysis. Results: Significant subpopulation differences (p A ( V257M) a nd g.21443A>G (E308G), have been identified in all populations examined so far with considerable frequency differences between the major ethnicities [9–12] . A variable extent of decreased FMO3 enzyme activity, often in a substratedependent manner, has been characterized as the functional impact of the three variants in vitro [12–15] . Linkage disequilibrium between E158K and E308G is observed in several populations. Simultaneous occurrence of K158 and G308 variants results in a more pronounced decrease in FMO3 activity than the individual ones do separately [9,12,16,17] . Carriers of the compound variant (K158/G308) have shown lower metabolic efficiency for N-oxidation of ranitidine [9,18] and sulindac [19,20] compared with nonc arriers. A transient or mild form of TMAuria can be triggered by exposure to trimethylamine overload [17,21] and hormonal changes related to menstruation [22] , in subjects homozygous for both K158 and G308. In contrast to all other known FMO3 SNPs, g.21599 T>C (L360P) is the only variant associated with enhanced catalytic efficiency in vitro [12,23] . However, P360, as well as another variant g.15089 G>C (D132H), has been found solely in individuals of African descent [24] . The latter variant, H132, is associated with decreased catalytic efficiency both in vitro and in vivo [12] .
10.2217/PGS.09.77 © 2009 Future Medicine Ltd
Pharmacogenomics (2009) 10(9), 1447–1455
Mao Mao1†, Alice Matimba2,3, Maria G Scordo1,4, Arzu Gunes1,5, Hakan Zengil5, Norio Yasui-Furukori6, Collen Masimirembwa3 & Marja-Liisa Dahl1 Author for correspondence: Department of Medical Sciences, Clinical Pharmacology, Uppsala University Hospital, Ent 61, SE-751 85 Uppsala, Sweden Tel.: +46 186 115 959; Fax: +46 186 113 703;
[email protected] 2 IIDMM, University of Cape Town, South Africa 3 African Institute of Biomedical Science and Technology, Harare, Zimbabwe 4 University of Messina, Messina, Italy 5 Gazi University, Ankara, Turkey 6 Hirosaki University, Hirosaki, Japan †
1
ISSN 1462-2416
1447
Research Article
Mao, Matimba, Scordo et al.
In light of the emerging knowledge on the functional impact of FMO3 polymorphisms, frequency data on SNPs and haplotypes are needed to understand their significance as a determinant of interindividual variability in pharmacokinetics and drug response of FMO3 substrates. As little is known regarding the frequency variation of FMO3 SNPs between populations within the same ethnicity, and data from native African populations is nonexistent, we aimed to explore the frequencies of the five FMO3 polymorphisms in multiple populations from Europe, East Asia and sub-Saharan Africa. Materials & methods
Study population A total of 13 defined groups from the three major ethnicities (Caucasian, Asian and African) with a total of 2152 unrelated subjects were enrolled in this study. The Caucasian population consisted of 410 subjects from Sweden, 279 from Italy and 300 from Turkey. The Asian population comprised 300 Japanese healthy volunteers. From the African continent, nine groups were included with a total of 863 subjects: 99 Luo, 143 Maasai and 97 Kikuyu from Kenya; 99 Shona and 63 San from Zimbabwe; 63 Venda from South Africa; and 100 Hausa, 100 Yoruba and 99 Ibo from Nigeria. All DNA samples were anonymized before analysis. The Swedish samples were provided by the local blood donor center. Subjects in the Italian, Turkish and Japanese groups were mainly students and employees from University of Messina, Italy, Gazi University, Turkey and Hirosaki University, Japan, respectively. The recruitment of the African population has been described in detail by Matimba and colleagues [25] . The ethnicity of the individuals was defined by self-reporting. The subjects had given written informed consent for genotype analysis. The study was approved by the local ethics committees and carried out in accordance with the Declaration of Helsinki. FMO3 genotyping With the exception of the Swedish subjects, from whom buffy coat was available, genomic DNA was purified from peripheral leukocytes and stored at -20°C. Standard extraction was performed using QIAamp DNA Blood Mini Kit (Qiagen, Hilden, Germany), according to the guidelines of the manufacturer. Genotyping was carried out by TaqMan® allelic discrimination with Fluorogenic 5´ nuclease 1448
Pharmacogenomics (2009) 10(9)
assays in an ABI Prism® 7000 Sequence Detection System (Applied Biosystems, CA, USA). The SNPs were analyzed by the validated TaqMan Genotyping Assays purchased from Applied Biosystems (for D132H, rs12072582, assay ID: C__30633935_10; for E158K, rs2266782, assay ID: C__2461179_30; for V257M, rs1736557, assay ID: C__8698544_30; for E308G, rs2266780, assay ID: C__2220257_30; for L360P, rs28363581, assay ID: C__30633936_20), according to the guidelines of the manufacturer.
Statistical analysis Data were assessed for consistency with Hardy–Weinberg equilibrium using exact tests (Haploview 4.0). Allele distributions were compared using two tailed c2 test or Fisher’s exact test (GraphPad Prism version 4, Graphpad Software, CA, USA). Haplotypes were inferred using the software package PHASE version 2.1. Evaluation of population differentiation was performed with the program ARLEQUIN version 3.1. Linkage disequilibrium analysis was quantified by D´ values calculated in Haploview 4.0. A p-value of less than 0.05 was considered statistically significant. Results
Allele frequencies The observed minor allele frequencies in the 13 populations studied are summarized in Table 1. No deviation from Hardy–Weinberg equilibrium was observed in any of the populations. The K158 variant was the most common variant allele in all populations. Among Caucasians, this allele was more frequent in Swedes (44.3%) than in either the Italian (33.5%, p A)
4.0 1.0 0.5 6.5 na 2.6 7 5
2.0 2.4 0.0
4.5 4.2 4.6
14.5 na 20.3 20
7.1 6.3 6.3 6.7 6.9 6.7 7 22 na
M257
1.3–6.7 0.0–2.4 0.0–1.5
0.0–4.0 0.0–5.1 –
1.6–7.4 1.9–6.5 1.7–7.5
11.7–17.3
5.3–8.8 4.3–8.3 4.4–8.2
95% CI
V257M (g.18281G>A)
0.5 0.5 0.5 5.2 4.5 5.6 4 0
1.5 1.6 0.8
0.0 0.0 1.0
21.0 18.3 14.8 14
22.4 10.8 6.0 21.6 22.5 15.7 16 22 20.2
G308
0.0–1.5 0.0–1.5 0.0–1.5
0.0–3.2 0.0–3.8 0.0–2.4
– – 0.0–2.4
17.7–24.3
19.6–25.3 8.2–13.4 4.1–7.9
95% CI
0.0 0.0 0.0 na na 1.9 na na
0.0 0.0 0.0
0.0 0.0 0.0
P360
– – –
– – –
– – –
95% CI
E308G (g.21443A>G) L360P (g.21599 T>C)
Table 1. Minor allele frequencies (%) of FMO3 SNPs from the present study and previously published studies.
FMO3 polymorphisms in 13 ethnic populations
Research Article
1449
Research Article
Mao, Matimba, Scordo et al.
significant subpopulation variation detected among Africans (c2 : 18.16; df: 8; p = 0.02) was again entirely due to the lower frequency of this allele in the San (0.0%) and the Ibo (0.5%) compared with the rest of the groups (1.0–4.6%). The frequency of the G308 allele also varied between populations. The frequency among Japanese (21.0%) was comparable to that among Swedes (22.4%), but significantly lower in both the Italian (10.8%, p G (E308G), four haplotypes being identified in all three major ethnicities and three being rare ones (Table 2) . The wildtype haplotype, G-G-A (E158–V257–E308), was highly prevalent (>45%) in all groups. The Japanese (1.8%) deviated from all the other groups (21.8–52.0%) regarding the frequency of the A-G-A (K158–V257–E308) haplotype. The inferred frequency of the A-G-G (K158–V257–G308) haplotype varied considerably across the ethnic groups, being lowest (G (E308G), and two African specific, g.15089 G>C (D132H), g.21599 T>C (L360P), FMO3 polymorphisms were determined in 13 groups representing Caucasians, Asians and Africans (Table 1) . In addition to confirming the known interethnic variation, novel findings of sub population and regional differences in allele and haplotype frequency within Caucasians and Africans were demonstrated, respectively. future science group
future science group
0.5
0.8 1.6 1.5 1.0
Base position in haplotypes: g.15167G>A (E158K)–g.18281G>A (V257M)–g.21443A>G (E308G). *
0.6 A-A-A A-A-G
0.2 20.8 6.0 10.8 22.4 G-G-G A-G-G
www.futuremedicine.com
Research Article
0.5
0.5
43.3 0.5
55.7 46.5
52.0 1.0 43.9 3.6
52.0 66.7
32.5 46.7 2.5
49.2 48.0
48.5 2.0 49.0 4.6
45.4 53.9
41.9 4.2 50.0 4.5
45.5 62.7
1.8 14.5 29.8 6.3
57.8
21.8 7.1 A-G-A G-A-A
22.1 5.6
48.7 G-G-A
60.9
Yoruba
West Africa
Hausa San Venda
Southern Africa
Shona Kikuyu Maasai
East Africa
Luo
East Asia: Japanese Southeast Europe: Turkish South Europe: Italian North Europe: Swedish Haplotype*
Table 2. Population frequencies (%) of inferred FMO3 haplotypes.
The finding of differences between European Caucasian populations (Swedish, Italian and Turkish) is further supported by comparisons of our data with those from other European groups, available in the literature (Table 1) . The reported allelic frequencies in German [26] and Irish [27] populations did not deviate from our Swedish group at any of the sites, but were significantly higher compared with both the Italian and Turkish groups for the K158 and G308 variants (p
