INTRODUCTION. Clinical Science (2005) 109, (Printed in Great Britain) 183

Transcription

1 Clinical Science (2005) 109, (Printed in Great Britain) 183 ATP-binding cassette transporter G8 M429V polymorphism as a novel genetic marker of higher cholesterol absorption in hypercholesterolaemic Japanese subjects Kenji MIWA, Akihiro INAZU, Junji KOBAYASHI, Toshinori HIGASHIKATA, Atsushi NOHARA, Masaaki KAWASHIRI, Shoji KATSUDA, Mutsuko TAKATA, Junji KOIZUMI and Hiroshi MABUCHI Molecular Genetics of Cardiovascular Disorders, Division of Cardiology, Graduate School of Medical Science, Kanazawa University, Takara-machi 13-1, Kanazawa , Japan, Department of Laboratory Sciences, School of Health Science, Faculty of Medicine, Kanazawa University, Takara-machi 13-1, Kanazawa , Japan, Department of Lifestyle-related Disease, Kanazawa University Graduate School of Medical Science, Takara-machi 13-1, Kanazawa Japan, and Department of General Medicine, Kanazawa University Hospital, Kanazawa University, Takara-machi 13-1, Kanazawa , Japan A B S T R A C T The ratio of serum plant sterols to cholesterol is positively correlated with the fractional cholesterol absorption, whereas serum precursors of cholesterol synthesis are positively correlated with cholesterol synthesis. Recently, two ABC (ATP-binding cassette) transporters, ABCG5 and ABCG8, have been described as playing an important role in the absorption and excretion of sterols. In the present study, we tested the hypothesis that genetic variation in ABCG5/ABCG8 influences the levels of serum plant sterol (sitosterol) and cholesterol precursor (lathosterol) in Japanese primary hypercholesterolaemic patients (n = 100). We identified a novel mutation [859T/C (C287R)] and a novel polymorphism [1285A/G (M429V)] at the ABCG5/ABCG8 loci, as well as four polymorphisms reported previously [1810C/G (Q604E), 161G/A (C54Y), 1199C/A (T400K) and 1895C/T (A632V)]. In carriers of the novel M429V variant, the serum level of sitosterol and the sitosterol/cholesterol ratio were significantly higher than those in non-carriers (3.64 compared with 2.56 µg/ml, and 1.45 µg/mg compared with 1.00 µg/mg respectively; P < 0.01 for both), and serum lathosterol tended to be lower (1.95 µg/ml compared with 3.03 µg/ml; P = 0.08), whereas no significant difference was observed in other lipid profiles. These four polymorphisms (1810C/G, 161G/A, 1199C/A and 1285A/G) generated six haplotypes, and the C/G/C/G haplotype was significantly associated with a higher sitosterol level and sitosterol/cholesterol ratio compared with the other five haplotypes (P < 0.05 for both). We conclude that, in 8 % of patients with hypercholesterolaemia, the novel ABCG8 M429V variant was associated with higher cholesterol absorption efficiency. Future studies should investigate whether these findings have implications for the optimal cholesterol-lowering drug treatment in hypercholesterolaemic patients. INTRODUCTION Whole-body cholesterol homoeostasis is strongly regulated by the balance of dietary cholesterol absorption, de novo synthesis, and biliary clearance and excretion. Serum plant sterols (campesterol, stigmasterol and sitosterol) come entirely from dietary intake (i.e. they are not synthesized endogenously), and the ratios of these Key words: ATP-binding cassette (ABC) transporter, cholesterol, genetic marker, polymorphism, sterol. Abbreviations: ABC, ATP-binding cassette; HDL-C, high-density lipoprotein-cholesterol; LDL-C, low-density lipoproteincholesterol; RFLP, restriction-fragment-length polymorphism; SSCP, single-strand conformational polymorphism. Correspondence: Dr Kenji Miwa ( miwakc2004@yahoo.co.jp).

2 184 K. Miwa and others Table 1 Method for RFLP screening of the ABCG5/ABCG8 gene Numbering from ATG in cdna. Lowercase letters indicate mismatch in primers to create a restriction site. Nucleotide Annealing Position Mutation change Forward primer (5 3 ) Reverse primer (5 3 ) temperature ( C) Enzyme Size (bp) ABCG5 Exon 7 C287R 859T C TCACACACTAACTACCTTCTGTTGTC ATGATGGGGAATGTGAAAGAAA 54 BstUI 191 Exon 13 Q604E 1810C G ATCTAGATTCACAATGAACTTTCTA GTCCCTGCAAGTTGTAAGAG 53 XhoI 193 ABCG8 Exon 2 C54Y 161G A GGAGGTCAGAGACCTCAAgT GCCCACCCTTTTATTTCCAC 56 RsaI 107 Exon 8 T400K 1199C A ACACCTGTGTGGAAAGGTAAGGT GCGGGTTCAGTAATAAAATGACAG 57 MseI 216 Exon 9 M429V 1285A G ATGCTGTTGCCTCAGCATCT AAGCTGTGTTCCTCTGAGCT 56 Tsp45I 306 Exon 13 A632V 1895C T ATGTCTGTGTCTCCAGATCCTCAGgG TACAGGACCATGAAGCCACCGCTGAcGCC 63 HaeIII 105 sterols to cholesterol are known to be positively related to cholesterol absorption and negatively to cholesterol synthesis [1 4]. On the other hand, serum also contains small amounts of precursors of cholesterol synthesis (squalene, desmosterol and lathosterol), and the ratios of these sterols to cholesterol are known to be positively related to cholesterol synthesis [1,2]. Recently, a critical component of the cellular machinery that limits accumulation of plant sterols was revealed by the molecular cloning of the underlying gene in patients with sitosterolaemia, a rare autosomal-recessive disorder of sterol metabolism [5 7]. Patients with sitosterolaemia are known to have increased intestinal absorption and markedly impaired biliary excretion of neutral sterols, which causes high plasma levels of cholesterol and plant sterols, leading to the development of xanthomas and premature coronary atherosclerosis [8]. Sitosterolaemia is reportedly caused by mutations in the two genes that encode the ABC (ATP-binding cassette) halftransporters ABCG5 and ABCG8 [5 7]. Plasma sitosterol concentrations vary over a 5- to 10-fold range among individuals consuming similar amounts of dietary sitosterol, and are weakly correlated with dietary plant sterol intake [9]. These findings suggest that interindividual variation in the plasma concentrations of plant sterols reflects interindividual differences in sterol metabolism. Indeed, it was previously proposed that common sequence variants in ABCG5 (Q604E) and ABCG8 (D19H and T400K) are associated with plasma plant sterol and lipid levels in normocholesterolaemic or mildly hypercholesterolaemic European American populations [10 12]. In this background, the aim of our present study was to screen the ABCG5 and ABCG8 genes for polymorphisms and to evaluate their contribution to the levels of cholesterol precursor (lathosterol) and plant sterol (sitosterol) variability in a sample of 100 Japanese patients with primary hypercholesterolaemia. MATERIALS AND METHODS Study subjects The study population consisted of 100 Japanese primary hypercholesterolaemic patients (48 men and 52 women) aged (means + S.D.) years (range, years), who attended the Department of Internal Medicine, Kanazawa University Hospital, Kanazawa, Japan. Hypercholesterolaemia was defined as present if the total cholesterol was 220 mg/dl or higher under dietary recommendation of NCEP/AHA (National Cholesterol Education Program/American Heart Association) Step I diet. None of the subjects had diabetes, hepatic disease, kidney disease, endocrinological disorders or malignant disease, or was receiving lipid-lowering therapy. Patients with familial hypercholesterolaemia were also excluded [13]. Blood samples were drawn after an overnight fast. Serum cholesterol and triacylglycerol (triglyceride) levels were determined enzymatically. Serum HDL-C (high-density lipoprotein-cholesterol) was determined by a direct method [14], and serum sitosterol and lathosterol were measured by GLC [15]. Informed consent was obtained from every participant, and the Ethical Committee of Kanazawa University Graduate School of Medical Science approved this study. Assay of DNA sequence variants in ABCG5 and ABCG8 genes PCR and SSCP (single-strand conformational polymorphism) analysis were performed to screen the DNA sequence variants in all 100 subjects. Genomic DNA was prepared from white blood cells by phenol/chloroform extraction. The ABCG5 and ABCG8 genes were then divided into 19 and 16 overlapping fragments of approx. 200 bp respectively, and subjected to PCR using specific primers under the following conditions: 94 C for 4 min; 35 cycles of 96 C for 30 s, the annealing temperature for 1 min, and 72 Cfor2min;and72 C for 10 min.

3 ABCG8 polymorphisms and cholesterol absorption in hypercholesterolaemic subjects 185 Specific primers were determined as described previously [16], and using Primer 3 software ( edu/cgi-bin/primer3/primer3 SSCP analysis was performed as described previously [16a]. Electrophoresis was performed on a % gradient polyacrylamide gel (ATTO) and run at a constant 100 V for 12 h in 0.5 Tris/borate/EDTA buffer under two conditions: (i) at 4 C without glycerol, and (ii) at room temperature with 5 % (w/v) glycerol. Fragments with a variant as detected by SSCP were re-amplified and sequenced with fluorescently labelled dideoxy terminators using the Thermo Sequenase Kit (Amersham Biosciences) on a genetic analyser (ABI 310; Applied Biosystems). After the mutation was characterized by direct sequencing, a mismatch technique allowing the analysis of RFLPs (restriction-fragment-length polymorphisms) was performed in all subjects to confirm the sequence variation (Table 1). Statistical analysis Statistical analysis was performed using Stat View 5.0 software (SAS Institute). ANOVA was used to evaluate the phenotypic characteristics among genotype groups. Contingency tables were used with standard χ 2 tests to compare observed genotype frequencies with those expected under Hardy Weinberg equilibrium. The haplotype frequencies were estimated and compared in the non-cholesterol levels using the expectation maximization algorithm for estimation of maximum likelihood haplotype frequencies from multilocus genotypic data without a known gametic phase using the Arlequin2000 software package ( The haplotypic effects on serum non-cholesterol concentrations were assigned in all subjects using PHASE software ( html) [17]. RESULTS The serum levels of total cholesterol, sitosterol and lathosterol are shown in Table 2, along with the clinical characteristics of the subjects. Genotype and allele frequencies A novel mutation of C287R and five SNPs (single nucleotide polymorphisms) (including a novel one ABCG8 M429V) were genotyped in our subjects. Frequency distributions for the genotypes in exon 7 (C287R) and exon 13 (Q604E) of ABCG5, and in exon 2 (C54Y), exon 8 (T400K), exon 9 (M429V) and exon 13 (A632V) of ABCG8 are shown in Table 3. All genotype frequency distributions were in Hardy Weinberg equilibrium. The allele frequencies differed somewhat from those described previously in European American populations [6,10]. For instance, the ABCG8 D19H variant was not Table 2 Clinical characteristics of the study subjects Values are means + S.D. BMI, body mass index. Parameters Value Age (years) Gender (men/women) 48/52 BMI (kg/m 2 ) Total cholesterol (mg/dl) Triacylglycerol (mg/dl) HDL-C (mg/dl) LDL-C (mg/dl) Sitosterol (µg/ml) Lathosterol (µg/ml) Table 3 Genotype distribution and allele frequencies of the polymorphisms in the ABCG5/ABCG8 gene Gene Mutation Nucleotide change Polymorphism Allele frequency ABCG5 Exon 7 C287R 859T C T 0.99 C 0.01 Exon 13 Q604E 1810C G C 0.89 G 0.11 ABCG8 Exon 2 C54Y 161G A G 0.82 A 0.18 Exon 8 T400K 1199C A C 0.88 A 0.12 Exon 9 M429V 1285A G A 0.96 G 0.04 Exon 13 A632V 1895C T C T found and the A632V variant was rare in our Japanese subjects. Effects of polymorphisms in the ABCG5/ABCG8 gene on serum non-cholesterol sterol concentrations We compared serum non-cholesterol sterols (sitosterol and lathosterol), both in absolute value (µg/ml) and in cholesterol-standardized ratio (µg/mg of cholesterol), among subjects with each genotype. Of the six mutation/ polymorphisms, one novel mutation (C287R) and one reported variant (A632V) were excluded from the study because of their rarity (allele frequency 0.01). Table 4 shows the serum non-cholesterol sterol concentration and its ratio to cholesterol for the four ABCG5/ABCG8 polymorphisms. The novel M429V variant of ABCG8 was found to be positively associated with both higher sitosterol concentrations and their ratiotocholesterol(p < 0.01 for both). The serum lathosterol concentrations tended to be lower in carriers of this

4 186 K. Miwa and others Table 4 Association of the ABCG5/ABCG8 gene polymorphisms with serum non-cholesterol sterol levels Values are means + S.D. chol, cholesterol. Sitosterol Lathosterol Sitosterol/chol Polymorphism Genotype n (µg/ml) P value (µg/ml) P value (µg/mg) P value Lathosterol/chol (µg/mg) P value ABCG5 Exon 13 Q604E (1810C G) QQ QE EE ABCG8 Exon 2 C54Y (161G A) CC CY YY ABCG8 Exon 8 T400K (1199C A) TT TK KK 0 ABCG8 Exon 9 M429V (1285A G) MM MV VV 0 Table 5 Effect of the four polymorphism haplotypes in the ABCG5/ABCG8 gene on serum non-cholesterol levels Values are means + S.D. The haplotype effects on serum non-cholesterol levels were assigned in all individuals using PHASE in 94 individuals. The values shown were calculated with the weighted average. P < 0.05, comparison of sitosterol/lathosterol levels between H1 and any other haplotype. chol, cholesterol; G5, ABCG5; G8, ABCG8. Polymorphism Haplotype G G8 161 G G n Sitosterol (µg/ml) Lathosterol (µg/ml) Sitosterol/chol (µg/mg) Lathosterol/chol (µg/mg) H1 C G C A H2 C A A A H3 G G C A H4 C A C A H5 C G C G H6 G A A A variant. None of the other polymorphisms were significantly associated with serum non-cholesterol sterol concentrations. In the two rare cases with the C287R mutation, their serum sitosterol and sitosterol/cholesterol levels were not significantly elevated (2.6 µg/ml and 3.1 µg/ml, and 0.97 µg/mg and 1.48 µg/mg respectively). Likewise, in a case with the A632V variant, serum sitosterol and sitosterol/cholesterol levels were not elevated (1.2 µg/ml and 0.6 µg/mg respectively). Haplotypes were assigned to all of the 100 subjects using the PHASE reconstruction method. However, six of the 100 subjects were excluded from the association analysis because of uncertain gametic phases. Of the 16 possible four-polymorphism haplotypes [1810C/G (Q604E), 161G/A (C54Y), 1199C/A (T400K) and 1285A/G (M429V)], six haplotypes were estimated to be present. The greatest effect on higher sitosterol concentrations and their ratio to cholesterol was the C/G/C/G haplotype compared with the main haplotype (Table 5; P < 0.05). The ABCG5 and ABCG8 polymorphisms were not significantly related to serum lipid concentrations [total cholesterol, LDL-C (low-density lipoprotein-cholesterol), HDL-C and triacylglycerols] or body mass index (results not shown). DISCUSSION The main findings of the present study performed in Japanese hypercholesterolaemic subjects are as follows: (i) two novel mutants or polymorphisms, C287R in ABCG5 and M429V in ABCG8, have been identified, and (ii) the M429V polymorphism is positively associated with serum sitosterol and sitosterol/cholesterol levels, whereas the other three polymorphisms are not

5 ABCG8 polymorphisms and cholesterol absorption in hypercholesterolaemic subjects 187 associated with serum non-cholesterol levels. Cholesterol absorption marker and serum plant sterol concentrations vary widely between individuals [18,19], but are rather stable over time in a given individual [10,20]. This suggests a strong effect of genetic background on plant sterol metabolism. Consistent with this, we found that genetic variation in ABCG8 explained part of this variability. Among the four polymorphisms examined, only the M429V variant was significantly associated with sitosterol concentration and its ratio to cholesterol. This result does not appear to be consistent with previous reports that the sequence variants in ABCG5 (Q604E) and ABCG8 (D19H and T400K) are associated with lower serum plant sterol levels in Western countries [10 12]. The results of our present study suggest that the frequencies of D19H and A632V variants of ABCG8 are rarer in Japanese than in European American populations, and that these Q604E and T400K variants may not be as important in the regulation of non-cholesterol sterol levels in Japanese populations. The absence of a significant association between the four polymorphisms examined in the present study and serum lipid levels suggests that common variants in ABCG5 and ABCG8 are not important determinants of the serum lipoprotein profiles in moderately hypercholesterolaemic patients. This agrees with observations in chow-fed ABCG5/ABCG8 knockout mice. In these animals, it has been reported that disruption of the ABCG5 and ABCG8 genes increases the fractional absorption of dietary plant sterols, but not of cholesterol [21 23]. As a consequence, serum and hepatic plant sterol concentrations are reported to be dramatically increased in ABCG5 knockout mice [21] and in Abcg5/ Abcg8 knockout mice [22,23], whereas serum cholesterol concentrations are unchanged. Moreover, overexpression of ABCG5/ABCG8 genes in the mouse is reported to have markedly different effects on plant sterols and cholesterol metabolism [24]. For instance, serum plant sterol concentrations were significantly lower in transgenic mice than in wild-type mice, whereas serum cholesterol concentrations were again not different. Therefore the findings in the mouse study suggest that ABCG5 and ABCG8 affect serum plant sterol concentrations, but not necessarily those of cholesterol. Genotype frequency in Japanese was different from that in Caucasians. This may be partly explained by ethnic difference and cholesterol levels in the population surveyed. Indeed, subjects in most of the previous studies in Caucasians are normolipidaemic, whereas in one report subjects were hypercholesterolaemic [25]. However, we found no difference in genotype frequency between hypercholesterolaemic subjects and normocholesterolaemic controls in our present study. Again, the polymorphisms do not account for the difference in plasma cholesterol levels. Serum lipid levels are strongly regulated by the balance of de novo synthesis, dietary cholesterol absorption, and biliary clearance and excretion. Consistent with this, statin treatment of patients with coronary artery disease, having low baseline ratios of cholestanol and plant sterol to cholesterol, was shown to produce greater reductions in cholesterol and reduced recurrences of coronary events in the Scandinavian Simvastatin Survival Study [26]. A recent study has shown that there is a statistically significant association between the ABCG8 D19H polymorphism and the proportional reduction in LDL- C during atorvastatin treatment [25]. Gylling et al. [11] showed that the higher efficacy of atorvastatin treatment in the DH/HH subjects, as found in an earlier study [25], might be related to higher endogenous cholesterol synthesis and lower cholesterol absorption in these subjects than in DD wild-types. In the present study, the novel M429V variant may have similar therapeutic implications of statins. Finally, we did not find a specific ABCG5/ABCG8 haplotype that was more significantly associated with non-cholesterol sterol concentrations than the M429V variant alone. This strongly suggests that the novel M429V variant itself, or another as yet undefined linked variant, has functional significance. In terms of putative topology, M429V is predicted to be located in the first transmembrane domain of an N-terminal site. From this point of view, the M429V variant is the marker rather than the cause of higher serum sitosterol concentration. In conclusion, in 8 % of Japanese patients with primary hypercholesterolaemia, the novel ABCG8 M429V variant is associated with higher serum sitosterol concentrations (probably due to higher cholesterol absorption efficiency), whereas no relationships with serum lipid concentrations are observed under dietary restriction. Future studies should investigate whether these findings have implications for the optimal cholesterol-lowering diet or drug treatment in hypercholesterolaemic patients. ACKNOWLEDGMENTS We thank Sachio Yamamoto, Mihoko Mizuno and Mayumi Yoshida for their excellent technical assistance. REFERENCES 1 Miettinen, T. A. (1970) Detection of changes in human cholesterol metabolism. Ann. Clin. Res. 2, Miettinen, T. A., Tilvis, R. S. and Kesaniemi, Y. A. (1990) Serum plant sterols and cholesterol precursors reflect cholesterol absorption and synthesis in volunteers of a randomly selected male population. Am. J. Epidemiol. 131, Salen, G., Ahrens, Jr, E. H. and Grundy, S. M. (1970) Metabolism of sitosterol in man. J. Clin. Invest. 49, Tilvis, R. S. and Miettinen, T. A. (1986) Serum plant sterols and their relation to cholesterol absorption. Am. J. Clin. Nutr. 43, Berge, K. E., Tian, H., Graf, G. A. et al. (2000) Accumulation of dietary cholesterol in sitosterolemia caused by mutations in adjacent ABC transporters. Science 290,

6 188 K. Miwa and others 6 Hubacek, J. A., Berge, K. E., Cohen, J. C. et al. (2001) Mutations in ATP-cassette binding proteins G5 (ABCG5) and G8 (ABCG8) causing sitosterolemia. Hum. Mutat. 18, Lee, M. H., Lu, K., Hazard, S. et al. (2001) Identification of a gene, ABCG5, important in the regulation of dietary cholesterol absorption. Nat. Genet. 27, Bjorkhem, I. J., Boberg, K. M. and Leitersdorf, E. (2001) Inborn errors in bile acid biosynthesis and storage of sterols other than cholesterol. In The Metabolic and Molecular Basis of Inherited Disease (Scriver, C. R., Beaudet, A. L., Sly, W. S. and Valle, D., eds.), pp , McGraw-Hill, New York 9 Kempen, H. J., de Knijff, P., Boomsma, D. I. et al. (1991) Plasma levels of lathosterol and phytosterols in relation to age, sex, anthropometric parameters, plasma lipids, and apolipoprotein E phenotype, in 160 Dutch families. Metab, Clin. Exp. 40, Berge, K. E., von Bergmann, K., Lutiohann, D. et al. (2002) Heritability of plasma non-cholesterol sterols and relationship to DNA sequence polymorphism in ABCG5 and ABCG8. J. Lipid Res. 43, Gylling, H., Hallikainen, M., Pihlajamäki., J. et al. (2004) Polymorphisms in the ABCG5 and ABCG8 genes associate with cholesterol absorption and insulin sensitivity. J. Lipid Res. 45, Plat, J., Bragt, M. C. and Mensink, R. P. (2005) Common sequence variations in ABCG8 are related to plant sterol metabolism in healthy volunteers. J. Lipid Res. 46, Yu, W., Nohara, A., Higashikata, T. et al. (2002) Molecular genetic analysis of familial hypercholesterolemia: spectrum and regional difference of LDL receptor gene mutations in Japanese population. Atherosclerosis 165, Warnick, G. R., Nauck, M. and Rifai, N. (2001) Evolution of methods for measurement of HDL-cholesterol: from ultracentrifugation to homogeneous assays. Clin. Chem. 47, Nikkila, K., Hockerstedt, K. and Miettinen, T. A. (1991) High cholestanol and low campesterol-to-sitosterol ratio in serum of patients with primary biliary cirrhosis before liver transplantation. Hepatology 13, Heimer, S., Langmann, T., Moehle, C. et al. (2002) Mutations in the human ATP-binding cassette transporters ABCG5 and ABCG8 in sitosterolemia. Hum. Mutat. 20, a Spinardi, L., Mazars, R. and Theillet, C. (1991) Protocols for an improved detection of point mutations by SSCP. Nucleic Acids Res. 19, Heinemyer, T., Chen, X., Karas, H. et al. (1999) Expanding the TRANSFAC database towards an expert system of regulatory molecular mechanisms. Nucleic. Acids Res. 27, Glueck, C. J., Speirs, J., Tracy, T. et al. (1991) Relationship of serum plant sterols (phytosterols) and cholesterol in 595 hypercholesterolemic subjects, and family aggregation of phytosterols, cholesterol, and premature coronary heart disease in hypercholesterolemic probands and their first degree relatives. Metab., Clin. Exp. 40, Kempen, H. J., Glatz, J. F., Gevers Leuven, J. A. et al. (1988) Serum lathosterol concentration is an indicator of whole-body cholesterol synthesis in humans. J. Lipid Res. 29, Li, J. H., Awad, A. B., Fink, C. S. et al. (2001) Measurement variability of plasma β-sitosterol and campesterol, two new biomarkers for cancer prevention. Eur. J. Cancer. Prev. 10, Plosch, T., Bloks, V. W., Terasawa, Y. et al. (2004) Sitosterolemia in ABC-transporter G5 deficient mice is aggravated on activation of the liver-x receptor. Gastroenterology 126, Yu, L., Hammer, R. E., Li-Hawkins, J. et al. (2002) Disruption of Abcg5 and Abcg8 in mice reveals their crucial role in biliary cholesterol secretion. Proc. Natl. Acad. Sci. U.S.A. 99, Yu, L., York, J., Hobbs, von Bergmann, K. et al. (2003) Stimulation of cholesterol excretion by the liver X receptor agonist requires ATP-binding Cassette transporters G5 and G8. J. Biol. Chem. 278, Yu, L., Li-Hawkins, J., Hammer, R. E. et al. (2002) Overexpression of ABCG5 and ABCG8 promotes biliary cholesterol secretion and reduces fractional absorption of dietary cholesterol. J. Clin. Invest. 110, Kajinami, K., Brousseau, M. E., Nartsupha, C. et al. (2004) ATP binding cassette transporter G5 and G8 genotypes and plasma lipoprotein levels before and after treatment with atorvastatin. J. Lipid Res. 45, Miettinen, T. A., Gylling, H., Strandberg, T. et al. (1998) Baseline serum cholestanol as predictor of recurrent coronary events in subgroup of Scandinavian simvastatin survival study. Finnish 4S Investigators. Br. Med. J. 316, Received 20 January 2005/5 April 2005; accepted 7 April 2005 Published as Immediate Publication 7 April 2005, DOI /CS