Genetics is a major determinant of expression of the human hepatic uptake transporter OATP1B1, but not of OATP1B3 and OATP2B1

Abstract

Background: Organic anion transporting polypeptide (OATP) 1B1, OATP1B3, and OATP2B1 (encoded by SLCO1B1, SLCO1B3, SLCO2B1) mediate the hepatic uptake of endogenous compounds like bile acids and of drugs, for example, the lipid-lowering atorvastatin, thereby influencing hepatobiliary elimination. Here we systematically elucidated the contribution of SLCO variants on expression of the three hepatic OATPs under consideration of additional important covariates.

Methods: Expression was quantified by RT-PCR and immunoblotting in 143 Caucasian liver samples. A total of 109 rare and common variants in the SLCO1B3-SLCO1B1 genomic region and the SLCO2B1 gene were genotyped by MALDI-TOF mass spectrometry and genome-wide SNP microarray technology. SLCO1B1 haplotypes affecting hepatic OATP1B1 expression were associated with pharmacokinetic data of the OATP1B1 substrate atorvastatin (n = 82).

Results: Expression of OATP1B1, OATP1B3, and OATP2B1 at the mRNA and protein levels showed marked interindividual variability. All three OATPs were expressed in a coordinated fashion. By a multivariate regression analysis adjusted for non-genetic and transcription covariates, increased OATP1B1 expression was associated with the coding SLCO1B1 variant c.388A > G (rs2306283) even after correction for multiple testing (P = 0.00034). This held true for haplotypes harboring c.388A > G but not the functional variant c.521T > C (rs4149056) associated with statin-related myopathy. c.388A > G also significantly affected atorvastatin pharmacokinetics. SLCO variants and non-genetic and regulatory covariates together accounted for 59% of variability of OATP1B1 expression.

Conclusions: Our results show that expression of OATP1B1, but not of OATP1B3 and OATP2B1, is significantly affected by genetic variants. The SLCO1B1 variant c.388A > G is the major determinant with additional consequences on atorvastatin plasma levels.

Figure 1

Hepatic expression of OATP1B1, OATP1B3 and OATP2B1 varies between individuals, but is coordinately regulated within one individual. (a) Representative immunoblot of liver membrane fractions (20 μg) quantified relative to a standard liver sample. (b-d) Correlation analyses of OATP protein and SLCO mRNA levels in human liver samples. (e, f) Correlation analyses of all three OATPs showed a coordinated expression of SLCO1B1, SLCO1B3, and SLCO2B1 mRNA and OATP1B1, OATP1B3, and OATP2B1 protein levels in the non-cholestatic liver sample set. r_S, Spearman rank-order correlation coefficient.

Figure 2

SLCO genetic variants affect expression of hepatic OATPs. (a, b) Pairwise linkage disequilibrium map of the SLCO1B3-SLCO1B1 genomic region (a) and the SLCO2B1 gene (b) including all variants detected in the 143 livers. Coloring corresponds to the standard Haploview (D'/LOD) [31]. (c,d) SLCO1B1 haplotypes, calculated based on the four missense variants present in the liver cohort (c.388A > G, c.463C > A, c.521T > C, c.1929A > C) and also previously described as key variants [8,80], affect OATP1B1 protein expression in the non-cholestatic liver samples (n = 117). Only haplotypes with frequencies ≥ 2% are given. The effect sizes indicate differences of OATP1B1 expression compared to reference haplotype SLCO1B1*1a. Boldface: significant P-values. (d) SLCO1B1 genotypes (percentage genotype frequencies in brackets) are sorted by median OATP1B1 protein expression. SLCO1B1 alleles harboring only variant c.388A > G and not c.521T > C (*1b, *14, *35, boldface) confer considerably higher OATP1B1 expression in a gene-dose dependent manner. Horizontal line indicates median; boxes indicate the 25th to 75th percentiles; whiskers indicate non-outlier range. (e) Percentage of interindividual variability of expression of OATPs in the non-cholestatic liver samples apportioned to non-genetic factors (gray), regulatory factors (blue), and SLCO variants (orange), and to the combination of all three categories (red) was calculated using multivariate linear regression analyses and stepwise model selection.

Figure 3

SLCO1B1 haplotypes affect atorvastatin pharmacokinetics in 82 healthy volunteers. (a) SLCO1B1 haplotypes calculated on the presence of the four previously described key variants and change in atorvastatin area under the plasma concentration-time curve (AUC), corrected for body weight, in comparison with the reference haplotype SLCO1B1*1a. Haplotypes are given with a frequency > 2%. (b) SLCO1B1 genotypes (percentage genotype frequencies in brackets) are sorted by median atorvastatin AUC. SLCO1B1 alleles harboring only variant c.388A > G and not c.521T > C (boldface) confer considerably lower AUCs with lowest levels in homozygous variant or compound heterozygous carriers for the SLCO1B1*14 or *1b alleles. Horizontal line indicates the median; boxes indicate 25th to 75th percentiles; whiskers indicate non-outlier range.