Pharmacogenomics of drug-metabolizing enzymes and drug transporters : studies on montelukast, fluvastatin, telmisartan, and celiprolol

Substantial interindividual variability exists in drug response. This variability often means differences in drug concentrations in the body between individuals receiving the same dose. Insufficient drug concentrations can cause poor efficacy whereas too high concentrations can predispose to adverse drug reactions. Genetic variants in pharmacokinetic genes, which encode drug-metabolizing enzymes and drug transporters, can affect drug concentrations, and thus, drug pharmacokinetics in the body. Knowledge of the effects of genetic variants in these genes can enable dose adjustments and more safe and efficient drug treatments. This doctoral thesis investigated the associations of genetic variants in pharmacokinetic genes with the pharmacokinetics of four drugs: montelukast, fluvastatin, telmisartan, and celiprolol. For this purpose, full pharmacokinetic profiles of the drugs were determined in 188-200 healthy volunteers after a single dose of each drug and 379 pharmacokinetic genes were fully sequenced using targeted massive parallel sequencing. Statistical analyses were performed using stepwise linear regression analysis with log-additive modelling. The analyses were done with all 379 genes applying Bonferroni-correction and with selected candidate genes without correction for multiple testing. The linear regression models of the candidate gene analyses were used to generate multigenic genotype scores, which predict the fold differences in areas under the plasma concentration-time curves (AUC0-∞) between carriers of different genetic variants and non-carriers. As a mechanistic validation, in vitro experiments were performed to further confirm the role of the identified metabolizing enzymes on the metabolism of montelukast and fluvastatin enantiomers, and the role of the identified transporter on the transport of fluvastatin enantiomers. Furthermore, the thesis investigated the associations of genetic variants, which were found to be associated with montelukast pharmacokinetics, with messenger ribonucleic acid (mRNA) expression of the respective genes in human liver samples. The increased-function UGT1A3*2 allele was associated with significantly decreased plasma concentrations of montelukast and with increased glucuronidation. The AUC0-∞ of montelukast was 18% smaller per copy of the UGT1A3*2 allele (P = 2.75 × 10-10). UGT1A3 variants have not previously been associated with montelukast pharmacokinetics. The gene encodes the uridine diphospho glucuronosyltransferase (UGT) 1A3 drug-metabolizing enzyme. In addition to UGT1A3, single nucleotide variations (SNV) in CYP2C8 (encoding cytochrome P450 (CYP) 2C8 enzyme) and ABCC9 (encoding sulfonylurea receptor 2) genes were associated with montelukast pharmacokinetics in the candidate gene analysis. In vitro, montelukast was highly metabolized by UGT1A3. Furthermore, variants in ABCC9 and UGT1A3 were associated with mRNA expression in human liver samples. The no function CYP2C9*3 allele was associated with significantly increased plasma concentrations of fluvastatin enantiomers. The AUC0-∞ of active 3R,5S-fluvastatin was increased by 67% (P = 3.77 × 10-9) and that of inactive 3S,5R-fluvastain by 94% (P = 3.19 × 10-12) per copy of the allele. CYP2C9 encodes the CYP2C9 drug-metabolizing enzyme. The no function c.521T>C SNV in SLCO1B1 gene, which encodes organic anion transporting polypeptide (OATP) 1B1 transporter, was associated with increased AUC0-∞ of 3R,5S-fluvastatin but not with that of 3S,5R-fluvastatin. Thus, the effect of this genetic variant appears to be enantiospecific although both enantiomers were transported similarly by OATP1B1 in vitro. On the other hand, SLCO2B1 transporter gene (encoding OATP2B1) SNVs were associated only with 3S,5R-fluvastatin AUC0-∞ in the candidate gene analysis. The decreased-function CYP2C9*2 allele was associated with increased AUC0-∞ of both fluvastatin enantiomers in the candidate gene analysis. The increased-function UGT1A3*2 allele was associated with significantly decreased plasma concentrations of telmisartan and increased plasma concentrations of telmisartan acyl-glucuronide. In addition, the telmisartan plasma concentrations were lower in men than in women. The AUC0-∞ of telmisartan in men heterozygous and homozygous for UGT1A3*2 was decreased by 64% (P = 1.21 × 10-16) and 63% (P = 5.21 × 10-8), respectively. In heterozygous and homozygous women, the decreases were 57% (P = 1.54 × 10-11) and 72% (P = 3.31 × 10-15), respectively. An SLCO1B3 SNV was associated with increased telmisartan AUC0-∞ in the candidate gene analysis. SLCO1B3 encodes the OATP1B3 transporter. No variants were associated with celiprolol pharmacokinetics in the analysis with the 379 pharmacokinetic genes. In the candidate gene analysis, variants in transporter genes SLCO1A2 and ABCB1, which encode OATP1A2 and P-glycoprotein, respectively, were associated with reduced celiprolol AUC0-∞. This thesis showed that genetic variants explain a significant proportion of interindividual variability in the pharmacokinetics of montelukast, fluvastatin, telmisartan, and celiprolol. The multigenic genotype scores combined the effects of the significantly associated variants. In the case of montelukast, telmisartan, and celiprolol, individuals with low genotype score values may have an increased risk of poor efficacy. In the case of fluvastatin, individuals with high genotype score for the active 3R,5S-fluvastatin may have an increased risk of adverse reactions. Furthermore, the studies found associations of genetic variants that have not previously been associated with drug pharmacokinetics.