MDR1 genotypes do not influence the absorption of a single oral dose of 1 mg digoxin in healthy white males

Abstract

Aims A noncoding single nucleotide polymorphism (SNP) in exon 26 3435C > T of the highly polymorphic MDR1 gene has been demonstrated to alter digoxin absorption after induction of the MDR1 gene product P-glycoprotein by rifampicin or after multiple oral dosing. The aim of the study was to investigate the effects of the major known MDR1 SNPs on the absorption of digoxin after a single oral dose in a large sample without drug pretreatment.

Methods Fifty healthy white male subjects between the age of 18 and 40 years were enrolled. Following an overnight fast, all subjects received a single oral dose of 1 mg digoxin. Venous blood samples were taken at intervals up to 4 h post dose to obtain a pharmacokinetic profile.

Results AUC(0,4 h), C_max and t_max, used as indices of digoxin absorption, were not significantly different in any of the genotype groups tested. In particular, there was no significant difference between homozygous carriers of the C and T allele in exon 26 3435 (AUC(0,4 h) 9.24 and 9.38 µg l⁻¹ h, C_max 4.73 and 3.81 µg l⁻¹, t_max 0,83 and 01.14 h).

Conclusions This lack of effect of the major MDR1 SNPs on digoxin absorption might be explained by saturation of the maximum transport capacity of intestinal Pgp at the dose used.

Introduction

The intestinal epithelium is not only involved in the absorption of nutrients, water, xenobiotics and drugs, but also has barrier and excretory functions. The latter are mainly provided by P-glycoprotein (Pgp), the product of the MDR1 gene, which is a major determinant of drug export from enterocytes into the gut lumen. Pgp is a member of the ATP-binding cassette (ABC) superfamily of membrane transporters [1] and was originally discovered through its ability to confer resistance to antineoplastic agents in tumour cells [2]. A broad range of diverse, structurally unrelated compounds, including the cardiac glycoside digoxin [3], anthracycline antibiotics, vinca alkaloids, epipodophyllotoxins [4, 5], daunomycin, and the immunosuppressant cyclosporin A [6] are transported from cells by Pgp in an ATP-dependent manner. Pgp is located as a 170 kDa integral membrane protein in the apical pole of luminal epithelial cells of the stomach, the small intestine, the colon [7], the biliary canaliculi of the liver, and the brush border of the renal proximal tubules [8]. The expression of Pgp in these tissues is related to its role in the excretion of substances into the gut, the bile, and the urine. Furthermore Pgp can be found in the endothelium of capillary blood vessels in brain, testis, and other blood–tissue barrier sites [9–11]. The extensive tissue distribution and the wide variety of therapeutically relevant compounds that are transported by Pgp, clearly indicate its important role in drug absorption, distribution, and elimination. Therefore, alterations of the transport function or the degree of expression of Pgp are likely to affect the pharmacokinetics of many drugs.

Recently a number of single nucleotide polymorphisms (SNPs) of the MDR1 gene have been identified [12–14]. Most of the SNPs are intronic or silent and therefore do not change the amino acid composition of Pgp. However, the silent polymorphism in exon 26 3435C > T was found to correlate significantly with the amount of intestinal Pgp protein and the extent of absorption of digoxin in a small number of subjects (n = 8) pretreated with rifampicin or given multiple doses (0.25 mg day⁻¹) of digoxin [12]. However, it is unclear whether this pharmacogenetic effect requires Pgp induction or steady state conditions. Therefore, in the present study we investigated the effects of the major MDR1 SNPs, including that in exon 26 3435C > T on the absorption of a single 1 mg oral dose digoxin in a large sample of 50 subjects without pretreatment with ­rifampicin.

Methods

Subjects

A total of 50 healthy unrelated Caucasian male volunteers (nonsmokers) with a median age of 26 years were enrolled from the Berlin area. They were considered to be healthy as determined by their medical histories, physical examination, electrocardiogram, urine analysis, and routine tests of biochemistry, haematology, hepatitis B and C, and HIV. Their weight ranged from 45.1 kg to 74.6 kg. All volunteers refrained from alcohol, coffee, tea or cola beverages consumption and did not take medications during the study. All subjects gave their written informed consent before entry into the study. The investigation was approved by the local ethics committee of the Charité Medical Center, Humboldt University of Berlin. MDR1 genotype was determined using polymerase chain reaction (PCR)–based restriction fragment length polymorphism (RFLP) assays [13].

Study protocol

After an overnight fast, each subject received a single oral dose of 1 mg digoxin (DilanacinTM, Arzneimittelwerk Dresden GmbH, Radebeul) along with 200 ml water. Venous blood samples (5 ml) for drug analysis were collected into syringes containing ethylene diaminetetraacetic acid (EDTA) before and 10, 20, 30, 35, 40, 45, 50, 60, 75, 90, 120, 180, and 240 min after digoxin intake. The plasma was separated by centrifugation and immediately stored in polypropylene tubes at −22 °C.

Digoxin concentration measurements

Digoxin plasma concentrations were determined by a microparticle enzyme immunoassay (IMx®Digoxin Assay, Abbott Laboratories; USA). The lower limit of quantification was 0.3 µg l⁻¹. The plasma samples of each volunteer were analyzed in duplicate together with calibration and quality control samples. Further calculations were done with the respective mean values. The inter­assay coefficients of variation at plasma concentrations of 0.9 µg l⁻¹, 1.9 µg l⁻¹ and 3.2 µg l⁻¹ were 8.37%, 5.75% and 4.68%, respectively.

Pharmacokinetic analysis

Previous investigations demonstrated that pharmaco­kinetic differences between the genotypic groups occur primarily during the initial hours of absorption and are best reflected by area under the plasma concentration time curves (AUCs) covering the early hours after oral drug intake [15–17]. The t_max for digoxin is known to be a variable parameter [18] ranging from 0.5 to 3 h in our sample. Thus, an AUC from 0 to 4 h AUC(0,4 h) was considered a suitable measure of digoxin absorption. AUC(0,4 h) was calculated by use of the trapezoidal rule, using WinNonlinTM (professional edition; version 1.5, Pharsight Corporation, Mountain View, CA, USA). C_max and t_max of digoxin were derived directly from the measured values.

Statistical analysis

The primary aim of this exploratory study was to evaluate the functional relevance of major known MDR1 SNPs on digoxin absorption represented by AUC(0,4 h). For a two groups comparison the Mann–Whitney-U two sample test was used. When more than two groups were being compared the Kruskal–Wallis test was used. A P value of <0.05 was considered significant. Calculations were done using SPSSTM software (version 9.0, SPSS Inc., Chicago, USA).

Results

To study the influence of MDR1 variant genes on digoxin pharmacokinetics we included SNPs occur­-ring at a frequency of at least 10% in the Caucasian ­population (exon 6 + 139C > T, exon 17–76T > A, 3435C > T), SNPs changing the amino acid composition (61 A > G, 1199G > A, 2677G > T > A), and a SNP that directly precedes the translation start codon of the MDR1 gene (exon 2–1G > A). The allelic frequency distribution of the investigated genetic variants in our sample (Table 1) is in good agreement with previous studies [12, 13] and did not show significant deviation from Hardy–Weinberg equilibrium.

Table 1.

Location, position, effects of protein expression, allelic frequencies and genotype prevalence of the MDR1 variants studied.

Location	Position	Allele	Effect	Allelic frequency (%)	Allelic frequency (14) (%)	Genotype	Genotype prevalence (observed) (%)	Genotype prevalence (95% confidence interval)
Intron 1	Exon 2–1	G	initiation of translation?	93.0	91.0	G/G	86.0	73.3, 94.2
		A		7.0	9.0	G/AA/A	14.00.0	5.8, 26.70.0, 0.1
Exon 2	cDNA 61	A	21Asn	87.0	88.8	A/A	74.0	59.7, 85.4
		G	21Asp	13.0	11.2	A/GG/G	26.00.0	14.6, 40.30.0, 0.1
Intron 6	Exon 6 + 139	C	?	61.0	62.8	C/C	36.0	22.9, 50.8
		T		39.0	37.2	C/TT/T	50.014.0	35.5, 64.75.8, 26.7
Exon 11	cDNA 1199	G	400Ser	98.0	94.5	G/G	96.0	86.3, 99.5
		A	400Asn	2.0	5.5	G/AA/A	4.00.0	0.5, 13.70.0, 0.1
Intron 16	Exon 17–76	T	?	53.0	53.8	T/T	28.0	16.2, 42.5
		A		47.0	46.2	T/AA/A	50.022.0	35.5, 64.511.5, 35.7
Exon 21	cDNA 2677	G	893Ala	52.0	56.5	G/G	24.0	13.1, 38.2
		T	893Ser	38.0	41.6	G/T	42.0	28.2, 57.0
		A	893Thr	10.0	1.9	T/T	14.0	5.8, 26.7
						G/A	14.0	5.8, 26.7
						T/A	6.0	1.3, 16.6
						A/A	0.0	0.0, 0.1
Exon 26	cDNA 3435	C	Wobble	49.0	46.1	C/C	24.0	13.1, 38.2
		T		51.0	53.9	C/TT/T	50.026.0	35.5, 64.514.6, 40.3
						T/T	26.0	14.6, 40.3

The positons of the MDR1 polymorphisms correspond to the MDR1 cDNA (GenBank accession no. AC002457/AC005068). The first base of the ATG start codon is set to no. 1. Intronic SNPs are described as (exon ± n), with n bases upstream (–) or downstream (+) of the exons. Eight polymorphisms were analysed.

After administration of a single oral dose of 1 mg digoxin the pharmacokinetics of the absorptive phase characterized by AUC(0,4 h), C_max and t_max was not ­significantly different between the genotypic groups. Comparison of the linked MDR1 polymorphisms in exon 21 2677 and exon 26 3435 also did not discriminate between subjects carrying the genotype exon 21 2677GG/exon 26 3435CC and exon 21 2677TT/exon 26 3435TT. Additional homozygous allelic combinations were not investigated due to low frequencies. Pharmacokinetic data and statistical analysis are summarized in Table 2. There were also no significant differences after normalization to ideal body weight to eliminate interindividual differences in weight and height of the subjects. There was considerable variation in the kinetic parameters within each genotypic group (Figure 1).

Table 2.

Pharmacokinetic parameters and statistical analysis.

Location	Position	Genotype		AUC(0,4 h) (µg l −1  h)			Cmax (µg l−1)			tmax(h)
			Pvalue	Mean	Difference between mean of SNP and wild type groups (95% CI)	Pvalue	Mean	Difference between mean of SNP and wild type groups (95%CI)	Pvalue	Mean	Difference between mean of SNP and wild type groups (95% CI)
Intron 1	Exon 2–1	G/G	0.426	9.17		0.780	4.00		0.232	1.02
		G/A		9.73	0.563 (−0.81, 1.94)		3.98	0.00 (−1.01, 0.97)		0.80	−0.22 (−0.59, 0.15)
Exon 2	cDNA 61	A/A	0.091	9.51		0.382	4.11		0.815	0.96
		A/G		8.51	−1.00 (−2.05, 0.06)		3.68	−0.43 (−1.20, 0.34)		1.09	0.13 (−0.17, 0.42)
Intron 6	Exon 6 + 139	C/C	0.785	9.22		1.000	4.10		0.907	0.96
		C/T		9.17	0.05 (−1.05, 1.15)		3.90	0.19 (−0.58, 0.97)		1.00	−0.04 (−0.33, 0.25)
		T/T		9.60	−0.38 (−1.84, 1.08)		4.08	0.01 (−1.21, 1.24)		1.06	−0.10 (−0.46, 0.25)
Exon 11	cDNA 1199	G/G	–	9.21		–	4.03		–	0.98
		G/A		10.21	–		3.26	–		1.18	–
Intron 16	Exon 17–76	T/T	0.718	9.01		0.294	3.88		0.228	1.07
		T/A		9.46	−0.45 (−1.58, 0.68)		4.21	−0.33 (−1.14, 0.47)		0.92	0,15 (−0.17, 0.47)
		A/A		9.07	−0.05 (−1.39, 1.28)		3.67	0.21 (−0.76, 1.17)		1.05	0.03 (−0.33, 0.38)
Exon 21	cDNA 2677	G/G	0.875	9.06		0.648	4.22		0.736	1.00
		G/T		9.10	−0.03 (−1.28, 1.21)		3.91	0.30 (−0.57, 1.17)		1.00	−0.06 (−0.36, 0.35)
		T/T		8.95	0.11 (−1.51, 1.74)		3.69	0.53 (−0.77, 1.83)		1.10	−0.10 (−0.50, 0.30)
		G/A		9.91	−0.84 (−2.65, 0.96)		4.00	0.21 (−1.30, 1.73)		0.85	0.15 (−0.20, 0.50)
		T/A		10.50	–		4.39	–		0.96	–
Exon 26	cDNA 3435	C/C	0.295	9.72		0.799	4.12		0.871	0.92
		C/T		8.89	0.83 (−0.39, 2.05)		3.87	0.26 (−0.64, 1.15)		1.05	−0.12 (−0.46, 0.22)
		T/T		9.51	0.20 (−1.10, 1.51)		4.14	−0.02 (−0.98, 0.94)		0.95	−0.03 (−0.30, 0.24)
Exon 21/	cDNA	2677 3435
Exon 26	2677/	G/G C/C	0.393	9.06		0.712	3.84		0.478	0.90
	3435	T/T T/T		9.06	0.00 (−1.78, 1.78)		3.69	0.15 (−1.07, 1.36)		1.11	−0.21 (−0.68, 0.26)

Pharmacokinetic parameters were compared between the wildtype groups and those carrying a SNP using the nonparametric Mann–Whitney-U two sample test or Kruskal–Wallis test. A P value of < 0.05 was considered significant. P values and 95% CI were not given for subject groups with a lower number than 4.

Figure 1.

AUC(0,4 h) data for digoxin after a 1 mg oral dose in relation to the MDR1 polymorphisms. n = number of subjects in each group.

Discussion

Very recently a number of MDR1 SNPs were detected and tested preliminarily for their ability to alter the pharmacokinetics of orally administered digoxin [12]. The main findings were that a polymorphism in exon 26 (3435C > T), although noncoding, resulted in a more than 2-fold decrease in duodenal epithelial Pgp protein content in homozygous T/T subjects compared with carriers of the C/C variant. This was accompanied by a significantly higher digoxin bioavailability in the T/T genotype. The finding was supported by a lowered efflux of rhodamine-123 from CD56⁺ natural killer cells, mediated by Pgp, in 3435TT compared with CC carriers [19]. These studies were limited because of the small number of subjects investigated. Accordingly the aim of the present study was to extend the sample number to seek for additional pharmacokinetic effects of the major MDR1 SNPs and to evaluate the impact of the known functional active exon 26 3435C>T polymorphism.

In contrast to the previously reported results [12] no significant differences in AUC(0,4 h), C_max and t_max were observed between MDR1 genotype combinations in this study.

A major finding of one of the previous studies was that the observed difference in the bioavailability of digoxin between genotypes with respect to exon 26 3435C > T SNP was observed only after pretreatment with rifampicin [12]. This antibiotic is reported to be a strong inducer of intestinal Pgp expression [16], increasing protein content 3.5 fold above control levels. The ability of rifampicin to increase epithelial Pgp content was significantly higher in individuals carrying the C/C genotype in MDR1 exon 26 3435C > T than in subjects homozygous for the T allele. Since digoxin is transported back into the intestinal lumen mainly by Pgp, induction of this efflux system may be an important prerequisite for the MDR1 SNP 3435C > T dependent modulation of digoxin bioavailability. This could also explain why there was a lack of association between the 3435C > T polymorphism and Pgp expression in 100 placentas obtained from Japanese women not treated with inducing agents [14].

The transport capacity of Pgp is a function of the amount of carrier protein expressed in epithelial cells. Digoxin is preferentially absorbed in the duodenum. If the local digoxin concentration exceeds the maximal Pgp secretory capacity, further digoxin absorption is merely dependent on passive diffusion or uptake carriers. Members of the organic anion-transporting polypeptide family in humans (OATP) and rats (Oatp) have been identified as candidate uptake transporters for amphipathic drugs, including digoxin [20, 21]. Localization of OATPs/Oatps to the basolateral pole of hepatocytes facing portal venous blood and to the apical membrane of enterocytes [22], suggests an involvement of this uptake system in digoxin absorption. Thus, the counteracting effects passive diffusion and OATPs could dramatically diminish the modulatory actions of MDR1 SNPs on digoxin pharmacokinetics after exceeding a ceiling dose. The latter could be increased by protein induction of the Pgp carrier system and would be dependent on the drug, as the maximal transport capacity V_max of a carrier is substrate specific. The oral dose of 1 mg digoxin used in our study might be higher than that needed to saturate the transport capacity of Pgp. A recent study investigating the effects of the noncoding SNP in exon 26 3435C > T on digoxin pharmacokinetics after a single low dose of 0.25 mg resulted in decreased digoxin plasma concentration in homozygous subjects of the T allele compared with carriers of the CC genotype [23].

In the present study AUC(0,4 h), C_max and t_max were determined after a single oral dose of digoxin. This experimental design was chosen to assess primarily the impact of MDR1 SNPs on the absorption of digoxin. Since drugs absorbed into the enterocytes are pumped back into the intestinal lumen by Pgp, one would expect the major effect of the latter to occur during the first hours after oral administration. However, in one of the previous studies, digoxin pharmacokinetics were monitored under steady state conditions [12]. In this case the observed differences between the 3435C > T genotypes may be due partly to genetic effects on digoxin distribution and elimination.

In our study there was a large variation in digoxin pharmacokinetics within each genotypic group. This high variability indicates that additional factors may influence the absorption of digoxin. These may include other as yet undetected SNPs or dietary and other environmental factors, that induce or inhibit the Pgp activity [5, 24, 25].

In summary, we conclude that the MDR1 SNPs studied, including that in exon 26 previously shown to have functional relevance, apparently do not contribute to the absorptive pharmacokinetics of a single oral dose of 1 mg digoxin. Our study suggests that other genetic and environmental factors may play an important role in MDR1 SNP mediated modulation of digoxin pharmacokinetics.

This work was supported by the German Federal Ministry of Education and Research, (grant ‘Pharmacogenetic Diagnostics’ no. 01 GG 9845/5 and grant no. 031U209B ‘Berlin Center for Genome Based Bioinformatics’).