|Home | About | Journals | Submit | Contact Us | Français|
hMATE1 (human multidrug and toxin compound extrusion-1; encoded by SLC47A1) is thought to have an important function in the renal and hepatic elimination of drugs, endogenous compounds and environmental toxins. The goals of this study were to identify genetic variants of hMATE1 and to determine their effects on hMATE1 transport function. We identified four synonymous and six nonsynonymous, coding region variants in DNA samples from 272 individuals (68 Caucasians, 68 African Americans, 68 Asian Americans and 68 Mexican Americans). The overall prevalence of hMATE1 nonsynonymous variants was relatively low with three singleton variants and three variants having allele frequencies ≥2% in a specific ethnic group. The nonsynonymous hMATE1 variants were constructed and stably transfected into HEK-293 cells. Uptake studies using four known hMATE1 substrates (paraquat, metformin, tetraethylammonium and oxaliplatin) were performed in cells transfected with hMATE1 reference or variants. We found that two singleton variants, G64D and V480M, produced a complete loss of function for all four tested substrates whereas three polymorphic variants (allele frequencies ≥2%), L125F, V338I and C497S, significantly altered the transport function in a substrate-dependent manner. Confocal microscopy studies were consistent with functional studies suggesting that the altered function of the variants was due to altered localization to the plasma membrane. These data suggest that nonsynonymous variants in hMATE1 may alter drug disposition and ultimately affect clinical drug response.
hMATE1 (human multidrug and toxin compound extrusion-1), encoded by SLC47A1, is a newly identified organic cation/H+ exchanger expressed in the apical membrane of renal tubule cells and the canalicular membrane of hepatocytes.1 hMATE1 has been shown to mediate the cellular uptake of structurally diverse small-molecular-weight organic cations such as tetraethylammonium (TEA), N-methylpyridinium, metformin,2 paraquat3 and oxaliplatin.4 hMATE2-K (human multidrug and toxin compound extrusion-2K, encoded by SLC47A2), the paralog of hMATE1, is specifically expressed in the kidney and interacts with various organic cations with a substrate specificity overlapping that of hMATE1.2, 5 The location in the kidney and the overlapping specificities suggest that hMATE1 and hMATE2-K may function somewhat redundantly in the kidney by mediating tubular transport of cationic organic molecules across the apical membrane. hMATE genes are located in tandem on chromosome 17, a region commonly deleted in Smith-Magenis syndrome, a genetic disorder with multiple congenital anomalies and mild mental retardation.6, 7
Recent studies have linked genetic variants in membrane transporters in the solute carrier superfamily (SLC) with altered drug disposition and response.8, 9, 10 In particular, such studies have taken a genotype to phenotype approach, focusing on variants that exhibited altered function in cellular assays. For example, Shu et al.8, 9 demonstrated that genetic variants in the hepatic influx transporter, organic cation transporter 1, OCT1, which exhibited altered function in cellular assays, affected the disposition and pharmacologic effects of the anti-diabetic drug, metformin, in healthy volunteers. Similarly, Urban et al.10 showed that a common variant of the novel organic cation transporter, OCTN1, which exhibited reduced uptake of the neuroleptic drug, gabapentin, in cellular assays was associated with a reduced renal clearance of gabapentin in healthy volunteers. These studies provide strong evidence that genetic variants in uptake transporters in the liver and kidney, which exhibit functional changes in cellular assays, may have important clinical effects and support a genotype to phenotype strategy in discovering clinically important variants.
The goal of these preclinical studies was to identify genetic variants in the coding region of hMATE1 and to determine whether these variants exhibit functional alterations in cellular assays. For variant discovery, we used DNA samples from a large ethnically diverse population. The function of nonsynonymous coding region variants of hMATE1 was characterized using a mammalian cellular expression system. In particular, we studied the interaction of four model substrates with the variant hMATE1 proteins.
To identify hMATE1 variants, we screened all 17 exons as well as 50–100 bp of flanking intronic sequence in genomic DNA samples from 272 ethnically diverse individuals. We identified four synonymous and six nonsynonymous, coding region variants (544 chromosomes) (Table 1 and Figure 1). Four variants were in dbSNP (Build 128) and the rs numbers are shown in Table 1. The position and population frequency of each variant will be deposited in dbSNP as well as PharmGKB (www.pharmgkb.org). For six nonsynonymous variants, three variants (G64D, V480M and Q519H) were singletons, identified on only one of the 544 chromosomes sequenced. Three variants (L125F, V338I and C497S) were found to be polymorphic, having allele frequencies >1% in a specific ethnic group. Notably, one variant (L125F) had an allele frequency of 5% in Mexican Americans and one variant (V338I) occurred at an allele frequency of 5% in African Americans. Overall, the genetic variation in the hMATE1 gene in the four ethnic groups is low. Three out of six nonsynonymous variants (G64D, L125F and Q519H) occur in evolutionary conserved amino acid residues identical among human, mouse and rat orthologs of hMATE1. The characteristics of the coding region variants and their population-specific allele frequencies are shown in Table 1. The population genetics statistics for the coding region of hMATE1 and four other representative membrane transporters in the SLC family are summarized in Table 2. The nucleotide diversity at synonymous sites (πS) for the SLC transporters was higher than that at nonsynonymous sites (πNS) consistent with purifying selection. The ratio of πNS/πS is frequently used as a measure of the extent of selective pressure on a gene, with low πNS/πS corresponding to a high degree of negative selection. The ratio of 0.21 for hMATE1 was close to the median value of that obtained for 11 SLC transporters surveyed by our group,11 suggesting a relative neutral evolutionary tolerance for alterations in hMATE1 protein structure.
We examined whether nonsynonymous genetic variants of hMATE1 would affect its function using model substrates. Six nonsynonymous hMATE1 variants identified in the population samples were constructed and stably transfected into human embryonic kidney HEK-293 cells. Uptake studies using four known hMATE1 substrates (paraquat, metformin, TEA and oxaliplatin) were performed in cells stably transfected with plasmids containing either hMATE1 reference or its variants. As shown in Figure 2a, the uptake of paraquat in HEK-hMATE1 reference cells was 14+4-fold greater than that in HEK-MOCK cells (P<0.01 by one-way analysis of variance followed by Dunnett’s two-tailed test). The uptake of paraquat was dramatically diminished in cells expressing the singleton variant transporters, hMATE1-G64D and hMATE1-V480M. These two variants resulted in complete loss of function. One variant L125F had significantly reduced transport of paraquat compared to the hMATE1 reference (P<0.01). To investigate the mechanism for the reduced function of this polymorphism, kinetic studies were performed in cells containing hMATE1 reference or the common variant transporter, hMATE1-L125F using paraquat as a substrate. As shown in Figure 3a and Table 3, for paraquat uptake, the Km and Vmax values for variant hMATE1-L125F were not significantly different from the reference hMATE1 (P=0.620 for Vmax, P=0.587 for Km). In addition, the ratios, Vmax/Km, for paraquat were not significantly different between the reference hMATE1 and hMATE1-L125F (P=0.804).
The uptake of the anti-diabetic drug, metformin, was significantly reduced in cells expressing variant transporters, hMATE1-G64D, hMATE1-L125F, hMATE1-V338I, hMATE1-V480M and hMATE1-C497S (P<0.01) (Figure 2b). Similarly, these same transporter variants except hMATE1-C497S had reduced function for the model quaternary ammonium compound, TEA, compared to reference (P<0.01) (Figure 2c). For oxaliplatin uptake, the variants hMATE1-G64D and hMATE1-V480M showed significantly reduced function whereas variant hMATE1-C497S had increased function (P<0.01) (Figure 2d).
To investigate the mechanism for the reduced function produced by the two polymorphisms of hMATE1, kinetic studies of metformin uptake were performed in cells containing hMATE1 reference or the more common variants (hMATE1-L125F or hMATE1-V338I). As shown in Figure 3b and Table 3, for metformin uptake, the Km and Vmax values for variant hMATE1-L125F and hMATE1-V338I were not significantly different from the reference hMATE1 (P=0.635 for Vmax, P=0.526 for Km). However, the ratios, Vmax/Km, for metformin were 3.07±0.35, 1.91±0.20 and 2.40±0.21 for the reference, variant hMATE1-L125F and variant hMATE1-V338I, respectively. The Vmax/Km ratios tended to be different among these variants (P=0.051), with reduced ratio in hMATE1-L125F.
The IC50 values of oxaliplatin in HEK-MOCK cells were 2.3- to 3.5-fold higher than those in HEK cells expressing hMATE1 reference and four of the hMATE1 variants (Table 4). These results indicated that hMATE1 and four of its coding region variants enhanced the cytotoxicity of oxaliplatin. Consistent with the uptake results, oxaliplatin did not exhibit enhanced cytotoxicity in cell lines expressing two of the reduced function hMATE1 variant transporters, hMATE1-G64D and hMATE1-V480M, in comparison to cell lines transfected with empty vector (Table 4).
We examined the cellular localization of two reduced function variants (L125F and V338I), two complete loss of function variants (G64D and V480M) (Figure 2) and reference hMATE1 to understand if the variants exhibited altered distribution to the plasma membrane. We were particularly interested in hMATE1-L125F because this variant was found in two populations, Mexican and Asian Americans, and had an allele frequency in the Mexican sample of 5%. Further, the variant exhibited a large difference in Vmax/Km ratio for metformin in comparison to the reference hMATE1. We performed uptake experiments using metformin as a substrate in cells transiently transfected with plasmids containing either hMATE1 reference or its variants. We found that the uptake of metformin was dramatically diminished in cells transiently expressing hMATE1-G64D and hMATE1-V480M variants. These two variants resulted in complete loss of function. The uptake of metformin was significantly reduced in cells transiently expressing hMATE1-L125F and hMATE1-V338I variants compared to the hMATE1 reference (data not shown). These data were consistent with those performed in cells stably transfected with hMATE1 reference or its variants. Next, HEK-293 cells were transiently transfected with green fluorescent protein GFP-labeled hMATE1 reference, hMATE1-G64D, hMATE1-L125F, hMATE1-V338I or hMATE1-V480M. Representative confocal images are shown in Figures 4a e. hMATE1 reference was mainly located on the plasma membrane (Figure 4a) whereas the reduced function variants hMATE1-L125F and hMATE1-V338I (Figures 4c and d) were located on both plasma membrane and in the intracellular spaces. The nonfunctional variants hMATE1-G64D and hMATE1-V480M appeared to have less protein expression on the plasma membrane compared with hMATE1 reference; instead, the nonfunctional variants were primarily localized intracellularly (Figures 4b and e).
hMATE1 was recently identified as an important transporter in the kidney and liver for structurally diverse cationic xenobiotics including therapeutic drugs and chemical toxins.1, 2, 3 It is abundantly expressed in the adrenal gland, suggesting that it also functions in the disposition of endogenous substrates that may be synthesized and secreted by the gland. Thus, genetic variation in hMATE1 could potentially contribute to variation in drug disposition and response as well as to physiologic and pathophysiologic variation among people. This study represents the first comprehensive screen of the coding region of hMATE1 for genetic and functional variation in an ethnically diverse sample population. We identified six nonsynonymous single nucleotide polymorphisms (SNPs) in hMATE1. Among these, three variants were singletons and three variants had allele frequency of >1% in the total population. Two common variants had allele frequencies of 5% in African American or Mexican American samples. No nonsynonymous variants were identified in samples from individuals with European ancestry. The overall nonsynonymous variation of hMATE1 is low compared to OCT1 and OCTN1 (see Table 2), but comparable to OCT2, another transporter in the kidney in the SLC superfamily, which is believed to work in concert with MATE1 in renal secretion of xenobiotics.12, 13 These data are consistent with purifying selection of MATE1 and are consistent with an important physiologic function for the transporter.
For our functional studies, we selected four model compounds: paraquat, metformin, TEA and the anticancer drug, oxaliplatin. TEA represents a classical model organic cation used to study the function of renal organic cation transporters. Paraquat is a toxic herbicide that has been banned from many countries. Nephrotoxicity, hepatoxicity and pulmonary toxicity, which occur after accidental or suicidal ingestion, are the major toxicities of this herbicide. hMATE1 is expressed in abundance in lung, kidney and liver; therefore, genetic variants in MATE1 may have a function in risk for toxicity after ingestion. Metformin is an important anti-diabetic drug, which is cleared almost exclusively in the kidney. hMATE1 appears to be an important protein in its secretion from tubule cell to tubule lumen. Genetic variation in hMATE1 may contribute to variation in renal clearance of metformin. Finally, oxaliplatin is an anticancer drug transported by hMATE1, which could potentially affect the tissue distribution and tubular secretion of this agent. Genetic variants in hMATE1 may have a function in anticancer effects and/or adverse effects of oxaliplatin.
Our cellular assays indicated that the six nonsynonymous variants were functionally heterogeneous. Such functional heterogeneity has been reported previously for SLC22A transporters such as OAT3 and OCT1.14, 15 Two of the hMATE1 variants (G64D and V480M) produced a complete loss of function for all the substrates tested and appeared as singletons only in our Asian population sample. From the sequence alignment of hMATE1 with its species orthologs, the G64 residue occurs in an evolutionarily conserved amino acid residue identical among human, mouse and rat orthologs of hMATE1. A change in an evolutionarily conserved amino acid residue would be expected to cause a functional change. Further, substitution of a chemically different amino acid at a functionally critical position could also cause a loss of function. For example, Otsuka et al.1 showed that amino acid replacement from glutamate to glutamine at position 273 (E273Q), the substitution of an essential amino acid residue found in the bacterial NorM protein, led to complete loss of function for TEA uptake. G64D, a substitution of a negatively charged aspartic acid for a neutral amino acid, glycine, in the first extracellular loop of MATE1, represents a large chemical change, which could have contributed to the loss of function. Furthermore, we found that hMATE1-G64D tagged with GFP is expressed mainly in the intracellular space and less on the plasma membrane (Figure 4b) compared to hMATE1 reference, which may explain its functional loss.
Because V480M is not an evolutionarily conserved residue and the substitution of methionine for valine is a conservative chemical change, the loss of function in this singleton variant was not expected. In further studies with a GFP chimera, we observed that the substitution appeared to change the intracellular localization of the transporter. That is, we found that hMATE1-V480M tagged with GFP was poorly localized to the plasma membrane (Figure 4), which rendered it functionally inert. Though further studies are needed, it is possible that valine at position 480 is critical for trafficking of hMATE1 to the plasma membrane.
We also identified two variants (L125F and V338I) that significantly reduced the transport function in a substrate-dependent manner. hMATE1-L125F had a reduced ability to transport all three model substrates, paraquat, metformin and TEA, but maintained the transport ability for oxaliplatin. hMATE1-V338I had reduced function with respect to metformin and TEA but not with respect to paraquat and oxaliplatin. Similarly, hMATE1-C497S exhibited a decreased transport of metformin and TEA; however it seemed to have an enhanced ability to transport paraquat and oxaliplatin. Chemically, paraquat has two positive charges and oxaliplatin after aquation will form monoaqua and diaqua complexes (two positive charges), whereas metformin and TEA carry a single positive charge. Thus, it is possible that different amino acid residues affect the substrate recognition and translocation of divalent organic cations by hMATE1 in comparison to monovalent cations.
Previous studies have demonstrated that hMATE1 expression enhances the cytotoxicity of oxaliplatin.16 In this study we examined the effect of genetic variants of hMATE1 on the cytotoxicity of oxaliplatin, and observed a 2.3- to 3.5-fold enhanced cytotoxicity in cell lines expressing the reference hMATE1 and four of the variant transporters. In contrast, cell lines expressing the hMATE1 variants, hMATE1-G64D and hMATE1-V480M, did not exhibit enhanced cytotoxicity or uptake of oxaliplatin. Because both of these variants are rare, their effects at the population level would be minimal; however, the studies suggest that rare variants of hMATE1 in the coding region could result in a modulation of the disposition and pharmacologic action of oxaliplatin.
The kinetics studies performed on the relatively common variants hMATE1-L125F and hMATE1-V338I did not clearly explain why these variants had reduced transport function for paraquat. The Km, Vmax and Vmax/Km values for hMATE1-L125F were not significantly different from the reference hMATE1 (see Table 3). However, because paraquat transport was only slightly reduced in the hMATE1-L125F variant (see Figure 2a), small differences in Km and Vmax values would be difficult to detect in kinetic studies. For metformin uptake, in which larger differences in uptake between the hMATE1 reference and the two polymorphic variants (hMATE1-L125F and hMATE1-V338I) were observed, the Vmax/Km ratios exhibited trends toward significantly lower values in cells expressing the variant transporters in comparison to the reference transporter (P=0.051). Confocal microscopy showed that hMATE1-L125F and hMATE1-V338I variants were located on both plasma membrane and in the intracellular spaces, in contrast to the reference hMATE1, which was predominately distributed to the plasma membrane (Figure 4). These data are consistent with the reduced function of hMATE1-L125F and hMATE1-V338I in comparison to the reference transporter.
In summary, we identified two rare variants of hMATE1 with no function and four variants with altered function. Three of the four variants were polymorphic in a particular ethnic population with allele frequencies equal to or greater than 2%. Because hMATE1 is expressed on the apical membrane of renal tubule cells and the canalicular membrane of hepatocytes, these genetic variants may contribute to variation in renal or biliary elimination of hMATE1 substrates. Clinical studies are clearly needed to determine whether these findings have implications to drug disposition and response.
Genomic DNA samples were collected from healthy volunteers in four major ethnic groups (68 each from Caucasians, African Americans, Mexican Americans and Asian Americans) as part of the Studies of Pharmacogenetics in Ethnically Diverse Populations (SOPHIE) project. Variants of hMATE1 were identified by direct sequencing of individual DNA samples as described previously.17
hMATE1 cDNA (GenBank accession number NM_018242) was amplified by reverse transcriptase PCR from human kidney RNA and cloned into pcDNA5-FRT (Invitrogen, Carlsbad, CA, USA).3 Six nonsynonymous hMATE1 variants identified in the population samples were constructed by site-directed mutagenesis using high-fidelity DNA polymerase PfuTurbo (Stratagene, La Jolla, CA, USA) according to manufacturer’s protocol. The variants were fully sequenced to verify that they had only the desired mutation.
HEK-293-Flp-In cells (Invitrogen) were transfected with pcDNA5/FRT empty vector, pcDNA5/FRT vector containing the full-length hMATE1 or each of the hMATE1 variants using Lipofectamine 2000 (Invitrogen). Forty-eight hours after transfection, 75 μg ml−1 hygromycin B was added to the medium, and stable clones were selected. All cell lines were grown at 37 °C in a humidified atmosphere with 5% CO2.
HEK-293 stably transfected cells were grown in monolayer in 24-well plate. When cells reach more than 90% confluence (generally within 24 h), they were incubated in buffer containing 125 mM NaCl, 4.8 mM KCl, 5.6 mM D-glucose, 1.2 mM CaCl2, 1.2 mM KH2PO4, 1.2 mM MgSO4 and 25 mM Tricine (pH 8.0). Transport assays were initiated by adding radiolabeled compounds including radiolabeled metformin, paraquat and TEA. The uptake was stopped after the indicated time by washing cells three times with ice-cold choline buffer containing 128 mM choline, 4.73 mM KCl, 1.25 mM CaCl2, 1.25 mM MgSO4 and 5 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid/Tris (pH 7.4). Cells were lysed and incubated with 0.1 N NaOH/0.1% SDS for 1 h. Radioactivity in aliquots of cell lysates was determined by scintillation counting. Protein was measured by bicinchoninic acid (BCA) protein assay (Pierce, Rockford, IL, USA). For kinetic studies, varying amounts of unlabeled compounds were added to the uptake solutions to give increasing total (radiolabeled plus unlabeled) substrate concentrations. Nonspecific cell-associated radioactivity was determined by measuring substrate uptake in empty vector transfected cells (MOCK) at each substrate concentration, and these values were then subtracted from the results in hMATE1 or hMATE1 variant-transfected cells to give the final data used for kinetic analysis. The Km and Vmax values were obtained by fitting the Michaelis-Menten equation V=Vmax[S]/(Km+[S]) using GraphPad Prism 4.0 (GraphPad Software, San Diego, CA, USA). V refers to the rate of substrate transport, Vmax the maximum rate of substrate transport, [S] the concentration of substrate and Km the concentration of substrate at the half-maximal transport rate.
HEK-293 stably transfected cells were incubated with 5 μM oxaliplatin at 37 °C in 5% CO2 for 2 h. Cells were then washed with ice-cold phosphate-buffered saline three times and harvested by centrifuge at 400 g for 10 min at 4 °C. The cell pellets were dissolved in 70% nitric acid at 65 °C for at least 2.5 h. Distilled water containing 10 p.p.b. of iridium (Sigma, Ronkonkoma, NY, USA) and 0.1% Triton X-100 was added to the samples and the final concentration of nitric acid was 7%. The platinum content was measured by inductively coupled plasma mass spectrometry in the Analytical Facility at the University of California at Santa Cruz. Cell lysates from a set of identical cultures were used for BCA protein assay.
The cytotoxicity of oxaliplatin was measured by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay in 96-well plates. HEK-293 stably transfected cells expressing either empty vector or hMATE1 reference or variants were seeded (7000 cells per well) the day before treatment with different concentrations of oxaliplatin. Cells were exposed to oxaliplatin for 6 h. The drug-containing medium was replaced with fresh medium and the cells were incubated for another 66 h. MTT assays were performed, and the IC50 values were obtained as described previously.3, 18
hMATE1 cDNA was first cloned into pEGFP-C1 vector (Clontech, Palo Alto, CA, USA). pEGFP gene was then subcloned into pcDNA5/FRT vector to make the GFP-tagged clone. The hMATE1-G64D, hMATE1-L125F, hMATE1-V338I and hMATE1-V480M variant transporters were constructed by site-directed mutagenesis. Cells transfected with GFP-containing vectors were seeded at 3 × 105 cells per well on 12-mm poly-D-lysine-coated glass coverslips (BD Biosciences—Discovery Labware, San Jose, CA, USA) in a 12-well plate. Cells were stained using the Image-IT labeling kit according to manufacturer’s protocol. Coverslips were mounted in VECTASHIELD antifade solution (Vector laboratories, Burlingame, CA, USA) on glass microscope slides and visualized by confocal microscopy using a Zeiss 510 laser scanning microscopes.
Data are expressed as mean±s.d. Unpaired Student’s t-test was used to analyse differences between two groups. Multiple comparisons were analysed using one-way analysis of variance followed by Dunnett’s two-tailed test. The data were analysed with GraphPad Prism 4.0 (GraphPad Software). A P-value less than 0.05 was considered statistically significant.
This work was supported by National Institutes of Health Grants GM61390 and GM74929, and a grant from NEDO, Tokyo. Y.C. was supported by National Research Service Award T32 GM07546 from the National Institutes of Health. We acknowledge James E. Shima for his helpful advice on this manuscript. We also thank Alexandra G. Ianculescu for her technical support on GFP images.
Duality of Interest