Abstract
The equilibrative nucleoside transporter 2 (ENT2; SLC29A2) is a bidirectional transporter that is involved in the disposition of naturally occurring nucleosides as well as a variety of anticancer and antiviral nucleoside analogs. The goal of the current study was to evaluate the function of genetic variants in ENT2 in cellular assays and to determine the haplotype structure of the coding and flanking intronic region of the gene. As part of a large study focused on genetic variation in membrane transporters (Leabman et al., 2003), DNA samples from ethnically diverse populations (100 African-Americans, 100 European-Americans, 30 Asians, 10 Mexicans, and 7 Pacific Islanders) were screened for variants in membrane transporters, including SLC29A2. Fourteen polymorphic sites in SLC29A2 were found, including 11 in the coding region. Five protein-altering variants were identified: three nonsynonymous variants, and two deletions. Each of the protein-altering variants was found at a very low frequency, occurring only once in the sample population. The nonsynonymous variants and the deletions were constructed via site-directed mutagenesis and were subsequently characterized in Xenopus laevis oocytes. All variants were able to take up inosine with the exception of ENT2-Δ845-846, which resulted in a frameshift mutation that prematurely truncated the protein. ENT2 showed very infrequent variation compared with most other transporter proteins studied, and it was found that five haplotypes were sufficient to describe the entire sample set. The low overall genetic diversity in SLC29A2 makes it unlikely that variation in the coding region contributes significantly to clinically observed differences in drug response.
Synthetic nucleoside analogs are widely used to treat a variety of diseases, including various types of cancer, HIV, hepatitis C, and other illnesses (Barreiro et al., 2004; Byrd et al., 2004; Li et al., 2004; Pearlman, 2004). Although nucleoside analogs are often the best available therapy, some common problems with nucleoside analog therapies occur, including lack of an initial response, or the development of resistance to therapy. One potential hypothesis for the ineffectiveness of some nucleoside analogs is genetic variation in nucleoside transporters, which function in the uptake of these compounds into cells. Genetic variation could lead to reduced function, or nonfunctional transporter proteins, which in turn could reduce the amount of drug that enters the cell and, therefore, the intracellular levels of the drug. Genetic variation in both concentrative and equilibrative nucleoside transporter members has been previously examined including CNT1 (Gray et al., 2004), CNT2 (Owen et al., 2005), CNT3 (Badagnani et al., 2005), and ENT1 (Osato et al., 2003).
In this report, we describe the functional characteristics of genetic variants of ENT2 that were identified previously in a large DNA sample set. ENT2 is thought to play a role in nucleoside analog therapy, since it has broad substrate specificity and is able to transport many of the currently used nucleoside analogs (Baldwin et al., 2004), including the pancreatic cancer drug gemcitabine (Garcia-Manteiga et al., 2003), and fludarabine (Molina-Arcas et al., 2003), used in the treatment of chronic lymphocytic leukemia. In a recent study, it was reported that expression levels of ENT2, but not ENT1, as measured by Western blot, correlated with cytotoxicity of fludarabine in cells isolated from patients with chronic lymphocytic leukemia (Molina-Arcas et al., 2005). This makes studies of genetic variation in ENT2 of particular interest for fludarabine therapy. The aim of this study was to evaluate the functional characteristics of variants of ENT2 and to describe its haplotype structure.
Materials and Methods
Genetic Analysis of ENT2. ENT2 variants were identified in the study of Leabman et al. (2003) through direct sequencing of its exons and flanking intronic regions in an ethnically diverse population sample of 247 individuals. The nucleotide diversity (π), was calculated as described by Tajima (1989). Synonymous and nonsynonymous polymorphisms were defined as described by Hartl and Clark (1997). Haplotypes were reconstructed from variant positions using PHASE, a Bayesian statistical method (Stephens et al., 2001). Before PHASE analysis, all singletons were removed from the analysis. Five haplotypes were identified through the PHASE analysis. The cladogram describing the ENT2 haplotypes was constructed by hand.
Construction of ENT2 Reference and ENT2-Variant Plasmids. Human ENT2 cDNA was subcloned into the amphibian high-expression vector pOX (Jegla and Salkoff, 1997). ENT2 reference was used as a template to create the three nonsynonymous variants and two deletions of ENT2 identified in the study of Leabman et al. (2003). Reference and variant sequences were confirmed by complete DNA sequencing at the University of California, San Francisco Biomolecular Resource Center.
Functional Screening and Kinetic Studies of Variants inXenopus laevisOocytes. Healthy stage V and stage VI X. laevis oocytes were injected with 30 to 50 ng of capped cRNA transcribed in vitro with T3 RNA polymerase (mCAP RNA Capping Kit; Stratagene, La Jolla, CA) from NotI-linearized pOX plasmids containing reference or variant ENT2 (NotI from New England Biolabs, Beverly, MA). Spectrophotometry was used to determine the concentration of cRNA, and an aliquot of each RNA preparation was run on a 1% agarose gel to ensure that the RNA was not degraded. Injected oocytes were stored in modified Barth's solution at 18°C (changed one to two times daily) for 2 to 3 days of expression before uptake studies. Seven to nine oocytes were incubated in Na+ buffer containing 1 μM 3H-substrate. Several different substrates were used including inosine, guanosine, uridine, hypoxanthine, fludarabine, and gemcitabine. All radiolabeled compounds were purchased from Moravek Biochemicals (Brea, CA). The injected oocytes were incubated with 3H-substrate (1 μM) for 20 min; ENT2 transport using X. laevis oocytes has been previously reported at 30 min (Yao et al., 2001). In the inhibition study, unlabeled inosine (2 mM) was used to inhibit the uptake of [3H]guanosine (1 μM). For the kinetic studies, 0.25 μM [3H]inosine was incubated with the unlabeled concentrations of inosine (1, 10, 50, 100, 500, 1000, 2000, and 4000 μM), and the uptake was examined for 30 min. The Vmax values are reported as pmol inosine/30-min uptake ± S.E. For all the studies with oocytes, uptake was terminated by the removal of buffer containing the radioligand and the oocytes were washed five times in ice-cold choline buffer. Oocytes were then individually lysed by the addition of 10% sodium dodecyl sulfate (100 μl), and the radioactivity associated with each oocyte was determined by scintillation counting. Uptake of all substrates in oocytes expressing each variant was determined in eight to nine oocytes from a single frog. The functional studies were repeated in oocytes from at least one other frog. Data are presented as pmol of substrate/oocyte/20-min uptake, and the error bars indicated are ±S.E. Uninjected oocytes incubated with the same reaction mix were used as a control.
Results and Discussion
The variable sites in SLC29A2 were identified through direct sequencing of all 11 SLC29A2 exons, and some flanking intronic regions. A summary of all of the total variants identified, as well as the frequency in which they were found in the sample population, is shown in Table 1 and can also be found at http://www.pharmgkb.org. Of the 14 variable sites, only five resulted in protein-altering variants, and were chosen for further study. All of the protein-altering variants were singletons, or variants that were found on only one chromosome in one individual; three of the five protein-altering variants were found in the African-American sample, and the other two were found in the European-American sample. ENT2 was unique among the 24 transporters studied by Leabman et al. (2003) in that it contained two deletion variants in the coding region. In total, 680 polymorphic sites were identified across the 24 transporter genes, but only five of these sites were deletions found in the coding region of the gene, two of which were found in SLC29A2. The two deletions likely arose from separate causes because one is near the middle of an exon, whereas the other is at the intron-exon boundary. The overall variation in SLC29A2 was much lower than the average found in the other genes in the studies of Leabman et al. (2003); the πT of SLC29A2, a measure of nucleotide diversity, is 1.64 × 10–4, versus the average πT of the genes, 5.09 × 10–4. However, the variation in SLC29A2 was similar to that observed for SLC29A1 (Osato et al., 2003). The low overall variation in the equilibrative transporter family suggests that these two transporter genes are under high selective pressure, with nonsynonymous variation being highly selected against. This contention is also supported by the greater frequency of synonymous variants in both genes, which would not result in a change in the encoded protein. Although there are no reports of an ENT2 knockout mouse, an ENT1 knockout mouse is viable and fertile (Choi et al., 2004). Consistent with its low variability, ENT2 has few haplotypes (in the coding and flanking intronic region), which are shown in the cladogram in Fig. 1. The *1 and *2 haplotypes account for nearly 95% of the overall haplotypes found, with *1 comprising about 75% of the population sampled. This haplotype profile is similar to that observed for ENT1, but showed considerably lower variability than the haplotype profiles of CNT family members (Gray et al., 2004; Badagnani et al., 2005; Owen et al., 2005).
Functional studies in oocytes revealed that ENT2-Δ845-846 was not able to take up inosine (or guanosine). In contrast, the other variants (as well as ENT2 reference) were able to transport both substrates (Fig. 2, a and b). ENT2-Δ845-846 was unable to transport inosine, because the deletion of two base pairs produced a change in the reading frame, which results in a severe truncation of the protein, and a subsequent loss of function. The variants ENT2-D5Y and ENT2-Δ551-556 showed reduced inosine uptake when compared with ENT2 reference, with ENT2-D5Y reaching statistical significance: p = 0.048 for ENT2-D5Y and p = 0.061 for ENTΔ551-556 (Fig. 2a). The loss of six base pairs in ENT2-Δ551-556 results in a two-amino acid deletion, and a nonsynonymous change of a third residue.
The uptake of a diverse array of substrates by ENT2 and its variants was examined (Fig. 3). Included in the analysis was a model purine (inosine), a model pyrimidine (uridine), and the nucleobase hypoxanthine, as well as the nucleoside analog drugs gemcitabine and fludarabine. Transport profiles were similar regardless of the substrate, with ENT2-D5Y exhibiting reduced function for all tested substrates, and ENT2-Δ845-846 not transporting any of the substrate panel. To gain insight into the mechanism of the reduced function of ENT2-D5Y, we performed kinetic studies with ENT2 reference and ENT2-D5Y with inosine and fludarabine. Both compounds showed a reduced Vmax for ENT2-D5Y relative to that of ENT2 reference. Representative curves of the inosine kinetics with ENT2 reference, ENT2-D5Y, and uninjected oocytes are shown in Fig. 4. The observed difference in Vmax (958 ± 53.6 pmol of inosine/oocyte/30-min uptake versus 706 ± 68.1 pmol of inosine/oocyte/30-min uptake for ENT2 reference and ENT2-D5Y, respectively) is statistically significant (p = 0.03), whereas the respective Km values were not statistically significant (p = 0.45). These data support the idea that the mechanism of reduction in activity of ENT2-D5Y is due to a reduced Vmax, possibly reflecting a reduced turnover rate constant or a reduction in the number of functional transporters expressed on the plasma membrane.
In summary, the low genetic and functional variation observed in ENT2 suggests a critical physiological role, similar to its homolog ENT1. Variants with altered function were observed in ENT2; however, because of their low frequency, these variants are unlikely to be a major source of variability in drug response. Our data do not explain previous studies in which enhanced ENT2 expression has been associated with response to anticancer drugs (Molina-Arcas et al., 2005). It is possible that polymorphisms in noncoding regions of ENT2 may explain variation in the expression levels of this gene.
Acknowledgments
We thank Ilaria Badagnani and Jennifer Gray for helpful discussions.
Footnotes
-
This work was funded by National Institutes of Health Grants GM61390 and GM42230.
-
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
-
doi:10.1124/dmd.105.006270.
-
ABBREVIATIONS: CNT, concentrative nucleoside transporter; ENT, equilibrative nucleoside transporter.
- Received July 4, 2005.
- Accepted October 5, 2005.
- The American Society for Pharmacology and Experimental Therapeutics