SLC2A9 is a high-capacity urate transporter in humans

Abstract

Background: Serum uric acid levels in humans are influenced by diet, cellular breakdown, and renal elimination, and correlate with blood pressure, metabolic syndrome, diabetes, gout, and cardiovascular disease. Recent genome-wide association scans have found common genetic variants of SLC2A9 to be associated with increased serum urate level and gout. The SLC2A9 gene encodes a facilitative glucose transporter, and it has two splice variants that are highly expressed in the proximal nephron, a key site for urate handling in the kidney. We investigated whether SLC2A9 is a functional urate transporter that contributes to the longstanding association between urate and blood pressure in man.

Methods and findings: We expressed both SLC2A9 splice variants in Xenopus laevis oocytes and found both isoforms mediate rapid urate fluxes at concentration ranges similar to physiological serum levels (200-500 microM). Because SLC2A9 is a known facilitative glucose transporter, we also tested whether glucose or fructose influenced urate transport. We found that urate is transported by SLC2A9 at rates 45- to 60-fold faster than glucose, and demonstrated that SLC2A9-mediated urate transport is facilitated by glucose and, to a lesser extent, fructose. In addition, transport is inhibited by the uricosuric benzbromarone in a dose-dependent manner (Ki = 27 microM). Furthermore, we found urate uptake was at least 2-fold greater in human embryonic kidney (HEK) cells overexpressing SLC2A9 splice variants than nontransfected kidney cells. To confirm that our findings were due to SLC2A9, and not another urate transporter, we showed that urate transport was diminished by SLC2A9-targeted siRNA in a second mammalian cell line. In a cohort of men we showed that genetic variants of SLC2A9 are associated with reduced urinary urate clearance, which fits with common variation at SLC2A9 leading to increased serum urate. We found no evidence of association with hypertension (odds ratio 0.98, 95% confidence interval [CI] 0.9 to 1.05, p > 0.33) by meta-analysis of an SLC2A9 variant in six case-control studies including 11,897 participants. In a separate meta-analysis of four population studies including 11,629 participants we found no association of SLC2A9 with systolic (effect size -0.12 mm Hg, 95% CI -0.68 to 0.43, p = 0.664) or diastolic blood pressure (effect size -0.03 mm Hg, 95% CI -0.39 to 0.31, p = 0.82).

Conclusions: This study provides evidence that SLC2A9 splice variants act as high-capacity urate transporters and is one of the first functional characterisations of findings from genome-wide association scans. We did not find an association of the SLC2A9 gene with blood pressure in this study. Our findings suggest potential pathogenic mechanisms that could offer a new drug target for gout.

Figure 1. Characterisation of Urate Fluxes and Kinetics Mediated by Human SLC2A9 Expressed in Xenopus laevis Oocytes and the Effect of Short-Chain Fatty Acids and Uricosurics on Urate Transport

(A) Time course of SLC2A9a-mediated urate uptake into Xenopus oocytes. Oocytes were injected with SLC2A9a cRNA or water and 4 d later incubated with 10 μM urate at 22 °C for the time periods indicated. Symbols represent the average uptake into ten oocytes per time point, with error bars representing the standard error of the mean (SEM). Solid circles show total uptake into oocytes expressing SLC2A9a cRNA, open circles urate uptake into water-injected oocytes, and inverted triangles the net uptake obtained by subtracting the water data from the total uptake for each time point. (B) Kinetics of human SLC2A9-mediated urate uptake in Xenopus oocytes. Symbols represent the mean data from six separate experiments each using ten oocytes per substrate concentration. Uptake was measured in oocytes injected with either SLC2A9a or SLC2A9b cRNA 4 d prior and corrected for uptake into water-injected oocytes. The curve was fitted by nonlinear regression analysis. The K _m = 981 μM and the V _max = 304 pmol/oocyte/20 min. Insert shows an Eadie-Hofstee plot of the same data. (C) Effect of uricosemics and short chain fatty acids on human SLC2A9a mediated urate uptake. Xenopus oocytes were injected with water or SLC2A9a cRNA and urate uptake measured 4 d later. Oocytes were incubated with 10 μM urate for 20 min, and bars represent net uptake determined by subtracting uptake into water-injected eggs from total uptake into SLC2A9a expressing oocytes, error bars are the SEM. Note: at no time was the urate uptake into the water-injected eggs more than 11% of the total uptake into SLC2A9-expressing oocytes. Compounds used were: no additional reagent (Cont), lactate 1 mM (Lac), pyruvate 1 mM (Pyr), butyrate 1 mM (But), acetate 1 mM (Ace), probenecid 1 mM (Prob), furosemide 100 μM (Furo), and phloretin 1 mM (Phlo). For comparison, a separate series of oocytes were injected with SLC2A1 (GLUT1) or SLC2A2 (GLUT2) cRNA and 100 μM urate uptake measured 4 d later using 30 min incubation times. Bars represent net uptake corrected for uptake into water-injected oocytes. *Significant inhibition at p ≤ 0.05. (D) Dose-dependent effect of benzbromarone on SLC2A9a-mediated urate uptake. Oocytes were injected with SLC2A9a cRNA or water and 4 d later were incubated with 10 μM urate for 20 min in the presence of 0, 1, 10, or 100 μM benzbromarone. Points represent the mean net uptake into ten oocytes corrected for uptake into water-injected eggs. Error bars were smaller than the data points.

Figure 2. Interaction between Hexoses and Human SLC2A9a-Mediated Urate Uptake in Xenopus Oocytes

Oocytes were injected with SLC2A9a cRNA 4 d prior to uptake experiments. Symbols represent mean net substrate uptake and error bars the SEM for measurements made in 6–10 oocytes per condition. Total uptake of substrates was measured into SLC2A9a cRNA-injected oocytes and then corrected for the uptake measured under identical conditions using water injected oocytes from the same batch of eggs. (A) Uptake over 30 min of 50 μM glucose in the presence of increasing concentrations of urate. (B) Uptake over 20 min of 5 μM urate in the presence of increasing concentrations of glucose. (C) Uptake over 30 min of 50 μM fructose in the presence of increasing concentrations of urate. (D) Human SLC2A9a-mediated glucose and urate exchange. Oocytes injected with SLC2A9a cRNA or water, 4 d prior, were incubated with non-radiolabelled 2 mM L-glucose or 2 mM urate for 1 h. Oocytes were then washed and then incubated at 22 °C for 30 min in ¹⁴C-labeled 10 μM D-glucose. Bars represent the average total uptake into 20 eggs expressing SLC2A9a or water-injected eggs, and the difference between the two, the net uptake. Error bars represent the SEM.

Figure 3. Human SLC2A9a-Mediated Urate Efflux from Xenopus Oocytes

(A) Comparison of urate and L-glucose efflux from SLC2A9a-expressing oocytes. Oocytes injected with SLC2A9a cRNA 4 d prior (triangles) or water-injected eggs (circles or squares) were injected with ¹⁴C-labelled urate (circles, triangles, or diamonds) or ¹⁴C-labelled L-glucose (filled squares) to give an estimated initial intracellular concentration of 200 μM. 20 oocytes per condition were incubated at 22 °C in efflux medium, which was sampled every 2 min. Efflux media contained 5 mM D-glucose (open circles), 5 mM L-glucose (filled squares, circles or triangles), or 2 mM urate (filled diamonds). Data points represent the log percentage of urate or L-glucose remaining in the oocytes for each time point. Lines were fitted by linear regression. (B) Acceleration of SLC2A9a mediated urate efflux by extracellular D-glucose or D-fructose. Oocytes injected with SLC2A9a cRNA 4 d prior (triangles) or water-injected eggs were injected with ¹⁴C-labelled urate to give an estimated intracellular concentration of 200 μM. 20 oocytes per condition were incubated at 22 °C in efflux medium, which was sampled every 2 min. Efflux media contained 5 mM D-glucose (filled circles), 5 mM L-glucose (filled triangles) or 5 mM D-fructose (open circles). Data points represent the log percentage of urate remaining in the oocytes for each time point corrected for the efflux of urate from the water-injected eggs under the same conditions. Lines were fitted by linear regression.

Figure 4. Specificity of Human SLC2A9a and SLC2A9b Antibodies and Urate Uptake into SLC2A9 Transfected Human Embryonic Kidney Cells

(A) Western blotting of hSLC2A9 expressed in HEK cells. Expression of hSLC2A9a was detected as a broad band at approximately 50 kDa by a polyclonal antibody raised against the N terminus of SLC2A9a in HEK293 cells overexpressing SLC2A9a (middle lane, Slc2a9a) but not detected in cells overexpressing SLC2A9b (right lane, Slc2a9b), or nontransfected HEK cells (left lane, HEK). (B) Expression of hSLC2A9b in transfected HEK293 cells. hSLC2A9b was detected by a polyclonal antibody raised against the N terminus of hSLC2A9b in HEK293 cells overexpressing hSLC2A9b (right lane, Slc2a9b), but not detected in cells overexpressing hSLC2A9a (middle lane, Slc2a9a). (C) Increase in urate uptake in HEK293 cells overexpressing either hSLC2A9a or hSLC2A9b. Radiolabelled urate uptake was measured in human embryonic kidney cells, which were stably overexpressing SLC2A9a or SLC2A9b as compared to their respective nontransfected controls. Uptakes of 120 μM urate were measured over 6 min at 37 °C. *p < 0.05; **p < 0.02.

Figure 5. Expression of SLC2A9 and Urate Uptake into Mouse Insulinoma MIN6 Cells

(A) Western blotting of mouse SLC2A9 in MIN6 cells. Cell lysates were analyzed by Western blot using antibodies to murine SLC2A9a and SLC2A9b of MIN6 cell lysates. (B) ¹⁴C-urate uptake into MIN6 cells is competitively inhibited by cold urate. Radiolabelled urate uptake was measured in mouse insulinoma cells, which endogenously express mSLC2A9a or mSLC2A9b. Uptake of ¹⁴C-urate, 120 μM, was measured for 6 min at 37 °C in the presence or absence of an additional 1mM cold urate. ** p < 0.02. (C) SLC2A9a and SLC2A9b expression is reduced by treatment of MIN6 cells with siRNA specific for mSLC2A9. Cells were treated with either scrambled RNA (scRNA) or mSLC2A9-specific siRNA and cell lysates run on a Western blot and probed with an antibody specific for either mSLC2A9a (mGLUT9a) or mSLC2A9b (mGLUT9b). (D) Reduction of urate uptake into MIN6 cells by transfection with siRNA targeted to SLC2A9. MIN6 cells were transfected with either SLC2A9-specific siRNA9 or scrambled RNA (scRNA). ¹⁴C-urate, 120 μM, uptake into transfected or untransfected cells was measured for 6 min at 37 °C. *p < 0.05.