SLC2A9 (GLUT9) mediates urate reabsorption in the mouse kidney

Abstract

Uric acid (UA) is a metabolite of purine degradation and is involved in gout flairs and kidney stones formation. GLUT9 (SLC2A9) was previously shown to be a urate transporter in vitro. In vivo, humans carrying GLUT9 loss-of-function mutations have familial renal hypouricemia type 2, a condition characterized by hypouricemia, UA renal wasting associated with kidney stones, and an increased propensity to acute renal failure during strenuous exercise. Mice carrying a deletion of GLUT9 in the whole body are hyperuricemic and display a severe nephropathy due to intratubular uric acid precipitation. However, the precise role of GLUT9 in the kidney remains poorly characterized. We developed a mouse model in which GLUT9 was deleted specifically along the whole nephron in a tetracycline-inducible manner (subsequently called kidney-inducible KO or kiKO). The urate/creatinine ratio was increased as early as 4 days after induction of the KO and no GLUT9 protein was visible on kidney extracts. kiKO mice are morphologically identical to their wild-type littermates and had no spontaneous kidney stones. Twenty-four-hour urine collection revealed a major increase of urate urinary excretion rate and of the fractional excretion of urate, with no difference in urate concentration in the plasma. Polyuria was observed, but kiKO mice were still able to concentrate urine after water restriction. KiKO mice displayed lower blood pressure accompanied by an increased heart rate. Overall, these results indicate that GLUT9 is a crucial player in renal handling of urate in vivo and a putative target for uricosuric drugs.

Keywords: GLUT9; Lithiasis; SLC2A9; Urate; Uric acid.

Fig. 1

GLUT9 is mainly expressed at the basolateral membrane of the DCT and slightly in the PCT. a Western blot of microdissected tubules from wild-type mice. GLUT9 is strongly detectable in the DCT. Some expression is also visible in the PCT after a longer exposure. NCX1 is used as positive control for the accuracy of the DCT microdissection. Protein loading can be evaluated by actin quantification. b Immunostaining of GLUT9 on wild-type kidney section. GLUT9 signal is restricted to cortical distal convoluted tubules (scale bar: 100 μm). c Co-immunostaining of GLUT9 and NCX1, and of GLUT9 and AQP2 on wild-type kidney sections. Both GLUT9 and NCX1 signals are co-localizing at the basolateral membrane of the DCT. There is no co-localization of GLUT9 and AQP2 (scale bar: 10 μm)

Fig. 2

Doxycycline-induced deletion of GLUT9 in the kidney: molecular analysis. a PCR on cDNA obtained 4 months after doxycycline induction from control and kiKO mouse kidneys. Recombination of both Glut9a and b isoforms is observed in the kidney of kiKO mice (n = 3). b Relative abundance of Glut9 transcript from total kidney 4 months after doxycycline induction, as measured by quantitative real-time PCR. Values are means ± SD relative to control (n = 10, *p < 0.05, by Student’s t test). c Renal GLUT9 protein expression levels in control and kiKO mice by Western blot 4 months after doxycycline induction. No GLUT9 protein is detectable in the kidney of kiKO mice (n = 3–4). Protein loading was evaluated by actin

Fig. 3

Doxycycline-induced deletion of GLUT9 in the kidney: functional analysis. a Time-course of the urate/creatinine ratio from spot urine after doxycycline induction (starting at day 0). Four days after the induction, the urate/creatinine ratio is significantly increased in kiKO mice compared to control. Four months after the induction by the doxycycline, the difference between control and kiKO mice is still present. Values are means ± SD (n = 6, *p < 0.05 by Student’s t test). b Twenty-four-hour urine collection of kiKO mice presents an important white deposit (arrow) when kept at room temperature. This deposit is made of uric acid crystals (not shown) and is absent in control urine. The 24-h volume of kiKO mice urine is higher than the volume of control urine and accordingly, urine is more diluted. c Measurement of 24-h urate excretion rate. Urate excretion rate for kiKO mice is higher than for controls. Values are means ± SD (n = 8, *p < 0.05). d SUA analysis. There is no difference between control (white bars) and kiKO (black bars) mice regarding plasma concentration of urate. Males had higher urate concentration in the plasma than females. Values are means ± SD (n = 8, *p < 0.05). e Fractional excretion of urate (FE urate). A significantly higher FE urate was measured in kiKO mice compared to controls. Values are means ± SD (n = 8, *p < 0.05)

Fig. 4

More diluted urine in the hyperuricosuric kiKO mice. a KiKO mice display an increase urine volume compared to control mice. Values are means ± SD (n = 16, *p < 0.05). b Urine osmolality is not changed between control and kiKO mice. Values are means ± SD (n = 23). c Urine concentration test. After 9 and 23 h of water deprivation, both control and kiKO mice are able to concentrate their urine the same way, with significantly increased urine osmolality compared to baseline, but no difference between the two genotypes. Values are means ± SD (n = 10). d. Measurement of osmolality in the cortex and the papilla of control (white bars) and kiKO (black bars) mice did not show any difference between both genotypes. A significant increase of osmolality is measured in the papilla by comparison with the cortex, for both control and kiKO mice. Data are expressed as mOsm/l per mg of renal tissue. Values are means ± SD (n = 20, *p < 0.05)

Fig. 5

Expression level of AQP2 and V2R in the kidney. No difference in the Aqp2 (a) and V2r (b) expression level was observed between control and kiKO mice by qPCR. Values are means ± SD (n = 19 for Aqp2 and n = 9 for V2r). c AQP2 protein expression level in control and kiKO mice. Values are means ± SD (n = 3 to 5)

Fig. 6

Possible compensatory mechanisms. qPCR analysis of the relative abundance of several known urate transporters and enzymes involved in urate metabolism in the kidney (a), the ileum (b), the colon (c), and the liver (d). Data are normalized to control expression. Values > 1 indicate higher expression in kiKO mice compared to controls. Values are means ± SD (n = 5, *p < 0.05)

Fig. 7

kiKO mice have lower blood pressure and higher heart rate. a Measurement of blood pressure in control (white bars) and kiKO (black bars) mice. A decrease of systolic (SBP) and diastolic (DBP) blood pressure is observed in kiKO mice. Values are means ± SD (n = 6, *p < 0.05). b Heart rate (in beats per minutes, b.p.m.) was higher in kiKO mice compared to control mice. Values are means ± SD (n = 6, *p < 0.05)