Increase in SGLT1-mediated transport explains renal glucose reabsorption during genetic and pharmacological SGLT2 inhibition in euglycemia

Abstract

In the kidney, the sodium-glucose cotransporters SGLT2 and SGLT1 are thought to account for >90 and ∼3% of fractional glucose reabsorption (FGR), respectively. However, euglycemic humans treated with an SGLT2 inhibitor maintain an FGR of 40-50%, mimicking values in Sglt2 knockout mice. Here, we show that oral gavage with a selective SGLT2 inhibitor (SGLT2-I) dose dependently increased urinary glucose excretion (UGE) in wild-type (WT) mice. The dose-response curve was shifted leftward and the maximum response doubled in Sglt1 knockout (Sglt1-/-) mice. Treatment in diet with the SGLT2-I for 3 wk maintained 1.5- to 2-fold higher urine glucose/creatinine ratios in Sglt1-/- vs. WT mice, associated with a temporarily greater reduction in blood glucose in Sglt1-/- vs. WT after 24 h (-33 vs. -11%). Subsequent inulin clearance studies under anesthesia revealed free plasma concentrations of the SGLT2-I (corresponding to early proximal concentration) close to the reported IC50 for SGLT2 in mice, which were associated with FGR of 64 ± 2% in WT and 17 ± 2% in Sglt1-/-. Additional intraperitoneal application of the SGLT2-I (maximum effective dose in metabolic cages) increased free plasma concentrations ∼10-fold and reduced FGR to 44 ± 3% in WT and to -1 ± 3% in Sglt1-/-. The absence of renal glucose reabsorption was confirmed in male and female Sglt1/Sglt2 double knockout mice. In conclusion, SGLT2 and SGLT1 account for renal glucose reabsorption in euglycemia, with 97 and 3% being reabsorbed by SGLT2 and SGLT1, respectively. When SGLT2 is fully inhibited by SGLT2-I, the increase in SGLT1-mediated glucose reabsorption explains why only 50-60% of filtered glucose is excreted.

Keywords: diabetes mellitus; glucose reabsorption; glucose transport; proximal tubule; sodium glucose cotransport inhibitor.

Fig. 1.

Maximum acute glucosuric response to sodium-glucose cotransporter SGLT2 inhibitor is enhanced 2-fold in Sglt1−/− vs. wild-type (WT) mice in metabolic cage studies. A: urinary glucose excretion (UGE) was greater in Sglt1−/− compared with WT mice following vehicle application. B: empagliflozin dose dependently increased UGE in WT. Please note different scale of y-axis vs. A. Compared with WT, the empagliflozin-induced UGE was shifted leftward and the maximum response doubled in Sglt1−/− mice. The difference between dose-response curves, which reflects the glucose reabsorption mediated via SGLT1 in WT mice, reached a maximum at a dose of ∼0.4 mg/kg (indicated at the left of the vertical lines) and was maintained (all vertical lines have same length) for higher doses up to 10 mg/kg, indicating a high selectivity of the SGLT2 inhibitor vs. SGLT1 in this dose range. Note that empagliflozin began to increase glucose excretion in WT when reabsorption via SGLT1 reached its maximum; n = 4–8/dose and genotype. *P < 0.05 vs. WT by ANOVA and unpaired Student's t-test.

Fig. 2.

Chronic glucosuric response to SGLT2 inhibitor is enhanced 1.5- to 2-fold in Sglt1−/− vs. WT mice, associated with greater increases in food and fluid intake. Depicted are effects of empagliflozin application (300 mg/kg of diet) for 3 wk on urinary glucose/creatinine ratios (A), absolute levels of and changes in blood glucose (B), and changes in food intake (C), fluid intake (D), and body weight (E). Spontaneous urine collections were made, and blood was collected by tail snip in awake mice; food and fluid intake was measured in regular cages; n = 9–11/group. *P < 0.05 vs. WT by ANOVA and unpaired Student's t-test. #P < 0.05 vs. day 0 same group by paired Student's t-test.

Fig. 3.

SGLT2 inhibitor empagliflozin reduces fractional renal glucose reabsorption to 44% in WT and completely prevented renal glucose reabsorption in Sglt1−/− mice. Inulin clearance studies were performed under terminal anesthesia in the late morning following empagliflozin application (300 mg/kg of diet) for 3 wk. Depicted is the relationship between filtered glucose and absolute renal glucose excretion (A), absolute renal glucose reabsorption (B), and fractional glucose reabsorption (C) in Sglt1−/− and WT mice. Each dot represents 1 clearance experiment period; n = 4–9/group. In A and B, the line of identity is included as a dashed line for ease of interpretation. Based on previous studies (3, 17), values for WT untreated mice are expected close to 0 for absolute renal glucose excretion (A), close to the line of identity for absolute renal glucose reabsorption (B), and close to 100% for fractional glucose reabsorption (C). D: free plasma concentrations of empagliflozin, corresponding to early tubular concentrations, were similar to reported IC₅₀ for mouse SGLT2 (∼1–2 nM) (4), when the drug was given “in the diet.” Additional application of empagliflozin 1 h before the study increased free plasma concentrations to 20–22 nM and reduced fractional glucose reabsorption in WT by ∼55–60%, similar to the phenotype of untreated Sglt2−/− mice (17), with the remaining glucose reabsorption (∼40–45%) being mediated by SGLT1.

Fig. 4.

Combined gene knockout of Sglt1 and Sglt2 completely prevented renal glucose reabsorption. Inulin clearance studies were performed under terminal anesthesia. Depicted is the relationship between filtered glucose and absolute renal glucose excretion (A), absolute renal glucose reabsorption (B), and fractional renal glucose reabsorption (C) in male and female Sglt1/Sglt2 double knockout mice. Each dot represents 1 clearance experiment period. In A and B, the line of identity is included as a dashed line for ease of interpretation. SGLT2 and SGLT1 explain renal glucose reabsorption. Values expected for WT untreated mice are depicted for comparison (see legend for Fig. 3 for details).