Dapagliflozin suppresses glucagon signaling in rodent models of diabetes

Abstract

Sodium-glucose cotransporter 2 (SGLT2) inhibitors are a class of antidiabetic drug used for the treatment of diabetes. These drugs are thought to lower blood glucose by blocking reabsorption of glucose by SGLT2 in the proximal convoluted tubules of the kidney. To investigate the effect of inhibiting SGLT2 on pancreatic hormones, we treated perfused pancreata from rats with chemically induced diabetes with dapagliflozin and measured the response of glucagon secretion by alpha cells in response to elevated glucose. In these type 1 diabetic rats, glucose stimulated glucagon secretion by alpha cells; this was prevented by dapagliflozin. Two models of type 2 diabetes, severely diabetic Zucker rats and db/db mice fed dapagliflozin, showed significant improvement of blood glucose levels and glucose disposal, with reduced evidence of glucagon signaling in the liver, as exemplified by reduced phosphorylation of hepatic cAMP-responsive element binding protein, reduced expression of phosphoenolpyruvate carboxykinase 2, increased hepatic glycogen, and reduced hepatic glucose production. Plasma glucagon levels did not change significantly. However, dapagliflozin treatment reduced the expression of the liver glucagon receptor. Dapagliflozin in rodents appears to lower blood glucose levels in part by suppressing hepatic glucagon signaling through down-regulation of the hepatic glucagon receptor.

Keywords: SGLT2 inhibition; dapagliflozin; diabetes; glucagon; glucagon receptor.

Fig. 1.

Response of alpha cells treated with dapagliflozin to increasing glucose. InR1-G9 cells were transfected with either a nontargeting (NT) or four different siRNAs targeting hamster SGLT2. Cells were cultured 3 d and transfected again for an additional 3 d. (A) An immunoblot shows the extent of SGLT2 knockdown. Un, untransfected control. Actin serves as a loading control. (B) Glucagon values from three biological replicates treated with each siRNA and cultured for 18 h in medium containing 5 mM glucose are compared with samples cultured for the same time in medium containing 25 mM glucose. The mean change in glucagon secretion is graphed for each sample with SEM. The control is the nontargeting siRNA. (C) Glucagon production by perfused pancreata in response to increased glucose in the perfusate. The red line on the graph shows glucagon secretion by pancreata treated with 20 µmol/L dapagliflozin. Dapagliflozin was added at 10 min after the perfusion began. At 25 min, the glucose in the perfusate was increased from 2.5 to 25 mM. The black line shows glucagon secretion by control pancreata in the same experiment. n = 6 for each group. STZ, streptozotocin; T1D, type 1 diabetes. (D) The mean plateau glucagon concentrations after challenge with 25 mM glucose in the presence or absence of 20 µmol/L dapagliflozin are compared with a challenge with 2.5 mM glucose without the drug. SDs are shown, and Student’s two-tailed t test was used to calculate P, the probability of a null hypothesis in this and other figures. **P < 0.015; ****P < 0.0001. Glucagon was measured with a Cisbio ELISA kit.

Fig. 2.

In the ZDF^fa/fa rodent model of type 2 diabetes, treatment with dapagliflozin significantly improves glucose disposal and glycemic control and reduces signs of hepatic glucagon action. (A) An oral glucose tolerance test was performed with conscious ZDF rats after 7-d treatment with dapagliflozin or placebo. The red line indicates mean glucose measurements ± SDs from five dapagliflozin-treated animals. The black line indicates similar measurements from four placebo-treated animals. *P < 0.02. (B) Six weeks of treatment of Zucker fatty rats with dapagliflozin significantly lowers HbA1c levels (P < 0.015). Means with SDs for blood HbA1c levels for six placebo-treated (black) and six dapagliflozin-treated (red bar, Dapa) ZDF rats are shown. (C) Liver samples from the placebo and dapagliflozin-treated animals after 7 wk were immunoblotted for the gluconeogenic transcription factor CREB, which is activated by phosphorylation. The upper band is labeled with an antibody for phosphorylated CREB, and the lower with an antibody recognizing all forms of CREB. The average ratio of phosphorylated to unphosphorylated CREB for the two sample sets measured by densitometry is graphed with SDs (P = 0.04). (D) Liver samples from the placebo and dapagliflozin-treated animals were immunoblotted for PEPCK and for tubulin (TUB) as a loading control. The average ratio of PEPCK to tubulin for the two sample sets measured by densitometry is graphed with SDs (P = 0.025). (E) Glycogen content of livers from 6 (control, black) or 9 (dapagliflozin, red) rats treated for 7 wk was measured. (F) Plasma glucagon was measured for placebo-treated (n = 11) or dapagliflozin-treated (n = 5) ZDF^fa/fa rats (P = 0.11) after 44 d, using a Mercodis ELISA kit. The lines on the scatter plot indicate the means of each group.

Fig. 3.

Dapagliflozin treatment improves glucose homeostasis and insulin sensitivity in db/db mice. (A) Consistent with the known action of dapagliflozin, treated animals had increased excretion of glucose compared with controls (P = 0.0015). The following parameters all decreased significantly in the dapagliflozin-treated group. (B) Blood glucose concentration (P = 0.0007). Urinary glucose and blood glucose were measured daily for 3 d for each animal. (C) Body weight (P = 0.003). (D) Whole-body fluid (P = 0.014). (E) Fat mass as a percentage of body mass (P < 0.05). (F) Lean mass of dapagliflozin mice increased (P = 0.03). n = 6 for each treatment group.

Fig. 4.

Dapagliflozin improves glucose disposal and suppresses endogenous glucose production without changing plasma glucagon levels in db/db mice. (A) Increased glucose infusion is required to maintain euglycemia in db/db mice treated for 4 wk with dapagliflozin compared with wild-type controls (P < 0.0005). Mean values ± SDs are graphed. (B) Endogenous glucose production is lower in dapagliflozin-treated (red bar) than in control (black bars) db/db mice in the basal (fasted) condition and in hyperinsulinemic (clamped) conditions. Mean values ± SDs are graphed. *P < 0.04; **P < 0.003; ***P < 0.0005. (C) The scatter plot shows plasma glucagon concentrations for the same mice as shown in A and B, measured with a Mercodis ELISA kit. Bars indicate means. P = 0.25. (D) Plasma insulin levels for the placebo (black) and dapagliflozin-treated animals (red) are for basal conditions and after the animals were clamped. The difference between two groups was not significantly different by two-tailed t test (P = 0.35 for basal condition; P = 0.76 for the clamped condition). (E) Liver samples from the these mice were probed by immunoblot with anti-glucagon receptor antibody, and the density of the protein bands for the receptor and tubulin loading control were measured by densitometry and graphed, with SDs in F. P < 0.002 by two-tailed t test.

Fig. 5.

Dapagliflozin improves cardiac function. (A) Dapagliflozin improves cardiac function in diabetic Zucker rats. At the end of the 7-wk drug treatment, cardiac fractional shortening of the left ventricles was measured. Means with SD for seven placebo-treated and six dapagliflozin-treated animals show a significant improvement (P = 0.008). (B) At the end of 4-wk treatment with dapagliflozin or placebo, cardiac fractional shortening of the left ventricles was measured in db/db mice. Means with SD for five placebo-treated and five dapagliflozin-treated animals show a significant improvement (P = 0.0014).