SGLT2 inhibition alters substrate utilization and mitochondrial redox in healthy and failing rat hearts

Abstract

Previous studies highlight the potential for sodium-glucose cotransporter type 2 (SGLT2) inhibitors (SGLT2i) to exert cardioprotective effects in heart failure by increasing plasma ketones and shifting myocardial fuel utilization toward ketone oxidation. However, SGLT2i have multiple in vivo effects and the differential impact of SGLT2i treatment and ketone supplementation on cardiac metabolism remains unclear. Here, using gas chromatography-mass spectrometry (GC-MS) and liquid chromatography-tandem mass spectrometry (LC-MS/MS) methodology combined with infusions of [13C6]glucose or [13C4]βOHB, we demonstrate that acute SGLT2 inhibition with dapagliflozin shifts relative rates of myocardial mitochondrial metabolism toward ketone oxidation, decreasing pyruvate oxidation with little effect on fatty acid oxidation in awake rats. Shifts in myocardial ketone oxidation persisted when plasma glucose levels were maintained. In contrast, acute βOHB infusion similarly augmented ketone oxidation, but markedly reduced fatty acid oxidation and did not alter glucose uptake or pyruvate oxidation. After inducing heart failure, dapagliflozin increased relative rates of ketone and fatty acid oxidation, but decreased pyruvate oxidation. Dapagliflozin increased mitochondrial redox and reduced myocardial oxidative stress in heart failure, which was associated with improvements in left ventricular ejection fraction after 3 weeks of treatment. Thus, SGLT2i have pleiotropic effects on systemic and heart metabolism, which are distinct from ketone supplementation and may contribute to the long-term cardioprotective benefits of SGLT2i.

Keywords: Cardiology; Glucose metabolism; Intermediary metabolism; Metabolism; Mitochondria.

Figure 1. Acute dapagliflozin treatment significantly increases relative rates of

βOHB oxidation at the expense of pyruvate oxidation in chow-fed male rats. (A) Outline of study design. Rats were given an intravenous bolus of ¹⁴C[2-DG] during the last 20 minutes of the infusion. (B–K) Weight change (B), plasma glucose (C), plasma insulin (D), plasma glucagon (E), plasma NEFAs (F), whole-body βOHB turnover (G), plasma βOHB (H), relative rates of myocardial βOHB oxidation (V_BDH) to total mitochondrial oxidation (V_CS) (I), myocardial glucose uptake (J), and relative rates of myocardial pyruvate oxidation (V_PDH) to total mitochondrial oxidation (V_CS) (K) in chow-fed male rats treated as in A. In panels (B–K), n = 24, 20, 20 (B and D and F); n = 22, 20, 20 (C); n = 24, 19, 19 (E); n = 11, 10, 11 (G); n = 22, 19, 20 (H); n = 12, 10, 11 (I); n = 11, 9, 9 (J); and n = 12, 10, 9 (K). All data are represented as mean ± SEM. P < 0.05 by 1-way ANOVA with Bonferroni’s corrections for multiple comparisons. Dapa, dapagliflozin; Tx, treatment; po, orally.

Figure 2. βOHB infusion significantly increases relative rates of βOHB oxidation in chow-fed male rats.

(A) Outline of study design. Rats were given an intravenous bolus of ¹⁴C[2-DG] during the last 20 minutes of the infusion. (B–K) Weight change (B), plasma glucose (C), plasma insulin (D), plasma glucagon (E), whole-body glucose turnover (F), plasma NEFAs (G), whole-body βOHB turnover (H), plasma βOHB (I), plasma acetoacetate (J), and plasma βOHB:AcAc in chow-fed male rats treated as in A. In panels (B–K), n = 11, 10, 9 (B); n = 9, 10, 13 (C); n = 8, 9, 12 (D); n = 9, 10, 11 (E, G, and I); n = 4, 5, 6 (F); n = 4, 4, 5 (H); n = 9, 10, 10 (J); and n = 8, 9, 11 (K). All data are represented as mean ± SEM. P < 0.05 by 1-way ANOVA with Bonferroni’s corrections for multiple comparisons.

Figure 3. Acute dapagliflozin-mediated increases in relative rates of

βOHB oxidation are driven by plasma βOHB levels. (A–E) Relative rates of myocardial βOHB oxidation (V_BDH) to total mitochondrial oxidation (V_CS) (A), myocardial glucose uptake (B), relative rates of myocardial pyruvate oxidation (V_PDH) to total mitochondrial oxidation (V_CS) (C), myocardial βOHB:AcAc (D), and myocardial acetyl-CoA content (E) in chow-fed male rats treated with vehicle, 1.5 mg/kg body weight dapagliflozin (po), or infused with 50 μmol/(kg-min) βOHB. (F–H) Representative Western blot analysis of GLUT1 and GLUT4 (F), pPDH/PDH (G), and BDH1 and SCOT (H) in the hearts of chow-fed male rats treated as in A–E. HSP90 was used as a loading control. Quantification of blots shown below each blot. (I) Correlation of plasma βOHB and V_BDH/ V_CS in rats treated as in A–E. In panels (A–I), n = 4, 4, 5 (A); n = 5, 5, 6 (B and C); n = 9, 9, 7 (D); n = 9, 12, 13 (E); n = 4 per group (F–H); and n = 4, 4, 5 (I). All data are represented as mean ± SEM. P < 0.05 y 1-way ANOVA with Bonferroni’s corrections for multiple comparisons.

Figure 4. Acute dapagliflozin treatment increases hepatic ketogenesis and plasma ketone levels in sham-surgery and MI rats.

(A) Outline of study design. Rats were given an intravenous bolus of ¹⁴C[2-DG] during the last 20 minutes of the infusion. (B–K) Weight change (B), urine glucose (C), plasma glucose (D), plasma glucagon (E), whole-body glucose turnover (F), plasma insulin (G), plasma NEFAs (H), whole-body βOHB turnover (I), plasma βOHB (J), and plasma acetoacetate (K) in chow-fed male rats treated as in A. In panels B–K, n = 26, 27, 33, 29 (B); n = 18, 16, 21,18 (C); n = 24, 26, 32, 29 (D); n = 22, 26, 28, 27 (E); n = 7, 9, 7, 9 (F); n = 25, 27, 32, 30 (G); n = 26, 27, 33, 30 (H and J); n = 9, 10, 10, 9 (I);and n = 16, 17, 21, 19 (K). All data are represented as mean ± SEM. P < 0.05 by 1-way ANOVA with Bonferroni’s corrections for multiple comparisons.

Figure 5. Acute dapagliflozin treatment reduces and increases relative rates of glucose and βOHB oxidation, respectively, in the ischemic/infarct area and nonischemic myocardium remote from the infarct area 2 weeks after MI.

(A–F) Glucose uptake, relative rates of myocardial pyruvate oxidation (V_PDH) to total mitochondrial oxidation (V_CS), and relative rates of βOHB oxidation (V_BDH) to total mitochondrial oxidation (V_CS) in the ischemic/infarct region (A–C) and nonischemic myocardium remote from the infarct area (D–F) in chow-fed male rats 2 weeks after MI or sham surgery. (G–I) Representative Western blot analysis of GLUT1 and GLUT4 (G), pPDH/PDH (H), and BDH1 and SCOT (I) in the LV of sham and MI rats treated as in A–F. HSP90 was used as a loading control. Quantification of blots shown in the right panels. (J) Correlation of plasma βOHB and V_BDH/V_CS in rats treated as in A–F. In panels A–J, n = 10, 8, 10, 9 (A); n = 10, 12, 7, 6 (B); n = 12, 12, 10, 10 (C); n = 11, 8, 6, 5 (D); n = 7, 9, 7, 9 (E); n = 11, 12, 7, 6 (F); n = 6 per group (G–I); and n = 8, 7, 7, 9 (J). All data are represented as mean ± SEM. P < 0.05 by 1-way ANOVA with Bonferroni’s corrections for multiple comparisons.

Figure 6. Acute dapagliflozin treatment increases plasma βOHB and improves mitochondrial redox in rats 2 weeks after MI.

(A–J) Weight change (A), plasma glucose (B), plasma insulin (C), plasma NEFAs (D), plasma βOHB (E), myocardial βOHB (F), myocardial acetoacetate (G), myocardial βOHB:AcAc (H), myocardial GSH:GSSG (I) and myocardial malondialdehyde (MDA) (J) in chow-fed male rats 2 weeks after MI or sham surgery and treated with vehicle or 1.5 mg/kg body weight dapagliflozin (po) for 6 hours. In panels A–J, n = 8, 8, 6, 7 (A); n = 7, 8, 7, 6 (B); n = 8, 10, 7, 7 (C); n = 7, 9, 8, 7 (D); n = 6, 7, 6, 5 (E); n = 8, 7, 7, 5 (F–H); n = 10, 9, 7, 5 (I); and n = 7, 7, 12, 10 (J). All data are represented as mean ± SEM. P < 0.05 by 1-way ANOVA with Bonferroni’s corrections for multiple comparisons.

Figure 7. Chronic dapagliflozin treatment increases plasma

βOHB and improves LV ejection fraction in rats 4 weeks after MI. (A) Outline of study design. Rats were given an intravenous bolus of ¹⁴C[2-DG] during the last 20 minutes of the infusion. (B–L) Body weight (B), plasma glucose (C), plasma glucagon (D), plasma insulin (E), plasma βOHB (F), plasma acetoacetate (G), plasma NEFAs (H), percentage change in heart rate (I), percentage change in diastolic volume (J), percentage change in stroke volume (K), and percentage change in LV ejection fraction (L) in chow-fed male rats 4 weeks after MI surgery and dapagliflozin (po) treatment (1.0 mg/kg body weight × 3 weeks). Panels B-H, were endpoint measures (4 weeks post-MI), while panels I-L represent percentage change from baseline (1 week post-MI compared with 4-weeks post-MI). In panels B–L, n = 15, 15 (B and I–L); n = 15, 13 (C); n = 11,10 (D); n = 9, 10 (E); n = 10, 8 (F); n = 10, 9 (G); and n = 11, 9 (H). All data are represented as mean ± SEM. P < 0.05 by unpaired Student’s t test.

Figure 8. Chronic dapagliflozin treatment reduces and increases relative rates of glucose and βOHB oxidation, respectively, in the LV myocardium remote from the infarct area 4 weeks after MI.

(A–D) Glucose uptake (A), relative rates of myocardial pyruvate oxidation (V_PDH) to total mitochondrial oxidation (V_CS) (B), relative rates of βOHB oxidation (V_BDH) to total mitochondrial oxidation (V_CS) (C), and relative rates of pyruvate, βOHB, and unmeasured free fatty acid and amino acid oxidation (D) in the nonischemic LV myocardium remote from the infarct area in chow-fed male rats 4 weeks after MI surgery and dapagliflozin (po) treatment (1.0 mg/kg body weight × 3 weeks). (E–F) Representative Western blot analysis of GLUT1, GLUT4, BDH1, and SCOT (E) and pPDH/PDH in the LV of MI rats treated as in A–D. n = 7 per group. HSP90 was used as a loading control. Quantification of blots shown in the right panels. (G) Correlation of plasma βOHB and V_BDH/V_CS in rats treated as in A–D. (H–J) Myocardial βOHB:AcAc (H), myocardial GSH:GSSG (I), and myocardial MDA (J) in the LV of MI rats treated as in A–D. In panels A–J, n = 8, 6 (A and J); n = 6, 8 (B); n = 8, 7 (C and G and H); n = 6–8 per treatment group (D); n = 7, 7 (E and F); and n = 10, 10 (I). All data are represented as mean ± SEM. P < 0.05 by unpaired Student’s t test.