Defective Branched-Chain Amino Acid Catabolism Disrupts Glucose Metabolism and Sensitizes the Heart to Ischemia-Reperfusion Injury

Abstract

Elevated levels of branched-chain amino acids (BCAAs) have recently been implicated in the development of cardiovascular and metabolic diseases, but the molecular mechanisms are unknown. In a mouse model of impaired BCAA catabolism (knockout [KO]), we found that chronic accumulation of BCAAs suppressed glucose metabolism and sensitized the heart to ischemic injury. High levels of BCAAs selectively disrupted mitochondrial pyruvate utilization through inhibition of pyruvate dehydrogenase complex (PDH) activity. Furthermore, downregulation of the hexosamine biosynthetic pathway in KO hearts decreased protein O-linked N-acetylglucosamine (O-GlcNAc) modification and inactivated PDH, resulting in significant decreases in glucose oxidation. Although the metabolic remodeling in KO did not affect baseline cardiac energetics or function, it rendered the heart vulnerable to ischemia-reperfusion injury. Promoting BCAA catabolism or normalizing glucose utilization by overexpressing GLUT1 in the KO heart rescued the metabolic and functional outcome. These observations revealed a novel role of BCAA catabolism in regulating cardiac metabolism and stress response.

Keywords: N-glcNacation; branched-chain amino acids; cardiac metabolism; glucose; ischemia-reperfusion; mitochondria; pyruvate dehydrogenase.

Figure 1. Defective BCAA catabolism alters cardiac glucose metabolism

(A) Contractile function of isolated perfused hearts from KO and WT mice measured as rate pressure product (the product of left ventricular developed pressure and heart rate) (n=15 per group). (B) Phosphocreatine to ATP ratio (PCr/ATP) assessed by ³¹P NMR spectroscopy in isolated perfused hearts (n=15 per group). (C) Relative contribution of glucose, fatty acids and other substrates (lactate, endogenous) to the TCA cycle in hearts perfused with ¹³C-labeled substrates without (n=6–7 per group) or with 0.429mM BCAAs (+BCAAs: n=13 per group). (D) The BCAA levels in cardiac tissue extracts measured by GC-MS (n=4–5 per group). (E) Glycogen content in WT and KO hearts (n=7–15 per group. (F) Enrichment of ¹³C3-glycogen normalized by ¹³C4-Glutamate in perfused hearts (n=5–6 No BCAAs, and n=10–11 +BCAAs). (G) Representative ³¹P NMR spectra showing time-dependent accumulation of 2-DG-P in the WT and KO hearts. 2-DG-P, 2-deoxyglucose-6-phosphate. PCr, phosphocreatine. Pi, inorganic phosphate. γ, α and β represent the 3 phosphates on adenosine in ATP. (H) Average time-dependent accumulation of 2-DG-P in KO and WT hearts (n=4–6 per group). (I) The 2-DG uptake rate in hearts perfused by insulin-free and insulin-containing perfusion buffer with or without 0.429 mM BCAAs (n=4–6 per group). Data are shown as mean ± SEM. *P<0.05 vs. WT, ^#P<0.05 vs. WT+BCAAs. See also Figure S1.

Figure 2. Mitochondrial pyruvate utilization is inhibited in the KO heart

(A) Representative electron microscopy images of WT and KO heart sections. Scale bar: 3 μm. (B) Fold differences in mitochondrial number. Mitochondrial number was counted in a total of 10 images per heart (45μm² per image at x12000 magnification, n=3 per group). Data were expressed as fold changes relative to WT. (C) Mitochondrial respiration, after inhibition of complex I by rotenone (1μM), was measured with sequential addition of succinate (5 mM, V₀), ADP (150 μM, V_ADP), Oligomycin (1 mg/ml, V_Oligo) and FCCP (1 μM, V_FCCP) (n=9 per group). (D) Representative oxygraph trace of WT and KO mitochondria after adding substrates and inhibitors as indicated. (E) State 3 respiration of WT and KO mitochondria stimulated with ADP (2.5 mM) in the presence of pyruvate/malate (5 mM/2 mM), glutamate/malate (10 mM/2 mM), or palmitoylcarnitine/malate (50 μM/2 mM) (n=4–9 per group). (F) The total complex I activity and rotenone-sensitive activity in isolated mitochondria from WT and KO hearts (n=4–5 per group). (G and H) Representative oxygraph traces of State 3 respiration in WT (G) and KO (H) mitochondria supported by pyruvate/malate (upper) or succinate (lower) without or with 0.429mM BCAAs. (I) The effect of BCAAs on pyruvate/malate-supported respiration in WT and KO mitochondria (n=8–9 per group). Data are shown as mean ± SEM. *P<0.05 for indicated comparisons, N.S.: no significant difference. See also Figure S2.

Figure 3. BCAA metabolism regulates PDH activity

(A) The PDH activity in the WT and KO hearts (n=5 per group). (B) Relative PDH flux (V_PDH/V_TCA) of the WT and KO hearts perfused by ¹³C-labelled substrate with or without 0.429 mM BCAAs (n=5–11 per group). (C) The activity of purified porcine PDH complex measured by the rate of NADH generation in the presence of increasing concentrations of BCAAs (0, 0.107 mM, 0.429 mM, 0.858 mM, 2.145 mM) (n=3 per group). (D) The average PDH activity at each [BCAAs] measured as nmol of NADH generated per minute at pH 7.5 at 37 °C (mU, n=3 per group). (E) The effect of Ca²⁺ on PDH activity of tissue lysates from WT hearts in the presence of indicated concentrations of BCAAs (n=3–4 per group). (F and G) Representative blots and quantitation of O-GlcNAc modified PDH subunits in WT and KO hearts (n=3 per group). Purified porcine PDH complex was used as positive control and PDH E1α subunit was blotted as input. (H) PDH activity of tissue lysates from WT and KO hearts incubated with increasing concentration of UDP-GlcNAc (n=3–4 per group). (I) The effect of UDP-GlcNAc on PDH activity of tissue lysates from WT hearts in the presence of indicated concentrations of BCAAs (n=3–4 per group). Data are shown as mean ± SEM. *P<0.05 vs. WT, #P<0.05 vs No BCAAs. See also Figure S3.

Figure 4. Protein O-GlcNAc modification in KO

(A and B) Western blot analysis of O-GlcNAc modified protein in hearts from WT and KO mice (n=7–10 per group). (C and D) Protein levels of HKII, GFAT1, GFAT2, OGT, OGA, GALE in hearts from WT and KO mice (n=4–9 per group). (E–G) Representative blots and a quantitation graph of O-GlcNAc modified protein in isolated adult rat cardiomyocytes treated with increasing concentration of BCAAs or BCKAs (n=3 per group). Data are shown as mean ± SEM. *P<0.05 vs. WT, #P<0.05 vs No BCAAs. See also Figure S4.

Figure 5. Cardiac function and responses to ischemia-reperfusion injury in KO hearts

(A–C) Echocardiographic data depicting fractional shortening (A), left ventricular-end diastolic posterior wall thickness (LVPW;d, B) and internal dimension (LVID;d, C) in WT and KO mice at 2 and 6 months (n=10–15 per group). (D) Heart weight normalized to tibia length (HW/TL) of WT and KO mice at 2 and 12 months (n=8–11 per group). (E) Atrial natriuretic peptide (ANP) and brain natriuretic peptide (BNP) mRNA values normalized to 18s and reported as fold changes over WT (n=4–5 per group). (F–H) Left ventricular end-diastolic pressure (LVEDP, F), left ventricular developed pressure (LVDevP, G) and rate pressure product (H) of isolated perfused hearts subjected to low flow ischemia (1% of baseline coronary flow) and reperfusion (n=7–9 per group). (I–K) Concentrations of ATP (I), phosphocreatine (PCr, J), and inorganic phosphate (Pi, K) measured by ³¹P NMR spectroscopy in isolated perfused hearts (n=7–8 per group). Data are expressed as mean ± SEM. *P<0.05 vs. WT by repeated measures ANOVA. See also Figure S5.

Figure 6. Excessive BCAAs worsen I/R injury which is rescued by promoting BCAA catabolism

(A) The time course of serum BCAA concentration after a single bolus of BCAA supplementation. Serum of WT mice was obtained before (BSL, n=11) and at different time intervals (30 min, 1 h, 1.5 h, 2 h, 3 h, 6 h and 12 h; n=4–7 for each time point) after a single bolus of BCAA gavage (1.5 mg/g body weight). (B) Representative photographs of Evans blue and TTC double-stained heart sections from mice treated with vehicle, BCAAs, BT2 and BCAAs+BT2. Blue, unaffected, viable tissue; white, infarct area; red+white, area at risk. Scale bar, 1 mm. (C and D) Infarct size (IS) relative to area at risk (AAR) and AAR relative to left ventricle (LV) were quantified. (E–G) Western blot analysis of BCKDH phosphorylation after 7 days gavage (n=6 per group). (H) Representative photographs of Evans blue and TTC double-stained heart sections from WT and KO mice after supplementation of BCAAs or BT2 for 7 days (n=4–5 per group). (I and J) Infarct size (IS) relative to area at risk (AAR) and AAR relative to left ventricle (LV) were quantified. Data are shown as mean ± SEM. *P<0.05 vs. BSL or for indicated comparisons. See also Figure S5.

Figure 7. Exacerbated I/R injury in the KO heart can be rescued by enhancing glucose metabolism

(A and B) LVEDP (A) and rate pressure product (B) of isolated perfused hearts subjected to low flow ischemia and reperfusion (n=5–9 per group). See legends of Figure 5 for details. (C and D) Concentrations of ATP (C) and PCr (D) measured by ³¹P NMR spectroscopy in isolated perfused hearts at baseline and the end of reperfusion (n=5–8 per group). (E) Representative photographs of Evans blue and TTC double-stained heart sections from WT, KO, KO/TG and TG hearts. Blue, unaffected, viable tissue; white, infarct area; red+white, area at risk. Scale bar, 1 mm. (F and G) Infarct size (IS) relative to area at risk (AAR) and AAR relative to left ventricle (LV) were quantified. (H) Glucose uptake in TG and KO/TG hearts perfused with buffer containing 50mU/L insulin (n=2–6 per group). (I) Relative contribution of substrates to the TCA cycle in KO, KO/TG and TG hearts (n=4–6 per group). The solid line represents the relative contribution of fatty acids and the dash line represents the relative contribution of glucose in WT heart. (J and K) Western blot analysis of protein O-GlcNAc modification after overexpression of GLUT1 (n=4–10 per group). (L and M) Western blot analysis of O-GlcNAc modified protein in hearts after low flow ischemia and reperfusion (n=3–4 per group). (N and O) Representative blots and quantitation of enzymes for the hexosamine biosynthesis pathway (HBP) in WT, KO, KO/TG and TG hearts (n=3–7 per group). (P) Metabolites of the HBP in WT, KO, KO/TG and TG hearts (n=4–6 per group). Data are shown as mean ± SEM. *P<0.05 for indicated comparisons, #P<0.05 vs KO. See also Figure S6.