BCAA catabolism targeted therapy for heart failure with preserved ejection fraction

Abstract

Rationale: Heart failure with preserved ejection fraction (HFpEF) is a major unmet medical need with limited effective treatments. A significant contributing factor to HFpEF, a multifactorial disease, is underlying metabolic dysfunction. While much of the prior research has been on glucose and fatty acid metabolic defects in the pathogenesis of HFpEF, other metabolic activities remain under investigated. Methods: System-based metabolomics and targeted mass spectrometry were employed to analyze serum and tissue samples from a deep-phenotyped human HFpEF cohort. A preclinical mouse model of HFpEF was developed by combined administration of a high-fat diet (HFD) and the nitric oxide (NO) synthase inhibitor N[w]-nitro-l-arginine methyl ester (L-NAME). The branched-chain amino acid (BCAA) catabolic activities were enhanced by genetic inactivation of branched-chain ketoacid-dehydrogenase kinase (BCKDK) or treatment with BT2 (3,6-dichlorobenzo[b]thiophene-2-carboxylic acid), a highly selective inhibitor of BCKDK. Cardiac function, myocardial remodeling and insulin signaling in the left ventricle were assessed across all experimental cohorts. Results: The systems-based metabolomics analysis of the deep-phenotyped HFpEF and non-HFpEF patients revealed that abnormal circulating BCAA levels were significantly associated with adverse outcomes. In the rodent model of HFpEF, significant impairment of BCAA catabolic activities in the heart and abnormal circulating BCAA levels were also observed. In adult mice, inducible knockout of BCKDK, the rate-limiting negative regulator of BCAA catabolic flux, markedly augmented BCAA catabolic activities. Compared with the controls, BCKDK inactivation blunted diastolic dysfunction, cardiac hypertrophy and myocardial remodeling in response to chronic treatment with HFD/L-NAME. This functional amelioration was associated with improved insulin signaling in the myocardium and reduced S-nitrosylation of cardiac proteins, without any impact on systemic blood pressure. Finally, pharmacological inhibition of BCKDK in HFpEF mice significantly reversed the diastolic dysfunction and cardiac hypertrophy associated with HFpEF. Conclusions: Our study provides the first proof-of-concept evidence that global catabolic impairment of BCAAs is an important pathogenic contributor and metabolic signature of HFpEF and restoring BCAA catabolic flux could be an efficacious therapeutic strategy for HFpEF.

Keywords: BCAA metabolism; BT2; HFpEF; heart failure; insulin resistance.

Figure 1

BCAA is significantly increased in human with HFpEF. A, Metabolic analysis results comparing HFpEF vs Control metabolite enrichments in KEGG pathways. X axis represents fold enrichment in HFpEF samples compared to control samples. Color reflects the p value. B, WGCNA dendrogram of 707 varying metabolites in our study population. C, P values showing the association of a subset of WGCNA modules to clinical traits. The yellow module shows significant enrichment for a number of traits. D, Major metabolites present within the yellow module. Abbreviation: Hoafib: Atrial fibrillation; BMI: Body Mass Index; BNP: Brain natriuretic peptide; BSA: Blood Serum Albumin; BUN: Blood urea nitrogen; CVD death: Cardiovascular death; DBP: Diastolic blood pressure; GRF: Glomerular filtration rate; Hgb: Hemoglobin; Hockd: Chronic kidney disease; Hodyslipid: Dyslipidemia; Hodm: Diabetes mellitus; Hohtn: Hypertention; HCO3: HCO-3; Plt: Platelet; SBP: Systolic blood pressure; K: Kalium; Na: Sodium; RDW: Red cell volume Distribution Width; WBC: White blood cell.

Figure 2

BCAA is significantly increased in HFpEF mice. A, Schematic view of experimental design. C57BL/6 mice were fed with HFD+L-NAME for 20 weeks with cardiac function and metabolic analyses conducted at indicated time points. B, Representative echocardiogram images (M-Mode) in control and mice fed with HFD+L-NAME for 14 weeks. C, Echocardiography analysis of left ventricle ejection fraction. n = 6-17. D, Mitral valve E/e' ratio. n = 6-17. E, Ratio of heart weight to tibia length (HW/TL). n = 5-9. F, Ratio of lung weight to tibia length (Lung/Tibia). n = 5-9. G, Treadmill analysis of running distance between the control and HFpEF mice. n = 5-9. H, Glucose tolerance test in Control and HFpEF groups. n = 5-9. I, Circulating BCAA levels at 20 weeks post HFpEF insult in Control and HFpEF mice. n = 5-9. J, Cardiac BCAA levels at 20 weeks post HFpEF insult. n = 5-9. K, Circulating BCKA levels at 20 weeks post HFpEF insult. n = 5-9. L-M, Circulating 3-HIB (L) and 3-aminoisobutyric acid (M) levels at 20 weeks post HFpEF insult. n = 5-9. N-O, Representative immunoblot (N) and quantification (O) of heart p-BCKDHA (Ser 293), BCKDHA and GAPDH protein. n = 5 per group. Two-way ANOVA followed by Turkey's test was used in C-D and One-way ANOVA followed by Turkey's test was used in I, J, K and L. Student t-test was used in E, F, G, M and O. Repetitive t-test was used in H.

Figure 3

Global BCAA catabolism activation prevents cardiac dysfunction in Two-Hit HFpEF model. A, Schematic view of experimental design. BCKDK^flox/flox and BCKDK-uKO mice were fed with tamoxifen diet for 2 weeks before treated with HFD+L-NAME for 14 weeks. Cardiac function and metabolic assessment were performed at indicated time points. B, Western blot analysis of p-BCKDHA (Ser 293), BCKDHA, BCKDK and GAPDH protein levels in BCKDK^flox/flox and BCKDK-uKO mice. n = 5. C-D, Serum BCAA (C) and BCKA (D) levels in BCKDK^flox/flox and BCKDK-uKO mice. n = 5. E, Cardiac BCAA levels in BCKDK^flox/flox and BCKDK-uKO mice. n = 5. F-I, Echocardiography analysis of left ventricle ejection fraction (F), fraction shortening (G), IVRT (H) and mitral valve E/e' ratio (I) in BCKDK^flox/flox and BCKDK-uKO mice. n = 5. J, Body weight of BCKDK^flox/flox and BCKDK-uKO mice. n = 5. K, Ratio of heart weight to tibia length (HW/TL) in BCKDK^flox/flox and BCKDK-uKO mice. n = 5. L, H&E staining for cardiac section in BCKDK^flox/flox and BCKDK-uKO mice post HFpEF insult. Scale bars, 600μm (H&E). M-N, Real-Time PCR analysis of mRNA expression of Col1a1 (M) and Col3a1 (N) in BCKDK^flox/flox and BCKDK-uKO mice post HFpEF stimulation. n = 5. O, Masson's trichrome staining for left ventricle section in BCKDK^flox/flox and BCKDK-uKO mice post HFpEF insult. Scale bars, 50μm (Masson). P-Q, Glucose tolerance test (P) and area under curve (Q) in BCKDK^flox/flox and BCKDK-uKO mice post HFpEF insult. n = 5. One-way ANOVA followed by Turkey's test was used for C-I. Student t-test was used in J-K, M-N, Q. Repetitive t-test was used for P.

Figure 4

Global BCAA catabolism activation alters insulin signaling pathway in heart. A-B, Western blot analysis (A) and quantification (B) of p-AKT (Ser 473), AKT and GAPDH in left ventricle tissues from different groups: Control without HFpEF insults, BCKDK^flox/flox and BCKDK-uKO following HFpEF insults, under insulin or PBS (control) treatment. n = 3. C-E, Western blot analysis (C) and quantification (D-E) of p-P70S6K, total P70S6K, p-AMPKα, total AMPKα and their corresponding GAPDH as loading control in BCKDK^flox/flox and BCKDK-uKO mice after HFpEF insults. n = 5. F-G, Western blot analysis (F) and quantification (G) of s-nitrosylation status in BCKDK^flox/flox and BCKDK-uKO mice following HFpEF insults. n = 5. One-way ANOVA followed by Tukey's test was used for B, D and E. Student t-test was used for G.

Figure 5

Targeting BCKDK to accelerate BCAA catabolism has no impact on blood pressure. A, Schematic view of experimental design. BCKDK^flox/flox and BCKDK-uKO mice were fed with tamoxifen and treated with BT2 or vehicle, and blood pressure was measured using telemetry, Average daily pressure over course of telemetry experiment from light cycle (7:00-19:00) and dark cycle (19:00-7:00). B-C, Mean arterial pressure measured via telemetry in BCKDK^flox/flox and BCKDK-uKO mice before (B) and after (C) tamoxifen treatment. n = 4-8. D, Mean arterial pressure measured via telemetry in BCKDK^flox/flox and BCKDK-uKO mice before and after BT2 treatment. n = 4-8. E-F, Pulse pressure measured via telemetry in BCKDK^flox/flox and BCKDK-uKO mice before (E) and after (F) tamoxifen treatment. n = 4-8. G, Pulse pressure measured via telemetry in BCKDK^flox/flox and BCKDK-uKO mice before and after BT2 treatment. n = 4-8. H-I, Heart rate measured via telemetry in BCKDK^flox/flox and BCKDK-uKO mice before (H) and after (I) tamoxifen treatment. n = 4-8. J, Heart rate measured via telemetry in BCKDK^flox/flox and BCKDK-uKO mice before and after BT2 treatment. n = 4-8. BT2 was administered via oral gavage once a day at 18:00. Two-way ANOVA followed by Turkey's test was used for B-J.

Figure 6

Activation BCAA catabolism using BT2 accelerates BCAA catabolism. A, Schematic view of experimental design. C57BL/6 mice were fed with HFD+L-NAME for 10 weeks before treated with BT2 or vehicle for 4 weeks, cardiac and metabolic analysis was performed as indicated. B-C, Representative immunoblot (B) and quantification (C) of heart p-BCKDHA (Ser 293), total BCKDHA and GAPDH protein levels in vehicle and BT2 treated HFpEF groups. n = 5. D, Circulating BCKA levels in BT2 or vehicle treated HFpEF groups. n = 4-6. E, Cardiac BCKA levels in BT2 or vehicle treated HFpEF groups. n = 4-6. F, Body weight of BT2 or vehicle treated HFpEF groups. n = 8-9. G, Ratio of heart weight to tibia length (HW/TL). n = 8-9. H-I, Glucose tolerance test (H) and area under curve (I) in BT2 or vehicle treated HFpEF groups. n = 8-9. One-way ANOVA followed by Turkey's test was used for C-E. Student t-test was used for F-G and I. Repetitive t-test was used for H.

Figure 7

Pharmacological activation of BCAA catabolism prevents cardiac dysfunction. A, Schematic view of experimental design. C57BL/6 mice were fed with HFD+L-NAME for 10 weeks before treated with BT2 or vehicle for 4 weeks, cardiac and metabolic analysis was performed as indicated. B, Representative echocardiogram images of mitral valve E/A and doppler e'/a' in HFpEF mice after 4 weeks of vehicle or BT2 treatment. C-E, Echocardiography analysis of LVEF (C), Isovolumic relaxation time IVRT (D) and mitral valve E/e' ratio (E) at baseline and after 4 weeks of vehicle/BT2 treatment. n = 8-9. F, Representative Pressure Volume Loops in HFpEF mice post vehicle or BT2 treatment for 4 weeks. G, Catheter analysis of relaxation time constant (Tau) in HFpEF mice after 4 weeks of vehicle and BT2 treatment. n = 8-9. H, Catheter analysis of dp/dt max in HFpEF mice after 4 weeks of vehicle and BT2 treatment. n = 8-9. I, Catheter analysis of dp/dt min in HFpEF mice after 4 weeks of vehicle and BT2 treatment. n = 8-9. J, Catheter analysis of cardiac output (CO) in HFpEF mice after 4 weeks of BT2 treatment. n = 8-9. Two-way ANOVA followed by Turkey's test was used for C-E. Student t-test was used for G-J.