Fig. 2: AZD1656 treatment improves cardiac metabolism in dbCM.
a, PCA plot of 1H NMR metabolomic profiling. b, Fold change db/db versus control 1H NMR metabolomic spectroscopy profiling (n = 6 per group). Metabolite fold change versus control (control = 1) with propagated error of mean by t-test; two tailed P < 0.05. c, Fold change AZD-treated db/db versus control 1H NMR metabolomic spectroscopy profiling. n = 6 per group metabolite fold change versus control (control = 1) with propagated error of mean by t-test; two tailed P < 0.05. PCr, phosphocreatine. d, Cardionet metabolic flux balance analysis based on the 1H NMR metabolomic profiling and plasma metabolomic analysis (n = 6 per group). Statistical analysis for Cardionet is described in the Methods. e, Visualization of CardioNet metabolic flux predictions using Cytoscape. Metabolites and reactions are depicted as nodes and lines, respectively. The line thickness corresponds to the calculate flux rate. Statistical analysis for Cardionet is described in the Methods. f, Representative annotated spectrum from in vivo 13C dynamic nuclear polarization magnetic resonance spectroscopy cardiac metabolic flux assessment. g, Prediction of oxygen consumption rates based on CardioNet simulations (control n = 5; db/db n = 4; AZD n = 5) by one-way ANOVA (control versus db/db P = 0.0001, control versus AZD P = 0.174, db/db versus AZD P = 0.0001). h,i, In vivo 13C DNP MRS measurement during tricarboxylic acid cycle flux (control n = 12, db/db n = 16, AZD n = 10). Multiple group comparison by one-way ANOVA (lactate/pyruvate: control versus db/db P = 0.399; control versus AZD P = 0.99; db/db versus AZD P = 0.99 (h); bicarbonate/pyruvate control versus db/db P = 0.0001, control versus AZD P = 0.99, db/db versus AZD P = 0.0147 (i)). n represents biological replicates (individual animals). Data are presented as mean ± s.e.m.
