Diabetic cardiomyopathy is a secondary complication of diabetes with an unclear etiology. Based on a functional genomic evaluation of obesity-associated cardiac gene expression, we previously identified and cloned the gene encoding apolipoprotein O (APOO), which is overexpressed in hearts from diabetic patients. Here, we generated APOO-Tg mice, transgenic mouse lines that expresses physiological levels of human APOO in heart tissue. APOO-Tg mice fed a high-fat diet exhibited depressed ventricular function with reduced fractional shortening and ejection fraction, and myocardial sections from APOO-Tg mice revealed mitochondrial degenerative changes. In vivo fluorescent labeling and subcellular fractionation revealed that APOO localizes with mitochondria. Furthermore, APOO enhanced mitochondrial uncoupling and respiration, both of which were reduced by deletion of the N-terminus and by targeted knockdown of APOO. Consequently, fatty acid metabolism and ROS production were enhanced, leading to increased AMPK phosphorylation and Ppara and Pgc1a expression. Finally, we demonstrated that the APOO-induced cascade of events generates a mitochondrial metabolic sink whereby accumulation of lipotoxic byproducts leads to lipoapoptosis, loss of cardiac cells, and cardiomyopathy, mimicking the diabetic heart-associated metabolic phenotypes. Our data suggest that APOO represents a link between impaired mitochondrial function and cardiomyopathy onset, and targeting APOO-dependent metabolic remodeling has potential as a strategy to adjust heart metabolism and protect the myocardium from impaired contractility.

Background: Diagnostic biomarkers for heart failure (HF) such as the natriuretic peptides (NPs) are widely used but have limitations. Innovative biomarkers could provide improved diagnostic performance. Methods: We launched a prospective case-control proteomic study and investigated for polypeptides specific to HF through a capillary electrophoresis-mass spectrometry (CE-MS) proteomic analysis. The putative biomarker was identified by Orbitrap liquid chromatography-MS, validated by western blot, then by ELISA using plasmas from multicentric international cohorts. A rat model of HF was tested for biomarker expression levels. Results: We identified insulin like growth factor binding protein 2 (IGFBP2) as a new diagnostic biomarker for HF with a high sensitivity and specificity (AUC = 0.93; 95% CI, 0.89-0.96; p b 0.0001) in the local cohort and IGFBP2 levels provided an AUC of 0.943 (95% CI, 0.860-1.026) which gave a 87 % sensitivity in AHF and 90 % specificity at the cut off value previously determined in the discovery cohort, i.e. 556 ng/ml. ROC curve analysis of IGFBP2 and NTproBNP showed an AUC of 0.784 (95% CI, 0.744-0.820) for IGFBP2 and a significantly higher AUC of 0.927 (95% CI, 0.900-0.949) for NT-proBNP, p b 0.0001 in a Dutch cohort. In this cohort, the optimal cut off value for IGFBP2 gave a sensibility of 71% (95% CI, 66-76) and a specificity of 75% (95% CI, 65-83). Conclusion: IGFBP2 is a new biomarker to diagnose HF which could be used to provide additional information to the NPs. Animals models will help in the evaluation of the putative IGFBP2 regulated mechanisms in HF. Clinical Trial Registration: ClinicalTrials.gov NCT01024049.

No abstract