Uncovering the molecular and cellular basis of cardiac recovery of the failing heart after mechanical unloading

Abstract Background Heart transplantation remains the only long-term therapy for patients with end-stage heart failure (HF). Nevertheless, donor heart availability is limited, making the need for alternative therapies urgent. Mechanical unloading via left ventricular assist device (LVAD) induces cardiac recovery in a small proportion of patients. The percentage of recovery increases when a bridge-to-recovery strategy is applied a priori. Prospectively identifying the patients most likely to benefit is however challenging due to the lack of efficient predictors of LVAD-associated recovery. Through bulk and single cell transcriptomics, cell-specific recovery signatures are beginning to emerge. However, the heterogeneity of the patient population in these studies precludes deeper mechanistic insights. Purpose We aim to identify the cell types, their gene expression changes and interactions involved in reverse remodelling after mechanical unloading of the failing heart. Methods Myocardial infarction (MI) was induced by acute coronary ligation in sheep. After 2 months HF development, animals were randomized to sham procedure (HF) or mechanical unloading via a temporary LVAD (Impella 5.5) implanted for 3 months (Unloaded). Controls included: 1. healthy sheep that had undergone surgery and similar housing conditions (Ctrl Sham) and, 2. healthy sheep (Baseline) not subjected to any intervention or prolonged housing. Cardiac function was assessed by MRI before and after HF and post mechanical unloading. Single nucleus (sn)RNA-Seq (Parse Biosciences) was conducted on nuclei isolated from the RZ of the myocardium (n=4-7 per group). Cardiomyocyte (CM) hypertrophy was assessed by measurement of cross-sectional area in WGA-stained sections. Relevance of the snRNA-Seq findings to human was assessed via analysis of published snRNA-Seq of pre/post LVAD patients (GSE226314). STAT3 post-translational modifications were studied via immunofluorescence analysis. Results Mechanical unloaded hearts displayed significantly reduced CM hypertrophy. We identified transcriptomic changes in each of the major cardiac cell clusters as well as specific gene markers in CMs sub-clusters. Pseudobulk of the main cell types indicated that the unloaded cells develop a distinct transcriptomic profile compared to the healthy cells. STAT3 pathway was upregulated in CMs from unloaded LV. We confirmed our findings via leveraging published snRNA-Seq database of pre/post LVAD patients. Conclusion Mechanical unloading results in structural and transcriptional reprogramming of the RZ leading to functional recovery. The adaptive remodelling is associated with STAT3 pathway alteration which may have a key role in CMs survival, metabolism and stress-response via both nuclear and mitochondrial pathways. The results achieved by this research will allow to better predict likelihood of recovery and to favour reverse remodelling after unloading.