Preview

PDF, English Download (6MB) | Terms of use

Abstract

Heart failure (HF) represents an immense health burden with currently no curative therapeutic strategies. Study of HF patient heterogeneity has led to the recognition of HF with preserved (HFpEF) and reduced ejection fraction (HFrEF) as distinct syndromes regarding molecular characteristics and clinical presentation. Until the recent past, HFrEF represented the focus of research, reflected in the development of a number of therapeutic strategies. However, the pathophysiological concepts applicable to HFrEF may not be necessarily applicable to HFpEF. HF induces a series of ventricular modeling processes that involve, among others, hallmarks of hypertrophy, fibrosis, inflammation, all of which can be observed to some extent in HFpEF and HFrEF. Thus, by direct comparative analysis between HFpEF and HFrEF, distinctive features can be uncovered, possibly leading to improved pathophysiological understanding and opportunities for therapeutic intervention. Moreover, recent advances in biotechnologies, animal models, and digital infrastructure have enabled large-scale collection of molecular and clinical data, making it possible to conduct a bioinformatic comparative analysis of HFpEF and HFrEF. Here, I first evaluated the field of HF transcriptome research by revisiting published studies and data sets to provide a consensus gene expression reference. I discussed the patient clientele that was captured, revealing that HFpEF patients were not represented. Thus, I applied alternative approaches to study HFpEF. I utilized a mouse surrogate model of HFpEF and analyzed single cell transcriptomics to gain insights into the interstitial tissue remodeling. I contrasted this analysis by comparison of fibroblast activation patterns found in mouse models resembling HFrEF. The human reference was used to further demonstrate similarities between models and patients and a novel possible biomarker for HFpEF was introduced. Mouse models only capture selected aspects of HFpEF but largely fail to imitate the complex multi-factor and multi-organ syndrome present in humans. To account for this complexity, I performed a top-down analysis in HF patients by analyzing phenome-wide comorbidity patterns. I derived clinical insights by contrasting HFpEF and HFrEF patients and their comorbidity profiles. These profiles were then used to predict associated genetic profiles, which could be also recovered in the HFpEF mouse model, providing hypotheses about the molecular links of comorbidity profiles. My work provided novel insights into HFpEF and HFrEF syndromes and exemplified an interdisciplinary bioinformatic approach for a comparative analysis of both syndromes using different data modalities.