| dc.description.abstracteng | Heart failure (HF) is the most common cardiovascular disease (CVD) and, despite significant pharmacological
progress, it remains the leading cause of morbidity and mortality globally. One of the hallmarks of HF
progression is transcriptional reprogramming, which involves the reactivation of the fetal gene program,
including the Wnt signaling pathway in cardiomyocytes (CMs). This pathway has been linked to the regulation
of genes involved in heart development, tissue regeneration, and disease. The relevance of Wnt activation in
hypertrophic remodeling, which triggers (i) a cell-autonomous CM reprogramming and (ii) a cell-
nonautonomous effect, was previously described. In this study, I aim to characterize the CM-nonautonomous
mechanisms in cardiac remodeling by investigating gene regulatory networks (GRNs) and cell interactions
based on transcriptional regulatory networks at single-cell resolution. Therefore, single-cell transcriptional
analysis of the heart in two disease models was conducted. This includes a previously described inducible CM-
specific β-catenin stabilization (β-catΔex3, gain-of-function (GOF)) and an experimental model of HF
development induced by trans-aortic constriction (TAC) in mice at different stages of hypertrophic remodeling
(compensated and failing stages).
Firstly, deep bioinformatic analysis of single CM transcriptomic of β-catΔex3 and control hearts confirmed
previous findings of cell cycling activation and pointed at an involvement of extracellular vesicles (EVs)
related processes. Next, a protocol for isolation and characterization of EVs from the whole heart was
established. A proteomic analysis of this in vivo cardiac derived EVs from β-catΔex3 hearts has identified
differentially enriched proteins involving 20 S proteasome constitutes, protein quality control (PQC),
chaperones, and associated cardiac proteins including -Cristallin B (CRYAB) and sarcomeric components.
An experimental hypertrophic model confirms that CMs reacted with an acute early transcriptional
upregulation of exosome biogenesis processes and chaperone transcripts. Finally, human induced pluripotent
stem cells-derived CMs (iPSC-CMs) subjected to pharmacological Wnt activation recapitulated the increased
expression of exosomal markers and PQC signaling, suggesting conservation between species. These findings
reveal that the secretion of EVs with a proteostasis signature contributes to the early pathophysiological
adaptation of CMs.
In order to define stage-specific mechanisms, a set of data from an early pathological remodeling and late
failing stage with their respective controls was added. Bioinformatic analysis revealed five subpopulations of
CMs in this data set, each subpopulation belonging to distinct conditions. The CM clusters predominantly
present in the diseased heart were enriched in genes categorizing to Acta1, Nppa, Nppb, Ankrd1, and Myh7,
which are characteristics of stress. The CM clusters predominantly present in the healthy control hearts were
enriched in genes classified into metabolism, energy consumption processes, and heart contraction.
Specifically, a CM subpopulation that was distinct from the compensatory stage showed gene enrichment
related to processes involved in vesicle production, further reinforcing the previous finding. Another
characteristic of the compensatory stage was an increase in the immune cell (IC) population. Accordingly,
analysis of cell-cell interactions showed that the IC population was significantly more active in sending and
receiving signals in the compensatory stage of the disease. To identify cell-specific master regulators of
phenotypic changes during cardiac remodeling, the transcription factor (TF) activity with major roles in this
process was inferred using single-cell sequencing data. This analysis showed that the activity of TFs associated
with the M2 phenotype was significantly changed in the compensatory stage in the IC population. In line with
this finding, these cells exhibited higher expression of genes characteristic of the M2 anti-inflammatory
phenotype. Next, the iPSC-CM model with Wnt activation, mimicking EV production upon CM stress, was
used to test the hypothesis that CMs may communicate with ICs via EVs. A system for tracking EV transport
containing fluorescent proteins was used to visualize exosome uptake or intake, which was introduced into
iPSC-CMs with and without Wnt activation by lentiviral particles. EVs derived from these cells were used to
treat THP1-derived macrophages. THP1 cells are a human leukemia monocytic cell line extensively used to
study monocyte/macrophage functions. Live imaging and molecular analysis showed an active uptake of EVs
into the THP1-derived macrophages and an increase of an M2 phenotype, similar to the change observed in
vivo upon compensatory remodeling.
Additionally, to define cell-specific master regulators of phenotypic changes of CMs for future studies, the TF
inference was run on the single CM data set. This approach identified Krueppel-like factor 15 (KLF15), a
previously described gene in pathological remodeling, as the most critical TF in the progression to hypertrophic
remodeling, which was normalized using a CRISPR/dCas9 approach in vivo in the frame of another study.
9
This resulted in transcriptional normalization of the CM targeted with KLF15 enhancement, which validated
the bioinformatic approach.
In summary, this project harnessed fundamental biology and bioinformatics expertise to better understand the
changes driven by major pathological pathways during HF progression, focusing on CM phenotypic changes
and their interaction with other cells. Specifically, this study (I) revealed a novel adaptive mechanism mediated
by EVs in hypertrophic remodeling, (II) established methods for investigating EVs derived from tissue in vivo,
and (III) established a pipeline for the identification of key TF activity at the single-cell resolution. | de |