Skip to main content

Biomechanics and signaling in models of congenital heart valve defects

Periodic Reporting for period 2 - EVALVE (Biomechanics and signaling in models of congenital heart valve defects)

Reporting period: 2018-06-01 to 2019-08-31

What is the problem issue being addressed?
Cardiovascular diseases (CVDs) take a huge toll on the world population. An estimated 19 million people died from CVDs in 2010, representing 30% of all global deaths. Abnormal blood circulation is widely recognized as a cardiovascular risk factor, and abundant data point to mechanotransduction as a trigger of common features of pathologies, such as atherosclerosis, cardiomyopathies and valvulopathies. Mechanotransduction is the conversion of a mechanical stimulus into a biological response. It is central to the coordination between mechanical forces generated by flowing blood and heart valve morphogenesis. In the developing heart, the heartbeat and the blood flow signal to endocardial cell (EdC) progenitors through mechanosensitive proteins which in turn modulate the genetic program controlling cardiogenesis. The precise mechanism through which this is achieved, however, is poorly understood due to a lack of biophysical and dynamic information on the forces and signaling cascades activated in these cells in vivo. It is essential to determine how mechanical forces control pathway activation and morphogenesis in vivo because the mechanical stimuli experienced by EdCs are too complex to be faithfully reproduced in vitro. Our study aims to uncover the mechanical cues and the molecular program activated in EdCs during normal and pathological heart valve development. My hypothesis is that endocardial mechanotransduction and mechanical forces are not only key for the morphogenesis of the valve, but also for the maintenance and repair of healthy valves. Specifically I aim to investigate the spatial and temporal nature of endocardial mechanosensitivity necessary for cardiac valve development and physiology in normal and pathological contexts.

Why is it important for society?
In humans, cardiac valve defects account for 20-30% of all congenital cardiovascular malformations, with an incidence of ≈2% in all live births. The origins of the defects are multifactorial but are usually associated with genes important for heart valve development, such as signaling factors (Notch1, TGF beta), and Filamin A (FLNA), an actin interacting protein. Diseased valves in patients carrying FLNA mutations display defects in extracellular matrix (ECM) deposition and the mitral valve specifically degenerates. My work in zebrafish (ZF) shows that notch1 expression is modulated by klf2a, a transcription factor activated by oscillatory wall shear stress (WSS) during valve development. Recently, I found that the mechanosensitive calcium channels TRPV4 and TRPP2 control the expression of klf2a. Since TRPP2 interacts with other stretch-activated channels (Piezos, TREKs) via the actin cytoskeleton and FlnA, these findings may provide a molecular explanation for the role of FlnA in valvulopathies, as well as for the impaired heart valve defects observed in patients with polycystic kidney disease where trpp2 can be mutated.
With these results in mind, we now propose a novel, multidisciplinary approach, combining developmental biology, live imaging, mathematical models, cell biology and whole genome sequencing methodologies, to study the mechanisms underlying mechanical force-EdC interactions in normal and pathological conditions. An important and unique aspect of our approach is to focus on in vivo dynamics of EdCs, in contrast to the extensive body of literature in cultured ECs. It is our expectation that the results will generate a holistic view of this extremely important biological process, and will provide novel alternatives for the treatment of valvulopathy-related pathogeneses.

What are the overall objectives?
The key objectives of this project are:
1. To quantify the mechanical forces associated with heart valve morphogenesis in normal and pathological hearts
2. To identify and characterize EdC mechanosensors
3. To decipher the downstream effect of mechanical sensing in EdCs
The cellular and molecular basis of valve development in response to mechanical forces. Valve morphogenesis is strictly dependent on the presence of mechanical forces and most heart contractility mutants display cardiac valve defects. In ZF, the transformation of EdC progenitors into valve leaflets happens between 48 and 96 hours post-fertilisation (hpf) in the atrioventricular canal (AVC) along with ECM deposition. Flow-induced AVC-specific Klf2a expression controls valve development (see preliminary results) though the specific EdC response at the cellular and molecular level remains unclear.
Endocardial mechanotransduction. In EC culture, shear stress activates different signaling pathways; atheroprotective Klf2 and endothelial nitric oxide synthase (enos) expression and a pro-inflammatory response mediated via beta-catenin and NFkappa B activation. Furthermore, ECs change their shape in response to tension, which is usually associated with deformation of the ECM, suggesting another potential role in dictating growth and cell shape. In fibroblasts, this response is mediated by the YAP signalling pathway . When considering EdCs, our recent work shows that TRPP2 and TRPV4 mutants lead to abnormal valve phenotypes but do not completely lack klf2a expression (50% reduction). Thus other mechanosensitive channels must be involved in the process.
Objective1 ongoing: Predicting endocardial forces associated with the heart beat and linking the effect of forces with specific valve defects in models of cardiomyopathies.
We performed live imaging of transgenic lines labelling cardiac cells and track each cell in three dimensions along with live imaging of the beating heart of zebrafish transgenic lines expressing genetically encoded calcium reporters. We developed a 3D image analysis approach allowing to capture the full three-dimensional process in time by introducing a centerline to characterize each cell with respect to its position in the heart. Part of the 3D approach was developed in and other elements are published. This approach also allows us to extract the blood flow pressure and estimate the induced cell-cell tension in the cardiac wall. The 2D flow velocities measured in task1a have been used as a validation of our 3D model. Elements of this task are published in.

Objective 2 ongoing: Molecular mechanism of endocardial mechanotransduction
A recent study from my laboratory demonstrated that TRPP2 and TRPV4 control klf2a expression and regulate EdC reorganization during embryonic valve development. Yet, these mutants do not fully phenocopy an absence of flow phenotype. Our hypothesis was that other mechanosensitive proteins are involved in the process. To identify them, we analysed the RNA sequencing datasets generated from WT endocardial cells at embryonic stages involving mechanosensitive processes. We could identify many known stretch sensitive channels from different families of channels including Transient Receptor Potential (TRP) and piezo channels. Since their function as mechanosensor in the endocardium is unknown, we generated compound piezo mutants and analyzed heart valve development and specific genes expressed in the valve forming area, including klf2a. We found that piezo1 mutants display valve defects and abnormal klf2a expression in the outflow tract. The next step is to explore potential compensatory mechanisms by other piezo related genes (piezo2a and piezo2b).

Objective 3 ongoing: To decipher the downstream effects of mechanical sensing during heart valve development of normal and pathological hearts
To assess the direct transcriptional targets of Klf2a we performed experiments on FACS-sorted EdCs. The mRNA seq and ATAC seq data have now been analysed. We report two novel transcription factors with a function in heart valve formation in zebrafish: egr1 and klf2b. Genome-wide analysis of gene expression reveals that the endocardial transcriptional programs modulated by klf2a, klf2b, and egr1 mainly contain non-redundant targets. Several of these targets have been implicated in endothelial-to-mesenchymal transition (EMT). Many of the deregulated genes exhibit changes in chromatin accessibility pointing to potential direct effects of these factors. Finally, in vivo phenotypic analyses show that VEGF receptor 1 (flt1) is a target of egr1 and klf2b during early valvulogenesis. These findings suggest that klf2a, klf2b, and egr1 cooperate for the activation of EMT program in response to mechanosensitive inputs. We propose that the combinatorial action of these factors mediates flow mechanotransduction to control the endocardial program, especially for valve development. In addition, the group of Mark Kahn (U. Penn, USA) collaborated with us to demonstrate that wnt9b expression is modulated by klf2a. This study is now published. We thus now established a clearer picture of the mechanosensitive network activated in response to fluid shear stress in endocardial cells.
We proposed to test the cellular behavior activated in EdCs in response to flow during valve development in vivo. So far, we assessed the impact of shear stress patterns and flow directionality on the behavior of endocardial cells, the specialized endothelial cells of the heart and obtained interesting results that are now published.

Further preliminary results on all three aims give us confidence but will require more time.
FIRST and SECOND YEAR and half: Building on our existing experience in imaging, numerical simulations and image processing, we focused on live imaging, analysis and optical tweezing assays to characterise the heart wall and mechanical force distribution in the heart, as well as initially testing force reporters in valve morphogenesis. To start we developed tools centered around cell scale analysis. Organogenesis involves extensive and dynamic changes of tissue shape during development. It is associated with complex morphogenetic events that require enormous tissue plasticity and generate a large variety of transient three-dimensional geometries that are achieved by global tissue responses. Nevertheless, such global responses are driven by tight spatio-temporal regulation of the behaviours of individual cells composing these tissues (MILESTONE 1).

Expected results for THIRD YEAR: We will publish mRNA sequencing datasets and ATAC sequencing on small EdCs quantities and analysis procedures (MILESTONE 2). We will consolidate successful approaches established in the two last years and, where necessary, implement back-up strategies for the previous set of targets (e.g. redesign force reporter constructs, knockin and tweezing strategies). We will initiate the pilot screens on the CRs by morpholino knockdown and the drug screen for Klf2a rescues (MILESTONE 3).
By year 4, we will complete the analysis of the mechanical force distribution and validate the model in the cardiomyopathic heart. The ISH and CR screen will be done. The information and tools will be released to the community as an open resource (MILESTONE 4). The results will be further integrated into the cellular context using live imaging of heart valve development and cell lineage experiments in klf2a, flnA mutants low contractility contexts (MILESTONE 5) and in the context of clonal analysis and regeneration (MILESTONE 6).

FOURTH AND FIFTH YEAR: If necessary, I will reprioritize and alter my research plan in response to my own and others results, but by this stage I aim to focus on fully developing and testing mechanistic hypotheses by gene knockout, overexpression and optical manipulation (MILESTONE 7). In addition, we will be alert to opportunities to use human samples and prioritize drug targets, which appear promising candidates for translational studies for anti-valvulopathy therapeutics (MILESTONE 8).