Abstract
Aims-During embryogenesis, the onset of circulatory blood flow generates a variety of hemodynamic forces which reciprocally induce changes in cardiovascular development and performance. It has been known for some time that these forces can be detected by as yet unknown mechanosensory systems which in turn promote cardiogenic events such as outflow tract and aortic valve development. PIEZO1 is a mechanosensitive ion channel present in endothelial cells where it serves to detect hemodynamic forces making it an ideal candidate to play a role during cardiac development. We sought to determine whether PIEZO1 is required for outflow tract and aortic valve development.
Methods and results-By analysing heart development in zebrafish we have determined that piezo1 is expressed in the developing outflow tract where it serves to detect hemodynamic forces. In particular, we have found that mechanical forces generated during the cardiac cycle activate Piezo1 which triggers nitric oxide to be released in the outflow tract. Consequently, disrupting Piezo1 signalling leads to defective outflow tract and aortic valve development and indicates this gene may be involved in the etiology of congenital heart diseases. Based on these findings, we analysed genomic data generated from a cohort of bicuspid aortic valve patients and identified 3 probands who each harboured a novel variant in PIEZO1. Subsequent in vitro and in vivo assays indicates that these variants behave as dominant negatives leading to an inhibition of normal PIEZO1 mechanosensory activity and defective aortic valve development.
Conclusion-These data indicate that the mechanosensitive ion channel piezo1 is required for OFT and aortic valve development and, furthermore, dominant negative variants of PIEZO1 appear to be associated with BAV in humans.
Introduction
In many species, the onset of circulation precedes the role it will play later in life as an oxygen and nutrient delivery system1. As the primitive heart initiates circulation, the forces generated by the blood flow are detected by mechanosensory systems present in the endothelium which lines both the heart and vasculature2. These extracellular forces are subsequently converted into intracellular signals which can trigger a variety of cellular responses3. Many lines of evidence suggest that the hemodynamic forces generated by the circulatory system act as epigenetic cues to drive developmental processes such as cardiogenesis and valvulogenesis forward4, 5. Although a number of mechanosensory systems with the potential to sense circulatory hemodynamics have been identified, our knowledge of how the endothelium detects mechanical stimuli is far from complete2, 6. PIEZO1 is a mechanosensitive ion channel present in the cell membrane. When the cell membrane is stretched, this opens the channel and allows an influx of cations7. Recently, PIEZO1 has been shown to confer mechanosensitivity to endothelial cells allowing them to detect hemodynamic shear stress and subsequently align themselves in the correct orientation during vasculogenesis8,9. Furthermore, deleterious mutations in PIEZO1 are associated with lymphedema in humans which is caused by defective lymphatic valve development10.
The OFT is a transient structure which is extensively remodeled during development and will give rise to a variety of cardiovascular structures including the aortic and pulmonary valves. Although it is well established that hemodynamic blood flow plays a role in OFT development and the formation of the aortic valves11, the mechanosensors that detect these forces have remained elusive. However, because of its role in sensing shear stress in the vasculature, PIEZO1 makes a promising candidate for detecting similar forces in the OFT.
Here we report that disrupting Piezo1 signalling in zebrafish leads to defective development of the OFT and aortic valves. Based on these findings we have been able to identify 3 independent predicted pathogenic PIEZO1 variants in patients with bicuspid aortic valve disease (BAV). Furthermore, in vitro electrophysiological analysis indicates that all variants are dominant negatives which significantly inhibit wild type PIEZO1 activity after stimulation.
Methods
Zebrafish strains and husbandry
Zebrafish were maintained under standardized conditions and experiments were conducted in accordance with local approval (APAFIS#4054-2016021116464098 v5) and the European Communities council directive 2010/63/EU. Embryos were staged as described 9. The Tg(fli1a:GFP)y1Tg was provided by the CMR[B] Centro de Medicina Regenerativa de Barcelona. The double transgenic line Tg(fli1a:GFP)y1;Tg(cmlc2a:RFP) was generated in house. All larvae were euthanised by administration of excess anaesthetic (Tricaine).
Aortic valve imaging
One day prior imaging, larvae were incubated in 0.2 μM BODIPY-FL Ceramide (Invitrogen D3521) in Embryo medium + PTU (0.003% 1-phenyl-2-thiourea). Larvae were then anesthetized with Tricaine (0.16g/L) and mounted in low melting agarose. Imaging was performed with a Zeiss LSM710 two-photon microscope coupled to a Ti:sapphire laser (Spectra-Physics, Santa Clara, CA, USA) and a water immersion 25× objective.
Electrophysiology
All electrophysiological experiments were performed after 2-6 days of culture for transfected HEK-293T cells seeded at a density of 20 000 cells/35mm dish and after 2-8 hours of culture of freshly dissociated embryonic zebrafish endothelial cells. Dishes were placed on an inverted microscope (DIAPHOT 300, Nikon). Recordings were performed in inside out or cell attached configuration for the patch clamp technique. PIEZO1 currents were elicited by a negative pressure step from 0 to −80mmHg with −10mmHg step increments at −80mV potential. Stimulation protocols and data acquisition were carried out using a PC (Hewlett Packard) with commercial software and hardware (pClamp 10.4) (supplemental information).
Exome sequencing
The exonic sequences were captured with the Agilent Sure Select All Exon v4 kit (Agilent, Santa Clara, CA, USA) and sequencing was performed on an Illumina HiSeq2000 sequencing apparatus (Illumina, San Diego, CA, USA). Raw Exome sequencing data were aligned against the human reference genome hs37decoy5 and duplicated reads were identified using the SNAP alignment tool version 1.0beta23. For detailed information regarding coverage etc, see (Suppl.table.2). GATK 3.4 was used to perform local indel realignment, score base recalibration and variant calling with the Haplotype Caller. Variations were then selected based on quality criteria using the Variant Filtration module from GATK. Variant annotation (ANNOVAR) and prioritization was performed with the VarAFT software (http://varaft.eu). Prioritization of the filtered-in variants was based on expression in aortic valve according to RNA-seq expression data. Patient recruitment was approved by the Comité de Protection des Patients (13.061).
Results
Identifying the zebrafish endothelial PIEZO1 ortholog
Previous research has identified a zebrafish PIEZO1 ortholog that does not appear to be expressed in endothelial cells12. We have subsequently identified a second PIEZO1 ortholog, piezo1b (Pz1b). At 24 hours post fertilization (hpf), we could detect a weak Pz1b signal in the developing heart tube (Suppl.fig.1.A). By 48hpf a strong expression of Pz1b appeared in the AV canal and OFT (Suppl.fig.1.B). By 4 days post fertilization (dpf) we were able to observe a strong expression of Pz1b in the OFT and vasculature (Fig.1.A,B and Suppl.fig.1.C). Furthermore, we were also able to co-localise Pz1b with GFP labelled endothelial cells in the OFT (Fig.1.C-E). These results indicate that Pz1b is expressed in endothelial cells during zebrafish development, similar to its mammalian counterpart9.
We next sought to determine whether Pz1b is a functional mechanosensitive ion channel in the developing zebrafish endothelium. To achieve this, we performed electrophysiological analysis of cultured zebrafish endothelial cells subjected to mechanical stimulation (Fig.1.F,G). Furthermore, we found that knockdown of Pz1b abolishes mechanically induced currents which was not the case when we targeted piezo213 (Fig.1.F,G). As an additional control, we performed the same electrophysiological analysis on WT endothelial cells treated with the PIEZO channel inhibitory peptide GsMTx414 and also observed a reduction in the amplitude of mechanically induced currents (Fig.1.F,G). These data indicate that Pz1b is a functional mechanosensitive ion channel present in the endothelial cells of developing zebrafish embryos.
Piezo1b is involved in cardiovascular development
Knockdown of Pz1b produced a phenotype characterized by defective cardiogenesis and associated oedema at 72hpf (Suppl.Fig.2. A-D’ and Suppl.fig.3.G-H’). Although we ensured the specificity of the observed phenotype to the knockdown of Pz1b by employing two different morpholinos targeting the same gene, we also analysed mRNA splicing, morpholino synergy and generated a CRISPR/Cas9 Pz1b knockout (Suppl.information and suppl.figs.S3 and S4). We next sought to rescue the Pz1b morphant phenotype using mouse piezo1 mRNA (mPz1) which is not targeted by the Pz1b-MO1. In this manner, we found that co-injection of Pz1b-MO1 with 20pg of mPz1 RNA could rescue the cardiac defects observed in Pz1b morphants (Suppl.fig.3.C-F). Analysis of a number of physiological parameters indicates that Pz1b morphants exhibit a decreased cardiac output along with blood regurgitation through the AV canal (Suppl.fig.2.E-I and suppl.information and suppl.fig.S6 and movies S1,2). Taken together these data indicate that loss of Pz1b results in defective cardiac development.
Piezo1b is required for OFT development
To determine whether OFT development has been affected, we analysed this structure in WT and Pz1b morphants. At 48hpf, we could not detect any discernible differences between WT and Pz1b morphant embryos (Suppl.fig.2.J,K). However, by 72hpf, the WT OFT had developed into a “pear shaped” structure which was noticeably wider at the interface with the myocardium when compared to the Pz1b morphant OFT (Suppl.fig.2.L-P). Furthermore we were able to rescue this defect by co-injecting mouse piezo1 RNA (suppl.fig.S3.Q,R).
Because the OFT is directly adjacent to the ventricle, it is most likely subjected to some of the most extreme hemodynamic forces which in turn could be detected by Pz1b to subsequently initiate the further development of this structure. To test this premise we analysed the dynamics of OFT in zebrafish larvae using confocal microscopy. At 48hpf we were able to observe the OFT stretch and relax during the cardiac cycle. At the end of systole, the OFT achieved a maximum peak inside diameter of 16.87μm (+/−0.24 SEM) (Suppl.fig.2.Q-S and movie S3). We also made similar measurements of the dorsal aorta as a comparison and found that its diameter increased by only 0.72μm (+/−0.20 SEM) during systole (Fig.5.S and movie S4). This indicates that indeed hemodynamic forces produce a strong dynamic response in the OFT endothelium in comparison to other regions of the vasculature.
Piezo1b is required for aortic valve development
During cardiac development, the aortic valves will form in the OFT, a process reliant on hemodynamic forces. Based on this notion we analysed the developing aortic valves to determine whether Pz1b is involved in this process. To achieve this we labelled 7dpf larvae with BODIPY and imaged them using 2 photon microscopy, a technique which has previously been employed to analyse the development of the atrioventricular valves in zebrafish15. Our analysis of WT larvae indicates that at this time point two defined leaflets have formed which function effectively to regulate the flow of blood as it is ejected from the ventricle (Fig.2A and Movie.S5). In comparison, the valves in Pz1b morphants are highly dysmorphic and appear to be enlarged and misshapen (Fig.2.B and Movie.S6). Previous research in mammals and humans has determined that NOTCH1 is required for aortic valve development and mutations in this gene are associated with BAV16. In this regard we next analysed the aortic valves in notch1b KO zebrafish larvae. Similar to Pz1b morphants, the aortic valves in notch1b mutants are also highly dysmorphic compared to WT controls (Fig.2.C and Movie.S7). The epidermal growth factor receptor (EGFR) gene has also been implicated in the etiology of BAV17, furthermore previous research in zebrafish has indicated that chemical inhibition of EGFR using either PKI166 or AG1478 results in defective OFT development18. Based on this we treated zebrafish embryos with either inhibitor for 7 days and again assessed aortic valve development. As with Pz1b morphants and notch1b mutants, inhibition of EGFR results in dysmorphic aortic valves (Fig.2.D,E). Lastly, Endothelial Nitric Oxide Synthase (eNOS) has also been linked to BAV in mice17, while treatment of zebrafish embryos with the NOS inhibitor 2-trifluoromethylphenyl imidazole (TRIM) results in defective cardiac development19. Therefore, we also treated zebrafish larvae with TRIM and analysed their aortic valves as before and found that TRIM treated larvae also display defective aortic valves (Fig.2.F). Taken together these data indicate that loss of Pz1b results in defective aortic valve development and, furthermore, the resulting phenotype is reminiscent of the valve phenotype caused by disrupting BAV associated genes/proteins in zebrafish.
Piezo1b regulates nitric oxide production and extracellular matrix composition in the outflow tract
During embryonic zebrafish development, there is a pronounced and sustained release of nitric oxide (NO) in the OFT, coincident with increasing hemodynamic load20, 21. To determine whether Pz1b is involved in NO release, we utilized a DAF-FM DA assay to detect the NO signal produced in the OFT 20, 21. Analysis of control 72hpf embryos indicates a clear NO signal in the developing OFT (Suppl.fig.5.A-B’and 5.I). Conversely, the NO signal was noticeably absent in the OFT of Pz1b morphants (Suppl.fig.5.C-C’and 5.I). We next assessed whether ablation of erythrocytes with phenylhydrazine (PHZ) to reduce shear stress, or reducing the heart rate substantially (108.3bpm +/−2.1SEM, n=10) by treatment with 10mM 2,3-butanedione monoxime (BDM) 22, 23, also had the same effect on NO production in the OFT (Suppl.fig.5.D-E’ and 5.I). Although each of these conditions leads to a significant reduction in NO signal, only complete cessation of the heart beat using 15mM BDM abolishes the NO signal (Suppl.fig.5.F-F’ and 5.I). To further confirm these results we also analysed NO production in troponin t2 (tnnt2) morphants which lack any discernible heartbeat24. Similarly to arresting the heart using BDM treatment, we also found that the loss of heart beat observed in tnnt2 morphants also significantly reduces the production of NO in the OFT (Suppl.fig.5.G-G’ and 5.I). Lastly, we also treated WT 72hpf embryos with the PIEZO1 inhibitor, GsMTx4, for 4 hours and found that this significantly reduced the amount of NO produced in the OFT (Suppl.fig.5.H-H’ and 5.I). This data indicates that hemodynamics forces produced in the OFT induce NO production and that Pz1b plays a role in detecting these forces. Previous research has identified that the extracellular matrix (ECM) component, Elastin, is produced in the OFT where it provides the elasticity to cope with increasing hemodynamic load and to dampen the force of blood as it is ejected from the ventricle25. We performed immunohistochemistry on 72hpf zebrafish larvae using a previously described ElastinB (ElnB) antibody25, and were able to determine that ElnB is produced by cells which ensheathe the developing OFT endothelium (Suppl.fig.5.J-L). Where these cells originate from is at present ambiguous however, both neural crest cells and second heart field progenitors do appear to contribute26, 27. We next assessed whether hemodynamic forces could play a role in triggering ElnB production. To achieve this we performed ISH using an elnB riboprobe on 72hpf zebrafish larvae. Under normal conditions we could detect a clear and specific expression of elnB in the OFT (Suppl.fig.5.M). However, analysis of tnnt2 morphants which lack a heartbeat, and so are devoid of hemodynamic forces, revealed an obvious loss of elnB expression in the OFT (Suppl.fig.5.N). Similarly, in Pz1b morphants, elnB expression is also appreciably reduced (Suppl.fig.5.O) and lastly, in embryos treated with TRIM to inhibit NO production, elnB expression in the OFT is clearly reduced (Suppl.fig.5.P). These data indicate that elnB expression in the OFT is hemodynamically dependent and that both Pz1b and NO are required for this process. Previously we have shown that the ECM component AggrecanA (Acana) is also expressed in the same population of cells as ElnB and in a hemodynamically dependent manner28. Furthermore, are data indicates that reduced AGGRECAN expression is associated with type 0 BAV in humans. Due to this, we next analysed whether Pz1b could regulate the expression of acana. ISH using an acana riboprobe on 72hpf WT zebrafish larvae indicates that acana is strongly expressed in the OFT at this timepoint (Suppl.fig.5.Q). Conversely, when Pz1b is knocked down the expression of acana in the OFT is clearly reduced (Suppl.fig.5.R). Taken together our data highlights a potentially novel signaling pathway initiated by hemodynamic forces in the OFT which are subseqeuntly detected by Pz1b, this, in turn, triggers NO to be released, which subsequently initiates the production of ElnB and AcanA (Suppl.fig.7).
Identification of human variants in PIEZO1 associated with BAV
To determine whether mutations in PIEZO1 are associated with valvulopathies such as BAV, we examined whole exome sequence data generated in-house from 19 patients diagnosed with isolated BAV. In parallel, we also analysed whole exome sequence data from 30 BAV patients provided by the National Heart, Lung, and Blood Institute (NHLBI) Bench to Bassinet Program: The Pediatric Cardiac Genetics Consortium (PCGC) dataset (dbGaP accession phs000571.v3.p2.). In this manner, we were able to identify 3 independent nonsynonymous variants in PIEZO1: p.Tyr2022His (c.6064T>C; p.Y2022H; located in a transmembrane C-terminal domain (in-house analysis)), p.Lys2502Arg (c.7505A>G; p.K2502R; located in the cytoplasmic C-terminal region of PIEZO1 (in-house analysis)) and p.Ser217Leu (c. 650C>T; p.S217L; located in N-terminal region (PCGC analysis)) (Fig.3.A-C). The heterozygous p.Y2022H and p.K2502R variants were subsequently validated by Sanger sequencing (Fig.3.B,C). Although this was not possible for the p.S217L variant, we were able to determine that one of the parents also harboured this mutation indicating that the proband is heterozygous for this variant. According to gnomAD29, all 3 variants are considered rare, with minor allele frequencies <1% (Table.1). By comparing PIEZO1 orthologs in different species it is apparent that Tyr2022, Lys2502 and Ser217 are evolutionary conserved amino acids (Fig.3.F). To determine the potential functional consequences these variants have on PIEZO1, they were analysed using the CADD30, Mutation Taster231 and UMD Predictor programs32 (Table.1). In this manner, all 3 variants are predicted to be pathogenic. Pedigree analysis for all 3 probands can be found in the supplemental information.
Functional analysis of BAV associated variants in PIEZO1
To determine whether the identified variants affected PIEZO1 protein function, we performed mechano-electrophysiology analysis of HEK-293T cells transfected with either wildtype (WT) human PIEZO1 or the p.Y2022H, p.K2502R and p.S217L variants. In this manner we were able to detect changes in current from HEK cells transfected with wildtype human PIEZO1 (Fig.4.A,E). Conversely, cells which had been transfected with either p.Y2022H, p.K2502R or p.S217L PIEZO1 variants showed significantly reduced mechano-stimulated currents (Fig.4.B-E). One explanation for the reduction in PIEZO1 activity associated with these variants is that the mutations result in a reduction of cell surface expression. To assess this possibility we performed immunohistochemistry (IHC) on HEK-293T cells transfected with either wildtype human PIEZO1 the p.Y2022H, p.K2502R or p.S217L PIEZO1 variants. Confocal images analysis of transfected cells indicates that none of the variants appear to significantly affect the localization of PIEZO1 to the cell surface when compared to wildtype PIEZO1 (Fig.4.F). To confirm this observation we measured the fluorescent intensity across the cell membrane. In this manner, wildtype PIEZO1 and all of the variants show a clear peak of intensity at the cell surface which drops sharply on the extracellular and intracellular sides when compared to GFP, which is expressed throughout the cell (Fig.4.G). This shows that aberrant trafficking of PIEZO1 to the cell surface is not the reason for the observed reduction in activity. These data indicate that the variants p.Y2022H, p.K2502R and p.S217L have deleterious effects on PIEZO1 protein function.
p.Y2022H, p.K2502R and p.S217L PIEZO1 variants are dominant negative isoforms
Because all 3 probands are heterozygous for their respective variants, we next sought to ascertain whether the association with BAV was due to haploinsufficiency or to a possible dominant negative effect of these mutations. When wildtype human PIEZO1 was co-transfected with either p.Y2022H, p.K2502R or p.S217L PIEZO1 variants, there was a significant decrease in current amplitude compared to the control (Fig.5.A-E). Taken together these data indicate that the p.Y2022H, p.K2502R and p.S217L PIEZO1 variants act as dominant negatives.
Next we sought to determine whether forced expression of the dominant negative human PIEZO1 variants in the zebrafish endothelium could also affect aortic valve development. To achieve this we generated a transgenic construct using the endothelial specific fliEP promoter33 to drive expression of either WT PIEZO1 or the variants specifically in endothelial cells in vivo. In this manner we were able to determine that expressing any of the dominant negative variants (p.Y2022H, p.K2502R or p.S217L) also had a significant impact on aortic valve development when compared to WT PIEZO1 expression. To assess this in more detail we performed a quantification of the defective valves by dividing the length of either leaflet by its width (Fig.6.A,B). In this manner we found that all the variants significantly reduced the overall length/width ratio in both leaflets when compared to the WT PIEZO1 expressing larvae. This indicates that expression of either Y2022H, p.K2502R or p.S217L PIEZO1 disrupts aortic valve development in zebrafish.
Discussion
Despite the differences in cardiac physiology, early OFT development is highly conserved between mammals and zebrafish, in particular this region will give rise to the aortic valves. Recently, it has been established that the coordinated actions of the mechanosensitive ion channels TRPV4 and TRPP2 are required to promote klf2a expression in the AV canal and subsequently drive valve morphogenesis in this region34. It appears that Pz1b may play a similar mechanosensory role in the OFT where it is required to trigger NO production in response to increasing hemodynamic load. How Pz1b triggers NO production is at present unclear, however previous research has highlighted a feedback mechanism that involves the release of ATP from the endothelium which activates a P2Y2 signalling cascade resulting in NOS activation35. Interestingly, treating zebrafish embryos with the Nitric Oxide Synthase (NOS) inhibitor TRIM also disrupts OFT development and leads to a similar cardiac phenotype36 while mice deficient in Endothelial Nitric Oxide Synthase (eNOS) develop BAV37, 38. Our results are also in line with mammalian data that indicates a direct regulation of eNOS by PIEZO18. NOTCH1 is one of the few genes directly linked to BAV in humans and evidence suggests that NO can also regulate NOTCH1 signaling during aortic valve development. Compound mutant mice which are eNOS−/−;Notch1+/− show a dramatic increase in the prevalence of BAV when compared to either eNOS−/− or Notch1+/− alone39. Our own data indicates that the zebrafish NOTCH1 ortholog notch1b also regulates aortic valve development. Indeed, zebrafish notch1b mutants display dysmorphic valves. Importantly, we also observed a very similar aortic valve phenotype in Pz1b morphant larvae indicating that this gene is also required for aortic valve development. Future investigation will be aimed trying to ascertain whether Pz1b mediated NO release acts synergistically with Notch1b during valve development.
In humans, disruption of ECM components has been linked to a variety of OFT pathologies such as aortic stenosis (AS) and bicuspid aortic valves (BAV) 40–42. Mutations and decreased expression of ELASTIN have been linked to supravalvar aortic stenosis (SVAS), a condition which leads to the narrowing of the aorta adjacent to the aortic valve 41. Furthermore, a chromosomal microdeletion which includes ELASTIN causes Williams Beuren syndrome which manifests with a variety of developmental defects including SVAS and BAV40. Similarly, ELASTIN haploinsufficiency in mice also leads to progressive aortic valve degeneration43. In chick embryos, ELASTIN production is initiated at day 3 by cells which surround the endothelium of the aorta directly adjacent to the myocardium, before spreading throughout the vasculature 44. In zebrafish, Elastin production in the OFT also commences at around 3 days post fertilisation and coincides with cardiogenic events such as coordinated contraction and AV canal development that place increasing hemodynamic loads on the OFT. Here we provide evidence of a novel signalling cascade initiated by hemodynamic forces in the OFT which are detected by Pz1b and ultimately leads to the expression of ECM components, such as Elastin. Future studies will be required to determine whether BAV patients who harbour deleterious PIEZO1 variants have a decreased expression of ECM components. It will also be interesting to determine whether other ECM components such as COLLAGEN are also regulated by a similar mechanism.
We have also identified 3 pathogenic PIEZO1 variants associated with BAV in humans. Although the variants we have identified inhibit PIEZO1 function, for the p.Y2022H mutant, the proband’s twin, presenting with a tricuspid valve, carried the same variant. This could be due to incomplete penetrance linked to modifiers. For example, mutations in NOTCH1 have been linked to CHD, however the BAV phenotype is not fully penetrant16. Furthermore, it has recently been shown that in monozygotic twins, despite the absence of any pathogenic genetic differences between them, only one of the pair developed BAV45. It should be noted that inhibitory variants in PIEZO1 are also associated with a novel form of hereditary lymphedema46, 47. Although neither report indicates the presence of BAV, it is also unclear whether the patients underwent echocardiography to detect this. However, it is interesting to note that in other conditions in which lymphedema is present, such as Turner syndrome, there is an increased prevalence of BAV (28.4%)48. Whether this is also the case for the PIEZO1 form of lymphedema will require more detailed and expansive analysis of this condition.
Funding
This work was supported by INSERM and CNRS. Work in the C.J lab is supported by a grant from the Fondation Leducq. Work in the C.J lab is supported by a grant from Fondation pour la Recherche sur le Cerveau “Espoir en tête 2017”. C.J was supported by an INSERM ATIP AVENIR grant and a Marie Curie CIG (PC 374 IG12-GA-2012-332772). H.M.M is supported by a grant from the Association Française contre les Myopathies (AFM-Telethon). A.F was supported by a Fondation Lefoulon-Delalande postdoctoral fellowship with previous support provided by a Fondation pour la Recherche Médicale (FRM) postdoctoral fellowship. N.N is supported by the LabexICST PhD program. A.F, H.M.M, N.N and C.J are members of the Laboratory of Excellence « Ion Channel Science and Therapeutics » supported by a grant from the ANR. A.P. received a PhD fellowship from the Association Française du Syndrome de Marfan et Apparentés (AFSMa). Work in the G.L lab is supported by a grant from the ANR (ANR 17 CE18 0001 06 AT2R TRAAK). Work in S.Z lab is supported by the INSERM and the Association Française contre les Myopathies (AFM-Telethon). IPAM acknowledges the France-BioImaging infrastructure supported by the French National Research Agency (ANR-10-INBS-04, «Investments for the future»), the Fondation pour la Recherche sur le Cerveau “Espoir en tête 2015”, and Fondation Leducq.
Conflicts of Interests
None declared
Supplemental figure legends
Movie S1. The beating heart of a 3dpf wildtype embryo.
High speed video recording capturing the beating heart of a 3dpf wildtype embryo. 5 heart beats are shown at 120fps then the same 5 heart beats are shown at 40fps. Blood can be seen entering the atrium and passing to the ventricle before ejection.
Movie S2. The beating heart of a 3dpf Pz1b morphant.
High speed video recording capturing the beating heart of a Pz1b morphant. 5 heart beats are shown at 120fps then the same 5 heart beats are shown at 40fps. Following the contraction of the ventricle blood can be seen regurgitating back into the atrium.
Movie S3. OFT dynamism of a 48hpf Tg(fli1a:GFP)y1 embryo.
Resonance laser imaging capturing the dynamic movement of the OFT during systole and diastole. Frames were acquired every 50ms.
Movie S4. Dorsal aorta dynamism of a 48hpf Tg(fli1a:GFP)y1 embryo.
Resonance laser imaging capturing the dynamic movement of the dorsal aorta during systole and diastole. Frames were acquired every 130ms.
Movie S5. Aortic valves in a 7dpf wildtype zebrafish larvae.
2 photon imaging of a BODIPY labelled 7dpf zebrafish larvae reveals the valves dynamically moving during the cardiac cycle.
Movie S6. Aortic valves in a 7dpf Pz1b morphant zebrafish larvae.
2 photon imaging of a BODIPY labelled 7dpf Pz1b morphant zebrafish larvae reveals the valves dynamically moving during the cardiac cycle.
Movie S7. Aortic valves in a 7dpf notch1b mutant zebrafish larvae
2 photon imaging of a BODIPY labelled 7dpf notch1b mutant zebrafish larvae reveals the valves dynamically moving during the cardiac cycle.
Movie S8. Echocardiogram of the aortic valve from the Y2022H proband.
Movie S9. Echocardiogram of the aortic valve from the K2502R proband.
Acknowledgements
We would like to acknowledge Dr Matteo Mangoni and Dr Joel Nargeot for their input and support. We would like to thank Prof Ardem Patapoutian for the kind gift of hPIEZO1-pIRES2-GFP. We would like to thank Dr Emmanuel Bourinet for the kind gift of mPiezo1. We would also like to acknowledge the Montpellier MGX Genomix platform for their input and support. Part of this data was generated by the Pediatric Cardiac Genomics Consortium (PCGC), under the auspices of the National Heart, Lung, and Blood Institute’s Bench to Bassinet Program <http://www.benchtobassinet.org/>. The Pediatric Cardiac Genomics Consortium (PCGC) program is funded by the National Heart, Lung, and Blood Institute, National Institutes of Health, U.S. Department of Health and Human Services through grants U01HL098123, U01HL098147, U01HL098153, U01HL098162, U01HL098163, and U01HL098188. This manuscript was not prepared in collaboration with investigators of the PCGC, has not been reviewed and/or approved by the PCGC, and does not necessarily reflect the opinions of the PCGC investigators or the NHLBI. We thank Anthony Pinot from the in vivo imaging platform IPAM-Biocampus Montpellier.
Footnotes
↵† Co-first authors.