BVES downregulation in non-syndromic tetralogy of fallot is associated with ventricular outflow tract stenosis

BVES is a transmembrane protein, our previous work demonstrated that single nucleotide mutations of BVES in tetralogy of fallot (TOF) patients cause a downregulation of BVES transcription. However, the relationship between BVES and the pathogenesis of TOF has not been determined. Here we reported our research results about the relationship between BVES and the right ventricular outflow tract (RVOT) stenosis. BVES expression was significantly downregulated in most TOF samples compared with controls. The expression of the second heart field (SHF) regulatory network genes, including NKX2.5, GATA4 and HAND2, was also decreased in the TOF samples. In zebrafish, bves knockdown resulted in looping defects and ventricular outflow tract (VOT) stenosis, which was mostly rescued by injecting bves mRNA. bves knockdown in zebrafish also decreased the expression of SHF genes, such as nkx2.5, gata4 and hand2, consistent with the TOF samples` results. The dual-fluorescence reporter system analysis showed that BVES positively regulated the transcriptional activity of GATA4, NKX2.5 and HAND2 promoters. In zebrafish, nkx2.5 mRNA partially rescued VOT stenosis caused by bves knockdown. These results indicate that BVES downregulation may be associated with RVOT stenosis of non-syndromic TOF, and bves is probably involved in the development of VOT in zebrafish.

tract area, which promotes outflow tract elongation and the correct fusion of the outflow tract myocardial wall and ventricular septum 13 . Ablation of SHF may result in TOF due to an abnormal outflow tract 8 . In animal models, the knockout or knockdown of SHF regulatory network genes, such as Gata4 14 , nkx2.5 12,15 , Tbx1 16 , Tbx20 17 , and Hand2 18 , leads to abnormal development of the outflow tract. Functional mutations of SHF genes, such as NKX2.5 [19][20][21] , TBX1 22,23 , GATA4 24,25 , TBX20 26 , and HAND2 27 , were found in TOF patients. The expression of these genes, including NKX2. 5 28 , GATA4 28 , and TBX1 29 , was significantly decreased in the cardiac tissue samples of TOF, indicating that the downregulation of SHF genes is associated with outflow tract stenosis.
BVES, a blood vessel epicardial substance, also known as POPDC1, belongs to the Popeye domain containing (POPDC) gene family, encodes a novel class of cyclic adenosine monophosphate (cAMP) effector proteins, is a highly evolutionarily conserved membrane protein and is highly expressed in adult heart and skeletal muscle in vertebrates [30][31][32] . In mice, Bves is expressed in the development of cardiomyocytes 33 and coronary endothelial cells 34 and is highly expressed in the adult conduction system 35 . Knockout of Bves in mice was shown to cause sinus bradycardia under stress in an age-dependent manner 35 . In humans, BVES was found to be more highly expressed in ventricles than in atria 36 . Functional mutations of the BVES gene were detected in a family affected by hereditary muscular dystrophy with arrhythmia 37 , and in non-syndromic TOF 38,39 . It has also been shown that the expression of BVES was downregulated in patients with heart failure 36 and non-syndromic TOF 39 . However, the relationship between BVES downregulation in non-syndromic TOF and outflow tract stenosis has not been determined.
Zebrafish have been used as an animal model for the development of VOT. In zebrafish embryos, SHF cells are derived from mesodermal nkx2.5-and gata4-positive cells 15 . Zebrafish bves expression began at the zygote transcript stage (1 h post fertilization, 1 hpf) and persisted through the developmental stages of the heart, including the SHF area and other cardiac tissues 40,41 . It was reported that the expression of the Aggrecan and Cyp26 genes was changed in cardiac tissues with aortic valve disease and TOF 42,43 , and was also associated with the development of the cardiac outflow tract in zebrafish 42,44 . However, whether bves is involved in the development of VOT has not been determined.
In this study, BVES downregulation was detected in most of the 83 tissue samples from TOF patients together with the downregulation of the key SHF genes related to TOF. bves downregulation in zebrafish led to abnormal cardiac looping and VOT stenosis and subsequently led to downregulation of the key SHF genes. In addition, our findings suggested that bves and nkx2.5 mRNA may partially rescue the VOT phenotype caused by the downregulation of bves to varying degrees.

Materials and methods
All methods were performed in accordance with the relevant guidelines and regulations.
Samples. This study was approved by the Ethics Committee of Hunan Normal University (NO. 014050), Guangdong General Hospital and the Institutional Ethics Committee of Guangdong Academy of Medical Sciences (GDREC2016186A). Written informed consent was obtained from each subject or their guardian.
TOF tissue samples were obtained from the hypertrophic muscle tissue of RVOT during open-heart surgery for TOF. Normal control samples were obtained from the RVOT tissue of individuals who died in accidents and agreed beforehand for their organs to be donated. Samples from cases with congenital heart-related diseases were excluded. Information about patients and normal controls is shown in Tables 1 and 2, respectively. After obtaining the tissue, it was quickly cut into small, 4-5 mm pieces with surgical scissors, placed into an Eppendorf Quantitative reverse-transcription polymerase chain reaction (qRT-PCR). Total cDNA was prepared from whole embryos or tissues and qRT-PCR was performed as previously described 39 . In patient samples, the data of Ct values were normalized to GAPHD, and the fold change between normal control samples and TOF tissue samples was quantified using the 2 −ΔΔCT Livak Method. The significance was analyzed by Student's t-test.
In the same plate, the expression in patient tissue that was lower than the average expression of normal samples was defined as 'downregulated' , while the expression in patient tissue was higher than the average expression of normal samples was defined as 'upregulated' , and the data are presented as scatter points and histogram that were generated by GraphPad Prism 5. In zebrafish samples, the control group and knockdown group included at least three samples for each time point. The data of Ct values were normalized to gapdh, and the fold change between control group and knockdown group was quantified using the 2 −ΔΔCT Livak Method. The significance was analyzed by Student's t-test. The data are presented in the form of a histogram that were generated by GraphPad Prism 5. All primer sequences are shown in Table S1.

Plasmid construction. The construction of overexpression plasmid with pCMV-BVES and luciferase
reporter plasmids with the human NKX2.5, GATA4 and MEF2C promoters were performed as previously described 39 . Promotors of TBX1, TBX20 and HAND2 were amplified by primer 1 (forward, ATC GGT ACC GAA TTCagaatgtccaacacaacatcc; reverse, GTG ATA TCA GAT CTCccatcaggcccagtctgagg, the capital letters are homologous arms on Vector), primer 2 (forward, ATC GGT ACC GAA TTCagaatgtccaacacaacatcc, reverse, GTG ATA TCA GAT CTCccatcaggcccagtctgagg, the capital letters are homologous arms on Vector), and primer 3 (forward, ATC GGT ACC GAA TTCacacgagtaaggccggtttt, reverse, GTG ATA TCA GAT CTCcggttagagctgtttggggt, the capital letters are homologous arms on Vector), repectively. pGL3-Bias plasmid were digested by restriction enzyme, XhoI, to linearize the plasmid. Promotor sequences were ligated into vectors using ClonExpress Ultra One Step Cloning Kit (Vazyme).
Luciferase reporter assays. The luciferase reporter assays were analysesd as previously described 39 .
HEK293T cell line was used, and the levels of firefly luciferase were standardised relative to that of Renilla luciferase.
Morpholinos and mRNA injections. Zebrafish embryos were injected as reported previously 40 , and the sequences of the bves morpholino oligos was designed intended to block the translation of bves (5′-GAT GTT GTG TTG GAC ATT CTG AGG C-3′, synthesised by GeneTools). pXT7-bves and pXT7-nkx2.5 were linearized and used for in vitro transcription (Ribo m7G Cap Analogue, RiboMAX Large Scale RNA Production Systerm-T7; Promega). A total of 150 ng and 80 ng of capped mRNA of bves and nkx2.5 was coinjected with bves morpholino, respectively.
Phenotypic analyses of zebrafish. Embryos were incubated at 28.5 °C in petri dishes containing fresh water and maintained as described 48 . 6% methylcellulose was used to restrict the movement of zebrafish and Axiocam of Zeiss Company was used to photograph and analyse the morphology of juvenile fish at 48 hpf and 72 hpf. The heart phenotype of the juvenile fish at these two stages were analysed as previously described 48 . The range of motion of the heart was clearly seen, and the approximate position of VOT was observed. Under normal beating conditions of zebrafish heart, continuous photographs were taken with green fluorescence. Finally, the diastolic and systolic images were selected for analysis. Zeiss AxioVision 3.0.6 software and Adobe Photoshop were used to process images. We measured the width of ventricle outflow tract at the end-systolic stage using Digimizer, which normalised the scales added by Zeiss AxioVision 3.0.6 software automatically.
Western blot. Total protein samples were prepared in radioimmunoprecipitation assay (RIPA) buffer, and the protein concentration was determined by BCA assay (Beyotime). Protein isolation was carried out in a 12% www.nature.com/scientificreports/ SDS polyacrylamide gel. Then, the protein was transferred to nitrocellulose membranes, blocked with 8% skim milk, and incubated with anti-BVES antibody (1:1,500 dilution; Absin) and anti-β-ACTIN antibody (1:3,000 dilution; Proteintech). The signal densities of BVES protein bands were quantified and normalised to β-ACTIN using ImageJ.
Statistical analysis. The scatter points and histograms were generated by GraphPad Prism 5. To assess whether the experimental data of two groups and three groups were significantly different from each other, we applied Student's t-test and one-way Anova, respectively. A p value of < 0.05 was considered statistically significant.

Results
Expression pattern of human BVES was related to pulmonary artery development. The RVOT in TOF patients showed a phenotype of RVOT stenosis in all samples examined (Fig. 1A). The parasternal short axis view of cardiac colour Doppler echocardiography is important to assess the size of, for instance, the RVOT, PV, main pulmonary artery, and ascending aorta 49 . In normal people, the blood flow of RVOT and PA was antegrade and presented a single colour, while in TOF patients, it was aliasing and turbulent because of stenosis and presented multiple colours 50 (Fig. 1A). The expression pattern of BVES in RVOT tissues with TOF was examined by qRT-PCR, and 79 of the 83 cases were valid for qRT-PCR analysis. Among them, the downregulation of BVES was detected in 57 cases, accounting for 72.2% of the total, while there was upregulation of BVES in the remaining 22 cases (Fig. 1B). Western blot analysis showed that the level of BVES protein expression in the three samples with BVES downregulation was approximately half that in the control (Fig. 1C,D). These results indicate that BVES downregulation could be related to TOF.
Expression pattern of a set of SHF genes in RVOT stenosis with TOF. It has been shown that the genes in the SHF regulatory network are required for the development of the cardiac outflow tract 2,5,6,9,10,51 . www.nature.com/scientificreports/ To examine the expression pattern of the SHF genes in RVOT stenosis tissue, the gene expression patterns in the RVOT tissues of TOF with BVES downregulation were analysed by qRT-PCR. The results showed that the expression of human GATA4, NKX2.5, TBX1, HAND2, SMYD1 and MEF2C was significantly downregulated compared with that in the control ( Fig. 2A-F), which is similar to the cases for GATA4, NKX2.5 and TBX1 observed in other studies 28,29 . Compared with the control samples, the mean expression of ISL1 was higher than that of the control group, which was significant (Fig. 2G), although approximately 41.7% of the RVOT stenosis samples (20/48) were downregulated. However, the expression of TBX20 was unexchanged (Fig. 2H). These results further indicate that the downregulation of the SHF genes is related to ROVT stenosis in TOF.
bves knockdown in zebrafish led to abnormal heart looping. Previous studies showed that abnormal looping of the heart (the change of outer or inner curvature) disrupts the alignment between the ventricle and artery, leading to congenital heart phenotypes, such as TOF 8,52,53 . To explore the causal relationship in vivo between BVES downregulation in ROVT stenosis and the development of heart looping, a zebrafish bves knockdown model with bves morpholino was used. We chose a bves morpholino blocking bves translation that was reported in a previous report 40 . Knockdown efficiency was verified by analysing the survival rate of the morphants at 72 hpf. The results showed that the morphants with 2.0 ng were suitable for analysing the cardiac phenotype, which is consistent with a previous report 40 , while the morphants with 1.25 ng were similar to the wild type, and the morphants with 5 ng with a death occurred at 5 dpf (Fig. 3A). Therefore, a bves knockdown model, the morphants injected with 2.0 ng of morpholino were used in the subsequent analyses, we called it bves Mo.
To determine the effect of bves knockdown on cardiac development, the embryos of Tg (cmlc2:dsRed) zebrafish were used for bves morpholino injection, in which RFP was localized at the cardiomyocytes, facilitating live imaging of heart morphology. At 48 hpf, the morphants showed looping defects, cardiac dysplasia, and cardiac oedema (Fig. 3B). Classification into four phenotypes could be performed according to the degree of cardiac abnormality 48 , and approximately 70% of bves morphants showed defects of various degrees of severity, and this phenotype was partially rescued by coinjection with bves morpholino and bves mRNA (Fig. 3C). At 72 hpf, approximately 6.7% (3/45) of the morphants showed a left-looping heart in the ventral view (Fig. 3D), and the morphant with left-looping heart was not found in the WT or coinjection with bves morpholino and bves mRNA. These results indicated that knockdown of bves led to abnormal cardiac looping. www.nature.com/scientificreports/ bves knockdown in zebrafish causes VOT stenosis. The RVOT is the site of circulation between the right ventricular and the pulmonary artery 5,54 . elnb, a marker of the outflow tract in zebrafish, showed high expression at 72 hpf 55 . Compared with the wild type, the expression of elnb was downregulated in bves morphants (Fig. 4A), consistent with TOF patient samples (Fig. 4B). The results showed that, in the bves morphants, the VOT was abnormal. Stenosis of the PA or the RVOT is an indicator of TOF. To study the role of bves in the development of the outflow tract, the embryos of Tg (flia:eGFP) zebrafish were used for the bves morpholino injection, in which GFP is expressed in all endothelial cells 46 . In the wild type (Fig. 4C, top row), the VOT at 72 hpf was open at the end-systolic stage, while the VOT was closed at the end-diastolic stage (Fig. 4C). In the bves morphants (Fig. 4C, middle row), the VOT was smaller and narrower than that in the wild type at the end-systolic stage. The width of VOT at the end-systolic stage was reduced approximately 50% (WT, 9.5 μm; bves Mo, 5.0 μm, p < 0.01) (Fig. 4D). The narrow phenotype was mostly rescued by bves mRNA (bves Mo + bves mRNA, 7.5 μm, p < 0.01 compared with WT; p < 0.01 compared with bves Mo) (Fig. 4C, bottom row; Fig. 4D). These results suggest that bves knockdown affected the development of outflow tract, leading to the phenotype of VOT stenosis.
bves regulated VOT development via SHF genes. The development of VOT at the late stage of cardiac looping is closely related to the SHF genes 2,5,6,9,10,51 . To detect the effects of SHF genes on the development of outflow tract, their expression in the bves morphants at 48 hpf was analysed by qRT-PCR. As seen in Fig. 5A, the expression of gata4, nkx2.5, hand2, tbx1, tbx20 and mef2c was significantly downregulated, which is consistent with the downregulation detected in the RVOT tissues of TOF (Fig. 2), except for tbx20. The expression of smayd1a, smayd1b and isl1 was unchanged (Fig. 5A). To study the regulatory relationship between bves and the SHF genes, GATA4, NKX2.5, TBX1, HAND2, TBX20 and MEF2C were chosen for dual-fluorescence reporter system analysis. The results showed that the overexpression of BVES in the HEK293T cell line significantly increased the transcriptional activities of the promoters of GATA4, NKX2.5 and HAND2 by tenfold, 1.5-fold and threefold, respectively (Fig. 5B). For other transcriptional activities, TBX1 was decreased by ninefold, and TBX20 and MEF2C were unchanged. These results suggest that genes in the SHF regulatory network, such as GATA4, NKX2.5 and HAND2, are potential factors involved downstream of BVES. www.nature.com/scientificreports/ To further confirm that bves regulates VOT development through the SHF genes, we coinjected bves morpholino and nkx2.5 mRNA into Tg (flia:EGFP) transgenic zebrafish and analysed the development of VOT at 72 hpf. The results showed that the width of VOT was partially rescued by nkx2.5 mRNA (Fig. 5C,D) (WT, 9.5 μm, bves Mo, 5.0 μm, p < 0.01 compared with WT; bves Mo + Nkx2.5 mRNA, 6.3 μm, p < 0.01 compared with WT, p < 0.05 compared with bves Mo). These results indicated that bves regulated the development of VOT via its downstream genes, such as nkx2.5, gata4 and hand2. However, the results need to be further studied.

Discussion
Our previous report showed that BVES allelic variants are associated with RVOT stenosis in TOF patients, which leads to the downregulation of BVES itself at the transcriptional and protein levels in the tissues of cases of RVOT stenosis with TOF containing allelic variants 38,39 . These results suggest that the downregulation caused by the BVES allelic variants was related to the phenotype of RVOT stenosis, consistent with the finding that gene downregulation is associated with RVOT stenosis, as demonstrated by other investigators 28,29,38 . However, the mechanism underlying the association between BVES downregulation and RVOT stenosis has not been elucidated. In this study, BVES downregulation was detected in approximately 72.2% of the samples (57/79) from cases of RVOT stenosis with TOF with decreases of both mRNA and protein levels by half. The results presented in this paper are consistent with a previous report 39 , suggesting that the downregulation of the BVES gene in RVOT stenosis with TOF is an aetiology, and not only the aetiology that lead to the phenotype of TOF.
It has been shown that, at the late stage of cardiac looping, cells from the anterior SHF are added to the outflow tract area for outflow tract elongation and the correct fusion of the outflow tract myocardial wall and ventricular septum 6,7,13 . Ablation of SHF may result in TOF due to an abnormal outflow tract 8 . In this study, compared with www.nature.com/scientificreports/ the normal RVOT samples, the expression of genes in the SHF regulatory network, such as NKX2.5 and GATA4, HAND2, TBX1, MEF2C and SMYD1, were significantly decreased in the TOF samples, with BVES being mostly downregulation (Fig. 2) and occasionally upregulated (data unpublished). In animal models, knockout or knockdown of SHF regulatory network genes, such as Gata4 14 and nkx2.5 12,15 , leads to abnormal outflow tract development. The expression of NKX2.5 and GATA4 in the open-heart surgery of CHD tissues is downregulated 28 , which is similar to that in the myocardial tissue at the earlier stage of embryonic development in the nkx2.5 and gata4 knockdown animal models 12,15 . These results suggest that the downregulation of gene expression detected in the tissues obtained during surgery may be similar to the gene downregulation in embryonic development. In our zebrafish model, knockdown of bves significantly decreased the expression of the five genes in the SHF regulatory network, including nkx2.5, gata4, hand2, mef2c and tbx1 (Fig. 5A), suggesting that bves downregulation may be the cause of the maldevelopment of heart looping and VOT. The results with a dual-fluorescence www.nature.com/scientificreports/ reporter system confirmed that BVES positively regulated the transcriptional activity of GATA4, NKX2.5 and HAND2 and negatively regulated the transcriptional activity of TBX1 promoters (Fig. 5B). The above results suggest that GATA4, NKX2.5 and HAND2 are potential downstream genes of the BVES gene. Nkx2.5 and Gata4, as marker of cardiac progenitors, together regulate the arrangement of ventricle outflow 14 . In Gata4+/− mice, a small number of heterozygote mutants were double outlet right ventricle (DORV), and in Gata4+/−;Gata5+/− double heterozygote mutants, almost all embryos were DORV, some of which had further aortic stenosis. Hand2, a heart and neural crest derivative, is important for the development of the heart, especially in the outflow tract 56 . In the mutant mice of Hand2f/f;Mef2c-Anf-Cre, which specifically deleted Hand2 in the SHF progenitors, the OFT lumen was narrowed, and the OFT wall was thickened 57 . In other mutant mice of Hand2f/-;Wnt1-Cre cKO, which specifically deleted Hand2 in the neural crest, a variety of arterial malformations, including pulmonary stenosis, were observed 18 . In this paper, nkx2.5 mRNA only partially rescued the OFT malformation caused by bves Mo, which indicated that bves caused OFT abnormalities by widely regulating the heart development gene, not one of the heart development genes. The cells of outflow tract are different from anterior lateral plate mesoderm (ALPM), where SHF progenitors are specified in higher vertebrates. The descendants of gata4+ or nkx2.5+ cell at ALPM are crucial for the development of outflow tract by compromised progenitor cell proliferation 15,58,59 . Nkx2.5 and Gata4 play an important role in the outflow tract by regulating the SHF genes, but they also play a vital role in early specialization of cardiomyocyte progenitor cells 60,61 . In our bves knockdown model in zebrafish, the early heart development was affected (Fig. 3), so bves leading to the abnormal phenotypes of outflow tract may be affected both by the outflow tract development via regulating SHF genes and the indirect consequence of early defects via regulating early cardiac progenitor specification. However, the molecular regulatory mechanism between bves and nkx2.5, gata4 and hand2 requires further study.
BVES, encodes a transmembrane protein, that contains three transmembrane domains and a Popeye domain. The Popeye domain makes up a large part of the cytoplasmic portion of the protein and functions as a cAMPbinding domain 62,63 . In both mice and zebrafish, Bves act as effector proteins of cAMP, which control the development of the conduction system by combining ion channel genes 35,37 . Furthermore, BVES has a functional mutant in AV-block patients 37,64 . However, in the process of carcinogenesis, BVES inhibits epithelial-mesenchymal transition (EMT) and cell adhesion of cancer cells to prevent the formation of cancer via GEFT/Rho signalling and WNT signalling 65 . In our previous study, BVES also had functional polymorphisms in TOF patients 39 , but whether BVES is related to the maldevelopment of OFT has not been studied to date. In this paper, we illustrate that bves is related to the development of OFT in zebrafish, and the potential target genes of bves are nkx2.5, gata4 and hand2. Whether bves acts as an effector protein of cAMP or as a signal transduction member, such as GEFT/ Rho signalling and WNT signalling, to regulate the expression of nkx2.5, gata4 and hand2 needs further study.
Zebrafish, as an animal model, has been used to study the pathogenesis of many diseases due to its simple developmental structure of the organs studied. In terms of studies on cardiac diseases, zebrafish have been used in research on heart failure 66 , congenital heart defects 67,68 , and outflow tract disease 42,69 . In this paper, we used morpholinos to downregulate bves expression in zebrafish. The downregulation of bves in zebrafish caused looping defects, cardiac dysplasia, cardiac oedema (Fig. 3B), stenosis of the VOT (Fig. 4C) and several heart rate abnormalities (data unpublished). Abnormal heart rhythm leads to abnormal ventricular contraction, which makes the cilia of the developing heart and outflow tract feel different shear stress, thereby affecting the development of the heart and outflow tract 70,71 . In our bves knockdown model, we found that some of the mutants had abnormal rhythms. Because the phenotype of outflow tract stenosis also appears in the mutants with normal heart development and rhythm, we think that the difference in outflow tract dynamics caused by ventricular contract was considerably less than outflow tract malformation. However, further studies need to be performed to support this hypothesis. In addition, the abnormal phenotypes of VOT and heart looping were partially rescued by bves mRNA (Figs. 3D, 4C), supporting the hypothesis that the phenotypes were partially caused by the downregulation of bves, and several might be caused by the toxicity of the morpholino. To further explore the molecular mechanism by which bves induced developmental abnormalities in the heart and VOT, it is necessary to establish a bves knockout zebrafish line.
In summary, we studied the relationship between BVES downregulation and the occurrence of RVOT stenosis in TOF using human RVOT stenosis samples and a zebrafish model of bves downregulation. Our results show that bves is required for the development of VOT in zebrafish, suggesting that BVES downregulation is associated with the occurrence of RVOT stenosis in non-syndromic TOF patients. www.nature.com/scientificreports/