Introduction

The heart develops with continuous contraction, and biomechanical cues play an essential role in cardiac morphogenesis^1,2. Blood flow is directly sensed by the surrounding endocardium, which undergoes multiscale remodeling during zebrafish heart development. In the atrioventricular canal (AVC) endocardium, oscillatory flow promotes valvulogenesis through transient receptor potential (TRP) channel-mediated expression of Krüppel-like factor 2a (klf2a)^3–5. Meanwhile, mechanical forces trigger ATP-dependent activation of purinergic receptors, inducing expression of nuclear factor of activated T cells 1 (nfatc1) and subsequent valve formation⁶. In the chamber endocardium, blood flow induces endocardial cells to adopt chamber- and region-specific cell morphology during cardiac ballooning⁷. A recent study further emphasizes that blood flow is essential for endocardial cell accrual in assembling the outflow tract⁸. Beyond their role in endocardial cells, proper biomechanical cues are indispensable for shaping the myocardium. For instance, in contraction compromised tnnt2a⁹ and myh6¹⁰ mutants, trabeculation is markedly reduced. Moreover, apart from the tissue-scale regulatory effect, the shape changes¹¹ and myofibril content¹² at the single-cardiomyocyte level are also sculpted by the interplay of contractility and blood flow in the developing heart.

In ventricular myocardium morphogenesis, biomechanical forces coordinate intra-organ communication between endocardial and myocardial cells by regulating BMP, Nrg/Erbb, and Notch signaling. The Nrg-Erbb axis stands as one of the most extensively studies signaling pathway mediating cell-cell communications in the heart^13–15. In particular, endocardial Notch activity induced by cardiac contraction promotes the expression of Nrg, which then secretes into the extracellular space, binding to Erbb2/4 receptor tyrosine kinases on cardiomyocyte and promoting their delamination^16,17. In agreement with the pivotal role of this signaling pathway in heart development, genetic mutations in zebrafish nrg2a and erbb2 result in severely compromised trabeculae formation^18–20.

The development of cardiomyocytes encompasses the specification of subcellular structure, metabolic state, gene expression profile, and functionality^21,22. The rhythmic contraction of cardiomyocytes relies on precisely regulated excitation-contraction coupling (E-C coupling), transducing electrical activity into contractile forces. This intricate signaling cascade involves membrane action potential, calcium signaling, and sarcomeric structure²³. Specifically, membrane depolarization triggers the opening of L-type calcium channel (LTCC), allowing calcium influx. The calcium signaling then activates the ryanodine receptor on the sarcoplasmic reticulum (SR) membrane, releasing additional calcium²³. E-C coupling is essential for heart development, as evidenced by the complete silence of the ventricle and reduced cardiomyocyte number in cacna1c (LTCC α1 subunit in zebrafish) mutant²⁴. Beyond its role in modulating cardiac structure formation, previous studies indicate that Nrg-Erbb2 signaling is also necessary for cardiac function, as erbb2 mutants exhibit severely compromised fractional shortening and an immature conduction pattern^18,25. However, the precise mechanism through which the blood flow-Nrg-Erbb2 axis controls cardiomyocyte functionality remains poorly understood.

Results

Transcriptome analysis identifies id2b as a blood flow sensitive gene

Blood circulation is dispensable for early embryonic development in zebrafish, presenting an ideal model to investigate biomechanical cues influencing heart morphogenesis. To identify genes affected by cardiac contraction or blood flow, we treated myl7:mCherry zebrafish embryos with tricaine to inhibit cardiac contractility from 72 hours post-fertilization (hpf) to 96 hpf. Hearts from control and tricaine-treated zebrafish embryos were manually collected under a fluorescence stereoscope as previously reported²⁶. Subsequently, approximately 1,000 hearts from each group were subjected to RNA sequencing (Figure 1A). A total of 4,530 genes with differential expression were identified, comprising 2,013 up-regulated and 2,517 down-regulated genes. With a specific focus on identifying key transcription factors (TFs) affected by perturbing biomechanical forces, differentially expressed genes (DEGs) encoding TFs were enriched into signaling pathways through KEGG analysis. Interestingly, our analysis identified several pathways known to be involved in heart development, including the transforming growth factor beta (TGFβ) signaling and Notch signaling pathways (Figure 1B). In particular, the scaled expression levels of the top 6 DEGs (|log₂FC| ≥ 0.585), exhibiting up- or down-regulation, were listed (Figure. 1C). Notably, as a component of the helix-loop-helix (HLH) family, id2b is the zebrafish homolog of the mammalian Id2 gene, functioning as a transcriptional repressor. Intriguingly, Id2 has been shown to regulate the formation of the murine AV valve, a process notably influenced by alterations in blood flow directionality. Consequently, we interrogated the expression and function of id2b in developing embryos.

Figure 1.

Through quantitative real-time PCR (qRT-PCR) analysis on purified embryonic hearts, we observed a ∼50% reduction in id2b mRNA level following tricaine treatment from 72 to 96 hpf compared to the control (Figure 1D). Furthermore, in situ hybridization was performed to visualize id2b expression under tricaine or 10 uM blebbistatin (an inhibitor of sarcomeric function and cardiac contractility) treatment from 48 to 72 or from 72 to 96 hpf. Consistently, our results showed a reduction in id2b signal in contraction-deficient hearts compared to the control (Figure 1E). Similarly, injection of a previously characterized tnnt2 morpholino at the 1-cell stage also led to compromised contraction and diminished expression of id2b (Figure 1F). Taken together, these results indicate that biomechanical cues are essential for activating id2b in embryonic hearts.

Visualization of the spatiotemporal expression of id2b in developing embryos

Due to technical challenges in visualizing the cell-type-specific expression of id2b in the developing heart using whole-mount in situ hybridization, we employed an intron targeting-mediated approach²⁷ to generate a knockin id2b:eGFP line. This method allowed us to achieve specific labeling without perturbing the integrity and function of the endogenous gene²⁷ (Figure 2A). Comparison of id2b:eGFP fluorescence with in situ hybridization at 24, 48, and 72 hpf revealed a substantial recapitulation of the eGFP signal with that of endogenous id2b expression. The fluorescence was notably enriched in the heart, brain, retina, and notochord (Figure 2B), mirroring observations from a previously reported id2b transgenic line generated through BAC-mediated recombination²⁸.

Figure 2.

To further elucidate the spatiotemporal expression of id2b in developing hearts, we crossed id2b:eGFP with myl7:mCherry or kdrl:mCherry, labelling cardiomyocytes or endocardial cells, respectively. Confocal images revealed minimal, if any, presence of id2b:eGFP in myl7:mCherry⁺cardiomyocytes (Figure 2C). In sharp contrast, clear co-localization between id2b:eGFP and kdrl:mCherry was evident at 48, 72, and 96 hpf (Figure 2D). Interestingly, there was an absence of id2b:eGFP signal in kdrl:mCherry⁺ endothelial cells in trunk blood vessel and brain vasculature (Figure 2E). In adult hearts, id2b:eGFP fluorescence was enriched in the chamber endocardium and the endothelium lining AVC, outflow tract (OFT), and bulbus arteriosus (Figure 2F).

Cardiac contraction promotes endocardial id2b expression through primary cilia but not BMP

Taking advantage of live imaging on developing embryos, we explored the in vivo dynamics of id2b in response to biomechanical force at single-cell resolution. When embryos were treated with tricaine or blebbistatin, the intensity of id2b:eGFP in atrial and ventricular endocardium was significantly reduced (Figure 3A, B). Similarly, injection of tnnt2a morpholino also markedly suppressed id2b:eGFP signal (Figure 3A, B), in agreement with the results obtained from in situ hybridization. Strikingly, the reduction in fluorescence intensity was particularly pronounced in AVC endothelial cells (Figure 3A, B, asterisks).

Figure 3.

We then explored how cardiac contraction modulated id2b expression. Given that endocardial cells can sense blood flow through primary cilia^29,30, we used a characterized morpholino²⁹ to knockdown ift88, an intraflagellar transporter essential for primary cilia formation. Previously work demonstrated a complete loss of primary cilia in endocardial cells upon ift88 knockdown²⁹. As expected, a significant decrease in id2b:eGFP intensity was observed in the chamber and AVC endocardium of ift88 morphants compared to control (Figure 3C, D), suggesting that biomechanical forces promote the expression of id2b via primary cilia. In the developing heart, a central hub for mediating biomechanical cues is the Klf2 gene, which includes the klf2a and klf2b paralogues in zebrafish^3–5,29,31. Previous studies in mammals and zebrafish have highlighted the essential role of Klf2 transcription factor activity in cardiac valve and myocardial wall formation^3,31. As a flow-responsive gene, klf2a expression has been observed throughout the entire endocardium, evidenced by mRNA expression and transgenic studies^3,4. Interestingly, in situ hybridization on 48 and 72 hpf klf2a^-/- and klf2b^-/- embryos unveiled a drastic decrease in id2b expression compared with wild-type zebrafish (Figure 3E), supporting the notion that klf2-mediated biomechanical signaling is essential for activating id2b expression.

Because id2b has been reported to be a target gene of bone morphogenetic protein (BMP) signaling, we explored whether BMP also played a role in regulating id2b when blood flow was perturbed. To this end, we knocked down bmp2b, bmp4, and bmp7a in 1-cell stage id2b:eGFP embryos and found a significant reduction in fluorescence signal in morpholino-injected hearts compared to control (Figure 3-figure supplement 1A, B), suggesting that id2b is a target gene of BMP signaling during early embryonic development. Considering that heartbeats in zebrafish commence at approximately 22 hpf, we treated embryos with the BMP inhibitor Dorsomorphin from 24 to 48 hpf or from 36 to 60 hpf. While the number of endocardial cells was slightly reduced upon Dorsomorphin exposure, as previously reported⁷, surprisingly, quantification of the average id2b:eGFP fluorescence intensity in individual endocardial cells revealed no significant differences between Dorsomorphin and DMSO controls (Figure 3-figure supplement 1C, D).

We further visualized BMP activity using the Bre:dGFP reporter line. Confocal images revealed strong fluorescence in the myocardium at 72 hpf, with minimal signal present in the endocardium except for the AVC endothelium (Figure 3-figure supplement 1E). Moreover, after tricaine treatment, endocardial Bre:dGFP slightly increased (Figure 3-figure supplement 1E), as opposed to the reduced id2b:eGFP signal (Figure 3A, B). Likewise, endocardial Bre:dGFP intensity was barely affected after completely blocking contraction with tnnt2a MO injection (Figure 3-figure supplement 1E). These observations align with previous work using pSmad-1/5/8 as a readout of BMP activity, indicating that endocardial BMP signaling is independent of blood flow⁷. Collectively, these results suggest that BMP signaling and blood flow modulate id2b expression in a developmental-stage-dependent manner.

Compromised AV valve formation in id2b mutants

To investigate the role of the contractility-id2b axis in zebrafish heart development, we generated a loss-of-function mutant line using CRISPR/Cas9. A pair of sgRNAs designed to target exon 1 was injected with zCas9 protein into 1-cell-stage embryos. Consequently, we identified a mutant allele with a 157 bp truncation, leading to the generation of a premature stop codon (Figure 4A). The overall morphology of id2b mutants (id2b^-/-) remained unaltered compared to id2b^+/+ siblings at 72 and 96 hpf (Figure 4B). However, id2b^-/- zebrafish were subjected to early lethality starting around 31 weeks post-fertilization (Figure 4C). Strikingly, pericardial edema was observed in 20% (9/45) of adult id2b^-/- zebrafish (Figure 4D, top). Upon dissecting hearts from these id2b^-/- zebrafish, a prominent enlargement in the atrium was detected (Figure 4D, bottom). Histological analysis further revealed malformation in the AV valves of these id2b^-/- mutants compared to the control (Figure 4E). Upon closer inspection of AV valve anatomy, we noted that the superior and inferior leaflets were significantly thinner, comprising only 1-2 layer of cells in id2b^-/-zebrafish with an enlarged atrium. This was in sharp contrast to id2b^+/+ zebrafish, which exhibited multilayers of cells (Figure 4E). Subsequent examination of the remaining 80% of id2b^-/- zebrafish (36/45) that did not display prominent pericardial edema also revealed an enlarged atrium and AV valve malformation, albeit to a lesser extent (data not shown). In contrast, id2b^+/+ controls (48/48) displayed normal heart morphology with intact AV valve (Figure 4E, left).

Figure 4.

To further interrogate the effect of id2b inactivation on AV valve formation and function, we analyzed the number of AVC endothelial cells using kdrl:nucGFP. At 96 hpf, a reduced number of kdrl:nucGFP⁺ cells was detected in the AVC region of id2b^-/- embryos compared with id2b^+/+ (Figure 4-figure supplement 1A, B). In contrast, the number of atrial and ventricular endocardial cells did not differ between id2b^-/- and id2b^+/+ (Figure 4-figure supplement 1C, D). Subsequently, we assessed hemodynamic flow by conducting time-lapse imaging of red blood cells labelled by gata1:dsred. Surprisingly, the pattern of hemodynamics was largely preserved in id2b^-/- compared with id2b^+/+ siblings at 96 hpf (Figure 4-figure supplement 1E, Video 1, 2). Additionally, we performed echocardiography to analyze blood flow in adult zebrafish as previously described³². In id2b^-/- hearts, prominent retrograde blood flow was detected in the AVC region (8/13) (Figure 4G), while unidirectional blood flow was observed in id2b^+/+ (10/10) (Figure 4F). Quantification analysis showed ∼32% retrograde blood flow in id2b^-/-, compared to 0% in id2b^+/+ zebrafish (Figure 4H). Consistently, the superior and inferior leaflets were much thinner in id2b^-/- exhibiting retrograde flow compared with control fish (data not shown). Overall, these histological and functional analyses indicate that id2b deletion leads to progressive defects in AV valve morphology and hemodynamic flow.

id2b deletion perturbs calcium signaling and contractile function in the myocardium

Although similar defects in AV valve formation have been reported in both klf2a and nfatc1 mutants, they do not display noticeable pericardial edema at the adult stage, nor do they experience early lethality^3,29,31–33. Therefore, we sought to investigate whether other cardiac properties have also been affected by id2b loss-of-function. To this end, we employed RNA-seq analysis on purified embryonic id2b^-/- and id2b^+/+ hearts (Figure 5-figure supplement 1). As expected, enrichment analysis of DEGs demonstrated that the top-ranked anatomical structures affected by id2b deletion included the heart valve, the compact layer of ventricle, and the atrioventricular canal (Figure 5-figure supplement 1A). Interestingly, id2b inactivation also impacted phenotypes such as cardiac muscle contraction and heart contraction (Figure 5-figure supplement 1B). Therefore, we investigated cardiac contractile function through time-lapse imaging on the myl7:mCherry background. At 120 hpf, a significant decrease in cardiac contractility was observed in id2b^-/-, as evidenced by reduced fractional area change and heart rate, compared with id2b^+/+(Figure 5A-C). Similarly, echocardiography analysis showed that the contractile function in adult id2b^-/- heart was dramatically reduced compared with age-matched id2b^+/+ (Figure 5D, E). These functional defects in id2b-deleted hearts could not be attributed to differences in cardiomyocyte number, as we counted cardiomyocytes using the myl7:H2A-mCherry line and found no apparent changes between id2b^-/- and id2b^+/+ embryos at 72 and 120 hpf (Figure 5-figure supplement 2A, B). Similarly, id2b^-/- also developed regular trabecular structures at embryonic (Figure 5-figure supplement 2C) and adult stages (Figure 4E). Through α-actinin immunostaining, we observed similar sarcomeric structures in 72 hpf and adult (115 dpf) id2b^-/- and control cardiomyocytes (Figure 5-figure supplement 2D), corroborating that the reduced contractility in id2b-depleted heart was independent of structural defects.

Figure 5.

The key functional unit that transmits electrical activity to contractile function is E-C coupling. Because id2b^-/- displayed reduced cardiac function without conspicuous myocardial structural defects, we visualized calcium signaling in the developing heart using actb2:GCaMP6s zebrafish (Figure 5G). Compared to id2b^+/+ control, id2b^-/- exhibited markedly decreased calcium amplitude at 120 hpf (Figure 5H), which is consistent with the reduced fractional area change. In cardiomyocyte, the entry of extracellular calcium is mainly mediated through the LTCC. As previously reported, a defect in zebrafish LTCC pore forming α1 subunit cacna1c leads to compromised cardiac function²⁴. We collected hearts from 72 hpf and 5 months post-fertilization (mpf) zebrafish and detected downregulated cacna1c in id2b^-/- compared to id2b^+/+(Figure 5I). In addition, we measured cardiac action potential using intracellular recording³⁴. Compared to id2b^+/+ zebrafish, the duration of the action potential in id2b^-/- was significantly shorter (Figure 5J, K), consistent with the decreased expression level of cacna1c. Together, these data indicate that id2b loss-of-function leads to compromised calcium signaling and cardiac contractile function.

Reduced expression of nrg1 mediates the compromised contractility in id2b^-/-

Because the deficiency of id2b in the endocardium disrupted the function of myocardium, we speculated that the crosstalk between these two types of cells was affected in id2b^-/-. Interestingly, comparing the differentially expressed genes in embryonic id2b^-/-and id2b^+/+ hearts identified a significant reduction in the expression level of nrg1, a key mitogen regulating the intra-organ communications between endocardial cells and cardiomyocytes (Figure 6A). Remarkably, analysis of a zebrafish single-cell database³⁵ revealed enriched expression of nrg1 in endocardial cells (Figure 6-figure supplement 1). However, attempts to detect nrg1 expression through in situ hybridization were unsuccessful, likely due to its low abundance in the heart. Alternatively, qRT-PCR analysis of purified 72 hpf embryonic hearts validated decreased nrg1 levels in id2b^-/- compared to control (Figure 6B). Previous studies have demonstrated that perturbations in Nrg-Erbb2 signaling, as seen in zebrafish erbb2 mutants, result in dysfunctional cardiac contractility¹⁸. Consistently, a decrease in heart rate was observed in embryos treated with the erbb2 inhibitor AG1478 (Figure 6E).

Figure 6.

Remarkably, injecting nrg1 mRNA at the 1-cell stage not only rescued the reduced expression of cacna1c in id2b^-/- hearts (Figure 6C) but also restored the diminished heart rate (Figure 6D). This is consistent with prior study showing that Nrg1 administration can restore LTCC expression and calcium current in failing mammalian cardiomyocytes³⁶. Overall, our data suggest that endocardial id2b promotes nrg1 synthesis, thereby enhancing cardiomyocyte contractile function.

id2b interacts with tcf3b to limit its repressor activity on nrg1 expression

We further interrogated how id2b promotes the expression of nrg1. As a HLH factor lacking a DNA binding motif, id2b has been reported to form a heterodimer with tcf3b and inhibit its function. Notably, we detected expression of tcf3b in endocardial cells by analyzing a zebrafish single-cell database³⁵ (Figure 7-figure supplement 1). To determine whether this interaction occurs in zebrafish, Flag-id2b and HA-tcf3b were co-expressed in HEK293 cells. Co-immunoprecipitation analysis confirmed the direct binding of id2b to tcf3b protein (Figure 7A). qRT-PCR analysis on purified 72 hpf embryonic hearts revealed a significant increase in the expression of socs3b and socs1a, target genes of tcf3b, in id2b^-/- compared to id2b^+/+ (Figure 7B). This suggests an elevation in tcf3b activity associated with the loss of id2b function. Notably, the expression levels of tcf3a and tcf3b remained consistent between id2b^-/- and id2b^+/+ hearts (Figure 7B).

Figure 7.

To understand how the altered interaction between id2b and tcf3b influences nrg1 expression, we analyzed the promoter region of zebrafish nrg1 using JASPAR and identified two potential tcf3b binding sites (Figure 7C). Subsequently, a DNA fragment containing the zebrafish nrg1 promoter region was subcloned into a vector carrying the luciferase reporter gene. Co-injection of this construct with tcf3b mRNA into 1-cell stage embryos resulted in a significant decrease in luciferase signal. Conversely, co-injection with a previously characterized tcf3b morpholino led to enhanced luciferase intensity (Figure 7D). These results suggest a possible mechanism by which tcf3b represses nrg1 expression in zebrafish.

Lastly, injecting tcf3b morpholino into id2b^-/-embryos was performed to assess whether attenuating the overactive tcf3b in id2b^-/- could restore the expression level of nrg1. qRT-PCR analysis of purified 72 hpf hearts revealed a partial restoration of the diminished nrg1 expression in id2b^-/- upon tcf3b inhibition (Figure 7E). Taken together, our results indicate that biomechanical cues activate endocardial id2b expression, leading to its interaction with tcf3b to alleviate repression on the nrg1promoter. Consequently, the depletion of id2b unleashes tcf3b’s repressor activity, leading to a reduction in nrg1 expression, which further acts through erbb2 to regulate cardiomyocyte function (Figure 7F).

Discussion

Biomechanical forces play an essential role in regulating the patterning and function of the heart. At AVC, oscillatory flow promotes the expression of klf2a and nfatc1 to modulate valve morphogenesis. In chamber endocardium, blood flow induces endocardial cells to acquire distinct cell morphology. However, it still lacks a systematic analysis of the transcriptome underlying compromised heartbeats. In the present study, we analyzed embryonic zebrafish hearts without contractility and identified genes that are regulated by biomechanical forces. Specifically, our results unveiled the endocardial-specific expression of id2b, which was tightly regulated by flow-sensitive primary cilia-klf2 axis. Genetic deletion of id2b resulted in compromised valve formation and progressive atrium enlargement. In addition, a reduction in heart rate and contractile force was observed in id2b^-/-, owing to decreased expression of LTCC α1 subunit cacna1c. Mechanistically, id2b interacts with bHLH transcription factor tcf3b to limit its repressor activity. Hence genetic deletion of id2b unleashes tcf3b activity, which further represses endocardial nrg1 expression. As a result, Injection of nrg1 mRNA partially rescues the phenotype of id2b deletion. Overall, our findings identify id2b as a novel mediator that regulate the interplay between endocardium and myocardium during heart development.

In mammals, the deletion of Id2 leads to malformations in the arterial and venous poles of the heart, as well as affects AV valve morphogenesis^37,38. Interestingly, approximately 20% perinatal lethality is reported in Id2 knockout mice, exhibiting AV septal defects and membranous ventricular septal defects³⁷. Remarkably, pericardial edema is evident in 20% of adult id2b^-/- zebrafish, with a prominent enlargement of the atrium. The superior and inferior leaflets of AV valves in id2b^-/- mutants are significantly thinner compared to the control. Therefore, our results suggest that id2b may play a similar role in regulating AV valve formation in zebrafish as its mammalian orthologue Id2. It is proposed that the loss of Id2 in mice results in compromised endocardial proliferation and aberrant endothelial-to mesenchymal transformation (EMT), collectively leading to defective valve morphogenesis³⁷. Nevertheless, the mechanism by which id2b loss-of-function causes a reduction in leaflet thickness in zebrafish remains to be determined in future studies.

id2b has been recognized as a target gene of the BMP signaling pathway. As expected, knockdown of bmp2b, bmp4, and bmp7a at 1-cell stage confirms that endocardial id2b expression is controlled by BMP activity during early embryonic development. Surprisingly, treatment with the BMP inhibitor Dorsomorphin at 24 and 36 hpf, when cardiac contractions have already initiated, fails to alter id2b expression in the endocardium, suggesting that BMP is dispensable for id2b activation at these stages. Instead, endocardial id2b expression is reduced upon loss-of-function of klf2a, klf2b, and ift88, suggesting an essential role of the primary cilia-klf2 axis in mediating id2b activation. In endocardial cells, Trp, Piezo, and ATP-dependent P2X/P2Y channels^4,6,39,40 are well-established sensors for biomechanical stimulation. The activation of these channels further promotes the activities of Klf2 and Nfatc1 to drive heart development and valvulogenesis. However, whether these channels are also required for the activation of id2b warrants further investigation.

The Nrg-Erbb signaling plays an essential role in regulating heart morphogenesis and function. In the mammalian heart, the genetic deletion of Nrg1 or Erbb2 results in severely perturbed cardiac trabeculae formation^13–15. Zebrafish erbb2 mutants exhibit a similar defect in cardiomyocyte proliferation and trabeculation¹⁸. Interestingly, nrg1 mutant zebrafish display grossly normal cardiac structure during early embryonic devolpment^19,41. Nevertheless, zebrafish nrg2a loss-of-function leads to defective trabeculae formation, suggesting that nrg2a is the predominant ligand secreted from endocardium, promoting ventricular morphogenesis¹⁹. In the adult stage, perivascular cells⁴² or regulatory T cells⁴³-derived nrg1 promotes cardiomyocyte proliferation during heart regeneration. Hence, the specific ligand/receptor and the spatiotemporal regulation of the Nrg-Erbb axis appear to be more complicated in both embryonic and adult zebrafish. Interestingly, the nrg1 mutant heart exhibits a defect in cardiac nerve expansion and heart maturation at the juvenile stage despite normal cardiac structure formation⁴¹, suggesting its potential role in regulating cardiac function. Our findings demonstrate that the expression of nrg1 in embryonic endocardial cells is influenced by biomechanical cues and id2b activity. This signaling axis is essential for coordinating endocardium-myocardium interaction and establishing proper cardiac function.

Materials and methods

Zebrafish handling and lines

All animal procedures were approved by the Animal Care and Use Committee of the Zhejiang University School of Medicine. Embryonic and adult fish were raised and maintained under standard conditions at 28 °C on a 14/10 hour day/night cycle. The following zebrafish lines were used in this study: Tg(myl7:mCherry)^sd744, Tg(myl7:H2A-mCherry)^sd1245, Tg(kdrl:mCherry)^S89646, Tg(kdrl: nucGFP)^{y7 47}, Tg(Bre:d2GFP)^{mw30 48}, and Tg(actb2:Gcamp6s). To generate the id2b mutant, two short guide RNAs (sgRNAs) targeting exon 1 were generated using the MAXIscript T7 transcription kit (ambion, AM1314). The sgRNAs were as follows: sgRNA1: 5’ - GAAGGCAGTCAGTCCGGTG - 3’; sgRNA2: 5’ - GAACCGGAGCGTGAGTAAGA - 3’. The two sgRNAs, along with zCas9 protein, were co-injected into one-cell stage embryos. Embryos were raised to adulthood and crossed to wild-type zebrafish to obtain F₁ progenies. Through PCR analysis, a mutant line with a 157 bp truncation was identified.

The knock-in id2b:eGFP line was generated using a previously reported method²⁷. Briefly, three sgRNAs were designed to target the intron of id2b (sgRNA1: 5’ - GAGACAAATATCTACTAGTG - 3’; sgRNA2: 5’ - GTTGAACACATGACGATATT - 3’; sgRNA3: 5’ - GCACAACTTAGATTTCAAGT - 3’). Co-injection of each individual sgRNA with zCas9 protein into one-cell stage zebrafish embryos yielded varying cleavage efficiency. Since sgRNA2 displayed the highest gene editing efficiency, it was selected for subsequent studies. Next, a donor plasmid containing the sgRNA targeting sequence of the intron, exon 2 of id2b, and P2A-eGFP was generated. Co-injection of sgRNA, donor plasmid, and zCas9 protein into one-cell stage embryos led to concurrent cleavage of the sgRNA targeting sites in both the zebrafish genome and the donor plasmid (Figure 2A). Accordingly, eGFP fluorescence was observed in injected 24 hpf zebrafish embryos, indicating the incorporation of the donor. The insertion of the id2b-p2A-eGFP donor into the genome was confirmed by PCR analysis with primers recognizing target site or donor sequences, respectively (Figure 2A). Embryos with mosaic eGFP expression were raised to adulthood and crossed with wild-type zebrafish to obtain F₁ progenies. Overall, two founders were identified. The junction region of the F₁ embryos was sequenced to determine the integration sites. Although the two founders had slightly different integration sites in the intron, the expression pattern and fluorescence intensity of eGFP were indistinguishable between the two lines.

Morpholinos

All morpholinos (Gene Tools) used in this study have been previously characterized. tnnt2a MO (5’ - CATGTTTGCTCTGATCTGACACGCA - 3’)⁴⁹; ift88 MO (5’ - CTGGGACAAGATGCACATTCTCCAT - 3’)²⁹; bmp2b MO (5’ - ACCACGGCGACCATGATCAGTCAGT - 3’)⁵⁰; bmp4 MO (5’ - AACAGTCCATGTTTGTCGAGAGGTG - 3’)⁵¹; bmp7a MO (5’ - GCACTGGAAACATTTTTAGAGTCAT - 3’)⁵⁰; tcf3b MO (5’ - CGCCTCCGTTAAGCTGCGGCATGTT - 3’)⁵². For each morpholino, a 1 nl solution was injected into one-cell stage embryos at the specified concentrations: 0.5 ng tnnt2a MO, 2 ng ift88 MO, 0.5 ng bmp2b MO, 2 ng bmp4 MO, 4 ng bmp7a MO, and 1 ng tcf3b MO.

Small molecules treatment

To inhibit cardiac contraction, embryos were incubated in 1 mg/mL tricaine (Sigma, A5040) or 10 uM blebbistatin (MedChemExpress, HY13441) PTU-added egg water for 12-24 hours. In order to inhibit erbb2 signaling pathway, 10 uM AG1478 (Sigma, 658552) were used to treat 4 dpf larvae. To inhibit BMP signaling pathway, 10 uM Dorsomorphin (Sigma, P5499) was used to treat 24 and 36 hpf embryos.

In situ hybridization

Whole mount in situ hybridization was performed as previously described³⁴. The probes were synthesized using the DIG RNA labeling kit (Roche). The primers used for obtaining the id2b probe template were as follows: Forward 5’ - ATGAAGGCAGTCAGTCCGGTGAGGT - 3’; Reverse 5’ - TCAACGAGACAGGGCTATGAGGTCA - 3’.

Embryonic heart isolation and RNA-seq analysis

Hearts were isolated from embryos carrying the Tg(myl7: mCherry) transgene following an established protocol²⁶. A minimum of 1,000 hearts for each experimental group were manually collected under a Leica M165FC fluorescence stereomicroscope and transferred into ice-cold PBS buffer. After centrifugation at 12, 000 g for 2 min at 4°C, the supernatant was removed, and hearts were lysed in cold Trizol buffer (Ambion, 15596). Total RNA was extracted for subsequent quantitative real-time PCR or RNA-seq analysis.

Duplicate samples from control and Tricaine-treated embryonic hearts underwent RNA-seq. Raw sequencing reads were preprocessed to remove adapters and filter low-quality reads. Clean sequencing reads were then mapped to the zebrafish reference⁵³ using STAR with default parameters⁵⁴. Subsequently, gene quantification was carried out with RSEM⁵⁵. The gene expected count was applied to identify differential expression genes (DEGs), retaining only genes with counts per million (CPM) of 10 in at least two samples. DESeq2⁵⁶ was employed for differential expression analysis, and P-values were adjusted using BH correction. DEGs were defined as those with |log₂ fold change| ≥ 0.585 and an adjusted P-value < 0.1. The primary focus was on genes related to transcription regulation, and gene enrichment analysis was conducted using ClusterProfiler⁵⁷. To analyze DEGs in id2b^-/- and control embryonic hearts, we performed enrichment analysis with the R package EnrichR, dissecting the potential anatomy expression pattern and underlying phenotypes. We mainly focused on genes with the heart related phenotypes, including cardiac muscle tissue development, cardiomyocyte differentiation, heart morphogenesis and cardiac chamber development. All the analysis on identifying DEGs were batch corrected.

Quantitative real-time PCR analysis

After extraction from isolated embryonic hearts, 50 ng-1 ug of mRNA was reverse transcribed to cDNA using the PrimeScript RT Master Mix kit (Takara, RR036A). Real-time PCR was performed using the TB Green Premix Ex Taq kit (Takara, RR420A) on the Roche LightCycler 480. Expression levels of the target genes were normalized to actb1 as an internal control. All experiments were repeated three times. The following primer sets were used: id2b Forward 5’ - ACCTTCAGATCGCACTGGAC - 3’, Reverse 5’ - CTCCACGACCGAAACACCATT - 3’; nrg1 Forward 5’ - CTGCATCATGGCTGAGGTGA - 3’, Reverse 5’ - TTAACTTCGGTTCCGCTTGC - 3’; cacnα1c Forward 5’ - GCCCTTATTGTAGTGGGTAGTG - 3’, Reverse 5’ - AGTGTTTTGGAGGCCCATTG - 3’; tcf3a Forward 5’ - CCTCCGGTCATGAGCAACTT - 3’, Reverse 5’ - TTTCCCATGATGCCTTCGCT - 3’; tcf3b Forward 5’ – CCTTTAATGCGCCGTGCTTC - 3’, Reverse 5’ - GCGTTCTTCCATTCCTGTACCA - 3’; socs1a Forward 5’ - TCAGCCTGACAGGAAGCAAG - 3’, Reverse 5’ - GTTGCACAGGGATGCAGTCG - 3’; socs3b Forward 5’ - GGGACAGTGAGTTCCTCCAA - 3’, Reverse 5’ - ATGGGAGCATCGTACTCCTG - 3’; actb1 Forward 5’ - ACCACGGCCGAAAGAGAAAT - 3’, Reverse 5’ - GCAAGATTCCATACCCAGGA - 3’.

Co-immunoprecipitation (Co-IP) and western blot

Zebrafish tcf3b and id2b were overexpressed in 293T cell for 48 hours. The transfected cells were then collected and lysed using IP lysis buffer (Sangon Biotech, C500035) containing protease and phosphatase inhibitors (Sangon Biotech, C510009, C500017). For the IP experiment, anti-Flag antibody (Cell Signaling Technology, 14793, 1:100) was incubated with cell lysates overnight at 4°C. Pretreated magnetic beads were bound with the antigen-antibody complex for 4 hours at 4°C, followed by washing with IP lysis buffer three times. For western blot, samples were denatured at 95°C for 10 min, separated on a 5%-12% gradient gel. Proteins were then transferred to a PVDF membrane (Sigma, ISEQ00010). The membrane was blocked for 1 hour with 5% nonfat milk or 5% BSA (Sangon Biotech) dissolved in TBST and then incubated with primary antibodies (anti-FLAG, Cell Signaling Technology, 14793, 1:1000; anti-HA, Sigma, H3663, 1:1000) overnight at 4°C, followed by three 10-min TBST washes. HRP-conjugated secondary antibodies (Invitrogen, 31430, 31460) were incubated for 1 hour at room temperature, followed by three 10-min TBST washes. The detection of immunoreactive bands was performed with a chemiluminescent substrate (Thermo Scientific, 34577) and imaged using the Azure Biosystems 400.

Immunofluorescence

For immunofluorescence on adult zebrafish hearts, we fixed the hearts overnight in 4% paraformaldehyde at 4 ℃, followed by equilibration through 15% and 30% sucrose in PBS solution. The hearts were embedded and frozen in O.C.T. compound (Epredia, 6502), and 10 μm thick cryosections were prepared using a CryoStar NX50 cryostat. Immunofluorescence experiments were performed as previously described⁵⁸. For immunofluorescence on embryonic hearts, embryos were fixed overnight in 4% paraformaldehyde at 4℃, washed twice quickly in 100% methanol, and then dehydrated overnight at -20℃ in 100% methanol. Subsequently, rehydration was performed through a methanol gradient (100%, 75%, 50%, 25%, 10 min each), followed by three washes in PBST (1% PBS/0.1% Triton-X100, 10 min each). The embryos were treated with 10 μg/mL proteinase K diluted in PBST for 20 min at room temperature, re-fixed in 4% paraformaldehyde for 20 min, washed in PBST, and immersed in blocking solution (PBST/1% BSA/2% goat serum) for 1 hour at room temperature. Following this, the embryos were incubated in the primary antibody diluted in blocking solution overnight at 4℃. After washing in PBST, they were incubated in the secondary antibody (1:200) for 2 hours at room temperature. The primary antibody used was: Anti-GFP antibody (Santa Cruz Biotechnology, sc9996, 1: 200), anti-α-actinin antibody (Sigma, A7811, 1: 200). The secondary antibody used was: Anti-mouse IgG-Alexa 488 (Invitrogen, A11011, 1: 400). DAPI was used to stain cell nuclei.

Cardiac function analysis

To assess cardiac function in embryonic hearts, embryos were embedded in 1% low melting agarose, and heart contractions were recorded for 1 min using a Nikon Ti2 microscope at a rate of 25 frames per second. Fractional shortening and heart rate were measured as described previously³⁴. For cardiac contractile functions in adulthood, zebrafish were fixed on a sponge soaked with system water with the belly facing up and echocardiography was performed⁵⁹. Videos and images in color Doppler mode and B-mode were obtained using the Vevo1100 imaging system at a frequency of 50 MHz. Nikon NIS elements AR analysis and Image J software were employed for data extraction. To evaluate AV valve function, the ratio of inflow and outflow area in the same frame was quantified³².

Calcium imaging

At 120 hpf, embryos were treated with 10 mM 2,3-butanedione monoxime (BDM, Sigma, B0753) and mounted in 1% low melting agarose. Time-lapse images were acquired using a Nikon Ti2 microscope at a rate of 50 frames per second. Data were analyzed using Nikon NIS elements AR analysis software.

Intracellular action potential recording

Electrophysiology study was performed on adult zebrafish ventricles as previously described³⁴. Briefly, hearts were mounted in a chamber containing Tyrode’s solution: NaCl 150 mM, KCl 5.4 mM, MgSO₄ 1.5 mM, NaH₂PO₄ 0.4 mM, CaCl₂ 2 mM, Glucose 10 mM, HEPES 10 mM, pH was adjusted to 7.4. Glass pipettes with tip resistance 30-40 MΩ were filled with 3M KCl solution. Intracellular action potentials were recorded using an HEKA amplifier and pClamp10.3 software (Molecular Devices).

Histology and HE staining

Adult hearts were dissected and fixed overnight at 4°C in 4% PFA, followed by three PBS washes. Dehydration involved an ethanol gradient (70%, 80%, 95%, 100%, 100%, 30 min each), followed by three soaks in dimethylbenzene at 65°C, before embedding in paraffin. Sections of 5 um thickness were prepared using the Leica RM2235 manual rotary microtome for hematoxylin and eosin (HE) staining.

Injection of mRNA

The embryonic zebrafish cDNA library was used as a template to amplify the nrg1 and tcf3b fragment, which was then subcloned into the pCS2 vector. The vector was linearized using Not I restriction endonuclease, and mRNA was transcribed in vitro using the mRNA transcription kit (Ambion, AM1340). 100 pg of purified mRNA was injected into one-cell stage embryos.

Luciferase assay

The LCR (luciferase reporter) plasmid was generated by subcloning the 5′ UTR of nrg1 into the upstream region of renilla luciferase on the psiCheck2 plasmid. Following construction, 25 pg of the LCR plasmid was co-injected with either 100 pg of tcf3b mRNA or 1 ng of tcf3b MO into 1-cell stage zebrafish embryos. At 48 hpf, twenty embryos were gathered into one group and fully lysed. Subsequently, firefly and renilla luciferase activities were sequentially measured using a microplate reader with the dual luciferase reporter gene assay kit (Yeasen, 11402ES60), according to the manufacturer′s instructions. The experiment was independently replicated three times. The relative renilla luciferase activity, normalized by firefly luciferase activity, served as an indicator of nrg1 expression level under the influence of tcf3b overexpression or reduction.

Image processing and statistical analysis

Whole mount in situ hybridization images were captured with a Leica M165FC stereomicroscope. Live imaging of zebrafish embryos involved mounting anesthetized embryos in 1% low melting agarose (Sangon Biotech, A600015) and manually oriented them for optimal visual access to the heart. Confocal images were obtained with a Nikon Ti2 confocal microscope. Fluorescence intensity and cell number counting were processed using Nikon NIS Elements AR analysis and Image J software. Statistical analysis was performed using Graphpad prism 8 software. No statistical methods were used to predetermine sample size. Unpaired two-tailed Student’s t-tests was used to determine statistical significance. Data are presented as mean ± s.e.m, *P < 0.05 was considered to be statistically significant.

Acknowledgements

We thank Dr. Pengfei Xu for providing morpholinos. We also thank Dr. Jia Li for the support in generating the id2b:eGFP line.

Author Contributions

Shuo Chen and Peidong Han contributed to the study design. Shuo Chen, Jinxiu Liang, Weijia Zhang, Peijun Jiang, Wenyuan Wang, Xiaoying Chen and Yuanhong Zhou conducted experiments. Jie Yin analyzed RNA-sequencing data. Peng Xia, Fan Yang, Ying Gu, and Ruilin Zhang provided key reagents and analytic tool. Shuo Chen, Jie Yin, and Peidong Han prepared the manuscript.

Source of Funding

This work was supported by the National Key R&D Program of China (2018YFA0800102, 2023YFA1800602, 2018YFA0800501), and the National Natural Science Foundation of China (32170823, 31871462).

Disclosures

The authors declare that they have no conflict of interest.