Introduction

Hey2, together with Hey1, HeyL and their presumptive Drosophila homologue, dHey, is a member of a subfamily of hairy-related basic helix-loop-helix (bHLH) transcription factors1,2, that are implicated in cell fate determination and boundary formation3. Hey2 mutations in both humans and mice cause a variety of cardiac morphogenetic defects, as well as a cardiomyocyte maturation defect. In humans, non-synonymous sequence changes in Hey2 correlate with atrioventricular septal defects and other cardiac defects4,5, and Hey2 duplication contributes to both congenital heart defects and neurodevelopmental defects6. Similarly, Hey2 knockout mice display defects including atrioventricular valvular defects, pulmonary stenosis, Tetralogy of Fallot, tricuspid atresia, and abnormal cardiac hemodynamics7,8,9, indicating that Hey2 is an essential regulator of cardiac morphogenesis and cardiac function. Furthermore, genome-wide association study associated Hey2 mutation with Brugada syndrome, a rare cardiac arrhythmia disorder in humans10.

Despite profound cardiac defects caused by Hey2 mutation in both humans and mice, the current knowledge of the Hey2 expression pattern—that Hey2 is expressed in the compact zone of the myocardium—cannot fully explain this broad range of congenital defects. It is unclear how Hey2 enrichment in the ventricular compact myocardium could contribute to atrioventricular morphogenesis and semilunar valvular morphogenesis. Using an RNAscope, an assay that enables single mRNA molecule detection and is compatible with immunofluorescent staining, we were able to determine the cell type- and developmental stage-specific expression pattern of Hey2 at the level of single mRNA molecule. Specifically, we found that, in addition to the previously reported expression in the ventricular compact myocardium and interventricular septum, Hey2 is also expressed in the endocardial cells of the atrioventricular canal (AVC), the outflow tract (OFT), and at the base of trabeculae, as well as in pro-epicardial cells and epicardial cells. The expressional level and pattern of Hey2 examined by the RNAScope were also consistent with the expression pattern determined by the Hey2CreERT2 knock-in mouse line.

Hey2 functions in cell fate specification and cardiomyocyte maturation. Gridlock/Hey2 is involved in adjudicating an arterial versus venous cell fate decision during the assembly of the first embryonic artery in zebrafish11. In mice, cardiac specific Hey2 knockout hearts display ectopic atrial gene expression12,13. Hey2 null cardiomyocytes displayed abnormal mitochondria, abnormal accumulation of glycogen particles, and disorganized myofibrils based on transmission electron microscope analysis8. Expression levels of β-MHC and ANF genes in the Hey2 knockout are increased8. The phenotypes of Hey2 knockout heart indicate that Hey2 might play roles in cardiomyocyte differentiation and maturation.

Previous work has suggested that trabecular cardiomyocytes, which take the major responsibility for pumping during the early stages of cardiac development, are more differentiated than the cardiomyocytes of the compact zone14. The facts that Hey2 is expressed in the cardiomyocytes of the compact myocardium and that Hey2 disruption results in a larger and wider sarcomere suggest that Hey2 might repress cardiomyocyte maturation and might serve as a marker for less differentiated and/or less mature cardiomyocytes. Previously, single cell lineage tracing studies revealed that asymmetric distribution of Hey2 during cardiomyocyte oriented cell division might contribute to the differential expression of Hey2 in compact and trabecular cardiomyocytes15. However, the signaling pathways that regulate the Hey2 expression pattern and its asymmetric distribution in a perpendicular oriented cell division during trabecular initiation are unknown. To identify the signaling pathways that regulate the differential expression levels of Hey2 in the compact zone and the trabecular zone, we assessed the expression of Hey2 in various mutants that are defective in several pathways implicated in trabeculation and compaction, including Nrg1/ErbB2/4 and Notch1. We found that these mutants showed decreased expression levels of Hey2, but displayed normal enrichment and repression of Hey2 in the compact zone and trabecular zone, respectively. FGF ligand stimulation changes the expression pattern of Hey2 in the ventricles temporally in both in vivo and ex vivo models, resulting in more trabeculation. We further found that FGF2 regulates the expression level of Hey2 in a Notch1 dependent manner.

Results

Hey2 is expressed in the endocardial cells of the AVC, OFT and the base of trabeculae in addition to its known localization to the compact zone

Previous in situ hybridization (ISH) studies using whole embryos or tissue sections demonstrated that Hey2 is expressed in the compact zone, AV cushion, and ventricular septum1,16. However, harsh treatment in traditional ISH protocols prevents successful simultaneous immunofluorescence staining, which is necessary to accurately identify the cell types that express Hey2. In this study, we used the RNAscope, which can detect single mRNA molecules and allows for ISH and immunofluorescence staining in the same samples15, and a Hey2CreERT2 knockin mouse line17, to determine the expression level and pattern of Hey2 and found a broader expression pattern of Hey2 than previously reported. In both left and right ventricles, Hey2 is enriched in the outer compact zone (OCZ), and is weakly expressed in the inner compact (ICZ) and trabecular zones (Fig. 1a,a1,b and b1). In addition to its expression in the outer compact zone, Hey2 is robustly expressed in endocardial cells of the atrioventricular canal (AVC) (Fig. 1a and a2) and outflow tract (OFT) (Fig. 1b and b2) at E9.5. Furthermore, Hey2 is weakly expressed in the endocardial cells at the base of trabeculae, and the endocardial cells that closely abut the cardiomyocytes contains 6.25 dots/section/cell (n = 6), which is relatively greater than the Hey2 expression in endocardial cells that are not adjacent to the cardiomyocytes, which contain 3.20 dots (n = 6) (Fig. 1a and a1). We also examined the Hey2 expression pattern by staining with estrogen receptor (ESR) using E9.5 hearts from Hey2CreERT2 knockin embryos, and consistently Hey2 was strongly expressed in the compact zone and in the endocardial cells of the AVC and OFT (Fig. 1c,c2 and c3), and weakly expressed in the cardiomyocytes of trabecular zone and the ventricular endocardial cells (Fig. 1c). However, via the ESR staining, we did not observe the differential expression of Hey2 in outer and inner compact zone, and its differential expression in ventricular endocardial cells at different regions cannot be distinguished (Fig. 1a–c), which might be due to the detection threshold of ESR proteins distingguishable by antibody-mediated immunostaining.

Figure 1
figure 1

Hey2 is expressed in the endocardial cells of the AVC, OFT and the base of trabeculae, in addition to its known localization to the compact zone. (a and b) Hey2 expression pattern in E9.5 control heart. Box a1,a2 and Box b1,b2 in Figures a,b were zoomed to a1,a2 and b1,b2 on the right. White and green arrows in a1,b1 point to the compact zone (OCZ) and inner compact zone (ICZ), respectively. Red and yellow arrows in a1 point to endocardial cells that attach or do not attach to cardiomyocytes, respectively. The red arrow in b2 points to endocardial cells of the OFT. Blue arrow in a points to pro-epicardial organ (PEO). (c) The expression pattern of Hey2 via ESR staining using the Hey2CreERT2 mouse line. White and green arrows in c1 point to the outer and inner compact zone, respectively. Red and yellow arrows in c1 point to endocardial cells that attach or do not attach to cardiomyocytes, respectively. The red arrow in c1 points to endocardial cells of the ventricle. LV: Left ventricle; RV: Right ventricle; AVC: Atrioventricular canal; EC: Epicardial cell; PEO: Pro-epicardial organ; OFT: Outflow tract; IFT: Inflow tract. Scale bars in (a–c) are 50 μm and 10 μm in zoomed boxes. Representative pictures from at least three embryos of each genotype at different ages were shown.

Full size image

Hey2 is expressed in the mesenchymal cells of the AVC and OFT

Consistently with its expression pattern at E9.5, Hey2 is highly expressed in the outer compact zone, endocardial cells of the AVC and OFT, and weakly expressed in ventricular endocardial cells in E10.5 hearts (Fig. 2a–c). The endocardial cells that closely abut against the cardiomyocytes display more Hey2 compared to the cells that are not adjacent to cardiomyocytes (Fig. 2a and a1). Hey2 is also expressed in the cells between endocardial cells and cardiomyocytes, which are mesenchymal cells derived from endocardial cells of the AVC (Fig. 2b and b1) and OFT (Fig. 2c and c1), at a lower intensity compared to the endocardial cells of the AVC and OFT. Via the Hey2CreERT2, we found that Hey2 is also expressed in the endocardial cells and mesenchymal cells of AVC and OFT cushion (Fig. 2d and d1). The expression of Hey2 in these endocardial cells and cushion tissues might suggest that Hey2 is involved in endocardial cell epithelial-mesenchymal transition (EMT) and that Hey2 disruption leads to the valvular morphogenesis defects previously reported in Hey2 knockouts16. In addition to its expression in the aforementioned endocardial cells, we also found that Hey2 is weakly expressed in pro-epicardial cells and epicardial cells (Figs 1a and 2d).

Figure 2
figure 2

Hey2 is expressed in the mesenchymal cells of AVC and OFT. (a–c) Hey2 expression pattern in an E10.5 control heart via RNAScope. The box regions in (a–c) are zoomed and put to the right. The red arrow in a1 points to a cell that closely attaches to cardiomyocytes and displayed more Hey2 molecules than a cell in a1 that does not attach to cardiomyocytes, as indicated by a white arrow. White arrows in b1 and c1 point to mesenchymal cells that are derived from the AVC and OFT, respectively. (d) The expression pattern of Hey2 via ESR staining using the Hey2CreERT2 mouse line at E11.5. Green arrows in d point to epicardial cells and white arrows in d1 point to the mesenchymal cells that are derived from the endocardial cells of AVC. LV: Left ventricle; RV: Right ventricle; AVC: Atrioventricular canal; EC: Epicardial cell; PEO: Pro-epicardial organ; OFT: Outflow tract. Scale bars in (a–d) are 50 μm and 10 μm in zoomed boxes. Representative pictures from at least three embryos of each genotype at different ages were shown.

Full size image

Notch signaling regulates Hey2 expression in ventricular endocardial cells

Previous work has shown that Hey2 expression is altered in Dll1 or Notch1 knockout mice during somitogenesis18,19. However, during cardiac morphogenesis, studies have demonstrated that Notch Intracellular Domain (NICD) overexpression activated Hey1 but not Hey2 expression in the heart using the NICD transgenic line20. Furthermore, Notch2 knockout mice displayed normal expression patterns of Hey1 and Hey2 in the myocardium21 and Rbpjk global knockout embryos did not alter the expressional level of Hey2 in the heart based on Q-PCR22. These reports indicate that Notch signaling does not regulate Hey2 expression in the heart. However, it is possible that Notch signaling regulates Hey2 expression in some cell types such as endocardial cells that cannot be detected by whole heart RNA quantification and traditional ISH. To test this possibility, we examined the expression of Hey2 in different mutants via RNAscope coupled with immunofluorescence staining. We used Nfatc1Cre/+, which is specifically expressed in the endocardial cells23, to mediate Notch1 deletion, and found reduced NICD expression (Fig. 3a and b). Nfatc1Cre/+; Notch1fl/fl knockout hearts showed a slight reduction in Hey2 in the endocardial cells of the AVC at E9.5, but did not show obvious reduction of Hey2 in the compact zone (Fig. 3c,c1,d and d1). Due to the incomplete deletion of Notch1 at early stages such as E9.5 (Fig. 3a), we examined the knockouts at E10.5 and consistently found that Hey2 expression was reduced in the endocardial cells of AVC and also in the endocardial cells that closely interact with cardiomyocytes at the base of trabeculae, while Hey2 expression in the compact zone was not obviously altered (Fig. 3e,e1,f and f1). Hey2 expression in the endocardial cells of the OFT was not changed in the Nfatc1Cre/+; Notch1fl/fl (Data not shown). The difference in expression of Hey2 between control and knockout hearts was not as large as expected, which might be due to late expression of Cre in the Nfatc1Cre/+ mouse (Fig. 3a,b). Therefore, we used Tie2Cre, which is expressed starting at E7.5, to delete Notch1 in endocardial cells. This knockout died at around E10.5 and displayed a trabeculation defect as previously reported24. Hey2 expression in the endocardial cells of this knockout was significantly reduced at E9.5 (Fig. 3g,g1,h and h1). To determine if other Notch receptors in addition to the Notch1 receptor regulate Hey2 in endocardial cells, we examined the expression of Hey2 in Tie2Cre-mediated Rbpjk knockout hearts. Surprisingly, the Hey2 expression level in the endocardial cells of the Tie2Cre; Rbpjkfl/fl knockout was not obviously different from that of the Tie2Cre; Notch1fl/fl knockout (Suppl. Figure 1a,b), suggesting that Notch1 is the major receptor in endocardial cells that regulates Hey2 expression. The incomplete abolishment of Hey2 expression in the endocardial cells, especially those of the AVC in Tie2Cre; Rbpjkfl/fl knockouts, indicates that an Rbpjk-independent signaling pathway also regulates Hey2 expression in the endocardium. We examined total Hey2 in control and Tie2Cre; Rbpjkfl/fl knockout hearts via Q-PCR and found that the knockouts displayed higher levels of Hey2 (Suppl. Figure 1c), possibly due to a relatively smaller trabecular zone and a larger compact zone, where Hey2 is enriched.

Figure 3
figure 3

Notch signaling regulates Hey2 expression in ventricular endocardial cells. (a and b) N1ICD staining in control and Nfatc1cre; Notch1fl/fl hearts at E9.5 and E10.5. White arrows show reduced NICD expression in a knockout compared to control in a&b. (c and d) Hey2 expression in control and Nfatc1cre; Notch1fl/fl hearts at E9.5. Endocardial cells pointed by white arrows in d and d1 show reduced Hey2 expression compared to c&c1. (e and f) Hey2 expression in control and Nfatc1cre; Notch1fl/fl hearts at E10.5. Endocardial cells pointed by white arrows in f1 show reduced Hey2 expression compared to e1. (g and h) Hey2 expression in control and Tie2cre; Notch1fl/fl hearts at E9.5. White arrows in f1 and h1 indicate reduced expression of Hey2 in ventricular endocardial cells and AVC endocardial cells compared to their controls (e1 and g1). The box regions in (c–h) are zoomed to separate pictures that are labeled with the same name of the box. Scale bars in (a–h) are 100 μm and 10 μm in zoomed boxes. Representative pictures from at least three embryos of each genotype at different ages were shown.

Full size image

Notch signaling regulates the expression level but not the pattern of Hey2 in the myocardium

Echocardiographic studies of Hey2 knockout mice revealed reduced left ventricular contractile function8. It is speculated that the presence of abnormal cardiomyocytes accounts for the decreased left ventricle fractional shortening, indicating the importance of Hey2 in compact zone cardiomyocyte maturation. The signaling pathways that regulate the enrichment of Hey2 in the compact zone and the low abundance in the trabecular zone have not been reported before. Our data showed that Notch1 or Rbpjk deletion in endocardial cells via Nfatc1Cre/+ or Tie2Cre did not affect the Hey2 enrichment in the compact zone (Fig. 3), indicating that Notch signaling in the endocardium does not regulate Hey2 in the myocardium non-autonomously—as it does with many other factors in the myocardium24,25. We then investigated if Notch1 and Rbpjk regulate Hey2 in the compact zone autonomously by examining Hey2 expression in the myocardium of both Nkx2.5Cre/+; Notch1fl/fl and Nkx2.5Cre/+; Rbpjkfl/fl knockouts. Nkx2.5Cre/+ mediated Notch1 or Rbpjk deletion did not change the Hey2 expression pattern between the control and knockout at both E9.5 and E12.5 (Fig. 4a–d, Suppl. Figure 2a,b). We then quantified the number of Hey2 molecules within cells in different layers of the myocardium with the enrichment index of Hey2 in different layers determined by their relative numbers of Hey2 molecules (Fig. 4a,e). We found that the intensity of Hey2 expression in the compact zone of the knockout is reduced compared to the control in both knockouts (Fig. 4a–e), while its intensity in other regions, such as endocardial cells, was not obviously affected (Fig. 4a–d). The Nkx2.5Cre/+; Rbpjkfl/fl heart showed slightly reduced Hey2 expression based on Q-PCR (Suppl. Figure 2c). We also examined whether the Notch1 deletion would disrupt the expression pattern of Hey2 in the compact and trabecular zone. The Hey2 expression pattern in the knockouts was compared to that of the controls by linear regression comparison, and we found that the expression pattern among the control and knockouts are not significantly different based on the slopes (Fig. 4e,f). Consistently, when ex vivo cultured hearts were treated with DAPT, a γ-secretase inhibitor which prevents Notch activation, the expression pattern of Hey2 was not changed, although the level was slightly reduced (Data not shown). These data suggest that Notch signaling mildly regulates the expression level of Hey2, but not its expression pattern in the myocardium.

Figure 4
figure 4

Notch signaling in myocardium regulates the expression level but not the pattern of Hey2 in the myocardium. (a and b) Hey2 expression in E9.5 control and Nkx2.5cre; Notch1fl/fl hearts. The white numbers in a1 indicate the cell layers from the 1st to the 5th layer of ventricular myocardium and the number of Hey2 expression in different layers were quantified and presented in (e). (c and d) Hey2 expression in E9.5 control and Nkx2.5cre; Rbpjkfl/fl hearts. The box regions in (a–d) are zoomed to separate pictures that are labeled with the same name of the box. (e) Hey2 enrichment index. The average number of Hey2 molecules from 10 cells from each layer was quantified, and their ratio to layer #1, which is considered as 1, was presented in (e) Quantification of Hey2 expression at different layers in Nkx2.5cre; Notch1fl/fl, and Nkx2.5 cre; RBPJkfl/fl hearts were shown. Expression level of Hey2 in different layer of cells of different genotypes were quantified and compared via student’s t-test. (f) The expression patterns of Hey2 in different treatments were analyzed by linear regression comparison and they were not significantly different from the control based on the lineage regression comparisons. Scale bars in (a–h) are 100 μm and 10 μm in zoomed box. Representative pictures from at least three embryos of each genotype or treatment at different ages were shown.

Full size image

The Nrg1/ErbB2 pathway does not regulate Hey2 expression pattern

We were then interested in identifying the signaling pathways that regulate the expression pattern of Hey2 in the compact zone. Previous work has shown that global deletion of the receptors or ligand of the Nrg1/ErbB2,4 pathway abrogates trabecular formation26,27,28. We used the Nkx2.5Cre/+ line to delete ErbB2 (Fig. 5a,b). The Nkx2.5Cre/+; ErbB2fl/fl knockout hearts displayed a lower proliferation rate and thinner compact zone (Data not shown). Previous work has shown that Notch1 signaling regulates cardiomyocyte differentiation through Nrg1/ErbB2,424. We hypothesized that Nrg1/ErbB2,4 might regulate Hey2 expression in the compact zone. We examined the expression of Hey2 in the Nkx2.5Cre/+; ErbB2fl/fl heart and found that the Hey2 expression level was slightly reduced (Suppl. Figure 3a), but its expression pattern and relative enrichment in the compact zone was not affected (Fig. 5c,d), suggesting that Nrg1/ErbB2,4 does not regulate the Hey2 expression pattern.

Figure 5
figure 5

Nrg/ErbB2,4 pathway does not regulate Hey2 expression pattern. (a and b) ErbB2 deletion in Nkx2.5cre; ErbB2fl/fl at E9.5 indicated by the absence of ErbB2 immunostaining in the knockout. (c and d) Hey2 expression in E9.5 control and Nkx2.5cre; ErbB2fl/fl hearts. The box regions in c,d are zoomed to separate pictures that are labeled with the same name of the box. Scale bars in (a–d) are 100 μm. Representative pictures from at least three embryos of each genotype were shown.

Full size image

FGF signaling regulates Hey2 expression pattern in the ventricles temporally

The studies above showed that the signaling pathways such as Notch and Nrg1 signaling that regulate trabeculation and/or compaction control the expression levels but not the expression pattern of Hey2. Indeed, the signaling pathway that regulates Hey2 expression pattern in the myocardium is unknown. In other developmental processes, such as somitogenesis, Notch regulates Hey219, while in the organ of Corti, FGF signaling regulates Hey2 expression via a Notch-independent mechanism to maintain pillar cell fate29. During cardiac development, FGF2, FGF9, and FGFR1 are the major fibroblast growth factors and receptor that are expressed in mouse and chick ventricles30,31,32. Fgfr1 and Fgfr2 cardiac-specific double knockout hearts show increased expression of sarcomeric actin and premature differentiation, indicating that FGF signaling functions to maintain the primitive status of cardiomyoblast32. Whether FGF signaling contributes to the relative immaturity of cardiomyocytes in the compact zone and more mature cardiomyocytes in the trabecular zone is not clear. Our RNAscope and immunofluorescence staining show that Fgfr1 is expressed in the myocardium and endocardium (Suppl. Figure 4a). Considering that multiple FGF receptors and FGF ligands are expressed in the heart, as well as the redundant functions of these ligands and receptors, we examined the effects of FGF signaling on Hey2 expression pattern by stimulating the hearts with FGF2 in vivo and ex vivo, instead of using a loss-of-function approach33. Ex vivo cultured hearts treated with FGF2 ligand for 2 hours showed an increased expression of Hey2 compared to the control based on the number of Hey2 mRNAs per cell (Fig. 6a,b). The FGF2 treated hearts also showed enrichment of Hey2 in both outer compact and inner/trabecular zones compared to the control (Fig. 6b,c). We then further examined if FGF2 stimulation in vivo would increase the Hey2 expression level and change its expression pattern by injecting FGF2 at a concentration of 20ng/gram body weight to the pregnant females when the embryos were at E8.5. We found that FGF2 stimulation increased the expression of pAkt and pErk compared to the vehicle treated control (Suppl. Figure 4b–e), indicating an effective FGF2 stimulation. We then quantified the Hey2 expression level and examined the enrichment of Hey2 in the compact zone in both in vivo and ex vivo models. We found that FGF2 stimulation increased the expression level of Hey2 and also changed the enrichment of Hey2 in the compact zone (Fig. 6c–f). The FGF2 induced expression pattern change was not maintained at later stages, as the Hey2 expression pattern in hearts treated by injecting FGF2 into the pregnant females for three consecutive days was not different from the expression pattern in vehicle treated controls. However, we did find that the FGF2 stimulation promotes trabeculation (Suppl. Figure 4f).

Figure 6
figure 6

FGF signaling regulates Hey2 expression pattern in the ventricles temporally. (a and b) FGF2 stimulation increases Hey2 expression in the inner compact zone as indicated by green arrows in the ex vivo cultured heart. (c) Quantification and comparison of Hey2 expression in cardiomyocytes at different layer of in vivo and ex vivo cultured E9.5 hearts with and without FGF2 stimulation. (d and e) FGF2 stimulation in vivo increases Hey2 expression in the inner compact zone as indicated by green arrows compared to the vehicle treated hearts. (f) Hey2 expression patterns of FGF2 stimulated hearts are significantly different from control hearts based on the linear regression comparison assay. Scale bars in (a–e) are 100 μm. Representative pictures from at least three embryos of each treatment were shown.

Full size image

FGF2 regulates Hey2 expression level in a Notch1 dependent manner

We then used mouse embryonic fibroblasts (MEF) from ROSA26CreERT2; Notch1fl/fl embryos to further determine whether FGF2 and Notch1 regulate Hey2 expression level (Fig. 7a–d, Suppl. Figure 5). ROSA26CreERT2; Notch1fl/fl MEF cells were either treated with vehicle or 4-Hydroxytamoxifen (4-OHT) to delete Notch1. qRT-PCR and western blot analyses indicated that the deletion of Notch1 is efficient (Fig. 7a,c,d). We stimulated the control and Notch1 null MEFs with FGF2, and the stimulation resulted in the higher level of pErk as expected33 (Fig. 7c). We then determined whether FGF2 and Notch1 regulate the expression level of Hey2. Notch1 deletion decreased the expression of Hey2 based on Q-PCR and Western blot analyses (Fig. 7b–d). FGF2 stimulation promotes Hey2 expression in control MEF, but not in the Notch1 null MEF (Fig. 7b–d). These data suggest that Notch1 regulates Hey2 expression and FGF2 regulates Hey2 in a Notch1 dependent manner.

Figure 7
figure 7

FGF signaling regulates Hey2 expression in a Notch1 dependent manner. (a) Notch1 is deleted efficiently in ROSA26CreERT2; Notch1fl/fl MEFs via 4-OHT treatment. (b) Q-PCR analysis shows that Hey2 mRNA level is increased with FGF2 stimulation in control MEF cell and decreased in Notch1 null MEF cell, while FGF2 stimulation doesn’t cause significant change in Notch1 null MEFs compared with control based on student’s t-test. (c and d) Notch1 is deleted efficiently in ROSA26CreERT2; Notch1fl/fl MEFs by 4-OHT treatment. p-Erk protein level decreased in Notch1 null MEFs. FGF2 stimulation increases p-Erk protein level in control MEFs but not in Notch1 null MEFs. FGF2 stimulation promoted Hey2 protein level in control MEFs but not Notch1 null MEFs. An unpaired, two-tailed student’s t-test was used for statistical comparison. The Representative pictures from three repeats were shown and the original pictures of all the blots are shown in Suppl. Figure 5.

Full size image

Discussion

Hey2 expression in endocardial cells of AVC, OFT and trabeculae might explain the valvular defects of Hey2 mutants

Hey2, Hey1, and HeyL comprise a subfamily of mammalian hairy/enhancer of split-related basic helix-loop-helix genes with a putative Drosophila homologue, dHey34, and a zebrafish homologue, Gridlock11. Hey2 displays a spatiotemporal expression pattern and is expressed in the somite, heart, craniofacial region, and nervous system at early embryonic stages34. Detailed examination of Hey2 expression in mouse heart shows that Hey2 is expressed in the ventricular septum and compact zone but not in the atrium at all. In this study, we used RNAscope and Hey2CreERT2 knockin mouse line to further characterize Hey2 expression level and pattern and found that Hey2 is also expressed in the endocardial cells of the AVC, OFT and at the base of trabeculae (Figs 1 and 2). With the RNAscope and immunofluorescent staining, we can infer that the expression of Hey2 detected in the cardiac OFT and aortic arch arteries1,34 was actually in the endothelial cells (Figs 1 and 2). The expression pattern of Hey2 via Hey2CreERT2 knockin mouse line is consistent with the results from RNAscope. The expression of Hey2 in the endocardial cells of the AVC and OFT and mesenchymal cells derived from those endocardial cells might explain the atrioventricular valve defects, pulmonary stenosis, tetralogy of Fallot, and tricuspid atresia found in the Hey2 knockout mice7,8. The expression of Hey2 in the ventricles but not in atria might explain why the Hey2 knockouts display ectopic atrial gene expression in the myocardium12,13. Hey2 null cardiomyocytes display malformed mitochondria, abnormal accumulation of glycogen particles, and disorganized myofibrils based on transmission electron microscope analysis8, suggesting that Hey2 might regulate cardiomyocyte differentiation and maturation, and is therefore critical for normal cardiac function. In summary, using RNAscope and Hey2CreERT2 knockin line, we have determined that the Hey2 expression pattern is broader than previously reported and the expression of Hey2 in the endocardial cells of the AVC and OFT might explain the variety of defects found in Hey2 knockout mice and human patients that bear Hey2 duplication or mutations.

Notch regulates Hey2 expression in endocardial cells

Despite Hey2’s high homology to other Hey family members18, the notion that Notch signaling regulates Hey2 expression is controversial. During mouse somitogenesis, Hey2 expression was altered in both the Dll1 and the Notch1 knockout mice19. NICD overexpression up-regulates Hey2 in an Rbpjk dependent manner in smooth muscle cells35, indicating that Notch signaling regulates Hey2 expression. However, in the heart, NICD overexpression activated Hey1 but not Hey2 expression in the NICD transgenic line20, the Notch2 knockout displayed a normal expression pattern of Hey1 and Hey221, and Rbpjk global knockout embryos did not alter the expression of Hey2 in the heart based on Q-PCR22. In addition, in the Numb and Numblike double knockout heart, in which Notch signaling was upregulated based on a Notch transgenic reporter and N1ICD protein levels36,37,38, there was no obvious Hey2 up-regulation based on mRNA deep sequencing and Q-PCR. These reports suggest that Notch does not regulate Hey2 in the heart. However, it is possible that Notch regulates Hey2 expression in some cell types that cannot be detected by total RNA in the heart or by traditional ISH. In this study, equipped with RNAscope coupled with immunofluorescence staining, we clearly demonstrated that Hey2 is expressed in endocardial cells as we discussed above. We used Nfatc1Cre or Tie2Cre to delete Notch1 or Rbpjk in the endocardial cells and found that Hey2 expression was reduced in endocardial cells that burrowed into the myocardium and closely contacted the cardiomyocytes. The expression of Hey2 in the endocardial cells of the AVC but not of the OFT was slightly reduced in the knockout compared to the control, indicating that Notch signaling at least partially regulates Hey2 in the endocardial cells of the AVC and those that contact the myocardium. This is consistent with the report that Notch1 is required for proper development of the semilunar valves and cardiac outflow tract39,40. Furthermore, the Hey2 expression in the endocardial cells attached to the myocardium or at the base of the trabecula is higher than that of other endocardial cells, which is consistent with the previous report that NICD is highly expressed in the endocardial cells at the base of the trabecula24. The Hey2 expression in these endocardial cells and its potential to be a target of Notch signaling might explain the trabeculation defect of the Hey1 and Hey2 double knockout41. Also, the higher expression of Hey2 and NICD in the endocardial cells attached to cardiomyocytes24 indicates a potential Notch signaling pathway from myocardium to endocardium; however, further work needs be done to identify the Notch ligand that regulates Hey2 expression in the endocardial cells.

Hey2 is essential for cardiomyocyte maturation in the left ventricle

The essential functions of Hey2 in the heart are revealed by the Hey2 loss of function studies. The Hey2 null hearts display a dilated left ventricular chamber with markedly diminished fractional shortening of the left ventricle8. Hey2 null cardiomyocytes displayed abnormal mitochondria, abnormal accumulation of glycogen particles, and disorganized myofibrils based on transmission electron microscope analysis8. Hey2 knockouts show ectopic expression of Tbx5, ANF, and Cx40 in the compact zone12,13. Furthermore, the expression levels of β-MHC and ANF genes in the Hey2 knockout are increased8,12, suggesting that Hey2 is required to repress atrial gene expression and to maintain ventricular identity. Other studies show that Hey2 null cardiomyocytes display altered action potentials and mild conduction system expansion42, and Hey2 null cardiomyocytes display defective myocardial calcium release and diminished fractional shortening in the setting of normal calcium stores and calcium sensitivity43. It was also reported that the Hey1 and Hey2 double knockout displayed a trabeculation defect41, and that Hey1 and Hey2 might function redundantly to prevent trabecular cardiomyocyte apoptosis41. These studies indicate that Hey2 might be involved in cardiomyocyte maturation during heart development. However, the functions of Hey2 enrichment in the compact zone but not in the trabecular zone remain unknown.

Pathways that regulate asymmetric distribution of Hey2 to compact zone and trabecular zones are unknown

Trabecular and compact cardiomyocytes display different features with trabecular cardiomyocytes exhibiting a lower proliferation rate and being more molecularly mature than cardiomyocytes of the compact zone44. Trabecular and compact zones can be distinguished by the differential expression of many genes. For example, p57, Irx3, BMP10, Sphingosine 1-phosphate receptor-1 and Cx40 are highly expressed in the trabecular zone, while Tbx20, Hey2, and N-Myc are highly expressed in the compact zone14,15,45,46,47,48. Our previous work has shown that asymmetric expression of Hey2 and Bmp10 occurs after but not before cytokinesis in perpendicularly oriented cell division, and that their asymmetric distributions are due to the differences in geometric location between the two daughter cells15. The daughter cell closer to cardiac jelly displays less Hey2 and more BMP10 compared to the other daughter cell that is relatively closer to the surface of the heart, indicating that potential instructive cues for trabecular regional specification might lie in the cardiac jelly and endocardium15. Indeed, the signaling pathway that regulates the asymmetric distribution of Hey2 or other genes between the two zones is unknown. We have showed that Numb Family Proteins regulate Irx3 expression pattern in the heart, but whether Numb also regulates Hey2 is not clear36,37. Previous work has shown that at later stages, direct Notch ligand/receptor interactions might be involved in the regulation of trabecular cardiomyocyte differentiation, as overexpression of glycosyltransferase manic fringe in the endocardium can cause abnormal expression patterns of Hey2 at later gestation stage49. However, the deletion of Notch1 or Rbpjk at early stage in endocardial cells or myocardium during trabecular formation did not affect Hey2 asymmetric distribution. Surprisingly, the signaling pathways that regulate trabecular initiation, including NRG1/ErbB2,426-28, Notch signaling24, Hand225 and YAP50,51, regulate the expression level of Hey2, but not the expression pattern of Hey2. This indicates that trabecular initiation and trabecular specification are two different processes and are regulated by different pathways. Instead, we found that FGF2 stimulation regulates Hey2 asymmetric distribution in myocardium, and results in over-trabeculation, but the mechanisms of how FGF2 regulates Hey2 and trabeculation will need further studies.

Methods

Experimental animals

Mouse strains Notch152 was purchased from Jackson Lab. Nfatc1Cre/+, which is specifically expressed in the endocardial cells23, was used to delete Notch152 and Rbpjk53 in endocardial cells. Tie2-Cre54, which is specifically expressed in endothelial cells, was used to delete Notch1 and Rbpjk in endothelial cells. Nkx2.5Cre/+ mice55, a gift from Dr. Robert Schwartz, were used to delete Rbpjk, Notch1 and ErbB256. Male heterozygous with Nkx2.5Cre/+ were setup to cross homozygous females to generate knockouts, e.g., Nkx2.5Cre/+; Notch1fl/+ males were mated to Notch1fl/fl females to generate Nkx2.5Cre/+; Notch1fl/fl. Their sibling embryos were designated as controls. Dr. Bin Zhou at Shanghai Institutes, Chinese Academy of Sciences, Shanghai, China, provided the Hey2CreERT2 mouse17. To isolate embryos, the pregnant females at the specified age will be euthanized by CO2 inhalation and cervical dislocation. CO2 with a filling rate of about 20% of the chamber volume per minute was added to the existing air in the chamber to achieve a balanced gas mixture to fulfill the objective of rapid unconsciousness with minimal distress to the animals. All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) at Albany Medical College and performed according to the NIH Guide for the Care and Use of Laboratory Animals.

RNAscope: In Situ hybridization and immunofluorescence staining

In Situ hybridization (ISH) and immunofluorescence staining (IFS) were performed according to the protocol of the RNAscope® Chromogenic (RED) Assay (Cat. No. 310036), which can detect single mRNA molecules15. Briefly, 24 hours after fixation, the embryos were frozen embedded in OCT compound. Sections of the frozen embedded samples were processed following the step-by-step protocol of the kit. The expressional level of mRNA in each cell was based on the number of mRNA molecules or signal intensity detected using the confocal scanned pictures, and three sections for each genotype were quantified. The number of mRNA molecules in cells at comparable locations was also quantified, and only the control and knockout samples that came from the same litter and underwent the same experiments were quantified and compared.

Quantification of Hey2 mRNA molecules in cardiomyocytes at different layers

Cardiomyocytes at different layers starting from the outer compact layer, to the inner compact zone and trabecular zone were selected to quantify the number of Hey2 molecules. For each sample, we quantified 3 sections, and for each section, 12 cells at 3 different regions per layer were quantified. The mean value of the 36 cells in each layer was presented. Only control and mutants that came from the same litter and underwent the same experiments were quantified and compared. Each experiment was repeated three times.

MEFs isolation and western blot analysis

Mouse embryonic fibroblasts (MEF) were isolated from E13.5-E14.5 embryos with the genotype of Notch1fl/fl; ROSA26CreERT2. Cultured MEFs were infected treated with 4-OHT as indicated in different experiments for 36 hours, and then starved with serum free medium overnight. The cells were then stimulated with FGF2 ligand for 15 minutes. After the stimulation, the cell lysate was harvested. Lysates from MEFs of different genotypes and treatment were processed for western blot. Antibodies used for western blots were listed above. The experiments were repeated three times and quantification data was shown in the figure.

Paraffin and frozen section immunohistochemistry (IHC)

Immunohistochemistry (IHC) and Immunofluorescence (IF) were performed as previously published57,58,59. The following primary antibodies were used: BrdU (1:50, Becton Dickinson, 347583), PECAM (1:50, BD Pharmingen), MF20 (1:100, Developmental Studies Hybridoma Bank (DSHB)), NICD (1:50, Cell Signaling, 4147 S), ErbB2 (1:100, Cell Signaling, 2165 S), ESR (Abcam, pre-diluted), Endomucin (1:50, SCBT), p-Akt (1:50, Cell Signaling, 9271 S) and p-Erk (1:100, Cell Signaling, 4376 S).

Ex vivo culture system

The ex vivo culture protocol was developed by Dr. C.P. Cheng and was utilized as reported36. Briefly, embryos lacking the head and tail were cultured in medium containing FGF2 at 2ng/gram body weight or vehicle for 2 hours. Then, the embryos were fixed and used for RNAscope.

Imaging

The following systems were used: for confocal imaging, Zeiss LSM880-META NLO confocal microscope equipped with Airyscan; for color imaging, Zeiss Observer Z1 with a Hamamatsu ORCA-ER camera; for fluorescence imaging, Zeiss Observer Z1 with an AxioCam MRm camera. All histology analysis and immune-fluorescent staining analysis were quantitated in a blinded way by at least two investigators.

Statistics

Data is shown as mean ± standard deviation. An unpaired, two-tailed student’s t-test or linear regression comparison as specified were used for statistical comparison. A P-value of 0.05 or less was considered statistically significant.