Hey2 restricts cardiac progenitor addition to the developing heart

A key event in vertebrate heart development is the timely addition of second heart field (SHF) progenitor cells to the poles of the heart tube. This accretion process must occur to the proper extent to prevent a spectrum of congenital heart defects (CHDs). However, the factors that regulate this critical process are poorly understood. Here we demonstrate that Hey2, a bHLH transcriptional repressor, restricts SHF progenitor accretion to the zebrafish heart. hey2 expression demarcated a distinct domain within the cardiac progenitor population. In the absence of Hey2 function an increase in myocardial cell number and SHF progenitors was observed. We found that Hey2 limited proliferation of SHF-derived cardiomyocytes in a cell-autonomous manner, prior to heart tube formation, and further restricted the developmental window over which SHF progenitors were deployed to the heart. Taken together, our data suggests a role for Hey2 in controlling the proliferative capacity and cardiac contribution of late-differentiating cardiac progenitors.

domain within the cardiac progenitor population. In the absence of Hey2 function an 23 increase in myocardial cell number and SHF progenitors was observed. We found that 24 Hey2 limited proliferation of SHF-derived cardiomyocytes in a cell-autonomous 25 manner, prior to heart tube formation, and further restricted the developmental 26 window over which SHF progenitors were deployed to the heart. Taken together, our 27 data suggests a role for Hey2 in controlling the proliferative capacity and cardiac 28 contribution of late-differentiating cardiac progenitors. 29

INTRODUCTION 30
Cardiac development is regulated by the activity of concerted signaling, 31 transcriptional and morphogenetic events. Subtle perturbations in these processes, 32 either genetic or environmental, can lead to congenital heart defects (CHD), the most 33 common class of congenital anomalies. It is now evident that the vertebrate heart is 34 built from two populations of progenitor cells, termed the first heart field (FHF) and 35 second heart field (SHF), which contribute to the heart in two successive windows of 36 differentiation. Cells of the FHF differentiate in an initial wave of cardiogenesis, 37 resulting in formation of the linear heart tube. Over a well-defined developmental 38 window, multi-potent, late-differentiating progenitors of the SHF migrate into the 39 poles of the heart tube to extensively remodel and add structure to the heart (Cai et al., Previously described primers were used for quantification of β-actin (Tang et al., 166 2007); ltbp3 (Zhou et al. 2011); tbx1 (Zhang et al., 2006) and amhc (Jia, et  The FGF receptor inhibitor SU5402 (Tocris 3300) was used at a concentration of 171 10μM from 16.5 to 20 hours post-fertilization (hpf) or from 19 to 24 hpf. BMP and 172 Notch signaling inhibitors dorsomorphin (Tocris 3093) and DAPT (Tocris 2634/10), 173 respectively, were used at a concentration of 10μM and 50μM between 16. 5

EdU incorporation 207
EdU incorporation assays were performed as previously described (Zeng and Yelon, 208 2014 To further examine the role of Hey2 in cardiac development, we first carried out a 225 detailed analysis of hey2 expression during key stages of cardiogenesis, spanning 226 early cardiac specification to the formation of the linear heart tube, using whole-227 mount RNA in situ hybridization. Interestingly, hey2 transcripts were found to 228 localize anteromedial to those of nkx2.5 and mef2cb in the anterior lateral plate 229 mesoderm (ALPM) at 16.5 hours post-fertilization (hpf, Fig. 1A and D, respectively). 230 At 20 hpf, when the primitive heart is organized into a cone of differentiating cardiac 231 cells (Yelon et al., 1999), hey2 expression was again evident anteromedial to that of 232 myl7 and mef2cb ( Fig. 1B and E). We further observed a domain of hey2 expression 233 lateral to the heart cone, in the region of the pharyngeal mesoderm ( Fig. 1B and E,  234 white arrowheads). Following formation of the linear heart tube at 24 hpf, hey2 235 transcripts were detectable both within and extending from the distal portion of the 236 ventricle, a region occupied by mef2cb-positive cells of the presumptive SHF ( Fig. 1C  237 and F; (Lazic and Scott, 2011). These results, as summarized in Figure 1G  inhibitor, SU5402, we noted the expected reduction in ventricular cardiomyocyte 298 differentiation but normal atrial differentiation, as shown by vmhc and amhc 299 expression (supplemental Fig. 2A and B). SU5402 treatment between 16.5 hpf and 20 300 hpf further resulted in reduced cardiac hey2 expression within the heart cone ( Fig. 2A

Loss of Hey2 affects cardiac function and maturation 358
Given the irregular morphology of hey2 mutant hearts, we next investigated 359 cardiomyocyte-intrinsic defects. The function of the heart is highly dependent on 360 cardiomyocyte morphology, specifically cell shape (Auman et al., 2007). 361 Cardiomyocytes initially have a uniform cuboidal characteristic that is altered as 362 ventricular chamber formation proceeds by both blood flow and cardiac contractility 363 (Manasek, 1981;Manasek et al., 1972;Taber, 2006). To examine potential defects in 364 cardiac maturation, the hey2 hsc25 ,allele was crossed into a Tg(myl7:nlsDsRedExpress) 365 background. IF staining using antibodies against DsRed and zn-8 was used to 366 visualize cardiomyocyte cell nuclei and cell membranes, respectively. As compared to 367 sibling controls, in hey2 hsc25 mutants cardiomyocytes failed to initiate cellular 368 elongation near the atrio-ventricular canal (AVC) ( Fig. 3H and I, arrowheads). 369 Instead, cell shape remained uniform throughout the ventricle. This was quantified 370 through axis ratio measurements, which showed a significant difference in cell shape 371 at the AV boundary in hey2 hsc25 mutants compared to control embryos ( Hey2 in establishing the appropriate number of late-differentiating progenitors that 450 will be added to the heart prior to heart tube formation. We next employed a transplantation approach to examine the cell autonomy of Hey2 490 activity in cardiac progenitors. To accomplish this, Tg(myl7:nlsKikGR) donor 491 embryos, either WT or injected with hey2 MO, were used. Donor cells at 4 hpf (50% 492 epiboly) were transplanted to the margin of WT host embryos (Fig. 7A), an approach 493 that has been shown to result in cardiac contribution of donor cells (Scott et al., 2007;494 Stainier et al., 1993). Transplant embryos were photoconverted at either 24 or 48 hpf 495 and imaged at either 48 or 60 hpf ( Fig. 7B and F, respectively). From our results we 496 observed that Hey2 morphant donor Tg(myl7:nlsKikGR) cells displayed a significant 497 increase in SHF contribution (shown by the ratio of green:yellow cardiomyocytes per 498 embryo), as compared to early myocardial addition, between 24 to 48 hpf (Fig. 7C-E). 499 The same result was observed between 48 to 60 hpf, with Hey2 deficient 500 myl7:nlsKikGR donors contributing significantly more green only cardiomyocytes 501 than controls (Fig. 7F-I). While the ratio in numbers of late versus early differentiated 502 cardiomyocytes per heart from hey2 donor cells was consistently increased at both 24-503 48 and 48-60 hpf, an analysis of cell numbers for each category revealed a bias in 504 progenitor populations. From 24-48 hpf, the increased late versus early CM addition 505 ratio observed in hey2 morphants was due to a decreased propensity for donor cells to 506 contribute early to the heart, evident by a significant decrease in total cell number of 507 yellow CMs in controls compared to morphant embryos (Fig. 7J). However, when 508 comparing total number of late differentiating CMs, no statistical significance was 509 observed (Fig. 7J). In contrast, the higher late versus early addition ratio observed at 510 48 hpf was due to a significantly higher amount of late (post 48 hpf) CM addition, 511 with a relatively equivalent amount of early (pre 48 hpf) CM addition as noted by no 512 significant change in early CM number between control and hey2 morphant embryos 513 (Fig. 7K). These results demonstrate a cell-autonomous function for Hey2, in 514 presumptive cardiac progenitors, that delays their addition to the heart as 515 where Hey2 acts specifically in SHF progenitors. While we did not observe 551 appreciable effects on FHF-associated markers in our study, this data is difficult to 552 interpret, as bona fide FHF-and SHF-specific markers that distinguish these 553 populations are poorly characterized. Our transplant assays strongly suggest that 554 cardiac contribution is at the very least delayed with the loss of hey2. The observed 555 phenotypes may reflect: 1) a shift in the balance between FHF and SHF progenitor 556 proliferation, favouring the SHF pool; 2) alteration in the timing of cardiac progenitor 557 differentiation; 3) changes in the number of progenitors allocated to the FHF and SHF 558 pools; or 4) a FHF/SHF-agnostic role for Hey2 in cardiac progenitor proliferation and 559 differentiation. This is a critical question that will require further study, but that may 560 provide insight into the diversity of early cardiac progenitors. 561

562
It is important to note that the function of hey2 in zebrafish cardiogenesis has been 563 previously addressed in an elegant study (Jia et al., 2007). However, while the overall 564 "large heart" phenotype observed is shared between our hey2 hsc25 and the grl m145 565 mutants, our data suggests a role for hey2 prior to 24 hpf, in cardiac progenitors, that 566 subsequently impacts cardiac development. In contrast, Jia and colleagues reported 567 that grl had minimal affect during this time, with grl m145 mutants having comparable 568 cardiomyocyte numbers to controls at 24 hpf. This discrepancy may reflect the nature 569 of the hey2 alleles used, with the hey2 hsc25 allele being, we believe, a true null. As the 570 SHF and late myocardial addition in zebrafish had not been described at the time of 571 the prior study, this would have also affected the interpretation of the results. This 572 highlights the fact that Hey2 likely acts at multiple steps of heart development.