The LIM Protein Ajuba Restricts the Second Heart Field Progenitor Pool by Regulating Isl1 Activity

Developmental Cell

Article The LIM Protein Ajuba Restricts the Second Heart Field Progenitor Pool by Regulating Isl1 Activity Hagen R. Witzel,1 Benno Jungblut,2 Chong Pyo Choe,3 J. Gage Crump,3 Thomas Braun,2 and Gergana Dobreva1,* 1Origin

of Cardiac Cell Lineages Group, Max Planck Institute for Heart and Lung Research, Parkstr. 1, 61231 Bad Nauheim, Germany Planck Institute for Heart and Lung Research, Ludwigstr. 43, 61231 Bad Nauheim, Germany 3Broad California Institute for Regenerative Medicine Center, University of Southern California Keck School of Medicine, 1425 San Pablo Street, BCC 406, Los Angeles, CA 90089, USA *Correspondence: [email protected] http://dx.doi.org/10.1016/j.devcel.2012.06.005 2Max

SUMMARY

Morphogenesis of the heart requires tight control of cardiac progenitor cell specification, expansion, and differentiation. Retinoic acid (RA) signaling restricts expansion of the second heart field (SHF), serving as an important morphogen in heart development. Here, we identify the LIM domain protein Ajuba as a crucial regulator of the SHF progenitor cell specification and expansion. Ajuba-deficient zebrafish embryos show an increased pool of Isl1+ cardiac progenitors and, subsequently, dramatically increased numbers of cardiomyocytes at the arterial and venous poles. Furthermore, we show that Ajuba binds Isl1, represses its transcriptional activity, and is also required for autorepression of Isl1 expression in an RA-dependent manner. Lack of Ajuba abrogates the RA-dependent restriction of Isl1+ cardiac cells. We conclude that Ajuba plays a central role in regulating the SHF during heart development by linking RA signaling to the function of Isl1, a key transcription factor in cardiac progenitor cells. INTRODUCTION Heart formation is a complex morphogenetic process involving specification, differentiation, and migration of cardiac progenitor cells. Defects in this program are responsible for the high rate of congenital cardiac abnormalities in humans underscoring the importance of understanding the molecular mechanisms regulating cardiogenesis (Bruneau, 2008; Srivastava, 2006a). Using different model systems, including Drosophila, Xenopus, zebrafish, chick, and mouse, substantial progress has been made to unveil the principal mechanisms and the key molecular players driving heart development. It has been demonstrated that the heart is generated by two distinct progenitor cell populations, referred to as first and second heart fields (FHFs and SHFs, respectively), which have different regional contribution to the developing heart (Buckingham et al., 2005; Cai et al., 2003; Kelly et al., 2001; Olson, 2006; Waldo et al., 2001). The FHF and SHF progenitor cells are distinguished by expression of specific 58 Developmental Cell 23, 58–70, July 17, 2012 ª2012 Elsevier Inc.

transcription factors, as well as signaling molecules, which have to act together in a coordinated manner for the proper development of the heart (Black, 2007; Buckingham et al., 2005; Evans et al., 2010; Laugwitz et al., 2008; Srivastava, 2006b; Vincent and Buckingham, 2010). Islet1 (Isl1), a LIM homeodomain transcription factor, has been regarded as a marker of the SHF (Cai et al., 2003; Laugwitz et al., 2005). Fate mapping experiments in mouse demonstrated a contribution of Isl1+ cells to the venous and the arterial poles of the heart (Cai et al., 2003). Ablation of Isl1 in mouse leads to severely misshaped, unlooped hearts, lacking the right ventricle and outflow tract (Cai et al., 2003). Similarly, Isl1 is necessary for heart morphogenesis, cardiac marker gene expression, and vasculogenesis of Xenopus embryos (Brade et al., 2007), whereas in zebrafish Isl1 is required to complete the cardiomyocyte differentiation process at the venous pole (de Pater et al., 2009). Isl1 is transiently expressed in SHF progenitors before migration into the heart tube and is downregulated during their differentiation (Cai et al., 2003; Laugwitz et al., 2008). Moreover, Isl1 is required for the proliferation, survival, and migration of these cells (Cai et al., 2003). Interestingly, it was also suggested recently that Isl1 negatively regulates the number of cardiac progenitor cells (Kwon et al., 2009). The mechanisms that coordinate these opposing functions of Isl1 are currently unclear. Another key regulator of the SHF is the transcription factor Nkx2.5. Studies in different model organisms indicated that Nkx2.5 and its homologs play important roles in cardiogenesis. Drosophila tinman, for example, is involved in the specification of the heart primordial cells (Azpiazu and Frasch, 1993; Bodmer, 1993), and mouse Nkx2.5 also plays a crucial role in heart morphogenesis (Lyons et al., 1995). More recently, it was shown that Nkx2.5 deficiency induces overspecification of cardiac progenitors at early stages of heart development followed by a failure to maintain proliferation of SHF progenitor cells and subsequent truncation of the outflow tract (Prall et al., 2007). In zebrafish, Nkx2.5 and Nkx2.7 are required to limit atrial cell numbers and establish proper numbers of ventricular cardiomyocytes (Targoff et al., 2008; Tu et al., 2009). Heart development is also critically regulated by extracellular signaling (Brand, 2003; Evans et al., 2010; Vincent and Buckingham, 2010). Several studies have demonstrated a role of retinoic acid (RA) signaling in the anteroposterior patterning of the heart (Hochgreb et al., 2003; Ryckebusch et al., 2008; Sirbu

Developmental Cell Ajuba Restricts the Cardiac Progenitor Pool

A

Isl1 HA-Ajuba

_ + _ + _ _ + +

B

10% input IP:

HA WB:

Isl1 Zyxin Myc-LPP family Myc-Zyxin Ajuba Myc-Limd1 family Myc-WTIP

Isl1

_ _

_ _ _ _ _ _ _ _ + _ + _ _ _ +

Figure 1. Ajuba Interacts with Isl1

+ _ + _ + _ + + _ _ _ + _ _ + _ _ _ _ +

10% input

WB:

myc

WB:

myc

WB:

Isl1

IP

C

IP: WB:

D

_ + _

Isl1

Ajuba

IP: Isl1

LIM1 LIM2 Isl1 Isl1 LIM1 Isl1 LIM2 Isl1HOMEO

HA-Ajuba Isl1 Isl1 LIM1

_ + + _ + _ + _ + _ _ +

HOMEO

HA-Ajuba Isl1 LIM2 Isl1HOMEO

10% input

10% input

IP: -Isl1

IP: -Isl1

E

+ + _ + _ +

HA-Ajuba HA-Ajuba LIM1 HA-Ajuba LIM2 HA-Ajuba LIM3 Isl1 HA-Ajuba HA-Ajuba LIM1 HA-Ajuba LIM2 HA-Ajuba LIM3

WB: -HA

WB: -HA

WB: -Isl1

10% input IP: -Isl1

10% input

10% input

_ + + + _ _ + + _ _ _ + _ _ _ _ _ _ _ _

WB: -Isl1

et al., 2008; Stainier and Fishman, 1992). Raldh2-deficient mouse embryos, which lack RA signaling, exhibit posterior expansion of Isl1 and Fgf8-positive populations, indicating that RA is important to establish the posterior boundary of the SHF (Ryckebusch et al., 2008; Sirbu et al., 2008). In the zebrafish embryo, RA signaling restricts the cardiac progenitor pool (Keegan et al., 2005; Waxman et al., 2008). The mechanisms linking RA to the transcriptional regulators of heart development, however, remain largely unclear. Here, we identified Ajuba, a LIM domain protein, as a crucial regulator of SHF progenitor cell specification and expansion. Furthermore, we show that Ajuba binds Isl1 and represses its transcriptional activity, which is required to downregulate the expression of key transcription factors in the SHF such as Mef2c, and to enable Isl1 to suppress its own expression. Moreover, we show that RA is a critical upstream regulator of this process because it controls the number of Isl1+ cells in the heart through an Ajuba-dependent mechanism. In addition, we find that Ajuba regulates Nkx-2.5 levels, which might help to limit cardiac specification. RESULTS Ajuba Interacts with Isl1 and Represses Its Transcriptional Activity To gain insight into the mechanisms underlying Isl1 function in cardiac progenitor cell regulation, we performed a screen for interaction partners of Isl1 (unpublished data). We identified the LIM domain protein Ajuba as a prominent Isl1-binding partner. Ajuba shuttles between the cytoplasm and the nucleus and affects proliferation and cell fate decisions, processes that are RA dependent (Kanungo et al., 2000). Thus, Ajuba potentially

WB: -HA

+ + _ _ _ _ _ + _ +

(A) FLAG-HA-tagged Ajuba and Isl1 were transiently expressed in HEK293T cells. Proteins were immunoprecipitated with an anti-HA antibody and detected by immunoblot with an anti-Isl1 antibody. IP, immunoprecipitation; WB, western blot. (B) Isl1 and Myc-tagged LPP, Zyxin, LIMD1, and WTIP were transiently expressed in HEK293T cells, and immunoprecipitation with an anti-Isl1 antibody followed by immunoblot analysis with an anti-myc antibody was performed. (C) Coimmunoprecipitation of extracts from day 5 EBs differentiated from ES cells using anti-Isl1 antibody and detected with anti-Ajuba antibody. (D) Schematic representation of wild-type Isl1 and Isl1 deletion constructs (top). FLAG-HA-tagged Ajuba and Isl1 or Isl1 deletion constructs were transiently expressed in HEK293T cells, and immunoprecipitation with an anti-Isl1 antibody followed by immunoblot analysis with an anti-HA antibody was performed (bottom). (E) Schematic representation of wild-type Ajuba and Ajuba deletion constructs (top). FLAG-HAtagged Ajuba or Ajuba deletion constructs and Isl1 were transiently expressed in HEK293T cells, and immunoprecipitation with an anti-Isl1 antibody followed by an immunoblot analysis with an antiHA antibody was performed. NES, nuclear export sequence.

provides a link between RA signaling and Isl1, both of which can negatively regulate progenitor cell numbers (Keegan et al., 2005; Kwon et al., 2009). To further investigate the interaction between Isl1 and Ajuba, we transfected HEK293T cells with a FLAG-HA-tagged construct of Ajuba alone or together with an Isl1 expression plasmid. Coimmunoprecipitation with an anti-HA antibody and subsequent immunoblot analysis with an anti-Isl1 antibody indicated that Isl1 was efficiently coimmunoprecipitated (Figure 1A). To determine whether other members of the Ajuba/ Zyxin family interact with Isl1 and to address the specificity of these interactions, we transfected Isl1 together with myctagged Ajuba LIM proteins (LIMD1, WTIP) and the closely related Zyxin LIM proteins (Zyxin, LPP) and performed coimmunoprecipitations. The Ajuba/Zyxin family is characterized by three homologous C-terminal LIM domains and a unique proline-rich N-terminal pre-LIM domain. We observed a strong interaction of Isl1 with WTIP and a weaker one with LIMD1 (Figure 1B). In contrast, Zyxin and LPP did not bind Isl1 (Figure 1B), indicating that Ajuba LIM proteins specifically interact with Isl1. To confirm that Ajuba and Isl1 interact in cardiac progenitors expressing endogenous levels of each protein, we performed immunoprecipitation of Isl1 from nuclear extracts of embryoid bodies (EBs) differentiated for 5 days, a stage enriched in cardiac progenitors. We found that Ajuba was efficiently coimmunoprecipitated together with Isl1 (Figure 1C). To delineate the domains of Isl1 necessary for its interaction with Ajuba, we transfected HEK293T cells with a FLAG-HA-Ajuba expression construct together with full-length Isl1 or Isl1 deletion constructs (Figure 1D). We observed a strong interaction of Ajuba with full-length Isl1 and truncated Isl1 protein lacking LIM1, but not with mutants lacking the LIM2 domain (Figure 1D), Developmental Cell 23, 58–70, July 17, 2012 ª2012 Elsevier Inc. 59

Developmental Cell Ajuba Restricts the Cardiac Progenitor Pool

A

B

HA-Ajuba

Isl1

4

Isl1/ Isl1/HA-Ajuba HA-Ajuba LIM1

C 5

HEK 293T

NIH3T3

C

4

3

Isl1 LIM2/ HA-Ajuba

3 2 2 1 0 Mef2c-luc Isl1 Isl1 Homeo Ajuba

1 0

+ + _ + _ _ _ _

+ _ _ +

+ _ + _

+ + _

+ + _

E

D Mef2c AHF enhancer IP Input mIgG Isl1

+ + Mef2c-luc + _ + + Isl1 _ _ Ajuba _

+ + _ + _ +

-Ajuba -Isl1

-Ajuba

+ + + + + +

1

+ +

2

Figure 2. Ajuba Colocalizes with Isl1 and Represses Its Transcriptional Activity (A) Subcellular localization of Isl1 and Ajuba in HEK293T cells. FLAG-HA-tagged Ajuba and Isl1 full-length or deletion constructs were transiently expressed and detected by immunofluorescence using anti-HA or anti-Isl1 antibodies. Scale bars, 50 mm. (B and C) HEK293T (B) and NIH 3T3 (C) cells were transiently transfected with a 40 ng (B) or 50 ng (C) luciferase reporter construct containing the Mef2c cardiac-specific enhancer fragment (Dodou et al., 2004) in front of a minimal fos promoter (Mef2cluc), alone or together with Isl1 (400 ng), Isl1 lacking the DNA-binding domain (Isl1DHOMEO, 400 ng), Ajuba (400 ng), or with constant amount of Isl1 (400 ng), and increasing amounts of Ajuba expression plasmid (100, 200, 300, 400 ng). The luciferase levels were normalized for the b-galactosidase activity of a cotransfected RSV-lacZ reporter and presented as fold activation relative to the luciferase levels of the Mef2c reporter construct alone. Transfections were performed at least three times in triplicates, and representative experiments with the SDs are shown. (D) ChIP of nuclear extracts from day 5 EBs, using mouse or rabbit IgG (mIgG and rIgG, respectively) as a control, anti-Isl1 and anti-Ajuba antibodies, respectively. PCRs were performed using primers flanking the two conserved Isl1-binding sites in the Mef2c AHF enhancer. (E) Confocal images of cryosections of E9 embryos stained with anti-Ajuba (green) and anti-Isl1 (red) antibodies. Scale bars, 100 mm. See also Figure S1.

2 -Isl1

affected the localization of either protein: Isl1 was partially retained in the cytoplasm, whereas significant amounts of -Ajuba IP Ajuba protein localized to the nucleus -Isl1 Input rIgG Ajuba DAPI (Figures 2A and S1A). In contrast, no colocalization was observed when Isl1 E9.0 was expressed together with a truncated Ajuba protein lacking the LIM1 domain or indicating that the LIM2 domain of Isl1 mediates the interaction when Isl1 protein lacking the LIM2 domain was coexpressed with Ajuba. An analogous approach was taken to identify the together with full-length Ajuba (Figure 2A). Furthermore, domains in Ajuba required for interaction with Isl1, revealing a similar colocalization was evident when Isl1 was transfected a strong association of Isl1 with full-length Ajuba and Ajuba together with WTIP, which strongly interacted in the coimmunotruncations lacking the LIM2 and LIM3 domains, but not the precipitation experiments (Figure S1B). As expected, we did LIM1 domain (Figure 1E). Thus, Isl1 specifically interacts with not detect a significant colocalization of Isl1 with LIMD1, Zyxin, the LIM1 domain of Ajuba. or LPP (Figures S1B and S1C; data not shown), which We next wanted to determine whether Isl1 and Ajuba coloc- showed weak or no interaction with Isl1 in coimmunoprecipitaalize in cells. Isl1 is a nuclear protein, and Ajuba was shown to tion experiments (Figure 1B). Moreover, the colocalization was shuttle between the cytoplasm and the nucleus (Kanungo et al., conserved between mouse and zebrafish (Figure S1D). 2000). When expressed alone, Isl1 was localized in the nucleus, Next, we sought to determine the effect of Isl1-Ajuba interacwhereas Ajuba was mainly detected in the cytoplasm (Fig- tion on Isl1 transcriptional activity. We transfected HEK293T and ure 2A). Interestingly, when cells were transfected with Isl1 NIH 3T3 cells with Isl1 and Ajuba, together with a luciferase and Ajuba together (Figure 2A) or with Isl1 and Ajuba deletion reporter construct containing the Mef2c cardiac-specific constructs that do not affect its binding to Isl1 (see Figure S1A enhancer, which is directly bound and activated by Isl1 (Dodou available online), a clear colocalization of Isl1 and Ajuba was et al., 2004). Overexpression of Isl1, but not an Isl1 mutant observed. Concomitant expression of both Isl1 and Ajuba lacking the DNA-binding homeodomain, led to upregulation of 1

60 Developmental Cell 23, 58–70, July 17, 2012 ª2012 Elsevier Inc.

Developmental Cell Ajuba Restricts the Cardiac Progenitor Pool

Figure 3. Ajuba Knockdown in Zebrafish Leads to an Increased Number of Cardiomyocytes at the Arterial and Venous Poles

A 1

23

45

6

78

1

2 3 4 5 67 8

AUG sp MO MO

B

Control

48h

Ajuba MO

48h

Control

48h Ajuba MO

D 180 Cell numbers in atrium

C

Control

myl7:dsRed myl7:mGFP Ajuba MO

120 100

myl7:dsRed myl7:mGFP

Ajuba MO

*

80

***

60 40

region 1), which was not seen in the epithelium (Figure 2E, boxed region 2).

20 0

Cell numbers in ventricle

Control

Control Ajuba MO

140

myl7:mGFP F

myl7:mGFP

E

***

160

48h

(A) Schematic representation of the Ajuba genomic locus (left) and the exons coding the three LIM domains (drawn to scale, right). The binding sites of the AUG and the splice morpholino are shown in red. (B) Control or AUG MO-injected Tg(myl7:EGFPHsHRAS)s883 embryos at 48 hpf. Lateral views, anterior to the left, are in the top-left and bottom panels. Frontal views of the same embryos are in the middle and right panels. (C) Confocal images of control and Ajuba morphant Tg(myl7:EGFP-HsHRAS)s883/Tg(
5.1myl7: nDsRed2)f2 hearts. (D) Total number of atrial cells and number of mGFP+DsRed+ and mGFP+DsRed
cells in the atrium at 48 hpf. Error bars represent SD (n = 5). *p < 0.05; ***p < 0.001. (E) Confocal images of control and Ajuba AUG morphant ventricle at 48 hpf. (F) Total number of ventricle cells. **p < 0.01. See also Figure S2 and Table S1.

Total mGFP+/ mGFP+/ dsRed+ dsRed-

240 220 200 180 160 140 120 100 80 60 40 20 0

luciferase reporter expression. Importantly, cotransfection of increasing amounts of Ajuba together with Isl1 led to a dosedependent downregulation of reporter gene expression indicating that Ajuba suppresses Isl1 transcriptional activity (Figures 2B and 2C). To determine whether Ajuba binds Isl1 at regulatory elements of Isl1 target genes, we performed chromatin immunoprecipitation (ChIP) using control (IgG) or anti-Isl1 and antiAjuba antibodies with extracts from day 5 EBs. Importantly, both Isl1 and Ajuba were bound to the Mef2c anterior heart field (AHF) enhancer, further corroborating our hypothesis that both proteins form a complex to regulate Isl1 target gene expression during cardiogenesis (Figure 2D). No binding was detected to a control sequence in the b-globin locus (data not shown). Next, we determined the expression and subcellular localization pattern of Ajuba during embryonic development focusing on the Isl1+ cardiogenic region. Interestingly, we found an accumulation of Ajuba in the nucleus of Isl1+ cells (Figure 2E, boxed

Ajuba Knockdown in Zebrafish Leads to an Increased Number of ** Cardiomyocytes at the Arterial and Venous Poles To investigate whether the transcriptional suppression of Isl1 by Ajuba is crucial for the restriction of SHF expansion and differentiation, we turned to the zebrafish system, where it was recently shown that a conserved SHF is required for cardiac development (Hami et al., 2011; Lazic and Scott, 2011; Zhou et al., 2011). We utilized the transgenic line Control Ajuba MO Tg(myl7:EGFP-HsHRAS)s883 that expresses membrane-bound GFP (mGFP) under the control of the myl7 (cmlc2; cardiac myosin light-chain 2) promoter in all differentiated cardiomyocytes (D’Amico et al., 2007). Two morpholinos (MOs) were designed to block either translation (AUG MO) or splicing (Sp MO) of Ajuba mRNA (Figure 3A) and tested for efficacy and specificity (Figure S2). Importantly, the two nonoverlapping MOs produced the same phenotype, confirming the specificity of the MO-mediated Ajuba knockdown (Figures 3B and S2D). Ajuba morphants exhibited developmental abnormalities including left-right axis malformations, enlarged hearts, and pericardial edema, whereas control MO had no discernible effects (Figures 3B and S2D; Table S1). The effects on left-right asymmetry of the heart were similar to results reported for the Ajuba homolog in medaka fish showing a situs inversus in 29% of Ajuba morphants (Nagai et al., 2010; Table S1). Interestingly, 86.2% of the morphants showed dramatically enlarged hearts (Figures 3B–3F; Table S1). Furthermore, the relative positioning of the atria and the

Developmental Cell 23, 58–70, July 17, 2012 ª2012 Elsevier Inc. 61

Developmental Cell Ajuba Restricts the Cardiac Progenitor Pool

ventricles was altered in the Ajuba morphants (Figure 3C). In addition, the beating frequency of Ajuba morphant hearts was reduced (data not shown). Previous studies using transgenic fish expressing both EGFP and dsRed in differentiating cardiomyocytes under the control of the myl7 promoter revealed that growth of the developing zebrafish heart is regulated by two distinct phases of cardiomyocyte differentiation and that Isl1 seems required to complete cardiomyocyte differentiation at the venous pole (de Pater et al., 2009). To investigate the potential role of Ajuba in this process, we injected Ajuba AUG MO into a double transgenic line Tg(myl7:EGFP-HsHRAS)s883/ Tg(
5.1myl7:nDsRed2)f2 (D’Amico et al., 2007; Mably et al., 2003) (Figure 3C). Because EGFP becomes fluorescent very soon after synthesis, but dsRed requires about 24 hr to mature, this assay allowed us to determine the timing of cardiomyocyte differentiation in response to Ajuba depletion. We observed a significant increase of the total number of atrial cells (control, 105 ± 1, n = 5; Ajuba morphant, 161 ± 7, n = 5; p < 0.001; Figures 3C and 3D). Interestingly, this increase seems to be due to a continuous generation of new cardiomyocytes because we detected only slight differences in the number of mGFP+/ dsRed+ cells in Ajuba morphants but a dramatic increase of the number of mGFP+/dsRed
cells:
23 ± 3 in control hearts and 75.5 ± 5 in Ajuba morphant hearts (Figures 3C and 3D). This finding corresponds nicely to the decrease of mGFP+/ dsRed
cells in Isl1 zebrafish mutants (de Pater et al., 2009), arguing for a role of Ajuba in the regulation of Isl1. Furthermore, we found a significantly enlarged arterial pole in Ajuba morphants (Figure 3E). Moreover, the total number of ventricular cardiomyocytes was significantly increased (control, 133 ± 6, n = 3; Ajuba morphant, 208 ± 18, n = 3; p < 0.01; Figures 3E and 3F). Ajuba and Isl1 Restrict the Number of Isl1-Expressing Cells in the Heart To address whether the increased number of cardiomyocytes at the arterial and venous poles is due to increased number of Isl1-expressing cells in the heart, we performed Isl1 immunostaining of control and Ajuba morphant embryos at different developmental stages. At the 24 somite stage, the number of Isl1+ cells located in and at the periphery of the future atrium (Figure 4A) (de Pater et al., 2009) was already significantly increased. Later during development, at 30 hr postfertilization (hpf), Isl1+ cells are found at both the arterial and venous poles (de Pater et al., 2009; Hami et al., 2011). Importantly, we detected a significant increase of Isl1+ cells at the periphery of both poles and in the future atrium and ventricle (Figures 4B, S3A, and S3B). At 48 hpf the number of Isl1+ cardiomyocytes, found at the venous pole of the atrium, was also significantly higher (Figure S3C), and the number of Isl1-positive tiers was expanded from three in control embryos to seven in the morphants (Figure 4C). To determine the effect of the additional Ajuba family members on the number of Isl1-expressing cells in the heart, we performed MO knockdown of LIMD1 and WTIP. Similar to the Ajuba morphants, the WTIP-deficient embryos showed a higher number of Isl1+ cells at the venous pole (Figure S3D). In contrast, no change in Isl1 levels was detected in embryos deficient for LIMD1 (Figure S3D). To further analyze whether Ajuba restricts the number of Isl1-expressing cells, we overex62 Developmental Cell 23, 58–70, July 17, 2012 ª2012 Elsevier Inc.

pressed Ajuba by injecting mRNA into early-stage embryos. Interestingly, increased concentrations of Ajuba resulted in significantly smaller hearts and a complete absence of Isl1+ cells at the venous pole of the heart (Figure 4D). To test whether the increased number of Isl1+ cells in the heart can be due to increased proliferation, we performed immunostaining with an antibody recognizing Ser10-phosphorylated histone H3 (phospho-H3). We could detect only very few (one to two) phosphoH3-positive cells in the myocardium of the entire heart tube in control embryos as reported previously (de Pater et al., 2009) and in Ajuba-deficient embryos (data not shown), suggesting that the increased number of Isl1+ cells is not due to increased proliferation. We next wanted to determine whether the observed phenotype in Ajuba-deficient embryos depends on its role in regulating Isl1 activity. To this end, we performed MO knockdown of Isl1 (Hutchinson and Eisen, 2006) either alone or in combination with Ajuba MO and performed anti-GFP (for myl7) and anti-Isl1 immunostaining. We did not detect any Isl1 staining at the venous pole of the heart in the Isl1 and Isl1/Ajuba morphants, confirming the efficacy of the MO-mediated Isl1 knockdown (data not shown). Importantly, the double Isl1/Ajuba-deficient zebrafish embryos did not show any increase in the total number of cardiomyocytes in contrast to the Ajuba morphants (Figures 4E and S3E). Because injection of Isl1 MO prevents production of Isl1 protein but does not affect transcription of the Isl1 gene, we were also able to analyze activity of the Isl1 promoter in absence of Isl1 protein. Surprisingly, absence of Isl1 protein or both Isl1 and Ajuba caused a significantly higher and broadened Isl1 expression similar to Ajuba-deficient embryos (Figure 4F), suggesting that Isl1 restricts the number of Isl1-expressing cells. Consistent with this finding, overexpression of Isl1, but not a truncated version lacking the LIM2 domain responsible for its interaction with Ajuba, led to an absence of Isl1+ cells at the venous pole of the atrium (Figure S3F). Immunostaining for myl7-mGFP combined with in situ hybridization for Isl1 revealed that the expanded Isl1+ cell population in double Isl1/Ajuba-deficient embryos was located outside of the heart tube (Figure 4G), which differs from the situation in control and Ajuba morphant embryos where Isl1+ cells are localized at the venous pole inside the atrium. Furthermore, we noticed mGFP+ cardiomyocytes in the pericardial wall adjacent to the heart in Isl1/Ajuba morphants (Figure 4E, arrows). Taken together, these data suggest that Ajuba together with Isl1 plays an important role in restricting the number of Isl1-expressing cells, but Isl1 is required for cardiomyocyte differentiation in the heart through Ajuba-independent mechanisms. Ajuba Restricts the Number of SHF Progenitor Cells Because we detected an increase of Isl1+ cells in the heart at 24 somites and did not observe increased proliferation of Isl1+ cardiomyocytes in Ajuba morphants, we wanted to determine whether Ajuba regulates cardiac progenitor cell numbers. To explore this possibility, we first refined the expression pattern of Isl1 and Ajuba during cardiogenesis by performing wholemount immunostaining and in situ hybridization at early developmental stages, and comparing it to the Nkx2.5-expressing domain using Nkx2.5:GFP transgenic embryos, which faithfully recapitulated Nkx2.5 expression (Figures 5A, S4, and S5)

Developmental Cell Ajuba Restricts the Cardiac Progenitor Pool

A

D

C

B

E

G

F

Figure 4. Ajuba and Isl1 Restrict the Number of Isl1-Expressing Cells in the Heart (A) Confocal images of control and Ajuba morphant Tg(myl7:EGFP-HsHRAS)s883 embryos stained with anti-GFP and anti-Isl1 antibodies at 24 somites. (B and C) Confocal images of control and Ajuba morphant Tg(myl7:EGFP-HsHRAS)s883 embryos stained with anti-Isl1 antibodies (red) and visualized by GFP fluorescence (green) for myl7 expression at 30 hpf (B) and at 48 hpf (C). (D) Control and zAjuba mRNA-injected Tg(myl7:EGFP-HsHRAS)s883 embryos at 48 hpf, frontal view (top); in situ hybridization for Isl1 in control or zAjuba mRNAoverexpressing embryos at 48 hpf (bottom). (E) Confocal images of hearts of control, Ajuba, Isl1, and Ajuba/Isl1 morphant Tg(myl7:EGFP-HsHRAS)s883 embryos, stained with anti-GFP antibody. Arrows show differentiated cardiomyocytes in the pericardial wall adjacent to the heart. (F) In situ hybridization for Isl1 in control, Ajuba, Isl1, and Ajuba/Isl1-deficient embryos. (G) Immunostaining with anti-GFP and in situ hybridization for Isl1 in control (left) and Ajuba/Isl1-deficient Tg(myl7:EGFP-HsHRAS)s883 (right) embryos from (F). See also Figure S3.

(Chen and Fishman, 1996). At 5 somites Isl1-positive cells were found in bilateral populations of cells lying within the anteriorlateral plate mesoderm, a region known to contain heart precursor cells (Figure S4), which is significantly earlier than a previously reported expression of Isl1 (Hami et al., 2011). Later, at 10 and 15 somites, Isl1 expression closely followed the established patterns of zebrafish heart morphogenesis (Chen and Fishman, 1996), including convergence toward the midline (Figures 5A, S4, and S5). At the 10 somite stage, Isl1 marked a progenitor population that partially overlapped with the Nkx2.5 progenitor population (Figure 5A). At 15 somites Isl1-expressing cells were found in the anterior-lateral plate mesoderm; some Isl1+ cells were positioned at the inner margin of the cardiogenic region, dorsal to the Nkx2.5-expressing cells, which start to involute at the inner margin (Figure S5C000 , arrows). At 18 somites Isl1-

expressing cells were found at the future arterial and venous pole (Figure S5D000 , arrows). At 24 somites Isl1+ cells are positioned at the base of the cardiac cone (Figures 4A and S5). In contrast, Ajuba was broadly expressed in the developing embryo, and covered the SHF, which is compatible with a role as a modulator of Isl1 function (Figure S4). At 48 hpf the relatively broad expression domain of Ajuba became confined to the head region but was not found in the heart (data not shown). Next, we analyzed the Isl1 expression pattern in control and Ajuba morphants using in situ hybridization (Figure 5B) at 3, 5, 10, and 15 somite stages. We detected a significant expansion of Isl1+ cells in the cardiogenic regions of Ajubadeficient embryos at all developmental stages. Already at 3 somites, we could detect Isl1+ cardiac progenitors in the Ajuba morphants, and at 5 somites the number was dramatically Developmental Cell 23, 58–70, July 17, 2012 ª2012 Elsevier Inc. 63

Developmental Cell Ajuba Restricts the Cardiac Progenitor Pool

A

-Isl1 -GFP Topro3

Tg

Isl1

Isl1

1B

NT

Isl1

Isl1

10s Control

15s

Isl1

Isl1

1 2 2 10s

-Isl1

Nkx2.5:GFP

C

Merge

10s Nkx2.5:GFP

10s Nkx2.5:GFP

Mef2ca

Control

Control -GFP

Mef2cb

10s Control

Mef2ca

10s Control

Mef2cb

3s Control

5s Control

Isl1

Ajuba MO 10s 3s Ajuba MO

10s

Tbx20

Tbx1

15s Control

10s Control

Tbx20

Isl1

Tbx1

Bmp4

15s

Bmp4

5s Ajuba MO 10s Ajuba MO 15s

D Relative mRNA expression

Nkx2.5:GFP

Control

Ajuba 1.2

1.6 1.4

1

1.2 0.8

1 0.8 0.6

0.6

* ***

***

0.4

0.4 0.2 0

*** ***

0.2 0

Ajuba MO 10s Ajuba MO 10s Ajuba MO 10s Ajuba MO 15s Ajuba MO 15s Figure 5. Ajuba Restricts the Number of SHF Progenitor Cells (A) Confocal images of Nkx2.5:GFP transgenic embryos stained with anti-Isl1 and anti-GFP antibodies at the 10 somite stage. The top-middle panel shows a higher magnification of the image in the top-left panel; the lines indicate the position of the transverse (xz) optical sections presented in top-right panel. NT, neural tube; Tg, trigeminal placode. (B) In situ hybridization of control and Ajuba morphant embryos at 3, 5, 10, and 15 somite stage with an Isl1 probe. (C) In situ hybridization of control and Ajuba morphant embryos with probes for Mef2ca, Mef2cb, Tbx20, Tbx1, and Bmp4. (D) Relative expression level of cardiac progenitor and cardiomyocyte marker mRNA normalized to GAPDH mRNA in day 5 EBs overexpressing either GFP only (control) or both GFP and Ajuba. Error bars represent SDs derived from two independent pools of control and Ajuba-overexpressing ES cells subjected to differentiation in EBs. See also Figures S4, S5, and S6.

increased compared to the control embryos (Figure 5B). In contrast, we detected only few proliferating phospho-H3-positive Isl1-expressing cells in both control and Ajuba morphants (Figure S6), suggesting that the increased number of Isl1+ cardiac progenitors is due to overspecification. Moreover, we found a significant expansion of cells in Ajuba morphants that express Mef2ca, Mef2cb, Tbx20, Tbx1, and Bmp4, which play important roles in the SHF (Figure 5C), suggesting that Ajuba restricts the SHF progenitor pool. To examine whether the role of Ajuba in SHF progenitor cells is evolutionarily conserved, we transfected mouse embryonic stem (ES) cells with an expression plasmid carrying GFP alone or together with Ajuba. GFP-expressing cells isolated by FACS were subjected to differentiation in EBs to facilitate generation of cardiac progenitor cells. RT-PCR revealed downregulation of pan-cardiac genes and genes specifically expressed in the SHF, whereas genes expressed in the FHF (Tbx5, Hand1) showed no change (Figure 5D). Furthermore, the expression of cardiomyocyte marker genes, such as Mlc2a and Mlc2v, was downregulated (Figure 5D). Taken together, these data suggest 64 Developmental Cell 23, 58–70, July 17, 2012 ª2012 Elsevier Inc.

that Ajuba may have a conserved function in the negative regulation of the number of SHF progenitor cells. The Isl1-Ajuba Transcriptional Complex Negatively Regulates Isl1 Expression in an RA-Dependent Mechanism RA signaling restricts the cardiac progenitor pool in zebrafish (Keegan et al., 2005; Waxman et al., 2008) and induces Ajuba expression and shuttling from the cytoplasm to the nucleus (Kanungo et al., 2000). Furthermore, Isl1 serves as a negative regulator of Isl1+ progenitor cells (Kwon et al., 2009), which is in line with our findings that Isl1 restricts the number of Isl1+ cardiac cells at the venous pole (Figure 4F). Hence, we postulated that Ajuba provides a link between RA signaling and Isl1. To test this hypothesis, we treated day 3 EBs, differentiated from ES cells, with RA and analyzed the expression of marker genes of cardiac progenitor cells at day 5. Real-time PCR analysis revealed a strong downregulation of Isl1 (Figure 6A, left), whereas Tbx5 was upregulated by RA (Figure 6A, right), consistent with previous reports (Liberatore et al., 2000).

Developmental Cell Ajuba Restricts the Cardiac Progenitor Pool

DMSO

0.3 M RA

0.1 M RA 8

1.2

***

7

1

-Ajuba

6

0.8

0.2

4

** **

2

Fold change

0

P19

IgG

Input

_

RA

P19

+

1.4

E

**

**

_

_

+

-Isl1

+

-Ajuba

_

P19+RA P19+RA

1.2 1 0.8 0.6

**

**

0.4

***

0.2

0.2

+ _ _

+ + _

+ _

+ +

+ +

+

Amhc

F

3642

Isl1-F

1

1.8 1.6 1.4 1.2 1 0.8 0.6 0.4

Isl1-luc Isl1 Ajuba

1

C

3

0

D

-LaminB

5

*

0.6 0.4

Figure 6. The Isl1-Ajuba Transcriptional Complex Negatively Regulates Isl1 Expression in an RA-Dependent Mechanism

B

Fold change

Relative mRNA expression

A

Amhc

0 Isl1-luc Isl1 Ajuba

+ _ _

+ + _ Isl1

G

+ _

+ +

+ +

+ Isl1

+

(A) Relative expression levels of cardiac progenitor marker mRNA normalized to GAPDH mRNA in day 5 EBs either nontreated or treated with 0.1 or 0.3 mM RA at day 3. Data are mean ± SD (n = 2). *p < 0.05; **p < 0.01; ***p < 0.001. (B) Western blot analysis of nuclear extracts from day 5 EBs either nontreated or treated with 0.1 or 0.3 mM RA at day 3. (C) ChIP of nuclear extracts from day 5 EBs either nontreated or treated with 0.1 mM RA at day 3, using anti-Isl1 and anti-Ajuba antibodies or IgG as a control. PCRs were performed using primers flanking a conserved Isl1 binding site in the Isl1 enhancer. (D and E) P19 cells were transiently transfected with 25 ng Isl1-luc reporter construct (Isl1-F; Kang et al., 2009), alone or together with Isl1 (400 ng), Ajuba (900 ng), or with constant amount of Isl1 (400 ng) and increasing amounts of Ajuba expression plasmid (500 and 900 ng). P19 cell were either nontreated (D) or treated with RA (E). Data are mean ± SD (n = 4). **p < 0.01; ***p < 0.001. (F) In situ hybridization for the atrial marker atrial myosin heavy chain (Amhc; top panels) and Isl1 (bottom panels) at 48 hpf in control and zebrafish embryos, treated with RA for 1 hr. WT, wild-type. (G) In situ hybridization for Isl1 in control and Ajuba morphant and RA-treated control or Ajuba morphant zebrafish embryos at 48 hpf. See also Figure S7 and Tables S2 and S3.

and Ajuba together with a luciferase reporter construct containing the Isl1 enhancer and a minimal fos promoter (Isl1-luc; Figure 6D). Importantly, expression of Isl1 or Ajuba alone did not significantly affect Isl1 reporter gene RA RA expression, but coexpression of both 48h Ajuba MO 48h Control WT 48h RA 48h proteins resulted in downregulation of reporter gene levels. RA treatment of P19 cells transfected with Isl1 alone was Furthermore, we monitored a strong upregulation of Ajuba sufficient to repress Isl1 reporter gene activity (Figure 6E). This protein levels (Figure 6B). Based on our findings that Isl1 and effect was not dependent on exogenously introduced Ajuba, Ajuba form a complex and that Ajuba restricts the number of presumably because the Ajuba gene was already strongly actiIsl1+ cardiac progenitor cells, we analyzed whether the Isl1- vated by RA in P19 cells (Figures 6B and 6E) (Kanungo et al., Ajuba complex might directly regulate Isl1 expression. 2000). To further determine whether the effects of RA on cardiac Sequence analysis of the Isl1 enhancer, which is sufficient to cells are mediated via the Isl1-Ajuba complex, we treated control direct expression of Isl1 in the SHF and its derivatives (Kang and Ajuba morphant embryos with different concentrations of et al., 2009), revealed the presence of a putative Isl1-binding RA at 50% epiboly. RA treatment led to shortening of the site (TTAATGG; Figure 6C). ChIP analysis of nuclear extracts atrium (Figure 6F; Table S2) and to a significant reduction of from day 5 EBs differentiated from ES cells demonstrated the number of Isl1+ cells at the venous pole of the atrium at 48 specific binding of Isl1 and Ajuba to the conserved Isl1-binding hpf (Figure 6F; Table S2), which went along with a decrease of site. Interestingly, we detected stronger binding of Isl1 and Ajuba Isl1 expression at the 10 somite stage (Figures S7A and S7B), upon RA treatment of EBs at day 3 (Figure 6C), supporting the similar to RA treatment of mouse embryos (Ryckebusch et al., idea that RA signaling modulates Isl1 expression through regu- 2008). Furthermore, treatment of embryos with DEAB, an inhiblating the abundance of the Isl1-Ajuba complex on the Isl1 regu- itor of the RA-producing enzyme retinaldehyde dehydrogenase, latory elements. To further test this hypothesis, we transfected led to expansion of Isl1 expression, confirming that RA plays P19 embryonic carcinoma cells with plasmids encoding Isl1 a role in restricting the number of Isl1+ cardiac cells (Figure S7C).

WT

48h RA Isl1

48h Isl1

Control

48h Ajuba MO 48h Isl1 Isl1

Developmental Cell 23, 58–70, July 17, 2012 ª2012 Elsevier Inc. 65

Developmental Cell Ajuba Restricts the Cardiac Progenitor Pool

A

B

Control

Ajuba MO

Nkx2.5 15s

15s

Control

Control Ajuba MO

Nkx2.5

Nkx2.5

Ajuba MO

Relative Nkx2.5 expression

Nkx2.5 10s

10s

1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

Isl1 5s

G 3s

Nkx-deficient

H

Nkx-deficient

_ +

+ +

α-Nkx2.5 WB: α-myc **

**

α-LaminB

E HA-Ajuba _ Nkx2.5-myc FLAG-H3.3 α-HA

+ +

F + +

+ +

+ + α-Nkx2.5

α-myc 10s 15s

α-FLAG

Ponceau

Isl1 15s

Control

Isl1 48h

Control

Isl1 15s

Nkx-deficient

RA

Nkx-deficient

Nkx-deficient

RA CF

SHF

Isl1 Isl1

SHF

Isl1

Isl1

Control

Isl1 48h

CF Ajuba

Figure 7. Ajuba Regulates Nkx2.5 Levels

D

IP: α-HA

Isl1 10s

Isl1 5s

3s

+ _

Nkx2.5-myc 10% Input

Isl1 10s

Control

Control

C HA-Ajuba

Isl1

Isl1

(A) In situ hybridization for Nkx2.5 in control and Ajuba morphant embryos at 10 and 15 somites. (B) Relative expression level of Nkx2.5 normalized to Rpl13a in pools of control and Ajuba morphant embryos at 10 and 15 somites. Error bars represent SDs derived from three independent pools of control (n = 10) and Ajuba morphant embryos (n = 10). **p < 0.01. (C) FLAG-HA-tagged Ajuba and Nkx2.5-myc expression plasmids were transiently expressed in HEK293T cells, and immunoprecipitation with an anti-HA antibody followed by immunoblot analysis with an anti-myc antibody was performed. (D) Western blot analysis of lysates from day 5 EBs overexpressing either GFP only or both GFP and Ajuba using Nkx2.5 antibody. The two lanes were spliced from the same exposure of the same blot. (E) HEK293T cells were transfected with constant amounts of Nkx2.5-myc (5 mg) and FLAG-H3.3 (1 mg, used as a control for transfection efficiency) expression plasmids and increasing amounts of an Ajuba plasmid (0.25, 0.5, 0.75 mg). Immunoblot analysis of equal amounts of total protein extracts was performed using either anti-myc, anti-HA or anti-FLAG antibodies. (F) Western blot analysis of whole-cell lysates from HEK293T cells treated with either DMSO or MG132 using Nkx2.5 antibody. Ponceau staining served as loading control. (G) In situ hybridization of control and Nkx2.5/ Nkx2.7 morphant (Nkx-deficient) embryos at 3, 5, 10, and 15 somite stages and at 48 hpf with an Isl1 probe. (H) Model for the control of the SHF progenitor pool by Ajuba.

Ub Ub Ub

Nkx2.5

Nkx2.5

?

?

Ajuba

WT

Importantly, administration of RA to control and Ajuba morphant embryos did not decrease the higher number of Isl1+ cells in Ajuba morphants, whereas treatment of control embryos led to a significant reduction of Isl1+ cells in the heart (Figure 6G; Table S3). Taken together, our data suggest that the Isl1-Ajuba transcriptional complex negatively regulates Isl1 expression in an RA-dependent manner. The Protein Level of Ajuba Correlates with that of Nkx2.5 To assess whether additional factors besides Isl1 may contribute to the phenotype of Ajuba morphants, we analyzed the expression of Nkx2.5, which has also been shown to prevent overspecification of cardiac progenitors (Prall et al., 2007). In situ hybridization analysis of 10 and 15 somite embryos indicated that the Nkx2.5 expression domain is reduced in Ajuba morphants (Figure 7A). Real-time PCR analysis of pools of control and Ajuba morphant embryos at 10 and 15 somites confirmed lower Nkx2.5 mRNA levels in the morphants (Figure 7B). 66 Developmental Cell 23, 58–70, July 17, 2012 ª2012 Elsevier Inc.

Because Ajuba was found to interact with another member of the Nkx2 class of homeodomain-containing proteins, Nkx2.1 (Missero et al., 2001), we wanted Ajuba MO to determine whether Ajuba also binds Nkx2.5. Coimmunoprecipitation experiments of HEK293T cells transfected with Nkx2.5 and Ajuba expression constructs demonstrated that Nkx2.5 indeed interacts with Ajuba (Figure 7C). Interestingly, we consistently observed dramatically higher Nkx2.5 protein levels in extracts (Figures 7C and 7D) and immunostaining (data not shown) of cells overexpressing Ajuba, which suggested that Ajuba may stabilize Nkx2.5, e.g., by protecting it from proteasomal degradation. To assess this possibility, we transfected HEK293T cells with Nkx2.5 alone or together with increasing amounts of Ajuba and constant amounts of histone 3.3 (H3.3), which served as a control for transfection efficiency (Figure 7E). Western blot analysis revealed that increasing levels of Ajuba led to a significant increase of Nkx2.5 protein levels (Figure 7E). Furthermore, treatment of HEK293T cells, which express low levels of endogenous Nkx2.5, with the proteasomal inhibitor MG-132 resulted in an increase of Nkx2.5 protein levels, indicating that Nkx2.5 is targeted for proteasomal degradation (Figure 7F). Interestingly, Nkx-deficient embryos that lack both Nkx2.5 and Nkx2.7 show enlarged atria (Targoff et al., 2008;

Developmental Cell Ajuba Restricts the Cardiac Progenitor Pool

Tu et al., 2009) similar to Ajuba morphants. To determine whether the enlarged size of the atrium in Nkx-deficient embryos might be caused by an increase of Isl1+ cardiac progenitors, we performed in situ hybridization for Isl1 using control and Nkx-deficient embryos at 3, 5, 10, and 15 somites and 48 hpf. We detected a dramatic expansion of Isl1+ cardiac progenitor cells as early as at the 3 somite stage, suggesting that Nkx2.5 regulates Isl1 progenitor specification in a similar manner as Ajuba (Figure 7G). Later, at 48 hpf, a large portion of the atrium consisted of Isl1+ cardiomyocytes (Figure 7G). Taken together, these findings suggest that downregulation of Nkx2 proteins may contribute to the Ajuba morphant phenotype. DISCUSSION The control of cardiac progenitor cell specification, expansion, and differentiation is essential for cardiac development and requires the coordinated function of different transcription factors and signaling pathways. The transcription factors Isl1 and Nkx2.5 as well as RA signaling play important roles in the regulation of the cardiac progenitor pool, but the precise mechanisms controlling their activity remain unclear. We reasoned that identification of molecules that control the activity of Isl1 might be a means to uncover the mechanisms that determine cardiac progenitor cell numbers and mediate RA-dependent restriction of SHF expansion. Our experiments identified Ajuba, a member of the Ajuba/Zyxin protein family, as an Isl1-binding partner. Ajuba is important for the formation and the maintenance of cell-cell junctions (Marie et al., 2003) and plays a role in cell migration (Kisseleva et al., 2005; Pratt et al., 2005). Ajuba shuttles into the nucleus and affects proliferation and cell fate decisions, a process that is RA dependent (Kanungo et al., 2000). Moreover, Ajuba acts as a corepressor of Snail/Slug (Hou et al., 2008; Langer et al., 2008) and of nuclear receptors (Hou et al., 2010). During embryonic development Ajuba is initially expressed rather broadly, including the SHF; however, it shows distinct subcellular localization in different tissues (Figures 2E and S4). Later, at E12.5 in mouse embryos and 48 hpf in zebrafish embryos (Goyal et al., 1999), it acquires a more specific expression pattern (data not shown). This temporal and spatial expression pattern suggests that Ajuba may play a regulatory role during early organogenesis. Inactivation of Ajuba led to defects in neural crest development in Xenopus (Langer et al., 2008) and left–right axis determination in medaka, due to defects in ciliogenesis (Nagai et al., 2010). No developmental defects in single knockouts of the Ajuba protein family members, Ajuba and LIMD1, were observed in mice, which suggests a potential compensation and redundant functions within the Ajuba/Zyxin protein family members during mouse development (Feng et al., 2007; Marie et al., 2003). More recently, however, it was reported that the Ajuba family member WTIP, which is also expressed in the developing heart, plays an important role during early embryonic development because no WTIP-deficient embryos were observed as early as E8.5 (Kim et al., 2012). We have gathered compelling evidence that Ajuba proteins, which interact with Isl1, restrict the activity of Isl1, thereby playing an important role in regulating SHF expansion and morphogenetic control of cardiogenesis. Notably, Ajuba-deficient

zebrafish embryos displayed an increased pool of Isl1+ cardiac progenitors and, subsequently, dramatically increased numbers of atrial and ventricular cardiomyocytes, derived from the SHF. The higher number of atrial cells was due to dramatically higher numbers of mGFP+/dsRed
Isl1-positive atrial cardiomyocytes at the venous pole of the heart tube, which are added later to the heart tube in a process that requires Isl1 activity (Figures 3C and 3D; de Pater et al., 2009). Furthermore, we detected a significantly enlarged outflow pole, which can be explained by enhanced contribution of the increased pool of Isl1+ cardiac progenitors to the arterial pole (Figures 3E and 3F; Hami et al., 2011). Additionally, the ability of Ajuba to limit cardiomyocyte numbers at both poles of the heart depends on Isl1 because the double Isl1/Ajuba-deficient zebrafish embryos did not show any increase in the total number of cardiomyocytes in contrast to the Ajuba morphants (Figure 4E; Figure S3E). Because previous reports have not observed defects in the arterial pole of Isl1-deficient zebrafish embryos (de Pater et al., 2009), our results might indicate that Ajuba controls additional factors besides Isl1, which are involved in the differentiation of the arterial pole, e.g., Nkx2.5 or other Isl1 homologs that we detect expressed at the arterial but not at the venous pole of the heart (unpublished data). Additionally, we observed cardiomyocytes in the pericardial wall adjacent to the heart in Isl1/ Ajuba morphants (Figure 4E, arrows). One possible explanation could be that Ajuba and Isl1 are required to direct the migration of cardiac cells into the heart tube, consistent with the proposed role of Ajuba and Isl1 in regulating migration (Cai et al., 2003; Pratt et al., 2005). Interestingly, we observed dramatically expanded Isl1+ cell population in Ajuba, Isl1, and Isl1/Ajuba morphants. However, in contrast to the control and Ajuba morphant embryos, where Isl1+ cardiomyocytes are localized at the venous pole, the dramatically expanded Isl1+ cell population in both Isl1 and the double Isl1/Ajuba-deficient embryos was found outside of the heart tube. Taken together, these data suggest that Ajuba and Isl1 restrict the number of Isl1+ cells, whereas Isl1 presumably in collaboration with other cofactors drives the differentiation of cardiomyocytes (Figure 4E–4G and 7H), consistent with previous reports (Kwon et al., 2009). The significant upregulation of Mef2ca, Mef2cb, Tbx20, Tbx1, and Bmp4, which play important roles in the formation of the SHF (Hami et al., 2011; Lazic and Scott, 2011; Vincent and Buckingham, 2010), after Ajuba depletion and the increased number of Isl1+ cardiac progenitors in Ajuba morphants suggests an important role for Ajuba in restricting the SHF progenitor cell pool. Furthermore, our data demonstrated dramatically increased specification of cardiac progenitors at early stages of heart development in Ajuba-deficient embryos (Figure 5B). Moreover, our results suggest that Ajuba directly represses Isl1 expression by binding to the Isl1 enhancer in an Isl1-dependent mechanism (Figure 6C–6E). Additionally, Ajuba suppresses Isl1-mediated activation of Mef2c in the SHF because Ajuba and Isl1 form a complex at the Mef2c AHF enhancer. The negative regulatory role of Ajuba for Isl1 and the SHF was further supported by gain-of-function experiments in the mouse system, where we detected a significantly decreased expression of Isl1 and other SHF markers in EBs overexpressing Ajuba. We assume that the function of Ajuba proteins is conserved during evolution, although the relative contribution of the different family Developmental Cell 23, 58–70, July 17, 2012 ª2012 Elsevier Inc. 67

Developmental Cell Ajuba Restricts the Cardiac Progenitor Pool

members and the level of redundancy between them might differ. The contribution of Ajuba to cardiovascular development in the mouse might be masked by the action of WTIP, which is highly related to Ajuba. Although Ajuba-deficient mice do not show a cardiovascular phenotype, ablation of WTIP, which is expressed in the developing heart, is embryonically lethal before E8.5 (Kim et al., 2012). Furthermore, WTIP depletion in zebrafish embryos results in an increase of Isl1+ cells at the venous pole, analogous to the one observed for Ajuba. Alternatively, the lack of a cardiac phenotype in mouse Ajuba mutants may be due to differences in the properties of the SHF cardiac progenitors in zebrafish and mouse because proliferating Isl1+ cardiac progenitor cells are rare in zebrafish (Figure S6), but abundant in mouse embryos (Cai et al., 2003), a property essential for the development of the four-chambered heart in the mouse. RA signaling regulates key aspects of heart development and is required to limit the posterior extent of the SHF (Ryckebusch et al., 2008; Sirbu et al., 2008). In the zebrafish embryo, RA has been shown to restrict the cardiac progenitor cell pool, and Raldh2 deficiency causes the formation of enlarged hearts composed of higher numbers of both atrial and ventricular cells (Keegan et al., 2005; Waxman et al., 2008). Our data suggest that Ajuba is an important downstream effector of RA signaling that regulates the cardiac progenitor pool, based on the following lines of evidence: (1) RA treatment led to upregulation of Ajuba and its accumulation in the nucleus (Figure 6B), which resulted in concomitant binding of Isl1 and Ajuba to the Isl1 enhancer and downregulation of Isl1 expression (Figures 6C–6E); and (2) RA-mediated restriction of the number of Isl1+ cells in the zebrafish heart depends on Ajuba because treatment of Ajuba morphants with RA failed to affect the number of Isl1+ cells (Figure 6G). Taken together, our data suggest that the Isl1-Ajuba transcriptional complex negatively regulates Isl1 expression downstream of RA signaling. It has been shown that a Nkx2.5/Bmp2/Smad1 negative feedback loop controls cardiac progenitor cell specification and proliferation (Prall et al., 2007). Interestingly, Nkx2.5 was downregulated in Ajuba morphants. Additionally, exogenously expressed Nkx2.5 interacted with Ajuba, and overexpression of Ajuba correlated with increased Nkx2.5 levels. Given the fact that Nkx2.5 levels are tightly regulated in cardiac progenitors and increase from progenitor to the differentiated state, it is tempting to speculate that Ajuba might be an important regulator of this process that is crucial for the early events of cardiac development (Prall et al., 2007). We suggest that downregulation of Nkx2.5 in Ajuba morphants may contribute to increased specification of Isl1+ cardiac progenitors, similar to the studies in Nkx2.5 knockout mice (Prall et al., 2007). However, further work will be required to determine the mechanisms through which Ajuba affects Nkx2.5 levels and how this is linked to the restriction of the Isl1+ progenitor pool. We propose a model in which Ajuba restricts the SHF progenitor pool by regulation of the activity and stability of two key transcription factors in heart development, Isl1 and Nkx2.5 (Figure 7H). Ajuba binds Isl1 and represses its transcriptional activity, which on the one hand is required to downregulate the expression of key transcription factors in the SHF such as Mef2c and on the other hand enables Isl1 to suppress its own expression, a process that is RA dependent. According to 68 Developmental Cell 23, 58–70, July 17, 2012 ª2012 Elsevier Inc.

this model, stabilization of Nkx2.5 by Ajuba may be involved in preventing cardiac progenitor cell overspecification. EXPERIMENTAL PROCEDURES Cell Culture and Transfection HEK293T, NIH 3T3, and p19 cells were grown in DMEM (Invitrogen) supplemented with 10% fetal bovine serum (FBS) (Invitrogen), 2 mM L-glutamine, 100 U/ml penicillin, and 100 mg/ml streptomycin (Invitrogen) at 37 C, 5% CO2. Undifferentiated ES cells were maintained on mouse embryonic fibroblast (MEF) feeder cells in DMEM supplemented with 15% FBS, 2 mM L-glutamine, 0.1 mM 2-mercaptoethanol (Sigma-Aldrich), 0.1 mM nonessential amino acids (Invitrogen), 1 mM sodium pyruvate (Invitrogen), 4.5 mg/ml D-glucose, and 1,000 U/ml of leukemia inhibitory factor (LIF ESGRO; Millipore). To induce EB formation, dissociated ES cells were cultured in hanging drops, with a density of 500 cells per drop in 15 ml of ES cell medium, in the absence of LIF. After 2 days in the hanging drop culture, the resulting EBs were transferred to bacterial culture dishes. Plasmids and Antibodies For detailed plasmid information, see the Supplemental Experimental Procedures. The following primary antibodies were used: rabbit anti-HA (Y-11; Santa Cruz Biotechnology); rabbit anti-Nkx2.5 (H-114; Santa Cruz Biotechnology); mouse anti-myc (9E11; Santa Cruz Biotechnology); mouse anti-Islet1 39.4D5 (Developmental Studies Hybridoma Bank); chicken anti-GFP (Millipore); goat anti-Lamin B (C-20; Santa Cruz Biotechnology); and anti-Ajuba (Cell Signaling; H-155 [Santa Cruz Biotechnology]). Immunoprecipitation, Immunoblotting, ChIP, and Immunofluorescence Detailed protocols are described in Supplemental Experimental Procedures. Luciferase Assay For luciferase assays, 1.5–2 3 104 HEK293T or NIH 3T3 cells were seeded in 24-well plates. Forty-eight hours after transfection, cells were lysed in 100 ml lysis buffer (Promega; Luciferase Assay System), and luciferase activity was measured on a Mithras LB 940 reader (Berthold Technologies) according to the Luciferase Assay System Manual. b-Galactosidase assays were performed using chlorophenolred-b-d-galactopyranoside (CPRG) substrate (Sigma-Aldrich). RNA Isolation, RT-PCR, and Real-Time PCR RNA was isolated using the Total RNA Isolation Reagent (ABgene). cDNA was synthesized with the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems), and real-time PCR was performed using the SYBR Green PCR Master Mix (Applied Biosystems). The cycle numbers were normalized to GAPDH (ES/EBs) or Rpl13a (zebrafish embryos). Primer pairs are described in Supplemental Experimental Procedures. Zebrafish Strains Animal experiments were performed in accordance with the institutional guidelines. Embryos and adult zebrafish were raised under standard laboratory conditions at 28 C. The following mutant and transgenic lines were used: Tg(myl7:EGFP-HsHRAS)s883 and Tg(
5.1myl7:nDsRed2)f2 (Mably et al., 2003; D’Amico et al., 2007). To generate the Nkx2.5:GFP line, the 6 kb upstream regulatory region of the nkx2.5 gene was amplified using primer pairs (50 -GGGGACAACTTTGTATAGAAAAGTTGCCCAAGCCCACTTACCAA TA-30 and 50 -GGGGACTGCTTTTTTGTACAAACTTGGGATAATCCGGTTGGG ATTT-30 ) and combined into pME-GFP together with p3E-polyA and pDestTol2pA2 to generate nkx2.5:GFP:pA construct, which was injected into one-cell stage embryos. The generation of the nkx2.5:GFP line is covered in an approved animal protocol by the University of Southern California Institutional Animal Care and Use Committee. Morpholinos MOs (GeneTools) were prepared at a stock concentration of 1 mM and diluted to the desired concentration for microinjection. To block translation, embryos

Developmental Cell Ajuba Restricts the Cardiac Progenitor Pool

were injected with 9 ng of the Ajuba AUG MO: 50 -TGAGTTTGATGCCAAGTC GATCCAT-30 . To block splicing, embryos were injected with 9 ng of Ajuba Sp MO: 50 -GATGGGTTTGTGATCTCACCGAAGT-30 . Uninjected and control MO-injected embryos served as controls. Nkx2.5/Nkx2.7 MO knockdown experiments were performed as described (Targoff et al., 2008). Isl1 MO knockdown experiments were performed as described (Hutchinson and Eisen, 2006). In Situ Hybridization, Whole-Mount Immunostaining, and Confocal Microscopy In situ hybridization was performed as described in Thisse et al. (2004). Wholemount staining was performed as described (Dong et al., 2007). Confocal images were acquired by a Zeiss LSM 710 system, and the z stacks were projected by Zeiss LSM 710 software. RA and DEAB Treatments For RA treatment embryos were exposed to 0.1 or 0.3 mM RA, 1% DMSO in E3 medium (fish water) for 1 hr beginning at 50% epiboly. DEAB (4-diethylaminobenzaldehyde; Sigma-Aldrich) in 1% DMSO, E3 medium was applied at final concentration of 5 mM. SUPPLEMENTAL INFORMATION Supplemental Information includes seven figures, three tables, and Supplemental Experimental Procedures and can be found with this article online at http://dx.doi.org/10.1016/j.devcel.2012.06.005. ACKNOWLEDGMENTS We thank Dr. Gregory Longmore for generously providing the pCS2-Limd1, pCS2-WTIP, pCS2-LPP, and pCS2-Zyxin plasmids. We thank Ingrid Konrad and Monika Mu¨ller-Boche for excellent technical assistance. We are grateful to Boyan Garvalov, Luca Caputo, Sonja Hundt, Carina Detzer, Aditi Mehta, and Joaquim Vong for their support and stimulating discussions. This work was supported by Emmy-Noether Program Grant DO 1323/1-1, the DFG SFB TRR 81, the Excellence Cluster Cardio-Pulmonary System EXC 147 of the DFG (Germany), and the LOEWE Universities of Giessen and Marburg Lung Center (UGMLC) and LOEWE Center for Cell and Gene Therapy (CGT) financed by the state of Hesse. Received: April 29, 2011 Revised: March 16, 2012 Accepted: June 4, 2012 Published online: July 5, 2012 REFERENCES Azpiazu, N., and Frasch, M. (1993). tinman and bagpipe: two homeo box genes that determine cell fates in the dorsal mesoderm of Drosophila. Genes Dev. 7 (7B), 1325–1340. Black, B.L. (2007). Transcriptional pathways in second heart field development. Semin. Cell Dev. Biol. 18, 67–76. Bodmer, R. (1993). The gene tinman is required for specification of the heart and visceral muscles in Drosophila. Development 118, 719–729. Brade, T., Gessert, S., Ku¨hl, M., and Pandur, P. (2007). The amphibian second heart field: Xenopus islet-1 is required for cardiovascular development. Dev. Biol. 311, 297–310. Brand, T. (2003). Heart development: molecular insights into cardiac specification and early morphogenesis. Dev. Biol. 258, 1–19. Bruneau, B.G. (2008). The developmental genetics of congenital heart disease. Nature 451, 943–948.

differentiation and contributes a majority of cells to the heart. Dev. Cell 5, 877–889. Chen, J.N., and Fishman, M.C. (1996). Zebrafish tinman homolog demarcates the heart field and initiates myocardial differentiation. Development 122, 3809– 3816. D’Amico, L., Scott, I.C., Jungblut, B., and Stainier, D.Y. (2007). A mutation in zebrafish hmgcr1b reveals a role for isoprenoids in vertebrate heart-tube formation. Curr. Biol. 17, 252–259. de Pater, E., Clijsters, L., Marques, S.R., Lin, Y.F., Garavito-Aguilar, Z.V., Yelon, D., and Bakkers, J. (2009). Distinct phases of cardiomyocyte differentiation regulate growth of the zebrafish heart. Development 136, 1633–1641. Dodou, E., Verzi, M.P., Anderson, J.P., Xu, S.M., and Black, B.L. (2004). Mef2c is a direct transcriptional target of ISL1 and GATA factors in the anterior heart field during mouse embryonic development. Development 131, 3931–3942. Dong, P.D., Munson, C.A., Norton, W., Crosnier, C., Pan, X., Gong, Z., Neumann, C.J., and Stainier, D.Y. (2007). Fgf10 regulates hepatopancreatic ductal system patterning and differentiation. Nat. Genet. 39, 397–402. Evans, S.M., Yelon, D., Conlon, F.L., and Kirby, M.L. (2010). Myocardial lineage development. Circ. Res. 107, 1428–1444. Feng, Y., Zhao, H., Luderer, H.F., Epple, H., Faccio, R., Ross, F.P., Teitelbaum, S.L., and Longmore, G.D. (2007). The LIM protein, Limd1, regulates AP-1 activation through an interaction with Traf6 to influence osteoclast development. J. Biol. Chem. 282, 39–48. Goyal, R.K., Lin, P., Kanungo, J., Payne, A.S., Muslin, A.J., and Longmore, G.D. (1999). Ajuba, a novel LIM protein, interacts with Grb2, augments mitogen-activated protein kinase activity in fibroblasts, and promotes meiotic maturation of Xenopus oocytes in a Grb2- and Ras-dependent manner. Mol. Cell. Biol. 19, 4379–4389. Hami, D., Grimes, A.C., Tsai, H.J., and Kirby, M.L. (2011). Zebrafish cardiac development requires a conserved secondary heart field. Development 138, 2389–2398. Hochgreb, T., Linhares, V.L., Menezes, D.C., Sampaio, A.C., Yan, C.Y., Cardoso, W.V., Rosenthal, N., and Xavier-Neto, J. (2003). A caudorostral wave of RALDH2 conveys anteroposterior information to the cardiac field. Development 130, 5363–5374. Hou, Z., Peng, H., Ayyanathan, K., Yan, K.P., Langer, E.M., Longmore, G.D., and Rauscher, F.J., 3rd. (2008). The LIM protein AJUBA recruits protein arginine methyltransferase 5 to mediate SNAIL-dependent transcriptional repression. Mol. Cell. Biol. 28, 3198–3207. Hou, Z., Peng, H., White, D.E., Negorev, D.G., Maul, G.G., Feng, Y., Longmore, G.D., Waxman, S., Zelent, A., and Rauscher, F.J., 3rd. (2010). LIM protein Ajuba functions as a nuclear receptor corepressor and negatively regulates retinoic acid signaling. Proc. Natl. Acad. Sci. USA 107, 2938–2943. Hutchinson, S.A., and Eisen, J.S. (2006). Islet1 and Islet2 have equivalent abilities to promote motoneuron formation and to specify motoneuron subtype identity. Development 133, 2137–2147. Kang, J., Nathan, E., Xu, S.M., Tzahor, E., and Black, B.L. (2009). Isl1 is a direct transcriptional target of Forkhead transcription factors in second-heart-fieldderived mesoderm. Dev. Biol. 334, 513–522. Kanungo, J., Pratt, S.J., Marie, H., and Longmore, G.D. (2000). Ajuba, a cytosolic LIM protein, shuttles into the nucleus and affects embryonal cell proliferation and fate decisions. Mol. Biol. Cell 11, 3299–3313. Keegan, B.R., Feldman, J.L., Begemann, G., Ingham, P.W., and Yelon, D. (2005). Retinoic acid signaling restricts the cardiac progenitor pool. Science 307, 247–249. Kelly, R.G., Brown, N.A., and Buckingham, M.E. (2001). The arterial pole of the mouse heart forms from Fgf10-expressing cells in pharyngeal mesoderm. Dev. Cell 1, 435–440.

Buckingham, M., Meilhac, S., and Zaffran, S. (2005). Building the mammalian heart from two sources of myocardial cells. Nat. Rev. Genet. 6, 826–835.

Kim, J.H., Mukherjee, A., Madhavan, S., Konieczkowski, M., and Sedor, J. (2012). Wt1 interacting protein (Wtip) regulates podocyte phenotype by cellcell and cell-matrix contact reorganization. Am. J. Physiol. Renal Physiol. 302, F103–F115.

Cai, C.L., Liang, X., Shi, Y., Chu, P.H., Pfaff, S.L., Chen, J., and Evans, S. (2003). Isl1 identifies a cardiac progenitor population that proliferates prior to

Kisseleva, M., Feng, Y., Ward, M., Song, C., Anderson, R.A., and Longmore, G.D. (2005). The LIM protein Ajuba regulates phosphatidylinositol

Developmental Cell 23, 58–70, July 17, 2012 ª2012 Elsevier Inc. 69

Developmental Cell Ajuba Restricts the Cardiac Progenitor Pool

4,5-bisphosphate levels in migrating cells through an interaction with and activation of PIPKI alpha. Mol. Cell. Biol. 25, 3956–3966. Kwon, C., Qian, L., Cheng, P., Nigam, V., Arnold, J., and Srivastava, D. (2009). A regulatory pathway involving Notch1/beta-catenin/Isl1 determines cardiac progenitor cell fate. Nat. Cell Biol. 11, 951–957. Langer, E.M., Feng, Y., Zhaoyuan, H., Rauscher, F.J., 3rd, Kroll, K.L., and Longmore, G.D. (2008). Ajuba LIM proteins are snail/slug corepressors required for neural crest development in Xenopus. Dev. Cell 14, 424–436. Laugwitz, K.L., Moretti, A., Lam, J., Gruber, P., Chen, Y., Woodard, S., Lin, L.Z., Cai, C.L., Lu, M.M., Reth, M., et al. (2005). Postnatal isl1+ cardioblasts enter fully differentiated cardiomyocyte lineages. Nature 433, 647–653. Laugwitz, K.L., Moretti, A., Caron, L., Nakano, A., and Chien, K.R. (2008). Islet1 cardiovascular progenitors: a single source for heart lineages? Development 135, 193–205. Lazic, S., and Scott, I.C. (2011). Mef2cb regulates late myocardial cell addition from a second heart field-like population of progenitors in zebrafish. Dev. Biol. 354, 123–133. Liberatore, C.M., Searcy-Schrick, R.D., and Yutzey, K.E. (2000). Ventricular expression of tbx5 inhibits normal heart chamber development. Dev. Biol. 223, 169–180. Lyons, I., Parsons, L.M., Hartley, L., Li, R., Andrews, J.E., Robb, L., and Harvey, R.P. (1995). Myogenic and morphogenetic defects in the heart tubes of murine embryos lacking the homeo box gene Nkx2-5. Genes Dev. 9, 1654–1666. Mably, J.D., Mohideen, M.A., Burns, C.G., Chen, J.N., and Fishman, M.C. (2003). heart of glass regulates the concentric growth of the heart in zebrafish. Curr. Biol. 13, 2138–2147. Marie, H., Pratt, S.J., Betson, M., Epple, H., Kittler, J.T., Meek, L., Moss, S.J., Troyanovsky, S., Attwell, D., Longmore, G.D., and Braga, V.M. (2003). The LIM protein Ajuba is recruited to cadherin-dependent cell junctions through an association with alpha-catenin. J. Biol. Chem. 278, 1220–1228. Missero, C., Pirro, M.T., Simeone, S., Pischetola, M., and Di Lauro, R. (2001). The DNA glycosylase T:G mismatch-specific thymine DNA glycosylase represses thyroid transcription factor-1-activated transcription. J. Biol. Chem. 276, 33569–33575. Nagai, Y., Asaoka, Y., Namae, M., Saito, K., Momose, H., Mitani, H., FurutaniSeiki, M., Katada, T., and Nishina, H. (2010). The LIM protein Ajuba is required for ciliogenesis and left-right axis determination in medaka. Biochem. Biophys. Res. Commun. 396, 887–893. Olson, E.N. (2006). Gene regulatory networks in the evolution and development of the heart. Science 313, 1922–1927.

70 Developmental Cell 23, 58–70, July 17, 2012 ª2012 Elsevier Inc.

Prall, O.W., Menon, M.K., Solloway, M.J., Watanabe, Y., Zaffran, S., Bajolle, F., Biben, C., McBride, J.J., Robertson, B.R., Chaulet, H., et al. (2007). An Nkx2-5/ Bmp2/Smad1 negative feedback loop controls heart progenitor specification and proliferation. Cell 128, 947–959. Pratt, S.J., Epple, H., Ward, M., Feng, Y., Braga, V.M., and Longmore, G.D. (2005). The LIM protein Ajuba influences p130Cas localization and Rac1 activity during cell migration. J. Cell Biol. 168, 813–824. Ryckebusch, L., Wang, Z., Bertrand, N., Lin, S.C., Chi, X., Schwartz, R., Zaffran, S., and Niederreither, K. (2008). Retinoic acid deficiency alters second heart field formation. Proc. Natl. Acad. Sci. USA 105, 2913–2918. Sirbu, I.O., Zhao, X., and Duester, G. (2008). Retinoic acid controls heart anteroposterior patterning by down-regulating Isl1 through the Fgf8 pathway. Dev. Dyn. 237, 1627–1635. Srivastava, D. (2006a). Genetic regulation of cardiogenesis and congenital heart disease. Annu. Rev. Pathol. 1, 199–213. Srivastava, D. (2006b). Making or breaking the heart: from lineage determination to morphogenesis. Cell 126, 1037–1048. Stainier, D.Y., and Fishman, M.C. (1992). Patterning the zebrafish heart tube: acquisition of anteroposterior polarity. Dev. Biol. 153, 91–101. Targoff, K.L., Schell, T., and Yelon, D. (2008). Nkx genes regulate heart tube extension and exert differential effects on ventricular and atrial cell number. Dev. Biol. 322, 314–321. Thisse, B., Heyer, V., Lux, A., Alunni, V., Degrave, A., Seiliez, I., Kirchner, J., Parkhill, J.P., and Thisse, C. (2004). Spatial and temporal expression of the zebrafish genome by large-scale in situ hybridization screening. Methods Cell Biol. 77, 505–519. Tu, C.T., Yang, T.C., and Tsai, H.J. (2009). Nkx2.7 and Nkx2.5 function redundantly and are required for cardiac morphogenesis of zebrafish embryos. PLoS One 4, e4249. Vincent, S.D., and Buckingham, M.E. (2010). How to make a heart: the origin and regulation of cardiac progenitor cells. Curr. Top. Dev. Biol. 90, 1–41. Waldo, K.L., Kumiski, D.H., Wallis, K.T., Stadt, H.A., Hutson, M.R., Platt, D.H., and Kirby, M.L. (2001). Conotruncal myocardium arises from a secondary heart field. Development 128, 3179–3188. Waxman, J.S., Keegan, B.R., Roberts, R.W., Poss, K.D., and Yelon, D. (2008). Hoxb5b acts downstream of retinoic acid signaling in the forelimb field to restrict heart field potential in zebrafish. Dev. Cell 15, 923–934. Zhou, Y., Cashman, T.J., Nevis, K.R., Obregon, P., Carney, S.A., Liu, Y., Gu, A., Mosimann, C., Sondalle, S., Peterson, R.E., et al. (2011). Latent TGF-b binding protein 3 identifies a second heart field in zebrafish. Nature 474, 645–648.