Proteolysis regulates cardiomyocyte maturation and tissue integration

Abstract

Tissue integrity is critical for organ formation and function. During heart development, cardiomyocytes differentiate and integrate to form a coherent tissue that contracts synchronously. However, the molecular mechanisms regulating cardiac tissue integrity are poorly understood. Here we show that proteolysis, via the E3 ubiquitin ligase ASB2, regulates cardiomyocyte maturation and tissue integrity. Cardiomyocytes in asb2b zebrafish mutants fail to terminally differentiate, resulting in reduced cardiac contractility and output. Mosaic analyses reveal a cell-autonomous requirement for Asb2b in cardiomyocytes for their integration as asb2b mutant cardiomyocytes are unable to meld into wild-type myocardial tissue. In vitro and in vivo data indicate that ASB2 negatively regulates TCF3, a bHLH transcription factor. TCF3 must be degraded for cardiomyocyte maturation, as TCF3 gain-of-function causes a number of phenotypes associated with cardiomyocyte dedifferentiation. Overall, our results show that proteolysis has an important role in cardiomyocyte maturation and the formation of a coherent myocardial tissue.

ARTICLE

Received 20 Aug 2016 | Accepted 4 Jan 2017 | Published 17 Feb 2017

Ryuichi Fukuda1, Felix Gunawan1, Arica Beisaw1, Vanesa Jimenez-Amilburu1, Hans-Martin Maischein1,

Sawa Kostin1, Koichi Kawakami2 & Didier Y.R. Stainier1

Tissue integrity is critical for organ formation and function. During heart development, cardiomyocytes differentiate and integrate to form a coherent tissue that contracts synchronously. However, the molecular mechanisms regulating cardiac tissue integrity are poorly understood. Here we show that proteolysis, via the E3 ubiquitin ligase ASB2, regulates cardiomyocyte maturation and tissue integrity. Cardiomyocytes in asb2b zebrash mutants fail to terminally differentiate, resulting in reduced cardiac contractility and output. Mosaic analyses reveal a cell-autonomous requirement for Asb2b in cardiomyocytes for their integration as asb2b mutant cardiomyocytes are unable to meld into wild-type myocardial tissue. In vitro and in vivo data indicate that ASB2 negatively regulates TCF3, a bHLH transcription factor. TCF3 must be degraded for cardiomyocyte maturation, as TCF3 gain-of-function causes a number of phenotypes associated with cardiomyocyte dedifferentiation. Overall, our results show that proteolysis has an important role in cardiomyocyte maturation and the formation of a coherent myocardial tissue.

DOI: 10.1038/ncomms14495 OPEN

Proteolysis regulates cardiomyocyte maturation and tissue integration

1 Department of Developmental Genetics, Max Planck Institute for Heart and Lung Research, Ludwigstrasse 43, 61231 Bad Nauheim, Germany. 2 Division of Molecular and Developmental Biology, National Institute of Genetics, Yata 1111, Mishima, Shizuoka 411-8540, Japan. Correspondence and requests for materials should be addressed to R.F. (email: mailto:ryuichi.fukuda @mpi-bn.mpg.de

Web End =ryuichi.fukuda @mpi-bn.mpg.de ) or to D.Y.R.S. (email: mailto:[email protected]

Web End [email protected] ).

NATURE COMMUNICATIONS | 8:14495 | DOI: 10.1038/ncomms14495 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 1

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14495

The development of functional organs requires tissue integrity, which is achieved through the coordinated regulation and positioning of cellcell contacts1,2.

Cardiac tissue integrity is particularly important for cardiac function as the myocardial wall must withstand stress from mechanical forces3. Understanding the molecular mechanisms underlying tissue integrity is also crucial to develop effective tissue engineering protocols to generate functional and coherent myocardial tissue4. During cardiomyocyte differentiation, progenitor cells initiate the expression of sarcomeric genes57, and as cardiomyocytes mature they continue to develop more organized sarcomeric structures8, as well as highly specialized cellcell junctions termed intercalated disks9. Mature sarcomeres and cellcell junctions are essential for mechanical and electrical coupling of cardiomyocytes, and allow for synchronized contractions9. Although cardiomyocyte specication and differentiation are well-studied5,6, the molecular mechanisms regulating the maturation of sarcomeres and cellcell junctions to maintain cardiac tissue integrity remain poorly understood. Proteolysis is essential for various cellular phenomena and this process is mainly mediated by the ubiquitin-proteasomal system (UPS)10,11. Ankyrin repeat and SOCS box containing 2 (ASB2) is an E3 ubiquitin ligase that functions to specify target proteins for proteasome-dependent degradation. ASB2 is expressed in haematopoietic and muscle cells, including the heart12; however, its function in the heart is unknown.

Here we show that ASB2 functions to regulate cardiomyocyte maturation and myocardial tissue integrity. Three-dimensional (3D) imaging of hearts using several zebrash transgenic lines shows that asb2b mutant cardiomyocytes exhibit disorganized sarcomeres and cellcell junctions, as well as defects in apico-basal polarity, resulting in reduced cardiac contractility and output. Moreover, we show that ASB2 induces the proteasome-dependent degradation of TCF3. TCF3 over-expression in cardiomyocytes leads to the disassembly of sarcomeres, disruption of cellcell junctions and upregulation of immature markers, phenotypes associated with cardiomyocyte dedifferentiation. We propose that proteolysis via ASB2 regulates cardiomyocyte maturation and myocardial tissue integrity through the regulation of TCF3.

Resultsasb2b mutant zebrash exhibit cardiac defects. We initially investigated the global effects of proteasome inhibition on cardiomyocyte maturation. We treated 24 h post fertilization (hpf) zebrash embryos, whose cardiomyocytes already exhibit nascent sarcomeres, with bortezomib, a proteasome inhibitor, for 72 h. We found that the treated animals exhibited mild pericardial oedema and defects in myolament maturation, without obvious cardiomyocyte death or morphological defects (Supplementary Fig. 1a). To gain insight into the molecular mechanisms that mediate proteasome-dependent proteolysis during cardiomyocyte maturation, we examined the expression of various E3 ubiquitin ligase genes in developing zebrash hearts. We found that asb2b, whose expression starts shortly after cardiac specication (Supplementary Fig. 1b), becomes more highly expressed at later stages in the heart tube and somites (Fig. 1a), whereas its paralogue asb2a is expressed in somites and the head region (Supplementary Fig. 1c). To better examine asb2b expression, we generated a green uorescent protein (GFP) bacterial articial chromosome (BAC) reporter line, TgBAC(asb2b:GFP-GAL4FF), whose expression mimics asb2b expression as observed by in situ hybridization (Supplementary Fig. 1d). When observed in combination

with a cardiomyocyte-specic reporter line, Tg(myl7:MKATECAAX)13, we found GFP expression in the cardiomyocytes (Fig. 1b). We also examined the subcellular localization of ASB2 in mouse embryonic cardiomyocytes by immunostaining and found it in the nucleus and cytoplasm (Fig. 1c). To examine the role of zebrash asb2a and asb2b, we generated loss-of-function mutants using the CRISPR/Cas9 technology (Fig. 1d and Supplementary Fig. 1e). We found that asb2b mutants (asb2bbns33) exhibit pericardial oedema, whereas asb2a mutants (asb2abns144) appear unaffected (Fig. 1e and Supplementary Fig. 1f). We also examined asb2a; asb2b double mutants and found no signicant differences in the severity and frequency of pericardial oedema compared with asb2b single mutants (Supplementary Fig. 1f). Thus, we focused our attention on the function of asb2b in cardiomyocytes and found that asb2b mutant hearts display reduced cardiac function, as evidenced by defects in fractional shortening and reduced blood ow velocity (Supplementary Fig. 1g). These phenotypes could not be attributed to defects in the cardiac conduction system, as we examined cardiac-specic calcium activation using the Tg(myl7:gCaMP) line14 and found no obvious changes in calcium excitation between wild type (WT) and asb2b mutant hearts (Supplementary Movies 1 and 2). Furthermore, the expression of several cardiac differentiation marker genes appeared to be unaffected in asb2b mutants (Fig. 1f, Supplementary Fig. 1h and Supplementary Table 1), suggesting that the cardiac phenotypes observed are not due to defects in the early specication of cardiac cells.

asb2b mutant cardiomyocytes exhibit structural defects. Cardiac contractility is generated by sarcomeres, the basic contractile units of muscle8. To examine sarcomeres in detail, we crossed the asb2b mutant with the Tg(myl7:LIFEACT-GFP) line15 to label myolaments. Atrial cardiomyocytes in WT hearts contain well-organized myolaments, which appear mostly aligned with one another. In contrast, myolaments in asb2b mutants are thinner, disorganized and misaligned (Fig. 2a). Accordingly, immunostaining revealed mislocalization of myosin heavy chain in asb2b mutant hearts (Supplementary Fig. 2a) and further ultra-structural analysis conrmed the thinner sarcomeres in asb2b mutant hearts (Supplementary Fig. 2b). To analyse myolament organization between neighbouring cardiomyocytes, we used the Tg(myl7:MKATE-CAAX) line to label cardiomyocyte cell membranes. We observed that myolaments in the WT atrium appear to be continuous across neighbouring cardiomyocytes, whereas those in asb2b mutants exhibit discontinuous myolaments, with noticeable breaks observed at cellcell borders (Fig. 2b). In concordance with these observations, analysis of a-Actinin3b-EGFP expression, which highlights z-bands13, reveals disorganized alignments between mutant cardiomyocytes (Fig. 2c). On the other hand, we found that forced overexpression of asb2b in cardiomyocytes also resulted in disrupted myolaments (Supplementary Fig. 2c), indicating that tight regulation of asb2b expression levels is crucial for cardiac development.

To examine cardiomyocyte morphology in greater detail, we used a cardiomyocyte-specic membrane-targeted GFP reporter line (Tg(myl7:ras-GFP))16. We observed that asb2b mutant cardiomyocytes exhibit irregular cellcell borders and membrane protrusions that overlap with neighbouring cardiomyocytes (Fig. 2d and Supplementary Fig. 3a), as well as ectopic Cdh2 (aka N-cadherin)-tdTomato punctae in the protruding membranes (Supplementary Fig. 3b). This phenotype was not accompanied by obvious changes in the

2 NATURE COMMUNICATIONS | 8:14495 | DOI: 10.1038/ncomms14495 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14495 ARTICLE

a 29 hpf

29 hpf

29 hpf

asb2b

asb2b

myl7

b

TgBAC(asb2b:GFP-GAL4FF);

Tg(UAS:mGFP)

50 hpf

At V

A

D

L

R

GFP

V

p

MKATE-CAAX

Merge

c E8.5 mouse heart DAPI cTnI DAPI ASB2

ASB2

cTnI

d e

exon3

WT

50 hpf

WT

asb2b

asb2b WT

Targeet(20bp) PAM

e

Target(20bp) PAM

5..CTGATGCTGGCGGTGCAGGCGCGGCAGCTGGAGTGTGTGAGGATCC..3

||||||||||||||||||||||||||||||||||||||||||||||

3..GACTACGACCGCCACGTCCGCGCCGTCGACCTCACACACTCCTAGG..5

asb2b/

asb2b 7

CTGATGCTGGCGGTGCAGGCGCGGCAGCTGGAGTGTGTGAGGATCCCTGATGCTGGCGGTGCAGGCGCGG-------AGTGTGTGAGGATCC

asb2b/

1 563

1 103 Asb2b 7

Asb2b WT

101

Ankyrin repeats SOCS box

f

WT asb2b/

24 hpf

myl7 myl7

Figure 1 | asb2b mutant zebrash exhibit cardiac defects. (a) In situ hybridization for asb2b and myl7 expression in 29 hpf embryos. asb2b is expressed in the myl7 heart tube (arrows) and in the somites. Lateral views, anterior to the left. Scale bars, 200 mm. (b) Three-dimensional images of 50 hpf

Tg(asb2b:GFP-GAL4FF);Tg(UAS:mGFP);Tg(myl7:MKATE-CAAX) hearts. Scale bar, 40 mm. (c) E8.5 mouse heart co-stained for ASB2 (red), cardiac troponin I (cTnI, green) and 4,6-diamidino-2-phenylindole (DAPI; blue), showing nuclear and cytoplasmic localization of ASB2; dashed lines outline nuclei. Scale bars, 10 mm. (d) Schematic representation of the asb2b locus. The gRNA target sequence, PAM and premature stop codon are highlighted in blue, red and green, respectively. Deleted nucleotides are indicated by dashes. Predicted structure of WT and mutant (Asb2b D7) proteins. The asb2b D7 allele (asb2bbns33) is predicted to encode a truncated polypeptide containing two incorrect amino acids (102103 aa). (e) Bright-eld micrographs of 50 hpf WT and asb2b mutant embryos in lateral and ventral views. Black arrows point to pericardial edema. Scale bars, 200 mm (f) in situ hybridization for myl7 expression in 24 hpf WT and asb2b mutant. Scale bar, 100 mm. At, atrium; A, anterior; D, dorsal; L, left; P, posterior; R, right; V, ventricle; V, ventral.

NATURE COMMUNICATIONS | 8:14495 | DOI: 10.1038/ncomms14495 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 3

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14495

expression levels of N-cadherin, a main junction molecule in cardiomyocytes3, or in the expression levels of other junctional components (Supplementary Fig. 3c,d and Supplementary Table 1). We also examined myocardial apico-basal polarity using a cardiomyocyte-specic transgenic

line expressing Podocalyxin, an apical protein17, fused to enhanced GFP (EGFP). EGFP-Podocalyxin localizes specically to the apical (abluminal) side of WT cardiomyocytes, but becomes mislocalized to the basal (luminal) side in asb2b mutant cardiomyocytes (Fig. 2e). We also

Atrium

a

WT asb2b/ asb2b/

WT

50 hpf 50 hpf

50 hpf

50 hpf

50 hpf

50 hpf

Myofilament thickness (m)

0 WT asb2b/

At At

V V

2.0 *

1.5

1.0

0.5

LIFEACT-GFP LIFEACT-GFP

Atrium

b

WT

LIFEACT-GFP

LIFEACT-GFP

MKATE-CAAX

MKATE-CAAX

Merge

Merge

asb2b/

l

c

WT

Actinin3b-EGFP MKATE-CAAX Merge

Actinin3b-EGFP MKATE-CAAX Merge

asb2b/

d

WT

50 hpf

50 hpf

WT asb2b/

At

asb2b/

At

V

V

ras-GFP ras-GFP

Atrium Atrium

e

WT Apical

Basal

Apical

Basal 0

100

80

EGFP-Podocalyxin MKATE-CAAX Merge

EGFP-Podocalyxin MKATE-CAAX Merge

% of cardiomyocytes

60

asb2b/

40

20

WT

**

asb2b/

4 NATURE COMMUNICATIONS | 8:14495 | DOI: 10.1038/ncomms14495 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14495 ARTICLE

transcriptional pathways by microarray analysis using isolated hearts from WT and asb2b mutants. Using Ingenuity Pathway Analysis (http://www.ingenuity.com

Web End =www.ingenuity.com), we found that TCF3 (also termed E2A), which was previously reported as a substrate of the haematopoietic-specic isoform ASB2a in mammals22, was a predicted upstream regulator of differentially expressed genes.Accordingly, downstream targets of Tcf3 were upregulated in asb2b mutant hearts (Fig. 4a and Supplementary Table 1). In mammalian muscle, the ASB2b isoform is predominantly expressed12, but there have been no reports that ASB2b targets TCF3. Thus, we performed immunoprecipitation and found that both E12 and E47, major splice isoforms of TCF3 (ref. 23), can interact with ASB2b in rat neonatal cardiomyocytes (NCMs,

Fig. 4b). We also found that ASB2b induces the proteasome-dependent degradation of both E12 and E47, which could be inhibited by the proteasome inhibitor MG132 (Fig. 4c).Furthermore, the point mutant ASB2bLA, which is defective in forming an E3 ubiquitin ligase complex24, failed to induce TCF3 degradation (Fig. 4c). It was previously reported that ASB2a cellular localization was dependent on the localization of its substrate24. We examined the cellular localization of EGFP-Asb2b and EGFP-Tcf3b, and found that both localized to the nucleus in zebrash cardiomyocytes (Fig. 4d). We also observed that EGFPAsb2b strongly accumulated in 4,6-diamidino-2-phenylindole-negative subnuclear structures (Supplementary Fig. 4c). Next, to test the hypothesis that the asb2b mutant phenotype was caused by an increase in Tcf3 levels, we generated a stable transgenic line overexpressing EGFP-tcf3b-2A-tdTomato in cardiomyocytes.Cardiomyocytes overexpressing Tcf3b exhibited severe myo-lament disassembly or disorganized myolaments, similar to the phenotype observed in asb2b mutant cardiomyocytes (Fig. 4e), as well as irregular cell shape and membrane protrusions (Fig. 4f).To quantitatively assess changes in the degree of myolament organization and cellular morphology in Tcf3b-overexpressing cardiomyocytes, we counted the number of myolament branch points, which represent the points where the direction of myolaments changes, and also measured cellular area and perimeter (Fig. 4g). We observed a signicant difference between WT and Tcf3b-overexpressing cardiomyocytes, suggesting that Tcf3b overexpression in asb2b mutant hearts causes the phenotype described above. Furthermore, we hypothesized that inhibition of Tcf3 function in vivo could rescue the cardiomyocyte phenotypes in asb2b mutants. The Id family of transcription factors is known to act as dominant-negative HLH proteins for TCF3 (ref. 23). To inhibit Tcf3 function, we generated a transgenic line overexpressing Id2b-2A-tdTomato in cardiomyocytes and crossed this line to the asb2b mutant.We found that Id2b overexpression led to a reduction in the severity of pericardial oedema (Supplementary Fig. 5a) as well as a recovery of cell morphology and myolament organization (Fig. 4g and Supplementary Fig. 5b), indicating that suppression

Figure 2 | asb2b mutant cardiomyocytes exhibit structural defects. (a) Three-dimensional images of 50 hpf Tg(myl7:LIFEACT-GFP) WT and asb2b mutant hearts reveal differences in myolament thickness and organization. Magnied views of myolaments in white boxes are shown on the right. Myolament thickness was measured in 50 hpf WT and asb2b mutant hearts (n 29 myolaments from 5 hearts). (b) Three-dimensional images of 50 hpf

Tg(myl7:LIFEACT-GFP);Tg(myl7:MKATE-CAAX) WT and asb2b mutant atria. The myolaments in WT are organized across cell borders (arrows), whereas asb2b mutant hearts show disorganized myolaments between adjacent cardiomyocytes (arrows). (c) Three-dimensional images of 50 hpf Tg(myl7:actn3b-EGFP) WT and asb2b mutant atria. Atrial cardiomyocytes in WT exhibit an organized z-band pattern across cellcell borders (arrows), whereas those in asb2b mutants exhibit a disorganized z-band pattern (arrows). (d) Three-dimensional images of 50 hpf Tg(myl7:ras-GFP) WT and asb2b mutant hearts. Magnied views of myolaments in white dotted boxes are shown on the right. Cardiomyocytes are outlined to dene shape. (e) Single-plane images of 50 hpf Tg(myl7:EGFP-Podocalyxin);Tg(myl7:MKATE-CAAX) WT and asb2b mutant atria. In WT, EGFP-Podocalyxin is localized on the abluminal (apical) side of cardiomyocytes, whereas in asb2b mutants, it appears to be localized on both the abluminal and luminal (basal) sides (arrows). Number of cardiomyocytes exhibiting EGFP-Podocalyxin localization on the basal side in WT and asb2b mutant hearts (n 5 hearts, with averages taken

from 20 cells per heart). *Po0.05 and **Po0.01 by one-way analysis of variance (ANOVA) followed by Tukeys honest signicant difference test. Error bars, s.e.m. Scale bars, 20 mm. At, atrium; V, ventricle.

observed that EGFP-Podocalyxin strongly accumulates in protruding membranes in asb2b mutant cardiomyocytes (Supplementary Fig. 3e). As the shape of asb2b mutant cardiomyocytes was reminiscent of that of migrating cells, we examined the localization of the pleckstrin homology domain of Akt1, which localizes to the leading edge of migrating cells18. We found enriched localization of pleckstrin homology-Akt in membrane protrusions in asb2b mutant cardiomyocytes (Supplementary Fig. 3f). Furthermore, we found upregulation in the expression of several genes involved in actin rearrangement and cell migration in asb2b mutant hearts (Supplementary Fig. 3g and Supplementary Table 1)19. These data suggest that asb2b mutant cardiomyocytes exhibit changes in cell morphology and cellcell junctions, as well as loss of apico-basal. These cell morphological phenotypes observed in asb2b mutant cardiomyocytes probably contribute to the functional defects described earlier.

asb2b functions cell autonomously in cardiomyocytes. Contractility and blood ow have both been reported to regulate cardiomyocyte shape during development20,21. To rule out the possibility that defects in asb2b mutant hearts arise due to secondary effects from decreased contractility and blood ow, we carried out cell transplantations to examine the behaviour of asb2b mutant cardiomyocytes in WT hearts (Fig. 3a), whose contractility and resulting blood ow velocity were not affected by the presence of mutant cells (Supplementary Fig. 4a and Supplementary Movies 3 and 4). We found that asb2b mutant cardiomyocytes in WT hearts failed to develop organized myolaments and also exhibited thin myolaments that appeared discontinuous with the myolaments of adjacent WT host cardiomyocytes (Fig. 3b and Supplementary Fig. 4b). At later stages, WT host cardiomyocytes assembled myolaments and connected exclusively with each other, while excluding asb2b mutant cardiomyocytes (Fig. 3c). Altogether, these data suggest a cell-autonomous function for Asb2b in cardiomyocyte maturation and integration. We also observed that asb2b mutant cardiomyocytes in WT hearts exhibited irregular borders and membrane protrusions (Fig. 3d). Thus, these results indicate that the phenotype observed in asb2b mutant hearts is due to a failure in cardiomyocyte integration and not secondary to cardiac dysfunction.

TCF3 is a target of ASB2 in cardiomyocytes. ASB2 is a subunit of the E3 ubiquitin ligase complex22 that functions to recognize specic substrates and target them for degradation by the proteasome. Thus, we hypothesized that the loss of asb2b function results in an increase of particular substrates, which causes the phenotype observed in the cardiomyocytes. We rst examined the global effect of the loss of asb2b on

NATURE COMMUNICATIONS | 8:14495 | DOI: 10.1038/ncomms14495 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 5

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14495

of Tcf3 results in partial rescue of the asb2b mutant phenotype. Altogether, these data suggest that an important function of Asb2b is to regulate the levels of Tcf3, and that higher than WT levels of Tcf3 during cardiomyocyte maturation causes defects in cardiomyocyte integrity.

TCF3 causes dedifferentiation of rat NCMs. Next, we wanted to determine whether TCF3 inhibits cardiomyocyte maturation and/or acutely induces disruptions in cardiomyocytes that have already established sarcomeres and cellcell junctions. To this end, we cultured rat NCMs until they exhibited synchronous

a

WT; myl7:LIFEACT-GFP or myl7:ras-GFP

orasb2b/; myl7:LIFEACT-GFP or myl7:ras-GFP

WT; myl7:LIFEACT-tdTomato or myl7:MKATE-CAAX

b

WT donor

LIFEACT-GFP

50 hpf

Myofilament thickness (m)

2.0

*

1.5

1.0

0.5

LIFEACT-GFP

LIFEACT-tdTomato - Host - WTnDsRed2 - Donor - WT Merge

LIFEACT-tdTomato- Host - WT nDsRed2 - Donor

WT

0

WT asb2b/

asb2b/ donor

WT donor

asb2b/ donor

WT donor

asb2b/ donor

Donor cells

LIFEACT-tdTomato -Host - WT

Merge

c

96 hpf

*

Myofilament thickness (m)

3.0

2.0

1.0

LIFEACT-GFP

Merge

0

WT asb2b/

Donor cells

LIFEACT-tdTomato -Host - WT

LIFEACT-GFP Merge

d

50 hpf

100

% of cardiomyocytes

80

60

40

20

ras-GFP Merge

ras-GFP Merge

MKATE-CAAX - Host - WT

MKATE-CAAX - Host - WT

0 WT asb2b/

Donor cells

6 NATURE COMMUNICATIONS | 8:14495 | DOI: 10.1038/ncomms14495 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14495 ARTICLE

contraction, indicative of coordinated sarcomeres and cellcell junctions. Overexpression of both E12 and E47 by adenoviral transfection resulted in the disassembly of sarcomeres (Fig. 5a) and this effect was rescued by co-transfection with ASB2b (Fig. 5a). We also observed a disruption of cellcell junctions (Fig. 5b), as observed in zebrash asb2b mutants. Moreover, we found that TCF3 protein levels in mouse hearts decreased during development (Supplementary Fig. 5c), further suggesting that TCF3 must be negatively regulated to allow cardiomyocyte maturation. Interestingly, we found the upregulation of markers associated with stem cells and cardiac progenitors, including gata4 (ref. 5), tbx5b5, kita (an orthologue of mammalian c-kit25) and vim (vimentin26) in asb2b mutant hearts compared with WT (Supplementary Fig. 5d). Furthermore, we examined the expression of immature markers in rat NCMs overexpressing TCF3 and found that a typical fetal marker, aa-smooth muscle actin27 (Fig. 5c), as well as Vimentin (Supplementary Fig. 5e), were upregulated. These phenotypes, observed both in vitro and in vivo, are reminiscent of cardiomyocyte dedifferentiation2528.

DiscussionIn this study, we identied an important role for ASB2 in cardiomyocytes for the maturation and organization of their sarcomeres and cellcell junctions, processes that are essential for myocardial tissue integrity. These ndings reveal key molecular mechanisms regulating cardiomyocyte maturation. In chimeric hearts, asb2b mutant cardiomyocyte failed to meld with WT myocardium, suggesting that asb2b functions cell-autonomously to ensure proper cell morphological characteristics that affect overall myocardial tissue integrity. Cell recognition and cellcell adhesion are mediated in part by cadherin molecules, and cells expressing the same type and similar levels of cadherin molecules preferentially adhere to one another29. However, asb2b mutant cardiomyocytes do not exhibit obvious changes in N-cadherin expression, suggesting that other mechanisms of cell recognition and adhesion are at play.

We identied TCF3 as a main target of ASB2 responsible for the defects in asb2b mutant cardiomyocytes. Our results show that overexpression of TCF3 recapitulates the asb2b mutant phenotypes. These phenotypes, namely disassembly of sarcomeres, disruption of cellcell junctions and re-expression of progenitor markers, have previously been reported in dedifferentiating cardiomyocytes in mouse neonatal and adult hearts, as well as during newt and zebrash heart regeneration2527,30. We also observed that the expression levels of cardiac differentiation marker genes were not affected in asb2b mutants, suggesting that the mutant cardiomyocytes may be in an intermediate, partially differentiated state. Taken together, our data suggest that in asb2b mutants, high levels of TCF3 block cardiomyocyte maturation. TCF3 may also cause an epithelial

to mesenchymal transition-like process in cardiomyocytes, as TCF3 can induce epithelial to mesenchymal transition in epithelial cells23.

During cardiomyocyte specication, cardiac progenitors migrate extensively to form the primitive heart tube6,7.After this initial migration, motility has to be suppressed to allow the maturation of cardiomyocytes and their full integration into a coherent tissue, to develop a fully formed and functional organ68. MESP1 plays a key role in the specication of the cardiac lineage31 and it was recently reported that it has to dimerize with TCF3 to regulate downstream target genes32.Our results indicate that TCF3 must subsequently be suppressed to allow cardiomyocyte maturation, and that this suppression is, in part, mediated by its proteolysis via the UPS. Regulation of TCF3 function via proteolysis may allow cardiomyocytes to quickly transition from one state to the next.

On the other hand, mechanical forces are known to be involved in cardiomyocyte development and maturation20,21,33. ASB2 is a member of the ankyrin repeat and SOCS box-containing protein family. It has been previously reported that the ankyrin repeats exhibit elastic properties and spring-like behaviour34,35 and recently the ankyrin repeats of NOMPC (No mechanoreceptor potential C), a transient receptor potential family protein, were shown to be essential for its role in mechanotransduction36. Moreover, ASB2 expression increases during mouse embryonic development and myogenic differentiation in C2C12 myoblasts12. Therefore, it is possible that the function of ASB2 is linked to mechanical forces during heart development and maturation.

In summary, our data reveal the importance of the UPS and its regulation of a specic transcription factor, for cardiomyocyte maturation and myocardial tissue integrity, and we speculate that proteolysis regulates other developmental transitions that need to occur quickly and efciently.

Methods

Zebrash. Zebrash were grown and maintained under standard conditions, as approved by institutional (Max Planck Society) and national ethical and animal welfare guidelines. Tg(myl7:MKATE-CAAX)Sd11 (ref. 13), Tg(myl7:EGFPHas.HRAS)s883 (ref. 37) (abbreviated Tg(myl7:ras-GFP)), Tg(myl7:LIFEACTGFP)s974 (ref. 15), Tg(-5.1myl7:DsRed2-NLS)f2Tg (ref. 16) (abbreviated Tg(myl7:nDsRed2)) , Tg(UAS:mGFP)m1230 (ref. 38), Tg(myl7:gCaMP)s878 (ref. 14), Tg(myl7:actn3b-EGFP)sd10Tg (ref. 13) and Tg( 0.2myl7: EGFP-Podocalyxin)bns103

(abbreviated Tg(myl7:EGFP-Podocalyxin)) lines were used. Tg(myl7:LIFEACT-tdTomato)bns141 was generated using the same strategy as Tg(myl7:LIFEACTGFP)s974 (ref. 15). To generate Tg(myl7:N-cadherin-tdtomato)bns78, Tg(myl7:EGFP-tcf3b-2A-tdTomato)bns142 and Tg(myl7:id2b-2A-tdTomato)bns143, complementary DNA fragments of N-cadherin (NCBI accession number: NM_131081), tcf3b (NCBI accession number: XM_005170492) and id2b (NCBI accession number: NM_199541 ) were isolated by RTPCR, using the following primers: N-cadherin forward 50-GAATTCACCATGTACCCCTCCGGAGGCGTG-30 and N-cadherin reverse 50-ACGCGTGTCGTCGTTACCTCCGTAC-30; tcf3b forward 50-GACGAGCTGTACAAGATGAACGAGCCACACCAG-30 and tcf3b reverse 50-GTCGACCATGTGACTGACGGAGTTG-30; id2b forward 50-AAGCTTCCA CCATGAAGGCAGTCA-30 and id2b reverse 50GTCGACACGAGACAGGGC

Figure 3 | asb2b functions cell autonomously in cardiomyocytes. (a) Schematic representation of the transplantation experiments. (b) Three-dimensional images of 50 hpf chimeric atria. Tg(myl7:LIFEACT-GFP);Tg(myl7: nDsRed2) WT or Tg(myl7:LIFEACT-GFP) asb2b mutant donor cells were transplanted into Tg(myl7:LIFEACT-tdTomato) WT host embryos. WT donor cardiomyocytes show integration of the myolaments into WT hearts (arrows), whereas those from asb2b mutants fail to meld into WT hearts (arrowheads). Myolament thickness was measured in 50 hpf WT and asb2b mutant donor cardiomyocytes (n 32 (WT) and 22 (mutant) myolament from 5 hearts). (c) Three-dimensional images of 96 hpf chimeric atria. asb2b mutant

cardiomyocytes in WT hearts fail to integrate their myolaments with those of adjacent cardiomyocytes. WT host cardiomyocytes appear to establish their myolaments underneath asb2b mutant cardiomyocytes (arrows). Myolament thickness was measured in 96 hpf WT and asb2b mutant donor-derived cardiomyocytes (n 39 (WT) and 26 (mutant) myolament from 5 hearts). (d) Three-dimensional images of 50 hpf chimeric atria resulting from

transplanting WT or asb2b mutant cells from Tg(myl7:ras-GFP) donors into Tg(myl7:MKATE-CAAX) WT hosts. asb2b mutant cardiomyocytes exhibit abnormal membrane protrusions (arrowheads). Number of WT and asb2b mutant donor-derived cardiomyocytes exhibiting membrane protrusions in 50 hpf WT hearts (n 5 hearts, with averages taken from 30 cardiomyocytes per heart.). *Po0.05 and **Po0.01 by one-way analysis of variance

(ANOVA) followed by Tukeys honest signicant difference test. Error bars, s.e.m. Scale bars, 20 mm.

NATURE COMMUNICATIONS | 8:14495 | DOI: 10.1038/ncomms14495 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 7

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14495

TATGAG-30. EGFP-tcf3b or id2b PCR fragments and 2A-tdtomato fragments were combined into one fragment via PCR and cloned into a vector containingTol2 elements and two I-SceI restriction enzyme sites under the control of the myl7 promoter element. The Tg(asb2b:GFP-GAL4FF)bns107 reporter line, encoding GFP fused GAL4FF39 and Tg(asb2b:GFP-asb2b)bns106, which carry GFP inserted in-frame with the coding region of asb2b, were generated using the BAC clone

(CH1073-416D2) using standard recombineering technology40. For whole-mount immunostaining, antibodies against myosin heavy chain 1:500 (MF20, eBioscience) and N-cadherin 1:250 (ab12221, Abcam) were used. After washing with PBS, samples were stained with uorescent secondary antibodies (Life Technologies). For quantifying cardiac function in embryonic zebrash, movies of beating hearts were taken using a Zeiss Spinning disc CSU-X1 confocal microscope with

a

WT

asb2b/

b c

IP: HA

2.4

Control +

+

Input: IB:Myc

IB:-tubulin 50

HA-ASB2

Myc-E47

+

Control

HA-ASB2 LA

MG132

+

+

+

+ + Mr (K)

HA-ASB2 WT

Myc-E47

+

+

+

+

+

+

2.0

Myc-E12

+ +

+

+

+

+

+ +

+

+

+

Myc-E12

1.6

Relative expression

IB: Myc

75

75

+ Mr (K)

75

1.2

IB: HA IB: Myc

IB: HA

IB:-tubulin

Control EGFP-tcf3b-2A-tdTomato

0.8

75

75

0.4

50

0 socs3b

id2a mef2cb

d

EGFP-Asb2b MKATE-CAAX

EGFP-Tcf3b

EGFP-Tcf3b

50 hpf MKATE-CAAX EGFP-Asb2b

50 hpf

e f

Control

50 hpf 50 hpf

0 Control EGFP-tcf3b

EGFP-tcf3b-2A-tdTomato

tdTomato

Merge

LIFEACT-GFP

EGFP-Tcf3b LIFEACT-GFP

ras-GFP

EGFP-Tcf3b ras-GFP

Mild

Severe

60

30

50

% of cardiomyocytes showing

sarcomere disassembly

% of cardiomyocytes showing

membrane protrusions

50

40

10

40

20

10

30

tdTomato

Merge

20

0

Control EGFP-tcf3b

g

**

**

Myofilament branch points

in cardiomyocytes

Cellular perimeter/cellular area

** **

** **

0.3

0.2

0.1

0.0

Overlapping area/cellular area

25 0.4

0.4

0.3

0.2

0.1

0.0

** **

**

0.5

20

15

10

5

0

asb2b/

WT Tcf3b asb2b/; id2b OE

asb2b/

WT Tcf3b asb2b/; id2b OE

asb2b/

WT Tcf3b asb2b/; id2b OE

8 NATURE COMMUNICATIONS | 8:14495 | DOI: 10.1038/ncomms14495 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14495 ARTICLE

in growth medium. The efciency of transfection was assessed by GFP expression, encoded within the adenovirus construct.

Imaging. Zebrash embryos were mounted in 1% low-melting agarose supplemented with 0.2% (w/v) tricaine to stop heartbeats. The images were acquired using a Zeiss spinning disk confocal microscope system with CSU-X1 (Yokogawa) and ORCA-ash4.0 sCMOS camera (Hamamatsu) or a confocal microscope (LSM800, Zeiss) using an LD C-Apochromat 40 (1.1 numerical aperture)

objective. Imaris (Bitplane) was used for 3D image processing. Myolament thickness was measured using the ZEN imaging software (Zeiss).

Protein analysis. Protein extracts were prepared in lysis buffer (150 mM Tris-HCl pH 7.5, 150 mM NaCl, 1% Triton X-100, 0.2% SDS, 1 mM EDTA, 5 mM NaF,0.1 mM orthovanadate, 1 mM phenylmethylsulfonyl uoride and 1 mg ml 1of aprotenin). Proteins were separated by SDSPAGE and transferred toa polyvinylidene diuoride membrane. The following primary antibodies were used: anti-HA 1:1,000 (3F10, Roche), anti-Myc 1:1,000 (9E10, Santa Cruz Biotechnology) and anti-a-tubulin 1:1,000 (T6199, Sigma). For immunoprecipitation, protein extracts were prepared in lysis buffer (150 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.1% NP40, 1 mM EDTA, 5 mM NaF, 0.1 mM orthovanadate,1 mM phenylmethylsulfonyl uoride and 1 mg ml 1 of aprotinin). Protein extracts were incubated with anti-HA 1:100 (3F10, Roche) and protein G beads (GE). Beads were washed four times with lysis buffer and precipitated proteins were analysed by immunoblotting. For densitometric analysis of western blottings, ImageJ (http://rsbweb.nih.gov/ij/

Web End =http://rsbweb.nih.gov/ij/) was used. All uncropped images related to western blotting data are available in Supplementary Fig. 6.

RNA analysis. Total RNA extraction was performed using miRNeasy Mini Kit (Qiagen) and puried RNA was used as a template for cDNA synthesis using the SuperScript second strand kit (Life Technologies), according to the manufacturers instructions. The gene symbols and gene names are listed (Supplementary Table 1).For quantitative real-time PCR, the CFX Connect Real-Time system (Bio-Rad) and DyNAmo colorFlash SYBR green qPCR kit (ThermoFisher Scientic) were used. mRNA expression levels were normalized to ef1a (Supplementary Fig. 1f) or ef1b (Fig. 4a and Supplementary Fig. 5c). Gene expression levels were relative to those of17.5 hpf WT embryos (Supplementary Fig. 1f) or 50 hpf WT hearts (Fig. 4a and Supplementary Fig. 5c). The Ct values are shown in Supplementary Table 2.The following primers were used: myl7 forward 50-GGAGAGAAGCT

CAATGGCACA-30and myl7 reverse 50-GTCATTAGCAGCCTCTTGAACTCA-30; vmhc forward 50-CTCCTGGTGCAATGGAGAAT-30 and vmhc reverse 50-CAGG TATGGCTGCAGGATTT-30; myh6 forward 50-TGGCGATGCTGACGTTTCT T-30 and myh6 reverse 50-TCACGGTGACACAAAACAGCC-30; asb2b forward 50-GACCCAGTGGGACTGAAGTG-30 and asb2b reverse 50-TGTAGAG TGCGCTTGTCCAC-30; socs3 forward 50-GGGACAGTGAGTTCCTCCAA-30 and socs3 reverse 50-ATGGGAGCATCGTACTCCTG-30; id2a forward 50-GACTCCAATTCGGCGATAAC-30 and id2a reverse 50-GTGTCCTGCTGTCCTCTG TG-30; mef2cb forward 50-CTATGGAAACCACCGCAACT-30 and mef2cb reverse 50-CCATTCCCTGTCCTGGTAGA-30; gata4 forward 50-TTACCTGTGCAA TGCCTGTG 30 and gata4 reverse 50-TGCAGACTGGCTCTCCTTCT-30; tbx5b forward 50-GCTCCACCATCTCCTGACTG-30 and tbx5b reverse 50-TGGTA GCTGAAGAGGGGGTA-30; kita forward 50-GCCCATGCAACAGAGAAAGA-30 and kita reverse 50-CGCAGGAAGTTCAACAGGTC-30; vim forward 50-TCCA TGAAGGAAACCAGACC-30 and vim reverse 50-CTCCAGCGTGGATCTTC

Figure 4 | TCF3 is a target of ASB2 in cardiomyocytes. (a) Relative mRNA expression (qPCR) of socs3b, id2a and mef2cb in isolated 50 hpf WT (blue bar) and asb2b mutant (orange bar) hearts (n 2 technical replicates; RNA samples obtained from 500 isolated hearts). (b) Co-immunoprecipitation (IP) of

HA-ASB2b and Myc-E12 or Myc-E47 immunoblotted (IB) for HA (arrow) or Myc. Rat NCMs were transfected with HA-Asb2b or control (mock), together with Myc-E12 or Myc-E47 adenovirus vectors. All adenovirus vectors used contain an independent GFP cassette to assess transfection efciency. After 24 h of transfection, cells were treated with 20 mM MG132 for 8 h and then IP was performed; lower bands in HA blot are HA antibody heavy chains. (c) Asb2b induced E12 and E47 degradation in a proteasome-dependent manner. Rat NCMs were transfected with control, HA-Asb2b or HA-Asb2b L595A mutant (Asb2bLA), together with Myc-E12 or Myc-E47 adenovirus vectors. After 48 h of transfection, NCMs were treated with 20 uM MG132 or DMSO for 12 h and these protein samples were extracted for immunoblotting. (d) Three-dimensional images of a 50 hpf TgBAC(asb2b:GFP-asb2b);Tg(myl7:MKATE-CAAX)

atrium, and a Tg(myl7:MKATE-CAAX) atrium expressing EGFP-Tcf3b. Both Asb2b and Tcf3 localize to the nucleus in cardiomyocytes.(e) Three-dimensional images of 50 hpf Tg(myl7:LIFEACT-GFP) and Tg(myl7:LIFEACT-GFP);Tg(myl7:EGFP-tcf3b-2A-tdTomato) atria. (The transgenic sh exhibit mosaic expression of EGFP-tcf3b-2A-tdTomato in the heart.) EGFP-Tcf3b-2A-tdTomato expressing cardiomyocytes exhibit disassembly of myolaments. Arrow points to a severe case of myolament disassembly, arrowhead to a mild case. Number of cardiomyocytes exhibiting sarcomere disassembly (n 5 hearts, with averages taken from 20 cardiomyocytes per heart). (f) Three-dimensional images of 50 hpf Tg(myl7:ras-GFP) or

Tg(myl7:ras-GFP);Tg(myl7:GFP-tcf3b-2A-tdTomato) atria. Tcf3b-2A-tdTomato expressing cardiomyocytes exhibit abnormal membrane protrusions (arrows). Number of cardiomyocytes exhibiting membrane protrusions (n 5 hearts, with averages taken from 20 cardiomyocytes per heart). (g) Quantication of

myolament branch points, cellular morphology and membrane protrusions in 50 hpf WT, asb2b mutant, Tg(myl7:GFP-tcf3b-2A-tdTomato) WT and Tg(myl7:id2b-2A-tdTomato) asb2b mutant atrial cardiomyocytes in Tg(myl7:LIFEACT-GFP) or Tg(myl7:ras-GFP) background. (n 15 to 34 cardiomyocytes

from 3 to 6 hearts). *Po0.05 and **Po0.01 by one-way analysis of variance (ANOVA) followed by Tukeys honest signicant difference test. Error bars, s.e.m. Scale bars, 20 mm.

a high-speed camera, which were then used to calculate fractional shortening as described previously41. For the analysis of myocardial-specic calcium activation, one-cell stage Tg(myl7:gCaMP)14 WT and asb2b mutant embryos were injected with tnnt2a MO42 to stop cardiac contraction for data acquisition and imagedat 50 hpf using a Zeiss spinning disk confocal microscope with CSU-X1 (Yokogawa) and ORCA-ash4.0 sCMOS camera (Hamamatsu). For electron microscopy (EM) analysis, zebrash embryos were xed with 3% glutaraldehyde (Merck), followed by 4% OsO4 in 0.1 mM sodium cacodylate buffer (pH 7.4). After dehydration in ethanol and propylene oxide, they were embedded in Epon as described. Ultrathin sections were stained with uranyl acetate and lead citrate, and viewed and photographically recorded under a JEM 1400 electron microscope. To count myolament branch points, Imaris (Bitplane) was used for 3D image processing and counting. To analyse cellular morphology, ImageJ (http://rsbweb.nih.gov/ij/

Web End =http:// http://rsbweb.nih.gov/ij/

Web End =rsbweb.nih.gov/ij/ ) was used to measure cellular area and perimeter.

Mouse. For immunostaining, E8.5 mouse embryos were xed with 4% paraformaldehyde. Antibodies against ASB2 1:100 (HPA001546, Sigma) and anti-cTnI 1:500 (ab56357, Abcam) were used. For western blotting, E8.5, E14.5, P1 andP7 mouse hearts were isolated. Antibodies against TCF3 1:1,000 (G-2, Santa Cruz Biotechnology) were used.

Genome editing. pT7-gRNA and pT3TS-nCas9n were used to generate sgRNA and Cas9 messenger RNA, respectively. Genomic RNAs (gRNAs) were designed targeting asb2a (Supplementary Fig. 1) and asb2b (Fig. 2) using CRISPR Design (http://crispr.mit.edu

Web End =http://crispr.mit.edu).

Cell culture analysis. Rat neonatal (P1) hearts were removed and placed in PBS; they were then minced into small pieces and placed in Hanks balanced salt solution (Gibco) supplemented with 0.0125% trypsin and incubated at 4 C overnight. The minced hearts were subsequently digested with 1.5 mg ml 1 collagenase/dispase mix (Roche) in L-15 medium (Sigma) for 20 min at 37 C with gentle agitation43. Cells were plated onto 0.1% gelatin-coated (Sigma) plates and cultured in DMEM/F12 (Gibco) supplemented with 5% horse serum, L-glutamine, Na-pyruvate, penicillin and streptomycin at 37 C and 5% CO2. For immunouorescence, cells were xed using 4% formaldehyde in PBS for 10 min. Samples were permeabilized with 0.3% Triton X-100 in PBS for 10 min and blocked with 5% fetal bovine serum in PBS for 1 h. Samples were incubatedwith the following antibodies: anti-cTnI 1:500 (ab56357, Abcam), anti-a-actinin 1:500 (EA-53, Sigma), anti-Vimentin 1:500 (c-20, Santa Cruz Biotechnology) and anti-N-cadherin 1:250 (ab12221, Abcam). After washing with PBS, samples were stained with uorescent secondary antibodies (Life Technologies) followed by 4,6-diamidino-2-phenylindole staining, to visualize nuclei. Adenovirus vectors for transfection were generated using the AdEasy system (Agilent Technologies). The cDNA fragments of HA-Asb2b (NCBI accession number: NM_001011984), myc-E12 (NCBI accession number: NM_133524) and myc-E47 (NCBI accession number: NM_001035237) were isolated by RTPCR, using the following primers: HA-Asb2b forward 50-GCGGCCGCACCATGTATCCATATGATGTTCCAGA

TTATGCTATGTCGACTGAGATCTCCAC-30 and HA-Asb2b reverse 50-TCT AGATTACTGTGTATTCTCGTATTTCAGG-30; myc-E12 and myc-E47 forward 50-GCGGCCGCaccatgGAACAAAAGTTGATTTCAGAAGAAGATCTGAT GATGAACCAGTCTCAGAGAA-30 and myc-E12 and myc-E47 reverse 50-tctaga TCACAGGTGCCCAGCTGG-30. Adenoviruses were applied to the cultured rat NCMs and incubated for 8 h. The transfected cells were then washed and cultured

NATURE COMMUNICATIONS | 8:14495 | DOI: 10.1038/ncomms14495 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 9

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14495

a

Control

E12

E47

Control

E12

E47

50

**

*

**

*

% of cardiomyocytes showing

sarcomere disassembly

40

30

-actinin

-actinin

20

10

0

E12

E12 + Asb2

E47 + Asb2

E47

-actinin

-actinin

Control

Mild Severe

-actinin

-actinin

b

Control

-actinin

-actinin

-actinin

N-cadherin

Merge

Merge

Merge

E12

E47

N-cadherin

N-cadherin

c

70

% of SMA and CTnI + cardiomyocytes

Control

E12

E47

*

*

60

50

cTnI SMA Merge

cTnI

cTnI

40

30

SMA

SMA

Merge

Merge

20

10

E12 Control

0

E47

Figure 5 | TCF3 causes dedifferentiation of rat NCMs. (a) Rat NCMs transfected with control, E12 or E47 adenovirus vectors, stained for sarcomeric-aactinin (red). Magnied views in white dotted boxes are shown on the right. Number of cardiomyocytes transfected with adenovirus vectors exhibiting sarcomere disassembly (n 2, with averages taken from 50 cardiomyocytes). (b) Rat NCMs transfected with control, E12 or E47 adenovirus vector,

co-stained for a-actinin (red), N-cadherin (magenta) and 4,6-diamidino-2-phenylindole (DAPI; blue). Cardiomyocytes expressing E12 and E47 exhibit disruptions of cellcell junction (arrowheads). (c) Rat NCMs transfected with control, E12 or E47 adenovirus vector, co-stained for cardiac troponin I (cTnI, red), a-smooth muscle actin (a-SMA, magenta) and DAPI (blue). Number of cardiomyocytes transfected with adenovirus vectors exhibiting both cTnI and a-SMA expression (n 2, with averages taken from 50 cells). *Po0.05 and **Po0.01 by one-way analysis of variance (ANOVA) followed by

Tukeys honest signicant difference test. Error bars, s.e.m. Scale bars, 20 mm.

10 NATURE COMMUNICATIONS | 8:14495 | DOI: 10.1038/ncomms14495 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14495 ARTICLE

AGT-30; ef1a forward 50-CTTCTCAGGCTGACTGTGC-30 and ef1a reverse 50-CCGCTAGCATTACCCTCC-30; ef1b forward 50-ATCTGTTTGGCTCCGA TGAG-30 and ef1b reverse 50-CAGGCTTCTTTGCCTTCTTG-30. For microarray analysis, total RNA samples were sent to Oak-Labs (Germany), where uorescent complementary RNA was generated using the Low Input QuickAmp Labeling kit (Agilent Technologies) and subsequently hybridized on a microarray. Fluorescence signals on microarrays were detected by the SureScan Microarray Scanner (Agilent Technologies). The detailed protocol is available from Oak-Labs (http://www.oak-labs.com/

Web End =http://www.oak-labs.com/). For whole-mount in situ hybridization, digoxygenin-labelled anti-sense RNA probes for myl7, asb2a and asb2b were generated using SP6 or T7 polymerase (Roche) and a DIG RNA Labeling Kit (Roche). cDNA encoding myl7 was a kind gift from Atsuo Kawahara (University of Yamanashi). To synthesize mRNA or RNA probes for in situ, cDNA encoding asb2a (NCBI accession number: XM_691298) and asb2b (NCBI accession number: NM_214725) were isolated by RTPCR with the following primers: asb2b forward 50-GAAT

TCATGACCCGGTTCTCTTATG-30 and asb2b reverse 50-CTAGATCACAGGT GGTGATGTTCC-30; asb2a forward 50-GGATCCatggcagcagccagagtgac-30; asb2a reverse 50-GAATTCctacatttcatattctcgttc-30

The PCR fragments were subcloned into the pCS2 vector.

Statistical analyses. Statistical analyses were performed by Students t-test or one-way analysis of variance with Tukeys test. Data were considered statistically signicant if the P-value was o0.05 (*).

Data availability. The data that support the ndings in this study are available within the article and its Supplementary Information les, and from the corresponding authors upon request.

References

1. Nelson, C. M. & Gleghorn, J. P. Sculpting organs: mechanical regulation of tissue development. Annu. Rev. Biomed. Eng. 14, 129154 (2012).

2. Hahn, B. S. & Labouesse, M. Tissue integrity: hemidesmosomes and resistance to stress. Curr. Biol. 11, 858861 (2001).

3. Vite, A. & Radice, G. L. N-cadherin/catenin complex as a master regulator of intercalated disc function. Cell Commun. Adhes. 21, 169179 (2014).

4. Vunjak-Novakovic, G., Lui, K. O., Tandon, N. & Chien, K. R. Bioengineering heart muscle: a paradigm for regenerative medicine. Annu. Rev. Biomed. Eng. 13, 245267 (2011).

5. Burridge, P. W., Sharma, A. & Wu, J. C. Genetic and epigenetic regulation of human cardiac reprogramming and differentiation in regenerative medicine. Annu. Rev. Genet. 49, 461484 (2015).

6. Staudt, D. & Stainier, D. Uncovering the molecular and cellular mechanisms of heart development using the zebrash. Annu. Rev. Genet. 46, 120913144909001 (2012).

7. Evans, S. M., Yelon, D., Conlon, F. L. & Kirby, M. L. Myocardial lineage development. Circ. Res. 107, 14281444 (2010).

8. Yang, X., Pabon, L. & Murry, C. E. Engineering adolescence: maturation of human pluripotent stem cell-derived cardiomyocytes. Circ. Res. 114, 511523 (2014).

9. Balse, E. et al. Dynamic of ion channel expression at the plasma membrane of cardiomyocytes. Physiol. Rev. 92, 13171358 (2012).

10. Schwartz, A. L. & Ciechanover, A. Targeting proteins for destruction by the ubiquitin system: implications for human pathobiology. Annu. Rev. Pharmacol. Toxicol. 49, 7396 (2009).

11. Komander, D. & Rape, M. The ubiquitin code. Annu. Rev. Biochem. 81,

203229 (2012).

12. Bello, N. F. et al. The E3 ubiquitin ligase specicity subunit ASB2beta is a novel regulator of muscle differentiation that targets lamin B to proteasomal degradation. Cell Death Differ. 16, 921932 (2009).

13. Lin, Y. F., Swinburne, I. & Yelon, D. Multiple inuences of blood ow on cardiomyocyte hypertrophy in the embryonic zebrash heart. Dev. Biol. 362, 242253 (2012).

14. Chi, N. C. et al. Genetic and physiologic dissection of the vertebrate cardiac conduction system. PLoS Biol. 6, 10061019 (2008).

15. Reischauer, S., Arnaout, R., Ramadass, R. & Stainier, D. Y. R. Actin binding GFP allows 4D in vivo imaging of myolament dynamics in the zebrash heart and the identication of ErBB2 signaling as a remodeling factor of myobril architecture. Circ. Res. 115, 845856 (2014).

16. Mably, J. D., Mohideen, M. A. P. K., Burns, C. G., Chen, J. N. & Fishman, M. C. Heart of glass regulates the concentric growth of the heart in zebrash. Curr. Biol. 13, 21382147 (2003).

17. Ojakian, G. K. & Schwimmer, R. The polarized distribution of anapical cell surface glycoprotein is maintained by interactions with the cytoskeleton of Madin-Darby canine kidney cells. J. Cell Biol. 107, 23772387 (1988).

18. Higuchi, M., Onishi, K., Masuyama, N. & Gotoh, Y. The phosphatidylinositol-3 kinase (PI3K)-Akt pathway suppresses neurite branch formation in NGF-treated PC12 cells. Genes Cells 8, 657669 (2003).

19. Long, A. & Freeley, M. Protein kinase C: a regulator of cytoskeleton remodelling and T-cell migration. Biochem. Soc. Trans. 42, 14901497

2014:

20. Auman, H. J. et al. Functional modulation of cardiac form through regionally conned cell shape changes. PLoS Biol. 5, 06040615 (2007).

21. Chi, N. C. et al. Cardiac conduction is required to preserve cardiac chamber morphology. Proc. Natl Acad. Sci. USA 107, 1466214667

2010:

22. Nie, L. et al. Notch-induced Asb2 expression promotes protein ubiquitination by forming non-canonical E3 ligase complexes. Cell Res. 21, 754769

2011:

23. Slattery, C., Ryan, M. P. & McMorrow, T. E2A proteins: regulators of cell phenotype in normal physiology and disease. Int. J. Biochem. Cell Biol. 40, 14311436 (2008).

24. Heuz, M. L. et al. ASB2 targets lamins A and B to proteasomal degradation. Blood 112, 51305140 (2008).

25. DUva, G. et al. ERBB2 triggers mammalian heart regeneration by promoting cardiomyocyte dedifferentiation and proliferation. Nat. Cell Biol. 17, 627638 (2015).

26. Laube, F., Heister, M., Scholz, C., Borchardt, T. & Braun, T. Re-programming of newt cardiomyocytes is induced by tissue regeneration. J. Cell Sci. 119, 47194729 (2006).

27. Kubin, T. et al. Oncostatin M is a major mediator of cardiomyocyte dedifferentiation and remodeling. Cell Stem Cell 9, 420432 (2011).

28. Wu, C. C. et al. Spatially resolved genome-wide transcriptional proling identies BMP signaling as essential regulator of zebrash cardiomyocyte regeneration. Dev. Cell 36, 3649 (2016).

29. Redies, C. Cadherins in the central nervous system. Prog. Neurobiol. 61, 611648 (2000).

30. Kikuchi, K. Dedifferentiation, transdifferentiation, and proliferation: mechanisms underlying cardiac muscle regeneration in zebrash. Curr. Pathobiol. Rep. 3, 8188 (2015).

31. Bondue, A. & Blanpain, C. Mesp1: a key regulator of cardiovascular lineage commitment. Circ. Res. 107, 14141427 (2010).

32. Chan, S. S. K. et al. Mesp1 patterns mesoderm into cardiac, hematopoietic, or skeletal myogenic progenitors in a context-dependent manner. Cell Stem Cell 12, 587601 (2013).

33. Happe, C. L. & Engler, A. J. Mechanical forces reshape differentiation cues that guide cardiomyogenesis. Circ. Res. 118, 296310 (2016).

34. Lee, G. et al. Nanospring behaviour of ankyrin repeats. Nature 440, 246249 (2006).

35. Sotomayor, M., Corey, D. P. & Schulten, K. In search of the hair-cell gating spring: elastic properties of ankyrin and cadherin repeats. Structure 13, 669682 (2005).

36. Zhang, W. et al. Ankyrin repeats convey force to gate the NOMPC mechanotransduction channel. Cell 162, 13911403 (2015).

37. DAmico, L., Scott, I. C., Jungblut, B. & Stainier, D. Y. R. A mutation in zebrash hmgcr1b reveals a role for isoprenoids in vertebrate heart-tube formation. Curr. Biol. 17, 252259 (2007).

38. Fernandes, A. M. et al. Deep brain photoreceptors control light-seeking behavior in zebrash larvae. Curr. Biol. 22, 20422047 (2012).

39. Asakawa, K. et al. Genetic dissection of neural circuits by Tol2 transposon-mediated Gal4 gene and enhancer trapping in zebrash. Proc. Natl Acad. Sci. USA 105, 12551260 (2008).

40. Suster, M. L., Abe, G., Schouw, A. & Kawakami, K. Transposon-mediated BAC transgenesis in zebrash. Nat. Protoc. 6, 19982021 (2011).

41. Walker, J. M. in Cardiovascular Development (eds Peng, X. & Antonyak, M.) Vol. 843 (Humana Press, 2012).

42. Sehnert, A. J. et al. Cardiac troponin T is essential in sarcomere assembly and cardiac contractility. Nat. Genet. 31, 106110 (2002).

43. Ehler, E., Moore-Morris, T. & Lange, S. Isolation and culture of neonatal mouse cardiomyocytes. J. Vis. Exp. 15, e50154 (2013).

Acknowledgements

We thank Toshihiko Ogura for bringing Asb2 to our attention; Atsuo Kawahara for the pCS2-myl7 plasmid; Thomas Braun, Michael Potente and Shuichi Watanabe for providing reagents for the AdEasy system; Patra Chinmoy for discussions and advice; Radhan Ramadass for image analysis; Renate Moehren for help with EM analysis;and Sabine Fisher, Carmen Bttner, Sharon Meaney-Gardian and Nana Fukuda for assistance. This work was supported by a fellowship from the Daiichi Sankyo Foundation of Life Science to R.F. and funds from the Max Planck Society to D.Y.R.S.

Author contributions

R.F. designed and conducted experiments. R.F. and D.Y.R.S. analysed data.V.J.A. contributed an unpublished reagent. H.-M.M. performed the transplantation experiments. S.K. performed EM analyses. K.K. contributed reagents and supported the

NATURE COMMUNICATIONS | 8:14495 | DOI: 10.1038/ncomms14495 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 11

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14495

early steps of the project. R.F., F.G., A.B. and D.Y.R.S. wrote the manuscript. D.Y.R.S. supervised the research.

Additional information

Supplementary Information accompanies this paper at http://www.nature.com/naturecommunications

Web End =http://www.nature.com/ http://www.nature.com/naturecommunications

Web End =naturecommunications

Competing nancial interests: The authors declare no competing nancial interests.

Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/

Web End =http://npg.nature.com/ http://npg.nature.com/reprintsandpermissions/

Web End =reprintsandpermissions/

How to cite this article: Fukuda, R. et al. Proteolysis regulates cardiomyocyte maturation and tissue integration. Nat. Commun. 8, 14495 doi: 10.1038/ncomms14495 (2017).

Publishers note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afliations.

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Web End =http://creativecommons.org/licenses/by/4.0/

r The Author(s) 2017

12 NATURE COMMUNICATIONS | 8:14495 | DOI: 10.1038/ncomms14495 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications