a PATTERNS & PHENOTYPES

The TGFb Type II Receptor Plays a Critical Rolein the Endothelial Cells During CardiacDevelopmentAndrew Robson,† Kathleen R. Allinson, Robert H. Anderson, Deborah J. Henderson,and Helen M. Arthur*

TGFb signalling is required for normal cardiac development. To investigate which cell types are involved,we used mice carrying a floxed Type II TGFb receptor (Tgfbr2fl) allele and Cre-lox genetics to deplete thisreceptor in different regions of the heart. The three target tissues and corresponding Cre transgenic lineswere atrioventricular myocardium (using cGata6-Cre), ventricular myocardium (using Mlc2v-Cre), andvascular endothelium (using tamoxifen-activated Cdh5(PAC)-CreERT2). Spatio-temporal Cre activity ineach case was tracked via lacZ activation from the Rosa26R locus. Atrioventricular-myocardial-specificTgfbr2 knockout (KO) embryos had short septal leaflets of the tricuspid valve, whereas ventricular myo-cardial-specific KO embryos mainly exhibited a normal cardiac phenotype. Inactivation of Tgfbr2 in endo-thelial cells from E11.5 resulted in deficient ventricular septation, accompanied by haemorrhage fromcerebral blood vessels. We conclude that TGFb signalling through the Tgfbr2 receptor, in endothelial cells,plays an important role in cardiac development, and is essential for cerebral vascular integrity. Develop-mental Dynamics 239:2435–2442, 2010. VC 2010 Wiley-Liss, Inc.

Key words: embryonic development; cardiogenesis; ventricular septal defect

Accepted 22 June 2010

INTRODUCTION

The mammalian heart develops in acomplex series of co-ordinated steps,involving differentiation of cardiacprogenitor cells, specification and sep-tation of the cardiac chambers, andformation of valves (Olson, 2006; Sri-vastava, 2006). TGFb signalling hasbeen shown to be important at anearly stage in cardiac development,with most previous investigationsfocussing on its role during formationof the cardiac cushions. These cush-ions, which develop between embry-onic day (E) E9.5 and E11.5 in the

mouse, are essential precursors of the

mature valves, and intimately

involved in septation. It is well known

that TGFb signalling promotes endo-

thelial to mesenchymal transition

(EMT), and subsequent migration of

mesenchymal cells into the cardiac

jelly (Armstrong and Bischoff, 2004).

Consistent with this role, it has been

shown that loss of the TGFb2 ligandis associated with deficient ventricu-

lar septation (Sanford et al., 1997;

Bartram et al., 2001), and that

TGFb1 mutant mice exhibit disorgan-

ised valves, albeit only in certain

genetic backgrounds (Letterio et al.,

1994). More recent advances in mouse

genetics have permitted in vivo analy-

sis of TGFb signalling during cardiac

development in a cell type–specific

manner, and have confirmed the im-

portance of the main type I and type

II TGFb receptors (Jiao et al., 2006;

Sridurongrit et al., 2008). Overall,

evidence produced to date suggests acritical role for these TGFb receptors

in endothelial cells (ECs), a role that

is required to promote formation of

the cardiac cushions (Sridurongrit

et al., 2008).

Dev

elop

men

tal D

ynam

ics

Institute of Human Genetics, Newcastle University, Newcastle, United Kingdom†Andrew Robson is presently at Institute of Cellular Medicine, Newcastle University, Newcastle, UK.Grant sponsor: Borwick Trust.*Correspondence to: Dr. Helen M. Arthur, Institute of Human Genetics, Centre for Life, Newcastle University, NE1 3BZ,UK. E-mail: helen.arthur@ncl.ac.uk

DOI 10.1002/dvdy.22376Published online 22 July 2010 in Wiley Online Library (wileyonlinelibrary.com).

DEVELOPMENTAL DYNAMICS 239:2435–2442, 2010

VC 2010 Wiley-Liss, Inc.

During TGFb signalling, a TGFbligand binds first to the TGFb type IIreceptor (TGFBR2) at the cell surface,causing it to activate, by phosphoryla-tion, an associated TGFb type I recep-tor, TGFBR1, also known as ALK5.This, in turn, phosphorylates SMAD2and SMAD3 proteins, enabling theirmigration to the nucleus (in associa-tion with SMAD4) to regulate tran-scription of TGFb-responsive genes.As TGFBR2 is one of the key recep-tors required in this pathway, wechose to investigate the role of TGFbsignalling in the developing heart byinactivating this receptor using a pre-viously described floxed Tgfbr2 mouse(Leveen et al., 2002). Our goal was touse a Cre-lox approach to depleteTgfbr2 in a tissue-specific manner.The selection of target tissues wasbased on a combination of availableCre lines, and the known spatio-tem-poral expression patterns of the threeTGFb ligands (TGFb1, 2, 3) and themain TGFb receptor proteins(Tgfbr1&Tgfbr2) during cardiac de-velopment. TGFb1 is expressed in theendocardium from E8 (Akhurst et al.,1990), while the TGFb2 ligand is

Fig. 1.

Fig. 2.

Dev

elop

men

tal D

ynam

ics

2436 ROBSON ET AL.

expressed in the myocardium of theatrioventricular (AV) canal at E10(Camenisch et al., 2002; Molin et al.,2003). TGFb3 is not expressed untilE11, and is limited to epicardium andthe mesenchymal cells in the cardiaccushions (Camenisch et al., 2002;Molin et al., 2003). The Tgfbr2 recep-tor is expressed in the primitive hearttube at E8, and in myocardium, car-diac endothelial cells, and the cardiaccushions, from approximately E10(Roelen et al., 1994; Wang et al., 1995;Mariano et al., 1998; Mummery,2001). Co-localisation of Tgfbr1 andTgfbr2 has been observed in thedeveloping endocardium and myocar-dium (Lawler et al., 1994) and aknock-in mouse that carries a lacZ re-porter in the Tgfbr1 gene has shownTgfbr1 expression in cardiac myocytesand trabecular muscles at E13.5 (Sekiet al., 2006). In addition, phosphoryl-ated Smad2 (pSmad2) is present indeveloping ventricular myocardium,consistent with, although not exclu-sive to, active TGFb signalling in thistissue (de Sousa Lopes et al., 2003).

The high levels of TGFb2 ligandexpression in the myocardial cells ofthe AV canal at E10 (Camenischet al., 2002; Molin et al., 2003) indi-cate that TGFb signalling may beoccurring in this region during thisstage of heart development. We,therefore, sought to determine therole of the Tgfbr2 receptor in this tis-sue using the cGata6-Cre mouse,which drives Cre expression from a

Gata6 enhancer sequence in the AVcanal myocardium from E8.5 (Daviset al., 2001). Using this Cre line, theBMP type I receptor, Alk3, has previ-ously been shown to be important forthe development of the cardiac valvesand conduction tissues (Gaussinet al., 2002, 2005). However, the roleof TGFb signalling in these processeshas not yet been determined. Further-more, in light of evidence that Tgfbr2is important in endothelial cells dur-ing cardiac development (Jiao et al.,2006), and to bypass the embryoniclethality at E10.5 to E11.5 seen inthe Jiao et al. (2006) study, we usedthe tamoxifen-inducible Cdh5(PAC)-CreERT2 line (Mahmoud et al., 2010).This permitted us to control the tim-ing of endothelial Cre activity, and toinactivate Tgfbr2 in ECs after E10.5.Finally, following the unexpectedfinding that loss of Tgfbr2 expressionin atrial and ventricular myocardiumgenerated only a mild phenotype incTnT-Cre/Tgfbr2fl/fl mice (Jiao et al.,2006), we sought to verify this pheno-type using an independent ventricu-lar myocardial-specific Cre line. Forthis experiment, we chose the Mlc2v-Cre mouse, in which Cre has beenknocked into the Mlc2v locus, anearly ventricular-restricted marker(Chen et al., 1998). Using this Cre-loxapproach, we aimed to analyse therole of TGFb signalling in the atrio-ventricular canal and ventricularmyocardium, and in the endothelium,during cardiac development.

RESULTS

To establish that Cre was expressedin the expected cell-specific locationsin the heart, we took advantage of theRosa26R line (R26R), which expresseslacZ in cells containing active Cre pro-tein, as well as in their descendantdaughter cells (Soriano, 1999). Moni-toring LacZ expression in cGata6-Cre/R26R mice at different stages of car-diac development, we observed a simi-lar spatio-temporal pattern of expres-sion to that previously reported forthis transgenic line. cGata6-Cre isfirst expressed as the atrioventricularcanal begins to form at E8. Lineagetracing experiments have shown thatthis tissue contributes to valvar leaf-let and conduction tissues in adulthearts (Davis et al., 2001; Gaussinet al., 2005). In cGata6-Cre/R26Rembryos, lacZ was expressed in the AVcanal myocardium, and can be clearlyseen at E10.5 (Fig. 1A,B). In addition,we observed that cGata6-Cre activatedLacZ expression occurred in the atrialvestibular musculature supporting theleaflets of the tricuspid valve in adulthearts (Fig. 1C). When Tgfbr2 wasdepleted in AV canal myocardium andits derivatives in cGata6-Cre/Tgfbr2fl/fl embryos, the septal leaflet of the tri-cuspid valve was significantly shorterin mutants than in controls at E17.5(Table 1, Fig. 1D–F). The averagelength of the septal leaflet in themutants was 167 mm, whereas the av-erage length in control mice was 238mm (P ¼ 0.007) (Fig. 1F). Analysis ofMendelian ratios (not shown) demon-strated there was no loss of viability incGata6-Cre/Tgfbr2fl/fl embryos oradults, suggesting this abnormalityhad no major detrimental effects oncardiac function.As the Mlc2v-Cre line was gener-

ated using a knock-in approach, Creexpression was expected to recapitu-late endogenous Mlc2v expression,which occurs in a ventricularly re-stricted manner from E8 (Chen et al.,1998). Cre activity was examinedusing the R26R transgene. Althoughexpression was mainly restricted to theventricles, it occurred in a patchy het-erogeneous pattern (Fig. 2A,B). Thephenotypes of the Mlc2v-Cre/Tgfbr2fl/Dembryos, examined at E14.5, werealso variable, with only two out ofthe 10 embryos analysed showing an

Fig. 1. cGata6-Cre expression and phenotype of cGata6-Cre/Tgfbr2fl/fl embryos. A,B: cGata6-Cre/R26R embryo at E10.5 shows LacZ expression (blue) in AV canal myocardium (arrows) inwhole mount view (A) and in a transverse section through the atrioventricular cushion tissue (B).C: cGata6-Cre is expressed in the atrioventricular node (outlined in red) and the overlying atrialvestibular myocardium (black arrows) at the orifice of the tricuspid septal leaflet in an adultheart. D,E: Examples of transverse heart sections of cGata6-Cre/Tgfbr2fl/fl (E) and control (D)embryos at E17.5 to show the shorter tricuspid septal valve leaflet in the mutant heart comparedwith the age-matched control. Leaflet lengths were measured, as indicated by the green arrows,from the hinge point of the leaflet to the distal tip through multiple serial sections of the heart,taking care to make the measurement in between the sites of insertion of the tendinous cordsto the free edge of the leaflet. F: Statistical analysis of the length of the tricuspid septal leafletthrough multiple serial sections of the heart of Gata6-Cre/Tgfbr2fl/fl mutant (n ¼ 8) and controlembryos (n ¼ 8) showed this leaflet was significantly shorter in mutant than in control hearts. *P< .05. cm, cushion mesenchyme; lv, left ventricle; ra, right atrium; rv, right ventricle; tsl, tricuspidseptal leaflet; v, ventricle.

Fig. 2. Mlc2v-Cre expression and phenotype of Mlc2v-Cre/Tgfbr2 fl/D embryos. A,B: Mlc2v-Cre is expressed in ventricular myocardium seen in whole mount view at E9.5 (A) and in trans-verse heart section at E16.5 (B) of Mlc2v-Cre/R26R embryos. Note that ‘‘patchy’’ lacZ expres-sion is evident in the ventricular myocardium (B). C,D: Heart of one abnormal Mlc2v-Cre/Tgbfr2fl/D embryo at E14.5 with a hypoplastic left ventricle and common arterial trunk (asterisk) over-ri-ding the interventricular septum (C). An age-matched control heart is shown for comparison (D).Abbreviations are the same as for Figure 1; a, atrium; ivs, interventricular septum; la, left atrium.

Dev

elop

men

tal D

ynam

ics

TGFBR2 AND CARDIAC DEVELOPMENT 2437

abnormal phenotype (Table 1). In oneof the affected embryos, there was aperimembranous ventricular septaldefect (VSD), but the other had moresevere defects, including a commonatrioventricular valve, a common arte-rial trunk, and a hypoplastic left ven-tricle (Fig. 2C).

Following tamoxifen activation ofthe Cdh5(PAC)-CreERT2 /R26R lineat E11.5, lacZ expression was seen inthe endothelial cells lining the valvesand coronary vessels, as well asthroughout the endocardium (Fig.3A–C). Endothelial-to-mesenchymaltransition is almost complete in thecushions by the time CreERT2 wasactivated at E11.5, and consequently

the cushion mesenchyme shows lowlevels of lacZ expression (Fig. 3A,D).We expect, therefore, that Cdh5(PAC)-CreERT2 /Tgfbr2 fl/D mice treated withtamoxifen from E11.5 will maintainTGFb signalling capacity in cushionmesenchyme, whilst suffering a majorloss of activity in endothelial cellsthroughout the cardiovascular system.The mutant embryos (Cdh5(PAC)-CreERT2 /Tgfbr2 fl/D) did not survivebeyond E15.5. Because of this, we ana-lysed the cardiac morphology 24 hr ear-lier, at E14.5. In 12 of the 13 heartsexamined, we found perimembranousventricular septal defects, which variedin size from small ‘‘suture line’’ defectsin two embryos (Fig. 4A,B) to large

defects, found in 10 out of 13 mutants(Fig. 4C,D). In embryos with thesmall suture line defects, there wasno evidence of fusion between adja-cent masses of cushion mesenchyme(Fig. 4A,B). Of the 10 embryos withlarge perimembranous ventricular sep-tal defects, three also had an overrid-ing aorta, including one with a grosslyabnormal heart with a double outletright ventricle (Table 1). In addition,all mutant embryos (13/13) had severecerebral haemorrhage, readily visibleexternally at E14.5 (Fig. 5B). Examina-tion of transverse sections through thehead revealed extensive cerebralhaemorrhage (Fig. 5D).

DISCUSSION

We have shown, using a floxed Tgfbr2mouse (Leveen et al., 2002) and threedifferent transgenic Cre lines thatexpress Cre recombinase in differentregions of the developing heart, thatTGFb signalling is important for nor-mal development of the tricuspidvalve, and for fusion of the cardiaccushions during closure of the embry-onic interventricular communication.In addition, we have demonstratedthat endothelial-specific TGFb signal-ling is required for integrity of thecerebral vasculature.The AV canal myocardial Cre line,

cGata6-Cre, expresses Cre in theatrial vestibular musculature sup-porting the septal leaflet of the tricus-pid valve, a site consistent withabnormalities of the septal leafletnoted in the AV canal myocardial

Fig. 3. Cdh5(PAC)-CreERT2 expression. A–D: Transverse sections of X-Gal-stained Cdh5(PAC)-CreERT2/R26R hearts at E15.5 following tamoxifen treatment. A: LacZ is expressed in the valveleaflet endocardium (red arrows) whilst the valve cushion mesenchyme (cm) is largely lacZ nega-tive. B, C: LacZ is strongly expressed in coronary vessels (green arrows, B) and trabecular en-docardium (arrowheads, C). D: The cushion mesenchyme (cm) in the outflow tract is negativefor lacZ expression.

TABLE 1. Summary of Embryonic Cardiac Defects in Tissue Specific Knockouts of Tgfbr2

Genotype Target Tissue Age of analysis Phenotype Frequency

cGata6-Cre/Tgfbr2fl/fl AV myocardium E17.5 Short septal leaflet oftricuspid valve

8/8 (100%)

Mlc2v-Cre/Tgfbr2fl/D Ventricular myocardium E14.5 Common AV valve, common arterialtrunk, hypoplastic left ventricle.

1/10 (10%)

Perimembranous VSDand overriding aorta

1/10 (10%)

Normal 8/10 (80%)VE-Cadherin(PAC)

CreERT2 Tgfbrfl/DEndothelial/endocardial

post E11.5E14.5 Slit-like perimembranous VSD 2/13 (15%)

Perimembranous VSD 7/13 (54%)Perimembranous VSD

and overriding aorta2/13 (15%)

Perimembranous VSD,overriding aorta and doubleoutlet right ventricle

1/13 (7%)

Normal 1/13 (7%)

Dev

elop

men

tal D

ynam

ics

2438 ROBSON ET AL.

Tgfbr2-specific KO mice (Fig. 1). Themyocardium involved in the develop-ment of the tricuspid valve derivesfrom two sources, the tricuspid gullycomplex and the developing supra-ventricular crest (Lamers et al.,1995). Our data show that thecGata6-Cre-labelled cells contributeto the atrial vestibular portion of thetricuspid gully and that loss of TGFbsignalling in the atrial vestibularmusculature supporting the tricuspidvalvar leaflet has an effect on itslength. In contrast, the myocytes thatderive from the AV canal at the mitralvalve orifice do not appear to dependon TGFb signalling for their normaldevelopment. The shortened leafletthat we observed in cGata6-Cre/Tgfbr2fl/fl mutants is almost thereverse phenotype of the abnormallylong leaflets of the tricuspid and mi-tral valves reported in cGata6-Cre/Alk3fl/fl mutants (Gaussin et al.,2002, 2005). Thus, loss of BMP signal-ling through Alk3, and loss of TGFbsignalling through Tgfbr2 in AV canalmyocardial derivatives, may haveopposing signalling effects, resultingin complementary defects in the for-mation of the valvar leaflets. Therewas no loss of viability in either theTgfbr2 or Alk3 AV canal myocardial-

Fig. 5.

Fig. 4.

Fig. 4. Cardiac phenotype of Cdh5(PAC)-CreERT2/Tgfbr2 fl/D embryos. A–D: Trans-verse sections through Cdh5(PAC)-CreERT2/Tgbfr2 fl/D hearts at E14.5 showing perimem-branous ventricular septal defects. The boxedareas in A and C are shown at high power inB and D, respectively. A small ‘‘suture line’’VSD is shown in A and B (arrow) whilst a largeVSD is illustrated in C and D (asterisk). E,F:Transverse section through an age-matchedcontrol heart shows that septation of the ven-tricles is normally complete at this stage andthat tamoxifen treatment had no detectableeffect on normal heart development. Abbrevi-ations are the same as for Figure 1.

Fig. 5. Cerebral phenotype of Cdh5(PAC)-CreERT2 /Tgfbr2 fl/D embryos. A,B: Wholemount views of E14.5 control and mutant(Cdh5(PAC)-CreERT2/Tgbfr2 fl/D) embryosshow clearly visible cerebral haemorrhage (B,arrow) compared with control (A). C,D: Trans-verse brain sections of E14.5 embryos showextensive haemorrhage in mutant embryos (D)whilst tamoxifen-treated controls were unaf-fected (C). Llv, left lateral ventricle; rlv, rightlateral ventricle.

Dev

elop

men

tal D

ynam

ics

TGFBR2 AND CARDIAC DEVELOPMENT 2439

specific KOs, suggesting no major del-eterious effect on cardiac function.

The majority (8/10) of the heartsfrom the ventricular myocardium–re-stricted Tgfbr2 KO embryos (pro-duced using the Mlc2v-Cre line) werenormal. The heart of one embryo,however, was grossly abnormal,exhibiting a common AV valve, com-mon arterial trunk, and hypoplasticleft ventricle (Fig. 2C). One furtherembryo had a perimembranous VSDwith an overriding aorta. The lowincidence of defects in the Mlc2v-Cre/Tgfbr2 fl/D embryos is similar to thatreported for the myocardial-restrictedTgfbr2 KO embryos generated usingcTnT-Cre/Tgfbr2fl/fl mice where only8% of embryos were affected, one witha VSD and one with a VSD plus dou-ble outlet right ventricle (Jiao et al.,2006). Two independent models target-ing Tgfbr2 expression in the myocar-dium, therefore, produced normal car-diac development in the majority ofcases. Although it remains possiblethat the low frequency of defects wasdue to the non-uniform expression ofMlc2v-Cre in our study, the fact thatmyocardial-specific depletion of therelated receptor Tgfbr1 also has nocardiac phenotype (Sridurongrit et al.,2008), suggests it is more likely thatTGFb signalling is not essential for de-velopment of the cardiac myocytes.

By using an inducible endothelialCre to inactivate Tgfbr2, we were ableto bypass the embryonic lethality atE10.5 to E11.5 caused by the constitu-tively expressed Tie2-Cre or Tie1-Crelines (Jiao et al., 2006; Carvalhoet al., 2007). We, therefore, activatedCre in ECs from E11.5, which alsocorresponds to the stage at which en-dothelial to mesenchymal transforma-tion is almost complete. This pre-sumption was confirmed by the lackof lacZ expression in mesenchyme ofthe arterial valves, and low levels ofexpression of lacZ in the atrioventric-ular valves in Cdh5(PAC)-CreERT2 /R26R embryos (Fig. 3A,D). In 93% ofendothelial-specific Tgfbr2 KOmutants, we observed perimembra-nous VSDs, often in combination withdelayed maturation of the AV cush-ions. We were unable, however, todetect any difference in proliferationor apoptosis of cells in the AV cush-ions or associated endothelial and en-docardial cells (not shown). Our evi-

dence, therefore, points to a failure inremodelling of the atrioventricularcushions. The cushion fusion defectsmay indicate a role for TGFb signal-ling in delamination of endothelialcells and eventual fusion of the atrio-ventricular and outflow cushions toclose the interventricular communica-tion. On the other hand, the largerVSDs are more likely to be due to fail-ures in gross remodelling of the cush-ion mesenchyme across the atrioven-tricular junction, an essential stepduring formation of the membranousportion of the ventricular septum(Wessels and Sedmera, 2003). Suchremodelling is likely to involve signal-ling in endothelial/endocardial cells.We observed defects that would beconsistent with this hypothesis,including delayed fusion of cushions,and slight misalignment of muscularventricular septum with the inferioratrioventricular cushion. Both thesedefects are known to contribute to de-fective ventricular septation (Webbet al., 1998). Alternatively, it is alsoformally possible that the high fre-quency of VSDs in these mutants maybe a secondary consequence of cere-bral haemorrhage leading to reducedblood flow that might, in turn, affectmaturation of the heart. On the otherhand, a primary defect caused byreduced TGFb signalling in the heartwould be consistent with the increasedfrequency of VSDs seen in embryosthat are non-haemorrhagic, but aredeficient in betaglycan, an auxiliaryreceptor that promotes TGFb2 signal-ling (Stenvers et al., 2003; Comptonet al., 2007). In one of these studies,Betaglycan null embryos died at E14,which was thought to be a result oflack of coronary vessels (Comptonet al., 2007). In contrast, coronary ves-sel development appeared to be unaf-fected in the endothelial-specificTgfbr2 knockout mice in our study(data not shown). Further investiga-tion of the underlying cause of theseVSDs is required to better understandthe cellular mechanisms involved invalvoseptal morphogenesis, especiallyin light of the fact that VSDs are themost common congenital heart defectsin newborn babies, and this is a highlypenetrant phenotype in our model.

Cerebral haemorrhage was seen inall inducible endothelial-specificTgfbr2 KO embryos. This was also a

reported phenotype in endothelial-specific Tgfbr1 KO embryos that dieat E13 (Sridurongrit et al., 2008),indicating that TGFb signalling iscritical for cerebral vascular integrity.(In this study, the cardiac interven-tricular septum and atrioventricularcushions were poorly developed inmutant embryos compared with con-trols, but it was not possible to for-mally investigate VSDs because thenormal ventricular septum is stillremodelling at E13.) Mice withoutone of the TGFb ligands (as seen inTGFb2, TGFb3, and certain TGFb1KO mice) can develop to birth; thiswas recently shown to be due toligand redundancy, because mice car-rying mutations in both Tgfb1 andTgfb3 genes show cerebral hemor-rhage from E11.5 and subsequent em-bryonic lethality (Mu et al., 2008; Alu-wihare et al., 2009). Interestingly,mice with mutations in alpha-V orbeta-8 integrins also develop cerebralhaemorrhage (Bader et al., 1998; Zhuet al., 2002). As integrins are requiredfor TGFb activation (Sheppard, 2005;ten Dijke and Arthur, 2007), thisstrengthens the evidence for a criticalrole of TGFb signalling in maintain-ing the integrity of the developing cer-ebral vasculature.Taken together, our findings sup-

port the importance of TGFb signal-ling in cardiovascular development.Signalling is required in AV canalmyocardial cells for normal develop-ment of the tricuspid valve, and in en-dothelial/endocardial cells for closureof the embryonic interventricularcommunication, as well as for integ-rity of the cerebral vasculature.

EXPERIMENTAL

PROCEDURES

Mice

All animals were maintained accord-ing to the requirements of the Animals(Scientific Procedures) Act 1986 of theUK Government. For timed matings,noon of the day of the copulation plugwas taken as 0.5 days embryonic de-velopment (E0.5). Pregnant femaleswere humanely killed at the appropri-ate day of gestation for embryo analy-sis. To activate the Cre recombinase inembryos carrying the Cdh5(PAC)-CreERT2 transgene, pregnant females

Dev

elop

men

tal D

ynam

ics

2440 ROBSON ET AL.

were given an intraperitoneal injec-tion of 1 mg Tamoxifen (dissolved inpeanut oil) at embryonic day 11.5 andon the following two days (E12.5 andE13.5). All the mouse lines used inthis work (R26R, Gata6-Cre, Mlc2v-Cre, Cdh5(PAC)-CreERT2 andTgfbr2fl/fl ) and the primers used forgenotyping have been previouslydescribed (Chen et al., 1998; Soriano,1999; Davis et al., 2001; Mahmoudet al., 2010; Leveen et al., 2002). Thefloxed Tgfbr2 allele was converted toa null (D) allele by PGK-Cre mediatedrecombination. As mice that wereubiquitously heterozygous for this al-lele (Tgfbr2 fl/D) had no detectablephenotype, they were used in some ofour breeding programs to reduce thenumber of Cre-mediated recombina-tion events required to remove Tgfbr2expression and increase the efficiencyof the Cre/lox approach.

Tissue Staining and Analysis

Freshly harvested embryonic tissueswere fixed in 4% PFA overnight at4�C and processed to paraffin. Formorphological analysis, transversesections that had been carefullymatched for stage and position withinthe heart were stained with haema-toxylin and eosin. For XGal staining,tissues were fixed for 30 min in 0.2%gluteraldehyde, 0.1M PhosphateBuffer, pH 7.2, 5 mM EGTA, 2 mMMgCl2, and 0.02% NP40 and thenprocessed for XGal staining, paraffinsections, and eosin counterstaining aspreviously described (Arthur et al.,2000). All sections were mounted ontoslides with histomount and photo-graphed using a digital cameraattached to a Zeiss Axioplan micro-scope. Tricuspid valve leaflet lengthin 8 serial sections from each heartwas measured using Axiovision soft-ware version 4.4. Mean tricuspid sep-tal valve leaflet lengths were com-pared between eight controls andeight mutants using an unpaired Stu-dent’s t-test and graphpad Prism sta-tistical software.

ACKNOWLEDGMENTSThis work was supported in part byBHF studentship FS/04/019 and BHFsenior fellowship FS/08/001.

REFERENCES

Akhurst RJ, Lehnert SA, Faissner A, Duf-fie E. 1990. TGF beta in murine mor-phogenetic processes: the early embryoand cardiogenesis. Development 108:645–656.

Aluwihare P, Mu Z, Zhao Z, Yu D, Wein-reb PH, Horan GS, Violette SM,Munger JS. 2009. Mice that lack activ-ity of alphavbeta6- and alphavbeta8-integrins reproduce the abnormalitiesof Tgfb1- and Tgfb3-null mice. J CellSci 122:227–232.

Armstrong EJ, Bischoff J. 2004. Heartvalve development: endothelial cell sig-naling and differentiation. Circ Res 95:459–470.

Arthur HM, Ure J, Smith AJ, Renforth G,Wilson DI, Torsney E, Charlton R,Parums DV, Jowett T, Marchuk DA,Burn J, Diamond AG. 2000. Endoglin,an ancillary TGFbeta receptor, isrequired for extraembryonic angiogene-sis and plays a key role in heart devel-opment. Dev Biol 217:42–53.

Bader BL, Rayburn H, Crowley D, HynesRO. 1998. Extensive vasculogenesis,angiogenesis, and organogenesis pre-cede lethality in mice lacking all alphav integrins. Cell 95:507–519.

Bartram U, Molin DG, Wisse LJ, Moha-mad A, Sanford LP, Doetschman T,Speer CP, Poelmann RE, Gittenberger-de Groot AC. 2001. Double-outlet rightventricle and overriding tricuspid valvereflect disturbances of looping, myocar-dialization, endocardial cushion differ-entiation, and apoptosis in TGF-beta(2)-knockout mice. Circulation 103:2745–2752.

Camenisch TD, Molin DG, Person A, Run-yan RB, Gittenberger-de Groot AC,McDonald JA, Klewer SE. 2002. Tempo-ral and distinct TGFbeta ligandrequirements during mouse and avianendocardial cushion morphogenesis.Dev Biol 248:170–181.

Carvalho RL, Itoh F, Goumans MJ, Leb-rin F, Kato M, Takahashi S, Ema M,Itoh S, van Rooijen M, Bertolino P, TenDijke P, Mummery CL. 2007. Compen-satory signalling induced in the yolksac vasculature by deletion of TGFbetareceptors in mice. J Cell Sci 120:4269–4277.

Chen J, Kubalak SW, Chien KR. 1998.Ventricular muscle-restricted targetingof the RXRalpha gene reveals a non-cell-autonomous requirement in cardiacchamber morphogenesis. Development125:1943–1949.

Compton LA, Potash DA, Brown CB, Bar-nett JV. 2007. Coronary vessel develop-ment is dependent on the type IIItransforming growth factor beta recep-tor. Circ Res 101:784–791.

Davis DL, Edwards AV, Juraszek AL,Phelps A, Wessels A, Burch JB. 2001. AGATA-6 gene heart-region-specificenhancer provides a novel means tomark and probe a discrete componentof the mouse cardiac conduction system.Mech Dev 108:105–119.

de Sousa Lopes SM, Carvalho RL, vanden Driesche S, Goumans MJ, ten DijkeP, Mummery CL. 2003. Distribution ofphosphorylated Smad2 identifies targettissues of TGFbeta ligands in mouse de-velopment. Gene Expr Patterns 3:355–360.

Gaussin V, Van de Putte T, Mishina Y,Hanks MC, Zwijsen A, Huylebroeck D,Behringer RR, Schneider MD. 2002.Endocardial cushion and myocardialdefects after cardiac myocyte-specificconditional deletion of the bone morpho-genetic protein receptor ALK3. ProcNatl Acad Sci USA 99:2878–2883.

Gaussin V, Morley GE, Cox L, Zwijsen A,Vance KM, Emile L, Tian Y, Liu J,Hong C, Myers D, Conway SJ, Depre C,Mishina Y, Behringer RR, Hanks MC,Schneider MD, Huylebroeck D, Fish-man GI, Burch JB, Vatner SF. 2005.Alk3/Bmpr1a receptor is required fordevelopment of the atrioventricularcanal into valves and annulus fibrosus.Circ Res 97:219–226.

Jiao K, Langworthy M, Batts L, BrownCB, Moses HL, Baldwin HS. 2006.Tgfbeta signaling is required for atrio-ventricular cushion mesenchyme remod-eling during in vivo cardiac development.Development 133:4585–4593.

Lamers WH, Viragh S, Wessels A, Moor-man AF, Anderson RH. 1995. Forma-tion of the tricuspid valve in the humanheart. Circulation 91:111–121.

Lawler S, Candia AF, Ebner R, Shum L,Lopez AR, Moses HL, Wright CV, Der-ynck R. 1994. The murine type II TGF-beta receptor has a coincident embryonicexpression and binding preference forTGF-beta 1. Development 120:165–175.

Letterio JJ, Geiser AG, Kulkarni AB,Roche NS, Sporn MB, Roberts AB.1994. Maternal rescue of transforminggrowth factor-beta 1 null mice. Science264:1936–1938.

Leveen P, Larsson J, Ehinger M, CilioCM, Sundler M, Sjostrand LJ, Holm-dahl R, Karlsson S. 2002. Induced dis-ruption of the transforming growthfactor beta type II receptor gene inmice causes a lethal inflammatory dis-order that is transplantable. Blood 100:560–568.

Mahmoud M, Allinson K, Zhai Z, Oaken-full R, Ghandi P, Adams R, Fruttiger M,Arthur H. 2010. Pathogenesis of arterio-venous malformations in the absence ofendoglin. Circ Res 106:1425–1433.

Mariano JM, Montuenga LM, PrenticeMA, Cuttitta F, Jakowlew SB. 1998.Concurrent and distinct transcriptionand translation of transforming growthfactor-beta type I and type II receptorsin rodent embryogenesis. Int J Dev Biol42:1125–1136.

Molin DG, Bartram U, Van Der Heiden K,Van Iperen L, Speer CP, Hierck BP, Poel-mann RE, Gittenberger-De-Groot AC.2003. Expression patterns of Tgfbeta1–3associate with myocardialisation of theoutflow tract and the development of theepicardium and the fibrous heart skele-ton. Dev Dyn 227:431–444.

Dev

elop

men

tal D

ynam

ics

TGFBR2 AND CARDIAC DEVELOPMENT 2441

Mu Z, Yang Z, Yu D, Zhao Z, Munger JS.2008. TGFbeta1 and TGFbeta3 are par-tially redundant effectors in brain vascularmorphogenesis. Mech Dev 125:508–516.

Mummery CL. 2001. Transforminggrowth factor beta and mouse develop-ment. Microsc Res Tech 52:374–386.

Olson EN. 2006. Gene regulatory net-works in the evolution and developmentof the heart. Science 313:1922–1927.

Roelen BA, Lin HY, Knezevic V, FreundE, Mummery CL. 1994. Expression ofTGF-beta s and their receptors duringimplantation and organogenesis of themouse embryo. Dev Biol 166:716–728.

Sanford LP, Ormsby I, Gittenberger-deGroot AC, Sariola H, Friedman R, Boi-vin GP, Cardell EL, Doetschman T.1997. TGFbeta2 knockout mice havemultiple developmental defects that arenon-overlapping with other TGFbetaknockout phenotypes. Development124:2659–2670.

Seki T, Hong KH, Oh SP. 2006. Nonover-lapping expression patterns of ALK1

and ALK5 reveal distinct roles of eachreceptor in vascular development. LabInvest 86:116–129.

Sheppard D. 2005. Integrin-mediated acti-vation of latent transforming growthfactor beta. Cancer Metastasis Rev 24:395–402.

Soriano P. 1999. Generalized lacZ expres-sion with the ROSA26 Cre reporterstrain. Nat Genet 21:70–71.

Sridurongrit S, Larsson J, Schwartz R,Ruiz-Lozano P, Kaartinen V. 2008. Sig-naling via the Tgf-beta type I receptorAlk5 in heart development. Dev Biol322:208–218.

Srivastava D. 2006. Making or breakingthe heart: from lineage determinationto morphogenesis. Cell 126:1037–1048.

Stenvers KL, Tursky ML, Harder KW,Kountouri N, Amatayakul-Chantler S,Grail D, Small C, Weinberg RA, Size-land AM, Zhu HJ. 2003. Heart andliver defects and reduced transforminggrowth factor beta2 sensitivity in trans-forming growth factor beta type III re-

ceptor-deficient embryos. Mol Cell Biol23:4371–4385.

ten Dijke P, Arthur H. 2007. Extracellularcontrol of TGFbeta signalling in vascu-lar development and disease. Nat RevMol Cell Biol 8:857–869.

Wang YQ, Sizeland A, Wang XF, SassoonD. 1995. Restricted expression of type-II TGF beta receptor in murine embry-onic development suggests a centralrole in tissue modeling and CNS pat-terning. Mech Dev 52:275–289.

Webb S, Brown NA, Anderson RH. 1998.Formation of the atrioventricular septalstructures in the normal mouse. CircRes 82:645–656.

Wessels A, Sedmera D. 2003. Develop-mental anatomy of the heart: a tale ofmice and man. Physiol Genomics 15:165–176.

Zhu J, Motejlek K, Wang D, Zang K,Schmidt A, Reichardt LF. 2002. beta8integrins are required for vascular mor-phogenesis in mouse embryos. Develop-ment 129:2891–2903.

Dev

elop

men

tal D

ynam

ics

2442 ROBSON ET AL.