Visualization of outflow tract development in the absence of Tbx1 using an FgF10 enhancer trap transgene

Authors

  • Robert G. Kelly,

    Corresponding author
    1. Department of Genetics and Development, Columbia University, New York, New York
    2. CNRS UMR 6216, Developmental Biology Institute of Marseilles - Luminy, Campus de Luminy, France
    • IBDML, Campus de Luminy Case 907, 13288 Marseille, Cedex 9, France
    Search for more papers by this author
  • Virginia E. Papaioannou

    1. Department of Genetics and Development, Columbia University, New York, New York
    Search for more papers by this author

Abstract

Tbx1, the major gene underlying del22q11.2 or DiGeorge syndrome in humans, is required for normal development and septation of the cardiac outflow tract. The fibroblast growth factor 10 gene (Fgf10) and an Fgf10 enhancer trap transgene are expressed in outflow tract myocardial progenitor cells of the anterior heart field. To visualize outflow tract development in the absence of Tbx1, we have analyzed the expression profile of the Fgf10 enhancer trap transgene during outflow tract development in Tbx1−/− embryos. Transgene expression confirms hypoplasia of the distal outflow tract in the absence of Tbx1, and altered expression in pharyngeal mesoderm reveals loss of specific bilateral subpopulations of outflow tract progenitor cells and disruption of the posterior boundary of the anterior heart field. Our results support the conclusion that Tbx1 controls deployment of Fgf10-expressing progenitor cells during heart tube extension. Furthermore, although normal Fgf10 levels are dependent on Tbx1, loss of Fgf10 alleles does not significantly modify the cardiac phenotype of Tbx1 heterozygous or homozygous mutant embryos. Developmental Dynamics 236:821–828, 2007. © 2007 Wiley-Liss, Inc.

INTRODUCTION

The mammalian heart has been recently found to develop from two cardiomyocyte progenitor cell populations. Cells of the cardiac crescent give rise to the linear heart tube, which subsequently elongates by addition of new myocardium from a population of Isl1-expressing cells in pharyngeal mesoderm at both the arterial and venous poles of the heart (Cai et al.,2003; Buckingham et al.,2005). Pharyngeal mesodermal progenitors expressing the fibroblast growth factor 10 gene (Fgf10) give rise to the right ventricle and outflow tract myocardium and have been termed the secondary or anterior heart field (AHF; Kelly et al.,2001; Mjaadtvedt et al.,2001; Waldo et al.,2001; Kelly and Buckingham,2002; Abu-Issa et al.,2004). Several mouse mutants, including Isl1 and Mef2c, have been shown to specifically perturb addition of pharyngeal mesodermal cells to the elongating heart tube (Cai et al.,2003; Verzi et al.,2005). Cardiac neural crest cells are required for normal outflow tract formation and septation and regulate secondary heart field development (Yelbuz et al.,2002; Ward et al.,2005).

Anomalies in outflow tract development underlie 30% of human congenital heart defects (Srivastava and Olson,2000). One of the major causes of outflow tract defects in humans is del22q11.2 or DiGeorge syndrome, characterized by craniofacial and cardiovascular defects and haploinsufficiency for a multigene deletion on chromosome 22 (Lindsay,2001). Extensive genetic analysis in human and mouse has demonstrated that the major gene responsible for the defects in del22q11.2 syndrome patients encodes the T-box–containing transcriptional activator Tbx1 (Jerome and Papaioannou,2001; Lindsay et al.,2001; Merscher et al.,2001; Yagi et al.,2003; Baldini,2005). Tbx1 is expressed in pharyngeal mesoderm and endoderm and plays a central role in development of the pharynx and associated structures in the mid-gestation embryo (Chapman et al.,1996; Jerome and Papaioannou,2001; Garg et al.,2001; Vitelli et al.,2002a). Tbx1+/− mice display defects in fourth aortic arch artery development, and homozygous mutant mice die at birth with a single ventricular outlet or persistent truncus arteriosus (Lindsay et al.,2001; Merscher et al.,2001; Jerome and Papaioannou,2001). Hypoplasia of the outflow tract is observed at mid-gestation in Tbx1−/− embryos, and genetic studies have demonstrated that mesodermal Tbx1 regulates fibroblast growth factor (FGF) gene expression in the pharyngeal region, leading to reduced proliferation within the AHF, the outflow tract myocardial progenitor population (Vitelli et al.,2002b; Hu et al.,2004; Xu et al.,2004; Zhang et al.,2006). Tbx1 is also required for formation of the aortopulmonary septum, which initiates outflow tract septation (Xu et al.,2004). However, the precise alterations in the AHF in the absence of Tbx1 and their impact on outflow tract development have not been previously examined.

To visualize outflow tract development in the absence of Tbx1, we carried out a genetic cross between Tbx1+/−mice and mice carrying an Fgf10-enhancer trap transgene expressing β-galactosidase in pharyngeal mesoderm and outflow tract myocardium (Kelly et al.,2001). Analysis of transgene expression confirmed AHF and outflow tract hypoplasia in mid-gestation embryos and further extends our understanding of how Tbx1 regulates heart development. Loss of specific AHF subpopulations and disruption of the posterior boundary of the AHF were observed in the absence of Tbx1. Furthermore, we present evidence that transgene and Fgf10 expression is initiated independently of Tbx1 and that there is no major genetic interaction between Fgf10 and Tbx1 during arterial pole development.

RESULTS AND DISCUSSION

The Mlc1v-nlacZ-24 (1v-24) transgene is expressed in myocardium at the arterial pole of the heart, in addition to contiguous pharyngeal mesoderm (Fig. 1). The pharyngeal mesodermal expression domain includes myocardial progenitor cells of the AHF and craniofacial muscle progenitors (Kelly et al.,2001,2004). This expression pattern is the result of a position effect due to transgene integration upstream of Fgf10 (Kelly et al.,2001). To visualize outflow tract development in the absence of Tbx1, we analyzed the expression profile of the 1v-24 transgene in Tbx1 mutant embryos (Jerome and Papaioannou,2001). At E8.5, transgene expression is reduced in pharyngeal mesoderm of Tbx1−/− embryos compared with Tbx1+/− embryos, including AHF cells adjacent to the distal region of the elongating heart tube (Fig. 1A–D). Transverse sections revealed that the 1v-24 transgene is expressed normally in ventral/medial regions of the splanchnic mesoderm of the dorsal pericardial wall of Tbx1−/− embryos, whereas expression in the lateral region of the dorsal pericardial wall and future pharyngeal arch region is reduced compared with Tbx1+/− embryos (Fig. 1E,F). No differences in transgene expression were observed between Tbx1+/+ and Tbx1+/− embryos (data not shown). 1v-24 transgene expression was maintained in differentiated cardiomyocytes of Tbx1−/− embryos at this stage and heart morphology appears normal.

Figure 1.

Mlc1v-nlacZ-24 transgene and other marker gene expression in Tbx1 mutant hearts between embryonic day (E) 8.5 and E10.5. E8.5 to E10.5 embryos after X-gal staining (A–L, S–X) or in situ hybridization (M–R): A–D, G, H, M, N, S, T, lateral views; I, J, U, V, ventral views; K, L, O–R, W, X, ventral views with the heart removed; E, F, and insets in I, J, transverse cryostat sections. A–F: At E8.5, Mlc1v-nlacZ-24 expression in pharyngeal mesoderm (arrowheads) is reduced in Tbx1−/− (B,D,F) compared with Tbx1+/− (A,C,E) embryos, whereas no differences are detected in the myocardium. Note in E and F that reduced transgene expression is evident in lateral (white arrowheads) but not medial/ventral (black arrowhead) splanchnic mesoderm. G–L: At E9.5 reduced transgene expression in pharyngeal mesoderm in addition to a shortened and narrowed distal outflow tract (arrowhead) is observed in Tbx1−/− (H,J,L) compared with Tbx1+/− (G,I,K) embryos. Note the distal outflow tract constriction in J and loss of bilateral positive cells in the dorsal pericardial wall in L (white arrowheads); an abnormal midline expression domain in the dorsal pericardial wall is observed in L (arrow). M–R: Whole-mount in situ hybridization with probes detecting Isl1 (M–P) and Tlx1 (Q,R) transcripts at E9.5. Reduced Isl1 and Tlx1 expression is observed in pharyngeal mesoderm of the dorsal pericardial wall of Tbx1−/− (N,P,R) compared with Tbx1+/− (M,O,Q) embryos (white arrowheads). S–X: At E10.5, altered Mlc1v-nlacZ-24 transgene expression is observed in pharyngeal mesoderm of Tbx1−/− (T,V,X) compared with Tbx1+/− (S,U,W) embryos, including loss of transgene expression at sites of branchiomeric myogenesis (arrowheads in S) and an abnormal domain of transgene expression contiguous with the distal outflow tract (arrows in T and X). Asterisks in W and X indicate pulmonary mesenchyme. Scale bars = 100 μm in E,F; insets in I,J.

Tbx1−/− embryos exhibit distal outflow tract hypoplasia at embryonic day (E) 9.5 (Vitelli et al.,2002a; Xu et al.,2004). In Tbx1−/−1v-24 hearts, normal development of the future right ventricle and proximal region of the outflow tract was observed, whereas the distal region of the outflow tract was reduced in length compared with Tbx1+/+ and Tbx1+/− embryos (Fig. 1G–J). In addition, the diameter of the distal outflow tract was reduced in Tbx1−/− embryos, resulting in a constriction proximal to the aortic sac (Fig. 1J). Our results are in agreement with the observations of Xu et al. (2005), who defined the temporal requirement of Tbx1 during heart development using tamoxifen-regulated conditional mutagenesis.

1v-24 transgene expression in pharyngeal mesoderm is severely reduced in Tbx1−/− embryos at E9.5 (Fig. 1G,H). Within arch mesoderm, this reduction is associated with abnormal patterning of the mesodermal core and failure of initiation of branchiomeric myogenesis (Kelly et al.,2004). Endogenous Fgf10 transcripts are absent from the mesodermal core of mutant embryos at E10.5 (Vitellli et al.,2002b; Kelly et al.,2004). Within the AHF, a reduced domain of 1v-24 transgene expression is consistent with AHF hypoplasia (Xu et al.,2004). The altered 1v-24 transgene expression profile permits visualization of the AHF defect in mutant embryos (Fig. 1L). We observed that bilateral streams of 1v-24–expressing cells in the dorsal wall of the pericardial cavity, which are present in Tbx1+/+ and Tbx1+/− embryos, are severely reduced or absent from Tbx1−/− embryos (Fig. 1K,L). Instead, a more medially situated population of transgene-expressing cells was observed in the dorsal pericardial wall, which extended abnormally in a posterior direction (arrow in Fig. 1L). Altered transgene expression in the dorsal pericardial wall could be a result of loss of a specific subpopulation of the AHF associated with caudal pharyngeal hypoplasia in Tbx1 homozygous mutant embryos or, alternatively, anomalous gene expression specific to the 1v-24 transgene.

To distinguish between these possibilities and to compare our observations with the 1v-24 transgene with other genes expressed in the AHF, we evaluated the distribution of Isl1 and Tlx1 in Tbx1−/− embryos (Cai et al.,2003; Dear et al.,1995). The expression domain of Isl1, like that of the 1v-24 transgene, was present but reduced in pharyngeal mesoderm of Tbx1−/− embryos (Fig. 1M,N). The bilateral streams of Isl1-expressing cells in the dorsal pericardial wall of wild-type and Tbx1+/− embryos were also replaced by more medially situated Isl1-positive cells (Fig. 1O,P). Tlx1 encodes a homeodomain containing transcription factor expressed in pharyngeal mesoderm of the dorsal pericardial wall (Dear et al.,1995; Kelly et al.,2004). Although maintained laterally, Tlx1 expression in the dorsal pericardial wall was no longer observed in the absence of Tbx1 (Fig. 1Q,R). Expression of the 1v-24 transgene, Isl1 and Tlx1 is, therefore, lost in bilateral populations of splanchnic mesodermal cells in the dorsal pericardial wall of Tbx1−/− embryos.

At E10.5, no 1v-24 transgene expression is observed in core pharyngeal arch mesoderm of Tbx1−/− embryos (Fig. 1S,T), associated with a failure of branchiomeric myogenesis and caudal pharyngeal hypoplasia (Jerome and Papaioannou,2001; Vitelli et al.,2002a; Kelly et al., 2004). The cardiac outflow tract is positioned ventral to the region of the second arch in Tbx1−/− embryos, rather than being connected to the caudal pharyngeal region (Fig. 1S,T). In addition, the outflow tract of Tbx1−/− embryos is narrower and straighter than in Tbx1+/− embryos (Fig. 1U,V). Abnormal confluence was observed between the transgene expression domain in the dorsal pericardial wall and right pulmonary mesenchyme, a site of endogenous Fgf10 and 1v-24 transgene expression (Fig. 1W,X; Mailleux et al.,2005). This result suggests that the posterior boundary of 1v-24–expressing cells in the dorsal pericardial wall is perturbed in Tbx1−/− embryos.

Division of the embryonic outflow tract into the aorta and pulmonary trunk is under way by E11.5 in Tbx1+/− embryos; in contrast, the narrow outflow tract of Tbx1−/− embryos remains unseptated (Jerome and Papaioannou,2001; Vitelli et al.,2002a). At this stage, in normal embryos, addition of AHF cells to the arterial pole of the heart is complete (Kelly et al.,2001). However, in Tbx1−/− embryos, a population of1v-24–expressing cells was observed in the midline of the dorsal pericardial wall, contiguous with distal outflow tract myocardium (Fig. 2B, arrow), suggesting that not all transgene-expressing cells have contributed to the mutant outflow tract. At E12.5, down-regulation of the 1v-24 transgene initiates in right ventricular myocardium (Kelly et al.,2001); a cuff of β-galactosidase–positive cells was observed around the base of the pulmonary trunk and aorta in Tbx1+/− hearts and the common trunk in Tbx1−/− hearts (Fig. 2C,D). The ventral myocardial domain adjacent to this cuff appears to be reduced in mutant hearts (Fig. 2C,D).

Figure 2.

Mlc1v-nlacZ-24 transgene expression in Tbx1 mutant hearts on embryonic days (E) 11.5 and E12.5. A–H: X-gal–stained hearts from E11.5 and E12.5; A,B, lateral views; C,D, ventral views; E,F, dorsal views; G,H, transverse cryostat sections. At E11.5, an abnormal persistent transgene expression domain is observed in the dorsal pericardial wall contiguous with the distal outflow tract in B (arrow). At E12.5, altered outflow morphology is observed in Tbx1−/− (D) compared with Tbx1+/− (C) embryos; note the cuff of transgene-positive cells around the base of the great arteries (white arrowheads) and the reduction in the adjacent myocardial domain of Tbx1−/− hearts (black arrowheads). At E12.5, transgene-expressing cells are observed in the dorsal right atrial wall of Tbx1−/− (F,H) but not Tbx1+/− (E,G) hearts; labeled cells are observed in the mediastinal component of the atrial wall between the venous valves (vv) and atrial septum (arrow in H).

In Tbx1+/+ and Tbx1+/−1v-24 embryos, β-galactosidase–positive cells are restricted to right ventricular and outflow tract myocardium and not observed in atrial myocardium (Fig. 2E,G). In Tbx1−/− embryos, ectopic 1v-24–expressing cells were observed connecting the distal outflow tract with the venous pole of the heart. The β-galactosidase–positive cells were observed in the mediastinal component of the dorsal wall of the right atrium, between the venous valve leaflets and the inter-atrial septum (Fig. 2F,H). We conclude that perturbation of the posterior boundary of 1v-24 expression in the absence of Tbx1 may result in an abnormal contribution of 1v-24–expressing cells to myocardium at the venous pole of the heart. Altered development of the dorsal mesocardium in the absence of Tbx1 may be implicated in this process. We found no evidence for abnormal atrial morphology in Tbx1−/− hearts, suggesting that the transgene-expressing cells integrate normally into atrial myocardium. The 1v-24–expressing population of AHF cells, which normally contributes exclusively to the arterial pole of the heart, is a subset of a lineage of Isl1-positive pharyngeal mesodermal progenitors that contribute to both poles of the elongating heart tube (Cai et al.,2003; Meilhac et al.,2004; Buckingham et al.,2005). Our results suggest that the segregation of Isl1-positive pharyngeal mesodermal progenitor cells to the arterial and venous pole is disrupted in the absence of Tbx1. It remains to be seen whether this patterning role for Tbx1 corresponds to a direct requirement in pharyngeal mesoderm or an indirect requirement in pharyngeal endoderm.

Fgf10 transcript abundance is severely reduced in pharyngeal mesoderm at E10.5 in the absence of Tbx1 (data not shown; Vitelli et al.,2002b; Kelly et al.,2004). Our observations suggest, however, that the initiation of transcription of the Fgf10 enhancer trap transgene 1v-24 is Tbx1 independent. We evaluated the expression pattern of Fgf10 at E8.5 to investigate whether the initiation of endogenous Fgf10 expression was also Tbx1 independent. Fgf10 transcripts were present in the region of the AHF in Tbx1−/− embryos, although transcript levels were severely reduced compared with Tbx1+/− embryos (Fig. 3A–D; n = 3). Together, our results suggest that the initiation of Fgf10 expression in the AHF is not entirely Tbx1 dependent, whereas subsequent expansion of the AHF and/or the maintenance of Fgf10 transcription is Tbx1 dependent, as is Fgf10 expression in pharyngeal arch core mesoderm.

Figure 3.

Lack of major genetic interaction between Tbx1 and Fgf10 during arterial pole development. A–D:Fgf10 in situ hybridization (A,B, left lateral views; C,D, ventral views). E–L: Embryonic day (E) 10.5 ink injections viewed in bilaterally dissected embryos (E–I,K, left; J,L, right). M–P: Dissected E17.5 hearts showing the great arteries (M,N,P, ventral views; O, superior view). Q–T: E9.5 hearts. At E8.5 Fgf10 transcripts are observed in pharyngeal mesoderm (arrowheads) of Tbx1−/− embryos (B,D), although they are less abundant than in Tbx1+/− embryos (A,C). Examples of embryos after ink injection into E10.5 hearts showing nonpatent fourth arch arteries in Tbx1+/− embryos carrying wild-type (E), heterozygous (F) or homozygous mutant (G) Fgf10 alleles. Normal pharyngeal arch artery development is observed in Fgf10−/− embryos (H). An example of a Tbx1+/−;Fgf10+/− embryo showing a hypoplastic left fourth arch artery (arrowhead) and absent right fourth arch artery in left (I) and right (J) views. An example of a Tbx1+/−;Fgf10+/− embryo showing nonpatent left fourth and sixth arch arteries and a hypoplastic right fourth arch artery (arrowhead) in left (K) and right (L) views. An example of a Tbx1+/−;Fgf10+/− embryo showing a left aortic arch and aberrant origin of the right subclavian artery and an abnormally angled pulmonary trunk (arrowhead) at E17.5 in ventral (N) and superior (O) views compared with the normal situation (M, arrowhead). A Tbx1+/−;Fgf10−/− embryo showing aberrant origin of the right subclavian artery (P). At E9.5 outflow tract development is normal in Tbx1+/+;Fgf10+/− (Q) and Tbx1+/−;Fgf10−/− (R) embryos; an indistinguishable hypoplastic distal outflow tract phenotype is observed in Tbx1−/−;Fgf10+/− (S) and Tbx1−/−;Fgf10−/− (T) embryos. A, aorta; P, pulmonary trunk.

Fgf8 is coexpressed with Fgf10 in the AHF and is required for normal development of the arterial pole of the heart in mice and fish (Reifers et al.,2000; Ilagan et al.,2006; Park et al.,2006). Fgf8 has been shown to be down-regulated in the pharyngeal region of Tbx1−/− embryos and to genetically interact with Tbx1 during great artery development (Vitelli et al.,2002b). Double heterozygous Tbx1+/−; Fgf8+/− mice were found to have a significantly elevated penetrance of aortic arch artery defects compared with Tbx1+/−;Fgf8+/+ mice (Vitelli et al.,2002b). To investigate whether there is a similar genetic interaction between Fgf10 and Tbx1, we generated double heterozygous Tbx1+/−;Fgf10+/− mice by crossing Tbx1+/− and Fgf10+/− mice (Sekine et al.,1999). Fgf10−/− mutant mice die at birth and display lung and limb aplasia in addition to defects in other tissues (Min et al.,1998; Sekine et al.,1999; Ohuchi et al.,2000). Despite expression of Fgf10 in the AHF, outflow tract elongation and septation is grossly normal in Fgf10−/− mutant mice, although abnormal positioning of the ventricular apex and loss of pulmonary arteries and veins are observed (Marguerie et al.,2006). Of 80 animals genotyped at weaning, Tbx1+/−;Fgf10+/− mice were obtained at the expected frequency (17/80, χ2 = 1.9; P > 0.5). Using intra-cardiac ink injection, we investigated whether heterozygosity or homozygosity for Fgf10 increased the severity or frequency of the fourth aortic arch artery phenotype of Tbx1+/− embryos at E10.5. Fourth arch arteries of 23 Tbx1+/−;Fgf10+/− and 4 Tbx1+/−;Fgf10−/− embryos were scored as normal, hypoplastic compared with wild-type littermates, or nonpatent (Fig. 3E–L; Table 1). The types of defects observed in fourth arch artery development were similar in Tbx1+/−;Fgf10+/+, Tbx1+/−;Fgf10+/−, and Tbx1+/−;Fgf10−/− embryos. A nonsignificant increase in the number of embryos with bilaterally absent fourth arch arteries was observed in the absence of one or both Fgf10 alleles (P = 0.2). Sixth arch artery development was also scored (Fig. 3K,L). Bilateral reduced or absent sixth arch arteries were noted in a proportion of Tbx1+/+;Fgf10+/− (2/10),Tbx1+/−;Fgf10+/− (4/23), and Tbx1+/−;Fgf10−/− (3/4) embryos; in addition, a unilateral sixth arch artery was observed in two wild-type embryos (Table 1). Our results show that there is no major genetic interaction between Tbx1 and Fgf10 during pharyngeal artery development; however, we do not discount the possibility that analysis of a significantly larger sample size may reveal a minor genetic interaction.

Table 1. Aortic Arch Artery Anomalies Observed by Ink Injection at E10.5a
GenotypeN4th Arch artery anomalies6th Arch artery anomalies
NormalAbnormalNormalAbnormal
UnilateralBilateralUnilateralBilateral
  • a

    E, embryonic day; hyp, artery hypoplastic; np, artery nonpatent.

   hypnphyp hyphyp npnp np hypnphyp hyphyp npnp np
              
Tbx1+/+;Fgf10+/+3300000111000
Tbx1+/+;Fgf10+/-10800011800011
Tbx1+/+;Fgf10-/-2200000200000
              
Tbx1+/-;Fgf10+/+9011232900000
Tbx1+/-;Fgf10+/-231125681711121
Tbx1+/-;Fgf10-/-4000013100012

Embryos generated by interbreeding Tbx1+/−;Fgf10+/− mice were analyzed at E14.5 and E17.5 to examine potential interactions between these genes during great artery individualization (Fig. 3M–P; Table 2). A total of 54 embryos were analyzed (28 at E14.5 and 26 at E17.5). Embryos of all 9 genotypes were recovered at the expected ratios, including Tbx1−/−;Fgf10−/− embryos (Table 2; χ2 = 12.3; 0.5 > P > 0.1). Tbx1+/+ embryos had normal great artery development with the exception of one Tbx1+/+;Fgf10+/+ embryo with a right aortic arch and one Tbx1+/+;Fgf10−/− embryo with a high aortic arch and rightward positioned ventricular apex. An elevated incidence of isolated dextrocardia has been found in Fgf10−/− hearts, although the incidence in the present cross (1/17) was considerably less than that found by Marguerie et al. (2006), suggesting that this phenotype may be less severe on a mixed than a C57BL/6J genetic background. In contrast, use of the mixed genetic background had no apparent effect on the described Tbx1 mutant phenotype. Fgf10−/− hearts displaying a compact medially located ventricular apex were observed for all three Tbx1 genotypes (data not shown); Tbx1+/+;Fgf10−/− and Tbx1+/−;Fgf10−/− hearts also displayed an abnormally angled pulmonary trunk (Fig. 3P), as reported by Marguerie et al. (2006). No aspects of the Fgf10−/− cardiac phenotype appeared to be aggravated by loss of Tbx1 alleles.

Table 2. Cardiovascular Phenotypes at E14.5 and E17.5a
GenotypeNbNormalAbnormal (%)Right aortic archcAbnormal RSAdHigh aortic archCommon arterial trunk
  • a

    E, embryonic day; RSA, right subclavian artery.

  • b

    Data pooled for 28 E14.5 and 26 E17.5 embryos.

  • c

    Right aortic arch with retro-esophageal connection to the descending aorta.

  • d

    Aberrant origin of the right subclavian artery.

        
Tbx1+/+;Fgf10+/+541 (20)1000
Tbx1+/+;Fgf10+/-990 (0)0000
Tbx1+/+;Fgf10-/-761 (14)0010
        
Tbx1+/-;Fgf10+/+642 (33)1100
Tbx1+/-;Fgf10+/-1046 (60)2310
Tbx1+/-;Fgf10-/-817 (87)3220
        
Tbx1-/-;Fgf10+/+505 (100)22-5
Tbx1-/-;Fgf10+/-202 (100)11-2
Tbx1-/-;Fgf10-/-202 (100)11-2

Anomalies detected in Tbx1+/− embryos of all three Fgf10 genotypes included a right aortic arch with a retroesophageal connection to the descending aorta (interruption of the aortic arch, type B), aberrant origin of the right subclavian artery and a high aortic arch (Fig. 3N,O; Table 2). This spectrum of defects has been previously observed in Tbx1 haploinsufficient embryos (Lindsay et al.,2001; Jerome and Papaioannou,2001; Lindsay and Baldini,2001). A nonsignificant increase in the percentage of embryos with great artery anomalies was observed in Tbx1+/−;Fgf10+/− (6/10) and Tbx1+/−;Fgf10−/− (7/8) embryos compared with Tbx1+/−;Fgf10+/+ (2/6) embryos (P = 0.1). Lindsay and Baldini have demonstrated that recovery of arterial growth accounts for the reduced penetrance of arterial pole anomalies in Tbx1 haploinsufficient hearts at fetal stages compared with the incidence of fourth arch artery anomalies at E10.5 (Lindsay and Baldini,2001). Although our data suggest that there is no major genetic interaction between Tbx1 and Fgf10 during arterial pole development, they do not rule out a mild interaction and it is conceivable that Fgf10 may intervene in this recovery process. A common arterial trunk was observed in all Tbx1−/− embryos examined (four with a right aortic arch and five with a left aortic arch); this phenotype was not enhanced with the loss of one or both Fgf10 alleles. Consistent with these results, two Tbx1−/−;Fgf10−/− embryos dissected at E9.5 displayed a similar distal outflow tract phenotype to Tbx1−/−Fgf10+/+ and Tbx1−/−Fgf10+/− embryos (Fig. 3S,T). Our results suggest that loss of Fgf10 does not aggravate the Tbx1−/− outflow tract phenotype and, conversely, that the Fgf10−/− phenotype is not enhanced in Tbx1+/− or Tbx1−/− backgrounds. This latter observation, given the potential redundancy between FGF ligands and the fact that Fgf8 plays a key role in outflow tract development, suggests that the reduction of Fgf8 expression in Tbx1−/− embryos does not unmask outflow tract elongation defects in the absence of Fgf10. This conclusion is consistent with the recent demonstration that Fgf8 and Tbx1 are required in different tissues or at different stages during outflow tract morphogenesis (Vitelli et al.,2006).

EXPERIMENTAL PROCEDURES

Mice carrying the null allele Tbx1tm1Pa (here referred to as Tbx1) have been previously described (Jerome and Papaioannou,2001), as has the Mlc1v-nlacZ-24 transgene (Kelly et al.,2001). Fgf10+/− mice were kindly provided by Drs. N. Itoh and S. Kato (Sekine et al.,1999). Mice were maintained on a mixed genetic background. Tbx1 and Fgf10 genotypes were determined by the polymerase chain reaction on isolated tail tip or yolk sac DNA as described by Jerome and Papaioannou (2001) and Sekine et al. (1999).

X-gal revelation of β-galactosidase activity, in situ hybridization, and histology were carried out as previously described (Jerome and Papaioannou,2001; Kelly et al.,2001).

Antisense ribroprobes were prepared from plasmids as described elsewhere: Isl1 (Pfaff et al.,1996); Tlx1 (Kelly et al.,2004); Fgf10 (Kelly et al.,2001).

India ink was injected into E10.5 hearts using drawn Pasteur pipettes. Fifty-eight embryos resulting from crosses between Tbx1+/−;Fgf10+/− mice and Tbx1+/+;Fgf10+/− or Tbx1+/−;Fgf10+/+ mice were scored. Injected embryos were subsequently fixed and bisected, and left and right arch arteries were scored as normal, hypoplastic (patent to ink but reduced in size compared with control arteries), or nonpatent. Statistics were carried out using Fisher's exact probability test.

Acknowledgements

We thank Elinor Pisano and Mathieu Dandonneau for technical support and Nobu Itoh and Shigeaki Kato for providing Fgf10+/− mice. R.K. is an Inserm research fellow and is supported by the Inserm Avenir programme, Fondation pour la Recherche Médicale, Association Francaise contre les Myopathies, and the Fondation de France. V.E.P. is supported by the NIH.

Ancillary