Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Mech Dev. Author manuscript; available in PMC 2009 September 1.
Published in final edited form as:
PMCID: PMC2654333



During neural tube closure, Pax3 is required to inhibit p53-dependent apoptosis. Pax3 is also required for migration of cardiac neural crest (CNC) from the neural tube to the heart and septation of the primitive single cardiac outflow tract into the aorta and pulmonary arteries. Whether Pax3 is required for CNC migration and outflow tract septation by inhibiting p53-dependent apoptosis is not known. In this study, mouse strains carrying reporters linked to Pax3 alleles were used to map the fate of CNC cells in embryos which were either Pax3-sufficient (expressing one or two functional Pax3 alleles) or Pax3-deficient (expressing two null Pax3 alleles), and in which p53 had been inactivated or not. Migrating CNC cells were observed in both Pax3-sufficient and –deficient embryos, but CNC cells were sparse and disorganized in Pax3-deficient embryos as migration progressed. The defective migration was associated with increased cell death. Suppression of p53, either by null mutation of the p53 gene, or administration of a p53 inhibitor, pifithrin-α, prevented the defective CNC migration and apoptosis in Pax3-deficient embryos, and also restored proper development of cardiac outflow tracts. These results indicate that Pax3 is required for cardiac outflow tract septation because it blocks p53-dependent processes during CNC migration.

Keywords: Cardiac neural crest, cardiac outflow tract, Pax3, p53, apoptosis


The neural crest is a transient population of cells that arises during vertebrate neurulation. The neural crest cells first become identifiable as they make an epithelial to mesenchymal transformation and delaminate from the neuroepithelium of the dorsal neural tube. They migrate to diverse, and in some cases distant, peripheral tissues, where they differentiate into a wide variety of cell types (Anderson, 1997; Sauka-Spengler and Bronner-Fraser, 2006). The resulting cell types exhibit remarkable diversity along the rostrocaudal neuraxis. For example, neural crest cells emanating from cranial regions produce sensory neurons, melanocytes, cartilage and bone, while trunk region neural crest cells produce sensory and autonomic neurons, melanocytes, and smooth muscle cells.

A sub-population of the neural crest, the cardiac neural crest (CNC), is essential for septation of the primitive outflow tract of the heart into the aortic and pulmonary tracts. CNC also promote the formation of the smooth muscle component of the tunica media, ‘middle layer’ of the aortic arch and its major branches. Kirby et al. showed, using quail–chick chimeras and diI-labeling techniques, that the CNC cells arise between mid-otic placode and the third somite (Kirby, 1987; Kirby et al., 1983). In the mouse, emigration of CNC from the neural tube initiates at the 7 somite stage, migrating around and between somites, through pharyngeal arches 3, 4, and 6, ultimately reaching their destination of the outflow tract by the 32 somite stage on E 9.5 (Chan et al., 2004). By E 12.5, the outflow tracts are septated, and neural crest can be identified within the aorticopulmonary septum and conotruncal cushions (Jiang et al., 2000). CNC can still be observed in the condensed mesenchyme of the outflow tract conus and the divided truncus on E 14.5 (Snider et al., 2007). If CNC cells are ablated during migration, neural crest derivatives in the developing cardiovascular system fail to arise, which results in defects of the cardiac outflow tract and the aortic arch arteries (Besson et al., 1986; Bockman et al., 1987; Nishibatake et al., 1987). Similar defects, such as aortic interruption, double outlet right ventricle, and persistent truncus arteriosus can occur in several human disorders, including Waardenburg syndrome, DiGeorge syndrome, and Hirchsprung disease, which affect various neural crest lineages (Creazzo et al., 1998). However, the molecular regulation of cardiac neural crest migration is not well understood.

Pax3 is one of nine members of the Pax transcription factor family, which are characterized by the paired box DNA-binding domain. The Pax proteins are classified into four subgroups according to the presence or absence of a partial or a complete homeodomain, and the presence or absence of a conserved octapeptide sequence (Robson et al., 2006). The Pax proteins are crucial regulators of development of many different organs and tissues, including the neural tube, neural crest derivatives, the eye, the kidney, lymphocytes, and α and β cells of the pancreas (Robson et al., 2006). The mutant mouse strain, Splotch, carries a loss-of-function Pax3 allele (Chalepakis et al., 1994; Epstein et al., 1991; Epstein et al., 1993; Vogan et al., 1993). Pax3Sp/+ mice develop normally, although they are distinguishable from wild type littermates by white splotches of fur on the belly due to the absence of some neural crest-derived melanocytes (Auerbach, 1954). However, Pax3Sp/Sp embryos develop open neural tube and cardiac outflow tract defects and lack limb musculature with 100% penetrance (Auerbach, 1954; Bober et al., 1994). The cardiac outflow tract defects of Pax3Sp/Sp embryos resemble those caused by neural crest ablation (Kirby et al., 1983), and embryos die around E 14.5 (Franz, 1989). The human PAX3 gene is responsible for type 1 Waardenburg syndrome, an autosomal dominant disorder, which is characterized by maldevelopment of certain neural crest derivatives, leading to deafness, white forelock, and displacement of the inner canthi of the eye. Some children with Waardenburg syndrome also display cardiac defects (Banerjee, 1986; Mathieu et al., 1990). Unlike the Splotch mutation of murine Pax3 that is null, the point mutations associated with human Waardenburg syndrome occur in the DNA binding domains, resulting in reduced or misdirected DNA binding (Lalwani et al., 1995).

While it is clear that Pax3 is required for CNC development and formation of the aorticopulmonary septum, it is not clear how it regulates these processes. Some studies have found that the onset of neural crest migration from the neural tube is delayed and that the numbers of migrating cells are reduced in Pax3-deficient embryos (Chan et al., 2004; Moase and Trasler, 1990). However, when cultured in vitro or grafted onto chick neural folds, Pax3-deficient CNC migration appears normal (Conway et al., 2000). Other studies indicate that the initial emigration is normal, but that CNC cells along the path leading to the cardiac field are reduced in number or absent (Conway et al., 1997b; Epstein et al., 2000). The reduction in numbers of migrating CNC cells has not been found to be associated with either decreased proliferation or increased cell death of migrating CNC, but may be due to decreased expansion of neural crest progenitors prior to emigration from the neural tube (Conway et al., 2000).

Pax3 is a DNA-binding transcription factor, and there is evidence that Pax3 regulates genes involved in differentiation or migration, for example, Myf-5, MyoD, myelin basic protein, c-Met, (Epstein et al., 1996; Kioussi et al., 1995; Maroto et al., 1997; Tajbakhsh et al., 1997). We identified two novel genes that are regulated by Pax3, cdc46, which participates in licensing DNA for replication, and Dep-1, the function of which is unknown (Cai et al., 1998; Hill et al., 1998). However, it is not known whether regulation of these genes by Pax3 is direct or indirect. Msx2 has also been reported to be regulated by Pax3, and gel shift analysis of a portion of the Msx2 transcription control element with in vitro-expressed Pax3 suggest that this effect is direct (Kwang et al., 2002). Nevertheless, comparison with the putative Pax3 binding site in the Msx2 promoter with optimal Pax3 binding sites that were identified using random oligonucleotides indicates that binding of Pax3 to that site would be low affinity (Phelan and Loeken, 1998).

We have previously shown that both the increased apoptosis and neural tube closure defects in Pax3Sp/Sp embryos are due to p53-dependent apoptosis (Pani et al., 2002). Steady state levels of p53 protein, but not mRNA, are increased in Pax3Sp/Sp embryos compared to w.t. embryos (Pani et al., 2002), suggesting that Pax3 is not required for neuroepithelial migration and neural tube closure, but it is required to inhibit p53 synthesis or stability in order to prevent apoptosis of neuroepithelium in the rostral and caudal neuropores. CNC derive from neuroepithelium, and so, Pax3 might also be required to inhibit p53-dependent processes in CNC. On the other hand, because the CNC arise from a portion of the neural tube that is not prone to defects in Pax3-deficient embryos, there might be different mechanisms by which Pax3 is involved in CNC development. Here we tested whether Pax3-dependent CNC migration and outflow tract septation depend upon inhibition of p53-dependent processes. Notably, p53 is not required for normal heart or cardiac outflow tract development (Donehower et al., 1992; Jacks et al., 1994). This permits testing of the role of p53 in mediating CNC and outflow tract defects in Pax3- deficient embryos.


CNC cell migration fails in Pax3-deficient embryos due to post-delamination apoptosis

Two mouse lines, carrying either a LacZ (β-galactosidase) or green fluorescent protein (GFP) reporter of Pax3 transcription, were used. In the Pax3LacZ/+ line, LacZ is inserted within the Pax3 coding sequence, so that crossing Pax3LacZ/+ mice generates Pax3LacZ/+ embryos, which, like Pax3Sp/+ embryos, are Pax3-sufficient and develop normally, and Pax3LacZ/LacZ embryos, which, like Pax3Sp/Sp embryos, are Pax3-null and display a Pax3-deficient phenotype (Relaix et al., 2003). In the Pax3GFP/+ line, the GFP reporter with an internal ribosome entry sequence is located 3’ of the Pax3 coding sequence. Neither Pax3 mRNA nor protein expression is reduced in Pax3GFP/+ embryos compared to w.t. embryos (data not shown). Crosses of Pax3GFP/+ and FVB mice were used to generate Pax3GFP/+ embryos. The Pax3GFP/+ line was used as a control for haploinsufficient Pax3LacZ/+ embryos.

The path of CNC cell migration in 14–35 somite stage Pax3LacZ/+ embryos is shown in Fig. 1 A. β-galactosidase-positive cells could first be seen migrating beyond the somites at approximately the 14 somite stage. The population subsequently divided at the 16 somite stage to give rise to two migrating cell populations. By the 35 somite stage, β-galactosidase-positive CNC cells could no longer be detected, although, β-galactosidase activity could still be detected in the hypoglossal extension, which will give rise to tongue musculature (Fig. 1 A viii).

Figure 1
CNC cell migration and apoptosis in Pax3LacZ/+ and Pax3GFP/+ (Pax3-sufficient) and Pax3LacZ/LacZ (Pax3-deficient) embryos

To determine whether CNC migration in haploinsufficient Pax3LacZ/+ embryos was characteristic of CNC in embryos with two functional Pax3 alleles, CNC in Pax3GFP/+ embryos was examined. As shown in Fig. 1 B, localization and expression of the GFP in Pax3GFP/+ embryos were similar to reporter expression in Pax3LacZ/+ embryos, indicating that Pax3LacZ/+ CNC cells are Pax3-sufficient.

To investigate the effect of Pax3 deficiency on CNC migration, Pax3LacZ/LacZ embryos obtained at the same stages of development as the Pax3LacZ/+ embryos were examined. As shown in Fig. 1 C i, β-galactosidase-positive cells could be detected migrating beyond the somites at the same developmental stage as Pax3-sufficient Pax3LacZ/+ embryos (the 14 somite stage). This suggests that the onset or rate of migration was not delayed by Pax3 deficiency. β-galactosidase staining did not appear to be reduced compared to Pax3-sufficient embryos at the 14 somite stage, but the numbers of β-galactosidase-positive cells were significantly reduced by the 18 somite stage (Fig. 1 C ii). In 18–29 somite stage embryos, the distance that CNC had migrated was not different from Pax3LacZ/+ embryos, but there was less β-galactosidase staining in Pax3LacZ/LacZ embryos.

To test whether the decreased β-galactosidase staining in Pax3LacZ/LacZ embryos was associated with increased apoptosis, embryos obtained on E 9.5 were analyzed by whole mount in situ TUNEL assay. There was a significant increase in TUNEL-positive cells in the region between the upper somites and otic pit in Pax3LacZ/LacZ embryos (coinciding with the locations of β-galactosidase-positive cells in Pax3-sufficient embryos) compared to Pax3LacZ/+ embryos (Fig. 1 D). This suggests that the depletion of β-galactosidase-positive CNC cells in Pax3LacZ/LacZ embryos with time was due to progressive apoptosis while they migrated between the otic pit and the somites. The embryos shown in Fig. 1D were obtained at the 26–27 somite stage, however, embryos obtained at earlier stages also showed that there was an inverse relationship between TUNEL-positive and β-galactosidase-positive staining (data not shown).

Inhibition of p53 rescues defective CNC migration in Pax3-deficient embryos

To determine whether CNC cell apoptosis and defective migration in Pax3LacZ/LacZ embryos was p53-mediated, Pax3LacZ/+ mice were crossed with mice carrying a germ line p53 null mutation (Jacks et al., 1994), and then double heterozygous progeny were crossed. As with the embryos shown in Fig. 1 C, CNC cell migration was defective in embryos that were Pax3-deficient but which carried two wild type p53 alleles (Pax3LacZ/LacZ p53+/+) (Fig. 2 A). Similarly, there was increased apoptosis at the site of the CNC path between the otic pit and upper somites (Fig. 2 B). Homozygous p53 mutation completely rescued CNC migration and apoptosis, and Pax3LacZ/LacZ p53−/− embryos were indistinguishable from Pax3LacZ/+ embryos at the 24–28 somite stage (Fig. 2 C and D, compared to Fig. 1 A and D), as well as at earlier somite stages (not shown). p53 heterozygosity rescued CNC migration, however, the rescue expressivity was variable among embryos: CNC migration was completely rescued in some embryos (indistinguishable from Pax3LacZ/+ embryos), not rescued at all in some embryos (indistinguishable from Pax3LacZ/LacZ p53+/+ embryos), and partially rescued in others (intermediate between Pax3LacZ/LacZ p53−/− and Pax3LacZ/LacZ p53+/+ embryos). Quantitation of the number of embryos with normal or defective CNC migration is shown in Table 1.

Figure 2
Inactivation of p53 rescues the CNC migration in Pax3-deficient embryos
Table 1
Numbers of Pax3-sufficient and –deficient E 9.5 embryos with differential p53 genotype exhibiting normal or defective CNC cell migration

As an alternative approach to inhibit p53, the effect of pifithrin-α (PFT-α), a specific inhibitor of p53 nuclear and mitochondrial localization (Endo et al., 2006; Komarov et al., 1999; Murphy et al., 2004), was examined. PFT-α was administered at noon on E 8.5 and E 9.5 as previously described (Pani et al., 2002) using a concentration which we found to be the highest effective dose that would rescue neural tube defects without having toxic effects on the mother or the embryo. As shown in Fig. 2 E and F, CNC migration was abnormal in Pax3LacZ/LacZ embryos treated only with vehicle (PBS), and this was associated with increased apoptosis. However, PFT-α rescued CNC migration in Pax3LacZ/LacZ embryos (Fig. 2 G), and apoptosis was reduced nearly to the level of wild type in embryos (Fig. 2 H). Like p53 haploinsufficiency, the expressivity of rescue by PFT-α was variable; CNC migration was completely rescued in some Pax3LacZ/LacZ embryos, not rescued at all in others, and partially rescued in others. Pax3-sufficient (Pax3LacZ/+) embryos were unaffected by PFT-α. Quantitation of the number of embryos with normal or defective CNC migration is shown in Table 2.

Table 2
Numbers of Pax3-sufficient and –deficient E 9.5 embryos treated or not with pifithrin-α exhibiting normal or defective CNC cell migration

The half-life of active PFT-α is 4.2 hours, as it spontaneously converts to a planar tricyclic derivative, which is very hydrophobic and exhibits poor solubility, under physiological conditions (Gary and Jensen, 2005). To investigate approximately when p53-dependent transcription must be inhibited in order to rescue CNC migration, PFT-α was administered at various times before the onset of Pax3 expression and until noon on E 9.5. In addition, since the half-life of PFT-α is only 4.2 hours, administering PFT-α only at noon on days 8.5 and 9.5 would not span the entire time period that p53 must be inhibited. Therefore, testing different times of PFT-α administration might improve the rescue CNC migration in Pax3-deficient embryos. As shown in Table 3, rescue of CNC migration in embryos in which PFT-α was administered at noon on both E 8.5 and 9.5 was similar to that in embryos in which PFT-α had been administered only at noon on E 8.5, and there was no effect of PFT-α administered at noon on E 9.5, suggesting that p53 must be inhibited during E 8.5. Administration of PFT-α at 8:00 AM and noon on E 8.5 (to inhibit p53 between approximately 8:00 AM-4:00 PM), or at noon and 4:00 PM (to inhibit p53 between approximately noon to 8:00 PM), or at noon, 4:00 PM, and 8:00 PM (to inhibit p53 between approximately noon to midnight), did not improve the rescue of CNC migration compared to administration of PFT-α at noon alone. Finally, administration of PFT-α at 6:00 PM rescued CNC migration in 20% of embryos, and was not significantly different from no PFT-α treatment at all. These results suggest that inhibition of p53-dependent activity by Pax3 must occur predominantly within approximately the first four hours after the onset of Pax3 expression in order for CNC migration to proceed normally, even though the loss of cells is not apparent until about 24 hours later. That it was not possible to completely rescue CNC migration in more than 55% of the PFT-α-treated embryos probably indicates that, at the highest nontoxic dose of PFT-α, it is not possible to achieve a high enough concentration of PFT-α in all CNC cells to completely suppress apoptosis. There is probably a minimum concentration of active p53 (i.e. localized at a cellular site) resulting in apoptosis in Pax3-deficient CNC cells and, like p53 haploinsufficiency, the reduction in “active” p53 by PFT-α in some cases is below that threshold, and in other cases, does not reach that threshold.

Table 3
Effect of time of PFT-α administration on CNC cell migration in Pax3-deficient embryos

Inhibition of p53 rescues cardiac outflow flow tract septation in Pax3-deficient embryos

The genotypes of embryos from Pax3LacZ/+ × Pax3LacZ/+ crosses occurred at approximately the predicted Mendelian frequency on E 12.5, with Pax3LacZ/LacZ fetuses representing 21.4% of total fetuses (Table 4). However, by E 16.5, Pax3LacZ/LacZ fetuses represented only 5.7% of the total fetuses, and only one Pax3LacZ/LacZ fetus (2.9% of all fetuses) was recovered on E 18.5. The Pax3LacZ/LacZ fetus that was recovered on E 18.5 was smaller than littermates, and the heart was not beating, suggesting that it had died prior to E 18.5 but had not yet reabsorbed. The failure to recover Pax3LacZ/LacZ fetuses near the end of gestation is consistent with previous reports that Pax3Sp/Sp embryos die around E 14.5 (Franz, 1989).

Table 4
Numbers of embryos or fetuses of each Pax3 genotype obtained on E 12.5, 16.5 or 18.5 from Pax3LacZ/+ × Pax3LacZ/+ crosses

Outflow tract septation is defective in Pax3Sp/Sp fetuses as a result of defective CNC migration. Because p53 loss of function rescued CNC migration in Pax3LacZ/LacZ embryos, we investigated whether p53 loss of function rescued outflow tract septation in Pax3LacZ/LacZ fetuses. Hearts of E 14.5 fetuses obtained from Pax3LacZ/+ p53+/− crosses, or from Pax3LacZ/+ crosses treated with PFT-α, were sectioned horizontally and examined for the presence or absence of separate pulmonary and aortic tracts. As shown in Fig. 3 A, a separated aorta and pulmonary tract could be identified in a control Pax3LacZ/+ p53+/+ fetus. In contrast, the outflow tract had failed to separate and there was no discernable muscular wall where the aorta should be in the Pax3-deficient, p53-sufficient Pax3LacZ/LacZ p53+/+ fetus (Fig. 3B). However, a separate aorta and pulmonary artery formed in the hearts of Pax3-deficient fetuses that were also p53-deficient (Pax3LacZ/LacZ p53−/−), or whose mothers had been treated with PFT-α (Fig. 3 C and D).

Figure 3
p53 loss of function rescues cardiac outflow tract septation in Pax3-deficient embryos

To test if p53 loss of function rescued outflow tract septation and fetal survival, fetuses from Pax3LacZ/+ p53+/− crosses, or from Pax3LacZ/+ crosses treated with PFT-α, were recovered on E 18.5, the day before normal term delivery, and hearts and outflow tracts were examined externally. As expected, all Pax3LacZ/+ fetuses exhibited normal outflow tract morphology, regardless of p53 genotype (Table 5). A few Pax3LacZ/LacZ p53+/+ fetuses were recovered, but they had already died; all displayed a single, malformed outflow tract. However, p53 deficiency rescued outflow tract septation in all Pax3LacZ/LacZ p53−/− and Pax3LacZ/LacZ p53+/− fetuses that were recovered. Similarly, the Pax3LacZ/LacZ fetuses from PBS-treated pregnancies all displayed outflow tract defects, and had already died, but all of the Pax3LacZ/LacZ fetuses from PFT-α-treated pregnancies exhibited normal outflow tracts (Table 6).

Table 5
Numbers of Pax3-sufficient and –deficient E 18.5 fetuses with differential p53 genotype exhibiting normal or defective outflow tract development
Table 6
Numbers of Pax3-sufficient and –deficient E 18.5 fetuses treated or not with pifithrin-α exhibiting normal or defective outflow tract development

Anatomical evidence for outflow tract septation in Pax3LacZ/LacZ fetuses in which p53-dependent apoptosis had been blocked was obtained from hearts of E 18.5 fetuses that were injected with a blue acrylic dye into the right ventricle, and with a red acrylic dye into the left ventricle. If outflow tracts were separate, pulmonary arteries would fill with blue dye, and aortas would fill with red dye, whereas, if the outflow tracts were not completely separated, mixed purple dye would fill the vessel(s). As shown in Fig. 4 A, the left pulmonary artery of a control Pax3LacZ/+ p53+/+ fetus is filled with blue dye, and is clearly separated from the aorta, which is filled with red dye. Few Pax3LacZ/LacZ p53+/+ fetuses were obtained, and the ventricles could not be injected, as the hearts were rigid and not beating, however, upon examination, separate outflow tracts could not be detected (Fig. 4 B). In contrast, circulation through the left pulmonary artery and aorta of a Pax3LacZ/LacZ p53−/− fetus and a Pax3LacZ/LacZ fetus from a PFT-α-treated pregnancy was completely separated (Fig. 4 C and D), indicating that inactivating p53 rescued outflow tract septation in Pax3-deficient fetuses.

Figure 4
Functional separation of cardiac outflow tracts in Pax3-deficient embryos rescued by p53 loss of function


The results of these experiments suggest that the reduction in migrating CNC cells in Pax3-deficient embryos is due apoptosis which is evident by TUNEL reaction approximately 24 hours after the onset of Pax3 expression. Loss-of-function germline mutation of p53, or administration of an inhibitor of p53 nuclear and mitochondrial translocation, blocked apoptosis, rescued CNC migration, and rescued cardiac outflow tract septation in Pax3-deficient embryos. These results indicate that, in CNC cells, Pax3 is not required to regulate expression of genes which control migration and cardiac outflow tract septation, but it is required to inhibit p53-dependent processes leading to apoptosis and consequent defective outflow tract development. While a simple explanation for a p53-dependent process leading to apoptosis would be transcriptional regulation of pro-apoptotic genes, p53 controls other processes, such as cell cycle arrest, migration, or angiogenesis, as well (Dameron et al., 1994; Kamijo et al., 1998; Linke et al., 1996; Roger et al., 2006). Thus, it is possible that, in the absence of Pax3, one or more of these p53-dependent processes is activated in CNC, which directly or indirectly results in apoptosis.

The mechanism by which Pax3 inhibits p53 is not well understood. We have previously shown that Pax3 has no effect on p53 mRNA levels, but that steady state p53 protein levels are increased in Pax3Sp/Sp embryos compared to w.t. embryos (Pani et al., 2002), suggesting that Pax3 inhibits p53 protein synthesis or stability. Similar observations have been made in mouse embryonic stem cells, in which Pax3 expression has been induced either by transfection of a Pax3 expression vector or differentiation of the embryonic stem cells to neuronal precursors (Morgan and Loeken, unpublished results). Further investigation is necessary in order to further understand how Pax3 regulates p53 protein on a molecular level.

Although it was not the focus of this study, we noted that Pax3 did not appear to regulate apoptosis in somitic mesoderm and derivatives. While somites were poorly formed in Pax3LacZLacZ embryos, and the hypoglossal cord was absent (compare Fig. 1 A viii and Fig. 1 C viii), there were no TUNEL-positive cells at the locations of these tissues, and blocking p53 did not rescue these structures. Moreover, limb musculature failed to develop in E 18.5 Pax3LacZ/LacZ p53−/− fetuses, and they were not viable, presumably due to failure of the diaphragm to develop. This indicates that Pax3 is required for developmental processes in somitic derivatives that are different from the role that it plays during development of neuroepithelial and neural crest derivatives.

Migrating CNC appeared beyond the somites at the same stage of development in Pax3LacZ/+ and Pax3LacZ/LacZ embryos, suggesting that the timing of initiation of CNC migration and the migratory speed do not seem to be impaired by Pax3 deficiency. Because inactivation of p53 completely rescued CNC migration, this suggests that Pax3 is not required to initiate or to pace CNC migration. Moreover, because TUNEL-positive cells were observed beyond the somites, located where CNC are found at the same stage of development in Pax3LacZ/+ embryos, this suggests that the CNC cells did migrate on time, but that their numbers are decreased because of attrition resulting from apoptosis. It was not possible to verify that the TUNEL-positive cells detectable on E 9.5 were CNC cells, because dead cells would not express cell-specific markers, but since the pattern, location, and numbers of TUNEL-positive cells were comparable to the pattern, location, and number of β-galactosidase-positive cells in Pax3-sufficient embryos, and β-galactosidase was present when p53-dependent apoptosis was inhibited, this strongly suggests that the TUNEL-positive cells were CNC cells.

Conway, et al. did not observe increased TUNEL-positive CNC cells in homozygous Splotch2H (Pax3Sp2H/Sp2H) embryos on E 9.5 (Conway et al., 2000). This may be due to differences between the Splotch2H and Splotch mutations that may differentially affect severity. Indeed, the effect of the Splotch2H mutation on outflow tract defects is only 60–85% penetrant (Conway et al., 2000; Conway et al., 1997a), whereas the Splotch defect is 100% penetrant (Auerbach, 1954). Alternatively, the fluorescent TUNEL assay used here may be more sensitive than the assay employed by Conway, et al. They concluded that there was reduced expansion of CNC progenitors in Pax3Sp2H/Sp2H embryos because there were fewer Wnt-1-expressing cells on E 8.5 in the dorsal edge of the neural folds (Conway et al., 2000). Wnt-1 is expressed in neural crest progenitors (Parr et al., 1993) and appears to be essential for neural crest cell expansion prior to emigration from the neural tube (Dorsky et al., 1998; Ikeya et al., 1997). Thus, a deficiency of Wnt-1-expressing cells in Pax3-deficient embryos could indicate that there was reduced expansion of CNC progenitors. On the other hand, cells expressing Wnt-3a, which is also expressed in neural crest progenitors, and appears to be required for initial neural crest expansion (Dorsky et al., 1998; Ikeya et al., 1997; Parr et al., 1993) were not reduced (Conway et al., 2000). Thus, it is possible that Pax3 positively regulates Wnt-1 gene expression, and a lack of Wnt-1 expression could result from deficient transcriptional regulation by the mutant Pax3.

p53 regulates cell cycle arrest as well as apoptosis (Kamijo et al., 1998; Linke et al., 1996). It is possible that Pax3 is required to override p53-induced cell cycle arrest to maintain neural crest cells, or their progenitors, in a proliferative mode. Thus, Pax3-deficient cells may undergo cell cycle arrest, and eventually undergo apoptosis, unless p53 is inactivated. It is possible that p53 also plays a role to inhibit migration of CNC as it may during tumor metastasis (Roger et al., 2006). Although the distance that CNC had migrated in Pax3-sufficient and Pax3-deficient embryos with w.t. p53 was not different, Pax3 might also play a role to override inhibition of migration.

The disappearance of β-galactosidase-positive CNC in Pax3LacZ/LacZ embryos and appearance of TUNEL-positive cells occurred more than 24 hours after the onset of Pax3 expression, and, based on timing of responsiveness to PFT-α, approximately 20–28 hours after the time period in which p53 needed to be inhibited. Furthermore, the loss of cells did not occur all at once, as if in response to a synchronized event, but progressively fewer cells were observed all along the CNC path in successively developing embryos. DNA strand breakage, which is detected by TUNEL assay, is a relatively late event in the course of apoptotic cell death. The time between the initiation of apoptosis and the morphological detection of DNA damage depends on the cell type, and is on the order of 3–4, and up to 14, hours (Bursch et al., 1990; Rodriguez and Schaper, 2005). Thus, this is consistent with a requirement for Pax3 to prevent a p53-dependent process within the first approximately 4 hours of Pax3 expression, and if this process is not prevented, cells will eventually undergo apoptosis while migrating toward the cardiac field.

We noticed that, while homozygous germ line p53 loss of function completely normalized CNC migration in Pax3LacZ/LacZ embryos, the rescue by p53 haploinsufficiency or PFT-α was variable. In contrast, there was no intermediate effect of p53 haploinsufficiency or PFT-α on neural tube defects; either neural tube closure was completely normal, or there was exencephaly and/or spina bifida, which was indistinguishable from Pax3Sp/Sp embryos (Pani et al., 2002). This suggests that the dose-dependency for p53 to activate apoptotic pathways in CNC and neuroepithelium may be different. On the other hand, we did not study migration of neuroepithelium before neural tube closure in Pax3Sp/Sp embryos with different p53 genotypes, and so, it is possible that there is a variable effect of p53 cellular concentration during neuroepithelial migration, but that there is simply a critical mass of neuroepithelium or CNC that is necessary to give rise to a normal neural tube or outflow tracts.

In summary, the data reported here show that Pax3 is required for CNC migration and outflow tract septation because it inhibits p53-dependent processes which culminate in apoptosis. Investigating how Pax3 regulates p53 on a biochemical level will be important in order to have a more detailed understanding of how CNC development is controlled.


Mouse strains and procedures

The Pax3IRESnLacZ/+ mouse line (referred to here as, Pax3LacZ/+) has been described (Relaix et al., 2003). This line carries a LacZ reporter in place of Pax3 exon 2, which contains the 5’ portion of the paired domain. Thus, LacZ serves as a Pax3 transcriptional reporter, but no functional Pax3 protein is produced by the recombined allele. This line, originally produced on a mixed C57Bl/6J-129Sv background, was made congenic with the FVB strain. Successful derivation of a congenic line was determined by marker-assisted analysis (Markel et al., 1997; Wakeland et al., 1997).

The Pax3-GFP+/− mouse line was generated to insert a Green Fluorescent Protein (GFP) coding sequence, containing a viral internal ribosomal entry site (IRES), into the 3’ UTR of the Pax3 coding sequence. Thus, GFP serves as a Pax3 transcriptional reporter, but Pax3 and GFP are translated independently and both proteins are produced. The 5′ homology arm of the targeting vector was a fragment obtained from 129/SvJ genomic DNA which spanned from the EcoRI site 5.827 kb upstream of the Pax3 stop codon-untranslated region boundary (nucleotide 1835 of reference sequence NM 008781) to an engineered EcoRI site at the coding-untranslated sequence boundary. The 3′ homology arm spanned 4.428 kb from the engineered site at the coding-untranslated sequence boundary to an AvrII site downstream of the last exon. An IRES-GFP-loxP-Neo-loxP cassette was inserted into the engineered EcoRI site at the coding-untranslated sequence boundary. The targeting construct was electroporated into CJ7 ES cells derived from blastocysts of the 129/Sv mouse strain. Correctly targeted clones were identified by Southern blot analysis using probes hybridizing to genomic DNA outside the targeting vector. Chimeric mice were obtained by injecting the IRES-EGFP-lox-Neo-lox targeted ES cells into C57Bl/6J blastocysts. The neo cassette was subsequently excised by micronuclear injection of pTurboCre plasmid (a generous gift of Dr. Tim Ley of Washington University in St. Louis) into fertilized eggs, resulting in a single loxP site downstream of the EGFP gene. Identification of pups carrying the properly targeted allele was performed by PCR of tail DNA, as described below. Pax3GFP/Sp pups are phenotypically indistinguishable from Pax3Sp/+ pups, and Pax3GFP/GFP pups are indistinguishable from w.t. pups, indicating that the Pax3GFP allele behaves like a w.t. Pax3 allele. The Pax3-GFP+/− line was made congenic with the FVB strain and successful derivation was confirmed by marker-assisted analysis (Markel et al., 1997; Wakeland et al., 1997).

Genotyping of mouse tail samples and yolk sac DNA was performed by PCR using REDExtract-N-Amp Tissue PCR Kit (Sigma). The forward and reverse primers used to detect a 627 base pair PCR product for the wild type Pax3 allele in embryos of Pax3LacZ/+ mice were (TAGACATCAGTCCTAGGTCTCCCTCC) and (TTTAGAACGCGCCCACTCTGGACCCGC), respectively. The forward and reverse primers used to detect a 320 base pair product for LacZ in the embryos of Pax3LacZ/+ mice were (GCTGAACGGCAAGCCGTT) and (GATTACGATCGCGCTGCACC), respectively. The forward and reverse primers used to detect a 367 base pair product for wild type Pax3 allele of Pax3-GFP+/− mice were (GACAAAGTAAGCCTTGGACG) and (CCTGAGACAAGTTTGACACAAC), respectively. The forward and reverse primers used to detect a 342 bp EGFP PCR product to genotype embryos of Pax3-GFP+/− mice were (GCACCATCTTCTTCAAGGACGAC) and (TCTTTGCTCAGGGCGGACTG), respectively.

Trp53tm1TyJ/Trp+ (Jacks et al., 1994) (referred to as, p53+/−) mice on a FVB background were obtained from the Jackson Laboratory and genotyped according to the Jackson Laboratory protocol (

Staging of pregnancies was performed using noon on the day on which a copulation plug was found to be day 0.5. Pifithrin-α (Calbiochem) was administered by intraperitoneal injection of 2.2 mg/kg dissolved in PBS daily at noon (or as otherwise indicated) on E 8.5–E 13.5 (Pani et al., 2002). Embryos were recovered on E 9.5 to assess CNC cell migration and apoptosis, and fetuses were recovered E 14.5–E 18.5 to assess cardiac outflow tract septation.

All procedures using animals conformed to the principles of laboratory animal care set forth by the National Institutes of Health (NIH) and were approved by the Joslin Diabetes Center Institutional Animal Care and Use Committee.

β-galactosidase staining

Dissected embryos were fixed for 15 min with 4% paraformaldehyde in PBS at 4°C and stained with X-gal as described (Relaix et al., 2003). Briefly, embryos were washed three times with PBS supplemented with 0.2% NP-40, then stained with X-Gal (Roche), using 0.4 mg/mL X-Gal, 2 mM MgCl2, 0.2% NP-40, 20 mM K4Fe(CN)6, and 20 mM K3Fe(CN)6 in PBS at 37°C 16 h. Embryos were extensively washed with PBS prior to visualization.

TUNEL assay

Apoptosis was assayed by TdT-mediated dUTP nick end labeling (TUNEL) in fixed whole-mount E 9.5 embryos as described (Phelan et al., 1997), using FITC-labeled anti-digoxigenin (Roche).


X-gal-stained E 9.5 embryos and E 14.5–E 18.5 fetuses were examined using a Nikon SMZ800 stereomicroscope. Images were captured using an attached SPOT RT CCD camera Model 2.2.1. and SPOT software Ver.3.0.5 (Diagnostics Instruments Inc.). Pax3-GFP+/− E 9.5 embryos and TUNNEL stained embryos were visualized using a Zeiss LSM META 510 upright confocal microscope. Images were captured using Zeiss LSM software. Images were imported into Adobe Photoshop CS2 Version 9.0.2 (Adobe Systems Inc.), and adjustments of image brightness or contrast were all performed equally.

Hematoxylin and Eosin Staining

Whole fetuses recovered on E 14.5 were fixed overnight in 4% phosphate-buffered paraformaldehyde, then stored at 4°C in PBS. Before embedding, the thorax was dissected from the rest of the fetus, submerged in 20% sucrose (in PBS) for 4 hours at 4°C, and then embedded in Tissue-Tek O. C. T. compound (Ted Pella, Inc., Redding, CA). Embedded tissues were stored at −80° C. Five µm serial horizontal sections from the atria through approximately the top half of the ventricles (approximately 40–50 sections per embryos) were obtained using a cryostat and stained with hematoxylin and eosin. Identification of outflow tracts was performed by comparison with Kaufman (Kaufman, 1992).

Dye Injection of Fetal Ventricles

Outflow tract morphology was visualized in situ in E 18.5 fetuses. Fetuses were decapitated immediately upon recovery, the hearts were exposed, and the left ventricles were injected with red acrylic and the right ventricles were injected with and blue acrylic cast (Batson’s No 17 acrylic, Polysciences) as described (Li et al., 1999).

Statistical Analysis

Numbers of β-galactosidase- or TUNEL-positive cells in E 9.5 embryos was determined by using the point selection tool in ImageJ 1.40e software ( The mean ± S.E.M. of numbers of cells from three separate embryos is reported. Data were analyzed by t-test, one way-ANOVA with Neuman Keuls post test, or Chi-square analysis, using Prism 4 software (GraphPad Software).


We are appreciative of Dr. Shoba Thirumangalathu for assistance with preliminary experiments. Michelle Ocana, Mark Chafel and Daniel Tom of the Optical Imaging Facility at the Harvard Center for Neurodegeneration and Repair provided assistance with confocal microscopy. LLS and JML acknowledge the contributions of Dr. Shirley Tilghman to the generation of the Pax3GFP/+ mice, and FR acknowledges the contributions of Dr. Margaret Buckingham to the generation of the Pax3IRESnLacZ/+ mice. We are extremely grateful to Dr. Sean Wu for providing expert advice in interpreting cardiac histological sections. This work was supported by grants from the American Diabetes Association (1-04-RA-65) and the National Institutes of Health (RO1 DK52865) to MRL.


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.


  • Anderson DJ. Cellular and molecular biology of neural crest cell lineage determination. Trends Genet. 1997;13:276–280. [PubMed]
  • Auerbach R. Analysis of the developmental effects of a lethal mutation in the house mouse. J. Exp. Zool. 1954;127:305–329.
  • Banerjee AK. Waardenburg's syndrome associated with ostium secundum atrial septal defect. J R Soc Med. 1986;79:677–678. [PMC free article] [PubMed]
  • Besson WT, Kirby ML, Mierop LHV, Teabeaut JR. Effects of the size of lesions of the cardiac neural crest at various embryonic ages on incidence and type of cardiac defects. Circulation. 1986;73:360–364. [PubMed]
  • Bober E, Franz T, Arnold HH, Gruss P, Tremblay P. Pax-3 is required for the development of limb muscles: a possible role for the migration of dermomyotomal muscle progenitor cells. Development. 1994;120:603–612. [PubMed]
  • Bockman DE, Redmond ME, Waldo K, Davis H, Kirby ML. Effect of neural crest ablation on development of the heart and arch arteries in the chick. Am J Anat. 1987;180:332–341. [PubMed]
  • Bursch W, Paffe S, Putz B, Barthel G, Schulte-Hermann R. Determination of the length of the histological stages of apoptosis in normal liver and in altered hepatic foci of rats. Carcinogenesis. 1990;11:847–853. [PubMed]
  • Cai J, Phelan SA, Hill AL, Loeken MR. Identification of Dep-1, a new gene regulated by the transcription factor Pax-3, as a marker for altered embryonic gene expression during diabetic pregnancy. Diabetes. 1998;47:1803–1805. [PubMed]
  • Chalepakis G, Goulding M, Read A, Strachan T, Gruss P. Molecular basis of splotch and Waardenburg Pax-3 mutations. Proc Natl Acad Sci U S A. 1994;91:3685–3689. [PubMed]
  • Chan WY, Cheung CS, Yung KM, Copp AJ. Cardiac neural crest of the mouse embryo: axial level of origin, migratory pathway and cell autonomy of the splotch (Sp2H) mutant effect. Development. 2004;131:3367–3379. [PubMed]
  • Conway SJ, Bundy J, Chen J, Dickman E, Rogers R, Will BM. Decreased neural crest stem cell expansion is responsible for the conotruncal heart defects within the splotch (Sp(2H))/Pax3 mouse mutant. Cardiovasc Res. 2000;47:314–328. [PubMed]
  • Conway SJ, Godt RE, Hatcher CJ, Leatherbury L, Zolotouchnikov VV, Brotto MA, Copp AJ, Kirby ML, Creazzo TL. Neural crest is involved in development of abnormal myocardial function. J Mol Cell Cardiol. 1997a;29:2675–2685. [PubMed]
  • Conway SJ, Henderson DJ, Copp AJ. Pax3 is required for cardiac neural crest migration in the mouse: evidence from the splotch (Sp2H) mutant. Development. 1997b;124:505–514. [PubMed]
  • Creazzo TL, Godt RE, Leatherbury L, Conway SJ, Kirby ML. Role of cardiac neural crest cells in cardiovascular development. Annu Rev Physiol. 1998;60:267–286. [PubMed]
  • Dameron KM, Volpert OV, Tainsky MA, Bouck N. Control of angiogenesis in fibroblasts by p53 regulation of thrombospondin-1. Science. 1994;265:1582–1584. [PubMed]
  • Donehower LA, Harvey M, Slagle BL, McArthur MJ, Montgomery CA, Jr, Butel JS, Bradley A. Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumours. Nature. 1992;356:215–221. [PubMed]
  • Dorsky RI, Moon RT, Raible DW. Control of neural crest cell fate by the Wnt signalling pathway. Nature. 1998;396:370–373. [PubMed]
  • Endo H, Saito A, Chan PH. Mitochondrial translocation of p53 underlies the selective death of hippocampal CA1 neurons after global cerebral ischaemia. Biochem Soc Trans. 2006;34:1283–1286. [PubMed]
  • Epstein DJ, Vekemans M, Gros P. Splotch (Sp2H), a mutation affecting development of the mouse neural tube, shows a deletion within the paired homeodomain of Pax-3. Cell. 1991;67:767–774. [PubMed]
  • Epstein DJ, Vogan KJ, Trasler DG, Gros P. A mutation within intron 3 of the Pax-3 gene produces aberrantly spliced mRNA transcripts in the splotch (Sp) mouse mutant. Proc Natl Acad Sci U S A. 1993;90:532–536. [PubMed]
  • Epstein JA, Li J, Lang D, Chen F, Brown CB, Jin F, Lu MM, Thomas M, Liu E, Wessels A, Lo CW. Migration of cardiac neural crest cells in Splotch embryos. Development. 2000;127:1869–1870. [PubMed]
  • Epstein JA, Shapiro DN, Cheng J, Lam PY, Maas RL. Pax3 modulates expression of the c-Met receptor during limb muscle development. Proc Natl Acad Sci U S A. 1996;93:4213. [PubMed]
  • Franz T. Persistent truncus arteriosus in the Splotch mutant mouse. Anat Embryol (Berl) 1989;180:457–464. [PubMed]
  • Gary RK, Jensen DA. The p53 inhibitor pifithrin-alpha forms a sparingly soluble derivative via intramolecular cyclization under physiological conditions. Mol Pharm. 2005;2:462–474. [PubMed]
  • Hill AL, Phelan SA, Loeken MR. Reduced expression of pax-3 is associated with overexpression of cdc46 in the mouse embryo. Dev Genes Evol. 1998;208:128–134. [PubMed]
  • Ikeya M, Lee SM, Johnson JE, McMahon AP, Takada S. Wnt signalling required for expansion of neural crest and CNS progenitors. Nature. 1997;389:966–970. [PubMed]
  • Jacks T, Remington L, William BO, Schmitt EM, Halachmi S, Bronson RT, Weinberg RA. Tumor spectrum analysis in p53-mutant mice. Current Biology. 1994;4:1–7. [PubMed]
  • Jiang X, Rowitch DH, Soriano P, McMahon AP, Sucov HM. Fate of the mammalian cardiac neural crest. Development. 2000;127:1607–1616. [PubMed]
  • Kamijo T, Weber JD, Zambetti G, Zindy F, Roussel MF, Sherr CJ. Functional and physical interactions of the ARF tumor suppressor with p53 and Mdm2. Proc Natl Acad Sci U S A. 1998;95:8292–8297. [PubMed]
  • Kaufman MH. The atlas of mouse development. San Diego, CA: Academic Press; 1992.
  • Kioussi C, Gross MK, Gruss P. Pax3: a paired domain gene as a regulator in PNS myelination. Neuron. 1995;15:55–62. [PubMed]
  • Kirby ML. Cardiac morphogenesis--recent research advances. Pediatr Res. 1987;21:219–224. [PubMed]
  • Kirby ML, Gale TF, Stewart DE. Neural crest cells contribute to normal aorticopulmonary septation. Science. 1983;220:1059–1061. [PubMed]
  • Komarov PG, Komarova EA, Kondratov RV, Christov-Tselkov K, Coon JS, Chernov MV, Gudkov AV. A chemical inhibitor of p53 that protects mice from the side effects of cancer therapy. Science. 1999;285:1733–1737. [PubMed]
  • Kwang SJ, Brugger SM, Lazik A, Merrill AE, Wu LY, Liu YH, Ishii M, Sangiorgi FO, Rauchman M, Sucov HM, Maas RL, REM Msx2 is an immediate downstream effector of Pax3 in the development of the murine cardiac neural crest. Development. 2002;129:527–538. [PubMed]
  • Lalwani AK, Brister JR, Fex J, Grundfast KM, Ploplis B, Agustin TBS, Wilcox ER. Further elucidation of the genomic structure of PAX3, and identification of two different point mutations within the PAX3 homeobox that cause Waardenburg syndrome type 1 in two families. Am J Hum Genet. 1995;56:75–83. [PubMed]
  • Li J, Liu KC, Jin F, Lu MM, Epstein JA. Transgenic rescue of congenital heart disease and spina bifida in Splotch mice. Development. 1999;126:2495–2503. [PubMed]
  • Linke SP, Clarkin KC, Di Leonardo A, Tsou A, Wahl GM. A reversible, p53-dependent G0/G1 cell cycle arrest induced by ribonucleotide depletion in the absence of detectable DNA damage. Genes Dev. 1996;10:934–947. [PubMed]
  • Markel P, Shu P, Ebeling C, Carlson GA, Nagle DL, Smutko JS, Moore KJ. Theoretical and empirical issues for marker-assisted breeding of congenic mouse strains. Nat Genet. 1997;17:280–284. [PubMed]
  • Maroto M, Reshef R, Munsterberg AE, Koester S, Goulding M, Lassar AB. Ectopic Pax-3 activates MyoD and Myf-5 expression in embryonic mesoderm and neural tissue. Cell. 1997;89:139–148. [PubMed]
  • Mathieu M, Bourges E, Caron F, Piussan C. Waardenburg's syndrome and severe cyanotic cardiopathy. Arch Fr Pediatr. 1990;47:657–659. [PubMed]
  • Moase CE, Trasler DG. Delayed neural crest cell emigration from Sp and Spd mouse neural tube explants. Teratology. 1990;42:171–182. [PubMed]
  • Murphy PJ, Galigniana MD, Morishima Y, Harrell JM, Kwok RP, Ljungman M, Pratt WB. Pifithrin-alpha inhibits p53 signaling after interaction of the tumor suppressor protein with hsp90 and its nuclear translocation. J Biol Chem. 2004;279:30195–30201. [PubMed]
  • Nishibatake M, Kirby ML, Mierop LHV. Pathogenesis of persistent truncus arteriosus and dextroposed aorta in the chick embryo after neural crest ablation. Circulation. 1987;75:255–264. [PubMed]
  • Pani L, Horal M, Loeken MR. Rescue of neural tube defects in Pax-3-deficient embryos by p53 loss of function: implications for Pax-3- dependent development and tumorigenesis. Genes Dev. 2002;16:676–680. [PubMed]
  • Parr BA, Shea MJ, Vassileva G, McMahon AP. Mouse Wnt genes exhibit discrete domains of expression in the early embryonic CNS and limb buds. Development. 1993;119:247–261. [PubMed]
  • Phelan SA, Ito M, Loeken MR. Neural tube defects in embryos of diabetic mice: role of the Pax-3 gene and apoptosis. Diabetes. 1997;46:1189–1197. [PubMed]
  • Phelan SA, Loeken MR. Identification of a New Binding Motif for the Paired Domain of Pax-3 and Unusual Characteristics of Spacing of Bipartite Recognition Elements on Binding and Transcription Activation. J Biol Chem. 1998;273:19153–19159. [PubMed]
  • Relaix F, Polimeni M, Rocancourt D, Ponzetto C, Schafer BW, Buckingham M. The transcriptional activator PAX3-FKHR rescues the defects of Pax3 mutant mice but induces a myogenic gain-of-function phenotype with ligand-independent activation of Met signaling in vivo. Genes Dev. 2003;17:2950–2965. [PubMed]
  • Robson EJD, He S-J, Eccles MR. A PANorama of PAX genes in cancer and development. Nat Rev Cancer. 2006;6:52–62. [PubMed]
  • Rodriguez M, Schaper J. Apoptosis: measurement and technical issues. Journal of Molecular and Cellular Cardiology. 2005;38:15–20. [PubMed]
  • Roger L, Gadea G, Roux P. Control of cell migration: a tumour suppressor function for p53? Biol Cell. 2006;98:141–152. [PubMed]
  • Sauka-Spengler T, Bronner-Fraser M. Development and evolution of the migratory neural crest: a gene regulatory perspective. Current Opinion in Genetics & Development. 2006;16:360–366. [PubMed]
  • Snider P, Olaopa M, Firulli AB, Conway SJ. Cardiovascular development and the colonizing cardiac neural crest lineage. Scientific World Journal. 2007;7:1090–1113. [PMC free article] [PubMed]
  • Tajbakhsh S, Rocancourt D, Cossu G, Buckingham M. Redefining the Genetic Hierarchies Controlling Skeletal Myogenesis: Pax-3 and Myf-5 Act Upstream of MyoD. Cell. 1997;89:127–138. [PubMed]
  • Vogan KJ, Epstein DJ, Trasler DG, Gros P. The Splotch-Delayed (Spd) Mouse Mutant Carries a Point Mutation within the Paired Box of the Pax-3 Gene. Genomics. 1993;17:364–369. [PubMed]
  • Wakeland E, Morel L, Achey K, Yui M, Longmate J. Speed congenics: a classic technique in the fast lane (relatively speaking) Immunology Today. 1997;18:472–477. [PubMed]