PIERCE1 is critical for specification of left-right asymmetry in mice

The specification of left-right asymmetry of the visceral organs is precisely regulated. The earliest breakage of left-right symmetry occurs as the result of leftward flow generated by asymmetric beating of nodal cilia, which eventually induces asymmetric Nodal/Lefty/Pitx2 expression on the left side of the lateral plate mesoderm. PIERCE1 has been identified as a p53 target gene involved in the DNA damage response. In this study, we found that Pierce1-null mice exhibit severe laterality defects, including situs inversus totalis and heterotaxy with randomized situs and left and right isomerisms. The spectrum of laterality defects was closely correlated with randomized expression of Nodal and its downstream genes, Lefty1/2 and Pitx2. The phenotype of Pierce1-null mice most closely resembled that of mutant mice with impaired ciliogenesis and/or ciliary motility of the node. We also found the loss of asymmetric expression of Cerl2, the earliest flow-responding gene in the node of Pierce1-null embryos. The results suggest that Pierce1-null embryos have defects in generating a symmetry breaking signal including leftward nodal flow. This is the first report implicating a role for PIERCE1 in the symmetry-breaking step of left-right asymmetry specification.

in the left side of the node are achieved by asymmetric expression of the NODAL antagonist, CERL2, in the right side of the node 5,23,24 .
The third step requires asymmetric expression of Nodal and Lefty2 in the left LPM. Transported NODAL activates its own expression in the left LPM through a NODAL-responsive asymmetric enhancer (ASE) of the Nodal gene 25,26 . Lefty2 expression is also induced through the ASE by locally produced NODAL in the left LPM 27 . LEFTY1 on the left side of the midline induced by NODAL 28 may function as a barrier to maintain the expression of Nodal and Lefty2 in the left LPM. The expression of Nodal and Lefty is dynamically regulated by the positive and negative feedback loops mediated by NODAL and LEFTY1/2 [25][26][27] .
The last step of LR asymmetry specification is situs-specific organogenesis. The key regulator of this step is Pitx2, which encodes a transcription factor with zinc finger motifs. PITX2 is expressed asymmetrically in the left LPM, and its expression is regulated by NODAL signaling 29,30 .
Retinoblastoma (Rb)/E2F transcription factors and p53 are integrated into a network that is critical for tumor suppression, and they mutually regulate each other through p16 Ink4a , p21 Waf1 , and p19 Arf31 . Both p53 and Rb/E2F have critical roles in suppressing tumor initiation and progression by controlling a plethora of genes. p53-induced expression in Rb −/− cells 1 (Pierce1) was first identified as an upregulated gene in Rb −/− mouse embryonic fibroblasts (MEFs) 32 . Pierce1 shows a cell cycle-dependent expression pattern and is upregulated in S and/or G2, and its expression is significantly down-regulated by spontaneous immortalization of Rb −/− MEFs 32 . Through subsequent analyses, we demonstrated that Pierce1 is not directly regulated by Rb/E2F, but is a novel p53 target gene 33 . Pierce1 is involved in regulating the DNA damage response, and its knockdown attenuates the expression of diverse p53 target genes upon ultraviolet C (UVC) irradiation 33 . This evidence indicates that Pierce1 plays a critical function in the regulation of cellular transformation and tumorigenesis. To analyze the physiological role of Pierce1, we generated a mouse model deficient for Pierce1 and unexpectedly observed interesting developmental defects. The mice exhibited diverse patterns of laterality, including SS, SI, and left and right isomerisms. We also found randomized expression of asymmetric genes, such as Nodal, Lefty2, and Pitx2, in the LPM of Pierce1-null embryos, indicating that PIERCE1 functions at the initial step of LR specification. This is the first report to identify PIERCE1 as one of the key regulators of LR asymmetry specification.

Results and Discussion
Pierce1 Gt is a null allele. We previously reported that Pierce1 is a downstream target of p53 and may play an important role in maintaining genomic integrity against genotoxic stresses 33 ; however, the in vivo role of Pierce1 has yet to be elucidated. To investigate the function of PIERCE1 in vivo, Pierce1-deficient mice were generated using an embryonic stem (ES) cell clone harboring a gene-trap cassette in intron 2 of the Pierce1 locus (Pierce1 Gt , Supplementary Fig. S1A). Mice homozygous for the gene-trapped Pierce1 allele (Pierce1 Gt/Gt ) were obtained by intercrossing heterozygotes (Supplementary Fig. S1B). Whereas Pierce1 expression is prominent in several tissues, including the brain, lung, kidney, and testis 32 , Pierce1 transcripts were undetectable in these tissues in Pierce1 Gt/Gt mice by reverse transcriptase (RT)-PCR ( Supplementary Fig. S1C), indicating that the Pierce1 Gt allele is a null allele of Pierce1.  Table S1), which deviates significantly from the expected Mendelian ratio (Chi square P = 0.0013). Because neonatal lethality was not evident, these results indicate that Pierce1 −/− mice are partially embryonic lethal. The surviving Pierce1 −/− mice appeared to be grossly normal, viable, and fertile. Upon autopsy, however, we found striking situs alterations in Pierce1 −/− mice ( Fig. 1A-D). Among 78 Pierce1 −/− adults examined, 44 mice (59%) showed SS and 32 (41%) showed SI. SI mutants exhibited complete mirror-image reversal of the position of the heart, stomach, spleen (Fig. 1A), and kidney ( Fig. 1D) along the LR axis. The lobation patterns of the lungs (Fig. 1B) and liver ( Fig. 1C) were reversed as well. This phenotype demonstrates that PIERCE1 is an important regulator of LR specification. PIERCE1 deficiency results in partial embryonic lethality associated with heterotaxia. As the number of Pierce1 −/− mice generated from heterozygous intercrosses was much less than the expected Mendelian ratio (Supplementary Table S1), we examined the possibility that Pierce1 −/− mice were embryonic lethal. Most of the E13.5 and E14.5 Pierce1 −/− embryos exhibited grossly normal appearance and were recovered at the expected Mendelian ratio (Chi square P = 0.5246), but 10 out of 30 Pierce1 −/− embryos at E14.5 were found dead (Supplementary Table S1). Interestingly, Pierce1 −/− embryos at E13.5 and E14.5 exhibited a wide spectrum of laterality defects, including SI and left and right isomerisms (Fig. 1E). SS embryos had normal lungs with one left lobe and four right lobes, whereas the lungs of SI embryos had a mirror image of the normal lung pattern. Most of the dead mutant embryos exhibited bilateral uni-lobed lungs (left pulmonary isomerism, LPI; Fig. 1E) or bilateral tetra-lobed lungs (right pulmonary isomerism, RPI; Fig. 1E) with defects, such as persistent truncus arteriosus in great arteries ( Supplementary Fig. S2).
To analyze LR asymmetry defects in detail, 20 E13.5 Pierce1 −/− embryos were examined using micro-computed tomography (μ CT). Pierce1 −/− embryos displayed SS (9/20), SI (2/20), and various forms of heterotaxy (9/20), including LPI and RPI, with associated cardiac anomalies (Supplementary Table S2). While the apex of the heart pointed leftward in all wild-type embryos, the cardiac apex in Pierce1 −/− mice pointed to the left (12/20), the right (dextrocardia, 4/20), or the midline (mesocardia, 4/20) (Supplementary Table S2 and Supplementary Fig. S3). Most embryos with SI or heterotaxy had various cardiac anomalies, such as interventricular septal defects, hypoplastic heart defects, mitral valve defects, atrophy of the left ventricle, and dilated atria. Laterality defects of the lung lobation pattern were also observed in mutant embryos (Supplementary Table S2 and Supplementary Fig. S3). Among nine mutant embryos with heterotaxy, three had bilateral tetra-lobed lungs (RPI), three had bilateral Scientific RepoRts | 6:27932 | DOI: 10.1038/srep27932 uni-lobed lungs (LPI), and one had a normal lung lobation pattern without a postcaval lobe (Supplementary Table  S2 and Supplementary Fig. S3). We found that the inferior vena cava was reversed (to the left side) in four heterotaxy mutants, as well as in all SI mutants (Supplementary Table S2). Reverse positioning of the stomach (to the right side of the abdomen) was found in some heterotaxy mutants as well as SI mutants (Supplementary Table S2 and Supplementary Fig. S3). Taken together, these data indicate that the partial embryonic lethality observed in Pierce1 −/− mice is closely associated with heterotaxy.
The randomized expression of asymmetric genes suggests that PIERCE1 functions during the initial step(s) of LR specification in the node. The expression of Pierce1 in E8.0 embryos was determined by whole-mount in situ hybridization using a Pierce1 riboprobe. We observed relatively strong expression of Pierce1 in the node area ( Supplementary Fig. S5), suggesting that PIERCE1 functions at the node for LR specification. Based on our results, PIERCE1 is likely involved in the earliest step(s) of LR asymmetry determination, such as symmetry breaking at the node.
Among numerous mutant mouse models that manifest laterality defects (Supplementary Table S3 35 , and 3) rigid and immotile cilia (iv/iv) 18 . ARL13B is a small regulatory GTPase involved in ciliogenesis 40 . NOTO is an essential regulator of multiple genes involved in ciliogenesis and ciliary motility, including Dnahc11 (iv), Dnahc5, and Nphp3 via the Foxj1 and Rfx3 transcription factors 36 .
To investigate if ciliogenesis is affected in Pierce1 −/− embryos, the nodes of E7.5 embryos were examined with scanning electron microscopy (SEM). We found no malformations of monocilia development, such as duplication, bifurcation, partial bifurcation, bulging, or disorganized alignment in the nodal cells of E7.5 Pierce1 −/− embryos (Fig. 3). Next, we examined if expression of Dnahc11 (iv) and Noto is affected in Pierce1 −/− embryos (n = 5 for Dnahc11; n = 13 for Noto). Expression of both genes was detected at the node of E8.0 Pierce1 −/− embryos, comparable to the levels detected in wild-type controls (Supplementary Fig. S6). As the earliest responding gene to leftward flow in the node 23,24 , Cerl2 is bilaterally expressed in the perinodal cells at the early headfold stage when leftward nodal flow is locally generated. The local flow is sufficient to induce down-regulation of Cerl2 on the left-side at the late headfold stage 24 . We observed right-side dominant expression of Cerl2 in the node of WT and Pierce1 +/− embryos, while bilateral expression of Cerl2 was detected in Pierce1 −/− embryos (Fig. 4). These results strongly support that Pierce1 plays a critical role in generating a symmetry breaking signal including leftward nodal flow.
In this study, we uncovered the role of PIERCE1 in the regulation of LR asymmetry. Pierce1 −/− mice at mid-gestational periods were recovered in the expected Mendelian ratio and exhibited a wide spectrum of laterality defects, including isomerisms. In contrast, the mice presented at postnatal periods deviated significantly from the expected Mendelian ratio and only exhibited SS and SI without isomerisms. These results indicate that the mice with isomerisms die in utero due to various cardiovascular malformations associated with isomerisms. The laterality phenotypes of Pierce1 −/− mice suggest that PIERCE1 plays a pivotal role in determining the LR axis at the symmetry-breaking stage. The phenotype of Pierce1 −/− mice most closely resembles that of mutant mice with impaired ciliogenesis and/or ciliary motility of the node. Especially, morphologically normal nodal cilia and bilateral expression of Cerl2 in the node of Pierce1 −/− embryos suggest that Pierce1 −/− has defects in generating a symmetry breaking signal including leftward nodal flow.
Even if the expressional difference of Cerl2 existed between the left and right sides of the node in Pierce1 −/− embryos, the difference was very minor compared with that in controls. Thus, Nodal activity was expected to be suppressed by Cerl2 in the both side of the node. However, we observed randomized expression of Nodal, Lefty2, and Pitx2 in the LPM (Fig. 2). Possible interpretation of this discrepancy is that the amount of bilaterally expressed CERL2 in Pierce1 −/− embryos is around threshold level to suppress NODAL activity, and thus the level of CERL2 in some Pierce1 −/− embryos is not sufficient to suppress NODAL activity in one or both side of the node. Nodal expression in the LPM is initially activated by transported NODAL protein from the node through the NODAL-responsive asymmetric enhancer of the Nodal gene, and then it is regulated through positive and negative feedback loops by NODAL itself and LEFTY2, respectively 25,26 . In Pierce1 −/− embryos, incomplete suppression of NODAL activity could initiate positive feedback loop for Nodal expression in the LPM. This discrepancy is also observed in Arl13b −/− embryos 34 . Five out of six Arl13b −/− embryos with 4-5 somites have bilateral expression of Cerl2, while Nodal expression is detected on the left side (5/19), right side (3/19), or both sides (11/19) 34 .
Notable expression of Pierce1 is detected in adult tissues including the brain, lung, kidney and testis as well as in the node of E8.0 embryos ( Supplementary Fig. S5) 32 . In fact, only distinct cells in these tissues have potential to generate motile cilia and flagella, implying that a dedicated genetic program controls gene expression for formation of motile cilia and flagella 42 . Previously we showed that Pierce1 knockdown compromises the transcriptional activation of several p53 target genes upon UVC irradiation 33 . In addition, the hemagglutinin (HA)-tagged PIERCE1 protein was localized in the nucleus as well as in the cytoplasm (Supplementary Fig. S7). Based on these lines of evidence, we speculate that PIERCE1 contributes to the transcriptional control of genes involved in ciliogenesis and/or ciliary motility in the node although any functionally-defined protein domain of PIERCE1 has not been defined yet.
Further investigation regarding the biochemical, molecular, and genetic mechanisms by which PIERCE1 interacts with known factors involved in LR specification would elucidate the precise role of PIERCE1 during LR asymmetry specification.

Methods
Generation of mice carrying targeted mutations in the Pierce1 gene. A Pierce1-deficient mouse model was generated using a gene-trapped mouse ES cell clone (Cell Line ID: OST3440, Lexicon Genetics, Inc.). The male chimeras were bred with C57BL/6J females, and germline F1 mice were bred with mice on a C57BL/6J or FVB/N background. In this study, most of the adult mice and embryos had mixed genetic backgrounds. The targeted allele was confirmed by PCR analysis. A three-primer PCR strategy was conducted to genotype mice and embryos using the following primers: 5′ -CGAAGGCCAATTAGTGAAGTCAAGC-3′ as a common primer (C), 5′ -CCAGAGAACAGGACTAAGAAGCACG-3′ as a wild-type-specific primer (WT), and 5′ -ATAAACCCTCTTGCAGTTGCATC-3′ for the gene-trap allele-specific primer (GT) ( Supplementary  Fig. S1A). To detect Pierce1 transcripts, complementary DNA (cDNA) samples were prepared using total RNAs from the brain, lung, kidney, and testis of control and Pierce1 −/− mice as previously described 32 , and RT-PCR was conducted using primer pairs specific for Pierce1 (5′ -CCAGTAACCAAACCTACGGA-3′  Noto (NM_001007472, nt682-1181), Pierce1 (IMAGE clone: 317678), and Pitx2 46 using a digoxigenin-UTP labeling kit (Roche) according to manufacturer's instructions.

Scanning electron microscopy (SEM).
To analyze node cilia formation, E7.5 embryos were removed from their extraembryonic tissues, fixed in 2.5% paraformaldehyde/2.5% glutaraldehyde/0.1 M Sorenson's phosphate buffer for at least 24 h, and cut at informative planes with a fine scalpel blade. Embryos were post-fixed with osmium tetroxide, dehydrated with a graded series of ethanol, and critical point dried with liquid CO 2 . Embryos were mounted onto double-sided adhesive tape on metal stubs, coated with gold/palladium, and photographed with a JEOL high-resolution scanning electron microscope (JSM-7401F).