Alteration of DNA Damage Response Causes Cleft Palate

Cleft palate is one of the most common craniofacial birth defects, however, little is known about how changes in the DNA damage response (DDR) cause cleft palate. To determine the role of DDR during palatogenesis, the DDR process was altered using a pharmacological intervention approach. A compromised DDR caused by a poly (ADP-ribose) polymerase (PARP) enzyme inhibitor resulted in cleft palate in wild-type mouse embryos, with increased DNA damage and apoptosis. In addition, a mouse genetic approach was employed to disrupt breast cancer 1 (BRCA1) and breast cancer 2 (BRCA2), known as key players in DDR. An ectomesenchymal-specific deletion of Brca1 or Brca2 resulted in cleft palate due to attenuation of cell survival. This was supported by the phenotypes of the ectomesenchymal-specific Brca1/Brca2 double-knockout mice. The cleft palate phenotype was rescued by superimposing p53 null alleles, demonstrating that the BRCA1/2–p53 DDR pathway is critical for palatogenesis. Our study highlights the importance of DDR in palatogenesis.


INTRODUCTION
Cleft palate occurs in about one in every 500∼700 live births (Stanier and Moore, 2004;Dixon et al., 2011). Surgical repair is effective and currently the major treatment option. However, affected individuals still face serious functional and social challenges throughout their lives, with cleft palate imposing tremendous psychological and financial burdens on patients and their families. During the last decades, tremendous progress has been made toward understanding the genetic control(s) of this craniofacial abnormality (Murray, 2002;Bush and Jiang, 2012;Marazita, 2012). These studies clearly showed that genetic and cellular signaling pathways are tightly linked during palatogenesis, but the molecular details of cleft palate remain elusive. To overcome these difficulties, it is critical to uncover novel etiological mechanisms underlying cleft palate, enabling us to prevent and develop potential therapeutic strategies for this malformation.
In this regard, recent studies, including ours, reveal the importance of the DNA damage response (DDR) during craniofacial development Calo et al., 2018;Kitami et al., 2018). The DDR encompasses multiple DNA repair and signaling pathways, damage tolerance processes, and cell-cycle checkpoints, which safeguard genomic stability and integrity (McKinnon and Caldecott, 2007;Venkitaraman, 2014). Among craniofacial abnormalities, Treacher Collins syndrome is one of the many great examples that link a dysfunction in DDR to craniofacial defects . Patients with heterozygous mutations in TCOF1 present multiple craniofacial defects (Dixon et al., 2006;Jones et al., 2008); importantly, approximately 40-50% of Tcof1 +/− mice exhibit cleft palate and/or a high-arched palate (also known as "pseudo-cleft") (Dixon et al., 2006;Conley et al., 2016), suggesting that an intact DDR is required for craniofacial development. To support this notion, we found that key components of the DDR, including tumor suppressor genes breast cancer 1 (BRCA1) and breast cancer 2 (BRCA2), are required for craniofacial bone development (Kitami et al., 2018). During early mouse embryogenesis, Treacle associates with the MRNM (MRE11, Rad51, NBS1, and MDC1) complex known as a DNA damage sensor and plays a critical role in DDR to limit oxidative stress induced neuroepithelial cell death . Interestingly, Brca1-labeled DNA damage-induced foci are significantly reduced in Tcof1 +/− cells. Therefore, Treacle plays an important role in the DDR/repair through BRCA1 recruitment to and/or maintenance at sites of DNA damage, but Treacle may only interact with BRCA1 indirectly. Mutation of Tcof1 results in severe craniofacial defects including cleft palate and hypoplasia of the facial bones, which represent several phenotypic similarities observed in neural crestspecific Brca1 mutant mice (Kitami et al., 2018). Therefore, while definitive proof is still required, DDR regulated by both Treacle-BRCA1 axis may play a pivotal role in craniofacial development. These studies reinforce the idea that exploring this previously unappreciated DDR component in craniofacial abnormalities is critical for understanding how DDR functions in palatogenesis.
The aim of this study is to investigate the mechanistic connection(s) between a dysfunctional DDR, due to disruption of BRCA1/BRCA2 and cleft palate, using pharmacological and mouse genetic approaches to dissect the etiology of this craniofacial malformation.

Treatment With a PARP Inhibitor Causes Cleft Palate in Mice
Poly (ADP-ribose) polymerase (PARP) inhibitors have garnered much enthusiasm for the treatment of breast cancer, but the inhibitors themselves may also induce DNA damage (Okazaki and Moss, 1999;Nathanson and Domchek, 2011). Therefore, we hypothesized that administration of a PARP inhibitor to mice may compromise DDR and induce craniofacial defects. To examine the influence of an altered DDR during mouse embryonic development, we used PARP inhibitor Olaparib, which is frequently used for the treatment of ovarian and breast cancer (Nathanson and Domchek, 2011;Chen et al., 2018). Olaparib (50 mg/kg) was administrated daily via intraperitoneal injection (i.p.) to pregnant C57BL/6 mice, from embryonic day (E) 10.5-E17.5; embryos were then harvested and analyzed at E18.5 ( Figure 1A). While Olaparib treatment caused minor embryonic delayed growth (data not shown), the treated embryos developed cleft palate (n = 30/33) (Figures 1B,C). Skeletal analysis revealed that the development of the palatine bone was severely attenuated in Olaparib-treated embryos ( Figure 1C), and histological analysis confirmed that cleft palate was morphologically obvious by E14.5 (Supplementary Figure 1). Because Olaparib-treated embryos also displayed shorter mandibles, we speculated that the cleft palate observed in these embryos could be attributed to abnormal mandible formation. To examine this possibility, we measured the length of the maxilla and mandible in both vehicle-and Olaparib-treated embryos. The ratio of the maxilla-mandible length was comparable until E15.5 (Supplementary Figure 2), indicating that cleft palate developed earlier than the mandibular defects observed in Olaparib-treated embryos. These results suggest that a functionally intact DDR is required for normal palatogenesis in mice.

Olaparib Treatment Causes Increased DNA Damage in the Palate
In addition to its role in the basic excision repair in response to single stranded breaks, PARP also plays critical roles in other repair mechanisms and DNA replication, thus inhibition of PARP will lead to a compromised DDR. Therefore, Olaparib traps PARP on DNA and blocks DNA replication and transcription, leading to DNA breaks that prevent palatogenesis. To explore this hypothesis, we first examined cell proliferation activity and cell survival in Olaparib-treated mouse embryos. While cell proliferation activity was comparable between vehicle-and Olaparib-treated palate tissues (Figure 2A), a large number of apoptotic cells was detected in Olaparib-treated palate tissues ( Figure 2B). This finding is consistent with other mouse models displaying cleft palate due to the massive amount of cell death in palatal mesenchymal cells (Liu et al., 2008(Liu et al., , 2015Goudy et al., 2010). To understand the etiology of cleft palate induced by Olaparib in wild-type embryos, we performed γ-H2AX/Caspase3 and γ-H2AX/phospho (p)-Chk2 marker analysis. Interestingly, over 80% of γ-H2AX positive cells were co-labeled with Caspase3 and/or p-Chk2 ( Figure 2C). This suggests that treatment of Olaparib in wild-type embryos may cause DNA damage with our dosing condition. We confirmed that the protein levels of both γ-H2AX and Caspase3 were significantly increased, as well as p53 levels, in the palate tissues of Olaparib-treated embryos at E13.5 ( Figure 2D). Because RNA-DNA hybrids influence genomic instability, the consequences of the presence of R-loops were examined with an anti-DNA-RNA hybrid (S9.6) antibody. We observed that the intensity of the S9.6 signals was increased in Olaparib-treated palate tissues compared with the controls (Figure 2E). These data demonstrate that DDR plays an important role in the cell survival of palatal tissues during palatogenesis.

Disruption of Brca1 and Brca2 in Neural Crest Cells Leads to Cleft Palate in Mice
The roles of the DDR have been studied extensively in cancer models (McKinnon and Caldecott, 2007;Riley et al., 2008;Venkitaraman, 2014), but little is known about their regulation and activity specifically during palatogenesis. This is partly because the deletion of DDR-associated genes frequently results in early embryonic lethality in mice (Gowen et al., 1996;Hakem et al., 1996;Suzuki et al., 1997;Xu et al., 1999). Therefore, it is not possible to explore how DDR functions during palate development. Since the DNA damage induced by Olaparib may cause cleft palate (Figures 1, 2), we asked which DDR pathways are predominantly involved in the etiology of this malformation. To understand the mechanisms of DDR in palatogenesis, we focused on key DDR elements BRCA1 and BRCA2, and designed RNA probes to examine the expression pattern of these genes in palate tissues in the mouse. Section in situ hybridization analysis showed that Brca1 and Brca2 were widely expressed in both palatal epithelial and mesenchymal cells ( Figure 3A). Because epithelial-mesenchymal interactions are critical for palatogenesis (Bush and Jiang, 2012;He and Chen, 2012;Parada and Chai, 2012) and Brca1 and Brca2 are expressed in both palatal epithelial and mesenchymal cells (Figure 3A), we asked whether BRCA1 and BRCA2 function in these types of cells. First, we disrupted Brca1 and Brca2 in a neural crest-specific manner using a wingless-related MMTV integration site1 (Wnt1)-Cre driver line (Danielian et al., 1998;Xu et al., 1999). This approach allowed us to examine the function of BRCA1 and BRCA2 in neural crest-derived ectomesenchymal cells in the palate (Chai et al., 2000;Bronner and LeDouarin, 2012). We observed that mice lacking Brca1 and/or Brca2 in neural crest cells ("Brca1 cKO" or "Brca2 cKO" hereafter) displayed severe cleft palate with 100% phenotypic penetrance ( Figure 3B). While Brca1 cKO and Brca2 cKO mutants were born at Mendelian ratios, they could not survive more than 24 h due to the cleft palate phenotype. Histological analysis confirmed that both Brca1 cKO and Brca2 cKO mutants showed cleft palate by E14.5 (Figures 3C-F). Next, we disrupted Brca1 and Brca2 using a Keratin14 (K14)-Cre driver line (Dassule et al., 2000), which allowed us to examine the function of BRCA1 and BRCA2 in palatal epithelial cells. In contrast to Brca1 cKO and Brca2 cKO mutants, mice with an epithelial cell-specific deletion of Brca1 or Brca2 did not show any overt craniofacial defects, including cleft palate FIGURE 2 | DDR is indispensable for regulating cell survival of palate tissues. (A) Coronal palate sections of vehicle-or Olaparib-treated mouse embryos were labeled with BrdU (green) to detect proliferative cells, which were quantified. PS, palatal shelf. (B) TUNEL assay (green) and corresponding quantification of palate sections from vehicle-or Olaparib-treated mouse embryos. Arrows show TUNEL-positive signals. PS, palatal shelf. (C) Immunostaining for γ-H2AX (magenta, left panels), Caspase3 (Green, left panels) and γ-H2AX (magenta, right panels), p-Chk2 (green, right panels), and corresponding quantification in palate sections from vehicle-or Olaparib-treated mouse embryos at E13.5. The yellow boxes show the high-magnification image of γ-H2AX/Caspase3 and γ-H2AX/p-Chk2-positive cells. PS, palatal shelf. (D) Protein levels of γ-H2AX, Caspase3 and p53 were examined by Western blotting and quantified. (E) S9.6 intensity was examined using palatal cell lysates from vehicle-or Olaparib-treated mouse embryos at E13.5. Data in panels (A-D) are represented as mean ± SD, n = 3 in each group. Data in panel (E) are represented as mean ± SD, n = 10 in each group. *p < 0.05; **p < 0.01; N.S., not significant.
Frontiers in Physiology | www.frontiersin.org  Figure 3). These results indicate that BRCA1 and BRCA2 play a critical role during palatogenesis in murine neural crest-derived ectomesenchymal cells, but not in epithelial cells.

Disruption of Brca1 and Brca2 Induces Palatal Cell Death
While BRCA1 and BRCA2 are known tumor suppressor genes, they may also function as regulators of cell proliferation and/or cell survival during mouse embryogenesis. Therefore, we asked whether the cleft palate observed in Brca1 cKO and Brca2 cKO mutants could be attributed to an absence of cell proliferation or excessive cell death. Whereas palatal mesenchymal cells showed similar proliferation activity in control and Brca1 cKO mutant mice ( Figure 4A), a large number of apoptotic cells were detected in the palate of Brca1 cKO mutants ( Figure 4B). Similar to Brca1 cKO mutants, cell proliferation activity was comparable ( Figure 4C), and cell survival was severely attenuated, in the palate of Brca2 cKO compared to that of the controls ( Figure 4D). Because cell death in pre-migratory neural crest cells frequently leads to craniofacial abnormalities (Noden and Trainor, 2005;Dixon et al., 2006), one might predict that increased cell death in pre-migratory neural crest cells may lead to cleft palate in Brca1 cKO and Brca2 cKO mutants. However, our recent findings show that BRCA1/2 is less likely to play a role in pre-migratory neural crest cells (Kitami et al., 2018). Altogether, these results suggest that the cleft palate phenotypes observed in Brca1 cKO and Brca2 cKO mutants are mainly due to excessive cell death, rather than a reduction in cell proliferation activity or defects in pre-migratory neural crests during craniofacial development.

Brca1 and Brca2 Genes Function Together in Palatogenesis
Breast cancer 1 plays a critical role in the response to DNA damage (Khanna and Jackson, 2001;Roy et al., 2011); additionally, BRCA2 is essential for maintaining genome integrity (McKinnon and Caldecott, 2007). Given the high similarity of both Brca1 cKO and Brca2 cKO cleft palate phenotypes (Figures 3, 4), it is important to examine whether Brca1 and Brca2 function redundantly during palatogenesis. To do so, we generated neural crest-specific Brca1 and Brca2 doubleknockout mice ("Brca1/2 dKO" hereafter). As in Brca1 cKO and Brca2 cKO mutants, deletion of both Brca1 and Brca2 in neural crest cells resulted in neonatal death (Supplementary Figure 4). Importantly, Brca1/2 dKO mutants developed much more severe cleft palate phenotypes than the Brca1 cKO and/or Brca2 cKO mutants ( Figure 5A). Measurement of the palatal defective area confirmed that the malformation seen in Brca1/2 dKO mutants was very severe (Figure 5B), and skeletal analysis revealed that palatine bone formation was severely attenuated compared with Brca1 cKO and/or Brca2 cKO mutants (Figure 5C). To examine whether the disruption of both Brca1 and Brca2 in palatal tissues causes severe cell death, TUNEL analysis was performed. As expected, compared with Brca1 cKO or Brca2 cKO mutants, increased cell death was observed in Brca1/2 dKO mutants ( Figure 5D, upper panels). Consistent with the observation of excessive apoptosis in Brca1/2 dKO mutants, the number of γ-H2AX-stained cells co-labeled with Caspase3 or p-Chk2 was significantly increased in Brca1/2 dKO, even when compared with either the Brca1 cKO or the Brca2 cKO mutants (Figure 5D, middle and lower panels). These results suggest that while BRCA1 has multiple functions in the repair process of DNA damage and while BRCA2 plays a pivotal role in homologous recombination (HR), both Brca1 and Brca2 function synergistically during palatogenesis in mice.

Deletion of p53 Partially Rescues Cleft Palate by Decreasing DNA Damage and Preventing Cell Death in Brca1 and Brca2 Mutants
It is known that DNA damage triggers p53 stabilization; if this damage cannot be repaired, it induces apoptosis (Riley et al., 2008). Therefore, we hypothesized that p53-mediated apoptosis may cause cleft palate in Brca1 cKO and Brca2 cKO mutants. Consistent with our hypothesis, the protein levels of p53 in palate tissues were much higher in Brca1 cKO and Brca2 cKO mutants than in the controls (Supplementary Figures 5A,B), suggesting that DNA damage-induced p53 stabilization may be responsible for the etiology of cleft palate in Brca1 cKO and Brca2 cKO mutants. To test this possibility in vivo, null alleles of p53 were superimposed into Brca1 cKO and Brca2 cKO mutants in a neural crest-specific manner (Marino et al., 2000). Consistent with our prediction, the p53 deletion rescued the cleft palate phenotype observed in Brca1 cKO and Brca2 cKO mutants (Figure 6A,B). Importantly, two copies of the p53 deletion alleles were able to rescue the cleft palate more efficiently in Brca1 cKO and Brca2 cKO mutants than one copy (Supplementary  Figures 5C,D). Phenotypic recovery was also confirmed by histological analysis (Supplementary Figure 6). We found that the p53 deletion sufficiently suppressed the enhanced cell death seen in Brca1 cKO and Brca2 cKO palates (Figures 6C,D,  upper panels, Figures 6E,F). This suppression of cell death in rescued mice was linked to a reduced number of γ-H2AX-stained cells co-labeled with Caspase3 (Figures 6C,D, lower panels,  Figures 6E,F). Thus, reduction of p53 levels in Brca1 cKO and Brca2 cKO mutants predominantly rescues the cleft palate by decreasing DNA damage-induced cell death. Altogether, these data demonstrate that BRCA1 and BRCA2 deficiency-induced cleft palate can be partially rescued by inactivating p53 through reduction of DNA damage-induced cell death.

DISCUSSION
The molecular etiology of cleft palate is very complex. Previous studies have suggested that altered transcriptional regulation, developmental signals, and/or epigenetic factors result in cleft palate. However, much still remains unclear about the regulation mechanisms of palate development. Recent studies, including ours, have shown the importance of the DDR during craniofacial development Calo et al., 2018;Kitami et al., 2018), but its role in normal palatogenesis is still not known. In this report, we found a novel connection between the etiology of cleft palate and the DDR, mainly via BRCA1/2-p53-dependent mechanisms. Our data indicate that maintaining genomic stability through the BRCA1 and BRCA2 DDR machinery may be key in preventing cleft palate.

Temporospatial DDR Regulation Is Important for Palatogenesis
While a fine-tuned DDR machinery is essential for orchestrating morphological events, the action of the DDR during embryogenesis is highly temporal (Vinson and Hales, 2002;McKinnon, 2017). This is expected because, as for transcription and replication, DDR is a chromatin-associated process that is tightly regulated in a temporospatial manner. For example, although the Brca1 and Brca2 genes are expressed throughout the entire embryonic development process in mice, their expressions peak around E12.5-E13.5 in the craniofacial region (Lane et al., 1995;Rajan et al., 1997). Importantly, mid-gestation (i.e., E12.5-E13.5) is the stage when cellular metabolism switches from being anaerobic to being oxidative Hales, 2001, 2002), and this metabolic change may increase genotoxic stress in mouse embryos. Therefore, a considerable amount of DNA damage likely needs to be repaired in the craniofacial region around mid-gestation; if cells in this region fail to complete this task, severe craniofacial abnormalities may result. In support of these hypotheses, we recently found that the onset of craniofacial bone abnormalities in neural crest-specific Brca1 and Brca2 mutant mice indeed occurred around mid-gestation (Kitami et al., 2018), suggesting that temporospatial regulation of the DDR may be critically important for craniofacial development, including palatogenesis. Consistent with this prediction, the alterations in DDR, by either pharmacological interventions or genetic manipulation of DDR components analyzed in our study, caused cleft palate, further supporting our hypothesis that temporospatial DDR regulation is indispensable for palatogenesis.
To understand the mechanisms of DDR during palatogenesis, we focused on two important DDR elements: BRCA1 and BRCA2. Both genes are well-known DDR key players, critical for repairing DNA double-strand breaks to maintain genome integrity (McKinnon and Caldecott, 2007;Venkitaraman, 2014). While little is known about how DDR functions during palate development, recent studies have shown that while it is modest, there are BRCA1 and BRCA2 variants in non-syndromic cleft lip and palate patients (Kobayashi et al., 2013;Rodriguez et al., 2018). Another study also shows that the genetic variants in other Immunostaining for γ-H2AX (magenta, middle panels) and Caspase3 (green, middle panels), γ-H2AX (magenta, lower panels) and p-Chk2 (green, lower panels), and corresponding quantification of palate sections from control, Brca1 cKO, Brca2 cKO, and Brca1/2 dKO mice. The yellow box shows the high-magnification image of γ-H2AX/Caspase3and γ-H2AX/p-Chk2-positive cells. PS, palatal shelf. Data in panels (B,D) are represented as mean ± SD, n = 3 in each group. *p < 0.05; **p < 0.01; N.S., not significant. DDR components including BRIP1 may contribute to the risk of non-syndromic cleft lip and palate (Mostowska et al., 2014). Therefore, it would be reasonable to speculate that BRCA1/2dependent DDR may be linked to palatogenesis. Since we previously reported that both BRCA1 and BRCA2 are required for craniofacial bone formation (Kitami et al., 2018), these two DDR components provide us with a premise to examine whether dysregulation of DDR via the BRCA1 or the BRCA2 pathway leads to cleft palate. Importantly, conventional disruption of Brca1 and Brca2 in mice results in embryonic lethality, and this is mainly due to increased cell death and/or restricted proliferation in neuroepithelial cells (Gowen et al., 1996;Hakem et al., 1996;Suzuki et al., 1997;Xu et al., 2001). In addition, mice lacking Brca1 and Brca2 in neural stem cells display severe brain defects (Frappart et al., 2007;Pulvers and Huttner, 2009;Pao et al., 2014). These data indicate that BRCA1 and BRCA2 may play a pivotal role in neuroepithelial cells, which are critical for developing craniofacial structures such as palate tissues. Our findings support this notion that both BRCA1 and BRCA2 are required for the differentiation of neuroepithelial lineage cells, i.e., neural crest derivatives. Here we show that neural crestspecific disruption of Brca1 and Brca2 results in increased cell death in palatal mesenchymal cells, with the dual deletion of Brca1/Brca2 leading to extensive apoptosis. Therefore, BRCA1 and BRCA2 in neural crest cells are most likely essential for the early onset of palatogenesis.
As in our neural crest-specific Brca1 and Brca2 mice, haploinsufficiency of the Tcof1 mutation results in enhanced p53 production, and genetic suppression of p53 in Tcof1 mutant mice rescues the craniofacial bone defects (Jones et al., 2008). Of note, Tcof1 haploinsufficiency causes neuroepithelial cell death, and loss of Tcof1 decreases the accumulation of BRCA1 at DNA damage sites , suggesting that TCOF1 and BRCA1 (and also possibly BRCA2) may be associated with the DDR during palatogenesis. It is worth noting that the Tcof1 mutation results in cleft palate approximately 40-50% of the time, while the remaining Tcof1 +/− mice have a high arched palate (Dixon et al., 2006;Conley et al., 2016). It remains to be determined whether TCOF1, BRCA1, and BRCA2 genetically interact in palate development, and for this reason it is important to study how these specific DDR-related elements are involved in the etiology of cleft palate.

BRCA1/2-Dependent DDR Is Critical for Palatogenesis
Whereas the importance of BRCA1/2-dependent DDR and repair has been adequately studied in tumorigenesis, how the BRCA1-and BRCA2-dependent DDR pathways function in palate development remains elusive. In addition to its role in regulating HR, BRCA1 acts in multiple aspects of DDR; on the other hand, BRCA2 facilitates HR. In this study, we found that neural crest-specific Brca1 and Brca2 mutants display an almost identical cleft palate phenotype. We initially hypothesized that a rapid neural crest cell proliferation may generate replication stress; this would render BRCA1 and BRCA2 essential to protect the replication fork's integrity, with their deficiency leading to increased DNA damage during palate development. Indeed, it is known that BRCA1 and BRCA2 play a pivotal role in protecting the stability of a stalled replication fork to maintain genome integrity (Prakash et al., 2015). Surprisingly, the cell proliferation activity in the palatal mesenchyme in both Brca1 and Brca2 mutants was normal, demonstrating that DDR via BRCA1 or BRCA2 may be involved in alternative mechanisms other than regulating proliferation activity. In our study, increased cell death was predominantly involved in the etiology of cleft palate, and palatal defects were rescued by p53 reduction. Given the essential role of BRCA1 and BRCA2 in embryonic development and the identical cleft palate phenotypes in mice with neural crest-specific deletions of these genes, we conclude that BRCA1 and BRCA2 act redundantly during palatogenesis. Altogether, our data highlight the requirement of DDR, via BRCA1 and BRCA2, for normal palate development.

PARP Inhibitor Therapy May Cause Fetal Cleft Palate
Breast cancer is the most common cancer diagnosed during pregnancy and affects about one in 3,000 pregnant women (Schedin, 2006;Letourneau et al., 2011). While curing breast cancer during pregnancy would be challenging, the use of PARP inhibitors has garnered much enthusiasm (Okazaki and Moss, 1999;Nathanson and Domchek, 2011), and several such agents have been approved by the FDA for the treatment of patients with ovarian and breast cancers harboring inherited BRCA1 and/or BRCA2 mutations. Since PARP is responsible for repairing single-strand breaks of damaged DNA, its inhibition leads to the formation of multiple double-strand DNA breaks (Lord et al., 2015;Chen et al., 2018). However, there is no indication whether pregnant women with breast cancer can be treated safely with these drugs, and little is known about the development of birth defects deriving from their use during pregnancy. In our study, we found that treatment with Olaparib results in severe cleft palate in wild-type mouse embryos. Out data suggest that Olaparib traps PARP on DNA, thus blocking DNA replication and transcription leading to DNA breaks, therefore it may lead to cleft palate. To the best of our knowledge, this is the first evidence of treatment with a PARP inhibitor causing cleft palate. Therefore, exploring the mechanisms of DDR may provide insights into the prevention of cleft palate when PARP inhibitor therapy is used in pregnant patients.
In summary, we demonstrated the critical function of DDR in palate development by pharmacological interventional and genetic approaches and determined that BRCA1/2-p53 dependent pathways are key in the regulation of normal palatogenesis in mice.
All mice were maintained in the animal facility of The University of Texas Medical School at Houston. The experimental protocol was reviewed and approved by the Animal Welfare Committee and the Institutional Animal Care and Use Committee of The University of Texas Medical School at Houston.

Treatment of Animals With PARP Inhibitor (Olaparib)
Olaparib (AZD2281, Selleck Chemicals) was dissolved in 5.7% DMSO (Rottenberg et al., 2008;Szabova et al., 2014;Henneman et al., 2015). A dose of 50 mg of the Olaparib per kg of body weight per day was injected intraperitoneally into C57BL/6 pregnant mice (12-16 weeks old). Olaparib dosing conditions were established based on the frequency of embryonic lethality and phenotypic penetration of cleft palate (Supplementary Table 1).

Section in situ Hybridization
PCR fragments containing regions of mouse Brca1 (nucleotides 2,302-3,028 bp) and mouse Brca2 (nucleotides 162-653 bp) cDNA were cloned into the pCR R II vector. Digoxygeninlabeled sense-and/or antisense-Brca1 and -Brca2 probes were generated by in vitro transcription (Sigma-Aldrichı). Section in situ hybridization was performed using standard procedures (Komatsu et al., 2014).

R-Loop Assay
DNA was extracted from vehicle-or Olaparib-treated palatal tissues at E13.5 (n = 10). DNA concentration was measured with a Nano-drop spectrophotometer, and 500 ng of DNA were applied onto a membrane using a dot blot apparatus. After cross-linking, the membrane was subjected to R-loop detection with the S9.6 antibody. S9.6 signals were normalized to the total DNA signals and quantified using the ImageJ software.

Statistical Analysis
The Student's t test and one-way ANOVA with post hoc Tukey HSD test were used for statistical analysis. A p value of less than 0.05 was considered statistically significant.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The animal study was reviewed and approved by The Animal Welfare Committee and the Institutional Animal Care and Use Committee of The University of Texas Medical School at Houston.

AUTHOR CONTRIBUTIONS
YK designed the study. HY, KK, XW, LH, and JW performed the experiments. HY, KK, XW, BW, and YK analyzed the data. BW and YK wrote the manuscript. All authors contributed to the article and approved the submitted version.

FUNDING
This study was supported by research grants NCI/NIH R01CA155025 (BW) and NIDCR/NIH R01DE025897 (YK), and by a fellowship from the Uehara Memorial Foundation (HY).