Conditional Mutation of Hand1 in the Mouse Placenta Disrupts Placental Vascular Development Resulting in Fetal Loss in Both Early and Late Pregnancy

Abstract

Congenital heart defects (CHD) affect approximately 1% of all live births, and often require complex surgeries at birth. We have previously demonstrated abnormal placental vascularization in human placentas from fetuses diagnosed with CHD. Hand1 has roles in both heart and placental development and is implicated in CHD development. We utilized two conditionally activated Hand1^A126fs/+ murine mutant models to investigate the importance of cell-specific Hand1 on placental development in early (Nkx2-5^Cre) and late (Cdh5^Cre) pregnancy. Embryonic lethality occurred in Nkx2-5^Cre/Hand1^A126fs/+ embryos with marked fetal demise occurring after E10.5 due to a failure in placental labyrinth formation and therefore the inability to switch to hemotrophic nutrition or maintain sufficient oxygen transfer to the fetus. Labyrinthine vessels failed to develop appropriately and vessel density was significantly lower by day E12.5. In late pregnancy, the occurrence of Cdh5^Cre+;Hand1^A126fs/+ fetuses was reduced from 29% at E12.5 to 20% at E18.5 and remaining fetuses exhibited reduced fetal and placental weights, labyrinth vessel density and placenta angiogenic factor mRNA expression. Our results demonstrate for the first time the necessity of Hand1 in both establishment and remodeling of the exchange area beyond early pregnancy and in patterning vascularization of the placental labyrinth crucial for maintaining pregnancy and successful fetal growth.

1. Introduction

Congenital heart disease (CHD) is the most common birth defect, affecting ~1% of all live births [1]. Babies born with a CHD often undergo corrective surgeries within days of being born, and survival from these surgeries often depends on size at birth [2]. Population studies have demonstrated that pregnancies complicated by CHD carry a higher risk of developing pathologies associated with abnormal placental development and function including growth disturbances [3,4,5], preeclampsia [6,7,8,9], preterm birth [10,11], and stillbirth [12]. The placenta serves as the mediator between mother and developing fetus to provide nutrition and gas exchange, remove fetal wastes, and prevents mixing of maternal and fetal blood [13]. It is unsurprising, while relatively recently demonstrated, that placental development is often affected in pregnancies with CHD as both the heart and placenta are vascular organs that develop concurrently very early in gestation, and shared pathways direct the development of both [14]. Placentas of fetuses with CHD often exhibit changes that disrupt the proper patterning and function of the maternal/fetal exchange area including underdeveloped vasculature and impaired nutrient transport [15,16].

Genome-wide associations studies (GWAS) undertaken previously have identified almost 400 genes associated with CHD [17,18,19]. However, despite decades of research with a primary focus on genetic etiology, the underlying cause of these CHD remains unknown in the majority of cases [20]. Additionally, modeling studies implicating genes and/or environmental influences responsible for abnormal heart development [21] often overlook the involvement of extraembryonic tissues or circumvent it by using ‘cardiac-specific’ conditional knockout mouse models. However, this does not truly reflect the situation in cases of CHD where genetic perturbations would occur in all cells/tissues expressing that gene and many of the identified genes are expressed in other cell types in addition to those found in the heart. Importantly, it has been shown that 68% of 103 knockout models of heart development that exhibited embryonic lethality at or after mid-gestation had abnormal placental development not previously investigated [22]. Such outcomes highlight the strong correlation between placental dysmorphology, angiogenesis and heart development.

One example of a gene involved in both cardiac and placental development is Hand1 [23]. Hand1 is a basic helix-loop-helix transcription factor found in multiple organ systems during embryogenesis. Hand1-null mice are embryonic lethal by E8.5 due to extraembryonic defects in the yolk sac, chorion, allantois, and trophoblast giant cells [23,24,25,26]. Lineage tracing verified Hand1 expression at E9.5 in these tissues [27]. Placental labyrinth vascular endothelium also expressed Hand1 at E14.5 as well as the endothelium of the umbilical vein [27]. Early pregnancy lethality of the global Hand1 knockout models increases the difficulty in studying its role on placental development as the developing conceptuses are lost before the placenta has fully formed. Furthermore, the relevance of the disruption to placentation by trophoblast giant cell failure is controversial in its applicability to the development of the human placenta due to species differences. There are a number of conditional knockout mouse models in which gene disruption is isolated within the fetus, which, depending on the Cre-driver, show varying developmental outcomes from mid-gestation lethality to viable offspring with mild phenotypes [28,29]. However, these studies focused solely on the fetus, despite the fact that some of the Cre-drivers function in extra-embryonic cells of the yolk sac and placenta, which may explain the wide disparities in outcomes.

The implications for Hand1 on heart development have been well characterized [23], however currently, there have been no investigations into the placental contribution to embryonic lethality. In the current study, we aimed to determine the placental contribution to embryonic lethality in the Hand1-mutant mice in early and late pregnancy using two conditional activation mouse models. Nkx2-5 is a well-known cardiac development gene, but is also required for yolk sac angiogenesis and is expressed in endothelial and hematopoietic cells within yolk sac mesoderm [30,31,32]. Previous investigations using the conditional activation knock-in Hand1^SFA126fs mice generated with Nkx2-5^Cre have shown that Hand1^A126fs/+ fetuses die at embryonic day 15.5 (E15.5) and display outflow tract abnormalities, thin myocardium and ventricular septal defects [28]. However, the extraembryonic tissue was not investigated. Cadherin 5 (Cdh5) is an endothelial cell specific gene expressed in the placental fetal vessels. Thus, the use of the Nkx2-5^Cre allows for the study of mutating Hand1 from E8.5 in the yolk sac and labyrinth trophoblast progenitor cells at a time essential for labyrinth establishment and the switch from histiotrophic to hemotrophic nutrition. The Cdh5^Cre model on the other hand, targets the requirements for Hand1 in placental vascular branching in pregnancy when Hand1 is expressed in endothelial cells from E12.5.

2. Results

2.1. Nkx2-5-Cre Is Expressed in Yolk Sac, Trophoblast Cells, and Cardiomyocytes

Conditionally activated Hand1^A126FS/+ females were time mated with homozygous Nkx2-5^Cre males to produce litters containing Nkx2-5^Cre;Hand1^A126FS/+ and Nkx2-5^Cre;Hand1^+/+ embryos. In order to verify the efficacy and spaciotemporal expression of the Nkx2-5^Cre, males were first mated with homozygous tdTomato reporting females. In control mice, immunofluorescence and microscopy confirmed trophoblast expression of the Nkx2-5-Cre at E8.5 as well as Hand1 expression in the trophoblast nucleus (Figure 1A); co-localization of Nkx2-5-Cre expression and the trophoblast marker cytokeratin 7 further confirmed trophoblast expression (Figure 1B). Co-expression of Nkx2-5-Cre and Hand1 in yolk sac, labyrinth trophoblast progenitor cells and syncytiotrophoblast was also confirmed at E9.5 (Figure 1C) and E10.5 (Figure 1D). Sinusoidal giant cells did not express Nkx2-5-Cre at E10.5 but did express cytokeratin 7 (Figure 1E). Analysis of the heart at E9.5 showed strong expression for Nkx2-5-Cre, and Hand1 was localized in the cardiac ventricular tissue (Figure 1F).

2.2. Hand1 Disruption under Nkx2-5-Cre Results in Fetal Demise

Hand1^A126FS/+ females mated with Nkx2-5^Cre males were sacrificed at E8.5, E9.5, E10.5, E12.5 and E14.5, and fetuses and placentas collected. The expected ratio of Nkx2-5^Cre;Hand1^+/+ to Nkx2-5^Cre;Hand1^A126fs/+ was 50% each. While Nkx2-5^Cre;Hand1^+/+ littermates showed consistent survival at all timepoints, marked fetal demise occurred after E10.5 in Nkx2-5^Cre;Hand1^A126fs/+ fetuses (Figure 2) and none survived to E14.5. Nkx2-5^Cre;Hand1^A126fs/+ fetuses were overrepresented at E10.5 (55%); however, they showed a marked decrease at E12.5 (36%) and no viable fetuses were recovered at E14.5. Resorptions were not genotyped due to a lack of available fetal tissue.

2.3. Early Placental Morphogenesis Is Impaired in the Nkx2-5^cre;Hand1^A126fs/+ Implantation Sites

Lack of viable Nkx2-5^Cre;Hand1^A126fs/+ fetuses by E14.5 indicated impaired early placental morphogenesis. Analysis of control litter mate implantation sites at E9.5 revealed invagination and folding of the chorion with scattered Hand1-positive labyrinthine progenitor trophoblasts and a single layer of trophoblast giant cells separating labyrinth from decidua (Figure 3A). Conversely, Nkx2-5^Cre;Hand1^A126fs/+ implantation sites had a disorganized layer of Hand1-positive trophoblast giant cells but lacked Hand1-positive labyrinthine progenitor trophoblasts (Figure 3B). While maternal erythrocytes were present in both genotypes in the developing labyrinthine area, maternal blood spaces were dilated in the conditional activated compared to control littermates and structural integrity of the area disrupted. In addition, yolk sac morphology appeared abnormal in Nkx2-5^Cre;Hand1^A126fs/+ fetuses at E9.5. The yolk sac was adjacent to the chorionic plate in the control littermates (Figure 3C), but was separated from the trophoblast layer in Nkx2-5^Cre;Hand1^A126fs/+ fetuses (Figure 3D).

2.4. Nkx2-5^Cre;Hand1^A126fs/+ Placentas Exhibit Failed Labyrinthine Vascularization and Syncytiotrophoblast Differentiation

Further analysis to identify trophoblast (cytokeratin 7) and fetal vessels (CD-31) in the Nkx2-5^Cre;Hand1^A126fs/+ placentas at E10.5 and E12.5 showed developmental failure of labyrinthine vasculature. CD-31 positive labyrinthine vessels were visible in the control littermate placentas at E12.5 and E14.5 (Figure 4A,B, respectively). Both syncytium and sinusoidal cytokeratin-7-positive trophoblast giant cells were present in the control labyrinth with clear delineation of fetal vasculature from maternal blood spaces (Figure 4A,B). In contrast, very few labyrinthine blood vessels were observed in the Nkx2-5^Cre;Hand1^A126fs/+ placentas at either timepoint (Figure 4C,D). Labyrinthine vessel density trended lower (P = 0.088; Figure 4E) in the conditional activated placentas at E10.5, and by E12.5, vessel counts were significantly lower in Nkx2-5^Cre;Hand1^A126fs/+ labyrinths compared to control (P = 0.001; Figure 4F). Additionally, the labyrinths at E10.5 showed disorganization of the trophoblasts and lack of syncytial layers while retaining sinusoidal trophoblast giant cells, demonstrating a lack of differentiation along the syncytiotrophoblast lineage (Figure 4C). By E12.5, Nkx2-5^Cre;Hand1A^126fs/+ placentas had very few remaining cytokeratin-7-positive trophoblasts (Figure 4D). There were no significant sex differences identified in labyrinth morphology, fetal vessel density or survival.

2.5. Conditional Activation of Mutant Hand1 in Placental Endothelium Resulted in a Reduced Percentage of Cdh5^Cre+;Hand1^A126fs/+ Fetuses by E18.5

Having confirmed the requirement for Hand1 to be expressed in yolk sac and labyrinth trophoblast progenitor cells during early pregnancy on successful development of the labyrinth exchange area, we chose to further assess the requirement for Hand1 in later pregnancy, examining vascular remodeling and expansion in the Cdh5^Cre+;Hand1^A126fs/+. Mating the Cdh5^Cre mice with tdTomato reporting mice confirmed Cdh5-Cre and Hand1 expression in the endothelial cells of the placental labyrinth at GD16.5 (Figure 5). Conditional activated Hand1^A126fs/+ females were then time mated to hemizygous Cdh5^Cre males to produce Cdh5^Cre+;Hand1^A126fs/+, Cdh5^Cre+;Hand1^+/+, Cdh5^Cre−;Hand1^A126fs/+ and Cdh5^Cre−;Hand1^+/+ fetuses. Fetuses and placentas were harvested at E12.5, E16.5 and E18.5. At E12.5, there was a 3% resorption rate which increased to 16% at E16.5 and to 24% at E18.5 (Figure 6). Meanwhile, the percentage of Cdh5^Cre+;Hand1^A126fs/+ fetuses was reduced from 29% at E12.5 to 25% at E16.5 and 20% at E18.5.

2.6. Fetal and Placental Weights Were Reduced in Cdh5^cre; Hand1^A126fs/+ by E18.5

At E16.5, no difference in fetal weight was found between Cdh5^Cre+;Hand1^A126fs/+ and Cdh5^Cre−;Hand1^+/+ littermates (P > 0.05; Figure 7A), however, placental weight of the Cdh5^Cre+;Hand1^A126fs/+ fetuses was significantly lower (P = 0.01; Figure 7B). By E18.5, both fetal and placental weight were lower in the Cdh5^Cre+;Hand1^A126fs/+ genotype when compared to their Cdh5^Cre−;Hand1^+/+ littermates (P < 0.001 and P = 0.002; Figure 7C,D, respectively). There was no difference in fetal or placental weight between males and females in either genotype.

2.7. Labyrinthine Vessel Density and Angiogenic Factor mRNA Expression Was Significantly Reduced in the Cdh5^Cre+/−;Hand1^A126fs/+ Placentas

Using immunohistochemistry, CD-31 positive fetal vessels in the placental labyrinth were identified and counted to assess placental vascularization at E16.5 and E18.5. There was no difference in in the number of labyrinth vessels in the placenta between Cdh5^Cre+;Hand1^A126fs/+ and littermate controls at E16.5 (Figure 8A). However, by E18.5, there was a significant reduction in the number of vessels in the labyrinth of the Cdh5^Cre+;Hand1^A126fs/+ fetuses (P < 0.05; Figure 8B). Whilst vascular density was not reduced in the Cdh5^Cre+;Hand1^A126fs/+ placentas at E16.5, there was a reduction in the mRNA expression of Angiopoietin 1 (Angpt1) and Angpt1 Receptor (Tie2) in the placentas of the Cdh5^Cre+;Hand1^A126fs/+ at E16.5 when compared to littermate controls (P < 0.001 and P = 0.012; Figure 9A,B, respectively). mRNA expression of Angpt2, Vascular endothelial growth factor α (Vegfα) and Placenta growth factor (Plgf) were also assessed however, there was no difference between the genotypes (Data not shown).

3. Discussion

CHDs often require complex surgeries at birth to correct the defect, and one critical predictor of survival is birth weight [2]. In pregnancies complicated by fetal CHD, abnormalities of placental development and function likely contribute to the growth restriction and prematurity that negatively impact clinical outcomes [3,4,5,10,11]. Nearly 400 genes known to be associated with CHD [17,18,19]. However, many of the more in-depth studies, aimed at determining the particular contribution of the genes or environment to cardiac defects, overlook the potential disruption to extraembryonic development that is crucial for maintaining the appropriate in utero environment and fetal organ development and growth. In the present study, we show the effect of targeted loss of Hand1, a gene known to be crucial in heart development, on placental development throughout gestation. Targeted loss of Hand1 in chorionic and labyrinthine progenitor trophoblasts at E8.5-9.5 led to abnormal formation of the placental labyrinth from both a trophoblast and endothelial perspective, ultimately contributing to embryonic lethality from E12.5. Interestingly, the loss of Hand1 in labyrinthine endothelial cells from E12.5 resulted in reduced vascular expansion and remodeling associated with reduced fetal and placental weights, and elevated risk of fetal demise by near-term.

Hand1 has been shown to be expressed during the differentiation of human trophoblast stem cells to syncytiotrophoblast cells in early pregnancy [33]. Interestingly at term, neither isolated human placental cytotrophoblasts nor villous tissue express HAND1 [34]. However, no studies have identified if vascular remodeling throughout gestation requires HAND1. In our present study, targeted early pregnancy loss of Hand1 in chorionic and labyrinthine progenitor trophoblasts resulted in embryonic lethality after E12.5. Co-expression of the Nkx2-5 reporter and Hand1 was limited to labyrinthine progenitor trophoblasts at gestation day 8.5 and expanded to include chorion by E9.5 and syncytium by E10.5. It was previously demonstrated that in Hand1-null mice, trophoblast stem cells fail to differentiate into trophoblast giant cells (TGCs) and exhibited reduced invasion, failed placentation and fetal demise around E6.5 [35], however in our study, Hand1 expression was maintained in the TGCs of Nkx2-5^Cre;Hand1^A126fs/+ placentas allowing us to study labyrinth components and development. Our histological analyses indicate that Nkx2-5^Cre-driven disruption to Hand1 does not significantly impact differentiation of the trophoblast stem cells into TGCs (sinusoidal, canal, and parietal) for establishment of the placenta, but does disrupt labyrinth formation and syncytialization. In mice, at E10-10.5 there is a switch in nutrient and oxygen supply mechanisms whereby support for continued embryonic survival changes from histiotrophic to hemotrophic [36]. In the case of the Nkx2-5^Cre;Hand1^A126fs/+ placentas, failure to form the appropriate layers of the exchange area of the placenta within the labyrinth (syncytium and vasculature) by E10/10.5 ultimately prevented this switch. While subsets of trophoblast cells within the labyrinth were still capable of receiving oxygen and nutrients carried in blood into maternal blood spaces, failure of syncytial layers and inadequate vascularization of the labyrinth resulted in a failure to transfer nutrients and oxygen to the fetal circulation, resulting in fetal demise.

Similar to targeted loss of Hand1 in early pregnancy, targeted loss in the placental endothelial cells after E12.5 resulted in disrupted placental vascular expansion and remodeling. Altering Hand1 expression in endothelium did not appear to cause early fetal loss, as survival at E12.5 was not significantly lower than the spontaneous loss levels in this mouse strain. However, later-gestation fetal demise was significantly higher in the Cdh5^Cre+;Hand1^A126fs/+ fetuses compared to wildtype littermates, pointing to a role for Hand1 in placental vascular remodeling and expansion of the placental labyrinth. The expansion of the labyrinth and increase in branching of the labyrinthine vasculature is vital to support the exponential fetal growth phase of late pregnancy and it’s failure, as we observed from E16.5 in the Cdh5^Cre+;Hand1^A126fs/+, leads to impaired fetal growth and in some cases late term fetal demise. Despite no differences in placental fetal vessel density at E16.5, we did identify reductions in the expression of Angpt1 and Tie2. Angpt1 is a growth factor which predominantly acts on endothelial cells, and whose signaling through its receptor Tie2 is capable of various vascular shaping functions including maintaining vascular integrity, vessel remodeling, cell migration and as anti-inflammatory potential [37]. Further studies, beyond the scope of the present one are required, but reduced expression of Angpt1 and Tie2, as well as reduced labyrinthine vessel density at E18.5 does provide strong evidence that Hand1 expression in placental endothelial cells is required for the angiogenic remodeling required to support fetal growth in late pregnancy.

Complete knockout of Hand1 results in embryonic lethality [23,24,25,26] and given that deletion of Hand1 in only cardiomyocytes results in fetal survival to term with only mild cardiac phenotypes [28], this suggests that the embryonic lethality is not solely due to issues with cardiac development. Results of the current study support this hypothesis; that abnormal placental development due to Hand1 disruption plays a primary role in the fetal demise in the Hand1-null mouse model. The role that the placenta may play in contributing to, or exacerbating the development of CHD, remains understudied. Many genes previously associated with CHD [17,20] have not been adequately investigated in placental development or function. The Deciphering the Mechanisms of Developmental Disorders (DMDD) mouse screen [22], and our study of placental gene expression in human CHD placenta samples identifying multiple ‘heart-specific’ pathways [15,38], underscore the importance of understanding the roles of developmental genes shared between placenta and heart. The dual impact of a genetic disruption to the development of embryonic (heart) and extra-embryonic (placenta, yok sac) organs, plus in utero environmental disruptions in oxygen and nutrient supply secondary to abnormal placenta development, may provide mechanisms that underlie early fetal loss, and most importantly those that underlie the high rate of miscarriage in humans associated with CHD directly or in prior/future pregnancy [39].

This study presents clear evidence that placental development is critical in driving fetal growth and survival in the context of CHD. One limitation to the study is that mouse and human placenta structure and development differs substantially [40], hence there are implications for the translatability of the findings to human pregnancy; although the placenta cell types in which Hand1 was manipulated are present in both mouse and human placenta. Additionally, the manipulation of Hand1 expression occurred either before or after establishment of the placental labyrinth (E10.5) and as such, future studies are required in order to fully understand the implications at this critical time in placental development.

In conclusion, by assessing placental development in the setting of a previously developed Hand1 mutation known to result in cardiac defects [28], we have begun to explore the mechanisms that may result in adverse pregnancy outcomes in the setting of CHD. Disruption of critical stages in placental development and function could contribute to the known clinical observations of stillbirth, fetal growth restriction, and prematurity. Our data highlights the necessity for future research to not only consider the contribution of genetic manipulation to the organ of interest but also the extraembryonic tissue and the in utero environment resulting from its disruption. By doing so, we can then better understand the broad overlap of placental and cardiac development which may ultimately drive novel therapies to improve outcomes for children with CHD.

4. Materials and Methods

4.1. Animal Procedures

All animal procedures were performed under protocols approved by the Institutional Animal Care and Use Committee of CCHMC (IACUC 2018-0065, 12 November 2018). Mice were housed and maintained under temperature and humidity-controlled environments on a 12:12 h light cycle with water and food provided ad libitum.

4.2. Cre Expression and Validation

To verify to efficacy and spaciotemporal expression of the Nkx2-5^IRESCre (Jackson Laboratory; 024637) [31] and Cdh5^cre (Jackson Laboratory, 017968), homozygous Nkx2-5^IRESCre males or Cdh5^Cre males were crossed with homozygous B6.Cg-Gt(ROSA)26Sor^{tm14(CAG-tdTomato)Hze}/J female mice (tdTomato: Jackson Laboratory; 007914). Fetoplacental units were then collected at E8.5, E9.5 and E10.5 for the Nkx2-5^Cre, and E14.5 and E16.5 for the Cdh5^Cre, and processed for cryopreservation and frozen-sectioning. Placentas and embryos were fixed in 4% paraformaldehyde (PFA) with 2.5% polyvinylpyrrolidone (PVP), in phosphate-buffered saline (PBS) for 4 h at room temperature on a rocker plate. Cryoprotecting was achieved by rinsing the fixed tissue in PBS before placing in a 30% sucrose solution until fully infused, then embedding the tissue in OCT and storing at −80 °C. At time of sectioning, blocks were warmed to −20 °C and 7 µm sections were obtained using a cryotome (Leica, Wetzlar, Germany) before being mounted onto slides for immunofluorescent analysis.

4.3. Mice Mating and Genotyping

Conditionally activated Hand1^A126FS/+ [28] females were time mated with homozygous Nkx2-5^IREScre males by placing the breeding pair together overnight and the following morning was designated GD0.5. The mating strategy produced litters containing Nkx2-5^cre;Hand1^A126FS/+ and Nkx2-5^cre;Hand1^+/+ embryos. For the Cdh5 study, conditional activated Hand1^A126fs/+ females were time mated to hemizygous Cdh5^cre males to produce Cdh5^Cre+;Hand1^A126fs/+, Cdh5^Cre+;Hand1^+/+, Cdh5^Cre−;Hand1^A126fs/+, and Cdh5^Cre−;Hand1^+/+ fetuses. Females were sacrificed via carbon dioxide (CO₂) asphyxiation and embryos and placentas weighed and collected at E8.5, E9.5, E10.5, E12.5, and E14.5 (Nkx2-5 study) and E12.5, E16.5 and E18.5 (Cdh5 study). At E8.5 and E9.5, implantation sites were collected, fixed and paraffin embedded for histological analysis. At E10.5 to E18.5, placentas were halved with one half fixed for histology and the other flash-frozen in liquid nitrogen and stored at −80 °C for molecular analyses.

Genotyping was performed on all feto-placental units within the litter by removing part of the yolk sac (E8.5, E9.5 and E10.5) or clipping the tail of fetuses (E12.5, E14.5, E16.5 and E18.5). Tissue was digested in 72.75 µL Direct PCR lysis buffer and 2.25 µL Proteinase K for 4 h at 56 °C, vortexed, then denatured for 30 min at 85 °C. PCR was then performed on 1 µg extracted DNA using FastStart PCR Master (Roche) and primers specific to Hand1 (Supplementary Table S1) under the following conditions: 95 °C for 4 min, 35 cycles of 95 °C × 30 s, 65 °C × 30 s, 72 °C × 1 min, and 72 °C for 7 min. To detect the frameshift mutation at the Hand1 locus, the Hand1 PCR product was digested with the restriction enzyme FauI (R0651, New England Biolabs, Ipswich, MA, USA) following standard protocol using 1 µL of restriction enzyme with 1.5 µL of PCR product, 2 µL CutSmart buffer, and 20 µL RNAse-free water and visualized on a 1% agarose gel. Fetal sex determination was performed using PCR following the protocol outlined in [41] with primers provided in Supplementary Table S1.

4.4. RNA Expression via Reverse Transcription Quantitative Polymerase Chain Reaction(qPCR)

For RNA extraction and qPCR gene expression, flash-frozen half placentas at E10.5 (Nkx2-5 study) and E16.5 (Cdh5 study) were placed into RLT lysis buffer (Qiagen, Hilden, Germany) and homogenized. RNA was extracted using a RNAEasy Mini kit (Qiagen) following standard protocol, and quantification and quality control assessed using Nanodrop Spectrophotometer (Thermo Fisher, Waltham, MA, USA). From each sample 1 µg of RNA was converted to cDNA using the Applied Biosystems high-capacity cDNA conversion kit per protocol. cDNA was stored at −20 °C. mRNA expression levels of Angpt1, Angpt2, Vegfα, Plgf and Tie2 were assayed in 25 µL SYBR Green (Applied Biosystems, Waltham, MA, USA) PCR Master Mix reactions containing 1/40th of the cDNA template and 300 nm/L of forward and reverse primer (Table S1). qPCR was performed in triplicate using the Applied Biosystems StepOne Plus Real-Time PCR System. Expression was normalized using the housekeeping gene Rps20. Relative quantification expression levels were calculated by comparative CT method using StepOne software v2.3 (Applied Biosystems).

4.5. Immunohistochemistry and Immunofluorescence

For paraffin-embedded implantation sites at E8.5 and E9.5, blocks were systematically sectioned until mid-sagittal placenta tissue could be identified and then 5 µm serial sections were obtained. Paraffin-embedded placenta tissue at E10.5 through to E18.5 were also serial sectioned at 5 µm. To assess gross morphology at E9.5 hematoxylin and eosin staining was performed. Briefly, sections were deparaffinized, rehydrated, placed in hematoxylin for 45 s, washed in running tap water, placed in 80% ethanol, dipped 3 times in eosin followed by dehydration, clearing and mounting in xylene-based mounting medium.

Immunohistochemistry for Hand1 was performed on sections of mouse placenta at E9.5 in the Nkx2-5 study, and for CD-31 at E10.5 and E12.5 in the NKx2-5 study and at E16.5 and E18.5 in the Cdh5 study. Sections were deparaffinized, rehydrated, and antigen retrieval was performed using Target Retrieval Solution (Dako, Santa Clara, CA, USA). Endogenous peroxidase activity was blocked using 3% hydrogen peroxide. Nonspecific binding was blocked with a 10% serum solution corresponding to the secondary antibody host species followed by overnight incubation with the primary antibody (antibodies and dilutions outlined Supplementary Table S1). Sections were then washed and incubated with appropriate biotinylated secondary antibodies. Antibody binding was amplified using the avidin-biotin complex (ABC) kit (Vector Laboratories, Inc., Burlingame, CA, USA) and visualized using the enzyme substrate DAB (Vector Laboratories, Inc.). Sections were counterstained briefly in hematoxylin before dehydration, clearing and mounting. Protein localization was observed by light microscopy (Nikon, Tokyo, Japan). In order to assess the number of fetal vessels within the placental labyrinth, two CD-31 stained sections from each placenta were analyzed. Ten high-powered images at least 50 µm apart across the placental labyrinth were obtained, the number of fetal vessels counted manually and then averaged to obtain a number per high powered field (HPF).

To verify the efficacy and spaciotemporal expression of Nkx2-5 and Cdh5, placenta samples from pregnancies between Nkx2-5 and Cdh5 males and tdTomato reporting females were stained for Hand1 or cytokeratin 7 (Nkx2-5 study only) using immunofluorescence. Sections were treated as described for immunohistochemistry, incubated with an appropriate fluorochrome-conjugated secondary antibody and nuclei counter stained with Dapi. For the Nkx-2-5 study, representative images were obtained using the Nikon Eclipse 80i fluorescent microscope. For the Cdh5 study, representative images were obtained using the Axioscan (Zeiss, Oberkochen, Germany).

To assess placental microstructure in the Nkx2-5 study, double-label immunofluorescence for cytokeratin 7 and CD-31 was performed on sections of mouse placental at E9.5, E10.5 and E12.5. Sections were treated as described for immunohistochemistry until the primary antibody was applied. Following, primary antibody incubation, sections were washed, incubated with appropriate fluorochrome-conjugated secondary antibodies and counterstained with DAPI before being mounted with antifade mounting media. Fluorescence microscopy was performed using the Nikon Eclipse 80i fluorescent microscope.

4.6. Statistical Analysis and Data Presentation

Statistical analysis was performed using Prism GraphPad (v8.4.3). Shapiro–Wilk test was used to test for normality in the data. Statistical significance in data which passed the normality test was determined using student’s t-test and expressed as mean ± SEM. Statistical significance in data which did not pass the normality test was determined using Mann–Whitney test and expressed as median ± interquartile range. n number refers to the number of litters. Level of significance (P-value) was defined as P ≤ 0.05.