HAND1 knockdown disrupts trophoblast global gene expression

Abstract Congenital heart disease (CHD) affects nearly 1% of births annually, and CHD pregnancies carry increased risk of developing pathologies of abnormal placentation. We previously reported significant developmental impacts of disrupting Hand1, a gene associated with CHD, expression in placenta trophoblast and endothelial cells in multiple mouse models. In this study, we aimed to build upon this knowledge and characterize the mechanistic impacts of disrupting HAND1 on human placenta trophoblast and vascular endothelial cell gene expression. HAND1 gene expression was silenced in BeWo cells, a choriocarcinoma model of human cytotrophoblasts, (n = 3–9 passages) and isolated human placental microvascular endothelial cells (HPMVEC; n = 3 passages), with HAND1 siRNA for 96 h. Cells were harvested, mRNA isolated and RNA sequencing performed using the Illumina NextSeq 550 platform. Normalization and differential gene expression analyses were conducted using general linear modeling in edgeR packages. Statistical significance was determined using a log2 fold change of >1.0 or < −1.0 and unadjusted p‐value ≤0.05. Panther DB was used for overrepresentation analysis, and String DB for protein association network analysis. There was downregulation of 664 genes, and upregulation of 59 genes in BeWo cells with direct HAND1 knockdown. Overrepresentation analysis identified disruption to pathways including cell differentiation, localization, and cell projection organization. In contrast, only seven genes were changed with direct HAND1 knockdown in HPMVECs. Disruption to HAND1 expression significantly alters gene expression profile in trophoblast but not endothelial cells. This data provides further evidence that future studies on genetic perturbations in CHDs should consider the extra‐embryonic tissue in addition to the fetal heart.


| INTRODUCTION
Congenital heart disease (CHD) affects nearly 1% of births annually, and often requires surgical intervention for repair and correction (van der Linde et al., 2011). Despite advances in care, CHD is still a leading cause of infant mortality from pregnancies complicated by birth defects accounting for approximately 4.2% of all neonatal deaths, and is associated with healthcare costs of approximately $6.1 billion dollars annually (Arth et al., 2017). About 25% of babies with CHD will have severe CHD (Oster et al., 2013), and the prevalence of CHD, particularly mild disease, is increasing. Pregnancies complicated with CHD also carry increased risk of developing pathologies of abnormal placentation including fetal growth restriction (FGR), preeclampsia, preterm birth, and stillbirth (Courtney et al., 2018). These adverse pregnancy outcomes greatly impact the after birth care required for the infant, specifically with regards to morbidity and mortality associated with surgery to correct the cardiac defect, as well as impacting childhood development and survival (Courtney et al., 2020).
Currently, we lack comprehensive understanding of the embryonic and fetal relationship between development of the placenta and the heart. The placenta is essential to fetal growth and development and placental dysfunction impacts perinatal outcomes (Burton & Fowden, 2015). This is because the placenta plays an essential role in regulating the transport of nutrients and oxygen from the mother to the fetus, and mediates maternal-fetal communication.
In utero, initial heart and placenta development occurs in parallel during the first 3 weeks of gestation. We and others have shown common molecular pathways in placental and heart development (Courtney et al., 2021;Firulli et al., 2010). Using a systematic computational approach, we have shown numerous commonly expressed genes between first trimester human heart and placenta cells, which if disrupted may concurrently contribute to the developmental perturbations resulting in CHD (Wilson et al., 2022). Additionally, our lab has previously reported disrupted vascular development, and morphologic abnormalities and placental insufficiency in placentas from human pregnancies with CHD (Courtney et al., 2020;Jones et al., 2015). However, in-depth knowledge of the regulation of these common molecular pathways, particularly in relation to vasculogenesis and angiogenesis, is lacking. Additionally, studies using mice models to better understand heart development have also been shown to exhibit abnormal placental development, although the latter is very rarely investigated (Perez-Garcia et al., 2018).
HAND1 is a transcription factor related to the basic helix-loop-helix (bHLH) with essential roles in embryonic placenta and heart development (Firulli et al., 1998), and is expressed in first trimester human placental trophoblast (James et al., 2015;Telugu et al., 2013) and first trimester cardiac cells (Cui et al., 2019). Hand1-null mice are embryonic lethal by E8.5 due to defects in the extraembryonic tissues (Cserjesi et al., 1995;Firulli et al., 1998;Morikawa & Cserjesi, 2004;Riley et al., 1998). In mouse models where a knock-in stop codon is expressed on one allele of the Hand1 gene (Hand1 A126fs/+ ) in specific cell types of the fetal heart because of different CRE drivers, embryonic lethality does not occur until gestational day 15.5 and fetuses exhibit outflow tract abnormalities, thin myocardium and ventricular septal defects (Firulli et al., 1998). More recently, we have shown that placentas of the Nkx2.5 Cre /Hand1 A126fs/+ mouse, in which the stop codon in the Hand1 gene is knocked-in in chorion and labyrinth trophoblast progenitor cells, to expand labyrinth trophoblast precursor population, or develop appropriate syncytiotrophoblast and endothelium in the labyrinth (nutrient exchange area), by gestational day 10 resulting in fetal demise (Courtney et al., 2021). However, using the Cdh5 cre /Hand1 A126fs/+ mouse model, which results in a stop codon in the Hand1 gene specifically in endothelial cells of the placenta and heart, the placentas are only affected in latergestation with reduced placental vascular branching, but little effect on fetal heart development (Courtney et al., 2021). Signaling between trophoblast cells and villous endothelium is necessary for placental development and function, however there is a paucity of data looking at how signaling occurs at a molecular level during development. In this study, we aimed to build upon our discoveries in the mouse models and characterize the impact of disrupting HAND1 expression on molecular signaling in human placenta trophoblast, and placental villous endothelial cells independently.

BeWo and HPMVECs
BeWo cells or HPMVECs were plated (2.5 × 10 5 cells/ well) onto Millicell hanging cell culture inserts (Millipore, Bedford, MA) in 12 chamber culture trays with respective culture media in both the well insert chamber. For HPMVECs, the inserts were pre-coated with attachment factor. After 24 h, the well culture media was removed, cells washed with PBS, and replaced with treatment media: minimum essential media (MEM; Sigma, St. Louis, MO) containing 1% L-glutamine (Gibco, Waltham, MA) and 1% penicillin-streptomycin. To knockdown HAND1 cells (n = 9 passages for BeWo cells and n = 3 passages for HPMVECs) were treated with 3 μl Lipofectamine +4 μl 10 μM HAND1 siRNA for 96 h as laboratory standard (Jones et al., 2009). Treatment of cells with 3 μl Lipofectamine +3 μl 10 μM Allstars negative siRNA was used as a negative control. After 6 h, 10% FBS was added to ensure cell survival without starvation effects of MEM. At 96 h, cells were harvested for RNA isolation following treatment.

| Isolation of RNA and confirmation of HAND1 knockdown via QPCR
Cells were lysed using RLT Buffer from Qiagen (Valencia, CA) following manufacturer's instructions. Total RNA was isolated using the RNeasy Mini Kit, QIAshredder, and on-column DNA digest (Qiagen) following the protocol provided by the manufacturer. For QPCR analysis, total RNA was quantified using a Nanodrop Spectrophotometer. One milligram of RNA was then converted to cDNA utilizing the Applied Biosystems High Capacity cDNA kit following manufacturer's protocol. QPCR was performed in duplicate reactions containing PowerUp SYBR Green (Applied Biosystems) as per manufacturer's instructions and with primers (Table S1) on the StepOne-Plus Real-Time PCR System (Applied Biosystems). Relative mRNA expression was calculated using the comparative CT method (Pfaffl, 2001) with the StepOne Software v2.3 (Applied Biosystems) normalizing genes to ACTB.

| Transcriptome generation and differential gene expression bioinformatic analyses
Total RNA was isolated from cells using same protocol as for QPCR. 50-100 μg of RNA from the various treated BeWo cells (n = 3), and HPMVECs (n = 3) was submitted to the University of Cincinnati Genomics, Epigenomics and Sequencing Core for RNA quality assessment and sequencing. RNA quality control (QC) was conducted on an Agilent 2100 Bioanalyzer for quality control. Only samples with a RNA integrity number (RIN) >8 were processed for sequencing. For each RNA sample, poly-A RNA libraries were generated using NEBNext Ultra Directional RNA Library Prep kits (New England Biolabs, Ipswich, MA), and the TruSeq SR Cluster kits (v3, Illumina). Transcriptomes were generated on the Illumina NextSeq 550 platform with ~25 million reads per sample with a single end read length between F I G U R E 1 HAND1 mRNA expression in BeWo cells and human placental microvascular endothelial cells (HPMVECs) following siRNA treatment. (a) In BeWo cells, HAND1 mRNA expression was reduced by 61% compared to Allstar negative control treatment after 96 h. (b) In HPMVECs, HAND1 mRNA expression was reduced by 69% compared to Allstar negative control treatment after 96 h. data are median ± interquartile range. Statistical significance was determined using Mann-Whitney test. BeWo n = 9. HPMVEC n = 3. 85-101 bp. Initial quality control for post-sequencing reads, read alignment, and read count generation were all performed in the public Galaxy Bioinformatic server (Boekel et al., 2015) utilizing the following tools: FASTQC (Andrews et al., 2010), trimmomatic (Bolger et al., 2014), Bowtie2 (Langmead & Salzberg, 2012), and featurecounts (Liao et al., 2014). All samples were then aligned utilizing the hg38 genome build via Bowtie2, which allowed for more precise alignments of the numerous homologous genes expressed in these specific cell lines. For each different experiment, gene count matrices were generated using featurecounts and utilized for T A B L E 1 Biological processes impacted by direct knockdown of the HAND1 gene in BeWo cells differential gene expression analyses. Differential gene expression analyses were conducted using the Empirical analysis of digital gene expression in R (EdgeR) package (Robinson et al., 2010). General linear modeling using the following pairwise comparisons were performed between, untreated controls, Allstars negative control treated, direct HAND1 knockdown treated BeWo and HPMVEC cells that were treated directly. Multiple corrections testing yielded no statistical differences in the pairwise comparisons. Therefore, we used the raw pvalues to determine genes to be used in overrepresentation analysis to identify pathways and processes rather than individual genes. RNA sequencing data have been deposited to NCBI GEO under the accession GSE209620.

| Panther DB evaluation
Lists of significantly differentially expressed genes identified between Allstar negative control and HAND1 siRNA treated BeWos and HPMVECs were analyzed by PantherDB (Panther15.0) to determine over-representation and identify pathways and processes involved in trophoblast-endothelial cell signaling. Gene names were submitted with statistical testing conducted using Fisher's Exact test with multiple corrections testing via Bonferroni correction. We conducted analyses using Panther pathways, Reactome pathways, and GO Biological Processes against the entire genome for Homo sapiens.

| In silico StringDB assessment of interaction networks
StringDB (version 11.0) was utilized to assess potential protein interactions affected by knockdown of HAND1 in BeWo cells. Seven significantly differentially expressed genes with large fold-changes were individually entered into StringDB and then functional interactions classified into biological pathways. Parameters were set at Homo Sapiens, Experimental and Database sources, Full Network Search.

| Statistical analysis
qPCR data were analyzed in Prism v8 (GraphPad) using either Kruskal-Wallis test with Dunn's multiple comparison test or Mann-Whitney test. Data for qPCR is presented as the median ± interquartile range.

BeWo cells resulted in changes to global gene expression
Compared to Allstar negative control, HAND1 knockdown in BeWo cells resulted in downregulation of 664 genes, and upregulation of 59 genes (Data S1). PantherDB was utilized to perform overrepresentation analysis against Panther pathways, Reactome pathways, and Gene Ontology (GO) Biological processes on the differentially expressed genes. No overrepresentation was seen compared to Panther and Reactome pathways, however many results were returned for GO Biological processes (Table S2). Groups of genes identified as significantly overrepresented in GO Biological processes were then reentered into PantherDB and indicated potential disruption to pathways including cell development, cellular projection, regulation and establishment of localization, and regulation of multicellular function (Table 1). There were several biological pathways over-represented, including GnRH releasing hormone pathways, cardiac conduction and signaling, TGF-beta signaling, and signaling of RHO GTPases. In addition, there was significant enrichment in MET activating PTK2 signaling, and pathways related to signal transduction. StringDB network analysis of seven genes with large (>2 or < −5) fold change differences in expression following HAND1 knockdown in BeWo cells were found to be involved in biological pathways with known importance in growth and development (Table 2). Upregulated genes were CALML5, NUBP1, TFAP2E, and WNT8A. Downregulated genes; FAM49B, CTTNBP2, and NFS1. Reactome pathways examined identified common relationships between the genes including Beta-catenin phosphorylation, TGF-beta signaling, and Pl3K-Akt signaling. Other notable biological pathways included cardiac conduction and calcium channel signaling, GnRH and Estrogen dependent gene expression, eNOS activation and regulation, and iron-sulfur and sulfur metabolism pathways and RHO GTPases (Figure 2).

HAND1 siRNA minimally disrupted global gene expression
Direct HAND1 knockdown in HPMVECs resulted in minimal disruption to global gene expression with differential F I G U R E 2 Network analyses of genes disrupted by HAND1 knockdown in BeWo. Expression of CALML5 (a), NUBP1 (b), TFAP2E (c), and WNT8A (d) was significantly upregulated in BeWo cells treated with HAND1 siRNA when compared to Allstar negative control. Expression of FAM49B (e), CTTNBP2 (f), and NFS1 (g) was significantly downregulated in BeWo cells treated with HAND1 siRNA when compared to Allstar negative control. Gray lines represent the network edges with thickness representing the confidence of the data support (thicker lines = higher confidence data). Color coded legends show genes in Reactome pathways that may be impacted by HAND1 knockdown. expression in just seven genes (Table 3). QPCR validation on RNA samples independent of those used for RNA sequencing, confirmed two genes of interest, GADD45g and NPPB as reduced and increased, respectively in HAND1 siRNA treated HPMVECs when compared to Allstar negative control (Figure 3).

| DISCUSSION
CHDs are often associated with pregnancy complications such as fetal growth restriction and preeclampsia, these conditions negatively impact clinical outcomes, increase the risk of neonatal morbidity and mortality, and are likely a consequence of abnormal placental development and function (Cnota et al., 2013;Laas et al., 2012;Puri et al., 2018;Ruiz et al., 2016;Tararbit et al., 2018). We have previously shown in mice, that targeted loss of Hand1 in chorionic and labyrinthine progenitor trophoblasts led to abnormal formation of the placental labyrinth, and ultimately embryonic lethality (Courtney et al., 2021). Histological analysis of the placenta indicated that loss of Hand1 in labyrinthine progenitor trophoblasts early in pregnancy significantly impacted the ability for the placenta to form syncytial layers, and impacted development of the labyrinthine vasculature. In this study we aimed to gain further mechanistic, translational understanding of the effects of HAND1 knockdown in models of human placenta trophoblast and villous endothelial cells. We demonstrated significant alterations to placental trophoblast gene expression following HAND1 knockdown, and identified potential pathways which may be significantly impacted by loss of HAND1 regulation. This study is the first to identify possible molecular signaling pathways that are impacted by disruption to HAND1 in the human placenta.
Overall, there were 664 genes differentially expressed in BeWo cells due to HAND1 knockdown. Overrepresentation analyses reveals several key GO Biological Processes including: cell development, establishment of localization, and regulation of multicellular function, as well as biological pathways including Pl3K-Akt signaling, signaling Rho GTPases, and TLR cascades. Trophoblast differentiation during the first trimester of pregnancy involves trophoblast proliferation, invasion, and extracellular matrix (ECM) remodeling. PI3K/Akt signaling reduction or inhibition plays an important role in trophoblast proliferation, migration, and survival. Disruption to PI3K/Akt signaling in early embryonic development is associated with growth restriction, preterm birth, and embryonic lethality (Yu & Cui, 2016), highlighting the importance of this signaling pathway to placental development and function. Additionally, inhibition of PI3K increases soluble fmslike tyrosine kinase 1 (sFlt1), a common biomarker of pre-eclampsia (Park et al., 2010). PI3K/Akt signaling has been closely linked to signaling Rho-GTPases which are known to play a role in trophoblast migration (Duquette & Lamarche-Vane, 2014;Gupta et al., 2016).
TLR cascades form the major family of pattern recognition receptors that are involved in innate immunity. The maternal-fetal interface immunologically is unique in that it must promote tolerance of the fetus while maintaining protection to the mother. Trophoblasts play an important role in modulating the maternal immune response throughout pregnancy, including through TLR signaling (Koga & Mor, 2010). Additionally, TLR signaling has been shown to potentially modulate angiogenesis as culture of trophoblasts with TLR2 ligand HKML have been shown to promote the expression of pro-angiogenic Placenta Growth Factor (Kato et al., 2017). Overall, poor migration of trophoblasts, and communication with resident immune cells, can impact invasion and establishment of a fully functional maternal-fetal interface.
Expression of CALML5 and NUBP1 was upregulated in BeWo cells following HAND1 knockdown. Both genes are involved with Calcium channel signaling, GnRH signaling, cAMP signaling, eNOS activation, and Pl3K-Akt signaling pathways. These are important biological pathways that impact trophoblast invasion, differentiation, development, resource control, and growth of the placenta and fetus (Liu et al., 2009;Moreau et al., 2002), and increased gene expression of CALML5 and NUBP1 may be a compensatory response to disruption of other signaling pathways. On the other hand, expression of CTTNBP2 and NSF1 was downregulated in BeWo cells in which HAND1 was knocked down. CTTNBP2 has been shown to have a direct relationship with the WNT signaling pathway (Kelly et al., 2017), and downregulation in WNT signaling in the placenta has been associated with pathological pregnancies (Zmijanac Partl et al., 2018). Similarly, NFS1 is a gene that has an essential role in iron-sulfur cluster processing making it important for electron transport, enzyme catalysis, and regulation of gene expression as well is iron homeostasis (Brzoska et al., 2006). Fetal growth is very dependent on energy metabolism in the placenta as it drives exchange of nutrients and plays a crucial role in DNA synthesis. Overall, our data indicates potential disruption to these pathways with HAND1 knockdown and provides further understanding of how a genetic perturbation in this gene may lead to growth issues, developmental defects, and lethality/miscarriage in the context of human pregnancies with CHDs. We sought to analyze the effect of HAND1 knockdown in cells within the villous environment. Interestingly, direct HAND1 knockdown in villous endothelial cells resulted in minimal impact to gene expression. This result however, is in agreement with our mouse model studies suggest that disrupting Hand1 expression directly in labyrinthine endothelial cells impacted vascular remodeling only in late pregnancy and non-branching angiogenesis mechanisms (Courtney et al., 2021) not as individual endothelial cells or vasculogenesis. HPMVECs are cultured as a single monolayer. Therefore, it would be interesting for future studies, beyond the scope of the current study, to assess angiogenesis and remodeling mechanisms in a 3D vascularized model when HAND1 is knocked down or cultured in "conditioned" media from BeWo cells treated with HAND1 siRNA. Cell-cell communication/signaling within the placenta villi in the human is believed to be important in establishment of the villous structure and exchange region but given our current data, the involvement of other cell types such as stromal fibroblasts in the communication process requires future investigation.
We and others have consistently shown that HAND1 is important to both fetal heart and placenta development (Courtney et al., 2021;Firulli et al., 2010), with the present study providing further mechanistic understanding of how HAND1 may influence the development of the placenta in the human. Given our data shows greater disruption to global gene expression in placenta trophoblasts then endothelial cells with HAND1 knockdown, this further highlights the importance of future research to consider analyzing the extra-embryonic tissue, as well as the heart, in the context of CHD.

AUTHOR CONTRIBUTIONS
RF conceived the study, performed experiments, analyzed the data, and wrote the manuscript. JC performed experiments, analyzed data, and edited manuscript. HB performed bioinformatic analysis and interpretation, and edited manuscript. RLW analyzed data and wrote manuscript. HNJ obtained funding, conceived study, and edited the manuscript. All authors approved final version of manuscript.

FUNDING INFORMATION
This study was funded by Eunice Kennedy Shriver National Institute of Child Health and Human Development award R01HD091527 (HNJ).

CONFLICT OF INTEREST
The authors report no conflict of interest. This article was prepared while Dr. Brockway was employed University of Florida. The opinions expressed in this article are the author's own and do not reflect the view of the National Institutes of Health, the Department F I G U R E 3 qPCR validation of two genes shown to be differentially expressed using RNA sequencing. (a) mRNA expression of GADD45G was shown to be reduced in HAND1 siRNA treated human placenta microvascular endothelial cells (HPMVECs) when compared to Allstar negative control treated cells. (b) Expression of NPPB was increased in in HAND1 siRNA treated HPMVECs when compared to Allstar negative control treated cells. Data are median ± interquartile range. Statistical significance was determined using Mann-Whitney test. n = 3 passages.
of Health and Human Services, or the United States government.

DATA AVAILABILITY STATEMENT
RNA sequencing data have been made publicly available in NCBI GEO, accession number: GSE209620. All other data can be provided upon request.