Abnormal expression of SLIT3 induces intravillous vascularization dysplasia in ectopic pregnancy

Introduction

Ectopic pregnancy (EP), in which a fertilized ovum is implanted outside of the uterine cavity, accounts for 1–2% of all human pregnancies. Ninety-eight percent of EPs occur in the fallopian tube, a phenomenon known as tubal EP (Shaw et al., 2010). Hemorrhaging as a result of EP is a major cause of maternal mortality in the first trimester, accounting for ∼10% of all pregnancy-related deaths (Farquhar, 2005).

The placental villi at the maternal-fetal interface are the bridge between mother and fetus and are therefore extremely important in embryo implantation and development (Cakmak & Taylor, 2011; Kokawa, Shikone & Nakano, 1998). Structural abnormalities or dysfunction of the villi have been implicated in pathological pregnancies, including preeclampsia, fetal growth restriction (FGR), and recurrent miscarriage (Tian et al., 2016). In tubal EP, the epithelial cells of the fallopian tube cannot undergo adequate decidualization to produce an appropriate microenvironment to support embryonic growth (Liu et al., 2020b), and most EP villi will be dysplastic or end with abortion (Cunningham et al., 2008). In a previous study, we showed that EP villi exhibit excessive oxidative stress, decreased mitochondrial DNA copy number, and mitochondrial dysfunction compared to normal pregnancy (NP) villi (Huang et al., 2020). Because mitochondria are the intracellular “energy factories”, mitochondrial dysfunction leads to metabolic dysfunction, triggers embryonic apoptosis, and prevents embryonic development (Guerin, Mouatassim & Menezo, 2001). Taken together, this evidence suggests that EP villi are dysplastic compared to NP villi, but the exact mechanism involved remains unclear.

Angiogenesis and vasculogenesis in placental villi are essential for the growth of trophoblast cells and the maintenance of NP because there is diffusion between the outer surface of the villi trophoblast and the inner surface of the villi vascular endothelium (Burton et al., 2009). Various factors, such as vascular endothelial growth factor (VEGF), acidic and basic fibroblast growth factor (aFGF and bFGF, respectively), and epidermal growth factor (EGF), reportedly affect angiogenesis and vasculogenesis through autocrine or paracrine mechanisms; these factors directly or indirectly stimulate endothelial precursor cell differentiation and proliferation, influencing angiogenesis and vasculogenesis and affecting the pregnancy outcome (Demir, Seval & Huppertz, 2007). However, it is unclear whether abnormalities in angiogenesis are responsible for dysplasia of EP villi.

To address these questions, we here compared villus morphology and capillary number between villi from EP and NP patients. Furthermore, we compared the transcriptome expression profiles of the two types of villi and identified key factors influencing EP villi angiogenesis through biological information analysis and experimental validation. This study provides new insights into the molecular mechanism underlying villous dysplasia in the tubal EP environment.

Material and Methods

Clinical tissue sample collection

The experimental workflow used in this study is shown in Fig. 1. This investigation was approved by the Institutional Ethics Committee of the International Peace Maternity and Child Health Hospital (IPMCH) in Shanghai, China (GKLW201909). Written informed consent was obtained from all enrolled patients. The study was conducted between May 2020 and March 2021. Villous tissues samples were collected from women undergoing laparoscopic salpingectomy for tubal EP and women electively terminating clinical NP for nonmedical reasons. All participants had singleton pregnancies and regular menstrual cycles, and were between six and eight gestational weeks as confirmed from the time of the last menstrual period. Tissue samples were collected immediately after surgical removal of the villi, stored in liquid nitrogen or 10% formalin, and transported from the operating theatre to the laboratory within 15 min. Voluntarily terminated NPs were confirmed by ultrasonography combined with serum levels of beta human chorionic gonadotropin (β-HCG). The diagnosis of tubal EP was first made by ultrasound combined with serum β-HCG, then confirmed and treated via laparoscopic salpingectomy. All EP samples were obtained without methotrexate treatment. Exclusion criteria included smoking, abnormalities in vital organ function or metabolic function (e.g., diabetes or obesity), hypertension or other cardiovascular pathologies, acute or chronic illness, NP with a previous history of abnormal pregnancy (e.g., preeclampsia or recurrent miscarriage), EP with obvious tubal inflammatory adhesions, previous fallopian tubal diseases, or a history of tubal surgery. Detailed baseline characteristics of the samples are provided in Table S1.

Results

Differences in morphology and capillary number between EP and NP villi

Differences between EP and NP villous capillaries were uncovered by HE staining of tissue sections (Fig. 2A); EP samples had fewer red cells and capillaries. Furthermore, IHC staining of the vascular endothelial cell marker gene CD31 confirmed the HE staining results. Compared with NP, EP placental villi exhibited higher mean and total cross-sectional areas (Fig. 2B). Nevertheless, the relative proportions of proliferating (KI67+TEAD4+) and resting (KI67-TEAD4+) cytotrophoblasts (CTBs) were similar in the EP and NP groups, implying that stem cell population proliferation was also comparable between the two types of placental villi (Liu et al., 2018) (Fig. S2A). EP villi also contained many fewer capillaries per unit area than NP villi (Figs. 2A, 2B), and the vascular number was strongly associated with β-HCG levels (r = 0.40, p = 0.02) (Fig. 2C). Taken together, these observations indicated that EP villi were more poorly developed than NP villi at comparable gestational ages.

Differences in miRNA and mRNA profiles between EP and NP villi

Ten of the villous samples (five EP and five NP) were selected for miRNA- and mRNA-seq analyses. Selections were made on the basis of RNA quality, copy number variation (CNV), and lack of maternal contamination (Table S4). There were 49 significant DE-miRNAs, including 30 up-regulated and 19 down-regulated miRNAs in tubal EP villous samples (Figs. 3A, 3C, Table S5), with a predicted total of 5939 putative target genes. mRNA-seq results revealed 342 up-regulated and 283 down-regulated genes in EP compared to NP villi (Table S6). Normalized counts and log₂(FoldChange) values were visualized with a volcano plot (Fig. 3B). Expression profiles of DE-mRNAs in the two groups were also evaluated with hierarchical clustering (Fig. 3D).

Construction of a miRNA–mRNA network

Based on the principle of mRNA and miRNA pairing, 32 up-regulated DE-mRNAs (Fig. 4A) and 71 down-regulated DE-mRNAs were visualized in a Venn diagram (Fig. 4B). The 103 overlapping genes were analyzed for enrichment of Gene Ontology (GO) terms (Fig. 4D) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways (Fig. S1) to understand their biological functions in EP. The down-regulated genes were significantly enriched in the following biological processes (BPs): collagen catabolic process, extracellular matrix (ECM) organization, cell adhesion, Roundabout (Robo) signaling pathway, and angiogenesis. The up-regulated genes were enriched in BPs including positive regulation of cell proliferation involved in kidney development, regulation of small GTPase mediated signal transduction, and positive regulation of intrinsic apoptotic signaling pathway. In addition to the GO term of angiogenesis, the Robo signaling pathway has also been reported to be associated with angiogenesis and placenta vascular remodeling (Li et al., 2015; Liao et al., 2012; Zhou et al., 2011). These results illustrated that pro-angiogenesis genes (e.g., SLIT3 and TAL1) were down-regulated in EP villi (Hu et al., 2016; Lazrak et al., 2004; Lelievre et al., 2001; Li et al., 2015; Liao et al., 2012; Zhou et al., 2011), consistent with the observed low capillary number. In addition, in agreement with dysplasia of EP villi compared to NP villi, we also found that pro-apoptotic genes were up-regulated in EP samples. The 103 overlapping genes were regulated by a total of 32 DE-miRNAs. A regulatory network of the 103 mRNAs and 32 miRNAs was constructed (Fig. 4C).

Validation of hub mRNAs associated with angiogenesis

Based on the GO term analysis and a review of the literature, we found that four of the 103 differentially expressed genes were associated with angiogenesis, namely HSPB6 (Zhang et al., 2012), TAL1 (Hu et al., 2016), PDZRN3 (Sewduth et al., 2017), and SLIT3 (Zhou et al., 2011) (Fig. 4D). To verify the accuracy of the RNA-seq results, expression levels of these four hub mRNAs were analyzed via qRT-PCR in the eight EP and eight NP samples. SLIT3 expression was significantly lower in the EP group than in the NP group (Fig. 5A). However, there were no significant differences in expression levels of the remaining three hub mRNAs between the two groups (Fig. 5A). IHC was performed to explore the cellular localization of SLIT3 in villous tissue and to further verify differences in expression between EP and NP (Fig. 5B). IHC scores for SLIT3 were significantly lower in patients with EP than in the NP controls (Fig. 5B), consistent with the RNA-seq results. Taken together, the results of the RNA-seq bioinformatic analysis, qRT-PCR validation, and IHC confirmed that SLIT3 may be involved in regulating angiogenesis of EP villi. Next, we validated the target gene ROBO4 (Xiao et al., 2022), which is downstream of SLIT3 and reportedly associated with angiogenesis, via qRT-PCR. This assay indicated that relative expression of ROBO4 was statistically lower in EP than in NP villous tissue (p = 0.007) (Fig. S4).

Validation of hub miRNAs targeting SLIT3

From the miRNA-mRNA regulatory network, four candidate miRNAs were found to potentially target SLIT3, namely miR-491-5p, miR-34a-5p, miR-34c-5p, and novel_829 (Fig. 4C). qRT-PCR showed that miRNA-491-5p and miRNA-34a-5p expression levels were significantly higher in EP villi than in the NP group. No significant differences in expression levels of miRNA-34c-5p or novel-829 were observed between the two groups (Fig. 6A). To verify whether these miRNAs directly inhibited SLIT3, miR-491-5p and miR-34a-5p were each overexpressed in the HTR-8 cell line, then SLIT3 expression was measured. As expected, miR-491-5p and miR-34a-5p were significantly up-regulated in cells transfected with the miR-491-5p and miR-34a-5p overexpression vectors, respectively (Fig. 6B). Overexpression of miR-491-5p resulted in a significant decrease in SLIT3 expression (Fig. 6B), whereas miR-34a-5p overexpression was not associated with a significant change in SLIT3 expression (Fig. 6B). Furthermore, a dual-luciferase reporter assay showed that miR-491-5p could bind the 3′ UTR of SLIT3 mRNA (Fig. S3A). Simultaneously, a significant increase in SLIT3 expression after transfection with the miR-491-5p mimic was revealed in an RNA-pull down assay (Fig. S3B), further suggesting that miR-491-5p could directly target and regulate SLIT3.

Discussion

In this study, we found that the mean and total cross-sectional areas of placental villi were significantly increased in EP patients compared to NP patients. The number of capillaries per unit area was significantly reduced in EP villi compared with NP controls, and was positively correlated with β-HCG levels. miRNA-491-5p-SLIT3 was also identified as involved in regulation of villous angiogenesis. Moreover, a positive correlation was confirmed between relative SLIT3 expression and vascular number.

Placental villous capillaries play a vital role in the transfer of nutrients, oxygen, and metabolites between maternal and fetal blood (Jirkovska et al., 2012). Compared to decidual angiogenesis, less is known about villous angiogenesis, which is a critical factor in functional placental formation and embryonic growth. It has previously been reported that the branching patterns of villous capillaries and structural alterations in placental villi are associated with gestational diabetes mellitus, preeclampsia, and FGR (Egbor et al., 2006; Jirkovska et al., 2012). Goddijn et al. (2000) indicated that there was no association between EP villous capillaries and chromosomal abnormalities. However, no previous studies have compared differences in villus morphology and capillary number between EP and NP samples.

Our results demonstrated that, compared to NP villi, EP villi tended to have a larger mean cross-sectional area, but fewer capillaries per unit area, and lower β-HCG levels. This indicated that EP is associated with a low potential to enlarge the villous surface and the vascular network that are key in maternofetal transport and embryonic growth. Unlike the endometrium, the epithelium of the fallopian tubes is poorly decidualized, which affects production and secretion of VEGF and vascular establishment of the embryo (Nowacek et al., 1999; Torry et al., 1996; Torry & Torry, 1997). The increase in villus cross-sectional area may be a compensation for the poor environment in which the embryo grows.

SLITs comprise a highly conserved family of secreted proteins that were originally discovered in the nervous system (Zhou et al., 2013a). Three SLIT genes (SLIT1, SLIT2, and SLIT3) have been identified in mammals. Slit signaling plays a vital role in axon guidance (Yu et al., 2014), angiogenesis (Zhou et al., 2011), and cell migration (Qi et al., 2014) via the Robo receptor. SLIT/ROBO systems have also been found in a variety of non-neuronal tissues, including the lung (Anselmo et al., 2003), kidney (Piper et al., 2000), ovary (Dickinson, Myers & Duncan, 2008), and placenta (Li et al., 2015; Liao et al., 2010). In those with preeclampsia, the SLIT/ROBO system is abnormally expressed in placental vascular endothelial cells and trophoblast cells, affecting placental development and function by altering the function of trophoblast cells and endothelial cells (Liao et al., 2012). These findings suggest that SLIT/ROBO signaling plays a vital role in placental angiogenesis and function during NP. Another study revealed that SLIT2 is expressed in extravillous trophoblasts (EVTs) of tubal pregnancy, and that changes in SLIT2 expression are linked to vascular remodeling of the fallopian tube (Li et al., 2015). SLIT3 is a new angiogenic factor that binds to ROBO4, promoting endothelial cell proliferation and migration and inducing angiogenesis in vivo (Xiao et al., 2022; Zhang et al., 2009). Moreover, expression of ROBO4 in placental arteries and veins (Huminiecki et al., 2002) and of SLIT3 in the placenta (Dickinson et al., 2004) have been previously reported. Our results revealed that relative expression levels of both SLIT3 and ROBO4 were decreased in EP villous tissue compared to NP, and that SLIT3 did promote angiogenesis, consistent with the poor angiogenesis of EP villi. EP is known to have faulty decidualization compared to NP, and it is uncertain whether altered SLIT3 expression in EP villi is associated with poor tubal decidualization.

Interestingly, SLIT3 expression was positively correlated with the number of villous capillaries and levels of β-HCG, and β-HCG levels were positively correlated with the number of villous capillaries. This confirmed our hypothesis that SLIT3 may be associated with the poor villous angiogenesis of tubal EP. It is unclear how SLIT3 regulates β-HCG levels, due to a lack of evidence in the literature, but our results led us to a hypothesis. In the human placenta, there are three major trophoblast subpopulations: CTBs, EVTs, and syncytiotrophoblast (STBs); of these three types, β-HCG is secreted only by STBs (Handschuh et al., 2009). Because early gestational trophoblasts are supplied with nutrients by the intravillous capillaries, we hypothesize that SLIT3 promotes trophoblast growth by enhancing villous capillary growth, which in turn increases β-HCG secretion. However, this hypothesis requires further verification. In addition, higher β-HCG levels (a result of high trophoblast activity) can increase the risk of tubal pregnancy rupture and decrease the likelihood of conservative treatment (Fukami et al., 2016). SLIT3 expression in the villous tissue is positively correlated with β-HCG levels, implying that high SLIT3 expression in EP villous tissue may be associated with poor prognosis of tubal pregnancy, possibly through promotion of angiogenesis in the villi. Thus, SLIT3 is a promising prognostic factor for tubal pregnancy and a potential therapeutic target for reducing the risk of EP.

In cancer cells, miR-491-5p induces apoptosis and suppresses proliferation and invasion (Denoyelle et al., 2014; Zhou et al., 2013b); in lung cancer, it restrains angiogenesis (Yang et al., 2021). miR-491-5p is also reportedly expressed in placental tissue and has potential roles in trophoblast proliferation, invasion, and angiogenesis (Liu et al., 2020a). Compared with normal placental tissue, miR-491-5p is upregulated in placental tissue collected from patients with preeclampsia, and overexpression suppresses trophoblast invasion and vascular remodeling (Liu et al., 2020a). Similarly, in the present study, miR-491-5p was expressed at higher levels in villous tissues from women with tubal EP compared to those with NP. Furthermore, miR-491-5p was found to suppress expression of SLIT3. We therefore speculate that miR-491-5p affected villous angiogenesis by targeting SLIT3 in villi, which may be associated with the mechanism underlying villous dysplasia in the fallopian tube environment. This hypothesis requires confirmation in further functional experiments.

There are several limitations of this study that should be noted. First, the sample size was relatively small. The correlation between SLIT3 and villous capillaries will be further confirmed by expanding the sample size in the future, with the aim of providing new biomarkers for predicting abnormalities in villi and villous vascular development. Additionally, although HTR8/SVneo is widely used as a model to study the invasion, migration, metabolism and other functions of placental trophoblast cells (Basak et al., 2018; Tan et al., 2022; Yin et al., 2022), it cannot fully reflect the real situation of trophoblast cells in vivo due to its own limitations; therefore, primary trophoblast cells should be used to further validated the findings of this study in the future. Furthermore, the specific molecular mechanisms by which miR-491-5p and SLIT3 regulate EP have not been fully investigated. Validation and functional experiments with larger sample sizes are urgently needed to elucidate the pathological mechanisms of villous dysplasia in EP.

Conclusion

Our results revealed that the morphology, capillary density, and transcriptome expression profiles of EP villi differed from those of NP villi. These changes may have been related to the biological roles of miR-491-5p and SLIT3. In addition, SLIT3 was identified as a putative predictor of abnormal chorionic villus development. However, its true functional significance requires further verification.

Supplemental Information