Analysis of gene expression difference and biological process in chorionic villi of unexplained recurrent spontaneous abortion

Abstract

Objective

To explore the biological relationship between the regulatory signal pathways involved in differentially expressed genes and recurrent spontaneous abortion (RSA) by analyzing the gene expression microarray data of unexplained RSA.

Methods

The gene expression profile data of chorionic villi from unexplained recurrent abortion with normal karyotype and selective induced abortion were compared. Differentially expressed genes were analyzed by the “Limma” package in R Studio, and Gene Ontology(GO) and Kyoto Encyclopedia of Genes and Genomes(KEGG) pathway enrichment analysis were carried out with “Cluster Profiler” and “org.hs.eg.db” packages. Finally, hub genes were identified through constructing the protein-protein interaction (PPI) network from the differentially expressed gene dataset in the STRING database. And the hub genes were verified by RT-PCR. The expression of TH1 and TH2 cytokines representing IL-2, IL-10 and their receptors related to hub gene immune regulation were detected by enzyme-linked immunosorbent assay (ELISA) and western blot (WB), respectively.

Results

A total of 295 differentially expressed genes were identified in the dataset GSE22490, with a significance level of P < 0.05 and an absolute log-fold change > 1.0, which included 166 up-regulated genes and 129 down-regulated genes. Go and KEGG enrichment analysis of these differentially expressed genes (P < 0.05,FDR < 0.05) revealed significant involvement in the regulation of inflammatory and immune responses. The PPI analysis revealed that the hub genes FCGR3A, TLR2, BTK, CLEC7A and CD163 were centrally located in the network cluster which were composed of the proteins encoded by differentially expressed genes associated with RSA. The mRNA levels of FCGR3A, TLR2 and CLEC7A in the RSA group were significantly higher than those in the NC group (P < 0.05). The protein expression level of TLR2 was also significantly increased in the RSA group (P < 0.05). The level of IL-2 in the RSA group was significantly higher than that in the NC group (P < 0.05), while the protein expression level of its receptor was not different(P > 0.05). There was no significant difference in the expression levels of IL-10 and its receptor between the two groups (P > 0.05).

Conclusion

Abnormal immune response plays an important role in unexplained RSA. The imbalance in immune regulation may be one of the most important reasons behind this phenomenon. These findings provide a foundation for further research into the mechanisms underlying RSA.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12884-024-07099-2.

Background

Recurrent spontaneous abortion(RSA) is defined as three or more consecutive natural abortions with the same sexual partner and is a serious reproductive disorder. The incidence rate of RSA is approximately 3 -5% among women of childbearing age and has been increasing year by year in recent years. Previous studies reveal that there are numerous factors that can lead to RSA, including advanced maternal age, genetic factors, anatomical abnormalities, infections, endocrine disorders, and so on [1]. However, in the majority of patients, the cause remains unknown. This study analyzed gene expression data from early placental villous tissue of RSA patients to detect changes in gene expression that occur in RSA embryos during early pregnancy and to explore the corresponding biological mechanisms. This evidence can provide a scientific basis for the study of RSA pathogenesis.

Materials and methods

Data acquisition

The mRNA expression microarray dataset GSE22490 was retrieved from the Gene Expression Omnibus (GEO) database, which is housed within the National Center for Biotechnology Information (NCBI) in the United States. The dataset was accessed utilizing the “GEOquery” package in R Studio. This dataset was derived from mRNA chip data of villous tissue samples taken from 4 RSA patients and 6 individuals with normal fertility who underwent elective abortions, based on the GPL570 platform. The villous tissue was collected from individuals with clinically normal pregnancies (age range 22 to 35, mean age 28 ± 3 years; gestational age at sampling was 52 ± 9 days, mean ± standard deviation), which were terminated for non-medical reasons. Additionally, we collected villus samples from RSA patients (aged 22 to 35, mean age 28 ± 4 years, with a gestational age of 53 ± 7 days at the time of abortion) who had experienced three or more consecutive unexplained spontaneous abortions.

Patients were excluded from the study if they had concurrent health issues, including: (1) chromosomal abnormalities, by performing a low-density microarray on the villi to rule out chromosomal abnormalities; (2) autoimmune diseases, such as Sjögren’s syndrome, antiphospholipid antibody syndrome, systemic lupus erythematosus, etc., by detecting immune markers; (3) infections, through monitoring body temperature and blood tests; (4) anatomical abnormalities, including adenomyosis, uterine fibroids, cervical incompetence, congenital malformations of the uterus, intrauterine adhesions, etc.; (5) endocrine or metabolic diseases, such as hyperthyroidism, hypothyroidism, hyperprolactinemia, polycystic ovary syndrome, diabetes mellitus, etc., by assessing thyroid function, hormone levels, and blood glucose levels. These criteria were applied to eliminate known causes or risk factors of RSA. However, there may still be some unidentified confounding factors that could lead to confusing confounding biases.

Differential gene analysis

The dataset was analyzed using the “Limma” package in R Studio to identify differentially expressed genes with a P-value < 0.05 and an absolute log-fold change > 1.0. The Fold Change (FC) reflected the ratio of gene expression in the villous tissue, where positive values indicated upregulation and negative values indicated downregulation. Heat map and volcano plot were generated to visualize the differentially expressed genes that were identified through this screening process.

Gene enrichment analysis

The Gene Ontology (GO) database is a repository established by the Gene Ontology Consortium, which characterized the roles of genes and proteins in cells by creating a dynamic, controlled vocabulary to comprehensively describe the attributes of genes and gene products in organisms. The GO database is divided into three main categories: Biological Process (BP), Cellular Component (CC), and Molecular Function (MF), which describe the molecular functions that gene products may perform, the cellular environments they are in, and the biological processes they are involved in. The Kyoto Encyclopedia of Genes and Genomes (KEGG) database integrates genomic, chemical, and systemic functional information, linking gene catalogs obtained from fully sequenced genomes to higher-level cellular, species, and ecosystem-level systemic functions. We performed GO and KEGG enrichment analyses on the differential genes using the “clusterProfiler” and “org.Hs.eg.db” packages in R Studio. A P-value < 0.05 and False Discovery Rate (FDR) < 0.05 indicate significant enrichment, which is considered statistically significant.

Protein interaction network analysis

The STRING database was employed to construct a network diagram illustrating the interactions among proteins encoded by the differentially expressed genes. Through this process, key proteins and their respective coding genes were identified and obtained.

Enzyme-linked immunosorbent assay(ELISA) detection

Villous tissue was collected from 14 RSA patients, designated as the RSA group, and 15 individuals with normal fertility undergoing elective abortions, known as the NC group, in the family planning clinic of obstetrics and gynecology department of the First Affiliated Hospital of Soochow University between January 2022 and August 2022. The inclusion criteria and case grouping adhered to the specifications mentioned previously. Following tissue homogenization, the supernatant was centrifuged, and an ELISA kit from Jiangsu Enzyme Immunoassay Company, China, was utilized to detect the levels of TH1 and TH2 cytokines, including IL-2 and IL-10, in accordance with the manufacturer’s instructions.

RT-PCR detection

According to the instructions provided with the kit, RNA was extracted from tissue homogenates using the RNeasy Mini kit supplied by Qiagen, Germany. Subsequently, a reverse transcription kit from Takara, Japan, was employed to convert the RNA into complementary DNA (cDNA). Real-time quantitative PCR was conducted using a SYBR Green kit also from Takara, Japan, in conjunction with the StepOnePlus system, which was procured from Thermo Fisher Scientific, USA. GAPDH served as the internal reference control for the assay. Each sample was subjected to the testing procedure on three separate occasions. The sequences of the primers utilized in the study were presented in Table 1.

Table 1.

Primer sequences

Target gene	Primer direction	Base sequence	Fragment size
GAPDH	F	5’-TCGGAGTCAACGGATTTGGTC-3’	146 bp
	R	5’-GCCATGGGTGGAATCATATTGG-3’
FCGR3A	F	5’-GCAGCTAGAAGTCCATATCGG-3’	148 bp
	R	5’-CTTCCTGCCTTTGCCATTCTG-3’
TLR2	F	5’ -CTCGGAGTTCTCCCAGTGTT-3’	99 bp
	R	5’ -TGTCTTCCTGCCTTCACTTGG-3’
CLEC7A	F	5’ -TGGGTACCATGGGGGTTCTT-3’	127 bp
	R	5’ -CCCAGTTGCCAGCATTGTCT-3’

Western blot(WB) detection

Villous tissue was ground and then dissolved in radioimmunoprecipitation assay (RIPA) buffer. The protein concentration was determined, followed by the separation of the protein through 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The separated proteins were transferred onto a membrane, which was then blocked with a serum blocking solution. The membrane was incubated with primary antibodies against TLR2 (1:1000, Affinity, USA), IL-2R (1:400, abcam, UK), IL-10R (1:1500, Affinity, USA), and GAPDH (1:10,000, abcam, USA) at 4 °C overnight. After three washes, the membrane was incubated with a secondary antibody at room temperature for one hour. The proteins were subsequently detected using a super enhancer chemiluminescence (ECL) kit from Absin, China. The results were analyzed with Image J software.

Statistical analysis

The results were presented as the mean ± standard deviation. T-tests between groups for ELISA, RT-PCR, and WB analyses were conducted using GraphPad 9.0, with a P-value < 0.05 being considered statistically significant. Gene differentiation analysis was carried out with the “Limma” package for Benjamini & Hochberg corrected t-tests, where a P-value < 0.05 indicated statistically significant differences.

Results

Results of differential gene expression analysis

A total of 295 differentially expressed genes were identified in the dataset GSE22490, with a significance level of P < 0.05 and an absolute log-fold change > 1.0, which included 166 up-regulated genes and 129 down-regulated genes. The heat map and volcano plot illustrating the differential gene expression were displayed in Fig. 1A&B, respectively.

Fig. 1.

Differential gene screening results. (A) Heat map of top 50 differential gene expression in dataset GSE22490; Note: Red stripes and blue stripes referred to the up-regulated and down-regulated genes, respectively. (B) Volcano plot of differential gene expression in dataset GSE22490; Note: Red dots and green dots referred to the up-regulated and down-regulated genes, respectively

Among the up-regulated genes, the top ten, ranked by the significance of their differential expression, were as follows: KCNJ11 (P = 0.0000), AQP9 (P = 0.0001), NPY (P = 0.0003), BCL2A1 (P = 0.0005), FLJ41455 (P = 0.0007), IFIT2 (P = 0.0010), BTK (P = 0.0011), LILRA2 (P = 0.0012), CXCL8 (P = 0.0012), and FAM26F (P = 0.0014). Similarly, among the down-regulated genes, the top ten, based on the significance of their differential expression, were: SCN4B (P = 0.0001), PARD3-AS1 (P = 0.0002), RP3-406C18.2 (P = 0.0002), MUC13 (P = 0.0003), CACNG1 (P = 0.0005), C15orf32 (P = 0.0007), PACRG-AS1 (P = 0.0010), USP29 (P = 0.0010), VIP (P = 0.0017), and LRIT1 (P = 0.0019).

Enrichment analysis results of differential genes

The GO analysis enriched the 295 differentially expressed genes, leading to the enrichment of BP, CC and MF categories. In the BP category, the significantly enriched genes were primarily associated with inflammatory and immune responses, particularly those involving immune response pathways that include white blood cells. For the CC category, the significantly enriched genes mainly pertained to components of the cell plasma membrane and vesicles, encompassing both whole and parts of these cellular structures. In the MF category, the significantly enriched genes predominantly involved the activation of cell membranes and transport receptors, with a specific emphasis on the activation of lipopolysaccharide receptors and Toll-like receptors (Fig. 2A-C).

Fig. 2.

Differential gene enrichment results. (A) GO-BP set of differential genes in dataset GSE22490. (B) GO-CC set of differential genes in dataset GSE22490. (C) GO-MF set of differential genes in dataset GSE22490. (D) KEGG enrichment results of differential genes in dataset GSE22490

The KEGG analysis enriched the 295 differentially expressed genes, which were significantly associated with several biological pathways at a P-value < 0.05. These pathways included the NFκB signaling pathway (P = 0.0148), neuroactive ligand-receptor interaction (P = 0.0258), interaction between viral proteins and cytokines and cytokine receptors (P = 0.0336), amoebiasis (P = 0.0148), leishmaniasis (P = 0.0219), rheumatoid arthritis (P = 0.0313), Staphylococcus aureus infection (P = 0.0316), osteoclast differentiation (P = 0.0026), IL-17 signaling pathway (P = 0.0313), and the Toll-like receptor pathway (P = 0.0359) (Fig. 2D).

PPI network analysis and mRNA validation of hub genes

A total of 295 differentially expressed genes were imported into the STRING database to construct the PPI network diagram (Fig. 3A). The analysis suggested that the hub genes FCGR3A, TLR2, BTK, CLEC7A and CD163 were centrally located in the network cluster which were composed of the proteins encoded by differentially expressed genes associated with RSA.

Fig. 3.

Analysis results of differential gene PPI network and validation of hub gene mRNA levels. (A) PPI network diagram of differential genes in dataset GSE22490. (B) Validation of hub gene mRNA levels in chorionic villi. NC, patients with normal fertility undergoing elective abortions; RSA, patients with unexplained recurrent spontaneous abortion. *P < 0.05

RT-PCR was conducted to validate the differential expression of hub genes. The relative expression levels of the hub genes FCGR3A, TLR2, and CLEC7A in the RSA group were found to be (1.35 ± 0.27), (2.38 ± 0.31), and (0.99 ± 0.21), respectively. These levels were significantly higher compared to those in the NC group, which were (0.94 ± 0.23), (1.18 ± 0.24), and (0.65 ± 0.21), respectively. Statistical analysis revealed significant differences (FCGR3A: t = 4.518, P = 0.0001; TLR2: t = 11.620, P = 0.0000; CLEC7A: t = 4.285, P = 0.0002) (Fig. 3B).

Cytokine and receptor protein expression in villous tissue: correlation with hub genes

The levels of IL-2 and IL-10, as representatives of TH1 and TH2 cytokines, respectively, in the tissue homogenate were determined. The IL-2 levels were found to be significantly elevated in the RSA group (3.40 ± 0.55) compared to the NC group (2.72 ± 0.44) (t = 3.64, P = 0.0011). In contrast, no significant difference was observed in IL-10 levels between the RSA group (9.57 ± 1.03) and the NC group (10.16 ± 0.84) (t = 1.686, P = 0.1032) (Fig. 4A).

Fig. 4.

Cytokine levels and receptor protein expression levels related to core genes in villous tissue. (A) Representatives of TH1 and TH2 cytokines levels in chorionic villi. (B) Expressions of receptor protein in chorionic villi. NC, patients with normal fertility undergoing elective abortions; RSA, patients with unexplained recurrent spontaneous abortion. *P < 0.05

A significant increase in the protein expression of TLR2 in the villous tissue of the RSA group was illustrated in Fig. 4B (RSA vs. NC: 0.17 ± 0.03 vs. 0.04 ± 0.02, t = 13.680, P = 0.0000), which was consistent with the observed mRNA levels (Supplementary Figs. 1–4). This correlation suggested that the elevated protein expression of TLR2 may be due to enhanced transcription. Furthermore, no significant differences in protein expression were detected for IL-2R and IL-10R between the RSA and NC groups (IL-2R: RSA vs. NC, 0.08 ± 0.03 vs. 0.06 ± 0.02, t = 1.573, P = 0.1352; IL-10R: RSA vs. NC, 0.09 ± 0.03 vs. 0.08 ± 0.02, t = 0.307, P = 0.7625).

Discussion

Bioinformatics effectively integrates biology with mathematics and computer science, leveraging comprehensive methods and tools from various disciplines including mathematics and information science to acquire, process, store, analyze, and interpret biological data. This approach aims to uncover the biological insights embedded within vast quantities of biological information. The GEO database is a gene expression database created by the National Center for Biotechnology Information (NCBI) in the United States in 2000, which includes gene expression data submitted by research institutions around the world, mainly including gene chips and high-throughput sequencing data. This study is based on the expression profiling dataset of villous tissue from RSA patients in the GEO database. Bioinformatics technology is used to help explore the biological mechanism of RSA. Through gene expression differential analysis, a total of 295 differentially expressed genes were selected, including 166 up-regulated genes and 129 down-regulated genes. And the majority of these differential genes were enriched within signaling pathways implicated in immune regulation.

Embryo implantation is the physical and physiological interaction between the activated blastocyst and the receiving endometrium, which is the initial stage of pregnancy. Recent study indicates that beyond pregnancy termination resulting from implantation failure, an abnormal immune response can also contribute to subsequent pregnancy loss during the embryo implantation phase [2]. Successful embryo implantation necessitates a normal and orderly immune dialogue at the maternal-fetal interface within the endometrium [3]. The immune response that occurs during the interaction between the mother and fetus in early pregnancy is exceedingly complex, involving a multitude of immune cells such as lymphocytes, decidual dendritic cells, macrophages, and others. These cells play a pivotal role in establishing a balance between inflammatory responses and immune tolerance [3, 4]. In numerous previous studies, the relationship between the immune microenvironment of the endometrium and reproductive disorders has been emphasized, and research findings have shown that these immune disorders are closely related to RSA and recurrent implantation failure(RIF) with unknown etiology [5–7]. The present research concentrates on the regulation of immunity within embryonic cells, achieved by identifying gene expression levels within villous tissue. The outcomes of GO enrichment and KEGG enrichment analyses suggest that the differentially expressed genes are predominantly involved in signaling regulatory pathways related to protein encoding for inflammatory and immune responses. These include inflammatory factors, immune cell composition proteins, as well as receptor and ligand proteins that mediate immune responses. The NFκB signaling pathway, identified through KEGG enrichment, has been verified to play a role in the regulation of embryo implantation by metallomatrix protein and is intimately associated with the occurrence of RSA [8]. PPI analysis revealed that proteins associated with RSA formed a robust network cluster, within which FCGR3A, TLR2, BTK, CLEC7A and CD163 were identified as hub genes.

The receptor encoded by the FCGR3A gene is expressed as a complete membrane glycoprotein anchored by transmembrane peptides on natural killer (NK) cells. It has been reported that the polymorphism of the FCGR3A gene can lead to abnormal immune responses by affecting NK cell function and response levels to monoclonal antibodies [9]. The CLEC7A gene encodes a member of the C-type lectin/C-type lectin like domain (CTL/CTLD) superfamily and is closely related to other CTL/CTLD superfamily members on chromosome 12p13 of the natural killer gene complex [10]. During normal pregnancy, the immune response is characterized by an increase in NK cells, which is accompanied by enhanced cytotoxicity and inhibitory receptor activity. This activity blocks the transmission of cytotoxic signals, thereby tempering the role of NK cells to support the maintenance of a normal pregnancy. Conversely, when the population of NK cells is abnormal, it may result in aberrant immune responses, potentially leading to pregnancy failure and miscarriage [11]. This study assessed the transcription levels of FCGR3A and CLEC7A using RT-PCR, and the findings indicated that both genes were up-regulated in patients with RSA, suggesting that aberrant activation of NK cells might play a role in the etiology of RSA.

The HE protein, encoded by the TLR2 gene, belongs to the Toll-like receptor (TLR) family and plays a crucial role in pathogen recognition and the activation of the innate immune response.Previous research has reported that activation of TLR2 can result in the release of TH1 cytokines [12], and an increase in TLR2 can be detected at the maternal-fetal interface in cases of RSA [13]. The results of the current study indicated that both the mRNA and protein expression levels of TLR2 were up-regulated in the RSA group. The level of the TH1 cytokine IL-2 was elevated, whereas the TH2 cytokine IL-10 did not exhibit a significant difference compared to the control group. The balance between TH1 and TH2 cells plays a crucial role in sustaining a normal pregnancy; an imbalance that favors TH1 can lead to abortion [14]. Abnormal activation of TLR2 can inflict damage on placental artery endothelial cells and directly disrupt the immune balance at the maternal-fetal interface, potentially leading to miscarriage. Our research further confirms that the activation of TLR2 may promote the secretion of TH1 cytokines, leading to the imbalance of TH1/TH2 and playing a role in the occurrence of RSA. Furthermore, elevated levels of TH1 cytokine can stimulate NK cells, resulting in aberrant immune responses associated with RSA.

The protein encoded by the BTK gene plays a crucial role in B cell development and serves as a key component in the BCR signaling pathway. Inhibition of BTK can block the BCR pathway, making it unable to activate NFκB [15], and the NFκB-mediated signaling pathway is involved in the regulation of metallomatrix proteins on embryo implantation, and is closely related to the occurrence of RSA [8]. The CD163 gene encodes a protein belonging to the scavenger receptor cysteine-rich (SRCR) superfamily, which is exclusively expressed in monocytes and macrophages and also serves as a marker for M2-type macrophages. At the maternal-fetal interface, there exists a dynamic equilibrium in the quantity and proportion of M1/M2 macrophages across various gestational stages. Research has indicated that a significant elevation in the M1/M2 ratio can induce a pro-inflammatory microenvironment within the endometrium, which hampers the induction of fetal tolerance and may lead to abortion [16, 17]. This aspect of immune regulation has not been confirmed in the present study, and we will pay further attention to it in future research.

In summary, among the differentially expressed genes in the dataset GSE22490 examined in this study, the up-regulated genes encompass a variety of receptor or cytokine component proteins implicated in immune response. Down-regulated genes encode proteins that facilitate cell division and differentiation, while suppressing cytokines and a series of organelles involved in triggering cell death. The signal pathway of differential gene enrichment is primarily associated with immune regulation, and the HUB gene identified through network analysis of PPI proteins is also linked to key proteins involved in immune regulation. We have successfully confirmed the presence of the hub gene and proceeded to explore the potential immunological mechanisms involved. All findings of this research indicate that an aberrant immune response at the maternal-fetal interface triggers the cell death, potentially serving as the primary biological mechanism behind RSA. Although we have conducted a simple research and discussion on the immunological mechanism, it must be acknowledged that the sample size utilized for differential gene analysis in this study is limited. In future research, we intend to expand the sample size and explore a greater number of possible mechanisms.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

L.S., L.W., D.H. and L.B. conceived the project. L.S. and L.W. wrote the main manuscript text and prepared figures and tables. L.S., M.L. and D.H. performed most of the experiments and analyzed the data. L.W. and L.B. performed image analyses. L.B. and D.H. supervised the project. All authors commented on previous versions of the manuscript and all the authors have read and approved the final manuscript.

Funding

This study was supported by grant from Postgraduate Research & Practice Innovation Program of Jiangsu Province (KYCX24_3340), Health Talent Plan of Suzhou (GSWS2022020).

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was performed in line with the principles of the Declaration of Helsinki, and written informed consent was obtained from all individual participants before being enrolled in the study. The study design, protocol and informed consent were approved and adopted by the Ethics Committee of the First Affiliated Hospital of Soochow University with approval number of 2023(322) on June 2, 2023.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.