Skip to main content
NCBI home page
Phenomics logoLink to Phenomics
. 2024 Oct 16;4(6):570–583. doi: 10.1007/s43657-024-00183-9

Integrated Transcriptome Analysis of lncRNA, miRNA, and mRNA Reveals key Regulatory Modules for Polycystic Ovary Syndrome

Rui Lin 1, Saihua Zheng 1,3, Haiyu Su 2, Guiying Wang 2,✉, Xuelian Li 1,✉, Chenqi Lu 1,✉
PMCID: PMC11889321  PMID: 40061822

Abstract

Polycystic ovarian syndrome (PCOS) is the most common reproductive metabolic disorder in women of reproductive age. However, the underlying mechanism is unclear, because the main symptoms vary with age and the pathogenesis is complex and multifactorial. In order to explore the gene expression and regulation networks, and identify potential biomarkers for diagnosis and treatment of PCOS, we conducted whole RNA sequencing of protein-coding genes, lncRNAs, and miRNAs in peripheral blood with case-control design. RNA sequencing and weighted gene co-expression network analysis (WGCNA) were performed on four pairs of PCOS cases and control peripheral blood samples. The results showed that there were significant differences in the expression levels of 341 mRNAs, 252 lncRNAs and 47 miRNAs between PCOS patients and control groups. Bioinformatics analysis showed that these differentially expressed genes (DEGs) were mainly involved in the metabolic, immune, endocrine, and nervous systems, and also identified potential WGCNA module related with PCOS. The DEGs of PCOS as reported in other published literatures were used to verify our DEGs in this study. These results suggest that the ceRNA regulatory relationship between miR-17-5p, LINC02213 and FCGR1A, the trans-regulatory relationship between RP11-405F3.4:IL1R1 and RP11-405F3.4:IL27, and a hub lncRNA of LINC02649 in core regulatory network, which have significant potential for PCOS research. We constructed the core WGCNA module of PCOS from the whole transcriptome of human peripheral blood and characterized the key gene characteristics of PCOS. These findings provide key insights into the candidate characteristics and mechanism elucidation of PCOS.

Supplementary Information

The online version contains supplementary material available at 10.1007/s43657-024-00183-9.

Keywords: Polycystic ovarian syndrome, Case-control study, Whole transcriptome, lncRNAs, miRNAs, Gene co-expression network

Introduction

As the most common reproductive-metabolic disorder of reproductive-aged women, polycystic ovary syndrome (PCOS) is characterized by hyperandrogenism, menstrual irregularity, and polycystic ovarian morphology (Dumesic et al. 2015). Most women with PCOS also have insulin resistance, which underlies metabolic dysfunction and accompanies increased abdominal body fat over a wide range of body mass index (BMI) (Corbould et al. 2005; Diamanti-Kandarakis and Dunaif 2012; Dumesic et al. 2016; Tosi et al. 2015). PCOS can arise from a myriad of factors, encompassing both genetic and environmental influences, along with lifestyle choices and their various combinations. Conditions such as thyroid dysfunction, hyperprolactinemia, androgen-secreting tumors, Cushing's syndrome (linked to elevated cortisol levels), and congenital adrenal hyperplasia (CAH) have been implicated in the pathogenesis of PCOS. Additionally, inadvertent or deliberate exposure to certain chemicals may heighten the risk of developing PCOS. Numerous genes and pathways have been proposed as potential mediators in the development of PCOS (Patel 2018). This syndrome leads to infertility, insulin resistance, obesity and cardiovascular problems, including a litany of other health problems. Moreover, PCOS patients present a higher risk of metabolic and cardiovascular diseases and their related morbidity, as compared to the general population (De Leo et al. 2016). Studies have further shown that there is physiologically relevant defect in PCOS related to the insulin receptor signaling pathway independent of obesity and type II diabetes (Dunaif et al. 2001). Since PCOS is the most common endocrine disorder in women, there are needs and challenges in diagnosis and treatment (Dokras et al. 2017; Teede et al. 2014). Because the main symptoms may vary with age, the pathogenesis is complex and multifactorial, with unknown underlying mechanisms (De Leo et al. 2016; Hoeger et al. 2021).

Recent advancements in high-throughput sequencing technology have facilitated comprehensive assessments of the entire transcriptome, including mRNAs, long non-coding RNAs (lncRNAs), and microRNAs (miRNAs), providing valuable insights into the underlying biological processes that contribute to complex phenotypes. Within the non-coding regions, constituting 98% of the human genome, miRNAs serve as regulatory non-coding RNAs, exerting negative control over gene expression primarily in the cytoplasm (Chae et al. 2016). MiRNAs serve as pivotal RNA molecules within the competing endogenous RNA (ceRNA) regulatory network, functioning as crucial mediators. A single miRNA has the capability to regulate multiple target genes, and concurrently, a given gene can be subjected to regulation by various miRNAs (Ma et al. 2023). LncRNAs share structural similarities with mRNAs. They are cell-specific, highly conserved, and play pivotal roles as transcriptional regulators. Predominantly situated in the nucleus or cytoplasm, lncRNAs exhibit minimal coding potential for proteins. Nevertheless, they exert regulatory influence on gene expression across various tiers through interactions with DNA, RNA, and protein (Hussen et al. 2021). LncRNAs subject to transcriptional inhibition can be categorized based on pre-transcriptional, transcriptional, and post-transcriptional levels. Each of these three distinct levels holds significant importance in governing various physiological functions (Adnane et al. 2022). LncRNAs can act as miRNA sponge to inhibit the normal biological function of miRNAs and regulate the expression level of miRNA targets. For example, a study on Parkinson's disease found that small nucleolar RNA host gene 1 (SNHG1), a lncRNA that can act as a ceRNA for miR-7, triggers neuroinflammation by up-regulating NOD-like receptor family pyrin domain containing 3 (NLRP3)nullinflammatory vesicles (Cao et al. 2018). Therefore, employing systems biology methodologies enables a more comprehensive depiction of intergenic interactions and the associated signaling pathways in diseases. This approach provides an enhanced capacity to discern co-regulatory patterns that underlie intricate phenotypes (Wang et al. 2020). Furthermore, weighted gene co-expression network analysis (WGCNA) has proven to be a valuable tool in various studies for enabling a systems-level characterization of expression patterns. This is achieved by clustering highly-correlated genes into co-expression modules with shared biological functions (Langfelder and Horvath 2008; Zhang and Horvath 2005). Additionally, WGCNA can serve as an effective feature selection method for exploring potential associations between phenotypes and gene modules, aiding in the identification of gene networks specifically linked to particular phenotypic traits.

To date, there has been limited research on the measurement of lncRNAs in both control subjects and women diagnosed with PCOS (Butler et al. 2019). Existing studies have demonstrated significant differential expression of miRNAs in altered follicular fluid excretory bodies associated with PCOS (Hu et al. 2020). Furthermore, differential expression of lncRNAs and mRNAs has been observed in both mature and immature follicles within the follicular fluid of PCOS patients (Jiao et al. 2018). Notably, the regulatory network involving miRNAs and mRNAs in PCOS has been linked to markers of insulin sensitivity and inflammation (Qin et al. 2021). Previous study utilized high-throughput sequencing technology to assess the expression profiles of both mRNAs and lncRNAs in the peripheral blood of PCOS patients. They unveiled clusters of mRNAs and lncRNAs displaying aberrant expression patterns in the peripheral blood of PCOS patients when compared with control subjects. Additionally, it identified several pairs of lncRNAs and mRNAs within the chemokine signaling pathway that may have genetic associations with PCOS (Sun et al. 2021). However, unlike the well-established role of miRNAs in PCOS, the sequencing of lncRNAs in PCOS remains relatively underexplored.

In this study, we comprehensively characterized the whole transcriptome landscape of human peripheral blood in a case-control study by whole-transcriptome sequencing, by implementing a network-based approach to construct core gene co-expression networks and delineate key lncRNA-miRNA-mRNA axis for PCOS. The results showed that there were significant differences in mRNA, lncRNA and miRNA expression levels between PCOS patients and control group, and these differentially expressed genes were mainly involved in metabolism, immune, endocrine and nervous systems. Our identification of potential WGCNA modules associated with PCOS may provide further mechanistic insights into the pathogenesis of PCOS. Through the construction of ceRNA regulatory network, we found the regulatory relationship between miR-17-5p, LINC02213 and Fc gamma receptor 1A (FCGR1A). Furthermore, the regulatory relationship of RP11-405F3.4:IL1R1, RP11-405F3.4:IL27, and a hub lncRNA LINC02649 was identified. All of these findings have potential significance for the mechanism exploration and treatment of PCOS.

Materials and Methods

Study Design

Four pairs of PCOS cases and controls obtained from Obstetrics and Gynecology Hospital (Fudan University, Shanghai, China) from 2017 to 2018 were used for RNA sequencing. PCOS cases and controls were matched according to age and maternal BMI (Table S1). All women provided written informed consent, and the study protocol was approved by the Ethics Committee of Obstetrics and Gynecology Hospital, Fudan University (No.201,545). All PCOS patients in this study were diagnosed according to Rotterdam criteria. To make a definitive diagnosis of PCOS, at least two of the following criteria need to be met: irregular menstruation (defined as fewer than eight menstrual cycles per year, or no menstruation for more than 35 days); clinical hyperandrogenism (defined as modified Ferriman-Gallwey score of more than six, or androgenic alopecia, or both); biochemical hyperandrogenism (defined as testosterone levels exceeding 2.81 nmol/l, or androstenedione levels exceeding 10.8 nmol/l, or both, which is in the 95th percentile of the normal range for this study population); polycystic ovary morphology (defined as 12 or more follicles 2–9 mm in diameter, or at least one ovary larger than 10 cubic centimeters in volume). Normal women of childbearing age who were treated in our hospital for azoospermia or asthenospermia were selected as the control group. They planned to undergo artificial insemination with donor sperm, had regular periods, were not menopausal, and had no history of adverse pregnancies and childbirth. Biochemical indexes were normal, ultrasound examination showed no polycystic ovaries, CAH, Cushing's syndrome, androgen secretory tumor and other Kaohsiung diseases, no endocrine diseases, metabolic diseases, family history.

Sample Collection

A total of 2.5 mL of maternal peripheral blood were collected into PAXgene whole blood RNA tubes (PreAnalytix) and stored at 25 °C for at least two hours, at -20 °C for 24 h, and at -80 °C until processing. According to the manufacturer’s instructions, the total RNA was extracted using a RNeasy Protect Animal Blood Kit (Qiagen). The RNA concentration and purity were measured using a NanoDrop ND100 spectrophotometer (Thermo Scientific) and BioAnalyzer 2100 system (Agilent).

RNA Sequencing Workflow

For lncRNAs and mRNAs, the RNA-sequencing library generation, workflow, and data analysis were performed as previously described (Weng et al. 2018). The small RNA sequencing including miRNAs was also performed. For lncRNAs and mRNAs, adapters and low-quality reads were removed with Trim Galore (0.4.1), and RNA paired-end reads were compared with the h38 human genome by HISAT (2.1.0) (Kim et al. 2015). RSeQC (2.6.4) was used for quality control (Wang et al. 2012). The reads counts were counted through featureCounts (Liao et al. 2014) in subread (1.5.3) using the annotations file (Homo_sapiens.GRCh38.108.gtf) in the Ensembl database. LncRNAs and mRNAs count matrices were obtained by refGenome package in R. For miRNAs, the adapter was first removed by cutadapt (1.9.1). Additionally, reads less than 18 nt and larger than 31 nt were removed. The mature.fa downloaded from mirbase database (v22) was filtered out into the human mature sequence (mature.human.fa), and the index was established by bowtie (0.12.9). Then, filtered reads were mapped to mature human miRNA mature sequences by bowtie. After that, samtools (1.7) was used to sort and quantify the later files to obtain the read counts matrix of miRNA. Differentially expressed genes (DEGs) were analyzed by DESeq2 package in R (Love et al. 2014), and defined as p < 0.05 and|log2FoldChange| > 1. BAM files were converted to BIGWIG files by bamCoverage (3.5.1) function in deepTools (Ramírez et al. 2014) with the parameter ‘–normalizeUsing RPKM’, then the representative differentially expressed protein coding genes and lncRNAs were demonstrated by IGVtools (2.16.1).

Bioinformatics and Stability Analyses

LncRNA targets were predicted with LncTar, RIblast, and RIsearch, while the miRNA targets were predicted with miRanda, Probability of Interaction by Target Accessibility (PITA), and RNAhybrid. The final regulation relationships were confirmed by at least two of three methods at same location. The gene ontology (GO) and signaling pathway enrichment analyses were completed using clusterProfiler package in R. Based on the log2FoldChange, we performed the gene set enrichment analysis (GSEA) for the significant signaling pathways and validation differentially expressed geneset from published dataset (Mao et al. 2021; Pan et al. 2018; Zou et al. 2022). Disease enrichment of target genes, miRNAs and pre-miRNAs was performed based on the Human MicroRNA Disease Database (HMDD). Clustering and family analysis of pre-miRNAs were performed based on miRBase database. The gene co-expression network was generated using the WGCNA package in R (Langfelder and Horvath 2008). The relative expression levels were calculated by rlog (regularized log) transformation in DESeq2 packages, then the expression matrix of top 5000 genes with higher Median Absolute Deviation (MAD) was used as input of WGCNA. The parameters of WGCNA were as follows: networkType = unsigned, corType = Pearson, Power = 13, minModuleSize = 80, mergeCutHeight = 0.25, randomSeed = 12,345, and the remaining default parameters. The regulatory network was illustrated by the cytoscape (Smoot et al. 2011). The stability of gene modules was assessed by R (Shannon et al. 2016).

Results

Significantly Different Transcriptome of PCOS and Control Peripheral Blood

To examine differences in the entire transcriptome between PCOS cases and normal controls, we performed RNA sequencing for protein-coding genes, lncRNAs, and miRNAs in a case-control study with four pairs of peripheral blood samples. The hierarchical clustering and heatmap of DEGsshowed that more DEGs significantly tended to be up-regulated for protein-coding genes (238 up vs. 103 down, p = 9.6e-14) (Fig. 1a) and lncRNAs (156 up vs. 96 down, p = 9.5e-05) (Fig. 1b). However, there was no significant tendency (25 up vs. 22 down, p = 0.39) for the differentially expressed miRNAs (Fig. 1c). All of the DEGs including protein-coding genes, lncRNAs, and miRNAs were shown in additional file (Table S2). The representative differentially expressed protein-coding genes (IL1R1, FCGR1A) and lncRNAs (LINC02213, LINC02649) were shown in Fig. 1d and e, respectively. These findings indicate that the entire transcriptome of peripheral blood from PCOS case is significantly different from the corresponding control.

Fig. 1.

Fig. 1

Open in a new tab

Significantly different transcriptome of PCOS and control peripheral blood a-c, Hierarchical clustering heatmap of 341 protein-coding genes (a), 252 lncRNAs (b), and 47 miRNAs (c), that are differentially expressed between PCOS and the corresponding control (d), Examples of reads distribution for critical protein-coding genes (IL1R1, FCGR1A) highly-expressed in PCOS that is lowly-expressed in normal control (e), Examples of reads distribution for critical lncRNAs (LINC02213, LINC02649) as rarely-expressed in normal control that becomes activated in PCOS

Differentially Expressed protein-coding Genes and Physiological Functional Signaling Pathways

Among the 341 differentially expressed protein-coding genes, 103 genes were down-regulated and 238 genes were up-regulated in PCOS cases (Fig. 2a, Table S2). The GO enrichment analysis showed that these down-regulated genes were mainly enriched in vitamin transmembrane transporter activity, acetyl-CoA biosynthetic process from pyruvate, and muscle myosin complex (Fig. 2b, Table S3). Further signaling pathway analysis indicated that these genes were mainly involved in diabetic cardiomyopathy, Wnt signaling pathway, and fluid shear stress and atherosclerosis (Fig. 2c, Table S3). Additionally, the up-regulated genes were mainly enriched in neutrophil degranulation, oxygen carrier activity, lipid translocation (Fig. 2d, Table S3), starch and sucrose metabolism and type II diabetes mellitus (Fig. 2e, Table S3). Further GSEA results showed that these genes involving in electron transport chain oxphos system in mitochondria (Fig. 2f) and Myc Targets V1 (Fig. 2i) tended to down-regulated in PCOS, while genes involving in activation of immune response (Fig. 2g) and interferon Alpha response (Fig. 2h) tended to up-regulated in PCOS. These findings indicate that the differentially expressed protein-coding genes are not only enriched in known PCOS-related processes, such as metabolism and the immune system, but also several potential processes associated with cell proliferation and growth.

Fig. 2.

Fig. 2

Open in a new tab

Differentially expressed protein-coding genes and physiological functional signaling pathways (a), Volcano plot of protein-coding genes (238 up-regulated and 103 down-regulated) (b), GO term enrichment of the down-regulated differentially expressed protein-coding genes. Green, cellular component; Blue, biological process cellular; Red, molecular function (c), Signaling pathway enrichment of the down-regulated differentially expressed protein-coding genes (d), GO term enrichment of the up-regulated differentially expressed protein-coding genes (e), Signaling pathway enrichment of the up-regulated differentially expressed protein-coding genes (f-i), GSEA plots of PCOS-related pathways, including electron transport chain (f), immune response (g), Alpha response (h) and Myc Targets (i)

Differentially Expressed miRNAs and Physiological Functional Signaling Pathways

Based on the 47 significantly differentially expressed miRNAs (Fig. 3a, Table S2), we predicted miRNAs targets by miRwalk2 database (Dweep and Gretz 2015). The top 30 popular target genes as regulated by these miRNAs were shown in Fig. 3b, including the coiled-coil domain containing 47 (CCDC47), which is associated with hair abnormalities and diabetic cardiomyopathy (Morimoto et al. 2018; Thapa et al. 2018). Further GO term (Fig. 3c, Table S4) and signaling pathway (Fig. 3d, Table S4) analyses showed that top 30 predicted targets of these miRNAs mainly focused on cellular response to interleukin-6, reproductive structure development, cellular response to leptin stimulus, AGE-RAGE signaling pathway in diabetic complications, Androgen receptor signaling pathway, and Thyroid hormone signaling pathway. An integrated analysis of miRNAs, targets and human diseases using HMDD (v3.2, http://www.cuilab.cn/hmdd) indicated that these miRNAs were significantly correlated with immune diseases, metabolic diseases and neuro-related diseases (Fig. 3e). The miRNA family analyses showed that these miRNAs precursors focused on the miR-17 family and the let-7 family (Fig. 3f).

Fig. 3.

Fig. 3

Open in a new tab

Differentially expressed miRNAs and physiological functional signaling pathways (a), Volcano plot of miRNAs (25 up-regulated and 22 down-regulated) (b), Top 30 popular targets as predicted for the critical miRNAs (c), GO enrichment of the top 30 targets of the critical miRNAs (d), Signaling pathway enrichment of top 30 targets of the critical miRNAs (e), Human diseases enrichment of top 30 targets of the critical miRNAs (f), miRNA family analyses of the precursors of the critical miRNAs

Core Regulatory Network and Validation of Differential Genes in PCOS

To comprehensively explore the core regulatory network of PCOS at the transcriptome level, we conducted RNA sequencing to profile lncRNA expression. Our analysis revealed 96 significantly down-regulated and 156 significantly up-regulated lncRNAs (Fig. 4a, Table S2). Further investigation of the regulatory landscape encompassing 341 DEG protein-coding genes and 252 DEG lncRNAs identified four lncRNAs predicted to cis-regulate five neighboring protein-coding genes within a 2-kb region (Fig. 4b, Table S6). We also identified 13,565 trans-regulatory relationships between lncRNA and mRNA by at least two of third prediction methods. To vividly depict these relationships, we presented the top 35 differentially expressed lncRNAs and their regulation targets in Fig. 4c (Table S7). Additionally, through a combined analysis of lncRNA-mRNA, miRNA-mRNA, and miRNA-lncRNA interactions, we identified three cis-, 140 trans-, and 43 miRNA regulatory relationships, applying a Pearson correlation coefficient (PCC) threshold (absolute value > 0.9) (Fig. S1, Table S8).

Fig. 4.

Fig. 4

Open in a new tab

Core regulatory network and validation of significantly different genes (a), Volcano plot of lncRNAs (156 up-regulated and 96 down-regulated) (b), Predicted cis-regulatory relationships among the differentially expressed protein-coding genes and lncRNAs. The color and shape of nodes indicate the biotype of genes: orange circle for lncRNAs and pink diamond for mRNAs (c), Representative trans-regulatory relationships among the top 35 differentially expressed protein-coding genes and lncRNAs. Blue represents down-regulation, red represents up-regulation, diamond represents mRNA, and circle represents lncRNA (d), Venn plot of verification differentially expressed geneset from two published PCOS datasets (e), GSEA plot of up-regulated genes in PCOS compared with up-regulated genes in verification geneset (f), GSEA plot of down-regulated genes in PCOS compared with down-regulated genes in verification geneset

Among them, we suggested that the regulatory relationship between miR-17-5p and LINC02213, FCGR1A was higher research potential for PCOS. Overexpressed miR-17-5p has been reported to promote cell proliferation and inhibit apoptosis of PCOS ovarian granulosa cells. The altered expression of miRNAs and their interacting mRNA pairs may play a key role in the expression of sex hormone-binding globulin and hormone levels in serum of women with PCOS (Liu et al. 2022). Moreover, we also predicted the regulatory relationship between RP11-405F3.4:IL1R1 and RP11-405F3.4:IL27, in which IL1R1 were not expressed in normal controls but were activated in PCOS (Fig. 1d). In patients with PCOS, IL-27 expression levels are significantly elevated, which is associated with a sustained increase in inflammation during ovulation. The study also indicated that the expression level of IL-27 is affected by the upregulation of estrogen and progesterone, which may promote endometrial decidualization through the signal transducer and activator of transcription 3 (STAT3) signaling pathway, thus affecting the normal function of the reproductive system (Zhang et al. 2022). In addition, the involvement of IL1R1 is also an important aspect of PCOS. IL1R1 is one of the receptors of the interleukin-1 family signaling pathway, and its interaction with inflammatory factors such as IL-1β may further exacerbate the disorder of the immune system. These inflammatory and immune responses may contribute to ovulation disorders, metabolic disorders, and other clinical manifestations in PCOS patients (Al-Obaidi et al. 2022).

Furthermore, we found LINC02649 as a hub node in the core regulatory network, which had two miRNA and six mRNA regulatory partners, especially for the trans-regulatory relationship of LINC02649:SLC26A8 (solute carrier family 26 member 8)and LINC02649:NAIP (NLR family apoptosis inhibitory protein) (Fig. S1). SLC26A8, as a member of anion transporters, plays indispensable roles in sperm motility and male fertility (Mariani et al. 2023). SLC26A8 is also highly expressed in the ovaries and uterus, and closely related to female reproductive health. The significant diagnostic potential of SLC26A8 in PCOS was supported by an integrated analysis (Yao and Wang 2022). For NAIP, the NLR family apoptosis inhibitory protein as critical for immunopathology of PCOS, it has been found a significantly positive correlation between NAIP and IL-1β especially in overweight patients (Rostamtabar et al. 2020), suggested that activation of NAIP inflammasome increased the production of IL-1β in PCOS. Thus, LINC002649 had significant potential for PCOS research.

As significantly differentially expressed protein-coding genes are critical in the occurrence of PCOS, we verified the significant up-regulation and down-regulation of protein-coding genes with data from previous PCOS transcriptome sequencing literatures (Mao et al. 2021; Pan et al. 2018; Zou et al. 2022) (GSE138518, GSE34526). The validation up-regulated and down-regulated gene sets we selected here were confirmed by two previous literatures whose absolute values of log2FoldChange were greater than 1, and the values were with same sign (Fig. 4d, Table S9). The GSEA results showed that genes from the validation up-regulated geneset tends to up-regulated in our dataset (Fig. 4e), genes from the validation down-regulated geneset tends to down-regulated in our dataset (Fig. 4f), indicating the accuracy and robustness of our findings.

Critical Gene Co-expression Network Modules Closely Correlated with PCOS

To clarify the significant gene co-expression network involved in PCOS, we performed WGCNA and clustered the entire transcriptome of PCOS cases and controls (Fig. 5a). All genes were clustered into 21 modules (Table S10), and most genes were observed in the turquoise module as enriched in immune response (Fig. 5b), where the differential genes FCGR1A and IL1R1 as mentioned before in this study was enriched. The cyan module is mainly concentrated in the response of cells to steroid hormone stimulation, the black module is mainly concentrated in peptide metabolism, and the pink module is mainly concentrated in the nervous system development. The Pearson correlation analysis of the relationships between the network modules and sample characteristics showed that the turquoise modules was significantly negatively correlated with age, while the turquoise and blue modules were significantly different between PCOS cases and controls (Fig. 5c). Similarly, we found that the salmon module was positively correlated with the expressed levels of luteinizing hormone (LH) and anti-mullerian hormone (AMH), and the black module was positively correlated with the T expression levels. Further investigation of the module stability suggested that the first four modules, including the turquoise, blue, brown, and yellow modules, exhibited much higher stability than the other modules (Fig. 5d). Further hierarchical clustering indicated that the turquoise module was significantly correlated with age and could clearly separate the PCOS and control groups (Fig. 5e). The GO enrichment and signaling pathway analyses suggested that the turquoise module was mainly enriched in chemokine signaling pathway and sphingolipid signaling pathway (Fig. 5f). These findings indicate that the critical module turquoise is significantly correlated with PCOS, but further confirmation of these genes in the module by a large sample size will provide more evidence for elucidating PCOS.

Fig. 5.

Fig. 5

Open in a new tab

Critical gene co-expression network modules closely correlated with PCOS (a), Results of the WGCNA and clustering of the entire transcriptome (b), Module size of the gene co-expression network and GO enrichment analyses of 21 modules (c), Pearson correlation analysis of the network modules and Clinical indicator variables in patients with PCOS. The color gradient indicates the direction, i.e., positive (red) and negative (blue), and the strength of the correlation. E2, Estradiol; P4, Progesterone; LH, Luteinizing hormone; FSH, Follicle-stimulating hormone; T, Testosterone; TSH, Thyroid stimulating hormone; AMH, Anti-Mullerian hormone (d), Boxplot of stability (Jaccard similarity coefficient) between each module of real data and modules from 1000 bootstrap resampling datasets. The line indexes the best-case stability of the random modules in the simulation (e), Hierarchical clustering heatmap of candidate genes in the representative turquoise module (f), Signaling pathway enrichment of the turquoise module

Discussion

Since the pathogenesis of PCOS is complex and multifactorial, diagnostic challenges, delayed diagnosis, and less-than-ideal treatment options still plague the disease. In this case-control study, we performed a systematic whole-transcriptomic analysis of functional regulatory networks in human PCOS cases and control peripheral blood samples. The differential expression of whole transcriptomes and weighted gene co-expression analyses revealed a wealth of functional processes and key modules related to the metabolic, endocrine, nervous, and immune systems.

Some studies have demonstrated that PCOS is a multifactorial disease with various genetic, metabolic, endocrine, and environmental abnormalities (De Leo et al. 2016; Franks et al. 2006), and there is increasing evidence that PCOS affects women throughout their lives and can begin in the womb of genetically susceptible subjects, present clinically during adolescence, and continue during the reproductive period. It can also put patients at increased risk of cardiovascular disease, high blood pressure, diabetes, and other metabolic complications, especially after menopause (Louwers and Laven 2020). During the reproductive period, it can lead to anovulatory infertility and may be associated with an increased prevalence of pregnancy complications such as miscarriage, gestational diabetes, and pre-eclampsia (De Leo et al. 2003). In newborns of women with PCOS, elevated T levels have been observed in umbilical vein blood (Barry et al. 2010; Mehrabian and Kelishadi 2012), and elevated maternal T levels during the second trimester predict higher anti-mullerian hormone (AMH) levels in adolescent daughters (Hart et al. 2010). A large number of studies have shown that excessive exposure of the fetal hypothalamic-pituitary-ovarian axis to androgens may trigger a series of events. These events may determine the onset of PCOS during adolescence (Abbott et al. 2002; Morimoto et al. 2018), when the hypothalamic-pituitary-ovarian pathway is activated and metabolic changes that lead to altered body fat distribution occur. In particular, during adolescence, insulin levels are physiologically increased, which on the one hand determines the decrease of SHBG levels, while expanding the role of circulating androgens, and on the other hand directly stimulates ovarian steroid production (Lewy et al. 2001). Therefore, early diagnosis is critical, but current diagnosis and management remain challenging. The accuracy of PCOS diagnosis model based on a single biomolecule is relatively limited, while a comprehensive prediction model from gene co-expression module and clinical features can help improve diagnostic reliability. Furthermore, these diagnostic models required validation with large number of clinical samples.

Given the heightened risk of cardiovascular disease, hypertension, diabetes, pregnancy complications, and other metabolic issues associated with PCOS, it is imperative to implement effective screening and treatment protocols to mitigate adverse health outcomes in affected women. In this study, we employed RNA sequencing to comprehensively assess the transcriptome, encompassing mRNAs, lncRNAs, and miRNAs. Non-coding RNAs, particularly lncRNAs and miRNAs, play a pivotal role in transcriptional regulation during the progression of PCOS. Moreover, lncRNAs and miRNAs exhibit close interplay in their regulatory mechanisms, exemplified by the ceRNA hypothesis. Therefore, we have simultaneously analyzed the expression levels of lncRNAs, mRNAs, and miRNAs in the same sample, to comprehensively elucidate the regulatory mechanisms of PCOS mediated by non-coding RNAs. This approach allowed us to uncover distinctive features and potential signaling pathways specific to PCOS. Employing a systems biology and network-based approach to analyze differential expression across the entire transcriptome, our investigation substantiates a significant enrichment of metabolic, immune, and endocrine processes in PCOS. Additionally, we unveiled a crucial correlation between differentially expressed transcripts and neurological conditions like medulloblastoma, epilepsy, and bipolar disorder. These revelations furnish promising leads for the development of more effective PCOS treatments. Considering the diverse cellular composition of human peripheral blood, including red blood cells, white blood cells, and platelets, it is conceivable that these components may exert varying effects on the transcriptomic profiles of protein-coding genes, lncRNAs, and miRNAs. A comprehensive examination of the specific contributions of distinct blood cell sources to RNA profiles could furnish more robust evidence for future studies.

To comprehensively capture intergenic relationships and discern co-regulatory patterns in PCOS, we employed WGCNA. This approach allowed us to systematically investigate expression alterations at a systems level and group genes exhibiting high correlations into co-expression modules (Langfelder and Horvath 2008; Zhang and Horvath 2005). This study identified nine modular-related clinical indicator variables for PCOS. Interestingly, we found that the turquoise module clearly separated the PCOS group from the control group and was inversely associated with age in PCOS cases. Similarly, we found that salmon module was negatively correlated with maternal BMI, while tan module was positively correlated with LH and ASH expression levels. Previous studies have highlighted a neuroendocrine hallmark of PCOS characterized by sustained rapid LH (GnRH) pulsation. This pattern favors the synthesis of LH over FSH in the pituitary gland, leading to an elevated LH concentration and LH: FSH ratio, which are typical features of PCOS (De Leo et al. 2016). However, the potential impact of patient age on the peripheral blood gene expression profile requires further investigation, particularly with larger sample sizes. Additionally, the influence of patient BMI on the genes enriched in the grey modules warrants additional confirmation. To assess the effectiveness of these modules in uncovering the molecular basis of PCOS, we conducted GO and signaling pathway enrichment analyses for each module. These analyses reveal that crucial PCOS modules are primarily associated with metabolic, immune system, neuroendocrine, and transcriptional disorders. Due to the limited sample size used in this study, the stability of gene modules defined by WGCNA was further studied through bootstrap method (Shannon et al. 2016). In line with Shannon et al. (2016), our analysis of the Jaccard similarity coefficient distribution for each module between real data and bootstrapped sets reveals that modules with a larger number of genes exhibit greater stability compared to those with fewer genes. Moreover, we validate that the turquoise, blue, brown, and yellow modules demonstrate superior stability when compared to other modules in the simulated optimal stability of the random module. Nonetheless, further research is essential to ascertain the clinical relevance of these genetic signatures for PCOS treatment and to gain deeper mechanistic insights into the onset of PCOS.

In addition, in order to systematically study the core regulatory network of PCOS at the whole transcriptome level, we conducted a comprehensive analysis of protein-coding genes, lncRNAs, and miRNAs. By combining lncRNA-mRNA, miRNA-mRNA, and miRNA-lncRNA analysis, we observed the core regulatory relationship between three cis-, 140 trans-, and 43 miRNAs. In addition, we used transcriptome sequencing data from other PCOS articles (Mao et al. 2021; Pan et al. 2018; Zou et al. 2022) to verify our data results with their up-regulated and down-regulated genes, and the results were consistent, which explained the accuracy of our results. Based on core regulatory network, we found miR-17-5p may regulate both FCGR1A and LINC00213, RP11-405F3.4 may interact with IL1R1 and IL27, and LINC002649 worked as a hub in core regulatory network, which had potential value for PCOS. However, these findings warrant further investigation into the diagnostic and therapeutic potential of the differentially expressed protein-coding genes, lncRNAs, and miRNAs, particularly in large sample sizes. Additionally, it is crucial to assess their relevance in both PCOS patients and healthy populations to comprehensively address the potential implications for other complications in PCOS patients.

Conclusion

In conclusion, we conducted a comprehensive analysis of human peripheral blood transcriptome, leveraging bioinformatics methods to uncover numerous differentially expressed signatures in lncRNAs, miRNAs, and mRNAs. This approach led to the establishment of a core gene co-expression network, which was further narrowed down to key pathways and molecules associated with PCOS. These findings were subsequently validated using previously reported PCOS transcriptome data, providing significant insights into the pathogenesis and potential biomarker candidates of PCOS. Our research objectives and methodologies exhibit a high degree of relevance and consistency. These results offer substantial support for our discoveries and verify the reliability of our RNA-seq findings. However, given the complexity of PCOS and its prevalence in large populations, further exploration in the diagnostic and therapeutic domains remains crucial. Additional evidence is necessary to deepen our understanding of the functional aspects and underlying mechanisms of PCOS.

Electronic Supplementary Material

Below is the link to the electronic supplementary material.

Supplementary Material 1 (4.7MB, eps)
Supplementary Material 2 (18.6KB, docx)
Supplementary Material 3 (17.5KB, docx)
Supplementary Material 4 (43.3KB, xlsx)
Supplementary Material 5 (18.9KB, xlsx)
Supplementary Material 6 (17.2KB, xlsx)
Supplementary Material 7 (60.1KB, xlsx)
Supplementary Material 8 (11.7KB, xlsx)
Supplementary Material 9 (17.2KB, xlsx)
Supplementary Material 10 (19.6KB, xlsx)
Supplementary Material 11 (19.8KB, xlsx)
Supplementary Material 12 (155.4KB, xlsx)

Acknowledgements

This work was supported by grants obtained from the National Natural Science Foundation of China (32070617, 32270591, and 31970639), and the Fundamental Research Funds for the Central Universities (22120240287).

Abbreviations

PCOS

Polycystic ovarian syndrome

WGCNA

Weighted gene co-expression network analysis

DEGs

Differentially expressed genes

lncRNAs

Long non-coding RNAs

miRNAs

MicroRNAs

BMI

Body mass index

GSEA

Gene set enrichment analysis

HMDD

Human microRNA disease database

GO

Gene ontology

ceRNA

Competing endogenous RNA

Authors' Contributions

R L, CQ L, GY W, and XL L designed the study. SH Z and XL L collected samples. R L, HY S, and CQ L collected data and ran the analyses. R L, SH Z, HY S, GY W, XL L, and CQ L interpreted the results and wrote the manuscript. All authors read and approved the final manuscript.

Data Availability

The RNA sequencing data were deposited in the National Omics Data Encyclopedia (NODE) database (https://www.biosino.org/node) under project ID: OEP004667. The data that support the findings of the current study are available from the corresponding authors on reasonable request.

Declarations

Ethical Approval

The study protocol was approved by the Ethics Committee of Obstetrics and Gynecology Hospital, Fudan University (No.201545).

Consent to Participate

Written informed consent was obtained from the participants.

Consent to Publish

All the participants approved to publish.

Competing Interests

The authors declare that they have no competing interests.

Contributor Information

Guiying Wang, Email: wgy@tongji.edu.cn.

Xuelian Li, Email: xllifc@fudan.edu.cn.

Chenqi Lu, Email: luchenqi@fudan.edu.cn.

References

  1. Abbott DH, Dumesic DA, Franks S (2002) Developmental origin of polycystic ovary syndrome - a hypothesis. J Endocrinol 174(1):1–5. 10.1677/joe.0.1740001 [DOI] [PubMed] [Google Scholar]
  2. Adnane S, Marino A, Leucci E (2022) LncRNAs in human cancers: signal from noise. Trends Cell Biol 32(7):565–573. 10.1016/j.tcb.2022.01.006 [DOI] [PubMed] [Google Scholar]
  3. Al-Obaidi MT, Thamer E, Ibraheem QA (2022) Interleukin-1β, 27 level in stimulated and non-stimulated PCOS infertile women (cross-sectional study). Open Access Macedonian J Med Sci 10(B):1555–1559. 10.3889/oamjms.2022.9973 [Google Scholar]
  4. Barry JA, Kay AR, Navaratnarajah R et al (2010) Umbilical vein testosterone in female infants born to mothers with polycystic ovary syndrome is elevated to male levels. J Obstet Gynaecol 30(5):444–446. 10.3109/01443615.2010.485254 [DOI] [PubMed] [Google Scholar]
  5. Butler AE, Hayat S, Dargham SR et al (2019) Alterations in long noncoding RNAs in women with and without polycystic ovarian syndrome. Clin Endocrinol (Oxf) 91(6):793–797. 10.1111/cen.14087 [DOI] [PubMed] [Google Scholar]
  6. Cao B, Wang T, Qu Q, Kang T, Yang Q (2018) Long noncoding RNA SNHG1 promotes Neuroinflammation in Parkinson’s Disease via regulating miR-7/NLRP3 pathway. Neuroscience 388:118–127. 10.1016/j.neuroscience.2018.07.019 [DOI] [PubMed] [Google Scholar]
  7. Chae DK, Ban E, Yoo YS, Baik JH, Song EJ (2016) Evaluation of inhibition of miRNA expression induced by anti-miRNA oligonucleotides. Anal Bioanal Chem 408(18):4829–4833. 10.1007/s00216-016-9611-z [DOI] [PubMed] [Google Scholar]
  8. Corbould A, Kim YB, Youngren JF et al (2005) Insulin resistance in the skeletal muscle of women with PCOS involves intrinsic and acquired defects in insulin signaling. Am J Physiol Endocrinol Metab 288(5):E1047–1054. 10.1152/ajpendo.00361.2004 [DOI] [PubMed] [Google Scholar]
  9. De Leo V, la Marca A, Petraglia F (2003) Insulin-lowering agents in the management of polycystic ovary syndrome. Endocr Rev 24(5):633–667. 10.1210/er.2002-0015 [DOI] [PubMed] [Google Scholar]
  10. De Leo V, Musacchio MC, Cappelli V, Massaro MG, Morgante G, Petraglia F (2016) Genetic, hormonal and metabolic aspects of PCOS: an update. Reprod Biol Endocrinol 14(1):38. 10.1186/s12958-016-0173-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Diamanti-Kandarakis E, Dunaif A (2012) Insulin resistance and the polycystic ovary syndrome revisited: an update on mechanisms and implications. Endocr Rev 33(6):981–1030. 10.1210/er.2011-1034 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Dokras A, Saini S, Gibson-Helm M, Schulkin J, Cooney L, Teede H (2017) Gaps in knowledge among physicians regarding diagnostic criteria and management of polycystic ovary syndrome. Fertil Steril 107(6):1380–1386e1381. 10.1016/j.fertnstert.2017.04.011 [DOI] [PubMed] [Google Scholar]
  13. Dumesic DA, Oberfield SE, Stener-Victorin E, Marshall JC, Laven JS, Legro RS (2015) Scientific Statement on the Diagnostic Criteria, Epidemiology, Pathophysiology, and Molecular Genetics of Polycystic Ovary Syndrome. Endocr Rev 36(5):487–525. 10.1210/er.2015-1018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dumesic DA, Akopians AL, Madrigal VK et al (2016) Hyperandrogenism accompanies increased intra-abdominal Fat Storage in Normal Weight Polycystic Ovary Syndrome women. J Clin Endocrinol Metab 101(11):4178–4188. 10.1210/jc.2016-2586 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Dunaif A, Wu X, Lee A, Diamanti-Kandarakis E (2001) Defects in insulin receptor signaling in vivo in the polycystic ovary syndrome (PCOS). Am J Physiol Endocrinol Metab 281(2):E392–399. 10.1152/ajpendo.2001.281.2.E392 [DOI] [PubMed] [Google Scholar]
  16. Dweep H, Gretz N (2015) miRWalk2.0: a comprehensive atlas of microRNA-target interactions. Nat Methods 12(8):697. 10.1038/nmeth.3485 [DOI] [PubMed] [Google Scholar]
  17. Franks S, McCarthy MI, Hardy K (2006) Development of polycystic ovary syndrome: involvement of genetic and environmental factors. Int J Androl 29(1):278–285 discussion 286–290. 10.1111/j.1365-2605.2005.00623.x [DOI] [PubMed] [Google Scholar]
  18. Hart R, Doherty DA, Norman RJ et al (2010) Serum antimullerian hormone (AMH) levels are elevated in adolescent girls with polycystic ovaries and the polycystic ovarian syndrome (PCOS). Fertil Steril 94(3):1118–1121. 10.1016/j.fertnstert.2009.11.002 [DOI] [PubMed] [Google Scholar]
  19. Hoeger KM, Dokras A, Piltonen T (2021) Update on PCOS: consequences, challenges, and Guiding Treatment. J Clin Endocrinol Metab 106(3):e1071–e1083. 10.1210/clinem/dgaa839 [DOI] [PubMed] [Google Scholar]
  20. Hu J, Tang T, Zeng Z, Wu J, Tan X, Yan J (2020) The expression of small RNAs in exosomes of follicular fluid altered in human polycystic ovarian syndrome. PeerJ 8:e8640. 10.7717/peerj.8640 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hussen BM, Azimi T, Abak A, Hidayat HJ, Taheri M, Ghafouri-Fard S (2021) Role of lncRNA BANCR in Human cancers: an updated review. Front Cell Dev Biol 9:689992. 10.3389/fcell.2021.689992 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Jiao J, Shi B, Wang T et al (2018) Characterization of long non-coding RNA and messenger RNA profiles in follicular fluid from mature and immature ovarian follicles of healthy women and women with polycystic ovary syndrome. Hum Reprod 33(9):1735–1748. 10.1093/humrep/dey255 [DOI] [PubMed] [Google Scholar]
  23. Kim D, Langmead B, Salzberg SL (2015) HISAT: a fast spliced aligner with low memory requirements. Nat Methods 12(4):357–360. 10.1038/nmeth.3317 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Langfelder P, Horvath S (2008) WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics 9:559. 10.1186/1471-2105-9-559 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lewy VD, Danadian K, Witchel SF, Arslanian S (2001) Early metabolic abnormalities in adolescent girls with polycystic ovarian syndrome. J Pediatr 138(1):38–44. 10.1067/mpd.2001.109603 [DOI] [PubMed] [Google Scholar]
  26. Liao Y, Smyth GK, Shi W (2014) featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics 30(7):923–930. 10.1093/bioinformatics/btt656 [DOI] [PubMed] [Google Scholar]
  27. Liu G, Liu S, Xing G, Wang F (2022) Retraction notice to: lncRNA PVT1/MicroRNA-17-5p/PTEN Axis regulates secretion of E2 and P4, proliferation, and apoptosis of ovarian granulosa cells in PCOS. Mol Ther Nucleic Acids 28:593. 10.1016/j.omtn.2022.05.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Louwers YV, Laven JSE (2020) Characteristics of polycystic ovary syndrome throughout life. Ther Adv Reprod Health 14:2633494120911038. 10.1177/2633494120911038 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Love MI, Huber W, Anders S (2014) Moderated estimation of Fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15(12):550. 10.1186/s13059-014-0550-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Ma B, Wang S, Wu W et al (2023) Mechanisms of circRNA/lncRNA-miRNA interactions and applications in disease and drug research. Biomed Pharmacother 162:114672. 10.1016/j.biopha.2023.114672 [DOI] [PubMed] [Google Scholar]
  31. Mao Z, Li T, Zhao H, Qin Y, Wang X, Kang Y (2021) Identification of epigenetic interactions between microRNA and DNA methylation associated with polycystic ovarian syndrome. J Hum Genet 66(2):123–137. 10.1038/s10038-020-0819-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Mariani NAP, Silva JV, Fardilha M, Silva EJR (2023) Advances in non-hormonal male contraception targeting sperm motility. Hum Reprod Update 29(5):545–569. 10.1093/humupd/dmad008 [DOI] [PubMed] [Google Scholar]
  33. Mehrabian F, Kelishadi R (2012) Comparison of the metabolic parameters and androgen level of umbilical cord blood in newborns of mothers with polycystic ovary syndrome and controls. J Res Med Sci 17(3):207–211 [PMC free article] [PubMed] [Google Scholar]
  34. Morimoto M, Waller-Evans H, Ammous Z et al (2018) Bi-allelic CCDC47 variants cause a disorder characterized by woolly hair, liver dysfunction, dysmorphic features and global Developmental Delay. Am J Hum Genet 103(5):794–807. 10.1016/j.ajhg.2018.09.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Pan JX, Tan YJ, Wang FF et al (2018) Aberrant expression and DNA methylation of lipid metabolism genes in PCOS: a new insight into its pathogenesis. Clin Epigenetics 10:6. 10.1186/s13148-018-0442-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Patel S (2018) Polycystic ovary syndrome (PCOS), an inflammatory, systemic, lifestyle endocrinopathy. J Steroid Biochem Mol Biol 182:27–36. 10.1016/j.jsbmb.2018.04.008 [DOI] [PubMed] [Google Scholar]
  37. Qin Y, Wang Y, Zhao H, Yang Z, Kang Y (2021) Aberrant miRNA-mRNA regulatory network in polycystic ovary syndrome is associated with markers of insulin sensitivity and inflammation. Ann Transl Med 9(18):1405. 10.21037/atm-21-1288 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Ramírez F, Dündar F, Diehl S, Grüning BA, Manke T (2014) deepTools: a flexible platform for exploring deep-sequencing data. Nucleic Acids Res 42:W187–W191. 10.1093/nar/gku365 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Rostamtabar M, Esmaeilzadeh S, Karkhah A et al (2020) Elevated expression of IL-18 but not IL-1beta gene is associated with NALP3 and AIM2 inflammasome in polycystic ovary syndrome. Gene 731:144352. 10.1016/j.gene.2020.144352 [DOI] [PubMed] [Google Scholar]
  40. Shannon CP, Chen V, Takhar M et al (2016) SABRE: a method for assessing the stability of gene modules in complex tissues and subject populations. BMC Bioinformatics 17(1):460. 10.1186/s12859-016-1319-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Smoot ME, Ono K, Ruscheinski J, Wang PL, Ideker T (2011) Cytoscape 2.8: new features for data integration and network visualization. Bioinformatics 27(3):431–432. 10.1093/bioinformatics/btq675 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Sun X, Liu Y, Gao X, Du M, Gao M, Zhong X, Wei X (2021) Analysis of LncRNA-mRNA co-expression profiles in patients with polycystic ovary syndrome: a pilot study. Front Immunol 12:669819. 10.3389/fimmu.2021.669819 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Teede H, Gibson-Helm M, Norman RJ, Boyle J (2014) Polycystic ovary syndrome: perceptions and attitudes of women and primary health care physicians on features of PCOS and renaming the syndrome. J Clin Endocrinol Metab 99(1):E107–111. 10.1210/jc.2013-2978 [DOI] [PubMed] [Google Scholar]
  44. Thapa K, Wu KC, Sarma A et al (2018) Dysregulation of the calcium handling protein, CCDC47, is associated with diabetic cardiomyopathy. Cell Biosci 8:45. 10.1186/s13578-018-0244-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Tosi F, Di Sarra D, Kaufman JM et al (2015) Total body fat and central fat mass independently predict insulin resistance but not hyperandrogenemia in women with polycystic ovary syndrome. J Clin Endocrinol Metab 100(2):661–669. 10.1210/jc.2014-2786 [DOI] [PubMed] [Google Scholar]
  46. Wang L, Wang S, Li W (2012) RSeQC: quality control of RNA-seq experiments. Bioinformatics 28(16):2184–2185. 10.1093/bioinformatics/bts356 [DOI] [PubMed] [Google Scholar]
  47. Wang G, Yu J, Yang Y et al (2020) Whole-transcriptome sequencing uncovers core regulatory modules and gene signatures of human fetal growth restriction. Clin Transl Med 9(1):9. 10.1186/s40169-020-0259-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Weng R, Lu C, Liu X et al (2018) Long noncoding RNA-1604 orchestrates neural differentiation through the miR-200c/ZEB Axis. Stem Cells 36(3):325–336. 10.1002/stem.2749 [DOI] [PubMed] [Google Scholar]
  49. Yao X, Wang X (2022) Bioinformatics searching of diagnostic markers and immune infiltration in polycystic ovary syndrome. Front Genet 13:937309. 10.3389/fgene.2022.937309 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Zhang B, Horvath S (2005) A general framework for weighted gene co-expression network analysis. Stat Appl Genet Mol Biol 4:Article17. 10.2202/1544-6115.1128 [DOI] [PubMed] [Google Scholar]
  51. Zhang XY, Shen HH, Qin XY et al (2022) IL-27 promotes decidualization via the STAT3-ESR/PGR regulatory axis. J Reprod Immunol 151:103623. 10.1016/j.jri.2022.103623 [DOI] [PubMed] [Google Scholar]
  52. Zou J, Li Y, Liao N et al (2022) Identification of key genes associated with polycystic ovary syndrome (PCOS) and ovarian cancer using an integrated bioinformatics analysis. J Ovarian Res 15(1):30. 10.1186/s13048-022-00962-w [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1 (4.7MB, eps)
Supplementary Material 2 (18.6KB, docx)
Supplementary Material 3 (17.5KB, docx)
Supplementary Material 4 (43.3KB, xlsx)
Supplementary Material 5 (18.9KB, xlsx)
Supplementary Material 6 (17.2KB, xlsx)
Supplementary Material 7 (60.1KB, xlsx)
Supplementary Material 8 (11.7KB, xlsx)
Supplementary Material 9 (17.2KB, xlsx)
Supplementary Material 10 (19.6KB, xlsx)
Supplementary Material 11 (19.8KB, xlsx)
Supplementary Material 12 (155.4KB, xlsx)

Data Availability Statement

The RNA sequencing data were deposited in the National Omics Data Encyclopedia (NODE) database (https://www.biosino.org/node) under project ID: OEP004667. The data that support the findings of the current study are available from the corresponding authors on reasonable request.

Articles from Phenomics are provided here courtesy of Springer

RESOURCES