Systematic prediction of key genes for ovarian cancer by co-expression network analysis

Collagen fiber density observed in metastatic ovarian cancer promotes tumor cell adhesion

Collagen type I, a key structural component of the extracellular matrix (ECM), is frequently altered in cancer, with altered fiber organization at the primary tumor site linked to metastasis and poor patient outcomes. Here, we demonstrate that collagen fibers are also altered in metastatic sites such as the omentum of patients with high-grade serous ovarian cancer (HGSOC). Specifically, we observed a significant increase in fiber density, alignment, and width. To determine if the increase in fiber density supports metastasis, we used a semi-interpenetrating methacrylated gelatin (gelMA) network in combination with increasing fibrillar collagen. Cancer cells had significantly increased adhesion as collagen fiber density increased. To determine the responsible mechanisms, we used orthogonal systems to examine 1) the different adhesion peptides exposed in collagen (GFOGER) and gelatin (RGD), and 2) the physical structure of fibers. Cells had minimal response to GFOGER, either alone or in combination with RGD, suggesting that increased adhesion did not result from this collagen-specific interaction. Cell adhesion was significantly higher on electrospun PCL-gelatin fibers compared to flat PCL-gelatin substrates, suggesting that increased cell adhesion resulted from fiber structure. We next investigated the cellular mechanisms involved in increased adhesion on gelMA/coll and found that actin polymerization, but not myosin II contractility, was needed. We further demonstrated that cells on fibrous gels had more robust actin polymerization, and that this resulted in greater adhesion strength. Combined, these results suggest that the increase in collagen fibers with tumor metastasis will support the development of additional metastases. STATEMENT OF SIGNIFICANCE: This work advances the evaluation of the matrisome of the omentum, the most common metastatic site in advanced ovarian cancer by characterizing how collagen fibers change with disease progression. To examine the effect of collagen fibers on metastasis, we utilized a suite of in vitro biomaterials to identify a novel role for collagen fibers in supporting cell adhesion through increased actin dynamics during nascent adhesion formation, which results in increased adhesion strength at later times.

Multi-dimensional characterization of cellular states reveals clinically relevant immunological subtypes and therapeutic vulnerabilities in ovarian cancer

BACKGROUND

Diverse cell types and cellular states in the tumor microenvironment (TME) are drivers of biological and therapeutic heterogeneity in ovarian cancer (OV). Characterization of the diverse malignant and immunology cellular states that make up the TME and their associations with clinical outcomes are critical for cancer therapy. However, we are still lack of knowledge about the cellular states and their clinical relevance in OV.

METHODS

We manually collected the comprehensive transcriptomes of OV samples and characterized the cellular states and ecotypes based on a machine-learning framework. The robustness of the cellular states was validated in independent cohorts and single-cell transcriptomes. The functions and regulators of cellular states were investigated. Meanwhile, we thoroughly examined the associations between cellular states and various clinical factors, including clinical prognosis and drug responses.

RESULTS

We depicted and characterized an immunophenotypic landscape of 3,099 OV samples and 80,044 cells based on a machine learning framework. We identified and validated 32 distinct transcriptionally defined cellular states from 12 cell types and three cellular communities or ecotypes, extending the current immunological subtypes in OV. Functional enrichment and upstream transcriptional regulator analyses revealed cancer hallmark-related pathways and potential immunological biomarkers. We further investigated the spatial patterns of identified cellular states by integrating the spatially resolved transcriptomes. Moreover, prognostic landscape and drug sensitivity analysis exhibited clinically relevant immunological subtypes and therapeutic vulnerabilities.

CONCLUSION

Our comprehensive analysis of TME helps leveraging various immunological subtypes to highlight new directions and targets for the treatment of cancer.

Tumor-associated fibrosis: a unique mechanism promoting ovarian cancer metastasis and peritoneal dissemination

Epithelial ovarian cancer (EOC) is often diagnosed in advanced stage with peritoneal dissemination. Recent studies indicate that aberrant accumulation of collagen fibers in tumor stroma has a variety of effects on tumor progression. We refer to remodeled fibrous stroma with altered expression of collagen molecules, increased stiffness, and highly oriented collagen fibers as tumor-associated fibrosis (TAF). TAF contributes to EOC cell invasion and metastasis in the intraperitoneal cavity. However, an understanding of molecular events involved is only just beginning to emerge. Further development in this field will lead to new strategies to treat EOC. In this review, we focus on the recent findings on how the TAF contributes to EOC malignancy. Furthermore, we will review the recent initiatives and future therapeutic strategies for targeting TAF in EOC.

Loss of mitochondrial pyruvate carrier 1 supports proline-dependent proliferation and collagen biosynthesis in ovarian cancer

The pyruvate transporter MPC1 (mitochondrial pyruvate carrier 1) acts as a tumour-suppressor, loss of which correlates with a pro-tumorigenic phenotype and poor survival in several tumour types. In high-grade serous ovarian cancers (HGSOC), patients display copy number loss of MPC1 in around 78% of cases and reduced MPC1 mRNA expression. To explore the metabolic effect of reduced expression, we demonstrate that depleting MPC1 in HGSOC cell lines drives expression of key proline biosynthetic genes; PYCR1, PYCR2 and PYCR3, and biosynthesis of proline. We show that altered proline metabolism underpins cancer cell proliferation, reactive oxygen species (ROS) production, and type I and type VI collagen formation in ovarian cancer cells. Furthermore, exploring The Cancer Genome Atlas, we discovered the PYCR3 isozyme to be highly expressed in a third of HGSOC patients, which was associated with more aggressive disease and diagnosis at a younger age. Taken together, our study highlights that targeting proline metabolism is a potential therapeutic avenue for the treatment of HGSOC.

HuR-dependent expression of Wisp1 is necessary for TGFβ-induced cardiac myofibroblast activity

Cardiac fibrosis is regulated by the activation and phenotypic switching of quiescent cardiac fibroblasts to active myofibroblasts, which have extracellular matrix (ECM) remodeling and contractile functions which play a central role in cardiac remodeling in response to injury. Here, we show that expression and activity of the RNA binding protein HuR is increased in cardiac fibroblasts upon transformation to an active myofibroblast. Pharmacological inhibition of HuR significantly blunts the TGFβ-dependent increase in ECM remodeling genes, total collagen secretion, in vitro scratch closure, and collagen gel contraction in isolated primary cardiac fibroblasts, suggesting a suppression of TGFβ-induced myofibroblast activation upon HuR inhibition. We identified twenty-four mRNA transcripts that were enriched for HuR binding following TGFβ treatment via photoactivatable ribonucleoside-enhanced crosslinking and immunoprecipitation (PAR-CLIP). Eleven of these HuR-bound mRNAs also showed significant co-expression correlation with HuR, αSMA, and periostin in primary fibroblasts isolated from the ischemic-zone of infarcted mouse hearts. Of these, WNT1-inducible signaling pathway protein-1 (Wisp1; Ccn4), was the most significantly associated with HuR expression in fibroblasts. Accordingly, we found Wisp1 expression to be increased in cardiac fibroblasts isolated from the ischemic-zone of mouse hearts following ischemia/reperfusion, and confirmed Wisp1 expression to be HuR-dependent in isolated fibroblasts. Finally, addition of exogenous recombinant Wisp1 partially rescued myofibroblast-induced collagen gel contraction following HuR inhibition, demonstrating that HuR-dependent Wisp1 expression plays a functional role in HuR-dependent MF activity downstream of TGFβ. In conclusion, HuR activity is necessary for the functional activation of primary cardiac fibroblasts in response to TGFβ, in part through post-transcriptional regulation of Wisp1.

MLLT11-TRIL complex promotes the progression of endometrial cancer through PI3K/AKT/mTOR signaling pathway

Endometrial cancer (EC) is a gynecological malignant tumor characterized by high incidence. EC occurrence and development are regulated by numerous molecules and signal pathways. There is a need to explore key regulatory molecules to identify potential therapeutic targets to reduce the incidence of EC. Treatment by targeting a single molecule is characterized by poor efficacy owing to the development of resistance and significant side effects. The current study explored potential candidates in EC by integrating bioinformatics analysis and in vivo and in vitro experimental validation to circumvent the limitation of low efficacy of currently used molecules. Molecular dynamics simulations provide details at the molecular level of intermolecular regulation. In the current study, MLLT11 and TRIL were identified as important regulatory molecules in EC. The two molecules formed a heteromultimer by binding to AKT protein, which induced its phosphorylation of threonine at position 308. Ultimately, the complex stimulates PI3K/AKT/mTOR signaling pathway, a pivotal pathway in tumors. The findings of the current study show a novel complex, MLLT11-TRIL, which can act as AKT protein agonist, thus inducing activity of PI3K/AKT/mTOR signaling pathway. Targeting MLLT11 and TRIL simultaneously, or blocking the formation of the MLLT11-TRIL complex, can abrogate progression of EC.

Lnc-CPLC promotes the progression of colorectal cancer via regulating ZBTB34 by competitively binding miR-4319

Long noncoding RNAs (lncRNAs) have been considered as regulatory molecules that play crucial roles in diverse biological processes, including the regulation of tumor progression. However, in colorectal cancer (CRC), due to the complex regulatory relationships involved in lncRNAs, the details of the specific mechanism still need to be elucidated. To discover the key regulatory role of lncRNA in CRC, we used bioinformatics analysis methods for preliminary screening. Through the combination of in vitro and in vivo verification, we further comprehensively analyzed the specific regulation of the key gene and the related key lncRNA in CRC. We found that ZBTB34 and lnc-CPLC (CRC progression-associated lncRNA) had a strong correlation, which plays a key role in the regulation of CRC. Furthermore, by exerting the "sponge" function, lnc-CPLC could bind to miR4319 and release its binding to the 3'UTR of ZBTB34 mRNA. Our results reveal the mechanism of the lnc-CPLC/miR-4319/ZBTB34 signal axis in CRC and provide evidence for elucidating the complex molecular mechanisms in tumors.

Integrated microRNA and mRNA signatures associated with overall survival in epithelial ovarian cancer

Ovarian cancer (OC), the eighth-leading cause of cancer-related death among females worldwide, is mainly represented by epithelial OC (EOC) that can be further subdivided into four subtypes: serous (75%), endometrioid (10%), clear cell (10%), and mucinous (3%). Major reasons for high mortality are the poor biological understanding of the OC mechanisms and a lack of reliable markers defining each EOC subtype. MicroRNAs (miRNAs) are small non-coding RNA molecules that regulate gene expression primarily by targeting messenger RNA (mRNA) transcripts. Their aberrant expression patterns have been associated with cancer development, including OC. However, the role of miRNAs in tumorigenesis is still to be determined, mainly due to the lack of consensus regarding optimal methodologies for identification and validation of miRNAs and their targets. Several tools for computational target prediction exist, but false interpretations remain a problem. The experimental validation of every potential miRNA-mRNA pair is not feasible, as it is laborious and expensive. In this study, we analyzed the correlation between global miRNA and mRNA expression patterns derived from microarray profiling of 197 EOC patients to identify the signatures of miRNA-mRNA interactions associated with overall survival (OS). The aim was to investigate whether these miRNA-mRNA signatures might have a prognostic value for OS in different subtypes of EOC. The content of our cohort (162 serous carcinomas, 15 endometrioid carcinomas, 11 mucinous carcinomas, and 9 clear cell carcinomas) reflects a real-world scenario of EOC. Several interaction pairs between 6 miRNAs (hsa-miR-126-3p, hsa-miR-223-3p, hsa-miR-23a-5p, hsa-miR-27a-5p, hsa-miR-486-5p, and hsa-miR-506-3p) and 8 mRNAs (ATF3, CH25H, EMP1, HBB, HBEGF, NAMPT, POSTN, and PROCR) were identified and the findings appear to be well supported by the literature. This indicates that our study has a potential to reveal miRNA-mRNA signatures relevant for EOC. Thus, the evaluation on independent cohorts will further evaluate the performance of such findings.

Adipocyte-specific deletion of HuR induces spontaneous cardiac hypertrophy and fibrosis

Adipose tissue homeostasis plays a central role in cardiovascular physiology, and the presence of thermogenically active brown adipose tissue (BAT) has recently been associated with cardiometabolic health. We have previously shown that adipose tissue-specific deletion of HuR (Adipo-HuR^-/-) reduces BAT-mediated adaptive thermogenesis, and the goal of this work was to identify the cardiovascular impacts of Adipo-HuR^-/-. We found that Adipo-HuR^-/- mice exhibit a hypercontractile phenotype that is accompanied by increased left ventricle wall thickness and hypertrophic gene expression. Furthermore, hearts from Adipo-HuR^-/- mice display increased fibrosis via picrosirius red staining and periostin expression. To identify underlying mechanisms, we applied both RNA-seq and weighted gene coexpression network analysis (WGCNA) across both cardiac and adipose tissue to define HuR-dependent changes in gene expression as well as significant relationships between adipose tissue gene expression and cardiac fibrosis. RNA-seq results demonstrated a significant increase in proinflammatory gene expression in both cardiac and subcutaneous white adipose tissue (scWAT) from Adipo-HuR^-/- mice that is accompanied by an increase in serum levels of both TNF-α and IL-6. In addition to inflammation-related genes, WGCNA identified a significant enrichment in extracellular vesicle-mediated transport and exosome-associated genes in scWAT, whose expression most significantly associated with the degree of cardiac fibrosis observed in Adipo-HuR^-/- mice, implicating these processes as a likely adipose-to-cardiac paracrine mechanism. These results are significant in that they demonstrate the spontaneous onset of cardiovascular pathology in an adipose tissue-specific gene deletion model and contribute to our understanding of how disruptions in adipose tissue homeostasis may mediate cardiovascular disease.NEW & NOTEWORTHY The presence of functional brown adipose tissue in humans is known to be associated with cardiovascular health. Here, we show that adipocyte-specific deletion of the RNA binding protein HuR, which we have previously shown to reduce BAT-mediated thermogenesis, is sufficient to mediate a spontaneous development of cardiac hypertrophy and fibrosis. These results may have implications on the mechanisms by which BAT function and adipose tissue homeostasis directly mediate cardiovascular disease.

Integrated study of miR-215 promoting breast cancer cell apoptosis by targeting RAD54B

BACKGROUND

MicroRNAs (miRNAs) are widely distributed in cells and participate in the regulation of the pathophysiological process of many diseases. As an important part of non-coding RNA, miRNAs regulate a variety of molecules and signal pathways in tumour cells. However, the evidence for regulatory mechanisms of specific miRNAs in tumour cells is still lacking.

METHODS

In this study, we used transcriptomics analysis and integrated a variety of public databases to screen miRNAs that have key regulatory effects on breast cancer (BC). In addition, we used in vitro and in vivo studies and combined clinical samples to verify its regulatory mechanism.

RESULTS

We found that among the specific miRNAs, miR-215-5p is a key regulator in BC. Compared with normal adjacent tissues, miR-215-5p has a lower expression level in BC tissues. Patients with high expression levels of miR-215-5p have a longer survival time. miR-215-5p can specifically target the 3'UTR region of RAD54B mRNA and down-regulate the expression of RAD54B, thereby inhibiting the proliferation of BC cells and promoting the apoptosis of BC cells.

CONCLUSIONS

Finally, we found that miR-215-5p can be used as an important biomarker for BC. We have clarified its function and revealed its mechanism of targeting RAD54B mRNA for the first time. This may provide important clues to reveal the deeper molecular regulation mechanism of BC.

Integrated Protein-Protein Interaction and Weighted Gene Co-expression Network Analysis Uncover Three Key Genes in Hepatoblastoma

Hepatoblastoma (HB) is the most common liver tumor in the pediatric population, with typically poor outcomes for advanced-stage or chemotherapy-refractory HB patients. The objective of this study was to identify genes involved in HB pathogenesis via microarray analysis and subsequent experimental validation. We identified 856 differentially expressed genes (DEGs) between HB and normal liver tissue based on two publicly available microarray datasets (GSE131329 and GSE75271) after data merging and batch effect correction. Protein–protein interaction (PPI) analysis and weighted gene co-expression network analysis (WGCNA) were conducted to explore HB-related critical modules and hub genes. Subsequently, Gene Ontology (GO) analysis was used to reveal critical biological functions in the initiation and progression of HB. Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis showed that genes involved in cell cycle phase transition and the PI3K/AKT signaling were associated with HB. The intersection of hub genes identified by both PPI and WGCNA analyses revealed five potential candidate genes. Based on receiver operating characteristic (ROC) curve analysis and reports in the literature, we selected CCNA2, CDK1, and CDC20 as key genes of interest to validate experimentally. CCNA2, CDK1, or CDC20 small interfering RNA (siRNA) knockdown inhibited aggressive biological properties of both HepG2 and HuH-6 cell lines in vitro. In conclusion, we identified CCNA2, CDK1, and CDC20 as new potential therapeutic biomarkers for HB, providing novel insights into important and viable targets in future HB treatment.

Detailed Molecular Mechanism and Potential Drugs for COL1A1 in Carboplatin-Resistant Ovarian Cancer

Carboplatin resistance in ovarian cancer (OV) is a major medical problem. Thus, there is an urgent need to find novel therapeutic targets to improve the prognosis of patients with carboplatin-resistant OV. Accumulating evidence indicates that the gene COL1A1 (collagen type I alpha 1 chain) has an important role in chemoresistance and could be a therapeutic target. However, there have been no reports about the role of COL1A1 in carboplatin-resistant OV. This study aimed to establish the detailed molecular mechanism of COL1A1 and predict potential drugs for its treatment. We found that COL1A1 had a pivotal role in carboplatin resistance in OV by weighted gene correlation network analysis and survival analysis. Moreover, we constructed a competing endogenous RNA network (LINC00052/SMCR5-miR-98-COL1A1) based on multi-omics data and experiments to explore the upstream regulatory mechanisms of COL1A1. Two key pathways involving COL1A1 in carboplatin resistance were identified by co-expression analysis and pathway enrichment: the "ECM-receptor interaction" and "focal adhesion" Kyoto Encyclopedia of Genes and Genomes pathways. Furthermore, combining these results with those of cell viability assays, we proposed that ZINC000085537017 and quercetin were potential drugs for COL1A1 based on virtual screening and the TCMSP database, respectively. These results might help to improve the outcome of OV in the future.

Identification of Potential Biomarkers Associated with Prognosis in Gastric Cancer via Bioinformatics Analysis

Background

Gastric cancer (GC) is one of the leading causes of cancer-related mortality worldwide. We aimed to identify differentially expressed genes (DEGs) and their potential mechanisms associated with the prognosis of GC patients.

Material/Methods

This study was based on gene profiling information for 37 paired samples of GC and adjacent normal tissues from the GSE118916, GSE79973, and GSE19826 datasets in the Gene Expression Omnibus database. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were used to investigate the biological role of the DEGs. The protein–protein interaction (PPI) network was constructed by Cytoscape, and the Kaplan-Meier plotter was used for prognostic analysis.

Results

We identified 119 DEGs, including 21 upregulated and 98 downregulated genes, in GC. The 21 upregulated genes were mainly enriched in extracellular matrix-receptor interaction, focal adhesion, and transforming growth factor-β signaling, while the 98 downregulated genes were significantly associated with gastric acid secretion, retinol metabolism, and metabolism of xenobiotics by cytochrome P450. Thirty hub DEGs were obtained for further analysis. Twenty-five of the 30 hub DEGs were significantly associated with the prognosis of GC, and 21 of the 25 hub DEGs showed consistent expression trends within the 3 profile datasets. KEGG reanalysis of these 21 hub DEGs showed that COL1A1, COL1A2, COL2A1, COL11A1, THBS2, and SPP1 were mainly enriched in the extracellular matrix-receptor interaction pathways.

Conclusions

We identified 6 genes that were significantly related to the prognosis of GC patients. These genes and pathways could serve as potential prognostic markers and be used to develop treatments for GC patients.

Bioinformatics Analysis of Key Genes and Pathways of Cervical Cancer

Background and Objective

Globally, cervical cancer (CC) is the fourth most common cancer affecting women. Although effective screening reduces its incidence, it remains one of the most serious cancers threatening the health of women. Therefore, the purpose of this study is to find new genes that can be used as potential biomarkers for the prognosis of CC.

Methods and Results

After downloading three datasets such as GSE6791, GSE63678, and GSE63514 from the Gene Expression Omnibus (GEO), we combined the expression matrixes and analyzed them to obtain the differential expressed genes (DEGs). Next, using the STRING website, we performed the protein interaction network analysis. Subsequently, hub genes were screened using the R and Cytoscape software. Then, the expression difference and survival analyses of the hub genes were confirmed using GIPIA. Here, we established that the KNTC1 gene was correlated to the overall survival prognosis of CC. Besides, the expression of the KNTC1 gene in the GSE63514 dataset was significantly different from that of the normal cervix, cervical pre-cancerous lesions, and CC. Consequently, immunohistochemistry analysis showed that the results have a definite diagnostic value.

Conclusion

The KNTC1 gene could be linked with the pathophysiology of CC and maybe one of the early diagnostic markers for the diagnosis of cervical pre-cancerous lesions.

TAF1A and ZBTB41 serve as novel key genes in cervical cancer identified by integrated approaches

Cervical cancer (CC) is the second most common cancer and the leading cause of cancer mortality in women. Numerous studies have found that the development of CC was associated with multiple genes. However, the mechanisms on gene level are enigmatic, hindering the understanding of its functional roles. This study sought to identify prognostic biomarkers of CC, and explore their biological functions. Here we conducted an integrated analysis to screen potential vital genes. Candidate genes were further tested by experiments in clinical specimens and cancer cell line. Then, molecular modeling was used to predict the three-dimensional structure of candidate genes’ proteins, and the interaction pattern was analyzed by docking simulation technique. Among the potential genes identified, we found that TAF1A and ZBTB41 were highly correlated. Furthermore, there was a definite interaction between the proteins of TAF1A and ZBTB41, which was affected by the activity of the p53 signaling pathway. In conclusion, our findings identified TAF1A and ZBTB41 could serve as biomarkers of CC. We confirmed their biological function and deciphered their interaction for the first time, which may be helpful for developing further researches.

Significance of hub genes and immune cell infiltration identified by bioinformatics analysis in pelvic organ prolapse

Objective

Pelvic organ prolapse (POP) refers to the decline of pelvic organ position and dysfunction caused by weak pelvic floor support. The aim of the present study was to screen the hub genes and immune cell infiltration related to POP disease.

Methods

Microarray data of 34 POP tissues in the GSE12852 gene expression dataset were used as research objects. Weighted gene co-expression network analysis (WGCNA) was performed to elucidate the hub module and hub genes related to POP occurrence. Gene function annotation was performed using the DAVID tool. Differential analysis based on the GSE12852 dataset was carried out to explore the expression of the selected hub genes in POP and non-POP tissues, and RT-qPCR was used to validate the results. The differential immune cell infiltration between POP and non-POP tissues was investigated using the CIBERSORT algorithm.

Results

WGCNA revealed the module that possessed the highest correlation with POP occurrence. Functional annotation indicated that the genes in this module were mainly involved in immunity. ZNF331, THBS1, IFRD1, FLJ20533, CXCR4, GEM, SOD2, and SAT were identified as the hub genes. Differential analysis and RT-qPCR demonstrated that the selected hub genes were overexpressed in POP tissues as compared with non-POP tissues. The CIBERSORT algorithm was employed to evaluate the infiltration of 22 immune cell types in POP tissues and non-POP tissues. We found greater infiltration of activated mast cells and neutrophils in POP tissues than non-POP tissues, while the infiltration of resting mast cells was lower in POP tissues. Moreover, we investigated the relationship between the type of immune cell infiltration and hub genes by Pearson correlation analysis. The results indicate that activated mast cells and neutrophils had a positive correlation with the hub genes, while resting mast cells had a negative correlation with the hub genes.

Conclusions

Our research identified eight hub genes and the infiltration of three types of immune cells related to POP occurrence. These hub genes may participate in the pathogenesis of POP through the immune system, giving them a certain diagnostic and therapeutic value.

Systematic prediction of key genes for ovarian cancer by co-expression network analysis

Ovarian cancer (OC) is the most lethal gynaecological malignancy, characterized by high recurrence and mortality. However, the mechanisms of its pathogenesis remain largely unknown, hindering the investigation of the functional roles. This study sought to identify key hub genes that may serve as biomarkers correlated with prognosis. Here, we conduct an integrated analysis using the weighted gene co‐expression network analysis (WGCNA) to explore the clinically significant gene sets and identify candidate hub genes associated with OC clinical phenotypes. The gene expression profiles were obtained from the MERAV database. Validations of candidate hub genes were performed with RNASeqV2 data and the corresponding clinical information available from The Cancer Genome Atlas (TCGA) database. In addition, we examined the candidate genes in ovarian cancer cells. Totally, 19 modules were identified and 26 hub genes were extracted from the most significant module (R² = .53) in clinical stages. Through the validation of TCGA data, we found that five hub genes (COL1A1, DCN, LUM, POSTN and THBS2) predicted poor prognosis. Receiver operating characteristic (ROC) curves demonstrated that these five genes exhibited diagnostic efficiency for early‐stage and advanced‐stage cancer. The protein expression of these five genes in tumour tissues was significantly higher than that in normal tissues. Besides, the expression of COL1A1 was associated with the TAX resistance of tumours and could be affected by the autophagy level in OC cell line. In conclusion, our findings identified five genes could serve as biomarkers related to the prognosis of OC and may be helpful for revealing pathogenic mechanism and developing further research.