Please wait a minute...
European Journal of Gynaecological Oncology  2021, Vol. 42 Issue (1): 50-65    DOI: 10.31083/j.ejgo.2021.01.2151
Original Research Previous articles | Next articles
Bioinformatic analysis identifies potential key genes in the pathogenesis of uterine leiomyoma
Yi-Chao Jin1, Tong-Hui Ji1, Xiong Yuan1, Ying Sun1, Yu-Jie Sun2, Jie Wu1, *()
1Department of Gynecology, The First Affiliated Hospital of Nanjing Medical University, Nanjing, 210029 Jiangsu, P. R. China
2Key laboratory of Human Functional Genomics of Jiangsu Province, Nanjing Medical University, Nanjing, 211126 Jiangsu, P. R. China
Download:  PDF(9049KB)  ( 57 ) Full text   ( 18 )
Export:  BibTeX | EndNote (RIS)      

Objective: The present study aimed to screen hub genes for pathology of uterine leiomyoma. Methods: The microarray data of GSE31699, containing 16 uterine leiomyoma tissue samples and 16 matched normal myometrium samples, were downloaded from the Gene Expression Omnibus database (GEO). The “limma” R language package was used to identify differently-expressed genes (DEGs) between uterine leiomyoma and myometrium. Gene Ontology (GO) and pathway enrichment analyses were performed by using clusterprofiler, the DEGs were mostly enriched in post-synapse assembly, response to glucocorticoid, extracellular matrix receptor interaction and coagulation cascades. Subsequently, a protein-protein interaction (PPI) network of DEGs was constructed by Search Tool for the Retrieval of Interacting Genes Database (STRING) and visualized by utilizing Cytoscape software. We screened hub clusters of PPI network by the plug-in Molecular Complex Detection (MCODE) in Cytoscape, then clusterprofiler was also utilized to analyze functions and pathways enrichment of the genes in the hub clusters. Furthermore, we employed the “WGCNA” package in R to conduct co-expression network for all genes in GSE31699. Ultimately, we selected the overlapped genes in hub clusters of DEGs’ PPI network and WGCNA. Results: Five genes (COL5A2, ALDH1A1, GNG11, EFEMP1, ANXA1) were finally validated in other GEO datasets (GSE64763, GSE764, GSE593) and Oncomine database. Gene set enrichment analysis (GSEA) was also performed for the hub genes. The expression of COL5A2 was significantly higher in uterine leiomyoma compared with that in myometrium, while the expression of the other hub genes was significantly lower in uterine leiomyoma. Conclusion: The results indicated that COL5A2, ALDH1A1, GNG11, EFEMP1 and ANXA1 may be the key pathological genes in uterine leiomyoma.

Key words:  Bioinformatics analysis      Uterine leiomyoma      PPI      WGCNA     
Submitted:  25 May 2020      Revised:  12 August 2020      Accepted:  28 August 2020      Published:  15 February 2021     
*Corresponding Author(s):  Jie Wu     E-mail:

Cite this article: 

Yi-Chao Jin, Tong-Hui Ji, Xiong Yuan, Ying Sun, Yu-Jie Sun, Jie Wu. Bioinformatic analysis identifies potential key genes in the pathogenesis of uterine leiomyoma. European Journal of Gynaecological Oncology, 2021, 42(1): 50-65.

URL:     OR

Fig. 1.  Study design and the flow diagram of study.

Fig. 2.  Heatmap of the top 273 DEGs according to the value of |logFC| after deleting the samples GSM786796 and GSM786789.

Fig. 3.  KEGG and GO enrichment of DEGs. (A) GO enrichment of the up-regulated DEGs. (B) GO enrichment of the down-regulated DEGs. (C) KEGG analysis of the up-regulated DEGs. (D) KEGG analysis of the down-regulated DEGs.

Fig. 4.  PPI network construction and clusters analyses. The red nodes represent the up-regulated genes and the blue nodes represent the down-regulated genes. (A) The PPI network of 273 DEGs was constructed via STRING that contained 267 nodes and 555 edges. (B) Cluster rank 1. This cluster consists of 24 nodes and 69 edges and has the highest score in those clusters. (C) Cluster rank 2. (D) Cluster rank 3. (E) Cluster rank 4.

Fig. 5.  Co-expression network creation and hub modules selection. (A) Dendrogram of all genes in GSE31699 clustered based on a dissimilarity measure (1-TOM). (B) A heatmap of selected genes. The intensity of the yellow color indicates the strength of the correlation between pairs of modules on a linear scale. (C) Correlation between modules and traits. The upper number in each cell refers to the correlation coefficient of each module in the trait, and the lower number is the corresponding P-value. Among them, the blue module was the most relevant modules with uterine leiomyoma traits.

Fig. 6.  Clustering of module eigengenes and eigengene adjacency heatmap (the red color indicates the strong correlation between different modules).

Fig. 7.  Scatter plots of GS for uterine leiomyoma versus the MM in the hub modules. (A) Blue module. (B) Turquoise module. (C) Tan module. (D) Cyan module.

Fig. 8.  PPI network construction of hub modules and identification of hub genes. (A) PPI network for genes in blue module. (B) PPI network for genes in turquoise module. (C) PPI network for genes in tan module. (D) Real hub genes belonging to both the hub modules and the hub clusters in PPI network of DEGs.

Fig. 9.  The relative expression of hub genes in other datasets from GEO (GSE64763, GSE764 and GSE593, *: P < 0.01, **: P < 0.001, ***: P < 0.0001, ****: P < 0.00001).

Fig. 10.  Transcriptional expression of hub genes in 20 different types of cancer diseases (ONCOMINE database). ‘Uterine corpus leiomyoma’ was included in ‘other cancer’. Difference of transcriptional expression was compared by students’ t-test. Cut-off of P value and fold change were as following: P value: 0.01, fold change: 1.5, gene rank: 10%, data type: mRNA. The number in cell means how many studies in the database meet the threshold. Red means higher expression in cancer while blue means lower expression.

Fig. 11.  Gene set enrichment analysis (GSEA) using GSE31699. The most enriched functional gene set in uterine leiomyoma samples with hub genes highly expressed was identified. (A) COL5A2. (B) ALDH1A1. (C) GNG11. (D) EFEMP1. (E) ANXA1.

[1] Bulun SE. Uterine fibroids. New England Journal of Medicine. 2013; 369: 1344-1355.
[2] Giuliani E, As‐Sanie S, Marsh EE. Epidemiology and management of uterine fibroids. International Journal of Gynecology & Obstetrics. 2020; 149: 3-9.
[3] Baranov VS, Osinovskaya NS, Yarmolinskaya MI. Pathogenomics of uterine fibroids development. International Journal of Molecular Sciences. 2019; 20: 6151.
[4] Stewart EA, Laughlin-Tommaso SK, Catherino WH, Lalitkumar S, Gupta D, Vollenhoven B. Uterine fibroids. Nature Reviews Disease Primers. 2016; 2: 16043.
[5] Islam MS, Ciavattini A, Petraglia F, Castellucci M, Ciarmela P. Extracellular matrix in uterine leiomyoma pathogenesis: a potential target for future therapeutics. Human Reproduction Update. 2018; 24: 59-85.
[6] McWilliams MM, Chennathukuzhi VM. Recent advances in uterine fibroid etiology. Seminars in Reproductive Medicine. 2017; 35: 181-189.
[7] Liu X, Liu Y, Zhao J, Liu Y. Screening of potential biomarkers in uterine leiomyomas disease via gene expression profiling analysis. Molecular Medicine Reports. 2018; 17: 6985-6996.
[8] Kim YJ, Kim YY, Shin JH, Kim H, Ku SY, Suh CS. Variation in MicroRNA Expression Profile of Uterine Leiomyoma with Endometrial Cavity Distortion and Endometrial Cavity Non-Distortion. International Journal of Molecular Sciences. 2018; 19: 2524.
[9] Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2009; 9: 559.
[10] Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Research. 2015; 43: e47-e47.
[11] Yu G, Wang L, Han Y, He Q. ClusterProfiler: an R package for comparing biological themes among gene clusters. A Journal of Integrative Biology. 2012; 16: 284-287.
[12] Szklarczyk D, Franceschini A, Wyder S, Forslund K, Heller D, Huerta-Cepas J, et al. STRING v10: protein-protein interaction networks, integrated over the tree of life. Nucleic Acids Research. 2015; 43: D447-D452.
[13] Shannon P. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Research. 2003; 13: 2498-2504.
[14] Bader GD, Hogue CWV. An automated method for finding molecular complexes in large protein interaction networks. BMC Bioinformatics. 2003; 4: 2.
[15] Horvath S, Dong J. Geometric interpretation of gene coexpression network analysis. PLoS Computational Biology. 2008; 4: e1000117.
[16] Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proceedings of the National Academy of Sciences. 2005; 102: 15545-15550.
[17] Mootha VK, Lindgren CM, Eriksson K-F, Subramanian A, Sihag S, Lehar J, et al. PGC-1α-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nature Genetics. 2003; 34: 267-273.
[18] Barlin JN, Zhou QC, Leitao MM, Bisogna M, Olvera N, Shih KK, et al. Molecular subtypes of uterine leiomyosarcoma and correlation with clinical outcome. Neoplasia. 2015; 17: 183-189.
[19] Quade BJ, Wang T, Sornberger K, Dal Cin P, Mutter GL, Morton CC. Molecular pathogenesis of uterine smooth muscle tumors from transcriptional profiling. Genes, Chromosomes & Cancer. 2004; 40: 97-108.
[20] Hoffman PJ, Milliken DB, Gregg LC, Davis RR, Gregg JP. Molecular characterization of uterine fibroids and its implication for underlying mechanisms of pathogenesis. Fertility and Sterility. 2004; 82: 639-649.
[21] Malik M, Norian J, McCarthy-Keith D, Britten J, Catherino W. Why leiomyomas are called fibroids: the central role of extracellular matrix in symptomatic women. Seminars in Reproductive Medicine. 2010; 28: 169-179.
[22] Giri A, Edwards TL, Hartmann KE, Torstenson ES, Wellons M, Schreiner PJ, et al. African genetic ancestry interacts with body mass index to modify risk for uterine fibroids. PLoS Genetics. 2017; 13: e1006871.
[23] Zeng X, Liu X, Liu T, Wang X. The clinical significance of COL5a2 in patients with bladder cancer: a retrospective analysis of bladder cancer gene expression data. Medicine. 2018; 97: e0091.
[24] Cao L, Chen Y, Zhang M, Xu D, Liu Y, Liu T, et al. Identification of hub genes and potential molecular mechanisms in gastric cancer by integrated bioinformatics analysis. PeerJ. 2019; 6: e5180.
[25] Zaitseva M, Vollenhoven BJ, Rogers PAW. Retinoic acid pathway genes show significantly altered expression in uterine fibroids when compared with normal myometrium. Molecular Human Reproduction. 2007; 13: 577-585.
[26] Xia L, Liu Y, Fu Y, Dongye S, Wang D. Integrated analysis reveals candidate mRNA and their potential roles in uterine leiomyomas. Journal of Obstetrics and Gynaecology Research. 2017; 43: 149-156.
[27] Zaitseva M, Vollenhoven BJ, Rogers PAW. Retinoids regulate genes involved in retinoic acid synthesis and transport in human myometrial and fibroid smooth muscle cells. Human Reproduction. 2008; 23: 1076-1086.
[28] Shveiky D, Shushan A, Ben Bassat H, Klein BY, Ben Meir A, Levitzky R, et al. Acetaldehyde differentially affects the growth of uterine leiomyomata and myometrial cells in tissue cultures. Fertility and Sterility. 2009; 91: 575-579.
[29] Downes GB, Gautam N. The G protein subunit gene families. Genomics. 1999; 62: 544-552.
[30] Yang G, Chen Q, Xiao J, Zhang H, Wang Z, Lin X. Identification of genes and analysis of prognostic values in nonsmoking females with non-small cell lung carcinoma by bioinformatics analyses. Cancer Management and Research. 2019; 10: 4287-4295.
[31] Haouas H, Haouas S, Uzan G, Hafsia A. Identification of new markers discriminating between myeloid and lymphoid acute leukemia. Hematology. 2010; 15: 193-203.
[32] Takauji Y, Kudo I, En A, Matsuo R, Hossain MN, Nakabayashi K, et al. GNG11 (G-protein subunit γ 11) suppresses cell growth with induction of reactive oxygen species and abnormal nuclear morphology in human SUSM-1 cells. Biochemistry and Cell Biology. 2017; 95: 517-523.
[33] Hossain MN, Sakemura R, Fujii M, Ayusawa D. G-protein gamma subunit GNG11 strongly regulates cellular senescence. Biochemical and Biophysical Research Communications. 2006; 351: 645-650.
[34] Marsh EE, Chibber S, Wu J, Siegersma K, Kim J, Bulun S. Epidermal growth factor-containing fibulin-like extracellular matrix protein 1 expression and regulation in uterine leiomyoma. Fertility and Sterility. 2016; 105: 1070-1075.
[35] Hu J, Duan B, Jiang W, Fu S, Gao H, Lu L. Epidermal growth factor-containing fibulin-like extracellular matrix protein 1 (EFEMP1) suppressed the growth of hepatocellular carcinoma cells by promoting Semaphorin 3B(SEMA3B). Cancer Medicine. 2019; 8: 3152-3166.
[36] Sheu M, Li C, Lin C, Lee S, Lin L, Chen T, et al. Overexpression of ANXA1 confers independent negative prognostic impact in rectal cancers receiving concurrent chemoradiotherapy. Tumor Biology. 2014; 35: 7755-7763.
[37] Duncan R, Carpenter B, Main LC, Telfer C, Murray GI. Characterisation and protein expression profiling of annexins in colorectal cancer. British Journal of Cancer. 2008; 98: 426-433.
[38] Foo SL, Yap G, Cui J, Lim LHK. Annexin-A1 - a blessing or a curse in cancer? Trends in Molecular Medicine. 2019; 25: 315-327.
[39] Anjum S, Sahar T, Nigam A, Wajid S. Transcriptome analysis of mRNA in uterine leiomyoma using next-generation RNA sequencing. Anti-Cancer Agents in Medicinal Chemistry. 2019; 19: 1703-1718.
[40] Guo C, Liu S, Sun M. Potential role of Anxa1 in cancer. Future Oncology. 2013; 9: 1773-1793.
[1] T. Tomimatsu, S. Mabuchi, T. Tsuboyama, Y. Hori, S. Sekine, T. Kimura. Malignant transformation of uterine leiomyoma: suggested by clinical, imaging, histological, and genetic findings[J]. European Journal of Gynaecological Oncology, 2019, 40(5): 879-882.
[2] Rong Wang, Chunyu Yin, Lei Fu, Jing Liu, Jinbin Li, Ling Yin. Expression profile analysis for epithelial-mesenchymal transition of breast cancer cell line DKTA based on microarray data[J]. European Journal of Gynaecological Oncology, 2019, 40(4): 579-584.
[3] G.M. Barelli, F. Carbonetti, V. De Sanctis, I. Martini, P. Bonome, C. Briani, E. Iannicelli. Diffusion magnetic resonance in cervical carcinoma: the role of apparent diffusion coefficient in the evaluation of treatment response[J]. European Journal of Gynaecological Oncology, 2019, 40(1): 91-96.
[4] S.Y. Yi, Y. Kuang, L.J. Zeng, F.F. Lu, Y. Zhang. Asymptomatic retroperitoneal benign metastasizing leiomyoma after laparoscopic uterine myomectomy: case report and review of the literature[J]. European Journal of Gynaecological Oncology, 2017, 38(6): 971-974.
[5] M.R. Asoglu, A.M. Rodriguez, M.A. Borahay, K. Yong-Fang, G.S. Kilic. Estimating risk for unexpected uterine leiomyosarcoma on the basis of uterine weight and age[J]. European Journal of Gynaecological Oncology, 2017, 38(4): 573-577.
[6] P. Pawłowicz, M. Dąbrowska, R. Bartkowiak, M. Dąbrowski. The role of sentinel node mapping with indocyanine green and endoscopic near-infrared fluorescence imaging in endometrial and cervical cancer[J]. European Journal of Gynaecological Oncology, 2017, 38(3): 441-443.
[7] J.I. Choi, H.J. Lee, Y.J. Shin, H.W. Lim, H.N. Lee, M.J. Kim. Rapid enlargement of endometrial stromal sarcoma after uterine fibroid embolization for presumed adenomyosis: a case report and literature review[J]. European Journal of Gynaecological Oncology, 2016, 37(6): 876-881.
No Suggested Reading articles found!