Exploring the potential role of ENPP2 in polycystic ovary syndrome and endometrial cancer through bioinformatic analysis

Abstract

Background: Growing evidence indicates a significant correlation between polycystic ovary syndrome (PCOS) and endometrial carcinoma (EC); nevertheless, the fundamental molecular mechanisms involved continue to be unclear.

Methods: Initially, differential analysis, the least absolute shrinkage and selection operator (LASSO) regression, and support vector machine-recursive feature elimination (SVM-RFE) algorithms were employed to identify candidate genes associated with ferroptosis in PCOS. Subsequently, the TCGA-UCEC data were utilized to pinpoint the core gene. Then, the expression of ENPP2 in granulosa cells and endometrium of PCOS was validated using real-time PCR (RT-qPCR). Additionally, we investigated the role of ENPP2 in the progression from PCOS to EC through western blotting (WB), colony formation assay, cell scratch assay, transwell assay, and immunofluorescence (IF). Subsequently, ENPP2 gene set enrichment analysis (GSEA) analyses were conducted to identify common pathways involved in PCOS and EC, which were then verified by RT-qPCR. Finally, immune infiltration and the tumor microenvironment (TME) were explored to examine the involvement of ENPP2 in EC progression.

Results: The datasets TCGA-UCEC (pertaining to EC), GSE34526, GSE137684, and GSE6798 (related to PCOS) were procured and subjected to analysis. The gene ENPP2 has been recognized as the shared element connecting PCOS and EC. Next, we observed a significant downregulation of ENPP2 expression in the granulosa cells in PCOS compared to the normal patients, while an upregulation of ENPP2 expression was observed in the endometrium of hyperandrogenic PCOS patients relative to the normal. In vitro, the WB revealed that 5-dihydrotestosterone (DHT) upregulated ENPP2 expression in Ishikawa and HEC-1-A cells. Additionally, we found that ENPP2 promoted the proliferation, migration, and invasion of Ishikawa and HEC-1-A cells. Subsequently, we discovered that overexpressed ENPP2 may lead to an increase in CYP19A1 (aromatase) and AR mRNA level. IF demonstrated that ENPP2 increased the expression of AR, suggesting a regulatory role for ENPP2 in hormonal response within PCOS and EC. Our findings indicated a significant correlation between ENPP2 expression and the modulation of immune responses.

Keywords: Bioinformatic analysis; ENPP2; Endometrial cancer; Ferroptosis; Polycystic ovary syndrome.

Figure 1. Flow chart of this study.

Abbreviations: PCOS, polycystic ovary syndrome; GEO, Gene Expression Omnibus; LASSO, least absolute shrinkage and selection operator; SVM-RFE, support vector machine–recursive feature elimination; TCGA, the cancer genome atlas; UCEC, uterine corpus endometrial carcinoma; CGs, candidate genes; FRGs, ferroptosis-related genes; DEGs, differentially expressed genes; TME, tumor microenvironment.

Figure 2. Volcano plot and heatmap of DEGs.

(A)The volcano plot shows the DEGs in integrated dataset (the red dots represent the upregulated genes, and the blue dots represent the downregulated genes). (B) Supervised hierarchical clustering for 741 unique probe sets was performed by Ward’s method and Euclidean Distance as the distance measure. Each row represents a single gene; each column represents a sample. Expression values were color coded: blue, transcript level below the median; and red, greater than median.

Figure 3. View of nine CGs in PCOS.

(A) The intersection of DEGs in PCOS and ferroptosis related genes. A Venn diagram was constructed using VennDiagram package. (B) Boxplot of the mRNA levels of nine CGs in PCOS samples. Boxplots are determined by the spacing between quarterbacks, with the median line representing the median and the whisker being 1.5 times the quarterback spacing. T test for difference significance. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 4. Diagnostic biomarkers of PCOS.

(A) Optimal lambda value was selected in the LASSO regression model based on 10-fold cross-validation. (B) Line graph shows the cross-validated accuracy based on different numbers of CGs in the SVM-RFE model. (C) The overlap genes of LASSO and SVM-RFE algorithms.

Figure 5. Construction of the nomogram model.

(A) Construction of the nomogram model based on HAMP, ENPP2, and HMOX1. (B) Calibration curve showed the accuracy of the nomogram model in the diagnosis of PCOS. (C) Decision curve analysis (DCA) indicated that the nomogram model had better clinical application value in the diagnosis of PCOS. (D) Clinical impact curve (CIC) suggested that the nomogram model had higher clinical application value in the diagnosis of PCOS.

Figure 6. Validating HAMP, ENPP2, and HMOX1 differential expression in the GSE6798 dataset.

Boxplots are determined by the spacing between quarterbacks, with the median line representing the median and the whisker being 1.5 times the quarterback spacing. T test for difference significance. ***p < 0.001.

Figure 7. Identification of the core gene.

(A) Different expression of HAMP, ENPP2, and HMOX1 in TCGA-UCEC cohort. Boxplots are determined by the spacing between quarterbacks, with the median line representing the median and the whisker being 1.5 times the quarterback spacing. T test for difference significance. (B) Receiver operating characteristic (ROC) curves were used to verify the accuracy of three CGs predictions (HAMP, ENPP2, HMOX1) in PCOS and TCGA-UCEC cohort. (C) qRT-PCR and western blot validated ENPP2 expression on granulosa cells in normoandrogen (NA)/hyperandrogen (HA) PCOS and normal women. (D) qRT-PCR and western blot validated ENPP2 expression on endometrium in NA/HA PCOS and normal women. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 8. DHT regulates the expression of ENPP2 in ISK and HEC-1-A cells.

(A) The transfection efficiency of AR decreased in mRNA level by qRT-PCR assay (B) 100 nM DHT-induced Up-regulation the expression of ENPP2 in ISK cells. (C) 100 nM DHT-induced Up-regulation the expression of ENPP2 in HEC-1-A cells. **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 9. Effect of ENPP2 on ISK cells proliferation, invasion and migration.

(A) The mRNA levels in Overexpression normal controls (OE-NC) and ENPP2 overexpression (OE-ENPP2) ISK cells. (B) OE-ENPP2 ISK cells were continuously cultured for 10 days or until more than 50 monoclonal cells had grown, and 10 nM ONO-8430506 was added to inhibit ENPP2 expression. (C) Migration rates of ISK cells at 12 and 24 h. (D) Transwell invasion rates of ISK cells at 24 h. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 10. ENPP2 expression and function.

(A, B) ENPP2 single-gene GSEA in PCOS and TCGA-UCEC. (C) AR and CYP19A1 mRNA levels in OE-ENPP2 ISK cells before and after treatment with 10 nM ONO-8430506. (D) ENPP2 and AR correlation analysis in GEPIA database. (D) The effect of ENPP2 on AR was assessed by immunofluorescence. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 11. The correlation analysis between TME score and ENPP2.

*p < 0.05, ***p < 0.001.

Figure 12. Association between ENPP2 and immune infiltration expression in TCGA-UCEC cohort.

(A, B) Correlation of ENPP2 expression with infiltration levels of Endothelial, T cell regulatory, Cancer associated fibroblast cell, Hematopoietic stem cell_XCELL, and Mast cell activated in EC available at TIMER2.0 database. (C–F) Correlation between ENPP2 expression and chemokines and chemokines receptors in EC available at TISIDB database. (G–J) Correlation between ENPP2 expression and immunostimulatory and immunoinhibitory in EC available at TISIDB database.