Construction of WGCNA network and clinical prediction model of lncRNA WAC-AS1 in breast cancer based on TCGA database
汪彦阳,周远洋,魏雪菲,陶志嵩,龚海燕
DOI: https://doi.org/10.3877/cma.j.issn.1674-0807.2023.04.004
2023-01-01
Abstract:Objective:To explore the biological function of long non-coding RNA (lncRNA) WAC antisense RNA1(WAC-AS1) in breast cancer and its impact on the prognosis of breast cancer based on the Cancer Genome Atlas (TCGA) and Gene Tissue Expression (GTE) databases.Methods:Based on the data of lncRNA WAC-AS1 expression in the TCGA and GTE databases, the expression of WAC-AS1 in breast cancer tissue was analyzed and compared with normal breast tissues. The breast cancer patients were divided into high expression group and low expression group according to the median value of WAC-AS1 expression. The OS, progression-free survival (PFS), disease-specific survival (DSS), DFS, and proportion of immune cells infiltrated in cancer tissues were compared between the two groups. WAC-AS1 related gene mutations were analyzed using the tumor somatic mutation detection tool VarScan. Gene set variation analysis (GSVA) and gene set enrichment analysis (GSEA) were also performed to explore WAC-AS1-involved signaling pathways in breast cancer. Finally, weighted gene correlation network analysis (WGCNA) was performed and a clinical prognostic model was constructed in breast cancer. The receiver operating characteristic (ROC) curve and the calibration curve of 3-year and 5-year OS were used to evaluate the prediction efficiency and accuracy of the model. The expression of WAC-AS1 was detected by fluorescence-based quantitative real-time PCR in MCF-10A, MCF-7 and MDA-MB-231 cells.Results:(1) WAC-AS1 expression in breast cancer tissue was significantly higher compared with normal breast tissues [4.17 (3.91, 4.41) vs 3.70 (3.37, 4.09), Z=3.880, P<0.001]. (2) The median OS in WAC-AS1 high expression group and low expression group was 10.0 years and 17.2 years, respectively, indicating a significant difference (HR=1.680, 95%CI: 1.208-2.338, P=0.002). There was no significant difference in PFS, DSS and DFS between those two groups(HR=1.105, 95%CI: 0.798-1.529, P=0.548; HR=1.303, 95%CI: 0.846-2.008, P=0.230; HR=1.092, 95%CI: 0.711-1.678, P=0.687). In breast cancer tissues with high expression of WAC-AS1, 4 kinds of immune cells (naive B cells, CD8+ T cells, follicular helper T cells and activated dendritic cells) were increased, while resting CD4+ memory T cells and resting mast cells were decreased (all P<0.050). (3)The genes with the highest frequency of WAC-AS1 related mutations were TP53, PIKCA and TTN. (4) The results of GSVA and GSEV showed a positive correlation between WAC-AS1 and DNA repair, MYC target V2, MYC target V1, E2F target, and mTORC1 signaling pathway. (5) WGCNA analysis found that WAC-AS1 had the highest correlation with 91 genes related to small molecule biosynthesis in the yellow green module (r=0.270, P<0.001). (6) The results of univariate analysis showed that age (HR=1.034, 95%CI: 1.021-1.047, P<0.001), gender (female vs male, HR=1.388, 95%CI: 0.188-9.929, P=0.870), clinical stage (Phase Ⅱ vs Phase Ⅰ, HR=1.457, 95%CI: 1.043-2.022, P=0.017; Phase Ⅲ vs Phase Ⅰ, HR=4.022, 95%CI: 2.804-5.739, P<0.001; and Phase Ⅳ vs Phase I, HR=16.130, 95%CI: 9.413-27.630, P<0.001) and WAC-AS1 expression (HR=1.032, 95%CI: 1.005-1.061, P=0.020) were influencing factors for OS. Multivariate analysis showed that the expression of WAC-AS1 was correlated with OS in breast cancer patients (HR=1.377, 95% CI: 1.021-1.872, P=0.039). (7) The C-index of the constructed clinical prediction model was 0.759, the area under the ROC curve (AUC) was 0.626 (95%CI: 0.58.0-0.673, P<0.001), the sensitivity was 76.9%, and the specificity was 46.8%. The predicted curve fit well with the ideal curve. (8) The CT values of WAC-AS1 detected by PCR in three cell lines MCF-10A, MCF-7, and MDA-MB-231 were 32.39±0.10, 30.55±0.25, and 30.82±0.07, respectively, indicating a significant difference (F=30.310, P<0.001). Pairwise comparison showed that WAC-AS1 expression in MCF-7 and MDA-MB-231 cells was significantly higher than that in MCF-10A cells (t=7.916, 7.431, both P<0.001).Conclusions:The lncRNA WAC-AS1 is related to the tumor microenvironment and immune infiltration in breast cancer. The high expression of WAC-AS1 affects the prognosis of breast cancer pateints. WAC-AS1 may be a potential target and prognostic marker in breast cancer treatment.