Identification of pyroptosis-related lncRNA signature and AC005253.1 as a pyroptosis-related oncogene in prostate cancer

Abstract

Background: Pyroptosis and prostate cancer (PCa) are closely related. The role of pyroptosis-related long non-coding RNAs (lncRNAs) (PRLs) in PCa remains elusive. This study aimed to explore the relationship between PRL and PCa prognosis.

Methods: Gene expression and clinical signatures were obtained from The Cancer Genome Atlas and Gene Expression Omnibus databases. A PRL risk prediction model was established by survival random forest analysis and least absolute shrinkage and selection operator regression. Functional enrichment, immune status, immune checkpoints, genetic mutations, and drug susceptibility analyses related to risk scores were performed by the single-sample gene set enrichment analysis, gene set variation analysis, and copy number variation analysis. PRL expression was verified in PCa cells. Cell Counting Kit-8, 5-ethynyl-2'-deoxyuridine, wound healing, transwell, and Western blotting assay were used to detect the proliferation, migration, invasion, and pyroptosis of PCa cells, respectively.

Results: Prognostic features based on six PRL (AC129507.1, AC005253.1, AC127502.2, AC068580.3, LIMD1-AS1, and LINC01852) were constructed, and patients in the high-score group had a worse prognosis than those in the low-score group. This feature was determined to be independent by Cox regression analysis, and the area under the curve of the 1-, 3-, and 5-year receiver operating characteristic curves in the testing cohort was 1, 0.93, and 0.92, respectively. Moreover, the external cohort validation confirmed the robustness of the PRL risk prediction model. There was a clear distinction between the immune status of the two groups. The expression of multiple immune checkpoints was also reduced in the high-score group. Gene mutation proportion in the high-score group increased, and the sensitivity to drugs increased significantly. Six PRLs were upregulated in PCa cells. Silencing of AC005253.1 inhibited cell proliferation, migration, and invasion in DU145 and PC-3 cells. Moreover, silencing of AC005253.1 promoted pyroptosis and inflammasome AIM2 expression.

Conclusions: Overall, we constructed a prognostic model of PCa with six PRLs and identified their expression in PCa cells. The experimental verification showed that AC005253.1 could affect the proliferation, migration, and invasion abilities of PCa cells. Meanwhile, AC005253.1 may play an important role in PCa by affecting pyroptosis through the AIM2 inflammasome. This result requires further research for verification.

Keywords: immunity; lncRNA; machine learning; prostate cancer; pyroptosis; tumor biomarkers.

Figure 1

Construction of pyroptosis-related lncRNA signatures. (A) Correlation gene screens for genes of univariate significance. (B, C) Six PRL signatures were constructed through a random forest and Lasso analysis. (D) Survival curves. (E) ROC curves. (F) Clinical feature scores. lncRNA, long non-coding RNA; PRL, pyroptosis-related lncRNA; Lasso, least absolute shrinkage and selection operator; ROC, receiver operating characteristic.

Figure 2

Correlation of risk scores with pyroptosis genes and immune infiltration. (A) The expression of LIMD1-AS1, AC127502.2, AC005253.1, AC068580.3, LINC01852, and AC129507.1. (B) Expression correlation plots of risk scores and model genes. (C) Heatmap of risk score associated with pyroptosis genes. (D) Heatmap of the relationship of the risk score to immune infiltration. *p < 0.05. **p < 0.01. ***p < 0.001 ****p < 0.0001.

Figure 3

Immune checkpoint. Immune checkpoint molecule expression in low- and high-score groups. *p < 0.05. **p < 0.01. ***p < 0.001 ****p < 0.0001. ns, not significant.

Figure 4

Functional analysis of risk score. (A) Heatmap for GO and KEGG analyses using GSVA package. (B) Correlation of risk score with functional enrichment pathways. GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; GSVA, gene set variation analysis. *p < 0.05. **p < 0.01. ***p < 0.001 ****p < 0.0001.

Figure 5

CNV and mutation analysis in high- and low-score groups. (A) CNV maps for groups with high and low score. (B) Top 30 gene mutation frequencies in two groups. CNV, copy number variation. *p < 0.05

Figure 6

Drug sensitivity analysis of risk scores in two groups. Information on the sensitivity of tumor cell lines to potential drugs was downloaded from CTRP v2 and PRISM. Prediction of drug susceptibility in two groups using pRRophetic. The lower the AUC of the cell line, the higher the sensitivity to the potential drug. AUC, area under the curve. ***p < 0.001

Figure 7

Risk score gene expression identification. RT-qPCR to detect the AC129507.1, AC005253.1, AC127502.2, AC068580.3, LIMD1-AS1, and LINC01852 expression in RWPE1, PC-3, and DU145 cell. *p < 0.05, vs. RWPE1.

Figure 8

Silencing of AC005253.1 affected PCa cell proliferation, migration, and invasion. (A) RT-qPCR detection of AC005253.1 expression in PC-3 and DU145 cells. (B) CCK-8 assay was used to measure the cell viability in PC-3 and DU145 cells. (C) EDU assay results showed the effect of si-AC005253.1 on cell proliferation. (D) Wound healing assay was performed to detect the migration in PC-3 and DU145 cells. (E) Transwell assay was used to detect the invasion of PC-3 and DU145 cells. #p < 0.05, vs. si-NC group. PCa, prostate cancer; CCK-8, Cell Counting Kit-8.

Figure 9

Silencing of AC005253.1 promoted pyroptosis of PCa cells. (A) AIM2, NLRC4, and NLRP3 levels were identified by Western blotting. (B) GSDMD-N, ASC, caspase-1, IL-18, and IL-1β proteins were identified by Western blotting in PC-3 and DU145 cells. #p < 0.05, vs. si-NC group. PCa, prostate cancer.