Integrative high-throughput studies to develop novel targets and drugs for the treatment of advanced prostate cancer

Abstract

Androgen deprivation therapies targeting the androgen receptor (AR) signaling pathway are the primary treatment strategy for prostate cancer. However, these therapies often lead to castration resistance. Developing novel agents targeting AR-independent oncogenes is critical to address this challenge, particularly for advanced castration-resistant prostate cancer. This study identified three potential tumor drivers of advanced prostate cancer, including CDC20, DTL, and RRM2, through integrative bioinformatic screening that considered gene dependency using CRISPRi/RNAi database, clinical relevance, and experimental validation with CRISPR-Cas13-mediated gene ablation. Further mechanistic studies revealed that CDC20, DTL, and RRM2 were transcriptionally regulated by the RB1/E2F1 axis, mediating cell cycle progression in prostate cancer. Additionally, we identified novel agents targeting these candidates through virtual screening and drug-sensitive tests, utilizing our established small-molecule library. These agents exhibited superior anti-tumor efficacy compared with AR antagonists in vitro. Our study identified novel prostate cancer therapeutic targets independent of the AR signaling pathway and established a research paradigm for developing anti-tumor agents through integrative cancer bioinformatics and network pharmacology analysis.

Keywords: Advanced prostate cancer; CDC20; DTL; RB1/E2F1 axis; RRM2; Structure-based virtual screening.

Figure 1

Integrative analysis revealed a prostate cancer-specific signature comprised of seven genes. (A) Visualization of representative gene dependency in 22Rv1 cells with CRISPR-Cas9 screening. (B) The Venn diagram indicates the genes whose ablation significantly suppresses prostate cancer cell growth in different prostate cancer cell models. (C) The heatmap shows the expression of genes differentially expressed in prostate cancer compared with normal prostate tissues in the TCGA-PRAD cohort. (D, E) Protein–protein interaction string (D) and expression correlation (E) of genes that were both dysregulated in prostate cancer and suppressed cell growth after being silenced in the TCGA-PRAD cohort. (F, G) Survival analysis of the gene signature comprised of 15 genes (F) as well as individuals (G) with the TCGA-PRAD prostate cancer cohort. (H) Survival analysis of the core gene signature comprised of 7 genes with the TCGA-PRAD prostate cancer cohort.

Figure 2

Enhanced CDRs correlated with prostate cancer prognosis. (A) Visualization of core genes' dependency in different prostate cancer cells with CRISPR-Cas9 and siRNA screening. (B, C) Survival analysis of CDRs in different CRPC patient cohorts. (D) The relative expression of CDRs in normal prostate tissues and prostate cancer tissues in the TCGA-PRAD cohort. Gleason score and tumor stage correlation analysis of CDRs in different prostate cancer patient cohorts. All prostate patient cohorts were indicated in the corresponding panels. CDRs refers to CDC20 (cell division cycle 20), DTL (denticleless E3 ubiquitin protein ligase), and RRM2 (ribonucleotide reductase M2).

Figure 3

CDRs linked with neuroendocrine features in prostate cancer cohorts. (A) The Venn diagram illustrates the genes specifically enhanced in prostate cancer compared with normal prostate tissues and NEPC compared with adenocarcinoma. (B, C) The relative expression of CDRs in NEPC compared with adenocarcinoma in different prostate cancer cohorts. (D) Correlation analysis of CDRs with NE score. The red indicates the higher NE score. (E, F) Pearson correlation analysis of CDRs with NE score CRPC cohort. (G) The heatmap indicates the relative expression of CDRs and AR, as well as KLK3, a transcription activity indication of AR. CDRs refers to CDC20 (cell division cycle 20), DTL (denticleless E3 ubiquitin protein ligase), and RRM2 (ribonucleotide reductase M2). NEPC, neuroendocrine prostate cancer; NE, neuroendocrine; AR, androgen receptor; KLK3, kallikrein-related peptidase 3.

Figure 4

CDRs ablation suppressed prostate cancer cell proliferation. (A) The circuit diagram illustrates the correlation among CDRs co-expressed gene sets. (B) Venn analysis of the overlap of CDRs co-expressed genes. (C) The heatmap shows the relative expression of CDRs overlapped co-expressed genes in patients with NE signature high and low groups. (D–F) The heatmap illustrates the correlation of CDRs co-expressed genes with CDRs in ADPC (D) and CRPC (E, F) patient cohorts. (G, H) KEGG pathway analysis (G) and GSEA analysis (H) of enriched biological processes of CDRs and CDRs co-expressed genes. (I, J) The diagram illustrates the working model of CRISPR-Cas13 for RNA silencing and knockdown efficiency of CDRs with the corresponding gRNAs. (K, L) Cell growth assays indicate the impact of CDRs knockdown on the viability of different prostate cancer cells. (M) Western blotting analysis of the expression of cell cycle-regulated genes, including CCND1, CDK1, and p-CDK1, after CDRs knockdown. ∗∗p < 0.01. CDRs refers to CDC20 (cell division cycle 20), DTL (denticleless E3 ubiquitin protein ligase), and RRM2 (ribonucleotide reductase M2). NE, neuroendocrine; KEGG, Kyoto Encyclopedia of Genes and Genomes; GSEA, gene set enrichment analysis; CCND1, cyclin D1; CDK1, cyclin-dependent kinase 1.

Figure 5

CDRs were transcriptionally regulated by the RB1/E2F1 axis in prostate cancer. (A) The diagrams show the canonical E2F1 motif location within the promoter of CDRs. (B) ChIP-sequencing E2F1 enrichment peak within the promoter of CDRs. (C) The relative enrichment of E2F1 within the promoter of CDRs was determined using standard ChIP-qPCR. (D) ChIP-sequencing peaks show the relative enrichment of E2F1 in CDRs' promoters after RB1 is known. (E–G) The relative expression of CDRs in CRPC patients with different RB1 deletion status. (H) CDRs expression in patients with RB1 deletion mutations from different prostate cancer cohorts. (I) qRT-PCR detected the relative expression of CDRs after RB1 or E2F1 knockdown with CRISPR-Cas13. ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001. CDRs refers to CDC20 (cell division cycle 20), DTL (denticleless E3 ubiquitin protein ligase), and RRM2 (ribonucleotide reductase M2). RB1, retinoblastoma tumor suppressor 1; E2F1, early 2 factor 1; ChIP, chromatin immunoprecipitation; qRT-PCR, quantitative real-time PCR.

Figure 6

Virtual screening identified compounds that suppressed advanced prostate cancer. (A) The diagram illustrates structure-based virtual screening strategies for CDRs-targeted compounds. (B) The heatmap shows the binding affinity of compounds with CDRs. The red indicated a higher binding affinity of compounds with the corresponding compounds. (C, D) Cell viability assays were used to determine the tumor suppressive effect of compounds with high CDRs-binding affinity in different prostate cancer cell models. (E–G) The 2D structure of Q199, XDD60, and A79, which exhibits the most significant anti-tumor efficacy in prostate cancer cell models. CDRs refers to CDC20 (cell division cycle 20), DTL (denticleless E3 ubiquitin protein ligase), and RRM2 (ribonucleotide reductase M2).

Figure 7

Compounds targeting CDRs exhibited superior anti-tumor efficacy compared with AR antagonists. (A–E) The tumor cell growth inhibition effects of different dosages of Q199, XDD60, and A79, as well as enzalutamide, were determined with CCK-8 assays (A–D), and the IC₅₀ of each agent was calculated with three independent experiments (E). (F, G) The histograms show the relative cell viability after being treated with 5 M of Q199, XDD60, or A79 alone, or a combination. (H) The Venn diagram shows the overlap of Q199, XDD60, and A79 potential targets predicted with SwissTargetPrediction (http://swisstargetprediction.ch/). Molecular docking shows the binding of CDRs with Q199, XDD60, and A79. (I) The lowest binding (LB) affinity of CDRs with Q199, XDD60, and A79. ns, not significant. ∗∗p < 0.01. CDRs refers to CDC20 (cell division cycle 20), DTL (denticleless E3 ubiquitin protein ligase), and RRM2 (ribonucleotide reductase M2). AR, androgen receptor.