Network Pharmacology and Experimental Validation Identify Paeoniflorin as a Novel SRC-Targeted Therapy for Castration-Resistant Prostate Cancer

Abstract

Background: Despite advances in prostate cancer treatment, castration-resistant prostate cancer (CRPC) remains clinically challenging due to inherent therapy resistance and a lack of durable alternatives. Although traditional Chinese medicine offers untapped potential, the therapeutic role of paeoniflorin (Pae), a bioactive compound derived from Paeonia lactiflora, in prostate cancer has yet to be investigated. Methods: Using an integrative approach (network pharmacology, molecular docking, and experimental validation), we identified Pae key targets, constructed protein-protein interaction networks, and performed GO/KEGG pathway analyses. A Pae-target-based prognostic model was developed and validated. In vitro and in vivo assays assessed Pae effects on proliferation, migration, invasion, apoptosis, and tumor growth. Results: Pae exhibited potent anti-CRPC activity, inhibiting cell proliferation by 60% and impairing cell migration by 65% compared to controls. Mechanistically, Pae downregulated SRC proto-oncogene, non-receptor tyrosine kinase (SRC) mRNA expression by 68%. The Pae-target-based prognostic model stratified patients into high- and low-risk groups with distinct survival outcomes. Organoid and xenograft studies confirmed Pae-mediated tumor growth inhibition and SRC downregulation. Conclusions: Pae overcomes CRPC resistance by targeting SRC-mediated pathways, presenting a promising therapeutic strategy. Our findings underscore the utility of network pharmacology-guided drug discovery and advocate for further clinical exploration of Pae in precision oncology.

Keywords: SRC; network pharmacology; paeoniflorin; prostate cancer; resistance.

Figure 1

Identification of paeoniflorin potential targets in prostate cancer: (A) Pae’s chemical structure; (B) Venn diagram displaying the intersection targets of Pae and prostate cancer; (C) PPI network of Pae and prostate cancer intersection targets (nodes represent proteins, and edge represents protein–protein association); (D) GO enrichment analysis showing top BP, CC, and MF functional attributes of Pae’s targets against prostate cancer; (E) Sankey diagram for KEGG enrichment analysis of top 20 signaling pathways of Pae against prostate cancer.

Figure 2

Exploration of clinical characteristics of ten target genes in prostate cancer: (A) the expression levels of 10 target genes in prostate cancer tissues and adjacent normal tissues in the TCGA PRAD cohort; (B) the expression levels of 10 target genes in prostate cancer tissues of different T stages (T2, T3, and T4 stage) in the TCGA PRAD cohort; (C) the expression levels of 10 target genes in prostate cancer tissues of different Gleason scores (Gleason score ≤ 7 and Gleason scores > 7) in the TCGA PRAD cohort; (D) survival analysis predicting the relationship between gene expression patterns of 10 target genes and patient biochemical recurrence-free time in PRAD. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Figure 3

Establishment of a prognosis signature based on Paeoniflorin target genes: (A) receiver operating characteristic (ROC) curve constructed using the TCGA PCa cohort; (B) biochemical recurrence-free survival curves of patients with high- and low-risk in the TCGA PCa cohort; (C) the survival status distribution was analyzed based on the Pae-related signature; (D) receiver operating characteristic (ROC) curve constructed using the DKFZ prostate cancer cohort; (E) multivariate Cox regression models were used to analyze the associations between risk score, clinical factors, and PCa prognosis; (F) nomogram for patients predicting survival outcomes. T stage, Gleason score, and risk are marked as “points.” The total points by adding the three points can predict survival outcomes; (G) the CIBERSORT algorithm was utilized to evaluate and compare the infiltration of immune cells between the high- and low-risk score groups; (H) immune cells were categorized into four main types: dendritic cells, lymphocytes, macrophages, and mast cells. We then compared the differences in infiltration levels between the high- and low-risk score groups; (I) the differences in gene mutations between the high- and low-risk score groups were compared using waterfall plots. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Figure 4

Molecular docking of Paeoniflorin target genes: Three-dimensional and two-dimensional docking patterns and interactions of (A) paeoniflorin–TP53; (B) paeoniflorin–AR; (C) paeoniflorin–TGFBR2; (D) paeoniflorin–EGFR; (E) paeoniflorin–AKT1; (F) paeoniflorin–ESR1; (G) paeoniflorin–SRC; (H) paeoniflorin–NQO1. Purple: protein encoded by the target gene, Blue: paeoniflorin, Green: binding sites.

Figure 5

Paeoniflorin reduces SRC expression levels in PCa cell lines and PCa cohort tissues: (A,B) different concentrations of Pae were incubated with 22Rv1 and C4-2 cells for 48 h, and the effect of Pae on the cell viability was assessed by CCK-8 assay; (C,D) relative expression of mRNA levels of 10 target genes in 22Rv1 and C4-2 cells treated with Pae; (E) WB detection of relative expression of SRC protein levels in 22Rv1 and C4-2 cells treated with Pae; (F) quantification of luciferase activity in 22Rv1 and C4-2 cells with different concentration Pae treatment; (G) genome-wide loss-of-function screen in PC-3 identified essential genes including SRC and AR for cell survival. Lower scores indicate higher dependency on the gene for cell viability. SRC is highly expressed in primary prostate tumors; (H) relative expression of mRNA levels of SRC in tumor tissue and adjacent tissues of patients in Tongji PCa cohort; (I) relative expression of mRNA levels of SRC in PCa patients with different T stages in Tongji PCa cohort; (J) biochemical recurrence-free survival curves of patients with high and low SRC expression in Tongji PCa cohort. Statistical analysis: two-tailed Student’s t-tests were employed for statistical significance, with significance levels indicated as * p < 0.05, ** p < 0.01, *** p < 0.001, ns: not significan.

Figure 6

Paeoniflorin effectively inhibits PCa cell proliferation and metastasis and induces apoptosis: (A) 22Rv1 and C4-2 cells were treated with Pae for 1, 2 or 3 days, followed by CCK8 assay to evaluate cell viability; (B) 22Rv1 and C4-2 cells were treated with Pae for 2 days, followed by EdU experiments to evaluate cell proliferation; (C) the impact of Pae on the colony formation of 22Rv1 and C4-2 cells; (D) changes in migratory and invasive capabilities in 22Rv1 cells treated with Pae; (E,F) changes in wound healing ability of 22Rv1 and C4-2 cells treated with Pae on days 0–3; (G,H) the apoptosis rate of 22Rv1 and C4-2 cells after Pae intervention for 48 h were measured by flow cytometry. Statistical analysis: All data points were evaluated for statistical significance using two-tailed Student’s t-tests. Compared to DMSO (0.01%), * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 7

Paeoniflorin has therapeutic effects on patient-derived organoids and subcutaneous xenograft tumors in nude mice: (A) light microscopy images revealed morphological changes in the organoids; (B) changes in organoid size of Pae-treated group (n = 20); (C) changes in luminescence signals indicating metabolic activity and cell survival rate (n = 3); (D,F) top: images of excised 22Rv1 (D) and C4-2 (F) xenograft tumors in nude mice treated with DMSO or Pae, bottom: tumor volume growth curve of 22Rv1 (D) and C4-2 (F) xenograft tumor in nude mice treated with DMSO or Pae (n = 5); (E,G) quantification of tumor weight in excised 22Rv1 (E) and C4-2 (G) xenograft tumors (n = 5); (H,I) IHC staining for SRC (H) and Ki67 (I) in 22Rv1 and C4-2 xenograft tumors; (J) quantitative analysis of SRC and Ki67 expression; (K) H&E staining for evaluation of the safety of DMSO or Pae on the heart, liver, spleen, lung and kidney. Statistical analysis: All data points were evaluated for statistical significance using two-tailed Student’s t-tests. Compared to DMSO, *** p < 0.001.