MIBC (T24 and 5637) provided by Professor Shengtian Zhao (First Clinical Medical College, Qilu Hospital of Shandong University, Jinan, Shandong) were used in the present study. The T24 and 5637 cell lines are both well-established models of MIBC and share a common TP53 mutation background. Using two distinct cell lines helps demonstrate that the observed pro-ferroptotic effect of PPII is not an isolated phenomenon limited to a single genetic context but can be reproduced across different models of MIBC. The cells were cultured in RPMI-1640 medium (cat. no. CM10041, Macgene Technology Ltd.) supplemented with 10% fetal bovine serum (cat. no. 10270–106; Gibco; Thermo Fisher Scientific, Inc.). The two cell lines were maintained at 37°C in a humidified incubator with 5% CO₂.

Cell viability was assessed using a CCK-8 assay kit (E-CK-A362, Elabscience Bionovation Inc.). T24 (3×10³) and 5637 (6×10³) cells were seeded into 96-well plates and incubated at 37°C with 5% CO₂ overnight, after which the cells were treated with cisplatin or PPII. Following treatment, 10% CCK-8 solution was added to each well and gently mixed, and the plates were then incubated for an additional 2 h. The absorbance was measured at 450 nm using a spectrophotometer (Type 1510; Thermo Fisher Scientific, Inc.).

To assess the proliferative capacity of the cells, a colony formation assay was performed. Cells (800 cells/well) treated with gradient concentrations of PPII (0–0.8 µM) were seeded into six-well plates and cultured in medium supplemented with 10% fetal bovine serum, after which they were allowed to grow at 37°C with 5% CO₂ for 10 days until visible colonies formed. The cells were then gently washed with PBS and fixed with 4% paraformaldehyde for 15 min at room temperature (RT). After fixation, the colonies were stained with 0.1% crystal violet for 30 min at RT, followed by rinsing with water to remove excess dye. The number of stained colonies was counted to evaluate the clonogenic potential of the cells.

T24 cells were treated with gradient concentrations of PPII [0 (NC group), 0.2 (L group) and 0.4 (H group) µM] for 48 h and total RNA were extracted with a MJzol animal RNA Extraction Kit (cat. no. Majorbio) following by the manufacturer's protocol. RNA quality and integrity were analyzed using a NanoPhotometer spectrophotometer (Implen GmbH) and an Agilent 2100 Bioanalyzer (Agilent Technologies, Inc.). To construct RNA-seq libraries, ribosomal RNA (rRNA) was removed from total RNA, and the remaining mRNA was then randomly fragmented. The RNA-seq libraries were constructed using an Illumina Truseq RNA Sample Prep Kit (Illumina, Inc.) and sequenced on a NovaSeq 6000 system (Illumina, Inc.). Prior to bioinformatics analysis, FastQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/) was used to assess the quality of the raw data and the raw data were preprocessed to obtain high-quality clean read data. Cleaned reads were then mapped to the human reference genome GRCh38/hg38 using the spliced-read aligner HISAT2 (14) and StringTie (15) to obtain raw read counts and transcripts per million (TPM).

Raw read count data were used for gene expression analyses. Genes with low counts might represent a sequencing bias and contribute less to further analysis; thus, genes with zero expression values were excluded. After data filtering, differential expression analysis was performed with the Bioconductor package DESeq2 (16). Any gene with a P-value <0.05 and a fold change >1.25 was regarded as a significantly differentially expressed gene (DEG).

TPM of DEGs were used for c-means clustering with the R package Mfuzz to characterize dynamic changes in expression patterns (17). Fuzzy c-means (FCM) clustering is a soft clustering method performed with the Mfuzz algorithm with two key parameters (c=number of clusters and m=fuzzification parameter). The algorithm iteratively assigns the profile to the cluster with the shortest Euclidean distance while minimizing any objective function. In the present study, the data were clustered with the parameters c=10 and m=2.

GO functional enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed using the R package clusterProfiler (18). GO terms were divided into three separate subgroups: Molecular functions (MFs), cellular components (CCs) and biological processes (BPs). Enriched GO terms and KEGG pathways were identified according to the cutoff criterion of P-values <0.05.

Western blotting was carried out following described protocols (19). Briefly, total cellular proteins were extracted using RIPA lysis buffer (cat. no. P0013C; Beyotime Biotechnology), and their concentrations were quantified with a BCA assay kit (cat. no. P0010S; Beyotime Biotechnology). An equal amount of protein (20 µg per lane) was separated by 10% SDS-PAGE gel and subsequently electrophoretically transferred onto a PVDF membrane (cat. no. IPVH00010; MilliporeSigma). The membrane was blocked with 5% skimmed milk powder in TBST buffer (0.1% Tween-20) at RT for 1 h and then probed with the primary antibody at 4°C overnight. After washing, it was incubated with an HRP-conjugated goat anti-mouse secondary antibody (1:3,000 for GPX4 and 1:10,000 for ACTB; cat. no. A0216; Beyotime Biotechnology) at RT for 1 h. Finally, the target protein bands were visualized using an enhanced chemiluminescence (ECL) substrate. The primary antibodies used in the present study included GPX4 (mouse-sourced antibody, 1:1,000; cat. no. 67763-1-Ig) (20) purchased from Proteintech Group, Inc. and ACTB (mouse-sourced antibody, 1:10,000; cat. no. A5441) (21), which was obtained from Sigma-Aldrich (Merck KGaA). The intensity of the protein bands was quantified using ImageJ (version 1.53q; National Institutes of Health).

Intracellular ROS levels were determined using a Reactive Oxygen Species Assay Kit (cat. no. S0033S; Beyotime Biotechnology), with DCFH-DA as the primary reagent. In brief, T24 and 5637 cells were plated in 6-well plates and treated with gradient concentrations of PPII (0, 0.4 and 0.8 µM for T24 cells and 0, 0.85 and 1.70 µM for 5637 cells) for 48 h. After treatment, the cells were incubated with DCFH-DA for 30 min at 37°C and the nuclei were stained with Hoechst 33342 (cat. no. C1029; Beyotime Biotechnology) for 10 min. The fluorescence was then visualized using a fluorescence microscope (Vert.A1; Zeiss GmbH) and the mean fluorescence intensity was quantified using ImageJ (version 1.53q; National Institutes of Health).

Fe²+ detection was performed according to the FerroOrange manufacturer's instructions (cat. no. F374; Dojindo Laboratories, Inc.). After 48 h of PPII treatment with various concentrations of PPII (as described in the ROS detection analysis section), the culture medium was removed and the BC cells were washed three times with HBSS. The cells were then incubated with FerroOrange working solution (1 µM) at 37°C in a 5% CO2 atmosphere for 30 min. Changes in fluorescence intensity were observed using a fluorescence microscope (Vert.A1; Zeiss GmbH).

T24 and 5637 cells were treated as aforementioned, followed by washing with PBS and lysis with lysis buffer. After centrifugation (19,480 × g), the supernatant was collected for the detection of malondialdehyde (MDA) levels using a MDA detection kit (cat. no. S0131S; Beyotime Biotechnology). The absorbance at 532 nm was measured with a spectrophotometer (Type 1510; Thermo Fisher Scientific, Inc.). Similarly, superoxide dismutase (SOD) levels were measured following the instructions of the corresponding reagent kit (cat. no. S0101M; Beyotime Biotechnology), with the absorbance recorded at 450 nm.

T24 and 5637 cells were treated with gradient concentrations of PPII for 48 h as aforementioned. The medium was then replaced with complete medium containing BODIPY 581/591 C11 (10 µM; cat. no. D3861; Invitrogen; Thermo Fisher Scientific, Inc.) and the cells incubated at 37°C with 5% CO2 for an additional 30 min. The cells were then washed three times with PBS (5 min per wash), followed by staining with Hoechst 33342 for 10 min at 37°C. After the samples were washed, the fluorescence intensities of oxidized and non-oxidized lipids were observed under a fluorescence microscope (Vert.A1; Zeiss GmbH), and the mean fluorescence intensity was quantified using ImageJ (version 1.53q; National Institutes of Health).

All the statistical analyses were performed using GraphPad Prism (version 5.0; Dotmatics). Data are presented as the mean ± standard deviation (SD) from at least three independent experiments. For comparisons between two groups, an unpaired two-tailed Student's t-test was used. For comparisons among more than two groups, one-way analysis of variance (ANOVA) was performed, followed by Dunnett's test for multiple comparisons when all groups were compared against a single control group. DEGs were identified using an adjusted P-value (False Discovery Rate; FDR) threshold. Furthermore, Pearson correlation analysis was conducted using R software (version 4.3.0; http://cran.r-project.org/bin/windows/base/old/4.3.0/) to assess the relationships between oxidative stress indicators. P<0.05 was considered to indicate a statistically significant difference.

The effect of PPII on bladder cancer cell viability was assessed using a CCK-8 assay. T24 and 5637 cells were treated with increasing concentrations of cisplatin (used as a positive control, Fig. 1A and C) or PPII (Fig. 1B and D) for 48 h. The CCK-8 results revealed that PPII could inhibited the viability of bladder cancer cells and the calculated IC₅₀ values of cisplatin were 11.70±1.31 µM for T24 cells and 20.17±0.82 µM for 5637 cells, whereas the IC₅₀ values of PPII were 0.86±0.09 µM for T24 cells and 1.72±0.21 µM for 5637 cells.

The effect of the PPII on the proliferative capacity of bladder cancer cells was evaluated through colony formation assays. The results suggested that PPII effectively reduced the number of colonies formed by T24 and 5637 cells in a dose-dependent manner (Fig. 1E). Statistical analysis revealed that 0.2, 0.4, and 0.8 µM PPII significantly decreased the number of colonies formed by T24 cells (P<0.01, Fig. 1F). Similarly, PPII at concentrations of 0.85 and 1.70 µM PPII markedly inhibited colony formation in 5637 cells (P<0.001, Fig. 1G).

Previous experiments had indicated that PPII effectively inhibits the proliferation of BC cells. To minimize the confusing effects of cytotoxicity from high concentrations of drugs, the present study chose PPII drug concentrations (0, 0.2, and 0.4 µM) that did not affect T24 cell viability for RNA transcriptome sequencing (Fig. S1A), aiming to elucidate the potential mechanisms involved in this process.

T24 cells were treated with high-concentration and low-concentration PPII for 48 h, after which the RNA was harvested in quadruplicate. A total of 12 samples were divided into three groups, namely, the NC, L and H groups, according to the concentrations of PPII (Fig. 2A). DEGs between each group were identified using a P-value <0.05 and a fold change greater than the cutoff. A total of 834 DEGs were identified between the NC, L and H groups. Principal components analysis revealed that the between-group samples were clearly separated from each other and revealed good clustering of samples within the same group, indicating distinct gene expression profiles upon PPII treatment (Fig. 2B). Volcano plots also revealed the DEGs that differed from each other in the groups (Fig. 2C and D). All the DEGs were included in the subsequent analyses.

To understand the dynamic alteration expression profiles in T24 cells during the progression of PPII treatment, 834 DEGs were subjected to soft clustering to determine their expression trends in the three series groups. The DEGs were divided into 10 clusters according to their trend similarity over time using a soft Mfuzz clustering algorithm and not all the clusters exhibited consistent expression trends with a distinct peak. A heatmap revealed that both Cluster 6 and Cluster 8 showed transcriptional upregulation response to PPII treatment, Cluster 6 demonstrated a more consistent concentration-dependent pattern, showing gradual upregulation with increasing PPII concentrations. By contrast, Cluster 8 exhibited a biphasic response with initial downregulation at lower concentrations followed by upregulation at higher doses (Fig. 3). On the basis of this clear concentration-dependent response, Cluster 6 was selected for further investigation.

GO and KEGG pathway analyses of the DEGs in Cluster 6 revealed enrichment of ribosome-related and RNA processing-related categories in the GO analysis (Fig. 4A). KEGG analysis revealed pathways related to ferroptosis and ROS (Fig. 4B), and ferroptosis-related factors, including PTGS2 and ACSL4, showed varying degrees of changes with dose-dependent manner in volcano maps (Fig. 2C and D). Given that ferroptosis is closely associated with the regulation of cell proliferation and that ROS play a crucial role in the ferroptosis process, subsequent experiments focused on validating the regulatory effects of PPII on ROS and ferroptosis.

The effect of PPII on ROS levels in bladder cancer cells was assessed using a ROS detection kit, as detailed in the methods section. The fluorescence intensity was used to quantify the ROS levels, revealing that T24 cells exhibited a significant increase in fluorescence intensity following PPII treatment. Notably, the fluorescence intensity in the 0.8 µM PPII treatment group was comparable to that in the positive control group (CRTL + Rosup; Fig. 5A). Statistical analysis demonstrated that as the concentration of PPII increased, the average fluorescence intensity significantly increased (Fig. 5C; P<0.001) in a dose-dependent manner. Similar results were observed in 5637 cells (Fig. 5B and C). Since MDA and SOD are critical regulators of ROS levels, their concentrations were measured. High concentrations of PPII (0.8 µM for T24 cells and 1.70 µM for 5637 cells) significantly increased MDA levels (Fig. 5D, P<0.05) but inhibited SOD activity (Fig. 5E; P<0.05). Furthermore, Pearson analysis revealed a negative correlation between ROS and SOD, whereas ROS and MDA showed a positive correlation (Fig. S1B). These findings suggested that PPII effectively increases ROS levels in bladder cancer cells.

Fe²+ accumulation is a key indicator of ferroptosis. To assess changes in Fe²+ levels after PPII treatment in BC cells, the FerroOrange kit was used. The results revealed that the Fe²+ staining intensity increased with increasing PPII concentration in T24 cells (Fig. 6A). Statistical analysis revealed that the average fluorescence intensity significantly increased in a dose-dependent manner (Fig. 6B; P<0.001). Consistent findings were also observed in 5637 cells (Fig. 6C and D).

The accumulation of oxidized lipids serves as a key marker in the process of ferroptosis. Using specific probes, the levels of both oxidized and non-oxidized lipids was assessed. Treatment with 0.8 µM PPII significantly elevated the levels of oxidized lipids in T24 cells (P<0.05), but did not markedly affect the levels of non-oxidized lipids (Fig. 6E and F). Similarly, in 5637 cells, treatment with 0.85 µM and 1.70 µM PPII markedly increased oxidized lipid levels (Fig. 6G and H).

Molecular docking suggests that PPII and GPX4, a key regulatory factor of Ferroptosis, have potential good binding ability (Fig. S1C and D; binding energy, −8.7 kcal/mol). Further analysis of the expression of GPX4 was conducted using western blotting. The results indicated that treatment with high concentrations of PPII markedly decreased GPX4 levels in both T24 and 5637 cells (Fig. 6I and J). Furthermore, PPII reduces the increase in GPX4 levels caused by the ferroptosis inhibitor Fer-1 (Fig. S1E). These findings suggested that the regulatory effect of PPII on ferroptosis is mediated by GPX4.