A cyclin D1–CDK4/6–E2F dependency represents an acquired vulnerability in cisplatin-resistant bladder cancer

Abstract

Background

Cisplatin resistance remains a major obstacle in the treatment of bladder cancer (BC). This study aimed to define molecular drivers of cisplatin resistance and to assess potential therapeutic targets that may help restore treatment responsiveness.

Methods

Integrative transcriptomic analyses were performed using The Cancer Genome Atlas (TCGA) cohort to compare cisplatin non-responders with responders, alongside a cisplatin-resistant BC cell model to identify dysregulated pathways. Functional studies, including siRNA-mediated knockdown, SRB assay, flow cytometry, western blotting, and microarray-based transcriptomics, sphere formation assay, and clonogenic assay, were used to characterize the dysregulated pathways.

Results

Both cisplatin non-responders in TCGA dataset and resistant BC cells exhibited upregulation of CCND1 and enrichment of E2F target genes. Silencing Cyclin D1 restored cisplatin sensitivity in resistant BC cells. Abemaciclib, a CDK4/6 inhibitor, selectively inhibited proliferation of cisplatin-resistant BC cells, reduced RB phosphorylation, induced sub-G1 accumulation, and suppressed expression of key regulators of cell-cycle progression and homologous recombination repair. Combined treatment with abemaciclib and cisplatin synergistically suppressed the proliferation of cisplatin-resistant BC cells in vitro and produced significantly greater tumor growth inhibition in an RT112 xenograft model in vivo. Furthermore, Abemaciclib reduced sphere-forming capability and enhanced the anti-clonogenic effect of cisplatin in cisplatin-resistant BC cells.

Conclusion

These findings identify a Cyclin D1/CDK4/6–E2F signaling dependency as a characteristic feature of the cisplatin-resistant state in BC. Targeting this acquired vulnerability provides a mechanistic rationale for combination strategies to enhance cisplatin responsiveness for refractory BC.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13062-026-00812-5.

Introduction

Bladder cancer (BC) is the second most common genitourinary malignancy worldwide, with an estimated 613,791 new cases and 220,349 deaths occurred according to the latest global cancer statistics in 2022 [1]. In the United States, approximately 83,190 people will be diagnosed with this disease, and 16,840 people are expected to die in 2024 [2]. About 70% of newly diagnosed patients present with non-muscle-invasive bladder cancers (NMIBC), which has a five-year survival rate of nearly 90%. However, up to 70% of NMIBC cases recur, with a higher chance of progression toward muscle invasive bladder cancers (MIBC) [3]. The remaining 30% of patients belong to MIBC with a five-year survival rate of about 50%, which decreases to 6% once the disease progress to the metastatic stage [4]. Despite decades of clinical efforts, outcomes for advanced or metastatic BC remain largely unchanged, reflecting an urgent need for improved therapeutic strategies [3]. Currently, the standard treatment for MIBC includes radical cystectomy combined with platinum-based chemotherapy, such as MVAC (methotrexate, vinblastine, doxorubicin, and cisplatin), dose-dense MVAC, and gemcitabine plus cisplatin [5, 6]. These regimens yield an overall survival rate of 9–15 months, yet the median survival rate is reduced to 5 to 7 months in patients encounter relapses after platinum-based chemotherapy [7]. For patients who fail or cannot tolerate cisplatin-based regimens, treatment options are limited. Immune checkpoint inhibitors (ICIs) have shown durable responses in a subset of patients [8]. Erdafitinib, the first targeted therapy, was approved for patients with advanced BC harboring alterations of FGFR2 or FGFR3 after progression on platinum-based chemotherapy [9]. Nonetheless, only a minority of patients benefit from ICIs or FGFR-targeted therapy, and the majority of advanced BCs remain unresponsive. Because of the lack of effective treatments for advanced or metastatic diseases, the high economic burden for caring bladder cancer remains a social problem [10]. Hence, the development of novel therapeutic agents and methods for early detection of resistance remain unmet medical needs in advanced BC patients.

Cisplatin-based chemotherapy has been a cornerstone of MIBC chemotherapy for decades [11]. Upon cellular entry, cisplatin needs to be activated by the replacement of its two chloride ligands with water molecules, which can attack the N7 atoms of purine bases to form a variety of adducts. The generation of DNA lesions subsequently triggers DNA damage response and the induction of mitochondrial apoptosis [12]. However, intrinsic and acquired resistance to cisplatin frequently emerge during treatment, contributing to therapeutic failure and disease progression. Multiple resistance mechanisms have been described, including reduced intracellular accumulation of cisplatin, enhanced DNA repair capacity, and impaired apoptotic signaling [13]. Despite these insights, no clinically approved drugs directly target these resistance mechanisms, underscoring the need to identify new molecular drivers of cisplatin resistance and pharmacological vulnerabilities that can be therapeutically exploited.

One of the hallmarks of cancer is the loss of growth control through dysregulation of the cell-cycle machinery [14]. Cyclin D associates with cyclin-dependent kinases 4 and 6 (CDK4/6) to phosphorylate the retinoblastoma (RB) protein, releasing E2F transcription factors that drive the expression of genes required for DNA replication and G1-to-S phase transition [15]. Under normal conditions, this process is restrained by endogenous CDK inhibitors (p16^INK4A, p15^INK4B, p18^INK4C, and p19^INK4D). In BC, CDKN2A, which encodes p16^INK4A, is frequently deleted on chromosome 9p21.3, leading to constitutive activation of the Cyclin D/CDK4/6–E2F axis, and thereby promoting uncontrolled proliferation and tumor progression [15–17]. Amplification or overexpression of CCND1 (Cyclin D1), CDK4, or CDK6 has been reported in multiple tumor types, including head and neck, breast, non-small-cell lung, esophageal cancers, melanoma, and glioblastoma [18]. The therapeutic relevance of this pathway has been supported by the development of CDK4/6 inhibitors, which have demonstrated significant clinical benefits and received FDA approval for the treatment of ER⁺/HER2^– advanced breast cancer patients. Palbociclib was the first CDK4/6 inhibitor approved by the U.S. FDA, either in combination with the aromatase inhibitor letrozole as first-line endocrine therapy or with fulvestrant after disease progression on prior endocrine treatment. Subsequently, Ribociclib and Abemaciclib were also approved for the management of advanced breast cancer. Among these, only Abemaciclib was approved as a monotherapy for patients with HR⁺/HER2^– advanced or metastatic breast cancer with progressive disease [19].

Recent integrative multi-omics approaches have begun to uncover the molecular complexity of cisplatin resistance. DNA methylome profiling identified hypermethylation and silencing of ASS1 and SAT1, leading to altered amino acid metabolism and chemoresistance, which could be reversed by gene re-expression [20]. Proteomic and phosphoproteomic analyses revealed CDK2, CHEK1, and ERBB2 as key regulators of cisplatin resistance signaling [21]. Furthermore, proteomic and miRNA profiling of extracellular vesicles from blood or urine have provided potential biomarkers for therapy response and disease monitoring [22, 23]. Despite these advances, the upstream transcriptional networks and cell-cycle–related mechanisms driving cisplatin resistance remain incompletely understood. Given the central role of Cyclin D/CDK4/6–E2F signaling in cell-cycle control and its potential intersection with DNA repair and chemoresistance pathways, we hypothesized that dysregulation of this axis contributes to cisplatin resistance in BC. Through integrative transcriptomic analyses of MIBC patient datasets and cisplatin-resistant cell models, we identified enrichment of E2F target genes and overexpression of CCND1 (Cyclin D1) in cisplatin non-responders. Functional studies further demonstrated that inhibition of Cyclin D1/CDK4/6 signaling by genetic knockdown or pharmacological blockade sensitizes BC cells to cisplatin. Collectively, our findings suggest that targeting the Cyclin D1/CDK4/6–E2F pathway represents a promising therapeutic strategy to overcome cisplatin resistance in BC.

Materials and methods

Cell culture and reagents

N/P(14) cells was a kind gift from Dr. Tzyh-Chyuan Hour (Kaohsiung Medical University) and cultured as previously described [24–26]. J82, RT112 and SV-HUC-1 cells were purchased from American Type Culture Collection (Manassas, VA, USA). The cisplatin-resistant subline of J82 (J82-Cis) cell was generated in our previous study [26]. The cells were cultured in RPMI 1640 (NTUB1, N/P(14)), EMEM (J82), or F12K (SV-HUC-1) growth median at 37 °C with 5% CO₂, supplemented with 10% FBS and 1X antibiotic-antimycotic (Thermo Fisher Scientific; Waltham, MA, USA). Abemaciclib (HY-16297 A), and Palbociclib (HY-50767) were purchased from MedChem Express (Monmouth Junction, NJ, USA). Cisplatin (13119) was purchased from Cayman Chemicals (Ann Arbor, MI, USA). Other chemicals used in this study were purchased from Sigma-Aldrich (St. Louis, MO, USA).

Sulforhodamine B (SRB) assay

Cells were seeded into 96-well plates and treated with the indicated drugs for 48–72 h. Following treatments, the cells were then fixed with 10% trichloroethanoic acid and stained with 0.4% SRB solution. Excess dye was removed by washing with 1% acetic acid, and the bound dye was subsequently solubilized in 10 mM Tris base. Absorbance at 515 nm was measured using a spectrophotometer. Growth inhibition 50% (GI₅₀) values, defined as the drug concentration that causes a 50% reduction in cell growth, were calculated as previously described [27].

Flow cytometry analysis

Cells were seeded in 6-well plates and treated with Abemaciclib for 24 h, 48–72 h. After treatment, the cells were trypsinized, washed with PBS, and fixed in 75% ethanol at -20 °C, and stained with propidium iodide (80 µg/mL) containing 0.1% Triton-X 100 and RNaseA (100 µg/mL) to analyze cell cycle. The samples were subsequently analyzed using a Guava Muse Cell Analyzer (Luminex; Austin, TX, USA) as previously described [28].

Western blot

Cells were treated with the indicated compounds, and proteins were extracted by using a SDS lysis buffer, separated through SDS-PAGE gels, and transferred onto a polyvinylidene difluoride (PVDF) membrane for subsequent analysis as previously described [28]. The antibodies used in this study were obtained from the following sources: Phospho-Rb (Ser807/811) (CS8516), Phospho-Histone H2A.X (Ser139) (γH2AX, CS9718), CDK2 (CS2546), Aurora A (CS4718), Aurora B (CS3094), FOXM1 (CS5436), and PARP (CS9542) were obtained from Cell Signaling Technology (Danvers, MA, USA); Rad51 (NB100-148) and caspase-3 (NB100-56708) were purchased from Novus (Littleton, CO, USA); Cyclin D1 (GTX108624) and GAPDH (GTX100118) were obtained from GeneTex (Irvine, CA, USA); Cyclin A (sc-596), Cyclin B1 (sc-594), Cyclin E (sc-247), CDK1 (sc-54), CDK4 (sc-601), PLK1 (sc-55504) and BRCA1 (sc-6954) were purchased from Santa Cruz Biotechnology (Dallas, TX, USA). The band intensities of each protein were determined by using the Image J software (Version 1.51, National Institutes of Health; Bethesda, MD, USA).

siRNA-mediated gene silencing

For siRNA-mediated knockdown experiments, cells were transfected with scrambled control siRNA (Thermo Fisher Scientific; catalog number: 4390843) or CCND1-targeting siRNA (Thermo Fisher Scientific; catalog no. 4427038; ID: s229) by using Lipofectamine RNAiMAX (Thermo Fisher Scientific), and incubated for 48 h. Knockdown efficiency was verified by western blot analysis.

Reverse transcription quantitative PCR

Toral RNA from treated cells was extracted using TRIzol reagent (Thermo Fisher Scientific). RNA was then reverse transcribed into complementary DNA using the HiScript I ^TM First Strand cDNA Synthesis Kit from Bionovas (Toronto, ON, Canada). Quantitative PCR was conducted using CCND1-specific primers (F: 5′-TCCTCTCCAAAATGCCAGAG-3′; R: 5′-GAGGGCGGATTGGAAATGAA-3′) and the RealQ Plus 2x Master Mix Green Kit (Ampliqon, Odense, Denmark).

RNA isolation and microarray analysis

N/P(14) cells were treated with Abemaciclib (0.625 µM) for 24 h, collected by scraping, washed with PBS, and subjected to total RNA extraction using TRIzol reagent (Thermo Fisher Scientific). RNA quality was assessed by the 2100 Bioanalyzer (Agilent Technologies; Santa Clara, CA, USA). DNA-free total RNA was hybridized to the human OneArray (Phalanx Biotech Group, Hsinchu, Taiwan). The resulting GPR files were loaded into Rosetta Resolver^® System (Rosetta Biosoftware) for data processing and analysis. The data generated in this study are publicly available in Gene Expression Omnibus (GEO) at GSE313088. Differentially expressed genes were identified using standard selection criteria of log2 |Fold change| ≥ 1 and P values < 0.05, as previously described [29]. Gene ontology analysis, Ingenuity Pathway Analysis (IPA), and Gene Set Enrichment Analysis (GSEA) were employed to interpret gene functions and networks [27].

Analysis of TCGA database and mRNA sequencing data

The TCGA bladder carcinoma cohort (TCGA-BLCA) transcriptomic dataset was obtained from the R package, TCGA biolinks, with 59 primary tissues included in the study. To identify genes with statistically significant differential expression between cisplatin responders and non-responders, all raw counts were normalized, and differential expression analysis was performed using the edgeR package. Genes with a false discovery rate (FDR) < 0.05 were considered significantly differentially expressed. For mRNA sequencing, total RNA was extracted from NTUB1 and N/P(14) cells using TRIzol reagent (Thermo Fisher Scientific), with quality and quantity assessed via bioanalyzer and Qubit. Poly-A–enriched mRNA was used for library construction and indexing with Illumina^® TruSeq™ stranded mRNA mRNA Library Prep Kit (San Diego, CA, USA). The data generated in this study are publicly available in Gene Expression Omnibus (GEO) at GSE313089. Raw fastq files were evaluated with FastQC, and reads were aligned to the human reference genome (hg19) using STAR. Gene expression quantification was performed with RSEM, and differentially expressed genes were identified using edgeR. Gene Ontology analysis, Ingenuity Pathway Analysis (IPA), and Gene Set Enrichment Analysis (GSEA) were performed to interpret gene functions and associated biological networks, as previously described [27].

Sphere formation assay

The cells were trypsinized and dissociated into single-cell suspensions, and 1.5 × 10⁴ cells were seeded into ultra-low attachment 6-well plates (Corning, Glendale, Arizona, USA). Cells were cultured in serum-free DMEM/F12 medium consisting 20 ng/mL EGF, 20 ng/mL bFGF, and 2% B27 supplement (Thermo Fisher Scientific). After 10 days, tumor spheres were counted using an inverted microscope, and mean diameter of the sphere was measure by ImageJ software (Version 1.53, National Institutes of Health; Bethesda, MD, USA).

Clonogenic assay

Cells were trypsinized and plated at 1500 cells/well in 6 well plates for 24 h, and treated with various concentrations of compounds or DMSO vehicle for 7 days. The colonies were fixed and stained with crystal violet (0.5% in 70% EtOH). The percentage of colony area was analyzed by Image J (Version 1.51, National Institutes of Health, Bethesda, MD, USA).

Homologous recombination detection

Homologous recombination (HR) was assessed using the Homologous Recombination Assay Kit (Norgen, Thorold, ON, Canada) as previously described [30]. Briefly, dl-1 and dl-2 plasmids were co-transfected into cells by Lipofectamine 3000 Transfection Reagent (Thermo Fisher Scientific). The cells perform HR will produce a plasmid recombination amplicon of 420 base pairs as a result of HR activity. The qPCR fluorescence intensity of amplicon can be directly correlated to the efficiency of HR from the template DNA. The assay primer mixtures are designed exclusively for the HR products, and the universal primer mixtures were used to detect the backbone plasmids (dl-1 or dl-2) as an internal control. To detect the homologous recombination product, thermal cycler was programmed into 3 min at 95 °C as initial denaturation, followed by 40 cycles of 15 s at 95 °C for denaturation, 30 s at 61 °C as annealing, 60 s at 72 °C for extension. Data analysis was based on the 2^−ΔΔCt method: [HR product/backbone plasmids]^N/P(14) versus [HR product/backbone plasmids]^NTUB1.

In vivo xenograft model

Six-week-old BALB/c nude mice were subcutaneously injected with 1 × 10⁶ RT112 cells. All animal procedures were conducted in accordance with institutional ethical guidelines and were reviewed and approved by the Animal Care and Use Committee of Taipei Medical University (IACUC approval no: LAC-2018-0474). Tumor size and mouse body weight were measured every 2 days. When tumors reached approximately 100 mm³, mice were randomly assigned to four treatment groups: vehicle, Abemaciclib (75 mg/kg, p.o., qd), cisplatin 5 mg/kg/i.p./qwk, and the combination treatment. Tumor volume (mm³) was calculated using the formula (W² × L) / 2, where W and L represent the tumor width and length, respectively.

Statistical analysis

Data analysis was performed using GraphPad Prism software (10.4.1). Statistical comparisons between two groups were conducted using an unpaired one-tailed or two-tailed Student’s t-test. One-way or two-way analysis of variance (ANOVA) with post-hoc analysis was used for comparisons among multiple groups. The data were considered significant when the p-value is lower than 0.05.

Results

E2F target genes are enriched in cisplatin resistance in bladder cancer (BC)

To elucidate the transcriptional alternation that might cause cisplatin resistance in MIBC patients, we firstly analyzed differential expressed genes (DSGs) between complete response (CR) patients and non-responders (partial response, stable disease, and progressive disease) in the RNA-sequencing (RNA-seq) data from TCGA Bladder cancer (BLCA) database, 2017 [31]. The results showed that 353 genes were upregulated and 450 genes were downregulated in non-responders. The gene list was further analyzed by Gene Set Enrichment Analysis (GSEA) and the pathways enriched in non-responders are shown in Table 1. Additionally, NTUB1 (a high-grade urothelial carcinoma cell line) and its corresponding cisplatin-resistant subline N/P(14) were used as representative models for cisplatin-sensitive and resistant BC cell lines, respectively [25]. Cisplatin resistance was confirmed by differential anti-proliferative activity by SRB assay as describe previously [26]. To analyze the possible mechanism underlying cisplatin-resistance in BC cells, DSGs between N/P(14) and NTUB1 cells were analyzed by RNA-seq and bioinformatics analysis. The GSEA results suggest that several pathways were enriched in N/P(14) cells (Table 2). Together, the data suggest that Wnt/β-catenin, cyclin and cell cycle regulation and E2F targets are commonly enriched pathways in cisplatin-resistant groups.

Table 1.

Pathways enriched in cisplatin non-responders of TCGA dataset

NAME	ES	NES	NOM p-val	FDR q-val	FWER p-val
HALLMARK_INTERFERON_GAMMA_ RESPONSE	0.796	2.398	0.000	0.000	0.000
HALLMARK_ALLOGRAFT_ REJECTION	0.796	2.376	0.000	0.000	0.000
HALLMARK_INTERFERON_ALPHA_ RESPONSE	0.822	2.240	0.000	0.000	0.000
HALLMARK_INFLAMMATORY_ RESPONSE	0.727	2.178	0.000	0.000	0.000
HALLMARK_MYC_TARGETS_V1	0.714	2.139	0.000	0.000	0.000
HALLMARK_EPITHELIAL_ MESENCHYMAL_TRANSITION	0.691	2.070	0.000	0.000	0.000
HALLMARK_COMPLEMENT	0.682	2.049	0.000	0.000	0.000
HALLMARK_E2F_TARGETS	0.666	1.964	0.000	0.000	0.000
HALLMARK_TNFA_SIGNALING_ VIA_NFKB	0.641	1.915	0.000	0.000	0.000
HALLMARK_IL6_JAK_STAT3_ SIGNALING	0.662	1.780	0.000	0.000	0.000
HALLMARK_MTORC1_SIGNALING	0.590	1.762	0.000	0.000	0.001
HALLMARK_COAGULATION	0.608	1.745	0.000	0.000	0.001
HALLMARK_APOPTOSIS	0.572	1.691	0.000	0.000	0.002
HALLMARK_KRAS_SIGNALING_UP	0.561	1.689	0.000	0.000	0.002
HALLMARK_G2M_CHECKPOINT	0.558	1.670	0.000	0.001	0.003
HALLMARK_IL2_STAT5_SIGNALING	0.555	1.670	0.000	0.001	0.003
HALLMARK_REACTIVE_OXIGEN_ SPECIES_PATHWAY	0.679	1.656	0.000	0.001	0.004
HALLMARK_HYPOXIA	0.518	1.555	0.000	0.003	0.020
HALLMARK_APICAL_JUNCTION	0.484	1.471	0.000	0.011	0.065
HALLMARK_GLYCOLYSIS	0.480	1.441	0.004	0.015	0.099
HALLMARK_MYC_TARGETS_V2	0.569	1.434	0.025	0.015	0.105
HALLMARK_OXIDATIVE_ PHOSPHORYLATION	0.476	1.430	0.000	0.015	0.110
HALLMARK_ANGIOGENESIS	0.597	1.398	0.029	0.021	0.154
HALLMARK_P53_PATHWAY	0.424	1.293	0.005	0.056	0.369
HALLMARK_APICAL_SURFACE	0.520	1.267	0.089	0.068	0.439
HALLMARK_UV_RESPONSE_UP	0.428	1.252	0.022	0.077	0.487
HALLMARK_KRAS_SIGNALING_DN	0.402	1.201	0.032	0.120	0.650

Table 2.

Pathways enriched in cisplatin-resistant N/P(14) cells of RNA-seq dataset

NAME	ES	NES	NOM p-val	FDR q-val	FWER p-val
HALLMARK_G2M_ CHECKPOINT	0.576	1.543	0.000	0.036	0.026
HALLMARK_E2F_TARGETS	0.564	1.517	0.000	0.021	0.030
HALLMARK_MYC_TARGETS_V2	0.623	1.440	0.000	0.034	0.067
HALLMARK_MITOTIC_SPINDLE	0.493	1.338	0.014	0.065	0.160
HALLMARK_MYC_TARGETS_V1	0.489	1.322	0.000	0.060	0.186
HALLMARK_UNFOLDED_PROTEIN_RESPONSE	0.485	1.242	0.026	0.095	0.322
HALLMARK_TGF_BETA_ SIGNALING	0.461	1.032	0.387	0.413	0.852
HALLMARK_WNT_BETA_CATENIN_SIGNALING	0.439	0.958	0.476	0.617	0.960
HALLMARK_NOTCH_SIGNALING	0.463	0.941	0.522	0.618	0.973

Cyclin D1/CDK4/CDK6 complex is a potential therapeutic target to overcome cisplatin resistance

To identify clinically relevant therapeutic targets, we further look for DEGs overlapping in the NGS-datasets of TCGA and N/P(14)/NTUB1. The results showed that 19 genes were upregulated and 32 genes were downregulated in cisplatin non-responders and N/P(14) cells (Fig. 1A and B). Uncontrolled cellular proliferation is achieved by the dysregulation of the cell-cycle regulators as well as cyclin-dependent kinases (CDKs) [14]. Cyclin D/CDK4/6 complex phosphorylates retinoblastoma (RB) protein and inactivates its function, resulting in the activation of the E2F transcription factor. This allows the transcription of E2F target genes that enable cell cycle progression from the G1 to the S phase. The transcriptomic analysis revealed that the gene expression of CCND1 (Cyclin D) is overexpressed in MIBC patients who did not respond to platinum-based therapy (TCGA dataset), as well as in cisplatin-resistant N/P(14) cells (Fig. 1A). The activity of Cyclin D/CDK4/6 is inhibited by CDK inhibitors, such as NK4 family proteins (p16^INK4A, p15^INK4B, p18^INK4C, and p19^INK4D). In BC, p16^INK4A is frequently underexpressed due to deletion on chromosome 9p21.3 [15–17]. We therefore focused on the enriched networks by interpretive phenomenological analysis (IPA), and the results demonstrated that Cyclin D1 is an important upstream regulator in N/P(14) cells as well as in cisplatin non-responders in TCGA dataset (Fig. 1C and D). Meanwhile, GSEA further revealed that E2F targets, the main downstream effector of cyclin D1, were upregulated in N/P(14) cells and in TCGA non-responders (Fig. 1E and F).The increased transcription of CCND1 was confirmed in N/P(14) cells (Fig. 1G). Knocking down of Cyclin D1 sensitized RT112 cells, which are reported to have low intrinsic sensitivity to cisplatin [32], as well as N/P(14) cells to cisplatin treatment (Fig. 1H and J). These data support our findings that Cyclin D1 and its working complex CDK4/6 are potential therapeutic targets for cisplatin-resistant BC. Therefore, we aim to determine the mechanisms of Cyclin D1/CDK4/6-mediated cisplatin resistance and validate the correlation between Cyclin D1 expression and cisplatin resistance in MIBC patient samples.

Fig. 1.

Bioinformatics analysis of TCGA and N/P(14)/NTUB1 datasets. Differentially expressed genes (DEGs) that were up-regulated (A) or down-regulated (B) in both datasets are shown. Cyclin D1, encoded by CCND1, was enriched in IPA network analysis in N/P(14) cells (C) and TCGA non-responders (D). Gene set enrichment analysis (GSEA) revealed significant enrichment of E2F target gene signatures in N/P(14) cells (E) and TCGA non-responders (F). (G) mRNA expression of CCND1 was increased in N/P(14) cells. Data are expressed as mean ± S.D. (n = 2). Statistical analyses were performed by unpaired two-tailed Student’s t-test. (H–J) Knockdown of CCND1 enhanced cisplatin sensitivity in RT112 and N/P(14) cells. Cells were transfected with siRNA and exposed to cisplatin for 72 h. Knockdown efficiency was examined by western blotting (H), and the growth of RT112 (I) and N/P(14) (J) cells was measured by SRB assay. Data are expressed as mean ± S.D. (n = 3), and the statistical analyses were performed using two-way ANOVA with Šídák’s multiple comparison test. *p < 0.05, **p < 0.01 and ***p < 0.001 versus the wild-type (WT) group

CDK4/6 inhibitors are potential therapeutic agents for cisplatin-resistant BC

From above mentioned data, we observed that Cyclin D1 is a potential therapeutic target for the treatment of cisplatin-resistant BC (Fig. 1). Cyclin D/CDK4/6-p16-Rb pathway is reportedly dysregulated in cancer, leading to abnormal cell proliferation. Therapeutic agents selectively targeting this pathway have demonstrated antitumor effects in preclinical and clinical studies [18]. As CDKN2A (p16^INK4A), the inhibitor of CDK4/6, is frequently deleted in BC due to deletion on chromosome 9p21.3, Cyclin D/CDK4/6 complex becomes an attractive therapeutic target in BC [15]. We therefore examined the anti-proliferative activities of two CDK4/CDK6 inhibitors (Abemaciclib, Palbociclib) in cisplatin-resistant (N/P(14)) and sensitive (NTUB1) cells as well as normal uroepithelial (SV-HUC-1) cells. The data suggested that these inhibitors preferentially inhibit cell proliferation in cisplatin-resistant N/P(14) cells (Fig. 2A and B; Table 3). Cisplatin-resistant J82 (J82-Cis) cells generated in our previous study were used as an additional cell model to validate this observation. J82-Cis cells exhibited marked resistance to cisplatin (Supplementary Fig. S1A), accompanied by increased CCND1 transcription compared with parental J82 cells (Supplementary Fig. S1B). Importantly, Abemaciclib exhibited stronger anti-proliferative activity in J82-Cis cells than in J82 cells (Supplementary Fig. S1C). Currently, Abemaciclib is the only CDK4/6 inhibitor approved as a monotherapy for patients with HR⁺/HER2^– advanced or metastatic breast cancer with progressive disease [19]. Meanwhile, it displayed higher potency in cisplatin-resistant BC cells with little toxicity to normal cells, and we therefore chose Abemaciclib to study the pharmacological mechanisms. In summary, we identified CDK4/CDK6 selective inhibitors are potential therapeutic agents for cisplatin-resistant BC.

Fig. 2.

Effects of CDK4/6 inhibitors on the proliferation of BC cells. (A, B) NTUB1, N/P(14), and normal uroepithelial SV-HUC-1 cells were treated with Abemaciclib (A) or Palbociclib (B) for 48 h (left panels) or 72 h (right panels), and cell growth was assessed by SRB assay. Data are expressed as mean ± S.D. (n = 3). Statistical significance was determined by two-way ANOVA followed by Dunnett’s multiple-comparison test. *p < 0.05, **p < 0.01 and ***p < 0.001 versus the control group

Table 3.

GI₅₀ values of CDK4/6 inhibitors in BC and uroepithelial (SV-HUC-1) cells

Drug	Abemaciclib (µM)		Palbociclib (µM)
Cell name	48 h	72 h	48 h	72 h
N/P(14)	0.47	0.06	0.92	0.43
NTUB1	1.10	0.42	2.61	0.91
SV-HUC-1	1.30	2.32	> 10	> 2.5

The pharmacological mechanisms of Abemaciclib in cisplatin-resistant cells

To elucidate the pharmacological mechanisms of Abemaciclib in N/P(14) cells, we firstly analyze its effect on cell-cycle distribution by propidium iodide staining. The results showed the accumulation of the sub-G1 population in a concentration-dependent manner after 72 h treatment (Fig. 3A and B). We also observed that the phosphorylation of RB was decreased in a concentration- and time-dependent fashion in N/P(14) cells, indicating the suppression of CDK4/6 activity (Fig. 4A). cDNA microarray revealed that Abemaciclib decreased the expression of genes regulating cell cycle and DNA repair in 24 h (Fig. 4B). Meanwhile, GSEA data suggested that the pathways of E2F targets, DNA repairs, mitotic spindle and G2/M checkpoint were suppressed in N/P(14) cells upon the exposure to Abemaciclin in 24 h (Fig. 4C). Ingenuity Pathway Analysis (IPA) revealed that the network regulating cell cycle, cellular assembly and organization, DNA replication, recombination, and repair was decreased by Abemaciclib (Fig. 4D). These data suggest that Abemaciclib transcriptionally suppressed the genes involved in cell-cycle progression and DNA repair. The phenomenon was further confirmed by western blotting showing that Abemaciclib significantly decreased the protein levels of Cyclin A, Cyclin B. CDK1, CDK2, PLK1, Aurora A, Aurora B, Rad 51 and BRCA1, while the expression of Cyclin D, Cyclin E and CDK4 were upregulated (Fig. 5A). However, Abemaciclib treatment displays no pronounced apoptosis in N/P(14) cells after 72 h (Fig. 5B). Collectively, the results suggest that Abemaciclib mainly suppresses the growth of cisplatin-resistant BC cells by inhibiting the transcription of genes related to cell-cycle progression and DNA-repair, but not the promotion of apoptosis.

Fig. 3.

Effects of Abemaciclib on cell-cycle progression in N/P(14) cells. (A) Cells were treated with Abemaciclib at the indicated concentrations and time points, followed by flow cytometric analysis of cell-cycle distribution. (B) Quantitative data of the histograms are expressed as mean ± S.D. (n = 3). Statistical significance was determined using one-way ANOVA followed by Dunnett’s multiple-comparison test. *p < 0.05, **p < 0.01 and ***p < 0.001 versus the control group

Fig. 4.

Effects of Abemaciclib on the transcriptome of N/P(14) cells. (A) N/P(14) cells were treated with Abemaciclib at 0.3125, 0.625, and 1.25 µM for 24, and 48 h. Protein expression levels were examined by western blotting. Band intensities were quantified using ImageJ, and normalized to GAPDH. Fold changes relative to the untreated control are presented as mean ± S.D. (n = 3). Statistical comparisons were performed by using one-way ANOVA followed by Dunnett’s multiple comparisons test. (B–D) N/P(14) cells were treated with Abemaciclib (0.625 µM) for 24 h, and total RNA was isolated for microarray analysis. Differentially expressed genes were further analyzed by pathway analysis (B), GSEA (C) and Ingenuity Pathway Analysis (IPA) (D)

Fig. 5.

Effects of Abemaciclib on cell-cycle regulators and DNA repair proteins. (A) N/P(14) cells were treated with Abemaciclib at indicated concentrations for 24 h. Protein levels associated with cell-cycle regulation, mitotic progression, and DNA repair were analyzed by western blotting. (B) N/P(14) cells were treated with Abemaciclib at indicated concentrations or with taxol (0.05 µM) as a positive control, for 72 h. Protein levels associated with apoptosis were analyzed by western blotting. Band intensities were quantified using ImageJ, and normalized to GAPDH. Data are presented as fold changes relative to the untreated control (mean ± SD, n = 3). Statistical comparisons were performed using one-way ANOVA followed by Dunnett’s multiple comparisons test

The efficacy of Abemaciclib in combination with chemotherapeutic agents in cisplatin-resistant BC cells

The above mentioned data revealed that Cyclin D-knockdown sensitized BC cells to cisplatin (Fig. 1H and J) and that Abemaciclib downregulated the expression of genes associated with DNA repair (e.g. Rad 51, BRCA1) in N/P(14) cells (Fig. 5A), we hypothesized that Abemaciclib may increase the efficacy of cisplatin in BC cells. Our results demonstrated that Abemaciclib synergistically enhances the therapeutic effects of cisplatin in N/P(14) cells and RT112 cells, as indicated by combination index (CI) values below 1 (Fig. 6A and B). Consistently, Abemaciclib significantly potentiated the anti-proliferative effects of cisplatin in J82-Cis cells (Supplementary Fig. S1D). We next attempted to establish xenograft tumors using N/P(14) cells in bulb/c nude mice; however, these cells failed to form tumors. Therefore, the combinatorial efficacy was alternatively evaluated using an RT112 xenograft model. The combination of Abemaciclib and cisplatin significantly reduced tumor volume compared with either treatment alone without affecting the body weight (Fig. 6C and D). Given that cancer stem-like cells (CSCs) are reportedly enriched following cisplatin treatment and contribute to therapeutic resistance through enhanced self-renewal and clonogenic capacity [33, 34], we further evaluated whether Abemaciclib affects CSC-associated phenotypes by analyzing sphere-forming and clonogenic capacity in N/P(14) cells. Abemaciclib concentration-dependently reduced sphere size of N/P(14) cells (Fig. 7A). Moreover, Abemaciclib enhanced the anti-clonogenic effect of cisplatin in N/P(14) cells (Fig. 7B). Collectively, these results suggest that Abemaciclib not only suppresses the growth of cisplatin-resistant BC cells, but also sensitized them to cisplatin treatments.

Fig. 6.

Synergistic interaction between Abemaciclib and cisplatin. N/P(14) (A) and RT112 (B) cells were treated with cisplatin in the presence or absence of Abemaciclib for 48–72 h, and subjected to SRB assay. Data are represented as mean ± S.D. (n = 3). Statistical significance was performed using two-way ANOVA followed by Tukey’s multiple-comparisons test. Combination index (CI) values were calculated by CompuSyn software. (C) The combination of cisplatin and Abemaciclib exhibited significant antitumor activity in the RT112 xenograft model, and no obvious bodyweight loss were observed (D). Tumor volume and body weight data are presented as mean ± S.E.M. (n = 5), and statistical analyses were performed using two-way ANOVA followed by Tukey’s multiple comparisons test

Fig. 7.

Effects of Abemaciclib on cancer stem cell properties in N/P(14) cells. (A) 1.5 × 10⁴ viable N/P(14) cells were seeded into a ultralow attach 6 well in the presence or absence of Abemaciclib for 10 days. Mean diameter of the sphere was measure by ImageJ and are expressed as mean ± S.D. (n = 3). Statistical significance was determined using one-way ANOVA followed by Dunnett’s multiple-comparison test. Scale bar = 100 µm. (B) 1.5 × 10³ viable N/P(14) cells were seeded into a 6 well for 24 h, and treated with cisplatin in the presence or absence of Abemaciclib for 7 days. The colonies were fixed and stained with crystal violet (0.5% in 70% EtOH). The percentage of colony area was analyzed by Image J. Data are represented as mean ± S.D. (n = 3). Statistical significance was performed using two-way ANOVA followed by Tukey’s multiple-comparisons test

Discussion

In this study, we demonstrated that dysregulation of the Cyclin D1/CDK4/6 axis contributes to cisplatin resistance in BC. Transcriptomic analyses from both TCGA datasets and cisplatin-resistant N/P(14) cells consistently revealed the enrichment of E2F target genes and aberrant upregulation of CCND1 (Cyclin D1) (Fig. 1A, E, F). Knockdown of Cyclin D1 restored cisplatin sensitivity in resistant cells (Fig. 1H ~ J), and pharmacological inhibition of CDK4/6 by Abemaciclib preferentially suppressed the growth of cisplatin-resistant BC cells while sparing normal urothelial cells (Fig. 2). Importantly, Abemaciclib treatment attenuated DNA repair gene expression and synergized with cisplatin both in vitro and in xenograft models (Fig. 6), suggesting that targeting the Cyclin D1/CDK4/6–E2F pathway may represent an effective therapeutic strategy to overcome cisplatin resistance.

Previous studies have highlighted multiple mechanisms of cisplatin resistance, including altered drug uptake and efflux, enhanced DNA repair capacity, and impaired apoptotic signaling [13]. However, few targeted therapies have been developed to directly address these mechanisms. Transcriptomic analysis shows that regulators of the cell cycle and DNA damage checkpoints, including CDK2 and CHEK1, are central nodes in cisplatin resistance [21]. The consistent identification of Cyclin D1/CDK4/6 dysregulation across patient samples and resistant cell lines underscores its clinical relevance. These data also align with the frequent loss of p16^INK4A in BC [15–17], which further sensitizes tumor cells to CDK4/6 activity. Together, these observations provide a mechanistic rationale for the use of CDK4/6 inhibitors in this disease setting.

Among clinically approved CDK4/6 inhibitors, Abemaciclib is unique in its ability to be administered as monotherapy in breast cancer patients, and has shown broader activity in solid tumors compared to Palbociclib and Ribociclib [18]. Our data showed that Abemaciclib suppresses the transcription of homologous recombination repair genes (RAD51, BRCA1) in cisplatin-resistant N/P(14) cells (Fig. 5A). Meanwhile, we observed that homologous recombination activity was higher in cisplatin-resistant N/P(14) cells than in parental NTUB1 cells (Supplementary Fig. S2A). Treatment with Abemaciclib resulted in increased levels of DNA double-strand breaks marker γH2AX in N/P(14) cells (Supplementary Fig. S2B). Previous studies have reported that activation of FOXM1 confers platinum resistance, whereas CDK4/6 inhibition reduces FOXM1 phosphorylation and synergized with cisplatin in BC [17, 35]. The data showed that Abemaciclib markedly reduced FOXM1 protein levels in N/P(14) cells (Supplementary Fig. S2C), suggesting that FOXM1 downregulation may represent a mechanism by which CDK4/6 inhibition attenuates DNA repair capacity in cisplatin-resistant BC cells. These findings are consistent with the notion that CDK4/6 inhibition impairs DNA damage repair, thereby enhancing cisplatin-induced cytotoxicity [36]. Furthermore, our combination studies demonstrated a synergistic interaction between Abemaciclib and cisplatin, which translated into significant tumor growth inhibition in vivo (Fig. 6). These results suggest that CDK4/6 inhibitors may function as both direct antiproliferative agents and as chemo-sensitizers by compromising DNA repair capacity.

A major clinical challenge in BC management is BCG-unresponsive NMIBC, as many patients ultimately progress to muscle-invasive disease requiring radical cystectomy or cisplatin-based chemotherapy [37, 38]. Increasing evidence suggests that BCG failure is associated with immune escape mechanisms [39], prompting the development of bladder-preserving strategies aimed at restoring anti-tumor immunity [40, 41]. Interestingly, CDK4/6 inhibition has been reported to exert immunomodulatory effects, including enhancement of antigen presentation, activation of interferon signaling, and reduction of regulatory T-cell populations [42, 43]. In BC models, CDK4/6 inhibitors have also been shown to increase CD8⁺ T-cell infiltration and enhance responses to PD-1 blockade [44, 45]. However, the relevance of CDK4/6 inhibition in BCG-unresponsive NMIBC remains unclear. Future studies are needed to determine whether activation of the Cyclin D1–CDK4/6–E2F axis contributes to BCG resistance and whether CDK4/6-based strategies could be integrated with immunotherapy or cisplatin-containing regimens.

Nevertheless, several limitations should be acknowledged. First, although our in vitro data were validated in xenograft models using RT112 cells, the lack of tumorigenicity of N/P(14) cells limited the direct in vivo evaluation of cisplatin-resistant cells. Second, while Abemaciclib preferentially suppressed resistant cells, the molecular heterogeneity of BC suggests that not all patients may derive equal benefit. Biomarkers such as CCND1 overexpression or CDKN2A loss may help stratify patients most likely to respond. Third, our analyses focused on transcriptomic and protein-level changes; further studies incorporating phosphoproteomics and single-cell profiling may reveal additional resistance mechanisms and therapeutic targets.

In conclusion, our findings provide evidence that dysregulation of the Cyclin D1/CDK4/6–E2F pathway contributes to cisplatin resistance in BC, and that pharmacological inhibition of CDK4/6 with Abemaciclib suppresses resistant cell growth and enhances cisplatin efficacy. These results highlight the therapeutic potential of CDK4/6 inhibitors as a strategy for overcoming cisplatin resistance and support further clinical evaluation of Abemaciclib, alone and in combination with cisplatin, in patients with advanced or refractory BC.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization and study design: T.-H.H., H.-P.K., M.-C. C and C.-H.C.; data acquisition: T.-H.H., C.-H.P., Y.-C.L. and B.-J.L.; data analysis and interpretation: T.-H.H., M.-C.C. and Y.-C.L.; Writing - original draft preparation: T.-H.H., M.-C.C. and C.-H.C; Writing - review and editing: T.-H.H., C.-H.P., M.-C.C., H.-P.K., M.-C.C. and C.-H.C. Funding acquisition: H.-P.K., M.-C.C. and C.-H.C. Supervision: M.-C.C. and C.-H.C. All authors have read and agreed to the published version of the manuscript.

Funding

This study was supported by the National Science and Technology Council of Taiwan (NSTC 108-2320-B-038-056, NSTC 109-2320-B-038-029, NSTC 114-2320-B-038-037, NSTC 114-2320-B-038-034-MY3), and Taipei Medical University Hospital (110TMU-TMUH-01-3).

Data availability

The datasets generated and analyzed during the current study are available in the Gene Expression Omnibus (GEO) repository, [GSE313088 and GSE313089].

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.