Targeting HNRNPU to overcome cisplatin resistance in bladder cancer

Abstract

Purpose: The overall response of cisplatin-based chemotherapy in bladder urothelial carcinoma (BUC) remains unsatisfactory due to the complex pathological subtypes, genomic difference, and drug resistance. The genes that associated with cisplatin resistance remain unclear. Herein, we aimed to identify the cisplatin resistance associated genes in BUC. EXPERIMENTAL DESIGN: The cytotoxicity of cisplatin was evaluated in six bladder cancer cell lines to compare their responses to cisplatin. The T24 cancer cells exhibited the lowest sensitivity to cisplatin and was therefore selected to explore the mechanisms of drug resistance. We performed genome-wide CRISPR screening in T24 cancer cells in vitro, and identified that the gene heterogeneous nuclear ribonucleoprotein U (HNRNPU) was the top candidate gene related to cisplatin resistance. Epigenetic and transcriptional profiles of HNRNPU-depleted cells after cisplatin treatment were analyzed to investigate the relationship between HNRNPU and cisplatin resistance. In vivo experiments were also performed to demonstrate the function of HNRNPU depletion in cisplatin sensitivity.

Results: Significant correlation was found between HNRNPU expression level and sensitivity to cisplatin in bladder cancer cell lines. In the high HNRNPU expressing T24 cancer cells, knockout of HNRNPU inhibited cell proliferation, invasion, and migration. In addition, loss of HNRNPU promoted apoptosis and S-phase arrest in the T24 cells treated with cisplatin. Data from The Cancer Genome Atlas (TCGA) demonstrated that HNRNPU expression was significantly higher in tumor tissues than in normal tissues. High HNRNPU level was negatively correlated with patient survival. Transcriptomic profiling analysis showed that knockout of HNRNPU enhanced cisplatin sensitivity by regulating DNA damage repair genes. Furthermore, it was found that HNRNPU regulates chemosensitivity by affecting the expression of neurofibromin 1 (NF1).

Conclusions: Our study demonstrated that HNRNPU expression is associated with cisplatin sensitivity in bladder urothelial carcinoma cells. Inhibition of HNRNPU could be a potential therapy for cisplatin-resistant bladder cancer.

Keywords: Bladder urothelial carcinoma; Cisplatin; Genome-wide CRISPR screening; HNRNPU.

Fig. 1

The cytotoxicity of cisplatin in 6 bladder cancer cell lines. Cell viability curves for T24, RT4, HT1197, SW780, RT112, and HT1376 cells. IC50 values are calculated using GraphPad software (Prism 7.0). Data are expressed as mean ± SD from a representative of three independent experiments

Fig. 2

CRISPR-Cas9 screen to investigate mediators of cisplatin response in T24 cells. (A) Cell number count differences among different cisplatin/sgRNA combination treatments in the high-content screening assay. (B) Cell number foldchange of different cisplatin and sgRNA combination treatments compared to day 0. Error bar = SD

Fig. 3

Loss of HNRNPU in T24 cells mediates cisplatin sensitivity in vitro. (A) Cells were transfected with six sgRNAs per gene, and the fold change of sensitivity enhancement (FSE) of the 21 top decreased genes in T24 cells from our screen were determined using HCS analysis. Results were normalized to the control group. FSE was calculated as the cell count fold change of ((NC + drug) /NC) / ((Treatment + drug) / Treatment), NC, negative control (B) sgRNA of four selected genes were transfected into cells, and the FSE of four selected genes, including HNRNPU, CCDC1, PRKCDBP and OCLM, were detected. (C) CRIPSR-based sgRNA editing efficiency of the top 9 sgRNA that targets the selected genes. (D) HNRNPU protein expression levels in T24, RT4, HT1197, SW780, RT112, and HT1376 cells. Simple linear regression of HNRNPU expression levels and cisplatin IC₅₀ values in bladder cancer cells were calculated in GraphPad Prism 7.0 (E) Representative blot of HNRNPU in the cells after different treatments

Fig. 4

Knockout of HNRNPU regulated BUC progression and cisplatin sensitivity. (A) T24 cells were transfected with control sgRNA or sgHNRNPU prior to cisplatin treatment for 5 days, and cell viability was measured using CCK-8 assay. Arrows showed the error bars = 100 uM. (B) The cell cycle profiles of T24 cells with different treatments were detected by flow cytometry. Student’s t test was performed to detect the differences among treatments *: p < 0.05 **: p < 0.01(C) The apoptosis rate under the indicated treatments were examined by flow cytometry. (D) Cell migration of T24 cancer cells was determined by trans well assay. The data indicates mean ± SD from three independent experiments. (E) Photo of the isolated tumors with different treatments on day 28. (F) The tumor growth curve of different treatments, NC = negative control, KO = knockout of HNRNPU. (G) Weight of tumors in the indicated group

Fig. 5

TCGA bladder cancer cohort analysis on HNRNPU expression and its impact on survival. (A) Expression profile of HNRNPU in bladder tumors versus control tissues in TCGA dataset. (B) Pan-cancer analysis of HNRNPU expression in multiple cancer types. (C) Promoter methylation level of HNRNPU in TCGA dataset (D). Disease free survival analysis of high HNRNPU patients versus low HNRNPU patients. (E) Overall survival analysis of high HNRNPU patients versus low HNRNPU patients (high = top 20%). (F) Co-expression correlation between HNRNPU and HNRNPK. (G) Co-expression correlation between HNRNPU and AHCTF1; R = Pearson’s r (H) Overall survival analysis between AHCTF1 high patients versus AHCTF1 low patients (high = top 50%). (I) Expression profile of HNRNPU in TCGA bladder cancer immune cell subsets. GEPIA2021 was used to generate the plot

Fig. 6

Differential expression analysis of HNRNPU-depleted cells. (A) Volcano plot showing the differentially expressed genes in HNRNPU-depleted cells. (B) Pathway analysis of the down-regulated genes and up-regulated genes. (C) CNEplot, which depicts the linkages of genes and biological concepts, reveals the top enriched pathways and genes involved in the indicated pathways

Fig. 7

ATAC-seq analysis of HNRNPU-depleted cells. (A) Volcano plot showing the genomic regions with significant ATAC-seq signal alternations. Pathway analysis (B) and motif enrichment analysis (C) of the down-regulated regions in HNRNPU-depleted cells

Fig. 8

Knockout of HNRNPU regulated the BUC cell proliferation, apoptosis, and migratory through regulating NF1 expression. T24 cells were transfected with sgHNRNPU, with or without the NF1 knockdown, and then treated with or without cisplatin. (A) T24 cell viability was measured using CCK-8 assay. (B) Cell invasion of T24 cancer cells was determined by transwell assay. (C) Cell migration of T24 cancer cells was determined by wound healing assay. (D) The cell cycle of T24 cells with different treatments were detected by flow cytometry. (E) The apoptotic rate was examined by flow cytometry. The data indicate mean ± SD from three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001