A Dual Approach with Organoid and CRISPR Screening Reveals ERCC6 as a Determinant of Cisplatin Resistance in Osteosarcoma

Abstract

Osteosarcoma (OS), the most prevalent primary bone malignancy in adolescents, is typically treated with cisplatin-based chemotherapy. However, the development of cisplatin resistance often leads to relapse or metastasis, significantly impairing therapeutic efficacy. To tackle this issue, patient-derived osteosarcoma organoids (OSOs) is established that accurately reflect the cellular composition and heterogeneity of the original tumors, as validated by single-cell RNA sequencing, bulk RNA sequencing, and histopathology analysis. Cisplatin resistance is successfully induced in these OSOs, creating a clinically relevant model for investigating chemoresistance. Utilizing RNA sequencing in cisplatin-resistance OSOs and CRISPR screening in OS cell line, ERCC6 is identified as a pivotal regulator of cisplatin resistance. Knockdown of ERCC6 markedly enhanced cisplatin sensitivity in vitro and in vivo. Mechanistically, ERCC6 interacts with HNRNPM, influencing the PI3K/AKT signaling pathway and alternative splicing of pre-mRNA for BAX. Notably, the knockdown of ERCC6 and HNRNPM increased expression of full-length BAX and reduced skipping of exon 2, thus promoting apoptosis. This exon skipping in BAX results in a frameshift and introduces a premature stop codon (TGA) within the BH3 domain. These findings underscore the utility of OSOs in elucidating resistance mechanisms and highlight ERCC6 and HNRNPM as potential therapeutic targets.

Keywords: ERCC6; alternative splicing; chemoresistance; organoid; osteosarcoma.

Figure 1

Establishment and Characterization of OSOs. A) Schematic representation of the standardized protocol for constructing OSOs, including tissue collection, enzymatic digestion, and culture. B) Growth curves of two OSOs (OS189‐O and OS198‐O), demonstrating the capacity for sustained growth over time, scale bars, 50 µm. C) PCA of RNA‐sequencing data from three patients, showing distinct clustering of normal, tumor, and OSO samples. D) Hierarchical clustering heatmap of gene expression in normal tissues, tumors, and OSOs. E) Heatmap of the top 25 upregulated and downregulated genes in tumor samples compared to normal tissues, showing that OSOs recapitulate the differential expression patterns of tumors. F,G) UMAP and bar plots comparing the cellular composition of paired tumor and OSO samples at the single‐cell level. H,I). UMAP and bar plots show the presence of CD45+ immune cells in OSOs. J) UMAP plot of pooled OSO samples from three patients, demonstrating heterogeneity in cellular composition across different patients. K) UMAP with patient sample annotation, highlighting patient‐specific differences in cell type distribution. L) Violin plots depict the expression of key genes (CD99, MKI67, RB1, SATB2, TP53, and VIM) across OSOs from three patients. They show variability in gene expression and reflect patient‐specific heterogeneity. M. IHC staining of tumor tissue and IF staining of OSO sections for CD99, VIM, CD68, KI67, and SATB2, showing positive expression in both tumor and OSO tissues, scale bars, 20 µm. ****P < 0.001.

Figure 2

Identification of ERCC6 as a Cisplatin Resistance Gene. A) Time course of cisplatin treatment to induce resistance in three OSOs (OS395‐O, OS396‐O, OS403‐O). (Created by FigDraw, ID: UOURT03073). B) Representative bright‐field images comparing cisplatin‐resistant and non‐resistant OSOs. Scale bars, 50 µm. C) IC50 comparison between cisplatin‐resistant and non‐resistant OSOs for three patient‐derived organoids (OS395‐O, OS396‐O, OS403‐O), showing a significant increase in IC50 in resistant OSOs. (+), with cisplatin; (−), without cisplatin. D) Correlation heatmap of gene expression in resistant and non‐resistant OSOs, showing high intra‐group correlation. (+), with cisplatin; (−), without cisplatin. E) Volcano plot of differentially expressed genes between cisplatin‐resistant and non‐resistant OSOs, highlighting key upregulated and downregulated genes. F) Schematic of CRISPR screen workflow, combining cisplatin‐induced resistance in OSOs with CRISPR/Cas9 screening. (Created by FigDraw, ID: UYWPPcc232). G,H) RT‐qPCR showing increased ERCC6 mRNA expression in cisplatin‐treated 143B and HOS cells. I) WB analysis of ERCC6 protein levels in cisplatin‐treated 143B and HOS cells. J) IHC analysis of ERCC6 expression in chemotherapy‐responsive versus chemotherapy‐resistant patient tumor samples, with representative images from each group. K) IHC scoring reveals that ERCC6 expression is significantly higher in chemotherapy‐resistant patients. Data are presented as the mean ± SD. ns > 0.05, ***P < 0.001, ****P < 0.0001.

Figure 3

Functional Validation of ERCC6 in Cisplatin Sensitivity. A) WB confirming ERCC6 knockdown in 143B and HOS cell lines using two sgRNAs. B) IC50 values showing increased cisplatin sensitivity in ERCC6 knockdown cells compared to control cells. C) Clonogenic assay results showing reduced colony formation in ERCC6 knockdown cells under cisplatin treatment. (0.5 µm Cisplatin for 143B and HOS cells). D) CCK8 demonstrating decreased proliferation rates in ERCC6 knockdown cells when exposed to cisplatin. E) Flow cytometry analysis showing increased apoptosis rates in ERCC6 knockdown cells treated with cisplatin, as indicated by Annexin V staining. (2 µm Cisplatin for 143B and HOS cells) F) Following sgERCC6 (#1) knockdown, WB results and IC50 detection were conducted on three patient‐derived organoids (OS551‐O, OS559‐O, and OS560‐O). G) Representative bright‐field images and cell viability assays of patient‐derived organoids (OS551‐O, OS559‐O, and OS560‐O) following sgERCC6 (#1) knockdown under 5 µm cisplatin treatment. H) Western blot analysis of caspase‐3 and cleaved caspase‐3, showing increased levels of cleaved caspase‐3 in ERCC6 knockdown cells under cisplatin treatment. I) Experimental design of an OS model under ERCC6 knockdown conditions (Cisplatin, 2 mg kg⁻¹, once every three days, intraperitoneally). (Created by FigDraw, ID: IITWPf1f3f). J) Tumor images. K) Tumor growth curves showing that ERCC6 knockdown significantly inhibited tumor growth in the presence of cisplatin. L) Final tumor weights from the xenograft model, confirming the inhibitory effect of ERCC6 knockdown on tumor growth under cisplatin treatment. Data are presented as the mean ± SD. ns > 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Figure 4

ERCC6 Regulation of AKT Pathway and Identification of Binding Partner HNRNPM. A) Volcano plot of differentially expressed genes from RNA‐seq analysis of ERCC6 knockout 143B cells. B) KEGG pathway analysis shows significant enrichment of the PI3K/AKT pathway in ERCC6 knockout cells. C,D) Gene set enrichment analysis showing reduced expression of “Nucleotide Excision Repair” and “ABC transporters” pathways in ERCC6 knockout cells. E) WB analysis of AKT and p‐AKT, showing reduced p‐AKT levels in ERCC6 knockdown 143B and HOS cells. F) Coomassie blue staining of Co‐IP experiment identifying ERCC6 binding partners, including HNRNPM. G) The IC50 value of cisplatin for potential co‐acting proteins in 143B cells. H) Mass spectrometry confirming HNRNPM as an ERCC6 binding partner. I) Co‐IP experiments in 293T and 143B cells confirming ERCC6‐HNRNPM interaction. J) GST pull‐down assay identifying the interaction domains between ERCC6 and HNRNPM. K) IF co‐localization of ERCC6 and HNRNPM in 143B cells, scale bars, 20 µm. L) Positive correlation between ERCC6 and HNRNPM expression, as shown by analysis of multiple public datasets (TARGET, TCGA, GSE21257, and GSE218035).

Figure 5

Functional Validation of HNRNPM in Cisplatin Sensitivity. A) WB confirming HNRNPM knockdown in 143B and HOS cell lines using sgRNAs. B) IC50 values showing increased cisplatin sensitivity in HNRNPM knockdown cells. C) Clonogenic assay results showing reduced colony formation in HNRNPM knockdown cells under cisplatin treatment. D) CCK8 demonstrating decreased cell proliferation in HNRNPM knockdown cells under cisplatin treatment. E,F) Flow cytometry analysis showing increased apoptosis rates in HNRNPM knockdown cells treated with cisplatin. (2 µm Cisplatin for 143B and HOS cells). G) Tumor images from the xenograft assay in mice. H,I) IHC staining of mouse tumor sections showing decreased Ki‐67 and increased cleaved caspase‐3 expression in HNRNPM knockdown tumors treated with cisplatin, scale bars, 50 µm. J) Tumor growth curves from the xenograft model showing reduced tumor growth in HNRNPM knockdown tumors treated with cisplatin. K) Final tumor weights, confirming the tumor‐inhibitory effect of HNRNPM knockdown under cisplatin treatment. Data are presented as the mean ± SD. ns > 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Figure 6

Interaction between ERCC6 and HNRNPM in Modulating PI3K/AKT Pathway. A) WB showing that ERCC6 knockdown reduces HNRNPM expression under cisplatin treatment, while HNRNPM knockdown increases ERCC6 expression. B) Reanalysis of data from Passacantilli et al. shows that PI3K inhibitor BEZ235 elevates ERCC6 and HNRNPM levels. C) Volcano plot of differentially expressed genes from RNA‐seq analysis of HNRNPM knockdown cells, highlighting significant changes. D,E) KEGG pathway analysis showing enrichment in the PI3K/AKT and cisplatin resistance pathways in HNRNPM knockdown cells. F) Western blot showing reduced p‐AKT levels in HNRNPM knockdown cells, with unchanged total AKT levels. G) p‐AKT levels in ERCC6 knockdown, HNRNPM knockdown, and double knockdown cells, with no significant difference between single and double knockdowns.

Figure 7

Alternative Splicing Modulation by ERCC6 and HNRNPM. A) Pie chart and volcano plot showing significant alternative splicing events following ERCC6 knockdown. B) Pie chart and volcano plot showing alternative splicing events following HNRNPM knockdown, highlighting the role of BAX. C) Bar chart categorizing 56 common alternative splicing events between ERCC6 and HNRNPM knockdowns. D) IGV visualization of BAX exon 2 skipping across control and knockdown groups. E) Schematic illustrating two BAX splicing variants: full‐length BAX and the truncated version resulting from exon 2 skipping, leading to a premature stop codon. F) Primer designs for exons 1 and 3 (for agarose gel electrophoresis) and exons 2 and 3 (for qRT‐PCR). G) qRT‐PCR results showing increased full‐length BAX expression in knockdown groups compared to controls. H) Agarose gel electrophoresis showing full‐length and skipped BAX bands in all four groups, with increased full‐length BAX in knockdown groups. I) Western blot showing increased BAX protein levels in knockdown groups, with no significant change in BCL2 expression. J) Mechanistic diagram summarizing the role of ERCC6 and HNRNPM in modulating BAX splicing and the PI3K/AKT pathway, influencing cisplatin resistance. Data are presented as the mean ± SD. *P < 0.05, **P < 0.01, and ***P < 0.001.