Trans-lesion synthesis and mismatch repair pathway crosstalk defines chemoresistance and hypermutation mechanisms in glioblastoma

Abstract

Almost all Glioblastoma (GBM) are either intrinsically resistant to the chemotherapeutical drug temozolomide (TMZ) or acquire therapy-induced mutations that cause chemoresistance and recurrence. The genome maintenance mechanisms responsible for GBM chemoresistance and hypermutation are unknown. We show that the E3 ubiquitin ligase RAD18 (a proximal regulator of TLS) is activated in a Mismatch repair (MMR)-dependent manner in TMZ-treated GBM cells, promoting post-replicative gap-filling and survival. An unbiased CRISPR screen provides an aerial map of RAD18-interacting DNA damage response (DDR) pathways deployed by GBM to tolerate TMZ genotoxicity. Analysis of mutation signatures from TMZ-treated GBM reveals a role for RAD18 in error-free bypass of O⁶mG (the most toxic TMZ-induced lesion), and error-prone bypass of other TMZ-induced lesions. Our analyses of recurrent GBM patient samples establishes a correlation between low RAD18 expression and hypermutation. Taken together we define molecular underpinnings for the hallmark tumorigenic phenotypes of TMZ-treated GBM.

Fig. 1. TMZ activates the RAD18 pathway in astrocytes and glioblastoma cells.

Representative immunoblot showing levels of mono-ubiquitinated PCNA and indicated DDR markers at different times after TMZ treatment in NHARas (a), U87 and U87 RAD18 ^−/− cells (b), or U373 and U373 RAD18 ^−/− cells (c). These experiments had been performed at least three times with similar results. d Immunofluorescence microscopy images showing localization of CFP-RAD18 to nuclear foci in U373 cells after 48 h treatment with 50 µM TMZ; Green: CFP-RAD18, Red: RPA32 (coating ssDNA), Blue: DAPI (nucleus). Scale bars, µM. The bar charts show the quantification of cells with CFP-RAD18 nuclear foci and RPA32-co-localizing nuclear foci. Data points represent mean ± SD of four independent experiments for RAD18 foci percentage (***p = 7.2992e-06) and ≥8 individual nuclei for RAD18 foci overlap with RPA (**p = 0.0091); e FACS analysis showing BrdU incorporation profiles in U373 cells transfected with siCon or siRAD18, followed by 50 µM TMZ treatment for 0, 3, 24, 48 h. BrdU median intensities are shown for gated late S and G2M populations of the 48 h treatment condition. f Representative immunoblots of two independent experiment showing levels of mono-ubiquitinated PCNA and other DDR markers in LN18 and LN18 RAD18 ^−/− cells after 200 μM TMZ-treatment in the presence or absence of 20 µM O6BG (MGMT inhibitor). g Quantification of primary TMZ-induced lesions in LN18 and LN18 RAD18^−/− cells following conditional O6BG treatment. Data points represent mean ± SD of three biological replicates. p values were determined by two-sided t test (d, g). NS no significance. Source data are provided as a Source Data file.

Fig. 2. RAD18 promotes TMZ tolerance in Astrocyte and GBM cell lines.

a Representative immunoblot showing the effect of 100 µM TMZ treatment (24 h) on PCNA mono-ubiquitination in NHA and NHARas cells transfected with siCon or siRAD18. b Effect of RAD18 knockdown on TMZ sensitivity in NHA and NHARas cells. (**p = 0.002, *p = 0.0174). Clonogenic survival assays showing TMZ-sensitivities of parental (RAD18 ^+/+) and RAD18 ^−/− U373 (c, ***p = 0.0006) and U87 cells (d, ***p = 0.0009). Clonogenic survival assays showing TMZ-sensitivities of RAD18 ^+/+ and RAD18 ^−/− LN18 cells treated with or without 20 µM O6BG (e, *p = 0.0128) or 5 mM MeOX (f, *p = 0.0111). g Clonogenic survival assay TMZ-sensitivities of RAD18 ^+/+ and RAD18 ^−/− U373 cells treated with (**p = 0.004) or without (***p = 1.0460e-05)10 mM MeOX. h, i TMZ sensitivity of RAD18 ^+/+ and RAD18 ^−/− U373 cells cultured on viable organo-type brain slice explants. The middle panel shows representative bioluminescence images of firefly luciferase-expressing U373 RAD18 ^+/+ and RAD18 ^−/− cells cultured on brain slices and treated with indicated concentrations of TMZ for 4 days. To measure TMZ sensitivity, bioluminescence on day 4 post-TMZ treatment was normalized to day 0 for each treatment group (n = 4). 100 µM TMZ ***p = 0.0003, 250 µM TMZ ***p = 9.4745e-05. Data points represent mean ± SD of triplicate determinations; P values were determined by two-sided t test (b–g, i). Experiment of (a–g) and (I) had been repeated at least three times with similar results. Source data are provided as a Source Data file. (h) was created with BioRender.com.

Fig. 3. RAD18 promotes tolerance of MMR-dependent DNA damage in TMZ-treated GBM.

a FACS analysis showing cell cycle profiles of RAD18^+/+ and RAD18^−/− U373 cells at different times post-treatment with 10 µM TMZ. b Cell cycle distribution of RAD18^+/+ and RAD18^−/− U373 cells after treatment with 10 µM TMZ for 48 h. Cells in G2 + M or M-phase were enumerated based on staining with mitosin and phospho-histone H3 respectively. *p = 0.0107. c FACS analysis of γH2AX and pRPA32s33 in WT, RAD18^−/−, MLH1^−/− and RAD18^−/−MLH1^−/− U373 cells after a 48 h treatment with TMZ. In the upper panel, the gated cell populations highlighted in red represent the pRPA32s33/γH2AX double-positive cells with 10 µM TMZ in U373 WT cells. The lower panel shows quantification of pRPA32s33/gH2AX double-positive cells in different genotypes after a 48 h treatment with 2 µM TMZ. *p = 0.027, **p = 0.0037, ***p = 0.001, NS, no significance. d Uniform Manifold Approximation and Projection (UMAP) scatterplots derived from scRNASeq analysis of GBM cells in patient samples. The UMAP plots show the clustering of individual cells based on RAD18 expression (left) and cell cycle phase (right). e Immunoblot showing levels of PCNA mono-ubiquitination and other DDR markers in WT, RAD18^−/−, MLH1^−/−, and RAD18^−/−MLH1^−/− U373 cells treatment with 10 µM TMZ. f Clonogenic survival assays showing TMZ-sensitivities of WT, RAD18^−/−, MLH1^−/−, and RAD18^−/−MLH1^−/− U373 cells. Cultures received a single treatment with TMZ daily for 5 days. ***p = 0.001. g Representative immunoblot of PCNA mono-ubiquitination in U373 cells transfected with siCon, siRAD18, siMSH2 or siRAD8 + MSH2, and treated conditionally with 50 µM TMZ. h Representative immunoblot showing levels of mono-ubiquitinated PCNA and indicated DDR proteins in RAD18^+/+ and Rad18^−/− LN18 cells after treatment with 100 µM TMZ for the indicated times. i Representative immunoblot showing levels of TMZ-induced PCNA mono-ubiquitination in U373 cells transfected with siCon or siEXO1. The experiments described in (e) and (g)–(i) were performed at least three times with similar results on each occasion. All data points represent mean ± SD of three biological replicates (b) and (c) and triplicate determinations from a single experiment (representative of three independent experiments) (f); p values were determined by two-sided t test (b) and two-sided t test with Tukey HSD adjust (c, f). Source data are provided as a Source Data file.

Fig. 4. RAD18-dependent TMZ-tolerance is mediated by Polκ.

a Schematic of TLS DNA polymerase recruited by RAD18 involved in ssDNA gap filling generated by futile MMR cycle. b Clonogenic survival assays showing effects of siPOLK, siPOLH, siPOLI, or JH-RE-06 treatments on TMZ sensitivity of U373 cells. The experiment was repeated twice with similar results. c Quantification of TMZ-induced GPF-Polκ foci in U373 RAD18^−/− (left) and U373 MLH1^−/− cells (middle) from at least three biological repeats. The right panel shows representative images of GFP-POLK foci in U373 and U373 RAD18^−/− cells treated with or without 20 µM TMZ for 48 h; Scale bars, 10 µm. d Clonogenic survival assay showing TMZ-sensitivities of WT, RAD18^−/−, POLK^−/− and RAD18^−/−POLK^−/− U373 cells. Cells were treated with a single dose of TMZ daily for 5 days. e Quantification of TMZ-induced YFP-Polη foci in RAD18^+/+ and RAD18^−/− U373 cells from at least 3 biological repeats. The right panels show representative images of cells containing YFP-Polη foci; Scale bars, 10 µm. f Clonogenic survival assays showing effect of POLK or POLH siRNAs on TMZ sensitivity of RAD18^+/+ and RAD18^−/− U373 cells. Cells were treated with the indicated doses of TMZ twice daily for 2 days. g Immunofluorescence microscopy images showing the distribution of GFP-Polκ and PCNA in U373 transfected with siCon, siPOLH and siRAD18, followed by 50 µM TMZ treatment for 48 h; Red: PCNA, Green: GFP-Polk, Blue: (DAPI); Scale bars, 8 µm. The bar chart shows quantification of foci containing co-localized GFP-Polκ and PCNA from three biological repeats. Experiment (d) had repeated at least three times and (f) had repeat twice with similar results. All data points in (d,f) represent the mean ± SD of triplicate determinations; p values of (b, d) were determined by two-sided t test, and (c,e,f,g) were determined by two-sided t test with Tukey HSD adjust. Source data are provided as a Source Data file.

Fig. 5. Results of CRISPR screen for RAD18-interacting DDR genes.

a Workflow of genetic screen. MOI Multiplicity of infection, BSD Blasticidin, PD Proliferation doubling, LD Lethal dose. b Normalized counts (Log10) of sgRNAs targeting DNA Damage Response (DDR) genes and non-targeting control across indicated samples in RAD18^+/+ (Blue) and RAD18^−/− (purple) U373 cells. ***p = 0.001. c Volcano plot showing Gene Abundance Change Scores (Sigma FC) vs -Log10 adjusted p-value for sgRNA depletion or enrichment in PD20 groups when compared with PD0. The −Log10 p value was calculated using a permutation test for DDR gene-targeting sgRNAs relative to non-targeting control sgRNAs. Black dashed lines indicate thresholds for statistical significance. Enriched sgRNAs targeting MMR genes, and depleted sgRNAs targeting POLD3, CHEK2 and PRKDC are highlighted. d Heatmap showing relative dropout of sgRNAs grouped by DDR pathway in RAD18^+/+ and RAD18^−/− cells cultured with or without TMZ for 20 PD. The numbers on the scale indicate −Log10 of paired t test p value (up) and Log2 fold change (down) of pooled sgRNA counts. BER Base excision repair; TLS Trans-lesion DNA synthesis; FA Fanconi Anemia; HR Homologous recombination; NHEJ Non-homologous end joining; NER Nucleotide excision repair; CS Checkpoint signaling; M/ASC Mitosis/spindle assembly checkpoint; PARP Poly ADP ribose polymerases; NM Nucleotide metabolism; TS Template switch. e Radar plot showing relative dropout (up: p value; down: Log2 fold change) of sgRNAs in DDR pathway between TMZ and DMSO control at PD20 in RAD18^+/+ and RAD18^−/− cells. f, g Dose response matrices and synergy heatmaps showing effects of pairwise combinations of TMZ with CHK2i (e) or DNA-PKi (f) on inhibition of viability in RAD18^+/+ and RAD18^−/− cells. Mean value of triplicates used to generate heatmaps. h Clonogenic survival assays showing TMZ sensitivity of WT, RAD18^−/−, POLD3^−/−, and RAD18^−/−POLD3^−/− U373 cells. Cultures received a single treatment with TMZ daily for 5 days. All data points represent the mean of triplicates ± SD. ***p = 0.001, *p = 0.036 Experiment had been repeated twice with similar results. i Heatmap showing sigmaFC of sgRNAs targeting Polζ complex genes and MIDAS/BIR pathway genes. p values were determined by two-sided t test with Tukey HSD adjust (b, h). Source data are provided as a Source Data file. a was created with BioRender.com.

Fig. 6. Effect of RAD18 on TMZ-induced mutagenesis.

a Experimental workflow of experiments to define impact of RAD18 and MMR on TMZ-induced mutagenesis. b Bar charts showing numbers of total SNV arising de novo following TMZ treatment over the course of 20 PD in cell lines differing with respect to RAD18 and MMR status. c, d Bar charts showing numbers of each individual SNV arising de novo following TMZ treatment over 20 PD in MLH1^+/+ (c) and MHL1^−/− (d) cells. Pattern I (O⁶mG-induced) and Pattern II (N³mA- and N⁷mG-induced) SNVs are indicated; NS no significance. e Stacked bar-chart showing contribution of individual COSMIC mutation signatures to the overall mutational patterns of U373 clones obtained from different treatment groups. f Bar chart showing effect of RAD18 and MLH1 status on mutational Signature 11 counts in clones of TMZ-treated U373 cells. NS, no significance. All data points represent mean ± SD of 6 single clone samples; p values were determined by two-sided t test (b, c, d) and two-sided t test with Tukey HSD adjust (f). (a) was created with BioRender.com.

Fig. 7. RAD18 suppresses hypermutation in recurrent GBM patients.

a Scatterplot showing the contribution of Signature 11 mutations to individual tumors from a cohort of TMZ-treated recurrent GBM (rGBM) patients. Tumor samples were stratified by MGMT status and RAD18 expression. RAD18 relative expression was corrected for proliferation and was designated high (upper tertile), medium, or low (lower tertile). Based on accepted convention, tumors harboring >500 signature 11 counts (indicated by the black dashed line) were considered “hypermutation”. b Model describing roles of RAD18, MMR, and MGMT in Hypermutation. (b) was created with BioRender.com.