PRMT5 inhibition sensitizes glioblastoma tumor models to temozolomide

Abstract

Background: Despite multi-model therapy of maximal surgical resection, radiation, chemotherapy, and tumor-treating fields, glioblastoma patients show dismal prognosis. Protein Arginine Methyltransferase 5 (PRMT5) is overexpressed in glioblastoma and its inhibition imparts an anti-tumor effect. Even though Temozolomide (TMZ) is the standard chemotherapeutic agent in the treatment of glioblastoma, tumor cells invariably develop resistance to TMZ. However, the mechanistic role of PRMT5 in glioblastoma therapy resistance is unknown.

Methods: Patient-derived primary glioblastoma neurospheres (GBMNS), treated with PRMT5 inhibitor (LLY-283) or transfected with PRMT5 target-specific siRNA were treated with TMZ and subjected to in vitro functional and mechanistic studies. The intracranial mouse xenograft model was used to test the in vivo antitumor efficacy of combination treatment.

Results: We found that PRMT5 inhibition increased the cytotoxic effect and caspase 3/7 activity of TMZ in GBMNS suggesting that apoptosis is the potential mode of cell death in the combination treatment. PRMT5 inhibition abrogated the TMZ-induced G2/M cell cycle arrest. Unbiased transcriptomic studies indicate that PRMT5 inhibition negatively enriches DNA damage repair genes. Importantly, combination therapy increased DNA double-strand breaks (H2AX foci) and enhanced the DNA damage (comet assay), suggesting that the combination treatment increases the TMZ-induced DNA damage. Specifically, the LLY-283 treatment blocked homologous recombination repair in GBMNS. In vivo, LLY-283 and TMZ combination significantly curbed the tumor growth and prolonged the survival of tumor-bearing mice.

Conclusion: Concomitant treatment of LLY-283 and TMZ has significantly greater antitumor efficacy, suggesting that PRMT5 inhibition and TMZ combination could be a new therapeutic strategy for glioblastoma.

Keywords: DNA damage repair; Glioblastoma; LLY283; PRMT5; TMZ.

Figure 1. PRMT5 inhibition increases the sensitivity of TMZ in GBMNS:

(A) Indicated GBMNS were treated with either TMZ and/or LLY283. 4 days post-treatment viability of the cells was measured by cell-titer glo assay. (B) GBMNS transfected with scrambled (Cntrl), or PRMT5-target specific siRNA (P5i) were treated with increasing doses of TMZ. 4 days post-treatment, cells were subject to viability assay (C) GBMNS were treated with TMZ and/or LLY283 for 48 hours and caspase3/7 activity was measured. (D) PRMT5-intact and depleted GBMNS were treated with increasing doses of TMZ, and caspase3/7 activity was measured 48 hours post-treatment with TMZ. n=3 (** p≤0.001).

Figure 2. PRMT5 inhibition abrogates TMZ-induced G2/M cell cycle arrest:

(A-D) PRMT5-intact and depleted GBMNS treated with TMZ for 48 hours were analyzed for the cell cycle progression using PI. (Upper panels) The graph represents the percent of the cell population in each stage of the cell cycle (**p ≤ 0.001). (Lower panels) Representative cell cycle histogram for each treatment condition across all the cell types that were analyzed for cell cycle progression.

Figure 3. PRMT5 inhibition downregulated DNA damage repair genes in GBMNS:

(A) Heatmap showing the correlation between PRMT5 and DNA repair genes across multiple tumor types. (B) GSC082209 treated with Control (DMSO) or LLY-283 for 24 hours were subjected to RNA sequencing. GSEA enrichment plot shows the correlation of LLY-283 with pan-DNA damage repair genes. (C) GSC040815 and GSC082209 treated with LLY-283 (50 μM) were probed for indicated DNA damage repair proteins by western blot. (D) Top differentially expressed genes based on the RNA sequencing analysis of panel B. (E) GSEA enrichment analysis showing a negative correlation between LLY-283 treatment and HR repair gene set. (F) Heatmap showing the genes that are differentially expressed based on panel E gene enrichment analysis. (G) Scatter plot from TIMER2.0 database showing correlation between PRMT5, and HR genes (RAD51 and POLD1) based on glioblastoma TCGA data sets. p-value computed for each data set. (H) Schematic representation of qPCR-based HR assay protocol. (I) GSC040815 and GSC082209 treated with LLY-283 were subjected to HR activity. n=3 (** p≤0.001).

Figure 4. PRMT5 inhibition enhances the TMZ-induced DNA damage:

(A, B) GBMNS treated with LLY-283 and/or TMZ for 48 hours were probed for H2AX foci by immunofluorescence and quantified. (**p ≤ 0.001). (C, D) PRMT5 transfected cells (P5i) treated with TMZ were probed for H2AX foci by immunofluorescence and the number of foci were quantified manually. (**p ≤ 0.001). (E, F) GBMNS treated with LLY-283 and/or TMZ for 48 hours were subjected to comet assay. DNA damage was graded/quantified from 0 to 4 based on the tail length and comet head size. (0 = no, 1= mild, 2 = moderate, 3 = high, 4= very high DNA damage. (**p ≤ 0.001). (G, H) GBMNS transfected with PRMT5 (P5i) were treated with TMZ for 48 hours and were subjected to comet assay. DNA damage was graded/quantified from 0 to 4 based on the tail length and comet head size. (0 = no, 1= mild, 2 = moderate, 3 = high, 4= very high DNA damage. (**p ≤ 0.001).

Figure 5. LLY-283 blocks the TMZ-induced HR repair in GBMNS:

(A) GSEA enrichment analysis showing negative enrichment of DNA repair gene sets in GSC082209 treated LLY-283 and TMZ combination. (B) Heatmap for showing differential gene expression based on panel A. (C) Top gene ontology terms of differentially expressed genes pointing out HDR through HR. (D) GSC040815 and GSC082209 treated with LLY283 (50 μM), TMZ (50 μM), or the combination of LLY-283 and TMZ, were probed for HR marker RAD51 by western blot. (E) Quantification of panel D showing the expression of RAD51 (**p ≤ 0.001).

Figure 6. Combination of LLY-283 and TMZ enhances the in vivo antitumor efficacy:

(A) Schematic representation of the in vivo study. (B) Mice were implanted with GSC040815 that expresses GFP-Luciferase. Post-implantation mice were treated with different treatment conditions and the Kaplan–Meier survival curve was plotted at the end of the study. (C) Quantification of the tumour volume based on the luciferase images generated during the study (D) Shown is the working model for the PRMT5-inhibition triggered apoptosis in the TMZ-treated GBMNS.