Pro-survival AKT and ERK signaling from EGFR and mutant EGFRvIII enhances DNA double-strand break repair in human glioma cells

Abstract

The epidermal growth factor receptor (EGFR) is frequently dysregulated in malignant glioma that leads to increased resistance to cancer therapy. Upregulation of wild type or expression of mutant EGFR is associated with tumor radioresistance and poor clinical outcome. EGFR variant III (EGFRvIII) is the most common EGFR mutation in malignant glioma. Radioresistance is thought to be, at least in part, the result of a strong cytoprotective response fueled by signaling via AKT and ERK that is heightened by radiation in the clinical dose range. Several groups including ours have shown that this response may modulate DNA repair. Herein, we show that expression of EGFRvIII promoted gamma-H2AX foci resolution, a surrogate for double-strand break (DSB) repair, and thus enhanced DNA repair. Conversely, small molecule inhibitors targeting EGFR, MEK, and the expression of dominant-negative EGFR (EGFR-CD533) significantly reduced the resolution of gamma-H2AX foci. When homologous recombination repair (HRR) and non-homologous end joining (NHEJ) were specifically examined, we found that EGFRvIII stimulated and CD533 compromised HRR and NHEJ, respectively. Furthermore, NHEJ was blocked by inhibitors of AKT and ERK signaling pathways. Moreover, expression of EGFRvIII and CD533 increased and reduced, respectively, the formation of phospho-DNA-PKcs and -ATM repair foci, and RAD51 foci and expression levels, indicating that DSB repair is regulated at multiple levels. Altogether, signaling from EGFR and EGFRvIII promotes both HRR and NHEJ that is likely a contributing factor towards the radioresistance of malignant gliomas.

Figure 1

Epidermal growth factor receptor signaling is important for the resolution of γ-H2AX foci. U87/HRR-GFP cells were grown in chambered glass slides and 24 h after seeding cells were: (A) Treated with AG1478 (5 µM) or PD184325 (3 µM) for 1 h or (B and C) Infected with adenovirus expressing EGFRvIII, EGFR-CD533 or LacZ (3 MOI) for 48 h. Cells were then exposed to 2 Gy of IR, fixed with para-formaldehyde at the indicated times and prepared for confocal microscopy. γ-H2AX foci per nucleus are depicted in graph form. Data points, γ-H2AX foci per nucleus; Error bars, SEM for data sets n = 10. *p < 0.05: **p < 0.01; ***p < 0.001; ns = non significant. (C) Shows representative images of γ-H2AX foci. Parallel cells were stained with anti-EGFR antibody. DAPI staining shows nuclei.

Figure 2

Genetic manipulation of EGFR status modulates HRR and NHEJ. U87/HRR-GFP or U87/NHEJ-DsRed (clone A2) cells were infected with `adeno-viruses expressing LacZ, EGFRvIII or EGFR-CD533 (3 MOI) and Ad-SceI (30 MOI). (A) GFP-HRR⁺ or (B) DsRed-NHEJ⁺ events were determined by FACS 48 h after Ad-SceI infection. Fold (x) indicates changes in the relative repair levels when compared to the Ad-LacZ control. *p < 0.05: **p < 0.01; ***p < 0.001; ns = non significant. (Columns) GFP⁺ or DsRed⁺ cells/60 K cells; (Error bars) SEM for data sets n = 3. (C) Parallel cultures were collected for western blot analysis 48 h after Ad-SceI infection. Fold depicts changes in phospho-ERK1/2 levels compared to control (Ad-LacZ) and normalized to total ERK2 protein levels. (D) Parallel cultures were collected for cell cycle analysis. The table depicts number of cells in each cell cycle phase as determined by propidium iodide staining and FACS.

Figure 3

EGFR-CD533 blocks NHEJ repair at unique I-SceI sites. (A) NHEJ repair kinetics. U87/NHEJ-DsRed (clone A3) cells were infected with Ad-SceI adenovirus and collected at 4, 8, 16 and 24 h post-infection. DNA extraction and qPCR was carried out as described in the Materials and Methods. (Columns) Relative NHEJ levels normalized to β-actin; (Error bars) SEM for data sets n = 3. Fold (x) indicates changes in the relative repair levels when compared to the 8 h sample. *p < 0.05: **p < 0.01; ***p < 0.001. (B) EGFR-CD533 blocks NHEJ. U87 cells were infected with an MOI of 10 with either Ad-LacZ or EGFR-CD533. Two days later cells were super-infected with Ad-SceI, and cells collected 24 h later. Columns, relative NHEJ levels normalized to β-actin; Error bars, SEM for data sets n = 3. Fold (x) indicates changes in the relative repair levels when compared to the Ad-LacZ control. *p < 0.05: **p < 0.01; ***p < 0.001.

Figure 4

EGFR signaling affects the localization of phosphorylated DNA-PKcs (T2609) and ATM (S1981) to the DSB. U87/HRR-GFP cells were grown in chambered glass slides, and infected with adenovirus expressing EGFRvIII, EGFR-CD533 or LacZ (3 MOI). Forty-eight h after infection cells were exposed to 5 Gy of IR, fixed with paraformaldehyde 15 min post IR, and prepared for confocal microscopy. (A) p-DNA-PKcs (T2609) foci. p-DNA-PK foci per nucleus are depicted in graph form. (B) p-ATM (S1981) foci. p-ATM foci per nucleus are depicted in graph form. DAPI staining shows nuclei. Fold (x) indicates changes in foci formation compared to Ad-LacZ control. Columns, foci per nucleus; Error bars, SEM for data sets n = 200. *p < 0.05: **p < 0.01; ***p < 0.001; ns = non significant.

Figure 5

EGFR signaling stimulates the formation of RAD51 foci and expression of RAD51. U87/HRR-GFP cells were grown in chambered cover slips, and infected with adenovirus expressing EGFRvIII, EGFR-CD533 or LacZ (3 MOI). Forty-eight h after infection cells were exposed to 5 Gy IR. 8 h after IR cells were fixed with paraformaldehyde and prepared for confocal microscopy. (A) Cells were exposed to anti-RAD51 antibody. Cells were scored as either positive or negative for RAD51 foci and depicted in graph form. Fold (x) indicates changes in the percentage of cells positive for RAD51 foci compared to the Ad-LacZ control. (Columns) Percentage of cells positive for RAD51 foci; (Error bars) SEM for three data sets with n = 200. *p < 0.05: **p < 0.01; ***p < 0.001; ns = non-significant. (B) Parallel cultures were collected for western blot analysis 48 h after infection. Fold (x) depicts RAD51 expression when compared to Ad-LacZ control after normalization to ERK2 protein levels.

Figure 6

Inhibition of EGFR, AKT and MEK/ERK suppresses NHEJ. (A) U87/NHEJ-DsRed (A3 clone) cells were infected with Ad-SceI and exposed to AG1478 (5 µM), PD184352 (3 µM), or SH-5 (20 µM) 2 h post infection. The drugs were kept in the medium throughout the experiments. Cells were collected 10 h post infection and prepared for qPCR. (Columns) Relative NHEJ levels normalized to β-actin; (Error bars) SEM for data sets n = 7. Fold (x) indicates changes in the relative repair levels when compared to the Control (+Ad-SceI, −drug). *p < 0.05: **p < 0.01; ***p < 0.001. (B) Western blot of extracts from U87 cells exposed to SH-5 at either 10 or 20 µM and then irradiated (5 Gy) or not. Cells were collected after 15 min and prepared for western blotting using p-(S473) AKT and AKT (loading control) antibodies. Fold (x) depicts changes in AKT phosphorylation levels compared to control (-IR, -SH-5) and normalized to total AKT protein expression.