Poly ADP Ribose Polymerase Inhibitors Potentiate Proton Therapy End-of-Range Effects by Accelerating Replication Forks and Promoting Transcription Conflict

Abstract

Purpose: Poly ADP ribose polymerase inhibitors (PARPi) are being combined with photon and proton radiation therapy in clinical trials. We sought to investigate mechanisms of PARPi radiosensitization at varying linear energy transfer (LET) levels after observing an extreme normal tissue response in an 18-year-old with high-grade glioma, without a germline alteration predictive of heightened radiosensitivity, treated with veliparib and proton therapy.

Methods and materials: BRCA1/2 wild-type noncancerous and cancerous cells were treated with PARPi plus photons or protons at the entrance (dose-averaged LET 2.2 keV/µm) or the Bragg peak (BP, dose-averaged LET 7.0 keV/µm) of the proton beam profile. DNA fiber, immunofluorescence, and other DNA damage signaling assays were used to evaluate replication fork progression, gap formation, transcription-replication conflicts, and DNA damage signaling and repair.

Results: PARPi modestly sensitized cells to photons and low LET protons; however, PARPi-treated cells were hypersensitive to high LET protons administered at the BP. Unexpectedly, cells treated with PARPi plus BP protons displayed accelerated replication fork progression, enhanced single-stranded DNA gap formation, and greater transcription-replication conflict-induced DNA double-strand breaks. Despite evidence of more single-stranded DNA and cells arrested in G2/M following PARPi plus proton BP, PARPi decreased RAD51 recruitment to break sites and enhanced cytotoxic error-prone repair by nonhomologous end joining.

Conclusions: PARPi renders cells hypersensitive to end-of-range high LET proton irradiation. The potential enhanced proton radiosensitization should be considered during clinical trial design with efforts to limit high physical dose and high LET overlap within normal tissues. Planning techniques that increase the LET within tumors warrant further investigation in combination with PARPi as a novel strategy to overcome therapeutic resistance.

Figure 1.. Radiation necrosis centered in an area of high LET within the target volume

(A–C) Axial T1-weighted contrast-enhanced MRI obtained prior to treatment (A), with corresponding overlays of the proton radiotherapy plan showing physical dose distribution (B) and Monte Carlo–based dose-weighted LET (MCBD) (C). (D–F) Axial T1-weighted contrast-enhanced MRI acquired seven months post-treatment (D), overlaid with the same physical dose distribution (E) and MCBD map (F) as in B–C. White arrows indicate the area of developing radiation necrosis, which corresponds spatially to regions receiving the highest LET as shown in the MCBD overlay.

Figure 2.. PARPi treated cells are hypersensitive to high LET proton BP irradiation

(A) U2OS cells were treated with olaparib (40 nM), talazoparib (0.2 nM), veliparib (2 μM), and varying doses of low-LET ENT and high-LET BP protons and the surviving fraction was assessed. (B–C) The ratio between the surviving fraction after a delivered physical dose of 2 Gy (B) or 4 Gy (C) ENT protons relative to BP protons in U2OS cells.

Figure 3.. PARPi accelerates replication forks and induces TRCs after proton BP irradiation

U2OS cells were pre-treated for 12 hours with olaparib (1 μM) or talazoparib (2 nM) and then subjected to mock treatment or 4 Gy delivered with the ENT or BP of the proton beam profile. The length of dual-color tracks in the DNA fiber assay with olaparib (A) and talazoparib (B) is shown. (C) U2OS cells were pre-treated with control or olaparib (10 μM) and then irradiated with 4 Gy protons delivered with the ENT or the BP of the proton beam profile. Cells were fixed for assessment of γH₂AX foci number 60 minutes after irradiation in the presence or absence of DRB (100 μM). Data represent three independent biological replicates. Each red circle denotes the median value from a single replicate, and the red bar represents the mean of these medians. Statistical comparisons were performed across biological replicates.

Figure 4.. PARPi induces ssDNA gaps and activates the replication stress response

U2OS cells were pre-treated with control or olaparib (10 μM) and then irradiated with 4 Gy protons delivered with the ENT or the BP of the proton beam profile. (A) Quantification of the length of the CldU track with or without S1 nuclease following olaparib (10 μM) monotherapy or combination therapies. Percent reduction in CldU track length after S1 treatment is shown above each bar, calculated relative to the matched untreated group. (B) The number of RPA32 foci were evaluated using IF at 4- and 24 hours following treatments. Data represent three independent biological replicates. Each red circle denotes the median value from a single replicate, and the red bar represents the mean of these medians. Statistical comparisons were performed across biological replicates.

Figure 5.. Enhanced replication associated DSBs following PARPi plus proton BP combination therapy

U2OS cells were treated with control, olaparib (40 nM), or veliparib (2 μM) with or without 4 Gy delivered with the ENT or the BP of the proton beam. (A-B) The number of γH₂AX foci (A) and 53BP1 foci (B) were evaluated using IF at 4- and 24 hours. For immunofluorescence, data represent three independent biological replicates. Each red circle denotes the median value from a single replicate, and the red bar represents the mean of these medians. Statistical comparisons were performed across biological replicates. (C) Flow cytometry analysis of the cell cycle 24- and 72 hours following irradiation.

Figure 6.. NHEJ repair pathway was upregulated by PARPi and proton BP irradiation

U2OS cells were treated with PARPi and irradiation 4Gy delivered by proton beam profile. (A) The number of RAD51 foci was evaluated using IF at 4- and 24 hours. Data represent three independent biological replicates. Each red circle denotes the median value from a single replicate, and the red bar represents the mean of these medians. Statistical comparisons were performed across biological replicates. (B) Survival fractions of cells treated with olaparib (40 nM), talazoparib (0.2 nM), veliparib (2 μM) in combination with proton ENT or BP after silencing XRCC4.