FIGURE 2.

RAD50 is important to limit distal end use, whereas H2AX appears dispensable. A, RAD50 depletion causes a shift toward distal end use. RAD50 was depleted from U2OS cells by transfecting a pool of four targeting siRNAs, which are compared with transfection of a nontargeting siRNA (siCTRL). Shown are immunoblot signals for RAD50 and GAPDH from such transfections. Subsequent to this depletion, the siRNAs were co-transfected with an expression vector for the I-SceI-Trex2 fusion. Shown are Distal-EJ, proximal EJ, and relative distal end use (siCTRL = 1) values from these experiments. Asterisks denote a statistical difference from siCTRL samples (p < 0.0001). B, genetic disruption of H2AX has no clear effect on distal end use. The EJ5-GFP reporter was integrated into the pim1 locus into WT (H2AX^f/f) and H2AX^−/− mES cells, and these cell lines were co-transfected in parallel with expression vectors for I-SceI and Trex2, and treated with ATMi (10 μm), PKi (20 μm), or vehicle (DMSO). Shown are immunoblot signals for H2AX along with Ponceau S staining of the same blot for histone extracts from these cell lines. Shown are relative distal end use values for these experiments (WT, DMSO = 1), calculated from the ratio of distal EJ and proximal EJ frequencies for individual samples. Asterisks denote a statistical difference from DMSO-treated samples (p < 0.0001). The bracket depicts no statistical difference for the frequency of distal end use between WT and H2AX^−/− samples.