Fidelity of end joining in mammalian episomes and the impact of Metnase on joint processing

Double Stranded Breaks (DSBs) are the most serious form of DNA damage and are repaired via homologous recombination repair (HRR) or non-homologous end joining (NHEJ). NHEJ predominates in mammalian cells at most stages of the cell cycle, and it is viewed as 'error-prone', although this notion has not been sufficiently challenged due to shortcomings of many current systems. Multi-copy episomes provide a large pool of genetic material where repair can be studied, as repaired plasmids can be back-cloned into bacteria and characterized for sequence alterations. Here, we used EBV-based episomes carrying 3 resistance marker genes in repair studies where a single DSB is generated with virally-encoded HO endonuclease cleaving rapidly at high efficiency for a brief time post-infection. We employed PCR and Southern blot to follow the kinetics of repair and formation of processing intermediates, and replica plating to screen for plasmids with altered joints resulting in loss of chloramphenicol resistance. Further, we employed this system to study the role of Metnase. Metnase is only found in humans and primates and is a key component of the NHEJ pathway, but its function is not fully characterized in intact cells.

Our careful re-examination of fidelity of NHEJ repair in mammalian cells carrying a 3' cohesive overhang at the ends revealed that the repair is efficient and highly accurate, and predominant over HRR. However, the background of the cells is important in establishing accuracy; with human cells perhaps surprisingly much more prone to generate deletions at the repaired junctions, if/when Metnase is abundantly expressed.

We found that repair of episomes by end-joining was highly accurate in 293 T cells that lack Metnase. Less than 10% of the rescued plasmids showed deletions. Instead, HEK293 cells (that do express Metnase) or 293 T transfected with Metnase revealed a large number of rescued plasmids with altered repaired joint, typically in the form of large deletions. Moreover, quantitative PCR and Southern blotting revealed less accurately repaired plasmids in Metnase expressing cells. Conclusions: Our careful re-examination of fidelity of NHEJ repair in mammalian cells carrying a 3' cohesive overhang at the ends revealed that the repair is efficient and highly accurate, and predominant over HRR. However, the background of the cells is important in establishing accuracy; with human cells perhaps surprisingly much more prone to generate deletions at the repaired junctions, if/when Metnase is abundantly expressed.