Regulation of DNA ligase activity by poly(ADP-ribose)

Nuclear ADP-ribosyl transferase (ADPRT)1,2 catalyses the transfer of ADP-ribose from NAD+ to chromatin proteins to form mono-, oligo- and poly (ADP-ribose)-modified chromatin proteins3–7. The enzyme is entirely dependent on DNA for activity8 and is activated by nicks in the DNA9–11. Both radiation and alkylating agents deplete cells of NAD which is used to make (ADP-ribose)n (refs 1, 2, 12, 15–17), and this depletion is prevented by inhibitors of ADPRT13,15,18–25. ADPRT inhibitors also retard DNA strand-rejoining after damage1,2,26,27, and potentiate the lethality of DNA-damaging agents1,2,28. N-methyl-N′-nitro-N-nitrosoguanidine, an alkylating agent, transiently increases intracellular (ADP-ribose)n over 100-fold17. These data suggest that (ADP-ribose)n biosynthesis is necessary for effective repair of some types of DNA damage1. The question remains of which step in DNA repair requires ADPRT activity. The enzyme inhibitors do not prevent incision events1,2,27 nor do they inhibit repair replication29–31. Here we describe experiments which support the hypothesis that ADP-ribosylation activates DNA ligase activity. We suggest that ADPRT has a general function in the control of the breaking and rejoining of DNA.