Acetylation blocks DNA damage-induced chromatin ADP-ribosylation

Abstract

Recent studies report serine ADP-ribosylation on nucleosomes during the DNA damage response. We unveil histone H3 serine 10 as the primary acceptor residue for chromatin ADP-ribosylation and find that specific histone acetylation marks block this activity. Our results provide a molecular explanation for the well-documented phenomenon of rapid deacetylation at DNA damage sites and support the combinatorial application of PARP and HDAC inhibitors for the treatment of PARP-dependent cancers.

Figure 1.. DNA damage-induced H3S10 ADP-ribosylation is blocked by H3K9 acetylation.

a, ADP-ribosylation of recombinant mononucleosomes by recombinant PARP1 in the absence and presence of HPF1 (cofactor = unlabeled NAD⁺). b, Recombinant PARP1/HPF1 ADP-ribosylation activity on recombinant mononucleosomes substrates wherein known H3 Ser-ADPr sites were mutated as indicated (cofactor = unlabeled NAD⁺). For biotin-NAD⁺ cofactor-based assays and product quantification see Supplementary Fig. 3. c, Binding efficiency and ADP-ribosylation activity (relative to wild type nucleosomes) for each member of the mononucleosome library in PARP1/HPF1 (left) and PARP2/HPF1 (right) assays. The red circle indicates all significant (two-tailed P-value ≤ 0.01; n = 3 from three independent assays) hits that exhibit a > 2-fold effect on activity. d, Quantification of recombinant PARP1/HPF1 ADP-ribosylation activity on recombinant, mutated/modified mononucleosome substrates as indicated for each lane. Replicate western blots used for quantification can be found in Supplementary Fig. 9 and 26. Center bar represents the mean and error bars represent standard error of the mean (n = 3 from three independent assays). e, ADP-ribosylation analysis of H3-Flag constructs from H₂O₂-treated U2OS cells. f, Recombinant PARP1/HPF1 ADP-ribosylation assays on the indicated recombinant, modified/mutated mononucleosomes substrates in the presence of the recombinant SIRT6 deacetylase (cofactor = unlabeled NAD⁺). Both ADP-ribosylation and H3K9 deacetylation activities are shown. Uncropped blots and replicates can be found in Supplementary Fig. 26. Unlabeled NAD⁺ cofactor-based assays and live cell assays were repeated once to ensure reproducibility.

Figure 2.. Acetylation of PARP1 regulates its auto- and trans-ADP-ribosylation activities.

a, Recombinant PARP1 ADP-ribosylation assays comprising wild type or 7xKQ PARP1 constructs, and different stimulatory factors (HPF1, free DNA, mononucleosomes) as indicated above each lane (cofactor = unlabeled NAD⁺). b, Immunoblot analysis of wild type and 7xKQ PARP1 construct auto- and trans-ADP-ribosylation activities in a PARP1 knockout (KO) U2OS cell line. Note: the H₂O₂-dependent ADP-ribosylation observed in the PARP1-knockout cell line represents PARP2 activity. c, Model wherein acetylation of H3K9 (substrate) or the PARP1 AM domain (enzyme) blocks ADP-ribosylation of the H3S10 residue. Uncropped blots can be found in Supplementary Fig. 26. Unlabeled NAD⁺ cofactor-based assays and live cell assays were repeated once to ensure reproducibility.