More than just a passive brick in the wall: the nucleosome facilitates DNA polymerase β activity in linker DNA and its PARP-dependent regulation in the BER pathway choice

Reviewer #1 (Public review):

Summary:

One of the most important fundamental questions in base excision repair (BER) is how chromatin structure affects the action of specific components of the BER pathway. Previous work from this and other groups has began to address this question. In this report, the authors study the activity of Pol beta on a gapped or nicked DNA substrate 23 bases from the entry/exit site of a 603 nucleosome core particle in the presence and absence of PARP1, PARP2, HPF1, or FEN1. They show that H1 and PARP block pol beta incorporation, which is relieved by NAD+.

Strengths:

They show, not unexpectedly, that HPF1 and PARP activity help to displace H1, allowing Pol beta incorporation. PARP1 and PARP2 suppress Pol beta activity, which is mitigated by autoparylation. PARP2 has a strong impact on strand displacement synthesis. This is an important contribution to the field.

Weaknesses:

This present work incrementally builds upon their previous work, and what has been known previously about the activity of PARP1/2, HPF1, and the modification of histones.

Reviewer #2 (Public review):

Summary:

The authors have shown some interesting data on DNA repair synthesis by PolB, acting on a BER substrate in the presence of a core nucleosome, and the effects of some accessory chromatin proteins. FEN1 and PARP proteins were also assessed for their effects on repair synthesis by PolB. However, the story for the PARP proteins seems a bit underdeveloped, or perhaps it just needs additional clarity in the writing. The concept that strand displacement synthesis by PolB in linker DNA and into the NCP is limited by these interactions is useful, although we need to bear in mind that the study does not address the role of the final repair enzyme, DNA ligase, which might itself limit the products. Likewise, the possible effects of competing DNA polymerases remain unexplored, notably the replication enzymes delta and epsilon. There are circumstances where these appear to be the main DNA repair polymerases for BER substrates. Addressing these and other issues, as listed below, would greatly improve a paper that is already fairly strong.

Specific Points:

(1) Substrates:

The gap substrate was prepared by treating a U-containing substrate with UDG + APE1. Consequently, it is not exactly a gap, but a repair intermediate with a 5-abasic site on one side of the break. It should be described more clearly in the text.

The nicked substrate was prepared by incubating the "gap" substrate with PolB and dTTP, the nucleotide to replace the excised U. It is expected that this substrate has the 5'-abasic site removed by the PolB lyase, and only one dTMP residue inserted. Has either of these expectations been verified? For example, PolB can insert more than one nucleotide in a prolonged incubation, and the enzyme has no intrinsic 3'-exonuclease to trim the extension.

Finally, it appears that these procedures were performed with the NCP already in place; therefore, the presence of the nucleosome is expected to influence the processing done to prepare the gap and nick substrates. What do we know about that?

(2) Figure 1c:

The rate difference for gap vs. NCP is modest, perhaps 2-fold in the data shown. Some statistical analysis is needed to solidify this observation.

(3) As noted on page 4, the histone tails might be important for some of the observed effects. While individual histones had no effect, the critical test would be in the context of the NCP. There are many modified or mutant histones now available that would enable this. While such experiments would be more for future work, the possibility should be mentioned in this paper.

(4) What are the molar ratios of the various enzymes to the substrates? Can we say whether that reflects the levels that might be found in vivo? For the in vitro studies, the stoichiometry would also influence competing binding reactions. Indeed, Figure 2 indicates that the NCP substrate has multiple, competing binding sites for PolB. Why are the multiple NCP-PolB species not better resolved in EMSA (Supplementary Figure 2a)? Perhaps the higher-order ones are more unstable in the gel? That would be consistent with Table 1.

(5) Wouldn't the incremental 3-nucleotide steps seen with PolB + FEN1 be a relatively inefficient process? Of course, one expects that the presence of a DNA ligase would effectively limit this process to just one synthesis/excision cycle. Hasn't that been tested with these substrates?

(6) In many of the gel images, it can be hard to tell S from the +1 products, especially further from the side of the gel. Is there an independent way to verify that just a single nucleotide was replaced?

Reviewer #3 (Public review):

This manuscript by Shtanov et al. attempts to define how DNA Polymerase β performs gap-filling DNA synthesis and strand displacement synthesis in linker DNA adjacent to a nucleosome. The authors show that DNA Polymerase β strand displacement synthesis activity is stimulated in linker DNA when the 1-nt gap is positioned 23 bp away from a nucleosome core particle. The authors further show that histone H1, known to bind linker DNA, disrupts the ability of DNA Polymerase β to perform strand displacement synthesis within this context. They then provide some evidence that PARP1 and PARP2 regulate DNA Polymerase β strand displacement synthesis in linker DNA adjacent to a nucleosome, possibly pointing to a role for PARP1 and PARP2 in base excision repair sub-pathway choice. While this study has some intriguing observations, these observations are severely underdeveloped, and many of the stated conclusions are inadequately justified by the experimental data.

Strengths:

(1) The authors have identified that DNA Polymerase β strand displacement synthesis is stimulated in linker DNA by the presence of an adjacent nucleosome, though the generalizability of this finding is unclear (see weaknesses).

(2) The authors convincingly show that the presence of histone H1 negatively regulates DNA Polymerase β strand displacement synthesis in linker DNA adjacent to a nucleosome core particle.

Weaknesses:

(1) Throughout the manuscript, the authors perform a variety of enzyme kinetic assays to show that DNA Polymerase β strand displacement synthesis is stimulated in linker DNA by the presence of an adjacent nucleosome, and that other chromatin factors (PARP1, PARP2, and histone H1) regulate strand displacement synthesis. The enzyme kinetic experiments presented have several issues that severely impact their interpretability. This includes the lack of proper substrate controls, a general lack of quantification and statistical analysis, the use of varied enzyme kinetics regimes that impede comparison between experiments, and a general lack of clarity regarding experimental replication/reproducibility.

(2) The general context where an adjacent nucleosome core particle would stimulate DNA Polymerase β strand displacement synthesis is severely underdeveloped, which limits the generalizability of these findings. It's unclear if this stimulation is dependent on the linker DNA length, the distance of the 1-nt gap from the nucleosome core particle, or the directionality of strand displacement synthesis (towards vs away from the nucleosome core particle). Given the data presented, it's possible that stimulation of DNA Polymerase β strand displacement synthesis by an adjacent nucleosome is a phenomenon that is unique to a 1-nt gap precisely 23 nts away from the nucleosome core particle.

(3) The conclusion that the N-terminal histone tails do not stimulate DNA Polymerase β strand displacement synthesis comes from an experiment where Gap-DNA227 was incubated with free core histones, and a reduction in strand displacement synthesis was observed. As designed, this experiment is simply unable to prove that the N-terminal tails do not stimulate DNA Polymerase β strand displacement synthesis.

(4) The observation of apparent cooperativity in DNA Polymerase β binding to Gap-NCP227 from the mass photometry data is intriguing. However, the relationship between this observation and the stimulation of DNA Polymerase β strand displacement synthesis in linker DNA adjacent to a nucleosome core particle is unclear.

(5) The general claims regarding differential specificity of PARP1 and PARP2 for nicks and gaps in linker DNA adjacent to the nucleosome come from experiments lacking a proper control using an undamaged linker-nucleosome substrate. This is particularly problematic as PARP1 and PARP2 are known to engage the terminal ends of DNA as they partially mimic DNA double-strand breaks.

(6) While the authors clearly show that PARP1 and PARP2 regulate DNA Polymerase β strand displacement synthesis in linker DNA, the interpretation that this is through direct competition for 1-nt gap binding cannot be proven from the experiments presented.

(7) The claim that the presence of histone H1 changes the yield and length of PARylated core histones is overstated. The quantification would suggest a subtle difference (particularly for PARP1), but the lack of statistical analysis related to the experiments makes interpretation challenging.