Uracil occurs at replication forks via misincorporation of deoxyuridine monophosphate (dUMP) or via deamination of existing cytosines, which occurs 2-3 orders of magnitude faster in ssDNA than in dsDNA and is 100% miscoding. Tethering of UNG2 to proliferating cell nuclear antigen (PCNA) allows rapid post-replicative removal of misincorporated uracil, but potential 'pre-replicative' removal of deaminated cytosines in ssDNA has been questioned since this could mediate mutagenic translesion synthesis and induction of double-strand breaks. Here, we demonstrate that uracil-DNA glycosylase (UNG), but not SMUG1 efficiently excises uracil from replication protein A (RPA)-coated ssDNA and that this depends on functional interaction between the flexible winged-helix (WH) domain of RPA2 and the N-terminal RPA-binding helix in UNG. This functional interaction is promoted by mono-ubiquitination and diminished by cell-cycle regulated phosphorylations on UNG. Six other human proteins bind the RPA2-WH domain, all of which are involved in DNA repair and replication fork remodelling. Based on this and the recent discovery of the AP site crosslinking protein HMCES, we propose an integrated model in which templated repair of uracil and potentially other mutagenic base lesions in ssDNA at the replication fork, is orchestrated by RPA. The UNG:RPA2-WH interaction may also play a role in adaptive immunity by promoting efficient excision of AID-induced uracils in transcribed immunoglobulin loci.
we demonstrate that uracil-DNA glycosylase (UNG), but not SMUG1 efficiently excises uracil from replication protein A (RPA)-coated ssDNA
matched biochemical assays support excision, with the SMUG1 contrast restricted to one substrate
the UNG2 N-terminal helix (starting at residue 66) and the RPA2-WH domain are both necessary for uracil excision from RPA-coated ssDNA.
domain deletions, helix mutants, and competition converge on the interaction requirement within the defined biochemical system
phosphorylation at T60 mediated two-fold reduction in RPA binding whereas double phosphorylation at T60 and S64 resulted in a further decrease.
peptide binding measurements and affinity capture support reduced interaction without directly establishing full-length cellular regulation
ubiquitination of the UNG RPA-binding helix does not block RPA binding but rather modestly stimulates the capability of UNG2 to excise uracil from RPA-coated ssDNA.
parallel activity comparisons show modest stimulation in truncated enzymes modified under E3-independent conditions
UNG has a unique capability among the human UDGs to excise uracil from RPA-bound ssDNA
comparison with SMUG1 does not establish uniqueness across all human uracil-DNA glycosylases
Derived from the full evaluation — not a separate score.
Strengths
RPA-WH deletion, UNG truncations, helix point mutations, and free-domain competition converge on the same molecular dependence. Parallel naked-DNA assays help distinguish impaired access from loss of catalytic function.
↳ Results, Figures 2–3 and Supplementary Figure S9
The assays place uracil within RPA-coated ssDNA rather than restricting it to the duplex portion of a junction substrate. This directly addresses the access question left open by the closest prior work.
↳ Introduction; Results, Figure 2
The Discussion identifies unresolved structural access, exchange of WH-binding factors, and downstream AP-site processing. These uncertainties are reflected in the conditional framing of the replication-fork model.
↳ Discussion, opening paragraph and downstream-processing model
Limitations
The Methods describe the RPA-blocking titration as data not shown and contain no statistics section despite significance language in Results. Exact replication counts and statistical outputs cannot be established from the supplied text alone.
↳ Methods, Uracil excision assays; Results, Figures 3F and 5I–J
Phosphorylation evidence relies principally on peptides, while functional ubiquitination experiments use N-terminally truncated proteins modified without an E3 ligase. These experiments support biochemical modulation but do not directly establish its operation in cells.
↳ Results, Figures 4–5; Methods, In vitro ubiquitination
The uniqueness statement extends from a SMUG1 comparison to all human uracil-DNA glycosylases. The Discussion also attributes a 1000-fold contrast involving deletion of the entire UNG N-terminal domain specifically to the WH interaction, a narrower attribution than that comparison alone identifies.
↳ Results, domain-dependence section, final paragraph; Discussion, paragraph 3
The core biochemical conclusion rests on matched substrate assays and convergent perturbations of both interaction partners in Figures 2–3. Phosphorylation and ubiquitination experiments extend the mechanism, but peptide-based binding measurements and E3-independent modification of truncated proteins limit physiological interpretation. The Discussion engages prior work and explicitly treats coordinated replication-fork repair as a model rather than a demonstrated cellular pathway. Generalizability remains limited by purified proteins and short oligonucleotide substrates, while the unpresented RPA-excess calibration creates a transparency note rather than a demonstrated numerical error.
Nabu’s assessment, alongside the field’s view.
Are you an author of this paper?
Strong3.8
Confidence highThe study extends prior UNG–RPA interaction work by testing excision within RPA-coated ssDNA and identifying dependence on the RPA2-WH domain and UNG helix. PTM regulation adds mechanistic detail, but its physiological consequences and the broader cellular pathway remain unresolved.
Parallel matched assays and independent domain, truncation, point-mutant, and competition experiments support the central biochemical inference. Narrow substrate coverage, E3-independent ubiquitination of truncated enzymes, and incomplete access to quantitative reporting limit assessment beyond that inference.
The argument separates experimentally tested biochemical dependencies from the proposed replication-fork model. The Abstract's PTM wording and the Results statement about uniqueness among human UDGs are broader than the most directly tested evidence.
The paper distinguishes its contribution from junction-substrate work and engages conflicting interaction findings and binding-affinity estimates. Mechanistic unknowns constrain the proposed model, although assay stoichiometry and substrate generalisability receive less attention.
Lower confidence on Contribution, Methodological Rigour — domain match limited.
Caveats4 of 4 checks
The reported numbers and statistics are consistent with the stated results.
The calibration supporting the RPA excess used in coated-substrate assays is explicitly described as data not shown. This limits transparency without establishing that the biochemical conclusion is incorrect.
103 references in manuscript 103 of 103 checkable references found in an index 1 reference confirmed by manual review
References not verified automatically are resolved manually within 24 hours.
No retraction or expression of concern found in Crossref (includes Retraction Watch).
Sources: Crossref (incl. Retraction Watch) ✓
Where this paper’s evidence sits on the path from initial observation to real-world use.
The evidence characterises a molecular mechanism under controlled conditions rather than an intervention or diagnostic application. Cellular and disease-relevant validation would be required before an actionable pathway emerges.
AI-generated, human-governed. Something look off? Contact us to request a review.
Keep reading
Other papers Nabu has evaluated blind, starting with the same field.
Ellis, Mills, Niedzwiedz, et al. · 2026 · Clinical medicine
Loss of IGF-1R impaired DNA-dependent protein kinase catalytic subunit (DNA-PKcs) localisation to chromatin, resulting in defective non-homologous end-joining (NHEJ) and a compensatory reliance on alternative repair pathways, including microhomology-mediated end-joining (MMEJ).
SupportedKolas, Chapman, Nakada, et al. · 2007 · Basic medicine
Iveland, Hagen, Sharma, et al. · 2020 · Basic medicine
UNG2 degradation appeared to be ubiquitous and was observed across several mammalian cell lines of different origin and with several HDACis.
Supported