DNA double-strand breaks are among the most lethal forms of genetic damage, arising from replication stress, ionizing radiation, or chemotherapeutic drugs. Cells prefer to mend them through homologous recombination, an error-free mechanism that operates during the S and G2 phases of the cell cycle, when a sister chromatid is available as a template. In this pathway, the MRN complex recruits repair machinery to the broken ends, nucleases such as EXO1 and DNA2 chew back the DNA to create single-stranded overhangs, and the RPA complex coats and stabilizes those exposed strands before BRCA2 swaps in RAD51 to search for homology. When this process goes awry, the consequences include genomic instability, tumor progression, and therapeutic resistance.

RPA1 is the largest subunit of the heterotrimeric RPA complex, and it is the subunit that directly grips single-stranded DNA through its central DNA-binding domains while its N-terminal region mediates contacts with other proteins. Loss or mutation of RPA1 impairs double-strand break repair and has been linked to genomic instability and to cancers such as breast and ovarian carcinoma. The protein is already known to be regulated by phosphorylation, SUMOylation, and ubiquitination, but the researchers suspected another modification might be at play: lysine lactylation, a chemically distinct mark derived from lactate that has recently been found on non-histone proteins as well as histones.

The team first showed that RPA1 is indeed lactylated. In HEK293T cells, treatment with sodium lactate increased RPA1 lactylation without changing its acetylation, while sodium oxamate, an inhibitor of lactate dehydrogenase, had the opposite effect. Depleting the lactate dehydrogenase isoforms LDHA and LDHB also reduced the modification, and exposure to several DNA-damaging agents, including ionizing radiation, cisplatin, and etoposide, raised RPA1 lactylation levels. A screen of candidate enzymes pointed to KAT8, a histone acetyltransferase better known for its role in chromatin remodeling. A KAT8 inhibitor called MG149 and KAT8 knockdown both sharply reduced RPA1 lactylation, and purified KAT8 catalyzed lactylation of RPA1 in vitro without acetylating it. Co-immunoprecipitation and a GST pull-down assay with purified proteins confirmed that KAT8 and RPA1 bind each other directly.

Mass spectrometry identified four candidate lactylation sites on RPA1, at lysines 88, 163, 167, and 267. When each was individually mutated to arginine, only the K167R mutant markedly reduced overall lactylation, identifying K167 as the major site. The researchers then generated an antibody specific to lactylated K167 and used it to confirm that cisplatin-induced damage and KAT8 overexpression increased the modification on wild-type RPA1 but not on the K167R mutant. They also found that SIRT6, a sirtuin deacetylase, appears to be the enzyme that removes the mark, since only SIRT6 overexpression among the sirtuin isoforms tested reduced RPA1 lactylation.

Functionally, the K167 mark proved to be a boost for RPA1’s core job. The K167R mutant formed fewer foci at damage sites, associated less with chromatin, and bound DNA more weakly in an electrophoretic mobility shift assay using purified proteins. In laser micro-irradiation experiments in U2OS cells, wild-type RPA1 rapidly accumulated at damage stripes while the K167R mutant lagged behind. The mutant also reduced homologous recombination efficiency, and sodium lactate could no longer enhance repair in its presence, while KAT8 inhibition no longer suppressed it further. Chromatin fractionation after cisplatin treatment showed that cells expressing the mutant had reduced RPA1 loading and impaired RAD51 recruitment, suggesting the modification helps the handoff from RPA to RAD51 that is a hallmark of productive recombination.

How does the cell switch this system on when DNA breaks occur? The answer lies in ATM, a master kinase of the DNA damage response. Cisplatin enhanced the interaction between KAT8 and RPA1, and an ATM inhibitor reversed it, while inhibitors of the related kinases DNA-PKcs and ATR did not. ATM phosphorylated KAT8 on serine 392, a site within the SQ/TQ motif consensus that ATM and ATR recognize, and mutating that serine to alanine abolished the damage-induced phosphorylation signal in an in vitro kinase assay. The KAT8 S392A mutant bound RPA1 poorly, failed to support RPA1-K167 lactylation foci after cisplatin, and reduced homologous recombination efficiency even when lactate was added. The authors describe this as an ATM-KAT8-RPA1 axis in which damage-activated ATM licenses KAT8 to lactylate RPA1.

The therapeutic implications emerged in breast cancer models. RPA1 lactylation levels were elevated in several breast cancer cell lines, particularly Hs578T and HCC1806, compared with the non-tumor MCF 10A control. Inhibiting lactate dehydrogenase or KAT8 sensitized Hs578T, HCC1806, and SUM-159 cells to olaparib, a clinically approved PARP inhibitor, while sodium lactate made SUM-159 cells more resistant. Critically, cells expressing the K167R mutant were more olaparib-sensitive than wild-type cells, and the inhibitors no longer added any sensitization in the mutant background, indicating that the drugs work through this specific modification. In a breast cancer patient-derived xenograft model confirmed to be homologous recombination proficient, combining either inhibitor with olaparib reduced tumor weight and volume, lowered markers of RPA engagement and proliferation, and increased gamma-H2AX, a marker of unrepaired DNA damage.

Going further, the team designed a cell-penetrating peptide, K167-pe, modeled on the sequence around the K167 site, intended to compete with RPA1 for lactylation. Among six candidate peptides, K167-pe was the most potent inhibitor of RPA1 lactylation, acting in a dose-dependent manner, and a scrambled control peptide with the lysine replaced by arginine had no effect. K167-pe reduced the modification after cisplatin-induced damage, suppressed homologous recombination, and increased the sensitivity of MDA-MB-231 and Hs578T breast cancer cells to olaparib and cisplatin, while the control peptide did none of these things. In the PDX model, K167-pe significantly increased tumor sensitivity to olaparib. The authors note that the effects of the inhibitors and the peptide were more pronounced in the PDX models than in cultured cells, possibly because the xenografts preserve tumor microenvironment features and heterogeneity absent from cell culture, though they caution that this remains to be investigated.

The study has acknowledged limits. The structural basis of the interaction between phosphorylated KAT8 and RPA1 is unknown, and the efficacy and safety of targeting RPA1 lactylation require testing in additional preclinical models and tumor types. The work also builds on a recent report that HAT1, another acetyltransferase, mediates the same K167 lactylation in lung adenocarcinoma and promotes radioresistance; the present study adds the ATM-dependent phosphorylation step, identifies KAT8 as the enzyme in this context, and demonstrates a peptide therapeutic in breast cancer PDX models. Together, the findings suggest that in tumors with intact homologous recombination, the Warburg effect is not merely a fuel supply but a direct chemical signal that arms the DNA repair machinery, and that selectively blocking that signal, rather than disabling RPA entirely, may offer a way to resensitize tumors that currently resist DNA-damaging therapy.

Subject of Research: Lactate-dependent lactylation of RPA1 by KAT8 in homologous recombination repair and breast cancer chemoresistance

Article Title: Metabolic regulation of ATM-KAT8-RPA1 axis in DNA damage repair and chemoresistance

Article References: Ding, X., Yang, J., Wu, J., Jin, M., Wang, Z., Yang, X., Wang, X., Song, Y., Huang, B., Li, R., Li, Y., He, Y., Zheng, W., Wang, Q., Zeng, Z., Chen, Y., & Yuan, J. (2026). Metabolic regulation of ATM-KAT8-RPA1 axis in DNA damage repair and chemoresistance. iScience, 29(11), Article 117908. https://doi.org/10.1016/j.isci.2026.117908

Image Credits: AI Generated

DOI: 10.1016/j.isci.2026.117908

Keywords: RPA1, KAT8, lysine lactylation, homologous recombination, ATM kinase, DNA double-strand break repair, chemoresistance, olaparib, breast cancer, patient-derived xenograft, lactate dehydrogenase, cell-penetrating peptide