The dark side of DNA repair: when genome maintenance runs out of fuel

Wait 5 sec.

In post-mitotic neurons, the high nucleotide demand of nucleotide excision repair (NER) becomes a problem, as limited neuronal deoxynucleoside triphosphate (dNTP) pools prevent the completion of repair synthesis, leading to persistent single-stranded DNA repair intermediates, that promote genome instability and neurotoxicity. This unexpected neuron-specific vulnerability reveals new opportunities to mitigate chemotherapy-induced neurotoxicity.As advances in cancer therapy have substantially improved patient survival across many cancer types, preserving long-term quality of life has become an increasingly important priority. However, chemotherapy-induced neurotoxicity, manifested predominantly as chemotherapy-induced peripheral neuropathy and cancer-related cognitive impairment, chemobrain, remains one of the most frequent and severe long-term complications experienced by cancer patients and survivors.1 Platinum-based chemotherapeutics, including cisplatin, carboplatin, and oxaliplatin, are amongst the most widely used anticancer agents. However, their therapeutic benefit comes at a cost, as up to 70% of patients develop neurotoxic symptoms ranging from peripheral neuropathy to cognitive impairment. These complications often persist for years after chemotherapy, and profoundly impair quality of life and can necessitate dose reduction or premature treatment discontinuation, ultimately compromising treatment efficacy. Despite the substantial burden of chemotherapy-induced neurotoxicity, no effective neuroprotective therapies are currently available, largely because the molecular basis for the exceptional vulnerability of post-mitotic neurons to chemotherapy-induced DNA damage remains poorly understood.1Platinum-based chemotherapeutics induce bulky DNA adducts, predominantly 1,2-intrastrand crosslinks and monoadducts, which distort the DNA helix and obstruct DNA replication and gene transcription. These lesions are primarily repaired by nucleotide excision repair (NER), a versatile DNA repair pathway consisting of two complementary lesion-recognition mechanisms.2 Global-genome NER (GG-NER) continuously surveys the entire genome for helix-distorting DNA lesions,3 whereas transcription-coupled NER (TC-NER) is initiated when elongating RNA polymerase II stalls at DNA damage, thereby specifically repairing the transcribed strand of expressed genes.4 Following lesion recognition, both pathways utilize a common repair mechanism that excises a ~24–32 nucleotide fragment containing the damage.5 The resulting single-stranded DNA gap must be filled by DNA synthesis and sealed by ligation, requiring one of the longest repair synthesis patches among the major DNA repair pathways and therefore rendering NER particularly demanding on cellular deoxynucleoside triphosphate (dNTP) pools for accurate repair.Interestingly, Nathan et al. reveal in their recent study in Cell6 that this high dNTP demand of NER becomes the pathway’s Achilles’ heel in post-mitotic neurons. To define the genomic landscape of cisplatin-induced DNA repair in neurons, the authors combined EdU-based detection of DNA repair synthesis with the newly developed single-cell synthesis-associated with repair sequencing (scSAR-seq) to map NER activity genome-wide. Because post-mitotic neurons no longer replicate their genome, maintaining genome-wide repair has generally been considered metabolically less advantageous than preserving transcription, leading to the long-standing view that differentiated neurons rely predominantly on TC-NER whereas GG-NER is largely attenuated following differentiation.4,7 Indeed, repair was initially concentrated within actively transcribed genes but unexpectedly expanded at later time points into lowly transcribed and intergenic regions, demonstrating that differentiated neurons retain substantial GG-NER activity. Genetic disruption of the TC-NER factor CSB and the GG-NER factor XPC confirmed that while TC-NER dominates the early repair response, GG-NER removes cisplatin lesions throughout the rest of the neuronal genome. Together, these findings reveal that in post-mitotic neurons, GG-NER is far more active than previously appreciated, raising the question of why such robust repair nevertheless fails to protect neurons.The answer proved remarkably counterintuitive. Rather than protecting neurons, NER-mediated repair of cisplatin lesions becomes neurotoxic. Genetic ablation of GG-NER strongly protected neurons from cisplatin-induced cell death, identifying GG-NER as the principal mediator of neurotoxicity, whereas TC-NER disruption yielded only modest protection. Mechanistically, the authors show that post-mitotic neurons lack sufficient dNTP pools to fuel the repair synthesis required for GG-NER. Quantitative metabolomics revealed that post-mitotic neurons contain substantially lower dNTP pools than proliferating cells, which decline further following cisplatin treatment, limiting the capacity of neurons to sustain the extensive repair synthesis required for NER and resulting in the accumulation of persistent single-stranded DNA repair intermediates.6 Consistent with this model, pharmacological depletion of the dNTP pools using hydroxyurea further sensitized neurons to cisplatin and increased DNA breaks, demonstrating that nucleotide shortage itself promotes the accumulation of toxic repair intermediates. These unresolved single-stranded gaps subsequently give rise to DNA double-strand breaks (DSBs) and trigger neuronal death (Fig. 1). Ironically, preferential protection of actively transcribed genes comes at the expense of repairing the rest of the neuronal genome, as the initial wave of TC-NER exhausts the nucleotide fuel required to complete GG-NER, resulting in neurotoxic DNA breaks.Fig. 1: NER drives cisplatin-induced neurotoxicity by exhausting neuronal dNTP pools.Full size imageFollowing cisplatin treatment, DNA lesions are preferentially repaired by TC-NER (green) in active genes, whereas lesions elsewhere in the genome are repaired with delayed kinetics by GG-NER (red). In cycling cells, abundant dNTP pools support efficient DNA repair synthesis, thereby promoting cell survival. However, in post-mitotic neurons, the limited dNTP pools are exhausted during the initial wave of TC-NER, leaving insufficient dNTPs to complete subsequent GG-NER repair synthesis. As a result, persistent single-stranded DNA repair intermediates accumulate and result in DNA DSBs, ultimately leading to chemotherapy-induced neurotoxicity. Created in BioRender. https://BioRender.com/rez2nsk.The discovery that chemotherapy-induced neurotoxicity arises from incomplete DNA repair caused by nucleotide shortage opens a promising opportunity for therapeutic intervention by refueling the neuronal dNTP pools. Capitalizing on this insight, the Nussenzweig laboratory demonstrates that both pharmacological deoxyribonucleoside supplementation and genetic enhancement of nucleotide biosynthesis restore repair synthesis, suppress break formation and improve neuronal survival following cisplatin treatment.6 Moreover, genetically increasing dNTP availability alleviates sensory deficits in mouse models of chemotherapy-induced peripheral neuropathy, highlighting nucleotide replenishment as a promising strategy to mitigate cisplatin-induced neurotoxicity. Importantly, these protective effects were selective for post-mitotic neurons, suggesting that replenishing dNTP pools could mitigate neurotoxicity without compromising the anti-tumor efficacy of chemotherapy in proliferating cells, although clinical translation remains to be demonstrated.The study also raises intriguing mechanistic questions. It is striking that the approximately 10-fold lower dNTP pools of post-mitotic neurons profoundly impair NER, given that repair synthesis requires filling only ~24–32 nucleotide gaps rather than replicating an entire genome. One possibility is that dNTP shortage slows gap filling, prolonging the lifetime of NER-generated gaps, thereby giving exonucleases such as EXO1 the opportunity to further resect the NER-intermediates into 100–1000 nucleotide gaps before they are sealed.8,9 These extended gaps increase repair-associated DNA synthesis and the probability that opposing gaps converge to subsequently form DSBs. Such a feed-forward mechanism would simultaneously amplify nucleotide consumption and genomic instability, potentially explaining how relatively modest reductions in dNTP availability could lead to neurotoxicity. Likewise, whether neuronal death is triggered primarily by persistent repair gaps or by the resulting DSBs remains to be determined. Resolving these questions will further clarify how metabolic constraints shape DNA repair outcomes in post-mitotic tissues. Beyond its therapeutic implications, this work could explains why GG-NER needs to be attenuated during neuronal differentiation. Prioritizing TC-NER may represent an evolutionary adaptation to conserve the limited dNTP reserves in neurons under physiological levels of DNA damage. However, under high damage loads during platinum chemotherapy, the remarkable efficiency of TC-NER rapidly exhausts these limited dNTP reserves, effectively causing GG-NER to run out of fuel with detrimental neurotoxic consequences.ReferencesWas, H. et al. Front. Pharmacol. 13, 750507 (2022).Article  PubMed  PubMed Central  Google Scholar Marteijn, J. A., Lans, H., Vermeulen, W. & Hoeijmakers, J. H. Nat. Rev. Mol. Cell Biol. 15, 465–481 (2014).Article  PubMed  Google Scholar Kusakabe, M. & Sugasawa, K. DNA Repair. 165, 103951 (2026).Article  PubMed  Google Scholar van Sluis, M. et al. Nat. Rev. Mol. Cell Biol. 27, 234–251 (2026).Article  PubMed  Google Scholar Hu, J., Adar, S., Selby, C. P., Lieb, J. D. & Sancar, A. Genes Dev. 29, 948–960 (2015).Article  PubMed  PubMed Central  Google Scholar Nathan, W. J. et al. Cell 189, 4005–4021.e11 (2026).Article  PubMed  PubMed Central  Google Scholar Nouspikel, T. & Hanawalt, P. C. Mol. Cell Biol. 20, 1562–1570 (2000).Article  PubMed  PubMed Central  Google Scholar Sertic, S. et al. Proc. Natl. Acad. Sci. USA 108, 13647–13652 (2011).Article  PubMed  PubMed Central  Google Scholar Giannattasio, M. et al. Mol. Cell 40, 50–62 (2010).Article  PubMed  Google Scholar Download referencesAuthor informationAuthors and AffiliationsDepartment of Molecular Genetics, Oncode Institute, Erasmus MC Cancer Institute, Erasmus University Medical Center, Rotterdam, The NetherlandsMarjolein van Sluis & Jurgen A. MarteijnAuthorsMarjolein van SluisView author publicationsSearch author on:PubMed Google ScholarJurgen A. MarteijnView author publicationsSearch author on:PubMed Google ScholarCorresponding authorCorrespondence to Jurgen A. Marteijn.Additional informationPublisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.Rights and permissionsReprints and permissionsAbout this article