Abstract

Mislocalization and aggregation of the DNA/RNA binding protein, TDP-43, is seen in most cases of amyotrophic lateral sclerosis-frontotemporal dementia (ALS-FTD). Accumulating DNA damage in neurons is also a common feature of ALS-FTD. TDP-43 has several characterized roles in the regulation of the DNA damage response (DDR). This review systematically explored the relationship between TDP-43, DNA damage and the DNA damage response in various models of ALS-FTD, facilitating comparison of findings between studies using similar models. Twelve peer-reviewed papers, covering eight TDP-43 mutations out of nearly 40, were reviewed and five experimental models included: cell lines, patient-derived iPS cells, organoids, and rodent models, plus post-mortem cortex and spinal cord tissue from ALS-FTD patients. Across the studies and models, depletion of TDP-43 or ALS-linked mutations consistently increased genomic instability. Q331K-expressing cells showed a 2-3-fold reduction in DNA repair activity and a 4-6-fold increase in DDR activation, while TDP-43-depleted cells showed a 20-fold rise in double strand breaks. TDP-43 normally binds to damaged chromatin, participates in early DDR signaling and scaffolds core DNA damage repair factors, including Ku70, XRCC4 and DNA ligase 4. This systematic review and narrative synthesis sheds light on mechanisms that explain how TDP-43 dysfunction impairs genome maintenance. When TDP-43 is mislocalized, mutated or aggregated, these interactions are disrupted, resulting in impaired DNA repair. DNA damage is also caused by increasing R-loops, dysregulation of mismatch repair gene transcription, and sequestering of repair proteins into cytoplasmic inclusions. Upstream DNA damage can further drive TDP-43 mislocalisation, creating a feed-forward loop. Given the ubiquity of TDP-43 pathology across neurodegenerative diseases, targeting the DDR mechanisms affected by TDP-43 may offer new therapeutic opportunities.

1 Introduction

Amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) form a spectrum of neurodegenerative diseases with overlapping clinical and molecular characteristics (ALS-FTD). ALS is marked by the gradual deterioration of motor neurons resulting in muscle weakness. ALS classically presents first with weakness of the lower motor neurons with later degeneration of the upper motor neurons that leads eventually to respiratory failure. ALS is an aggressive form of neurodegeneration with an average survival time of 3 to 5 years following diagnosis (Eshima et al., 2023; Tzeplaeff et al., 2023). Frontotemporal dementia is characterized by changes in behavior and difficulties with language, stemming from the atrophy of the frontal and temporal lobes (Neumann et al., 2009; Tzeplaeff et al., 2023). Neuropathologically, TDP-43 cytoplasmic aggregates are observed in 90% of ALS cases and 50% of FTD cases (Jo et al., 2020), highlighting TDP-43 proteinopathy as a characteristic of ALS-FTD spectrum disorders.

TDP-43, encoded by the TARDBP gene, is a predominantly nuclear protein that plays crucial roles in RNA processing, including splicing regulation, mRNA stability maintenance, and transport (Koike, 2024). The protein contains an N-terminal dimerization domain that also contains the nuclear localization sequence (NLS), with two central RNA recognition motifs essential for the role of TDP-43 as a splicing regulator. The unstructured C-terminal mediates binding with other proteins and is considered central to the mislocalization and aggregation of TDP-43 to the cytoplasm, which compromises its functional capacity, resulting in genomic instability and heightened vulnerability to neurodegenerative processes (Dang, et al., 2025).

Neurons are thought to be more vulnerable to DNA damage due to their high metabolic activity and their limited capacity to repair such damage, coupled with their longevity. Post-mitotic neurons experience primarily three forms of DNA damage: (a) oxidative lesions, which result from reactive oxygen or nitrogen species and lead to base modifications and single-strand breaks (SSB); (b) double-strand breaks (DSB), considered to be the most harmful type of DNA damage and which occur due to failed SSB repair or oxidative stress; and (c) the activation of transposable elements, linked to the loss of TDP-43 function in specific ALS subtypes (Martin, 2008; Chatterjee and Walker, 2017; Eshima et al., 2023). In post-mitotic cells, homologous recombination (HR) is inactive; consequently, these cells rely on error-prone non-homologous end-joining (NHEJ) and base excision repair (BER) to address DSB and oxidative damage, respectively (Martin, 2008; Chatterjee and Walker, 2017). TDP-43 also appears to regulate the splicing and transcription of two genes involved in the mismatch repair pathway (MMR): MLH1 and MSH6. This suggests a broader role for TDP-43 in the DNA damage response (DDR) that extends beyond conventional pathways (Provasek et al., 2025).

To date, three pharmaceutical treatments for ALS have received FDA approval: Riluzole, Edaravone, and Sodium phenylbutyrate/Taurursodiol. These treatments primarily function to slow disease progression and can slightly extend patient survival (Tzeplaeff et al., 2023). In contrast, FTD currently lacks any approved treatment options. Certain genetic therapies have been designed to address particular familial ALS conditions. For instance, the antisense oligonucleotide, Tofersen, targets SOD1 mRNA, triggering its degradation and consequently reducing production of the toxic SOD1 protein (Tzeplaeff et al., 2023). However, SOD1-associated ALS is only a tiny fraction of ALS cases, which leaves most patients with no effective treatments. An alternative therapeutic approach that might be more generalizable for ALS-FTD patients is to target the genomic instability and downstream effects associated with TDP-43. To aid understanding of how TDP-43 dysfunction leads to genomic instability and to highlight potential therapeutic targets, this review systematically explores the experimental findings from both in vitro and in vivo pre-clinical research to evaluate the involvement of TDP-43 in DNA damage and repair mechanisms in models of ALS-FTD.

This review follows the same methodology as two similar reviews that focused on connections between DNA damage, the DNA damage response and two other genes associated with familial ALS-FTD that have been studied using a wide variety of pre-clinical models: C9orf72 and FUS (Almalki et al., 2025a; Almalki et al., 2025b). When viewed together, the conclusions reached provide insight into common pathophysiological mechanisms at play across a range of disorders of the ALS-FTD spectrum.

2 Methods2.1 Search strategy

This systematic review and narrative synthesis forms part of a larger review of the association of ALS-FTD and DNA damage in the nervous system that has been written as three related articles. The methodology used for the review is described in detail in the first article of the three that focuses on ALS-FTD associated with C9orf72 expansions (Almalki et al., 2025a). In brief, our literature search used the Boolean terms ‘amyotrophic lateral sclerosis’ OR ‘ALS’ AND ‘DNA’ AND ‘double strand breaks’ across PubMed, EMBASE and Web of Science databases for peer-reviewed primary research articles, in English and published to February 2025. This search yielded 91 publications after removing of duplicates. Of these, 41 articles were retained for full-text screening and only the 12 (Provasek et al., 2025; Gianini et al., 2020; Guerrero et al., 2019; Fang et al., 2023; Mitra et al., 2019; Nogami et al., 2022; Mitra et al., 2025; Lee and Rio, 2024; Konopka et al., 2020; Tamaki et al., 2023; Marques et al., 2024; Pal et al., 2021) that specifically focused on TDP-43 proteinopathy or dysfunction and its association with DNA damage, the DNA damage response (DDR), and DNA repair pathways were included for analysis (Figure 1).

The PRISMA flow diagram indicating inclusion and exclusion criteria used in the systematic review.

3 Results3.1 Basic information of included studies

The study selection and characteristics are detailed in the accompanying systematic review (Almalki et al., 2025a).

3.2 Risk of bias assessment results

Each of the 12 studies included in vitro results. To assess Risk of Bias (RoB) in these studies we used the Office of Health Assessment and Translation (Ohat, 2019; Figure 2). The tool covers seven domains. Overall, 9 out of 12 studies (Provasek et al., 2025; Fang et al., 2023; Gianini et al., 2020; Guerrero et al., 2019; Marques et al, 2024; Mitra et al., 2025; Lee and Rio, 2024; Konopka et al., 2020; Tamaki et al., 2023) were classified as Tier 1, indicating a low RoB. In contrast, three studies (Nogami et al, 2022; Mitra et al., 2019; Pal et al, 2021) were identified as having moderate RoB and were classified as Tier 2. This suggests generally strong methodological rigor across the studies. Domains 5 (incomplete outcome data) consistently showed high confidence across all studies, with nearly all cases rated as “++.”

The OHAT tool for rating the risk of bias in in vitro studies. (A) Risk of bias assessment across included studies. (B) Risk of bias domains and classification criteria.

Only one study presented vertebrate in vivo results (Mitra et al, 2025). For this study we employed the SYRCLE tool (Hooijmans et al, 2014) to assess RoB (Figure 3). Although this demonstrates low RoB in six critical domains, including the blinding of participants and outcome assessment, it identifies four domains with high RoB: baseline characteristics, random housing, allocation concealment and specification of primary outcome data.

SYRCLE risk of bias assessment of the solo in vivo study.

In summary, while the in vitro studies exhibited overall low risk of bias, the in vivo study showed a high risk of bias in some domains.

3.3 Results of studies

The systematic review and narrative synthesis identified 12 publications that investigated the role of DNA damage, the DDR, and DNA repair mechanisms in TDP-43 proteinopathy in both neuronal and non-neuronal models. Five experimental systems were used to establish TDP-43 models, including immortalized cell lines of either human or rodent origin (Fang et al., 2023; Gianini et al., 2020; Guerrero et al., 2019; Konopka et al., 2020; Lee and Rio, 2024; Mitra et al., 2019; Mitra et al., 2025; Nogami et al., 2022; Provasek et al., 2025), organoids (Tamaki et al., 2023), induced pluripotent stem cells (iPSCs) derived from ALS cases (Fang et al., 2023; Gianini et al., 2020; Konopka et al., 2020; Marques et al., 2024; Mitra et al., 2019; Pal et al., 2021; Provasek et al., 2025), rodent tissues (Konopka et al., 2020; Mitra et al., 2025; Provasek et al., 2025), and postmortem tissues from ALS patients (Fang et al., 2023; Guerrero et al., 2019; Marques et al., 2024; Mitra et al., 2019; Provasek et al., 2025). The 12 studies were published from 2019 to 2025. Seven studies were carried out in the USA (Fang et al., 2023; Guerrero et al., 2019; Lee and Rio, 2024; Marques et al., 2024; Mitra et al., 2019; Mitra et al., 2025; Provasek et al., 2025) and the remaining five were in Italy/Spain (Gianini et al, 2020), Japan (Nogami et al, 2022), Australia (Konopka et al., 2020), Canada (Tamaki et al., 2023), and Germany (Pal et al., 2021).

ALS-FTD patient-associated point mutations in TDP-43 were investigated in eight studies via either overexpression or via point mutations engineered into the TARDBP gene (Fang et al., 2023, Gianini et al., 2020, Guerrero et al., 2019, Konopka et al., 2020, Lee and Rio, 2024, Marques et al., 2024, Mitra et al., 2025, Provasek et al., 2025, Pal et al., 2021). Three publications explored the role of TARDBP depletion or knockdown (Gianini et al., 2020, Marques et al., 2024, Mitra et al., 2019). Only one study investigated the overexpression role of wild-type TDP-43 (Provasek et al., 2025). Finally, one study used samples from patients with sporadic rather than familial ALS-FTD (Tamaki et al, 2023).

Three primary methodologies were employed to assess the extent of DNA damage: immunostaining for γH2AX, which accumulates at sites of DSBs (Fang et al., 2023; Gianini et al., 2020; Guerrero et al., 2019; Konopka et al., 2020, Marques et al., 2024, Mitra et al., 2019, Mitra et al., 2025, Nogami et al., 2022, Pal et al., 2021, Provasek et al., 2025, Tamaki et al., 2023) (11/12 studies); Comet assays (Fang et al., 2023; Guerrero et al., 2019; Konopka et al., 2020; Mitra et al., 2019; Mitra et al., 2025; Provasek et al., 2025) (6/12 studies); and long amplicon-PCR (Guerrero et al., 2019; Mitra et al., 2019; Mitra et al., 2025) (3/12 studies) (Table 1). In addition, single studies used immunostaining for FANCD2 (Gianini et al, 2020), or 53BP1 (Pal et al, 2021), while TUNEL assay was used in one study (Mitra et al, 2025).

StudyCountryModel systemInterventionsTDP-43 mutationsDNA repair pathwaysMethod of detection(A) Human and rodent cell line-based studiesGianini et al. (2020)Italy/SpainHeLa/SH-SY5YRNAi75% KD/ G294V/ A382TFanconi Anemia pathwayH2AX and FANCD2 staining/(DRIP)-qPCRGuerrero et al. (2019)USASH-SY5YIR 3 GyQ331KNHEJH2AX assay/Comet assay/LA-PCR/PLA assayFang et al. (2023)USAA549/HEK293/3C4shRNAKDNHEJ/HRH2AX assay/NHEJ, and HR assaysMitra et al. (2019)USASH-SY5YRNAi/ CRISPR-Cas9/ ETO/IR 6 Gy30–75% KD/KONHEJH2AX assay/Comet assay/LA-PCRNogami et al. (2022)JapanU251CLM-NHEJ (DNA-PK)Live cell imagingMitra et al. (2025)USASH-SY5YRNAi/ ETO(mNLS)/KDNHEJH2AX assay/MTT assay/Comet assayProvasek et al. (2025)USAHEK293/SH-SY5YRNAi/CRISPR-Cas9 /GO∼50% KD/ ΔNLS/ OE/ Q331K/ A315T/M337VMMRWB/Comet assayLee and Rio (2024)USASH-SY5Y-KD/ N352S-Gene Ontology (GO) Term AnalysisKonopka et al. (2020)AustraliaNSC-34/Cortical neuronsETO/H2O2/DNA-PK inhibitorA315T/ Q331KNHEJH2AX assay/NHEJ assay/Comet assay(B) Brain organoid-based studiesTamaki et al. (2023)CanadaCerebral organoidsInjection of extracted protein from ALS patientssALS containing pathogenic TDP-43-H2AX assay/WB/IHC(C) iPSC-derived motor neurons from ALS patientsGianini et al. (2020)Italy/SpainLCLs-sALS/ A382TFanconi Anemia pathwayH2AX assay/co-IP/Flow cytometryFang et al. (2023)USAALS patientsETO/5-fluorouracilM337V/ Q343RNHEJ/HRRNAi screening/comet assay/ survival assay/NHEJ/HR assaysMitra et al. (2019)USANPCs/MNsRNAi/ETO∼80% TDP-43 KDNHEJH2AX assay/Co-IP/WB/PLAMarques et al. (2024)USAfALS-KD/ TDP-43+/G298SSTING pathwayH2AX assayPal et al. (2021)GermanySpinal MNs-C9ORF72-ALS-H2AX and 53BP1 assayProvasek et al. (2025)USANPCs-KD/ Q331KMMRH2AX assayKonopka et al. (2020)AustraliaFibroblasts cells from ALSfibroblasts derived from pre symptomatic and ALS patientsM337 mutationNHEJH2AX assay/ICC/IB(D) Rodent studiesMitra et al. (2025)USATDP-43 KI miceCRISPR/Cas9KI (hTARDBP NLS)NHEJH2AX assay /TUNEL assay/LA-PCR/ PLA assayProvasek et al. (2025)USANPCs/MNsCRISPR-FLeXOE (murine) TDP-43/ΔNLSMMRH2AX assay/WBKonopka et al. (2020)AustraliatgTDP-43 rNLS mice-cytoplasmic hTDP-43 with NLSNHEJH2AX assay/IHC/ WB(E) Post-mortem brain and spinal cord tissue from ALS patientsGuerrero et al. (2019)USASC tissue-Q331K mutationFanconi Anemia pathwayWB/ IHC/LA-PCRFang et al. (2023)USAPrecentral gyrus Frontoinsular cortex-FTLD-TDP (C9ORF72 and sFTD-ALS)NHEJ/HRH2AX assayMitra et al. (2019)USASC tissue (cervical region)-sALS cases (veterans and Chamorro)NHEJIHC/IB/LA-PCR/TUNEL assayMarques et al. (2024)USAMotor cortex and SC tissues-A315T and sALSSTING pathwayH2AX assayProvasek et al. (2025)USACNS tissues-Guamanian ALS patientsMMRWB/RT-PCR

Characteristics of the included studies focussing on TAR DNA-binding protein 43 (TDP-43).

Genes and proteins: TDP-43, TAR DNA-binding protein 43; TARTDP, gene encoding TDP-43; FUS, fused in sarcoma protein; C9ORF72, chromosome 9 open reading frame 72. Disease classification: sALS, sporadic amyotrophic lateral sclerosis; fALS, familial amyotrophic lateral sclerosis, FTLD-TDP, frontotemporal lobar degeneration with TDP-43 pathology. Genetic tools and models: CRISPR/Cas9, clustered regularly interspaced short palindromic repeats/CRISPR-associated protein 9; RNAi, RNA interference; shRNA, short hairpin RNA; KI, knock-in; OE, overexpression; KD, knockdown; KO, knockout; tg, transgenic; FLeX, flip-excision system; rNLS, reduced or mutated nuclear localization signal. Experimental methods: IF, immunofluorescence; IHC, immunohistochemistry; WB, western blot; IB, immunoblotting; IP, immunoprecipitation; Co-IP, co-immunoprecipitation; ICC, immunocytochemistry; DRIP-qPCR, DNA–RNA immunoprecipitation followed by quantitative PCR; ChIP-seq, chromatin immunoprecipitation followed by sequencing; RT-PCR, reverse transcription polymerase chain reaction; Comet assay, single-cell gel electrophoresis for DNA damage; LA-PCR, long-amplicon PCR; PLA assay, proximity ligation assay; TUNEL assay, terminal deoxynucleotidyl transferase dUTP nick end labeling (apoptosis assay); MTT assay, colorimetric assay for cell viability; Gene Ontology (GO) analysis (bioinformatics method). Cell and tissue sources: iPSC, induced pluripotent stem cells; NPCs, neural progenitor cells; MNs, motor neurons; LCLs, lymphoblastoid cell lines; CNS, central nervous system; SC, spinal cord. DNA repair and mutagenesis assays: NHEJ, non-homologous end joining; HR, homologous recombination; EMS, ethyl methanesulfonate (mutagen). Chemical reagents and stressors: CLM, calicheamicin, ETO, etoposide (a topoisomerase II inhibitor); IR, ionizing radiation; H₂O₂, hydrogen peroxide (oxidative stress inducer), Glucose oxidase (GOx).

Outcomes were categorized as follows. The primary outcome aimed to evaluate the extent of DNA damage accumulation across the five models. Secondary outcomes examined dysfunction in both DDR and DNA repair mechanisms, focusing on the association between TDP-43 proteinopathy, increased DNA damage, DDR activation, and impaired repair processes.

3.4 Primary outcome: detection or accumulation of DNA damage in the CNS

The accumulation of DNA damage was investigated in each of the 12 studies. DSB were identified as the most prevalent form of DNA damage in neurons in 10 of the 12 studies (Fang et al., 2023; Gianini et al., 2020, Guerrero et al., 2019, Konopka et al., 2020, Mitra et al., 2019; Mitra et al., 2025; Nogami et al., 2022; Pal et al., 2021, Provasek et al., 2025, Tamaki et al., 2023) but other types of DNA damage were also identified: increased SSB (Provasek et al., 2025; Fang et al, 2023) and abnormal R-loop formation (Gianini et al, 2020), plus reduced DNA amplification (Gianini et al., 2020), elevated DNA mutation rates (Fang et al, 2023; Guerrero et al., 2019) and increased reactive oxygen species (ROS) levels (Guerrero et al, 2019). The methods used to identify these changes are described below to facilitate comparison between studies.

3.5 DNA damage quantification

Significant quantitative evidence of the accumulation of DNA damage was shown, most notably in the studies involving TARDBP knockdown or mutations linked to ALS-FTD. Mitra et al. (2019) demonstrated a 20-fold increase in DNA damage when TDP-43 was depleted in differentiated SH-SY5Y cells compared to the controls. Moreover, Fang et al. (2023) revealed a correlation between DNA damage and the aggregation of TDP-43 in ALS-FTD patient brain tissues. In the precentral gyrus of ALS brains, 62% of neurons with TDP-43 cytoplasmic inclusions also showed γH2AX foci, while the remaining 38% of the neurons retained normal TDP-43 (p = 0.0229) (Guerrero et al., 2019). Similarly, in the frontoinsular cortex of FTD patient brains, 35% of neurons with TDP-43 cytoplasmic inclusions also showed γH2AX foci, while only 21% of neurons with normal nuclear TDP-43 exhibited such damage (p < 0.001) (Fang et al, 2023). Mitra et al. (2025) used TUNEL analysis to detect DNA damage in the cerebral cortex and spinal cord of 12-month-old transgenic mice expressing mislocalized TDP-43 (MN-Tdp-43∆NLS) compared to sham control. The number of TUNEL-positive nuclei exhibited a 24.40 ± 3.54% increase in the cortex and a 31.10 ± 3.26% increase in the spinal cord compared to sham (cortex and spinal cord both p = 0.0001) (Table 2).

Author/YearDNA damageDDR components affectedDNA repair(A) Human and rodent cell line-based studiesGianini et al. (2020)R-loops causing DSBsATM, H2AX, 53BP1, FANCD2Significant increase in H2AX and FANCD2Guerrero et al. (2019)SSB/ DSB /reduction in DNA amplificationATM, H2AX, 53BP1 (>4-6-fold increase in DDR factors)Delayed in DSB repair after IR/ cytosolic accumulation of the XRCC4-DNA ligase 4 complex