DNA damage burden causes selective CUX2 neuron loss in neuroinflammation
Article Date: 01 April 2026
Article URL: https://www.nature.com/articles/s41586-026-10310-3
Article Image: (not provided)
Summary
This Nature paper shows that upper-layer cortical excitatory neurons (layer 2/3 excitatory neurons; L2/3ENs), marked by CUX2, accumulate persistent DNA damage during neuroinflammation and demyelination and are selectively lost in multiple sclerosis (MS) and mouse models. The authors combine human post-mortem tissue, longitudinal mouse models (including DTA and Myrf-cKO), conditional genetics (Cux2 and Atf4 loss), single-nucleus RNA-sequencing and cellular assays to show that:
1) MS L2/3ENs have elevated DNA-damage markers (γH2AX, 53BP1) and a transcriptional DNA damage response (DDR) signature, yet show evidence of DDR exhaustion.
2) Mouse models of pan-cortical demyelination/neuroinflammation replicate selective thinning and loss of CUX2+ L2/3 neurons, with increased DNA double- and single-strand breaks and oxidative nucleic acid damage.
3) CUX2 (and ATF4) functionally promote DNA repair — including non-homologous end joining (NHEJ) — and drive expression of RPA3; loss of Cux2 impairs DDR and increases neuron vulnerability after injury.
4) Interferon-γ (IFNγ) drives reactive oxygen species (ROS), causes oxidative DNA damage in neurons, and is sufficient in vivo to induce selective L2/3 neuron loss; antioxidants or boosted repair (CUX2/RPA3/ATF4) reduce damage and improve survival.
Key Points
- Human MS L2/3 excitatory neurons show elevated DNA damage markers (γH2AX, 53BP1) and DDR activation but still bear a high damage burden.
- Multiple mouse models (DTA, Myrf-cKO, AS-IFNγ) reproduce selective loss and thinning of upper cortical layers (L2/3) during demyelination or chronic IFNγ exposure.
- CUX2 is required for neuronal DNA-repair competence: it transcriptionally activates RPA3 and supports efficient NHEJ and repair kinetics.
- ATF4 cooperates with CUX2 to promote resilience; Cux2/Atf4 double loss markedly increases persistent DNA damage and L2/3 neuron loss.
- IFNγ directly increases ROS in neurons and glia, generating oxidative DNA/RNA lesions and DSBs; antioxidants (NAC, Mito-TEMPO) mitigate this damage in vitro and improve neuron viability.
- Boosting DNA-repair pathways or lowering IFNγ/ROS are proposed therapeutic strategies to protect vulnerable cortical neurons in MS.
Context and relevance
The work links neuroinflammation, oxidative stress and an exhausted neuronal DNA damage response to the selective vulnerability of CUX2+ upper-layer cortical neurons in MS. It connects genetics (CUX2 as an MS-risk gene / RPA3 regulation), cell biology (NHEJ and BER pathways), and immunology (IFNγ-driven ROS) to explain cortical atrophy that persists despite modern immune-targeting therapies. For researchers this clarifies a cell-intrinsic mechanism of degeneration; for clinicians and translational scientists it suggests actionable approaches: reduce IFNγ/ROS signalling or enhance neuronal DNA repair capacity to slow cortical neurodegeneration and atrophy in progressive MS.
Why should I read this?
Short and blunt — this paper gives you a mechanistic bridge between inflammation and neuron loss in MS. If you care about why upper cortical layers thin in progressive MS, how IFNγ and ROS damage neurons, or whether DNA-repair pathways can be targeted to protect the brain, this one cuts to the chase. It’s packed with human tissue, solid mouse genetics and cell assays that show both the problem and routes to fix it (antioxidants, repair factors).
Author style
Punchy: the authors map an explanatory path from T cell/IFNγ-driven oxidative stress to neuron DNA damage, show that intrinsic repair factors (CUX2/ATF4/RPA3) determine whether those neurons survive, and demonstrate that bolstering repair or limiting ROS can blunt loss. If you work on MS progression, neuroinflammation or DNA repair in neurons, the detailed methods and multi-model evidence make this a high-impact read.
Implications
Therapeutic strategies that antagonise IFNγ signalling, reduce ROS, or directly enhance neuronal DNA-repair (for example via targets in the NHEJ/BER axis or upstream regulators such as CUX2/ATF4/RPA3) may help prevent or slow cortical atrophy and progression in MS. The data support trials of antioxidant or DNA-repair–focused interventions and motivate genetic/biomarker work to identify patients with heightened L2/3EN vulnerability.