In vivo base editing of Chd3 rescues behavioural abnormalities in mice

Steven

In vivo base editing of Chd3 rescues behavioural abnormalities in mice

Summary

This Nature paper reports a targeted in vivo base-editing approach that corrects a recurrent de novo CHD3 variant (c.C3073T, p.R1025W) associated with Snijders Blok–Campeau syndrome (SNIBCPS). The authors generated a humanised knock-in mouse (Chd3_hR1025W/+) that shows reduced CHD3 protein, impaired neuronal morphology and a range of SNIBCPS-like phenotypes: reduced pup ultrasonic vocalisations, cognitive deficits, social impairments, increased repetitive behaviours and hypotonia. The team engineered a TadA‑embedded adenine base editor (TeABE-1248) delivered via a dual AAV split-intein system (AAV-PHP.eB in mice; AAV9 intrathecal in nonhuman primates). Systemic (tail-vein) delivery in mice achieved region-wide neuronal editing, restored CHD3 protein levels, normalised locus-specific chromatin accessibility and transcription, and importantly rescued multiple behavioural and motor phenotypes. In nonhuman primates, intrathecal AAV9 delivery yielded widespread neuronal transduction and robust intein-mediated reconstitution of the editor with measurable A-to-G edits at cognate sites, supporting translational feasibility. Off-target and bystander editing were low in vivo, and the optimised TeABE narrowed the editing window to reduce unwanted edits.

Key Points

  • The recurrent CHD3 R1025W variant causes accelerated proteasomal degradation of CHD3, producing loss-of-function effects that disrupt neuronal morphology and behaviour.
  • A humanised Chd3_hR1025W/+ mouse recapitulates core SNIBCPS phenotypes: dysphonia-like vocalisation deficits, cognitive impairment, social deficits, repetitive behaviours and hypotonia.
  • The authors developed and optimised a TadA‑embedded ABE (TeABE-1248) with a narrowed editing window to reduce bystander edits while retaining robust on-target correction.
  • Dual AAV split-intein delivery (systemic in mice; intrathecal in NHPs) enabled efficient brain transduction and reconstitution of full-length TeABE in neurons.
  • In vivo editing restored CHD3 protein levels, reversed locus-specific chromatin closing and transcript dysregulation, and rescued behavioural and motor deficits in treated mice.
  • Nonhuman primate studies demonstrated broad parenchymal transduction and ~70–80% intein reconstitution, with detectable editing activity at an orthologous locus — an encouraging step toward clinical translation.

Context and relevance

CHD3 is a NuRD complex ATP-dependent chromatin remodeller; pathogenic CHD3 variants cause a dominantly inherited neurodevelopmental syndrome with speech and intellectual disabilities and frequent autistic features. This study combines precise base editing and AAV delivery to reverse a defined pathogenic SNV in the brain — a significant advance because CNS-targeted base editing faces unique delivery and safety challenges. By demonstrating molecular rescue, behavioural recovery in mice and editor activity in primates, the work provides a compelling preclinical proof of concept for treating dominant neurodevelopmental disorders that arise from single-nucleotide lesions.

Author’s note (punchy)

Big-picture: this is the kind of paper you flag for translational gene therapy. They didn’t just tinker with cells — they made a humanised mouse, fixed the mutation across the brain, showed behaviour and molecular readouts return towards normal, and then showed the editor actually reconstitutes and acts in the primate brain. It’s a rigorous, multi-level demonstration that base editing can be pushed toward real CNS applications — but delivery, dosing and durability remain the next, hard yards.

Why should I read this?

If you work in neurodevelopment, gene therapy, or genome engineering — or just want to know whether base editing can actually fix brain disorders — skim this now. The team sorted out editing specificity, delivered the editor across the blood–brain barrier in mice, reversed behaviour, and showed the approach is feasible in primates. In short: big step from concept to preclinical reality.

Source

Article Date: 18 February 2026
Source: In vivo base editing of Chd3 rescues behavioural abnormalities in mice — Nature (full article)