An Adnp frameshift variant disrupts Wnt signalling inducing chromatocytoskeletal defects and autism-related behaviour in male mice

Le résumé fourni par la source

BACKGROUND: Heterozygous de novo variants in the transcription factor Activity-Dependent Neuroprotective Protein (ADNP) cause a severe neurodevelopmental disorder, termed Helsmoortel-Van der Aa syndrome (HVDAS), characterised by autism, intellectual disability, and multisystem involvement. The ADNP gene is essential for embryonic development and interacts with components of several chromatin remodelling complexes. However, the precise pathophysiological mechanisms underlying the disorder remain incompletely understood. METHODS: We used CRISPR/Cas9 genome editing to create a 14-base pair deletion c.2463_2476del (p.Leu822Hisfs∗6) in the murine Adnp gene to establish a disease-relevant mouse model. Only male mice were used in this study to reduce variability associated with sex-specific differences. Molecular, transcriptomic (RNA-seq), chromatin accessibility (ATAC-seq), proteomic (mass spectrometry), and 3D genome architecture (Hi-C) analyses were performed in cortical brain tissue. Neuroanatomical and behavioural phenotyping, including Morris water maze, elevated plus maze, marble burying, social interaction testing, and the Live Mouse Tracker were conducted to assess cognitive and autism-related phenotypes. FINDINGS: Heterozygous mice are viable and fertile. Introduction of the 14-base pair deletion reduced cellular Adnp levels in the brain (p = 0.0008) and decreased its chromatin association (p = 0.001), parallelled by a genome-wide increase in chromatin accessibility. Morphological analyses revealed mild neuroanatomical alterations in regions associated with cognition, memory, learning, and motor function (p < 0.05). Behavioural testing confirmed cognitive impairment in the Morris water maze (p < 0.05), increased anxiety-like behaviour in the elevated plus maze (p = 0.0035), repetitive behaviour in the marble burying assay (p = 0.045), and impaired social interactions (p < 0.05). Transcriptome sequencing of the frontal cortex, an essential region involved in executive functions, cognition, and motor control, revealed predominant downregulation of the Wnt signalling pathway (p = 0.03). Cytoskeletal abnormalities were further coupled to synaptic plasticity deficits, dysregulation of transcription factors implicated in lineage specification, and alterations in neuronal cell numbers. Adnp directly regulated mechanisms of synaptic plasticity through interaction with Camk2a and Dbn1. In heterozygous mice, these protein interactions were disrupted, resulting in aberrant Camk2a phosphorylation at synapses (p = 0.012). Mass spectrometry identified changes in multiple chromatin-interacting proteins and cytoskeletal components, corroborating nuclear and cytoskeletal dysregulation observed at the transcriptomic level. Hi-C analysis detected locus-specific alterations in 3D genome architecture associated with transcriptional changes. INTERPRETATION: We generated a disease-relevant heterozygous Adnp mouse model that recapitulates key molecular and behavioural features observed in patients with Helsmoortel-Van der Aa syndrome. Our findings demonstrate a role for Adnp in chromatin regulation and Wnt signalling, coupled to aberrant expression of cytoskeletal components and synaptic dysfunction. This mouse model provides a mechanistic framework linking chromatin dysregulation to autism-related behaviours and represents a valuable platform for future preclinical studies. FUNDING: This work was supported by the Marguerite-Marie Delacroix Foundation (FFP240064 to C.P.D.), the Research Fund of the University of Antwerp (Methusalem grant "GENOMED" to R.F.K.), ERA-NET NEURON ("ADNPinMED"), and the European E-RARE programme ("IMPACT"). Additional support was provided by a crowdfunding initiative from the German ADNP parent community to support C.P.D.; E.P. was supported by the FWO grant G0A1Z24N. C.A.M. received support from the Marguerite-Marie Delacroix fellowship. D.S. was supported by an EMBO fellowship (ALTF 1564-2025). F.P. was supported by a Human Frontier Science Program Long-Term Fellowship (LT000111/2021-L) and an EMBO Long-Term Fellowship (ATLF 716-2020). W.V.B. acknowledges support from the University of Antwerp (ID46420/48076/50799) and COST Action CA18127 (International Nucleosome Consortium 5INC). W.V.B., F.K., E.P., and C.P.D. acknowledge the IMPULS BOF 2024. Infrastructure support was provided by FWO (including IRI grant I000123N, GOH4216N).

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.