ERBB4 Perturbation Supports a Developmental Cytoskeletal Reactivation Hypothesis in Alzheimer's Disease

Aberrant reactivation of developmental cytoskeletal programs may contribute to Alzheimer's disease (AD). We tested whether a prespecified Arp2/3-RAC1-MYH10-KIF5C module is regulated by Erb-B2 receptor tyrosine kinase 4 (ERBB4) in an AD-model neuronal state, providing a perturbational test of the hypothesis that aberrant re-engagement of developmental cytoskeletal machinery contributes to AD.

We performed a secondary analysis of publicly available single-nucleus RNA-sequencing data from ERBB4 perturbation experiments using pseudobulk analysis with the primary conservative excitatory-neuron classifier. A prespecified 10-gene module comprising Actr2, Actr3, Arpc1a, Arpc2, Arpc3, Arpc4, Arpc5, Rac1, Myh10, and Kif5c was tested using CAMERA and expression-matched permutation analyses.

ERBB4 disruption in 5 × FAD excitatory neurons downregulated all 10 module genes (CAMERA P = 0.0207). Arpc4, Arpc1a, and Kif5c remained significant after false-discovery-rate correction. Expression-matched permutation testing independently supported pathway downregulation (mean log-fold-change P = 0.0102; 10/10 genes down P = 0.0041; CAMERA permutation P = 0.0150). ERBB4 overexpression in wild-type neurons did not significantly alter the complete module, although Arpc5 was significantly induced, suggesting selective component responsiveness.

Independent perturbational data support ERBB4-dependent regulation of a prespecified Arp2/3-centered cytoskeletal program in AD excitatory neurons, with effects dependent on the disease context.