作者
Oscar G. Wilkins,Max Z. Y. J. Chien,Josette J Wlaschin,Simone Barattucci,Peter Harley,Francesca Mattedi,Puja R. Mehta,Maria Pisliakova,Eugeni Ryadnov,Matthew J. Keuss,David Thompson,Holly Digby,Lea Knez,Rebecca L. Simkin,Juan Antinao Díaz,Matteo Zanovello,Anna‐Leigh Brown,Annalucia Darbey,Rajvinder Karda,Elizabeth Fisher,Thomas J. Cunningham,Claire E. Le Pichon,Jernej Ule,Pietro Fratta
摘要
Loss of function of the RNA-binding protein TDP-43 (TDP-LOF) is a hallmark of amyotrophic lateral sclerosis (ALS) and other neurodegenerative disorders. Here we describe TDP-REG, which exploits the specificity of cryptic splicing induced by TDP-LOF to drive protein expression when and where the disease process occurs. The SpliceNouveau algorithm combines deep learning with rational design to generate customizable cryptic splicing events within protein-coding sequences. We demonstrate that expression of TDP-REG reporters is tightly coupled to TDP-LOF in vitro and in vivo. TDP-REG enables genomic prime editing to ablate the UNC13A cryptic donor splice site specifically upon TDP-LOF. Finally, we design TDP-REG vectors encoding a TDP-43/Raver1 fusion protein that rescues key pathological cryptic splicing events, paving the way for the development of precision therapies for TDP43-related disorders.