Engineered tRNA reduces vision loss in a mouse model of Leber congenital amaurosis

Signal Transduction and Targeted Therapy ARTICLE www.nature.com/sigtrans OPEN Engineered tRNA reduces vision loss in a mouse model of Leber congenital amaurosis Pawan K. Shahi1,2, Enes Akyuz1,2, Lionel Gissot3, Ahmad Al Saneh3, Sanjai K. Pillala1,2, Divya Sinha 2,4, Giovanni M. Hanstad4, Meha Kabra1,2, Maria A. Fernandez Zepeda4, David M. Gamm2,4,5, Samuel M. Young Jr.6,7, Christopher A. Ahern 3 and Bikash R. Pattnaik 1,2,5 ✉ 1234567890();,: Premature termination codons (PTCs) are a major class of pathogenic variants that underlie rare inherited disorders, including forms of childhood blindness. Therapeutic suppression of these “nonsense mutations” offers a gene- and position-agnostic strategy to restore protein function. Our previous work established that the W53X PTC in the KCNJ13 gene causes Leber congenital amaurosis type 16 (LCA16) by disrupting the inwardly rectifying potassium channel Kir7.1, leading to retinal pigment epithelium (RPE) dysfunction. Here, we present a proof-of-concept approach using anticodon-engineered transfer RNA (ACE-tRNA) to promote targeted translational readthrough. We engineered a suppressor tRNA encoding tryptophan (ACE-tRNATrp.UAG) to selectively recognize the UAG stop codon at the W53X site, enabling incorporation of the correct amino acid and restoration of full-length Kir7.1 protein. Delivery of ACE-tRNA via helper-dependent adenovirus (HDAd) resulted in robust rescue of channel function in heterologous systems expressing mutant KCNJ13 and in patient-derived human induced pluripotent stem cell (hiPSC)-RPE cells. Functional recovery was confirmed by electrophysiological assays demonstrating restored inwardly rectifying currents and membrane potential. Importantly, subretinal delivery of HDAd-ACE-tRNATrp.UAG in a W53X mouse model led to partial restoration of RPE physiology, as measured by electroretinography, without evidence of retinal toxicity. Together, these findings establish ACEtRNA-mediated suppression as a viable therapeutic strategy for nonsense mutations in multimeric ion channels and provide a translational framework for precision treatment of inherited retinal diseases. Signal Transduction and Targeted Therapy (2026)11:225 INTRODUCTION Nonsense mutations introduce premature termination codons (PTCs) into messenger RNA (mRNA), leading to truncated proteins that often lack normal structure and function. These variants represent a major class of disease-causing mutations across diverse human genetic conditions.1 Nonsense mutations account for approximately 15% of all known inherited disorders, including cystic fibrosis, Duchenne muscular dystrophy, and certain forms of congenital blindness.2 Under such conditions, the presence of a PTC disrupts normal translation, preventing the synthesis of fulllength proteins required for cellular function. The resulting lossof-function phenotypes frequently lead to progressive and severe clinical manifestations, particularly in tissues with limited regenerative capacity, such as muscle and retina. Given the broad contribution of nonsense mutations to human disease context, there is a critical need to develop therapeutic strategies capable of restoring functional protein expression from PTCcontaining transcripts. A range of therapeutic approaches has been explored over the past several decades to overcome the effects of nonsense mutations, each with distinct advantages and limitations. Some ; https://doi.org/10.1038/s41392-026-02793-3 of the recent developments are the pharmacological readthrough agents, including aminoglycoside antibiotics, and geneediting techniques.3 Both read-through drugs and aminoglycoside antibiotics, such as gentamicin, can induce partial readthrough of PTCs by allowing near-cognate tRNA incorporation.4 While this strategy can partially restore protein synthesis, it often introduces incorrect amino acids at the mutation site, raising concerns about protein stability and function, particularly in proteins that are sensitive to missense substitutions.5 Genome editing technology, such as CRISPR/Cas9, has enabled precise correction of pathogenic variants at the DNA level. However, these approaches are inherently mutation-specific, require efficient delivery and transient expression of editing machinery, and demand extensive evaluations of potential off-target effects. These technical and biological challenges are especially pronounced in post-mitotic tissues, such as the retina, where longterm safety and precision are paramount.6–8 Furthermore, many nonsense mutations occur as ultra-rare “n of 1”, patient-specific variants, underscoring the need for therapeutic platforms that are not only effective but also adaptable across mutations and genes. 1 University of Wisconsin-Madison, Department of Pediatrics, Madison, WI, USA; 2University of Wisconsin-Madison, McPherson Eye Research Institute, Madison, WI, USA; University of Iowa, Carver College of Medicine, Department of Molecular Physiology and Biophysics, Iowa City, IA, USA; 4University of Wisconsin-Madison, Waisman Center, Madison, WI, USA; 5University of Wisconsin-Madison, Department of Ophthalmology and Visual Sciences, Madison, WI, USA; 6Center for Molecular Medicine, University of North Carolina-Chapel Hill, Chapel Hill, NC 27599, USA and 7Department of Pediatrics, Department of Pharmacology, Department of Microbiology and Immunology, University of North Carolina-Chapel Hill, Chapel Hill, NC 27599, USA Correspondence: Bikash R Pattnaik () These authors contributed equally: Pawan K. Shahi, Enes Akyuz 3 Received: 21 November 2025 Revised: 9 April 2026 Accepted: 11 May 2026 © The Author(s) 2026 Engineered tRNA reduces vision loss in a mouse model of Leber congenital. . . Shahi et al. 2 Anticodon-engineered transfer RNAs (ACE-tRNAs) represent a promising alternative strategy that directly engages the endogenous translational machinery to suppress PTCs. By modifying the anticodon sequence, ACE-tRNAs can be designed to recognize specific stop codons and insert the correct amino acid during translation, thereby restoring full-length protein synthesis without altering the underlying DNA sequence.9–11 This approach addresses key limitations associated with small-molecule readthrough agents by ensuring accurate amino acid insertion and reducing the risk of generating dysfunctional proteins.12 ACEtRNAs have demonstrated their capability to rescue expression of PTC-containing transcripts in vitro across a range of diseaserelevant genes, including CFTR,13,14 CDKL5,15 and the cardiac potassium channel HERG.16 Although in vivo efforts using small molecules such as ataluren or aminoglycosides have shown limited success in achieving clinically meaningful mRNA or protein-level read-through, these approaches have not consistently yielded sufficient functional protein to produce meaningful physiological benefit.3,17–19 Similarly, early attempts at viral delivery of suppressor tRNAs have faced challenges achieving robust, sustain (...truncated)