ASO-mediated paternal UBE3A unsilencing for UBE3A c.67C>T (p.Arg23Ter): rugonersen Phase 1 evidence and variant-specific considerations
CONCLUSION
For UBE3A c.67C>T (p.Arg23Ter), an early nonsense variant that abolishes maternal UBE3A expression, antisense oligonucleotide (ASO)-mediated unsilencing of the intact paternal UBE3A allele is the most clinically advanced therapeutic strategy. Rugonersen (RO7248824), targeting UBE3A-ATS, has demonstrated an acceptable safety profile and dose-dependent EEG normalization with signals of clinical improvement in the Phase 1 TANGELO trial (NCT04428281, n=61). This approach bypasses the need to correct the maternal nonsense variant directly, instead restoring UBE3A from the epigenetically silenced but genetically intact paternal copy. The strategy is variant-agnostic for loss-of-function maternal alleles, making p.Arg23Ter an ideal candidate.
EVIDENCE
The TANGELO Phase 1 trial (Hipp et al., Nat Med 2025; PMID:40646322) enrolled 61 children aged 1-12 with Angelman syndrome and demonstrated that intrathecal rugonersen led to dose-dependent partial normalization of AS-associated EEG delta-power abnormality (2-4 Hz band), a validated pharmacodynamic biomarker of UBE3A deficiency. Exploratory clinical endpoints on Bayley-III and Vineland-3 showed improvement signals beyond natural history expectations. The mechanism — ASO-mediated degradation of UBE3A-ATS (the antisense transcript that silences the paternal UBE3A allele) — has been validated in mouse models where paternal Ube3a reinstatement restores behavior and proteome (Punt et al., Mol Autism 2025; PMID:40877933). Additionally, ION582 (Ionis/GeneTx) and GTX-102 (Ultragenyx) represent parallel ASO programs targeting the same biological mechanism, with ION582 in Phase 3, reinforcing the therapeutic class validity. ClinVar classifies c.67C>T as Pathogenic, consistent with complete loss of UBE3A function from the maternal allele.
UBE3A c.67C>T introduces a premature stop codon at position 23 of the 875-amino-acid UBE3A protein, effectively a null allele. Because UBE3A is subject to genomic imprinting with paternal silencing in neurons, loss of the maternal allele results in complete neuronal UBE3A deficiency. The paternal allele remains genetically intact but transcriptionally silenced by UBE3A-ATS, a long non-coding antisense transcript. ASO-mediated degradation of UBE3A-ATS via RNase H represents an elegant therapeutic logic: restore endogenous protein from the native locus without requiring exogenous gene delivery or mutation-specific correction. This mechanism has been independently validated by multiple groups and drug programs (Roche/rugonersen, Ionis/ION582, Ultragenyx/GTX-102), providing orthogonal confidence in the target biology.
LIMITATIONS
All ASO data for Angelman syndrome remain in early-phase trials; no Phase 3 efficacy readout for rugonersen is available yet. The TANGELO trial was open-label and uncontrolled, so clinical improvement signals cannot be definitively attributed to treatment versus developmental maturation. Intrathecal delivery requires repeated lumbar punctures, posing procedural burden especially in young children. The optimal therapeutic window is uncertain — earlier treatment may be critical, but the youngest cohort data are limited. Whether paternal UBE3A unsilencing can achieve sufficient protein levels to fully rescue the neurological phenotype, particularly in patients with established neuronal circuit deficits, remains an open question. Long-term durability and safety of chronic ASO administration in the developing CNS require further study. Finally, while the strategy is variant-agnostic, this analysis specifically applies to loss-of-function variants; dominant-negative variants (if any exist for UBE3A) would require different consideration.