Alternative splicing coupled to nonsense-mediated mRNA decay (AS-NMD) is described in this work as an evolutionarily conserved mechanism that controls gene expression. Dysregulation of AS-NMD has been implicated across a range of human diseases, including developmental defects and cancer. A central feature of AS-NMD is the presence of poison exons (PEs)—highly conserved alternative exons that introduce a premature termination codon when included in a transcript and thereby trigger NMD. Cancer cells can co-opt PE inclusion or exclusion to alter levels of tumor suppressors or oncoproteins, respectively. Because PEs can be used by malignancies to modulate gene expression, they represent a potential therapeutic target for cancer and other diseases.
The authors test a genome-editing approach to manipulate PE-mediated oncogenic AS-NMD. They used a CRISPR-based strategy employing paired guide RNAs to delete a PE from the genome. The therapeutic concept assessed is that removing a PE from a tumor suppressor locus would prevent its inclusion in mRNA, allow transcripts to escape AS-NMD, and restore protein expression and downstream function.
As a proof-of-concept, the study targeted a PE in the tumor suppressor EZH2 in models of leukemia bearing mutations in the splicing factor SRSF2. Paired guide RNAs were used to excise the EZH2 poison exon at the genomic level. Following this editing, cells produced EZH2 mRNAs that lacked the PE and therefore did not undergo AS-NMD. The edited cells showed restored EZH2 protein expression in contrast to the unedited, PE-containing transcripts that were subject to decay.
The manuscript reports that removal of the EZH2 PE by CRISPR editing resulted in molecular rescue of the EZH2 pathway: restored EZH2 protein led to recovery of downstream histone modification—specifically H3K27 methylation—consistent with reestablished EZH2 function.
Restoration of EZH2 protein and H3K27 histone methylation translated into rescue of defective chromatin regulation that had been associated with the splicing factor mutation. The authors report that these molecular changes were linked to recovery of impaired hematopoietic stem cell differentiation in their model system. These findings support the concept that correcting an oncogenic AS-NMD event at the genomic level can reverse both molecular and cellular defects driven by splicing factor mutations.
The authors note a prior antisense technology they developed targeting the same EZH2 PE. In this study they demonstrate what they describe as preferential advantages of the CRISPR deletion approach over that antisense method for targeting the EZH2 poison exon. Specific comparative data, quantitative metrics, or detailed experimental parameters for the comparison are not reported in the abstract and would require consulting the full preprint for additional experimental detail.
This work is presented as a preprint and has not been certified by peer review. The competing-interest statement indicates that M.A.R., M.R.I., and P.N. are inventors on a patent application not related to this study; other authors declare no conflicts. Funding sources declared for the study include the National Institute of General Medical Sciences (NIGMS, R35GM154991) and support from the Winthrop P. Rockefeller Cancer Institute.
The authors propose that CRISPR-mediated deletion of PEs provides compelling proof-of-concept evidence as a therapeutic strategy to correct oncogenic AS-NMD in cancer and potentially other human diseases. The abstract emphasizes molecular restoration of a tumor suppressor (EZH2), rescue of H3K27 methylation, and recovery of hematopoietic differentiation as key supportive outcomes.
The abstract does not report certain experimental details in this summary, including the specific guide RNA sequences, editing efficiencies, off-target analysis, in vivo validation, long-term outcomes, safety data, or comparative quantitative results versus the antisense approach. Those details would need to be obtained from the full preprint text and supporting data, which are referenced in the article information.
Note: This report is a preprint and has not undergone peer review. Readers should interpret the findings in that context and consult the full manuscript for methodological and data details.