One of the most persistent challenges in gene and cell therapy is not simply delivering a therapeutic gene to the right cell — it is ensuring that, once delivered, the gene’s expression can be turned on and off with sufficient precision to remain both safe and effective over time. Current inducible systems, which allow researchers to control transgene activity with a small molecule, have offered partial solutions to this problem; unfortunately, most of them depend on exogenous proteins that are foreign to the human immune system, raising the risk of adverse immune responses, or on compounds that have not yet been tested in clinical settings, complicating regulatory approval. A study published in May 2026 in Nature Communications by Mendel, Schwarz, Sun and colleagues from F. Hoffmann-La Roche and Genentech now reports a compact, molecularly elegant solution to this longstanding problem: a gene switch they call RisdiON, controlled by risdiplam, an orally bioavailable small molecule already approved for clinical use in the treatment of spinal muscular atrophy.
The central molecular logic of RisdiON rests on the cell’s own pre-mRNA splicing machinery. Risdiplam is a splicing modifier — it works by altering the activity of the spliceosome, the large ribonucleoprotein complex responsible for removing introns from precursor messenger RNAs and joining exons to generate mature transcripts. Mendel and colleagues engineered risdiplam-responsive intronic sequences directly into the regulatory region of a transgene construct. In the absence of the drug, the splicing event defaults to a configuration that prevents productive translation of the transgene; when risdiplam is added, its interaction with the spliceosomal machinery shifts the splicing pattern, enabling the correct joining of exons and generating a translatable mRNA. A further refinement, which the authors call the split-ATG architecture, places the translation initiation codon across the splicing junction, so that only correctly spliced transcripts produce a functional start codon and can therefore be translated. Importantly, because the system hijacks endogenous splicing factors rather than recruiting exogenous proteins, the expressed therapeutic protein emerges without artificial tags — a feature that reduces immunogenicity.
In a series of experiments spanning immortalized cell lines, human induced pluripotent stem cells, primary T cells, and mouse models, the team demonstrated that RisdiON achieves dose-dependent, reversible transgene expression across a wide range of promoter contexts. The system was used to drive inducible chimeric antigen receptor (CAR) expression in primary T cells — a configuration directly relevant to adoptive cell therapies for cancer — and to regulate Cas9 activity for precise genome editing, a context in which uncontrolled nuclease expression poses significant safety risks. In vivo delivery via adeno-associated virus (AAV) vectors yielded reversible transgene expression in mice, confirming the platform’s compatibility with one of the most clinically advanced gene delivery modalities available.
The therapeutic implications of these results are considerable, though several limitations must be acknowledged. First, risdiplam is not without pharmacological effects of its own: as an approved splicing modifier for spinal muscular atrophy, its chronic systemic use at the doses required to maintain inducible gene expression will need careful evaluation for off-target splicing perturbations in sensitive tissues. Second, the study was conducted primarily in cell-culture and murine models, and the efficiency, durability, and immunogenicity profile of the system in larger animal models and, ultimately, in human clinical trials remains to be established. Third, the degree to which the split-ATG architecture prevents leaky baseline expression across all promoter and chromatin contexts will require further characterization before deployment in highly safety-sensitive applications. Taken together, however, RisdiON represents an important conceptual advance: it demonstrates that the cell’s own RNA processing machinery can serve as a programmable molecular switch, unlocking a new design principle for next-generation gene and cell therapies that prioritize reversibility, titrability, and immunological compatibility.
Leave a Reply