Gene and Cell Therapy · global
Making Gene Delivery to the Brain and Muscles More Precise: Dyno Releases Animal Data on AI-Designed Capsids
Engineering viral shells could help therapeutic genes enter target tissues more efficiently while reducing interference from preexisting antibodies. Dyno’s primate and human serum experiments offer preliminary signals, but whether improved delivery efficiency will translate into patient benefit remains to be verified.
For gene therapy to work, it must deliver the appropriate gene to the cells that need it. Delivery efficiency in the brain and muscles, unintended exposure in other organs, and interception by preexisting antibodies in the human body all pose development challenges. Newly released research from Dyno Therapeutics seeks to address these barriers simultaneously by modifying the protein shell of adeno-associated virus (AAV).
On October 27, Dyno announced three sets of preclinical results, with presentations scheduled at the annual meeting of the European Society of Gene and Cell Therapy in Hamburg, Germany. The company uses AI to help design capsids, followed by experimental screening and refinement. AI’s specific role here is to adjust interactions between delivery vectors and receptors and antibodies, rather than directly determine efficacy against disease. Coverage of the same event by the Boston Life Sciences Times also summarized the three studies, but the data still come from company disclosures.
One capsid targeting the alkaline phosphatase ALPL showed brain gene transduction 152 times that of AAV9 in nonhuman primate experiments. Transduction refers to the process by which a vector delivers genetic material into cells; this figure cannot be interpreted as a 152-fold increase in therapeutic effect. The same capsid also delivered genetic material to skeletal muscle: approximately 55% of quadriceps muscle fibers were transduced, compared with approximately 5% for another external comparator capsid. AAV9 was not the comparator in this muscle comparison.
Another set of studies focused on muscle, combining integrin-binding capability with a muscle receptor mechanism whose details the company has not disclosed. At a dose of 5 × 10¹² vector genomes per kilogram of body weight, vector distribution in nonhuman primate skeletal muscle was approximately 20 times that of MyoAAV-4E, while average transduction was approximately 40 times higher. The former reflects the extent to which vectors reach tissue, while the latter more closely reflects the entry of genetic material into cells. These two measures are not interchangeable.
The third study addressed the immune barrier. Through three rounds of AI-assisted design, the team incorporated mapping of antibody recognition regions on the capsid. Among the human serum samples tested, the proportion that was seronegative reached as high as 85% for the new capsids, compared with 49% for AAV9, while interactions with receptors associated with the central nervous system and muscles were retained. This means that, under those experimental conditions, more samples showed no detectable antibody response. It cannot be directly extrapolated to mean that 85% of patients would be eligible for treatment, nor does it demonstrate that repeat dosing would be safe.
These findings still have clear evidentiary limits. Neither source provided sufficient information on animal numbers, serum sample size, statistical uncertainty, or long-term follow-up to allow a full assessment. The announcement also did not identify the external comparator capsid. Cross-species performance and human serum testing can help screen candidate vectors, but remain insufficient to establish safety or efficacy in humans.
For gene therapy development, the significance of these findings lies in providing delivery tools that may reach target tissues more efficiently. To translate them into usable therapies, further work must pair the capsids with specific therapeutic genes and establish effective doses, tissue distribution, and toxicity. Human trials must then test whether better delivery figures can translate into improvements that patients can experience.