SAN FRANCISCO—Kicking off J.P. Morgan week, Dyno Therapeutics announced that Roche has exercised its option to license a novel capsid for use in a gene therapy program for an undisclosed neurological disease.
The capsid license follows an agreement between Dyno and Roche originally announced in May 2020 and marks Dyno’s first adeno-associated virus (AAV) capsid licensing and partnership completion. Dyno will receive an option exercise fee of $7 million with the potential to earn more than $220 million in associated development, regulatory, and commercial milestone payments plus royalties.
“This latest advancement in one of our partnerships is further validation of the proven effectiveness of our gene delivery platform and continued and rapid advancement of our sequence design capabilities,” said Eric Kelsic, PhD, CEO of Dyno. Dyno was founded in 2018 and applies artificial intelligence (AI) and quantitative in vivo experiments to address gene therapy’s delivery challenge.
Naturally occurring AAV capsids lack precise targeting to deliver medicine to the right cells, are difficult to manufacture, and have pre-existing immunity. To address this gap, Dyno’s platform leverages AI to navigate large sequence space and create synthetic AAV capsids with optimized properties for therapeutics.
Kelsic states that the company’s approach has been data driven from the beginning.
“You could explore the capsid sequence space randomly but there would be a lot of trial and error. AI can automate decision–making so that we can make the most efficient library and optimize for improvement of function,” said Kelsic in an interview with GEN.
Last year, Dyno introduced Low-shot Efficient Accelerated Performance (LEAPSM) to efficiently generate capsid sequences with improved performance beyond any of the capsids in the training data. The technology enables Dyno to advance capsids more positioned to succeed into in vivo validation studies, thereby potentially enabling lower manufacturing costs.
Kelsic also emphasized gathering information from the right contexts as one important area of platform investment. For example, most of Dyno’s screening is focused on non-human primates to address the translational gap. At the same time, animal experiments can be challenging and expensive.
“In one animal, we can measure hundreds of thousands of different capsid sequences and to which organs they’re delivering their payloads,” Kelsic said. “Investing in the technology to get the most information from a single animal is one way we’ve differentiated ourselves at Dyno.”
Dyno continues to maintain a partnership-centric approach with several collaborations underway. In October 2024, Roche and Dyno announced a second collaboration agreement, with the potential to generate more than $1 billion, that provides Roche with further access to Dyno’s platform for the treatment of neurological diseases, including LEAPSM. Additionally, Dyno maintains partnerships with Sarepta Therapeutics and Astellas Pharma focusing on gene therapy delivery to the muscle.
“We’re not competing with our partners. We’re giving them a competitive edge. Finding ways to give our partners more options so that they can choose the right strategy is the idea that’s motivating how we evolve the platform now,” Kelsic told GEN.
Kelsic states the company is also planning to prioritize machine learning partnerships. In May 2024, Dyno announced a collaboration with NVIDIA’s optimized cloud and BioNeMo platform to accelerate the design of high-performance biological sequences.
Taken together, Kelsic emphasizes Dyno’s mission to build high-performance genetic technologies.
“Right now, we’re known as experts in delivery as an AAV platform but the technologies are much more generalizable,” Kelsic said. “We can point to our success in capsids and help other biopharma partners understand how they can make use of the capabilities that we spent several years developing.”
