Gene delivery is, arguably, the biggest bottleneck in gene therapy. Adeno-associated viruses (AAVs) are the historical mainstay, but they are expensive to manufacture and have payload limits and immunological issues. Lipid nanoparticles, which delivered the mRNA COVID-19 vaccines, are mostly limited to targeting the liver. To deliver gene-based therapeutics both cost-effectively and to a wider variety of tissue targets, a different approach is required.
Polymer nanoparticles may be the answer. With virtually unlimited options for chemical features, “you can design bespoke delivery vehicles easier than with a lipid-based system,” says Sean Kevlahan, PhD, co-founder and CEO, Nanite. “There’s (approximately) 1060 possible polymers you can synthesize, and perhaps 1025 known stars in the universe. Without the power of computation, machine learning, and artificial intelligence, you could never screen all those possible different formulations and combinations.” Polymer nanoparticles are also simpler to manufacture.
Nanite takes an engineering approach to polymer-based delivery discovery, Kevlahan says. “When we started Nanite in 2021, we found that many scientists don’t completely understand what chemical features drive certain polymers to specific destinations—the lung versus the liver, for example.”
Rather than focusing on one or two polymers, Nanite works with its pharma and biotech partners to build fit-for-purpose polymer delivery vehicles. An artificial intelligence (AI) platform called SAYER assesses polymers based upon first-principle assays, such as identifying the polymer’s composition, the charges on its surface, and how the polymer interacts with specific nucleic acid–based “cargos.”
“The power of this approach allows us to generate a huge corpus of data,” Kevlahan says. “[AI] gives us the ability to explore the polymer universe without having to actually, empirically, identify each star in that universe. Then we can use the latest computational methods to understand which chemical features drive localization to different organs and design the polymer nanoparticles accordingly with our partners.”
Likely candidates are then assessed in preclinical in vitro or animal-based experiments. To do this—at the simplest level—the team makes a diverse set of polymers, loads them with different types of nucleic acid–based cargo, pools them into one sample, and injects them into animals. Then they simply see what nanoparticles went where. All the data is fed back to SAYER to generate predictions for more accurate targeting and delivery performance.
“We haven’t identified a (cargo) size limit yet,” he reports. So far, the nanoparticles have carried molecules as small as 15 bases and as large as whole ribonucleoproteins. “You can do that easier with a polymer-based system because you don’t need complex formulation equipment to make polymer nanoparticle droplets.”
Experienced founders
Nanite’s co-founders are seasoned entrepreneurs. Kevlahan and Shashi Murthy, PhD, now CTO, previously co-founded Quad Technologies, which was acquired by Bio-Techne, and co-founder Thomas Neenan, PhD, now CBO, also co-founded AbFero Pharmaceuticals (acquired by Pharmacosmos) and Panbela Therapeutics. In addition, Murthy founded Flaskworks, acquired by Northwest Biotherapeutics. They’ve worked together in various capacities for more than a decade. Having a seasoned founding team that’s worked together such a long time resonates well with potential investors and partners.
They are also chemical engineers, so it was natural they would apply an engineering approach to building a workflow around such issues as data quality, throughput, and cost per data point. “We are each fascinated by the convergence of materials science, biology, and computation,” Kevlahan remarks.
Delivery was a pain point
“The nucleation point came during the COVID-19 vaccine research,” he says. “There was unbelievable momentum within mRNA and transient-based therapeutics … but if there isn’t a good delivery system, you don’t have a therapeutic.”
AAV vectors (whether recombinant or wild type) are limited by their small cargo size (approximately 5 kb), immunological issues (such as neutralizing antibodies that block AAV delivery), and manufacturing that is inefficient and expensive. However, rAAV gene therapy programs have successfully targeted the liver, striated muscles, and the central nervous system.
Lipid nanoparticles are the unsung heroes of the COVID-19 mRNA vaccines, he says. “They were the first big foray into nonviral mediated delivery.” They are limited, however, by their tendency to accumulate in the liver, low drug payloads, and—for solid lipid nanoparticles—drug expulsion or bursts. “To go after different diseases that are outside the liver, for example, cystic fibrosis (in the lungs) or Charcot-Marie-Tooth disease (in the peripheral nerves), you need different chemical features that are not liposomes.”
Polymers, Kevlahan thought, could be a next-generation, nonviral delivery vehicle with potential even greater than lipids. They are quite stable, are functionalized easily, can be tuned, have high loading capacities, and can deliver multiple agents with varying levels of hydrophilicity and molecular weight in one carrier simultaneously.
New company challenges
Nanite’s polymers are in early development stages with several different confidential partners. Consequently, Kevlahan says, “We’re well capitalized and we’re generating revenue.” For example, the company received $6 million in seed funding last year.
At about the same time, Nanite also received an investment of about $2.5 million from the Cystic Fibrosis Foundation and the Charcot-Marie-Tooth Research Foundation to develop a system to deliver gene therapy to the lungs and the peripheral nervous system, respectively. A priority of the cystic fibrosis program is designing a vehicle that resists the mucus that coats cystic fibrosis patients’ lungs. “Given the breadth of the design space, this goal can be readily incorporated,” he says.
A current challenge is managing version changes in successive iterations of the SAYER platform. Improvements in one area invariable necessitate adaptations in another. Kevlahan likens this to a huge flywheel. “It’s running at a certain speed. You standardize things around it. Then, when you implement a change, you also have to (consider those standards and) develop new protocols, ensure data quality is maintained, etc. Managing version changes is probably the biggest challenge when operating a platform.”
So far, Nanite connects with potential partners and investors primarily through conferences and one-on-one interactions. (Its website throughout 2023 was a single succinct landing page that will likely expand this year). “We’re an inch wide and a mile deep, producing more and more training data (for the AI) and more AI-based predictions,” Kevlahan says. In the coming year, he expects to advance Nanite’s relationships with the Cystic Fibrosis Foundation, the Charcot-Marie-Tooth Research Foundation, and additional partners. Noting that the possibilities for therapeutic delivery number in the novemdecillion (1060) range, he says that Nanite is confident that “polymers are going to take over the whole gene delivery space.”
