Alex Lugovskoy, PhD, an author of over 100 patents and an expert in antibody R&D, is working hard to unlock one of the most important areas of drug discovery and development: agonist antibodies for heteromeric receptor complexes.
Though Lugovskoy has guided multiple oncology, autoimmunity, and rare disease programs into the clinic, serving as COO of Dragonfly Therapeutics, CDO of Morphic Therapeutic, vice president of therapeutics at Merrimack Pharmaceuticals, and associate director of drug discovery at Biogen, agonist antibodies are an entirely different beast.
It’s a far easier problem to find antagonizing molecules that act like a wedge to jam the receptors and prevent them from forming an active complex than to find a molecule that cross-links transmembrane receptors in such a way that protein adapters on the other side of the membranes will be juxtaposed correctly and trigger a signaling cascade.
The complexity of discovering and developing agonist antibodies lies in the many unknowns, like how many binding sites are present in the receptors, let alone what the correct orientation of these binding sites needs to be engaged to signal.
“If you think through that combinatorially, there’s a lot of different possibilities,” said Lugovskoy, an entrepreneur in residence at Atlas Venture. “If you want to recruit two different receptors to talk to each other, you have to find places for two parts of a bispecific antibody to bind. With two receptors, there are thousands of binding sites, which leads to millions of possible conformations. If we were to permute all of them, that would give us billions and billions of combinations, which is not feasible to interrogate experimentally.”
But in the past two years, Lugovskoy has been putting the pieces together to solve the heteromeric receptor complex assembly problem in a stealth start-up called Diagonal Therapeutics. Lugovskoy and his team officially launched Diagonal Therapeutics today with a $128 million Series A funding round. This money will support their DIAGONAL platform, which uses advanced computational and experimental methods to find rare agonist antibodies that restore deficient signaling pathways and have created a pipeline of new therapeutics.
BVF Partners and Atlas Venture co-led the financing, with support from Lightspeed Venture Partners, RA Capital Management, Frazier Life Sciences, Viking Global Investors, Velosity Capital, and Checkpoint Capital.
Connecting the dots
The DIAGONAL platform uses experimentally generated deep immune repertoires of binders to receptor subunits of a target heteromeric receptor complex. To do so, the researchers at Diagonal Therapeutics start by using in vitro display libraries and in vivo animal immunization.
“We want to make sure that every single binding opportunity is captured,” said Lugovskoy. “We try to generate as many binders to receptor units as we want to bring together. At this point, our antibodies only bind to one receptor subunit or the other.”
The resulting catalogs are typically in the 100,000s of antibody binder sequences for each receptor, which means more than 10 billion possible combinations. With these libraries, Diagonal Therapeutics then uses a combination of artificial intelligence and physics-based computational methods to identify non-redundant subsets of antibodies to build a smaller set of bispecific antibodies for experimental testing.
This multi-step computational process to filter out redundancy to capture maximal diversity requires computational clustering of the antibodies by differences in their shape and then their complementarity to receptors using proprietary docking techniques developed in collaboration with Dima Kozakov, PhD, a core faculty member at Boston University’s BioMolecular Engineering Research Center (BMERC). These steps bring the numbers down to around twenty groups per receptor and, thus, 400 possible combinations.
From these groups, the Diagonal team then generates agonist bispecific antibodies based on all possible pairwise epitope bin combinations, which they experimentally profile and optimize to select leads for preclinical characterization. According to Lugovskoy, this platform reduces possible combinations from 1011 to 100 antibody candidates for testing.
“We’re using computational techniques for enrichment, and nature tells us which one of those combinations is productive,” said Lugovskoy. “If we find the smallest subsets that will still bind to all the possible combinations, we will build those molecules experimentally, use them in high-throughput assays, and we’ll let nature tell us which one of those clades is successful. Then, since every single antibody clade is computationally analyzed, we can very rapidly go back to do lead diversification and optimization. With any campaign, we can go from any heteromeric complex to a functional agonist in eight months or less, which is very comparable to regular antibody discovery campaigns.”
While the company name is rooted in an acronym for “digital antagonist antibody ligands,” Lugovskoy said the concept of a diagonal also captures how the company brings receptors together in asymmetric confirmations.
The clinical angle
With their DIAGONAL platform, the start-up has discovered agonist antibody candidates against four complex targets that incite key signaling events. Diagonal’s pipeline centers around genetic diseases, with lead candidates for two cardiovascular disorders: hereditary hemorrhagic telangiectasia (HHT) and pulmonary arterial hypertension (PAH).
Lugovskoy believes that the combination of novel technology, a medical need, and the pre-clinical data they have generated catalyzed this group of investors to support Diagonal.
“The company has come together in the last two years around how we can blend cutting-edge AI computational techniques and experimental engineering techniques to help us find those rare agonists,” said Lugovskoy. “I brought this idea to [Atlas partner] Michael Gladstone, and he immediately realized its potential and proposed that we form a company. We are very thankful to the bluest of blue-chip investors for their support in raising $128 million. These sophisticated investors are experienced company builders.”
The funding will enable Diagonal Therapeutics’ lead asset, discovered with its DIAGONAL platform, to undergo clinical proof of concept. At this point, Lugovskoy declined to provide any guidance on the timing of clinical milestones for the Boston-based company.
And now that the company is out of stealth, Lugovskoy expects to be engaged with potential partners. Given the uniqueness of the technology, Lugovskoy believes there are many potential mutually beneficial partnerships to be had that would allow partners to get access to innovation and for Diagonal to increase the reach of their platform in bringing new medicines to patients.
