When COVID-19 emerged as a global pandemic, the world mobilized to deliver vaccines and treatments at unprecedented speed. This massive achievement saved millions of lives and demonstrated the power of global scientific collaboration. However, SARS-CoV-2 continues to evolve, and other coronaviruses with pandemic potential remain a concern. To stay ahead of the next outbreak, we must develop antiviral medicines that work broadly across known and emerging coronaviruses.
Takeda and Schrödinger recently collaborated to make important progress toward that goal. Together, we discovered a highly potent noncovalent inhibitor of the coronavirus main protease, Mpro, which is a critical viral enzyme that enables coronaviruses to replicate inside human cells. This research, published in the Journal of Medicinal Chemistry, describes how our teams leveraged structure-guided design and predictive modeling to drive the rapid discovery of a next-generation antiviral candidate.
Mpro: The Virus’ Weak Spot
Coronaviruses such as SARS, MERS, and SARS-CoV-2 rely on the enzyme Mpro to reproduce, using it to cut long viral proteins into smaller pieces needed for replication. Since humans have no equivalent enzyme, Mpro is an especially attractive target for antiviral drugs that can block viral replication with minimal risk of off-target effects.
Existing oral Mpro inhibitors were developed in response to the urgent need for antiviral treatments during the COVID-19 pandemic. These medicines have helped millions of patients by reducing viral load and preventing severe disease. The field now has an opportunity to build on this foundation by advancing molecules with improved profiles, including reduced potential for drug–drug interactions and activity that extends across a wider range of coronavirus strains. Such progress can help ensure that future therapies remain effective even as the virus continues to evolve.
To prepare for emerging pandemic threats, our teams at Schrödiner and Takeda set a clear objective: design a next-generation noncovalent Mpro inhibitor that is potent, safe, and broadly active across coronaviruses.
An Integrated, “Predict-First” Design Strategy
Mpro is structurally well characterized, yet designing inhibitors with both high potency and broad coverage requires precise molecular engineering. Takeda brought deep expertise in medicinal chemistry and structural biology. Working together, our teams leveraged Schrödinger’s suite of solutions including physics-based FEP+ (Free Energy Perturbation) and LiveDesign, a cloud-native collaborative environment for molecular design. These tools enabled us to build an agile discovery engine that worked around the clock.
First, using Schrödinger’s Glide software, we performed large-scale virtual screening on Takeda’s chemical library of 1.5 million compounds. This approach rapidly narrowed the search to a handful of promising candidates, which would have taken several months using traditional experimental methods alone.
Takeda scientists determined the X-ray crystal structure of SARS-CoV-2 Mpro bound to one of the early hit molecules, which contained a piperazine scaffold, a common six-membered ring with two nitrogen atoms that serves as a basic chemical framework in many drugs. This structure revealed how the molecule nestled into the Mpro active site, providing precise guidance for further optimization.
At this stage, Schrödinger’s FEP+ simulations became the engine of innovation. FEP+ predicts changes in binding affinity with remarkable accuracy, accounting for protein flexibility and induced-fit motions of key amino acids such as Gln189, Met49, and Met165. These subtle shifts open hidden “cryptic” pockets that create new binding opportunities.
We ran our FEP+ calculations directly within LiveDesign, giving our global, interdisciplinary team a unified interface to design, predict, analyze, and collaborate. With all experimental data and in silico results centralized in one place, our work never stalled. When Takeda team members in Japan wrapped up a set of calculations, Schrödinger team members in New York, Portland or San Diego could immediately pick up the next steps. This kept the project moving 24 hours a day. LiveDesign’s integrated modeling tools and collaborative interface also helped us eliminate silos through real-time sharing of data, models, and results, leading to faster, more creative design cycles and seamless, around-the-clock decision-making. We evaluated more than 7,000 compounds computationally and only synthesized 85 of them — a massive time and resource savings compared to conventional discovery efforts that depend on large-scale synthesis and screening.
Introducing Compound 30
By leveraging Takeda’s medicinal chemistry expertise within a global cycle of design, synthesis, and testing, we arrived at compound 30, a noncovalent Mpro inhibitor with compelling properties:
- Potently inhibits Mpro from multiple human coronaviruses
- Shows high selectivity against human proteases
- Demonstrates strong cellular antiviral activity, including against Omicron
- Is approximately eight times more potent than nirmatrelvir in Omicron antiviral assays
- Exhibits a favorable safety and metabolic profile in early studies
Our next steps include optimizing the oral bioavailability and pharmacokinetics of compound 30, advancing toward clinical development, and expanding testing against a wider range of coronavirus strains.
Every coronavirus mutation and every animal spillover event reminds us that viral evolution is relentless, and the ideal time to develop antivirals is before we urgently need them. A broadly active, oral Mpro inhibitor could strengthen resilience against emerging variants and expand global access to treatments, since oral small-molecule drugs are easy to store and distribute. Compound 30 represents a blueprint for the development of next-generation antivirals that could provide durable protection against both current and future coronavirus threats. This progress brings us closer to having antivirals ready before the next threat appears.
