9Bio Therapeutics

Cohort CA1

Engineering next-generation cancer therapies by reprogramming proteins to act selectively within tumors

Current targeted cancer therapies often cause severe side effects due to off-target binding, limiting their effectiveness and the development of new treatments. 9Bio addresses this by engineering conditionally active therapeutics that specifically target the tumor microenvironment. Their computational platform uses advanced modeling and AI to design proteins that selectively bind to tumor cells, minimizing harm to healthy tissue. This innovative approach, which focuses on selective paratope mutations, aims to create safer and more effective treatments for currently underserved cancers.

What is the problem?

Targeted oncology therapeutics, such as antibody-drug conjugates (ADCs), often face significant challenges related to toxicity, pharmacokinetics (PK), and a limited druggable landscape. On-target, off-tumor binding can result in severe toxicities, causing dose-limiting side effects and treatment discontinuation. For example, HER2 expression in pulmonary tissue can lead to interstitial lung disease, a life-threatening condition that may cause permanent lung damage. Managing these toxicities often requires reducing doses or stopping treatment altogether, limiting therapeutic benefit. This unintended binding in healthy tissues also restricts the development of novel targeted oncology therapies. Consequently, the majority of ADCs are focused on a small set of validated targets, limiting options for hard-to-treat cancers.

What is their solution?

9Bio’s computational platform engineers conditionally active therapeutics by leveraging the unique metabolic characteristics of the tumor microenvironment (TME) in solid cancers, such as acidosis and hypoxia (the Warburg effect). Our platform accelerates the discovery of conditionally active targeted therapeutics by employing an integrated methodology that combines structural modeling, physics-based simulations, molecular dynamics modeling, and AI-driven predictions. These computational tools enable us to design therapeutic proteins that selectively engage tumor cells while minimizing off-target binding to healthy tissues. Once we have designed potential candidates, we express the proteins and validate their binding in vitro to confirm our predictions and further refine our models. By streamlining candidate selection and minimizing experimental cycles, our platform accelerates the development of targeted oncology therapeutics with improved tumor specificity, enhanced safety, and broader therapeutic potential.