The science of delivering cures straight to your cells | Eric Kelsic

Episode Overview

• The episode features Eric Kelsic, CEO of Dyno Therapeutics, discussing the primary challenges and future potential of gene therapy.
• A core theme is the "gene delivery problem"—the difficulty of getting therapeutic genetic material into the right cells in the human body.
• Kelsic explains how his company is using AI and machine learning to engineer viral protein shells (capsids) to create highly efficient and targeted delivery vehicles.
• The ultimate vision is to make gene therapy a mainstream, accessible, and even personalized form of medicine for thousands of genetic diseases.

Key Concepts

• Gene Therapy: A medical approach that aims to treat or cure diseases by correcting the underlying genetic problem. It offers the potential for a one-time, lifelong cure.
• Capsid: The protein shell of a virus. In gene therapy, the capsid from a non-harmful virus (like AAV) is used as a delivery vehicle to carry therapeutic DNA into a patient's cells.
• AAV (Adeno-Associated Virus): A small, non-disease-causing virus that is a leading delivery vector for gene therapies. Its small size allows it to access many parts of the body.
• The Delivery Problem: The single biggest challenge in gene therapy is getting the genetic payload to the target cells with high efficiency and specificity, while avoiding other tissues and the immune system.
• Directed Evolution: A traditional lab method of creating millions of random mutations in a protein (like a capsid) and screening them to find one with improved function, a process often compared to finding a "needle in a haystack."
• AI-Guided Design: Dyno Therapeutics' approach, which uses machine learning models to analyze vast amounts of experimental data. The AI learns the rules of capsid biology to predict which specific modifications will improve its function, replacing random screening with intelligent design.

Quotes

• At 00:39 - "So it's the potential for a one-time treatment for a disease where you wouldn't otherwise be able to reach the cells and solve for the root cause of the disease." - Kelsic explains the core promise of gene therapy: its ability to provide a permanent cure by fixing the genetic issue at its source.
• At 01:50 - "Capsids are the protein shells of Adeno-associated virus... AAVs, Adeno-associated virus, is a parasite of other viruses." - This quote introduces the fundamental biological tool being engineered—the AAV capsid—and explains its natural origin as a safe and efficient vehicle for entering cells.
• At 03:07 - "As a one-time treatment, it can be an effective cure." - Highlighting the ultimate goal of gene therapy, Kelsic emphasizes that once the therapeutic gene is delivered to long-lasting cells (like neurons), it can treat the disease for the patient's entire life.

Takeaways

• The bottleneck in innovation is often delivery, not the core technology. Gene therapy's potential has been understood for over 50 years, but its progress was stalled by the challenge of getting the therapy to the right place in the body. Solving the delivery mechanism is key to unlocking its full potential.
• AI can dramatically accelerate biological engineering. Instead of relying on slow, random trial-and-error (directed evolution), machine learning can analyze massive datasets to understand complex biological rules and predict successful designs, making the process of creating new therapies exponentially faster and more efficient.
• Future medicine will become increasingly personalized. By using AI to rapidly design bespoke delivery vehicles, it may become economically feasible to create gene therapies for rare and even "ultra-rare" diseases that affect only a handful of patients, moving medicine toward a truly individualized model.