Lewis Kay on NMR’s Expanding Role in the Post–AlphaFold Era

When Lewis Kay took the podium at the Eastern Analytical Symposium (EAS), he confronted a question that has hovered over structural biology since the advent of AlphaFold: If algorithms can predict protein structures with remarkable accuracy, what role remains for experimental techniques like nuclear magnetic resonance (NMR)?

Kay, a professor at the University of Toronto and senior scientist at the Hospital for Sick Children, offered a clear answer: NMR is not obsolete—it is indispensable. “AlphaFold frees NMR from the burden of determining small protein structures,” he told the audience. “That allows us to focus on what really matters: dynamics.”

His remarks came as he accepted the 2025 EAS Award for Outstanding Achievements in Magnetic Resonance, a recognition of decades of work advancing NMR technology and its applications in biology.

Kay’s research has long centered on the molecular machines that sustain life—proteasomes, apoptotic scaffolds, and other large complexes whose functions hinge on motion. While crystallography and cryo-electron microscopy excel at capturing static structures, they struggle to reveal the choreography of these systems. NMR, by contrast, can probe dynamics at atomic resolution, even in assemblies approaching a megadalton in size. This capability is hard-won. Large molecules tumble slowly in solution, causing conventional NMR signals to decay before they can be detected. Kay’s group overcame this barrier through a combination of isotope labeling and what he calls “spin alchemy”—engineering methyl groups within a deuterated environment to create narrow, long-lived signals. These innovations have extended NMR’s reach from small proteins to massive molecular machines.

One example Kay highlighted involves the apoptosome, a seven-part scaffold that orchestrates intrinsic apoptosis. The central question: How do caspase-9 protease domains activate once recruited to the complex? Competing hypotheses suggested either scaffold-induced activation or simple proximity-driven dimerization. Kay’s team combined biochemical assays with NMR spectroscopy to resolve the debate. Their data showed that caspase-9 remains monomeric on the scaffold until a substrate or inhibitor is present—at which point rapid dimerization occurs. This mechanism ensures that “molecules of death” remain inert until the cell commits to apoptosis, a safeguard against premature destruction.

Kay also turned to a phenomenon that has reshaped cell biology: biomolecular condensates. These membraneless organelles concentrate proteins and nucleic acids, creating environments where diffusion slows to a crawl. For folded proteins entering these viscous phases, the effect mimics the slow tumbling of large complexes—a scenario tailor-made for Kay’s NMR strategies. In one study, his group examined superoxide dismutase (SOD1), a protein implicated in familial ALS, within stress granules formed by the scaffold protein Caprin-1. NMR revealed that SOD1 partially unfolds upon entering the condensate, exposing regions that drive aggregation. This insight links phase separation to neurodegenerative pathology—a connection invisible to static structural methods.

Kay’s message was clear: computational breakthroughs like AlphaFold are not competitors—they are catalysts. By providing accurate structural models, these tools allow NMR to focus on what it does best: uncovering motion, interaction, and regulation at the atomic level.

“We can now focus on the dance, not just the pose,” he said.

For a discipline often defined by static pictures, this shift underscores a broader truth: understanding biology requires more than knowing what molecules look like. It demands insight into how they move, interact, and sometimes fail—a challenge NMR is uniquely equipped to meet.