The future of proteins

Episode Summary

Executive Summary: The episode explores how high-throughput protein biophysics is revealing that protein function depends on far more than active sites or folded cores. Polly Fordyce describes using nanoliter-scale assays to test thousands of variants at once, uncovering how mutations alter enzyme activity, protein structure, and DNA-binding behavior. The goal is to decode function well enough to design proteins for chemistry, therapeutics, and gene control.

Main Topics: Proteins as molecular machines (Priority: 5/5): Fordyce explains why proteins are central to cellular work: they catalyze reactions, move cargo, provide structure, and execute DNA instructions. Her fascination began with kinesin, a motor protein that literally walks along cellular filaments. High-throughput protein measurement (Priority: 5/5): Her lab replaces one-at-a-time test-tube experiments with nanoliter-scale platforms that can test roughly a thousand to fifteen hundred variants simultaneously, enabling systematic mapping of sequence-to-function relationships. Protein structure-function relationships (Priority: 5/5): The conversation emphasizes that amino-acid sequence encodes 3D folding, but function is distributed across much of the protein. In Fordyce's enzyme work, mutations far from the active site still meaningfully change catalytic behavior and conformation. Enzyme functional architecture (Priority: 4/5): Using the phosphatase PafE, Fordyce's lab found that many residues outside the active site matter, and that clusters of distant residues create a 'functional architecture' that shapes activity and ligand interactions. DNA recognition by transcription factors (Priority: 5/5): The discussion shifts to how proteins read DNA to turn genes on and off. Fordyce describes transcription factors like Fofor that sense phosphate levels, recognize specific DNA motifs, and recruit machinery that initiates transcription. Toward design and therapeutics (Priority: 5/5): A central future goal is not just predicting structure, but engineering proteins with desired functions for green chemistry, plastic cleanup, and drug development. Mapping allosteric surfaces could enable more specific therapies. Scaling and democratizing biophysical tools (Priority: 4/5): Fordyce aims to make these experiments easier for other labs through shared instruments and portable bead-based platforms whose colors encode attached molecules, enabling wider adoption of high-throughput measurements.

Key Arguments: Protein function cannot be understood by focusing only on active sites; residues far away can strongly affect catalysis and binding. High-throughput assays create a qualitative and quantitative leap because thousands of variants can be tested under comparable conditions. Absolute physical measurements such as binding energy and kinetics are valuable because they can be compared across labs and combined into shared datasets. Sequence-to-structure prediction is now much more advanced, but sequence-to-function remains the major frontier. DNA-binding proteins do not simply recognize a short motif; neighboring sequence context and 3D DNA shape also matter. Understanding functional architecture may reveal druggable sites outside conserved active sites, improving specificity for diseases such as cancer. Future bioengineering may turn proteins into designed molecular machines that solve environmental and medical problems.

Data Points: Protein variants tested simultaneously: ~1,000 to 1,500 - Fordyce describes her lab's nanoliter platforms for running many experiments at once. Kinesin step size: 8 nanometers - She cites the motor protein kinesin as a molecular machine that walks in discrete steps. Distance in the body: Base of spine to tip of toes - Used to illustrate why diffusion is too slow for some cellular transport tasks. Enzyme length: 526 amino acids - The PafE phosphatase studied by Fordyce and collaborator Dan Herschlag. Residue impact on function: About two thirds - She says roughly two-thirds of PafE amino acids affect function when mutated. Transition-state analog difference: Less than a tenth of an angstrom - Vanadate and tungstate are nearly identical yet can be differentially affected by distant mutations. Genome size: 25 megabases - Fordyce uses this as a scale example for how transcription factors locate their DNA targets. Target DNA motif: C-A-C-G-T-G - The consensus site for one transcription factor discussed in the DNA-binding section. Back catalog size: More than 200 interviews - Mentioned by the host at the end as a promotional note.

Pivotal Quotes: "What we really want to do next is we want to get to the next step where we could actually make machines that would do things that we wanted." — Russ Altman: Opening framing of the episode’s vision for engineered proteins. "It turns out that in order to get where we want to go to actually design a functional molecule, we're not going to be able to just take that active site and stick it in another protein and call it good." — Polly Fordyce: Explaining why distant residues and full-protein context matter for enzyme function. "The whole thing matters." — Polly Fordyce: Summarizing the recurring lesson that both protein cores and disordered regions contribute to function.

Implications: The episode suggests a shift from predicting protein shape to engineering protein behavior. If successful, this could enable better drugs, tailored enzymes, and scalable biology tools for medicine, manufacturing, and environmental cleanup.