Best of: The future of proteins

Episode Summary

Executive Summary: Russ Altman and Polly Fordyce discuss proteins as molecular machines that power cellular life, how her lab uses high-throughput nanoliter-scale experiments to link sequence, structure, and function, and why the next frontier is decoding how proteins read DNA and control gene expression. The conversation highlights broad mutational effects, transcription-factor specificity, and the promise of scalable tools for biology and therapy.

Main Topics: Proteins as molecular machines (Priority: 5/5): Fordyce explains proteins as the body's workhorses, emphasizing motor proteins like kinesin that convert chemical energy into movement and transport cargo inside cells. Structure-function relationships in proteins (Priority: 5/5): The discussion centers on how amino-acid sequences fold into 3D structures that determine biochemical activity, including the emerging impact of AI tools like AlphaFold on structure prediction. High-throughput protein engineering platform (Priority: 5/5): Fordyce describes her lab's fluidic, nanoliter-scale platform that enables thousands of experiments at once, replacing traditional one-at-a-time test-tube assays. Mapping enzyme function beyond the active site (Priority: 4/5): Using the phosphatase PAFE, the lab shows that mutations far from the active site can still strongly affect function and conformational state, revealing a broader functional architecture. Protein-DNA recognition and gene regulation (Priority: 5/5): The conversation shifts to transcription factors, how they locate DNA binding sites, and how DNA sequence context and structural grooves influence where proteins bind and what genes they activate. Toward scalable, portable measurement tools (Priority: 4/5): Fordyce outlines efforts to make measurements easier for other labs, including color-coded hydrogel beads that can encode many different binding reagents for distributed use.

Key Arguments: Proteins are essential molecular machines, not just dietary nutrients; they catalyze reactions, transport cargo, and provide structure in cells. Understanding protein function requires more than identifying the active site because mutations throughout the protein can alter activity and folding. High-throughput, quantitative measurements at physical scales like energy and rate are crucial because they enable comparison across laboratories and integration into larger datasets. AlphaFold and similar models have advanced structure prediction, but the harder and more important next challenge is predicting function from sequence. Transcription factors do not simply bind a motif; surrounding DNA context and protein domains outside the folded DNA-binding core strongly influence binding and gene activation. Therapeutic specificity may come from targeting distant surfaces or allosteric features rather than conserved active sites shared by many protein family members. Making instruments and portable assay systems available to more labs could accelerate discovery and standardize protein-function datasets worldwide.

Data Points: Protein transport step size: 8 nanometers - Kinesin motor protein walks in 8-nm steps along cellular filaments. Experimental scale: 1,000 experiments at a time - Fordyce describes her lab's nanoliter fluidic platform versus one test tube at a time. Variant library size: 1,500 different enzyme variants - Her team can systematically mutate and test many enzyme versions in parallel. Protein length: 526 amino acids - The enzyme PAFE is described as having 526 linear building blocks. Functional sensitivity: About two-thirds - Approximately two-thirds of PAFE's amino-acid positions affect function when changed. Comparative chemical scale: Less than a tenth of an angstrom - Vanadate and tungstate transition-state analogs differ by under 0.1 Å yet are differentially affected by distant mutations. Genome size: 25 megabase genome - Altman references the scale a protein must navigate when finding binding sites in cells. Portable encoding capacity: Up to 1,000 different colors - Color-coded hydrogel beads can encode many attached molecules for scalable assays.

Pivotal Quotes: "they turn chemical energy into motion" — Polly Fordyce: Describing kinesin motor proteins as nanoscale machines that walk along cellular filaments. "turns out that the whole thing matters" — Russ Altman: Summarizing the recurring lesson that protein function depends on regions beyond the obvious active site or core DNA-binding domain. "we want to get to the next level of the problem, which is going from how sequence encodes the shape to how it encodes the actual function" — Polly Fordyce: Explaining the next frontier after structure prediction.

Implications: The episode underscores a shift from describing biomolecules to quantitatively engineering them. Better scalable measurements could improve enzyme design, drug specificity, gene-regulation models, and the spread of standardized protein-function data across labs.