BPS2026 – Probing the rules of protein structure-function relationships through the design of functional proteins with fine-tuned stability

A central challenge in protein science is understanding how folding stability and conformational dynamics govern structure-function relationships, particularly in the context of protein-peptide interactions. We address this question through the de novo design of stimuli-responsive protein masks, termed Enfolds, which encapsulate and conditionally release bioactive peptides such as antimicrobial peptides (AMPs) and cell-penetrating peptides (CPPs). Although these peptides hold significant therapeutic potential, their clinical translation has been limited by off-target toxicity, poor serum stability, and inefficient delivery. Enfolds provide a tunable scaffold that balances two competing demands—maintaining robust peptide sequestration under basal conditions while enabling rapid and selective release in response to disease-relevant cues such as proteolysis, redox imbalance, or pH changes. To explore this stability-responsiveness trade-off, we developed an integrated design pipeline combining physics-based molecular dynamics, AI-guided sequence optimization, and structural refinement to systematically adjust hydrogen bonding, electrostatics, and hydrophobic interactions. Designs were tailored to their peptide cargos, embedding cleavable linkers or conformational switching motifs. Biophysical characterization using circular dichroism melts, urea denaturation, and proteolysis assays established quantitative links between folding energetics, conformational plasticity, and peptide release. Cell-based imaging further confirmed that optimized Enfolds achieve efficient and highly selective triggered release. This work establishes Enfolds as a tractable model for interrogating the fundamental coupling of protein stability, dynamics, and function. By demonstrating that folding landscapes can be rationally tuned to balance robustness with conditional activation, we highlight generalizable principles for programmable protein design. More broadly, these results point toward a new class of de novo protein systems that extend beyond static scaffolds, enabling dynamic control of bioactive molecules for therapeutic delivery and synthetic biology applications.