News

Scientists Build Programmable Protein Cages That Mimic Viruses

Published

on

<p>A virus has no blueprint&period; Over billions of years&comma; nature refined a trick&colon; build large&comma; complex shells from a single protein that plays different geometric roles depending on where it sits&period; On May 20&comma; 2026&comma; researchers at the University of Washington&&num;8217&semi;s <a href&equals;'https&colon;&sol;&sol;www&period;ipd&period;uw&period;edu&sol;' target&equals;'&lowbar;blank' rel&equals;'noopener'>Institute for Protein Design<&sol;a> &lpar;IPD&rpar; published two back-to-back studies in Nature showing that AI-powered computational methods can now replicate that trick from scratch&comma; with structures confirmed at atomic resolution and cages demonstrated inside living mammalian cells&period;<&sol;p>&NewLine;<p>The new protein cages are <strong>two to three times larger<&sol;strong> than anything the same laboratory previously built using strict symmetry&period; They can be loaded with molecular cargo&comma; enter cells&comma; and be tuned to different sizes by adjusting a single geometric parameter&period; The foundational problem of building virus-like nanoscale containers on demand&comma; from proteins designed entirely at a computer&comma; just shifted from open question to solved problem&period;<&sol;p>&NewLine;<h2>The Soccer Ball Principle&comma; Scaled to the Nanometer<&sol;h2>&NewLine;<p>Think of a soccer ball&period; Its surface panels come in two shapes&comma; pentagons and hexagons&comma; arranged so the pentagons introduce curvature and the hexagons fill the flatter regions in between&period; No single panel changes its size or chemistry&period; Together&comma; the arrangement closes a flat lattice into a sphere&period; Viral capsids have used this same geometric logic for billions of years to protect genetic material during transit between host cells&period;<&sol;p>&NewLine;<p>That logic has a name&colon; quasisymmetry&period; Structural biologist Aaron Klug and biophysicist Donald Caspar first formally described the principle in 1962&comma; analyzing how icosahedral viruses build shells far larger than a strictly symmetric icosahedron allows&period; In a strictly symmetric cage&comma; every protein subunit is surrounded by identical neighbors&period; In a quasisymmetric cage&comma; chemically identical subunits adopt subtly different backbone conformations depending on their local environment&comma; letting the shell grow far larger before it closes&period;<&sol;p>&NewLine;<p>Prior computational protein design had hit a ceiling at the strictly symmetric&comma; 60-subunit icosahedron&period; Quasisymmetric designs demand that one amino acid sequence accommodate multiple distinct local geometries simultaneously&comma; a requirement that earlier methods lacked the precision to satisfy&period; Shunzhi Wang&comma; now an assistant professor at New York University&&num;8217&semi;s &lpar;NYU&rpar; Grossman School of Medicine and the paper&&num;8217&semi;s lead author&comma; built the new strategy around an insight borrowed from soft-matter physics&period;<&sol;p>&NewLine;<p>The obstacle he exploited is called <strong>geometric frustration<&sol;strong>&colon; a flat hexagonal lattice cannot tile a spherical surface without distortion&period; Rather than treating that frustration as a problem to eliminate&comma; the team used it as a design lever&period; Accept the frustration&comma; allow pentagonal defects to appear at calculated positions in the growing hexagonal sheet&comma; and the lattice bends into a curve and then closes into a hollow sphere&period; The cage size is set by how many hexagons appear between each pair of pentagons&comma; a tunable quantity controlled by the geometry of the bridging linker protein&period;<&sol;p>&NewLine;<figure class&equals;"wp-block-image aligncenter featured-image" style&equals;"margin&colon;1&period;5em auto&semi;text-align&colon;center&semi;"><img class&equals;"aligncenter" src&equals;"https&colon;&sol;&sol;budgyapp&period;com&sol;wp-content&sol;uploads&sol;2026&sol;05&sol;computational-design-of-quasisymmetric-protein-cage-for-gene-therapy-delivery&period;webp" alt&equals;"Computational design of quasisymmetric protein cage for gene therapy delivery&period;" style&equals;"width&colon;100&percnt;&semi;max-width&colon;800px&semi;height&colon;auto&semi;border-radius&colon;8px&semi;display&colon;block&semi;margin&colon;0 auto&semi;" &sol;><figcaption style&equals;"text-align&colon;center&semi;font-size&colon;0&period;85em&semi;color&colon;&num;888&semi;margin-top&colon;0&period;5em&semi;">Computational design of quasisymmetric protein cage for gene therapy delivery&period;<&sol;figcaption><&sol;figure>&NewLine;<h2>How RFdiffusion Cracked the Geometry<&sol;h2>&NewLine;<h3>The Linker That Bends the Lattice<&sol;h3>&NewLine;<p>The design relies on two proteins produced separately and combined at controlled ratios&period; The first is a trimeric component designated C3-A&comma; which forms the hexagonal faces of the growing lattice&period; The second is a dimeric linker designated C2-B&comma; which bridges adjacent C3-A trimers at each lattice edge&period; The cone angle of the C2-B linker determines the local curvature of the entire assembly&period;<&sol;p>&NewLine;<p>A shallow cone angle keeps adjacent trimers nearly coplanar&comma; favoring flat sheets that never close&period; A steeper angle introduces positive curvature&period; At the angle corresponding to T&equals;3 icosahedral geometry&comma; pentagonal defects appear spontaneously in the growing hexagonal lattice and force it to close&period; The team designed a family of C2-B variants&comma; labeled by their cone angle in degrees&period; The α20 variant produces a small cage analogous to a regular dodecahedron&period; The α30 variant&comma; with a steeper angle&comma; yields a mixture of larger topologies&period; The α25 variant gives the confirmed icosahedral structure that the study characterized in full structural detail&period;<&sol;p>&NewLine;<h3>Where AI Enters the Pipeline<&sol;h3>&NewLine;<p>Designing C2-B required holding two C3-A trimers in geometrically precise&comma; rigid register using a protein scaffold with no natural equivalent&period; The team used <a href&equals;'https&colon;&sol;&sol;github&period;com&sol;RosettaCommons&sol;RFdiffusion' target&equals;'&lowbar;blank' rel&equals;'noopener'>RFdiffusion&comma; the deep-learning diffusion model for protein backbone generation<&sol;a> developed at the same institute&comma; to generate candidate backbones satisfying those geometric constraints&period; Given fixed positions of two input motifs extracted from the previously designed LHD101 heterodimer pair&comma; RFdiffusion ran symmetric diffusion sampling to produce C2-B scaffolds de novo&comma; with no natural starting structure to copy&period;<&sol;p>&NewLine;<p>ProteinMPNN then assigned amino acid sequences to those backbones&period; AlphaFold2 predicted folded structures and filtered candidates that passed structure-prediction quality thresholds for experimental testing&period; The pipeline mirrors the general framework used across recent IPD projects&comma; applied here to a problem that required generating a protein capable of bridging two fixed geometric anchors at a tunable angle while self-assembling correctly at nanomolar concentrations&period;<&sol;p>&NewLine;<ul>&NewLine;<li><strong>Two proteins&comma;<&sol;strong> C3-A &lpar;trimeric&rpar; and C2-B &lpar;dimeric&rpar;&comma; produced separately and mixed to trigger cage assembly<&sol;li>&NewLine;<li><strong>T&equals;3 cage<&sol;strong> confirmed by cryo-electron microscopy &lpar;cryo-EM&comma; which fires electrons through frozen protein samples for near-atomic-resolution imaging&rpar;&comma; with coordinates deposited in the Protein Data Bank &lpar;PDB&comma; the global repository for macromolecule structures&rpar; as accession 9OM3<&sol;li>&NewLine;<li><strong>2-3x larger<&sol;strong> internal volume than prior strictly symmetric two-component cages from the same laboratory<&sol;li>&NewLine;<li><strong>Multiple cage sizes<&sol;strong> accessible from one C3-A component by substituting C2-B linker variants carrying different cone angles<&sol;li>&NewLine;<&sol;ul>&NewLine;<h2>Two Papers&comma; One Day&comma; Two Routes to the Same Sphere<&sol;h2>&NewLine;<p>The Wang et al&period; paper published simultaneously with a companion study&comma; <a href&equals;'https&colon;&sol;&sol;www&period;doi&period;org&sol;10&period;1038&sol;s41586-026-10554-z' target&equals;'&lowbar;blank' rel&equals;'noopener'>Design of One-Component Quasisymmetric Protein Nanocages<&sol;a>&comma; led by Sangmin Lee&comma; assistant professor of chemical engineering at Pohang University of Science and Technology &lpar;POSTECH&rpar; and a former postdoctoral researcher in Baker&&num;8217&semi;s group&period; Where the two-component approach controls curvature through a tunable bridging linker&comma; Lee&&num;8217&semi;s approach encodes the required curvature into a single protein subunit designed to undergo spontaneous symmetry breaking as it self-assembles from one genetic construct&period;<&sol;p>&NewLine;<table>&NewLine;<thead>&NewLine;<tr>&NewLine;<th>Feature<&sol;th>&NewLine;<th>Two-Component Cages &lpar;Wang et al&period;&rpar;<&sol;th>&NewLine;<th>One-Component Cages &lpar;Lee et al&period;&rpar;<&sol;th>&NewLine;<&sol;tr>&NewLine;<&sol;thead>&NewLine;<tbody>&NewLine;<tr>&NewLine;<td>Building blocks<&sol;td>&NewLine;<td>Two proteins&colon; C3-A trimer and C2-B dimer<&sol;td>&NewLine;<td>Single protein subunit<&sol;td>&NewLine;<&sol;tr>&NewLine;<tr>&NewLine;<td>Assembly mechanism<&sol;td>&NewLine;<td>Geometric frustration via tunable cone angle<&sol;td>&NewLine;<td>Spontaneous symmetry breaking from programmed curvature<&sol;td>&NewLine;<&sol;tr>&NewLine;<tr>&NewLine;<td>Assembly trigger<&sol;td>&NewLine;<td>Mixing separately produced components<&sol;td>&NewLine;<td>Self-assembly from one protein species<&sol;td>&NewLine;<&sol;tr>&NewLine;<tr>&NewLine;<td>Cargo loading method<&sol;td>&NewLine;<td>Controllable at the mixing step<&sol;td>&NewLine;<td>Interior volume accessible by design<&sol;td>&NewLine;<&sol;tr>&NewLine;<tr>&NewLine;<td>Key practical advantage<&sol;td>&NewLine;<td>On-demand cargo packaging during co-assembly<&sol;td>&NewLine;<td>Single gene encodes the full cage<&sol;td>&NewLine;<&sol;tr>&NewLine;<&sol;tbody>&NewLine;<&sol;table>&NewLine;<p>Together the two papers mark a categorical shift&period; Two-component cages with strict symmetry existed from the Baker lab since 2016&period; Quasisymmetric designs&comma; which capture the actual architectural logic of real viral capsids&comma; have now arrived in two independent forms on the same day&comma; from overlapping teams&comma; using different underlying principles&period; That convergence argues for a design capability that is reproducible engineering rather than a singular lucky result&period;<&sol;p>&NewLine;<h2>Electron Microscopy Validates the Cage Structures<&sol;h2>&NewLine;<p>Structural confirmation drew on three independent imaging modalities&period; During early design screening&comma; negative-stain electron microscopy &lpar;nsEM&rpar; revealed multiple cage populations&comma; including assemblies consistent with dodecahedral geometry&comma; the target icosahedral topology&comma; and potential larger analogs whose morphology tracks known classes of fullerene-like carbon structures&period; The appearance of distinct size classes when the linker angle was varied confirmed that cage size was genuinely tunable by linker choice rather than fixed by the C3-A trimer itself&period;<&sol;p>&NewLine;<p>The icosahedral cage structure was resolved by cryo-EM at near-atomic resolution and deposited in the <a href&equals;'https&colon;&sol;&sol;www&period;rcsb&period;org&sol;structure&sol;9OM3' target&equals;'&lowbar;blank' rel&equals;'noopener'>Protein Data Bank as accession 9OM3<&sol;a>&period; Separately&comma; cryo-electron tomography &lpar;cryo-ET&rpar; produced a three-dimensional reconstruction of the same cage under near-physiological conditions&comma; deposited as PDB 9OP9&period; The X-ray crystal structure of the isolated C2-B-α20 linker&comma; solved at the National Synchrotron Light Source II at Brookhaven National Laboratory&comma; was deposited as PDB 9NDL and showed the designed fold achieved with high geometric fidelity&period;<&sol;p>&NewLine;<p>Defect analysis of nsEM micrographs matched theoretical predictions closely&period; For cage topologies below a triangulation number of nine&comma; only pentagonal defects appeared&comma; consistent with positive Gaussian curvature throughout the shell&period; For larger topologies&comma; both pentagonal and heptagonal defects were visible&period; Heptagonal defects produce negative curvature and are energetically unfavorable&comma; but the authors&&num;8217&semi; analysis indicates they can form under kinetic control during closure of large cages&comma; precisely the behavior documented in carbon fullerene chemistry&period; The authors draw that comparison explicitly&comma; placing their protein cages in the same geometric family as buckminsterfullerene and its larger carbon cousins&period;<&sol;p>&NewLine;<h2>Cargo Loading and Cellular Uptake Tests<&sol;h2>&NewLine;<p>Structural characterization was accompanied by functional demonstrations designed to establish what the cages can carry and where they can go&period; Experiments documented across the study included&colon;<&sol;p>&NewLine;<ul>&NewLine;<li>Fusing superfolder green fluorescent protein &lpar;sfGFP&comma; an engineered GFP variant that retains fluorescence even when attached to large protein assemblies&rpar; to cage components for live-cell tracking of cage position and movement inside mammalian cells<&sol;li>&NewLine;<li>Covalent cargo attachment via HaloTag &lpar;a protein tag forming an irreversible bond with synthetic ligands&rpar; and SpyTag&sol;SpyCatcher &lpar;an isopeptide-bond-forming pair that permanently links protein modules&rpar;&comma; enabling modular decoration of both cage interior and exterior surfaces<&sol;li>&NewLine;<li>Cellular uptake experiments showing that functionalized cages are internalized by mammalian cells&comma; the minimum prerequisite for any intracellular delivery application<&sol;li>&NewLine;<li>Live-cell imaging in collaboration with Liam J&period; Holt&&num;8217&semi;s research group at NYU Langone Medical Center&&num;8217&semi;s Institute for Systems Genetics&comma; using the cages as calibrated probes to study cytoplasmic crowding&comma; intracellular diffusion&comma; and protein localization<&sol;li>&NewLine;<&sol;ul>&NewLine;<p>Extended data also demonstrated successful cage assembly when the building-block proteins were fused to a broad set of additional proteins of interest&comma; suggesting the modular surface-attachment strategy generalizes beyond the specific cargo combinations tested in the main experiments&period; The cage&&num;8217&semi;s two-component design offers a particular logistical advantage for cargo loading&colon; molecular passengers can be encapsulated during the mixing step that triggers assembly&comma; without requiring disassembly and reassembly of a pre-formed structure&period;<&sol;p>&NewLine;<h2>Immunogenicity and the Path to Clinical Use<&sol;h2>&NewLine;<p>David Baker&comma; director of the Institute for Protein Design at the University of Washington and a 2024 Nobel laureate in Chemistry&comma; described the combined significance of the two papers in an IPD statement accompanying the publications&colon;<&sol;p>&NewLine;<blockquote>&NewLine;<p>These papers show that protein design is beginning to capture some of the architectural principles that nature uses to build at very large scales&period;<&sol;p>&NewLine;<&sol;blockquote>&NewLine;<p>That framing also marks where the work stops&period; The cages are de novo proteins whose sequences did not exist in nature before computational design produced them&period; Whether a human immune system will treat them as benign carriers or mount an inflammatory response is not yet established&period; <strong>Immunogenicity testing<&sol;strong>&comma; the systematic measurement of immune responses a new protein antigen triggers&comma; is the next required step before any human research application becomes feasible&period; Additional pre-clinical studies are needed&comma; the authors note&comma; before the cages enter human research protocols&period;<&sol;p>&NewLine;<p>Research was funded by a consortium spanning the Defense Threat Reduction Agency&comma; the Bill and Melinda Gates Foundation&comma; the Howard Hughes Medical Institute&comma; and the NIH&&num;8217&semi;s National Institute on Aging&comma; among others&period; A gift from Microsoft supported part of the computational work&period; That funding portfolio&comma; spanning national security&comma; global infectious disease&comma; and basic cell biology&comma; reflects the multiple application areas researchers envision for programmable protein containers at this scale&period;<&sol;p>&NewLine;<p>If immunogenicity data show the cages are well-tolerated&comma; the size range now accessible through quasisymmetric design places genetic payloads previously too large for strictly symmetric carriers within reach of a single engineered vessel&period; If immune reactivity proves significant&comma; engineering around that response becomes the next multi-year problem&period; Both paths forward start with the same experiment&comma; and none of those results exist yet&period;<&sol;p>&NewLine;

Leave a Reply

Your email address will not be published. Required fields are marked *

Trending

Exit mobile version