Virus Structure and Symmetry

A virion is the complete, infectious viral particle: a nucleic acid genome enclosed in a protein coat called the capsid, sometimes further wrapped in a lipid envelope. The capsid is built from repeating protein subunits called protomers (individual folded protein chains) that assemble into larger repeating units called capsomeres, visible as the bumps on the surface of many virions in electron micrographs. The entire nucleic-acid-plus-capsid unit, with or without envelope, is called the nucleocapsid. The central design constraint is genetic economy. A viral genome is typically a few thousand to a few hundred thousand nucleotides — vastly smaller than a bacterial or eukaryotic genome — and only a small fraction of its coding capacity can be spent on structural proteins without leaving too little room for replication and regulatory genes. Instead of encoding thousands of different, uniquely positioned proteins to build a large shell (as would be needed if every position in the coat were structurally distinct), a virus encodes one or a few small proteins and uses them repeatedly, in a symmetric arrangement, to close up a container around the genome. This principle, first articulated by Francis Crick and James Watson in 1956, is called the genetic economy principle: symmetry allows a shell of arbitrary size to be built from a minimal, reusable genetic instruction set. Only two families of symmetry satisfy the physical requirement that identical subunits, bonding to their neighbors in identical ways, can close into a stable, finite structure: helical symmetry (subunits arranged in a spiral, forming a rod or filament) and icosahedral symmetry (subunits arranged on the faces of a 20-sided polyhedron, forming a roughly spherical shell). A minority of viruses combine or elaborate on these motifs into "complex" architectures — bacteriophage heads-and-tails, brick-shaped poxviruses — but even these complex forms are typically built from an icosahedral or helical sub-assembly plus additional appendages. Layered onto any of these capsid types, some viruses acquire a host-cell-derived lipid bilayer, the envelope, studded with viral glycoproteins, which they obtain by budding through a cellular membrane. Viruses that never acquire this layer are called naked or nonenveloped.

Step 1 — Helical symmetry: winding a spiral of subunits around a genome. In a helical capsid, identical protein subunits bind to one another and to the nucleic acid in a repeating, staggered fashion, forming a continuous spiral staircase. Each subunit occupies an identical position relative to its neighbors (a fixed rotation angle and rise per subunit), so the resulting rod or filament can be extended indefinitely simply by adding more subunits at the growing end. The nucleic acid genome is not a passive passenger inside a tube; it lies in a helical groove between successive turns of the protein spiral, its phosphate backbone often forming direct ionic and hydrogen-bonding contacts with basic residues on the inner face of the subunits. This has two mechanistic consequences worth memorizing:

• The length of the capsid is dictated by the length of the genome (for rigid, non-enveloped helical viruses like TMV), since a fixed amount of nucleic acid must fit into the spiral groove at fixed base-pairs (or bases)-per-subunit.
• The diameter of the capsid is dictated by the size and packing angle of the subunit, essentially independent of genome length.

Tobacco mosaic virus is the paradigm case: a single 158-amino-acid coat protein, 2,130 copies of it, wound around one molecule of (+)-sense single-stranded RNA (6,395 nucleotides) in a helix with 16⅓ subunits per turn, producing a rigid rod 300 nm long and 18 nm in diameter with a 4 nm central hollow channel. Many enveloped helical viruses (influenza, rabies, Ebola, measles) use the same helical nucleocapsid principle but the helix is flexible and coiled up inside a lipid envelope rather than forming a rigid exposed rod.

Step 2 — Icosahedral symmetry: closing a shell with the minimum possible subunits. For a genome to be enclosed rather than just coated, the protein shell must close into a finite, roughly spherical container. Geometrically, the most efficient closed shell built from identical subunits packed with identical contacts is an icosahedron: a solid with 20 triangular faces, 12 vertices, and 30 edges, possessing 2-fold, 3-fold, and 5-fold rotational symmetry axes. The minimum number of identical protein subunits that can form a closed icosahedral shell is 60 — one subunit per asymmetric unit, with each subunit occupying geometrically and chemically identical bonding environments related by the icosahedron's symmetry operations. Most viral genomes, however, are far too large to be enclosed by a 60-subunit shell built from a single small protein; a bigger container is needed without abandoning economical repetition. Donald Caspar and Aaron Klug solved this puzzle in 1962 with the theory of quasi-equivalence: capsids larger than the minimal 60-subunit shell are built from multiples of 60 subunits (60T, where T is the triangulation number), with the same protein adopting several slightly different local bonding conformations — "quasi-equivalent" rather than strictly identical environments — to accommodate the different curvatures required at different positions in the shell. Allowed T numbers follow the formula:

T = h² + hk + k², where h and k are non-negative integers (0, 1, 2, 3…) giving permitted values T = 1, 3, 4, 7, 9, 12, 13… Each triangular facet of the icosahedron is then subdivided into T smaller triangles, each built from three subunits, so the total number of protein subunits in the shell is 60T. Capsomeres visible by electron microscopy are clusters of these subunits: pentamers (5 subunits, called pentons) sit exactly at the 12 vertices of the icosahedron in every T number, while hexamers (6 subunits, called hexons) fill in the faces, with the number of hexons increasing as T increases. Poliovirus and many small "simple" icosahedral viruses have T = 1 or T = 3; adenovirus has T = 25 (240 hexons plus 12 pentons, using several distinct capsid proteins rather than one, illustrating that quasi-equivalence is a flexible design principle, not a rigid law); the very large herpesvirus capsid has T = 16.

Step 3 — Self-assembly, not construction. Neither helical nor icosahedral capsids are built by any cellular assembly-line machinery analogous to ribosomal translation. Instead, capsid subunits fold, diffuse, and bind to one another (and to the genome, in many cases) driven purely by the thermodynamics of the interacting surfaces — hydrophobic patches, salt bridges, and shape complementarity that make the fully assembled shell the lowest free-energy state available to the components. This is why purified capsid protein and purified genome, mixed together outside any cell under the right ionic conditions, will spontaneously reconstitute infectious virions (see Evidence, below). In many complex viruses, assembly additionally uses scaffolding proteins — proteins that template the correct curvature of the procapsid during construction and are then cleaved out and discarded (as in herpesviruses and tailed bacteriophages), a strategy conceptually similar to the wooden form used to pour a concrete arch and later removed.

Step 4 — Genome packaging. Positive-sense RNA plant and animal viruses (like TMV and many picornaviruses) often condense their genome as an integral part of assembly, with capsid protein binding cooperatively to a specific "packaging signal" sequence on the RNA to nucleate the process. Many dsDNA viruses (herpesviruses, tailed phages) instead build an empty procapsid first, then actively pump the genome in through a portal vertex using an ATP-powered terminase motor — one of the most powerful molecular motors known, generating forces of tens of piconewtons against the enormous internal pressure of tightly packed DNA (up to ~60 atmospheres inside a phage head).

Step 5 — Envelope acquisition. Enveloped viruses assemble a nucleocapsid (helical or icosahedral) in the cytoplasm or nucleus, then acquire their lipid envelope by budding: viral transmembrane glycoproteins are trafficked to a host membrane (plasma membrane, or in some cases the endoplasmic reticulum or Golgi), the nucleocapsid associates with the cytoplasmic tail of these glycoproteins (often via a matrix protein bridging the two), and the membrane curves around the nucleocapsid and pinches off, carrying away a patch of host lipid bilayer studded with viral spike glycoproteins (like influenza hemagglutinin and neuraminidase, or HIV gp120/gp41). Because the envelope is a lipid bilayer, it is destroyed by desiccation, organic solvents, detergents, and many disinfectants — the mechanistic reason enveloped viruses (HIV, influenza, coronaviruses) are generally less stable in the environment and more easily inactivated by soap than nonenveloped viruses (poliovirus, norovirus, adenovirus), whose rigid protein shells resist these agents far better.

Step 6 — Complex architectures. Some virions do not reduce cleanly to a single helix or icosahedron. Tailed bacteriophages (e.g., phage T4, lambda) have an icosahedral head fused to a separate helical tail apparatus with a contractile sheath, baseplate, and tail fibers used for host recognition and genome injection. Poxviruses (e.g., vaccinia, variola) dispense with strict capsid symmetry altogether, instead building a brick-shaped or ovoid particle with a complex, multilayered coat of lipoprotein membranes and a dumbbell-shaped core containing the genome and internal enzymes — an architecture too large and too irregular to be described by any single symmetry group.

The classic demonstration: Fraenkel-Conrat and Williams reconstitute TMV from purified parts (1955). Heinz Fraenkel-Conrat and Robley Williams asked a deceptively simple question: is the assembly of a virus particle directed by some hidden biological process, or is it a purely spontaneous, physicochemical event? They took purified tobacco mosaic virus and treated it with mild alkali and detergent (sodium dodecyl sulfate), which dissociated the virion completely into its two separated components: naked coat protein subunits and free genomic RNA, each purified away from the other and shown separately to be non-infectious on tobacco leaves. They then simply mixed the purified RNA and purified protein back together in a buffered saline solution, with no cells, no enzymes, and no energy source added. The mixture spontaneously reassembled into rod-shaped particles that were, by electron microscopy, essentially indistinguishable from native TMV virions in size and shape — and, critically, these reconstituted particles were infectious, producing the characteristic mosaic lesions when rubbed onto tobacco leaves. Because the reconstitution occurred in a cell-free tube containing only two purified macromolecular species, the experiment proved that all the information needed to build a correctly shaped, infectious capsid is contained in the physical and chemical properties of the coat protein and RNA themselves — assembly is a thermodynamically favorable self-assembly process, not something requiring a cellular assembly machine, chaperone system, or template beyond the subunits' own shape complementarity. A further, elegant twist came when Fraenkel-Conrat mixed the RNA of one TMV strain with the coat protein of a different strain: the resulting hybrid particles had the protein coat of the donor strain, but the disease symptoms and progeny virus produced on infection matched the strain that had donated the RNA — independently confirming (alongside Hershey-Chase-type logic) that the nucleic acid, not the protein, carries hereditary information, while showing that the protein's role is purely structural and self-directed in assembly.

Corroborating structural evidence: X-ray diffraction and electron microscopy establishing icosahedral symmetry. Rosalind Franklin, in the years just before her death in 1958, used X-ray fiber diffraction to work out that TMV's RNA lies in a helical groove roughly 4 nm from the central axis of the rod, at a radius consistent with a single strand rather than a double helix — direct physical evidence for the helical architecture described above, obtained by essentially the same diffraction method she had earlier applied to DNA. For icosahedral viruses, Donald Caspar's 1956 X-ray diffraction studies of tomato bushy stunt virus first revealed icosahedral symmetry in a plant virus, and Caspar and Klug's 1962 theoretical paper ("Physical Principles in the Construction of Regular Viruses") used these diffraction patterns together with geometric reasoning to derive the quasi-equivalence theory and the T-number formula described above — a rare case in biology where a mathematical, first-principles argument (how many ways can identical subunits close a shell?) predicted structures that electron microscopy and, decades later, cryo-electron microscopy and X-ray crystallography of viruses like poliovirus, adenovirus, and bacteriophage HK97 subsequently confirmed atom-by-atom.

Modern confirmation: cryo-electron microscopy. Contemporary cryo-EM, which flash-freezes virus particles in vitreous ice and reconstructs three-dimensional structures from thousands of individual particle images without staining or fixation artifacts, has since resolved capsid structures to near-atomic (better than 3 Å) resolution for dozens of virus families, directly visualizing individual quasi-equivalent subunit conformations, portal proteins, and genome-packaging motors, and confirming in atomic detail the icosahedral and helical symmetries that Franklin, Caspar, and Klug inferred indirectly from diffraction data alone.

Example 1 — Predicting rod length from genome size in a helical virus. Tobacco mosaic virus packages 6,395 nucleotides of RNA in a helix with 16⅓ subunits per turn, a rise per turn (pitch) of 2.3 nm, and 3 nucleotides bound per protein subunit. Suppose a related helical virus in the same structural family has a genome of 9,600 nucleotides, uses the same coat protein (same subunits-per-turn and same 3 nucleotides bound per subunit), and the same 2.3 nm rise per turn. How long would you predict its rod to be?

• Subunits needed = 9,600 nt ÷ 3 nt/subunit = 3,200 subunits.
• Turns needed = 3,200 subunits ÷ 16.33 subunits/turn ≈ 196 turns.
• Predicted length = 196 turns × 2.3 nm/turn ≈ 451 nm.

Check against TMV itself: 6,395 ÷ 3 = 2,132 subunits (matches the observed ≈2,130); 2,132 ÷ 16.33 ≈ 130.5 turns; 130.5 × 2.3 nm ≈ 300 nm — matching TMV's actual measured length of 300 nm almost exactly. This confirms the mechanistic rule: for a rigid helical virus, particle length scales linearly and predictably with genome length, because each subunit occupies a fixed length of genome and a fixed increment of helical rise. This is precisely why helical viruses with larger genomes (like some filamentous plant viruses over 1 μm long) are visibly longer rods than TMV under the electron microscope, without any need to invoke a different assembly mechanism.

Example 2 — Reading a T-number and predicting subunit count and capsomere layout. A newly discovered icosahedral virus is imaged by cryo-EM and its capsid surface lattice shows h = 2, k = 1 in the Caspar-Klug lattice. A virologist needs to predict: (a) the triangulation number, (b) the total number of capsid protein subunits, and (c) the number of hexon versus penton capsomeres, to compare it against known viruses.

• T = h² + hk + k² = 2² + (2×1) + 1² = 4 + 2 + 1 = 7.
• Total subunits = 60T = 60 × 7 = 420 subunits.
• Number of pentons = always 12 (one at each of the icosahedron's 12 vertices), regardless of T.
• Number of hexons = (10T − 2) for a T-number lattice minus the 12 pentonal positions already accounted for within that count; concretely, total capsomeres = 10T + 2, of which 12 are pentons and the remainder are hexons: 10(7) + 2 = 72 capsomeres total, 12 pentons + 60 hexons.

A virologist recognizing "T = 7, 420 subunits, 60 hexons + 12 pentons" would immediately place this hypothetical virus in the same structural class as bacteriophage HK97 and the hepatitis B virus core (both genuine T = 7 capsids), predicting a mid-sized icosahedral shell roughly 55–60 nm in diameter — larger than a T = 1 or T = 3 picornavirus (which cannot accommodate as much genetic material) but far smaller and simpler than a T = 25 adenovirus or T = 16 herpesvirus capsid, illustrating how the single number T lets a structural biologist immediately estimate genome-packaging capacity from capsid geometry alone.

Aaron Klug was a South African-born, Lithuanian-descended crystallographer who trained partly in Rosalind Franklin's laboratory group at Birkbeck College, London, after Franklin's death in 1958, inheriting and extending her X-ray diffraction studies of tobacco mosaic virus. Working with the American biophysicist Donald Caspar, Klug combined X-ray diffraction data from several plant and animal viruses with rigorous geometric and group-theoretic reasoning to ask a purely mathematical question with enormous biological payoff: given that identical protein subunits must bond to their neighbors in essentially identical ways, what closed shells are geometrically possible, and how can a shell be scaled up without abandoning that identical-subunit economy? Their answer — quasi-equivalence and the T-number rule — turned viral architecture from a descriptive, case-by-case cataloguing exercise into a predictive, quantitative science, allowing structural biologists to predict subunit counts and capsomere arrangements for viruses whose atomic structures had not yet been solved. Klug went on to develop crystallographic and electron-microscopic image reconstruction techniques more broadly, extending them to chromatin structure and transfer RNA, and was awarded the Nobel Prize in Chemistry in 1982 "for his development of crystallographic electron microscopy and his structural elucidation of biologically important nucleic acid-protein complexes."

Vaccine design. Understanding capsid self-assembly directly enabled the recombinant, virus-like particle (VLP) vaccines now in routine clinical use: the hepatitis B vaccine and the HPV vaccines (Gardasil, Cervarix) are made by expressing a single viral structural protein in yeast or insect cells, letting it self-assemble into empty, noninfectious icosahedral shells that mimic the real virus's surface closely enough to elicit protective neutralizing antibodies without any viral genome present at all — a direct, deliberate exploitation of the same self-assembly logic Fraenkel-Conrat first demonstrated with TMV.

Antiviral drug design and disinfection policy. Knowing whether a pathogen is enveloped or naked directly determines rational infection-control strategy: alcohol-based hand sanitizers and soap reliably inactivate enveloped viruses (influenza, HIV, coronaviruses) by dissolving the lipid envelope, but nonenveloped viruses like norovirus and poliovirus require bleach-based or oxidizing disinfectants because their rigid icosahedral protein shells resist detergents.

Gene therapy and nanotechnology vectors. Adeno-associated virus (AAV) and lentiviral capsids are engineered by structural biologists exploiting the same T-number and quasi-equivalence principles to redesign capsid surface loops, retargeting which host-cell receptors the shell engages while preserving the underlying icosahedral scaffold's stability — the basis of modern gene-therapy vector engineering (e.g., AAV-based therapies for inherited retinal and neuromuscular disease).

Agriculture and plant pathology. Plant virologists use the predictable relationship between helical capsid length and genome size (Worked Example 1) as a practical diagnostic tool: electron-microscopic measurement of a newly isolated filamentous plant virus's particle length gives an immediate, testable estimate of its genome size before any sequencing is performed, guiding crop-disease surveillance in agriculture.

Structural virology and pandemic preparedness. Cryo-EM structural determination of capsid and spike architecture — built on the same symmetry principles established by Caspar and Klug — was central to the rapid structural characterization of SARS-CoV-2 in 2020, informing both vaccine antigen design (stabilized spike constructs) and structure-based antiviral drug development.

Ask the tutor: How does capsid symmetry class (helical vs. icosahedral vs. complex) constrain or predict which cell-entry strategy a virus can use?