The Nature of Viruses

A virus consists, at minimum, of two components: a genome (a single molecule or a small number of molecules of DNA or RNA, never both in the same virus) and a capsid, a shell built from many copies of one or a few types of protein subunits called capsomeres. Together, genome plus capsid form the nucleocapsid. Some viruses add a third layer, a lipid envelope derived from a host membrane and studded with viral glycoproteins, acquired as the virus buds out of the cell. The complete infectious particle — whatever combination of these parts it has — is called a virion. Crucially, a virion is not a cell. It has no cytoplasm, no ribosomes, no ATP synthase, no independent metabolic pathways, and in the vast majority of cases no membrane-bound compartments of its own (the envelope is stolen, not synthesized by the virus). This is why viruses are described as acellular: they are not simply very small or very simple cells, they are categorically different objects that lack the machinery cell theory requires. Whereas the smallest free-living bacteria (such as Mycoplasma, at roughly 200–300 nm) still carry a full complement of ribosomes and enzymes for translation and energy metabolism, even a "large" virus like a poxvirus (up to ~300 nm) or a giant Mimivirus (up to 750 nm) carries none of a bacterium's biosynthetic autonomy — it still must enter a cell and hijack ribosomes to make protein. This dependency defines the term obligate intracellular parasite: obligate because there is no alternative reproductive strategy, and no free-living stage; intracellular because the parasitism happens inside, not merely attached to, a host cell; parasite because the relationship benefits the virus at the host's expense. This single property — total metabolic dependence — is the thread that connects viral structure (Virus Structure and Symmetry), the diversity of genome types and replication strategies (The Baltimore Classification System), and the practical difficulty of treating viral disease: because a virus uses host enzymes for so much of its life cycle, drugs that block viral replication often risk blocking host processes too.

Because a virion carries no functional metabolism, its "life cycle" is really a host-cell takeover cycle. The steps are conserved in outline across essentially all viruses, though the molecular details differ enormously between, say, a bacteriophage and a human retrovirus.

1. Attachment (adsorption). A protein on the virion surface — a capsid protein spike, or a glycoprotein embedded in the envelope — binds specifically to a receptor molecule on the host cell surface. This receptor is normally there to do something else for the cell (e.g., CD4 on human T-helper cells normally binds MHC class II during immune signaling; HIV's gp120 protein exploits it). Receptor specificity is the molecular basis of host range and tissue tropism: a virus can only infect cells that display the right receptor.

2. Penetration (entry). The virion, or just its genome, crosses the host plasma membrane. Bacteriophages typically inject naked DNA through the cell wall/membrane, leaving the empty capsid outside — this is the very fact the Hershey-Chase experiment exploited. Enveloped animal viruses commonly fuse their envelope directly with the host membrane, or enter by receptor-mediated endocytosis and then fuse with the endosomal membrane once triggered by low pH.

3. Uncoating. The capsid is degraded or opened, releasing the naked genome into the host cytoplasm or nucleus, where it becomes accessible to enzymes.

4. Genome replication and gene expression. This is the step where genome chemistry (DNA vs. RNA, single- vs. double-stranded) dictates strategy, because the host cell only natively provides machinery for DNA → mRNA → protein (the standard direction of the Central Dogma of Molecular Biology). A DNA virus can often simply present its genome to host RNA polymerase. An RNA virus, by contrast, must supply its own RNA-dependent RNA polymerase (host cells have none), either packaged in the virion or made anew from viral mRNA. Retroviruses go one step further, carrying reverse transcriptase to convert their RNA genome into DNA, which then integrates into the host chromosome. This diversity of strategies is systematized later in The Baltimore Classification System.

5. Assembly. Newly synthesized capsid proteins self-assemble around newly replicated genome copies — a largely spontaneous, thermodynamically favorable process driven by protein-protein and protein-nucleic acid interactions, requiring no external enzyme in many simple viruses (demonstrated dramatically for tobacco mosaic virus, see below).

6. Release. Progeny virions exit by lysing the host cell (common in phages and many "naked," non-enveloped viruses) or by budding through the plasma membrane, acquiring an envelope in the process (common in enveloped animal viruses such as influenza and HIV), often without immediately killing the cell. A single infected cell can release anywhere from dozens to tens of thousands of progeny virions, depending on the virus and cell type — a burst size that makes exponential epidemic spread possible from an initially tiny inoculum.

Filterable agents (Ivanovsky, 1892; Beijerinck, 1898). Adolf Mayer showed in the 1880s that tobacco mosaic disease could be transmitted by rubbing sap from a diseased plant onto a healthy one, implicating an infectious agent. Dmitri Ivanovsky in 1892 passed the infectious sap through a Chamberland filter — a porcelain filter with pores fine enough to trap all known bacteria — and found the filtrate remained infectious. Ivanovsky initially assumed this meant a bacterial toxin or an unusually small bacterium had slipped through. It was Martinus Beijerinck, repeating and extending the experiment in 1898, who drew the radical conclusion: the agent was not merely a small microbe but a fundamentally different kind of infectious entity, which he called a contagium vivum fluidum ("contagious living fluid") — later shortened to virus (Latin for "poison"). Beijerinck also showed the agent multiplied only in actively growing plant tissue, and could not be cultured on any nutrient medium — an early clue to obligate intracellular parasitism.

Crystallization (Wendell Stanley, 1935). For decades it remained unclear whether the tobacco mosaic agent was even a coherent chemical entity. Stanley purified tobacco mosaic virus (TMV) from infected leaves and, extraordinarily, obtained it in crystalline form — a property previously thought exclusive to non-living chemical compounds. The crystals, when redissolved and rubbed onto tobacco leaves, caused the disease all over again, proving the crystallized material retained full infectivity. This forced biologists to confront the paradox directly: something that could be crystallized like a simple salt could also reproduce like a living organism. Stanley shared the 1946 Nobel Prize in Chemistry for this work. Later refinement by Frederick Bawden and Norman Pirie showed TMV crystals were actually a nucleoprotein (protein plus RNA, not pure protein as Stanley first thought) — an important correction that pointed toward nucleic acid, not protein, as the likely hereditary material even before DNA's role was established in cellular organisms.

The Hershey-Chase blender experiment (1952). By the early 1950s biologists knew phage particles were made of DNA and protein, but not which of the two carried the genetic instructions for making new phage. Alfred Hershey and Martha Chase used bacteriophage T2, which infects E. coli, and exploited a simple chemical fact: DNA contains phosphorus but essentially no sulfur, while proteins contain sulfur (in cysteine and methionine) but no phosphorus. They grew separate batches of T2 phage in media containing either:

Each labeled phage stock was allowed to infect fresh E. coli, and the mixtures were then agitated violently in a kitchen blender to shear off any phage material remaining attached to the outside of the bacterial cells, followed by centrifugation to separate the (heavier) bacterial cell pellet from the (lighter) free phage debris in the supernatant. The result was unambiguous:

• Bacteria infected with ³²P-labeled phage retained most of the radioactive phosphorus (DNA had entered the cells).
• Bacteria infected with ³⁵S-labeled phage retained almost none of the radioactive sulfur (the protein coats stayed outside, blended free).
• Progeny phage produced from the infected cells contained newly synthesized ³²P-labeled DNA but essentially no ³⁵S from the parental protein coats.

This proved that DNA, not protein, is the genetic material injected by the phage and responsible for directing synthesis of new phage particles — the empty protein "ghost" capsids left outside contributed nothing to heredity. Combined with the contemporaneous work of Rosalind Franklin, Watson and Crick, and Avery-MacLeod-McCarty's transformation experiments in pneumococcus, Hershey-Chase (for which Hershey shared the 1969 Nobel Prize) cemented DNA's role as the universal hereditary molecule and simultaneously revealed the essential logic of viral infection: the capsid's job is delivery, the genome's job is instruction.

Example 1 — Estimating burst size and epidemic potential from plaque data. A student infects a lawn of E. coli with a diluted stock of bacteriophage T4. She plates 0.1 mL of a phage dilution of 10⁻⁶ from an original stock and counts 47 plaques (each plaque marks one initial infection event that lysed a patch of bacteria). She separately measures that the initial burst produced, on average, an increase from 2.0 × 10⁵ infected cells to 1.9 × 10⁷ free phage particles after one 25-minute lytic cycle. Step 1: Calculate the original phage titer. Plaque-forming units (PFU) per mL = (number of plaques) / (volume plated × dilution factor) = 47 / (0.1 mL × 10⁻⁶) = 47 / (10⁻⁷ mL) = 4.7 × 10⁸ PFU/mL Step 2: Calculate burst size. Burst size = progeny phage / infected cells = (1.9 × 10⁷) / (2.0 × 10⁵) = 95 phage per infected cell Step 3: Extrapolate exponential growth. If each of those 95 progeny phage goes on to infect a new cell and produce another 95 progeny, after just 3 cycles (~75 minutes) one initial phage could in principle yield 95³ ≈ 8.6 × 10⁵ phage — illustrating why viral titers in an unchecked bacterial culture, or in an unchecked human infection, can rise by many orders of magnitude within hours, far outpacing typical bacterial or host-cell division times (E. coli divides roughly every 20–30 minutes, comparable to a single phage lytic cycle, but phage burst size multiplies numbers far faster than binary fission).

Example 2 — Diagnostic reasoning: is the agent a virus or a bacterium? A public health lab receives a sample from a patient with severe gastroenteritis. Three observations are reported: 1. The agent passes through a 0.22 µm filter (standard bacteria-retaining filter) but is retained by a 0.02 µm ultrafilter. 2. It fails to grow on any standard bacteriological agar medium, even nutrient-rich blood agar, over 5 days of incubation. 3. Electron microscopy reveals particles roughly 70–90 nm in diameter with an icosahedral capsid and no visible cell wall, ribosomes, or internal organelles. Reasoning: Observation (1) places the agent's size between about 20 and 220 nm — consistent with a virus (typical range 20–300 nm) and too small for essentially all bacteria (typically 0.5–5 µm, i.e., 500–5000 nm), immediately ruling out a bacterial cause. Observation (2) is diagnostic of an obligate intracellular parasite: unlike bacteria, which can be cultured axenically (alone) on nutrient media because they carry their own metabolic machinery, a virus requires living host cells to replicate and will show zero growth on any acellular medium no matter how rich. Observation (3) confirms the absence of cellular structures. Taken together, these three independent lines of evidence converge on the conclusion: this is a viral pathogen, most consistent in size and morphology with a norovirus or rotavirus-type agent, not a bacterium — and the correct next step is inoculation of a susceptible cell culture line (or, historically, an animal model) rather than further bacteriological subculturing.

Martinus Willem Beijerinck was a Dutch microbiologist working at the Delft Polytechnic in the 1890s, a contemporary and rival of Sergei Winogradsky, both pioneers of what would become environmental and general microbiology. Beijerinck was known for a rigorous, almost obsessive insistence on experimental control and for the enrichment culture technique, still used today to isolate microbes with specific metabolic capabilities from environmental samples. He came to the tobacco mosaic problem somewhat by accident, repeating Ivanovsky's filtration experiments independently (apparently unaware of Ivanovsky's 1892 publication) and pushing the logic further than Ivanovsky had: rather than assuming a very small bacterium had passed the filter, Beijerinck showed the agent diffused through agar gel like a fluid, could not be cultured on any medium, and multiplied only when tissue was actively growing and dividing — hallmarks of an entity fundamentally unlike a bacterial cell. His willingness to name a genuinely new category of biological entity, the contagium vivum fluidum, rather than force the observation into the existing framework of bacteriology, is what makes 1898 the conventional founding date of virology as a discipline distinct from bacteriology.

Understanding viruses as obligate intracellular parasites directly shapes modern medicine and biotechnology. Antiviral drug design exploits the few virus-specific enzymes that differ from host enzymes — reverse transcriptase inhibitors and protease inhibitors for HIV, neuraminidase inhibitors (oseltamivir) for influenza, and RNA-dependent RNA polymerase inhibitors (remdesivir) for coronaviruses — precisely because these enzymes exist only in the virus's replication cycle, not in uninfected human cells. Vaccine design works by presenting the immune system with capsid or envelope proteins (or attenuated/inactivated whole virions, or, more recently, mRNA that directs host cells to make a single viral protein) so that adaptive immunity can recognize and neutralize the real pathogen on first exposure, discussed further under Adaptive Immunity and Vaccination. Bacteriophage therapy turns the parasitism around, using phages that specifically lyse pathogenic bacteria (including multidrug-resistant strains) as a targeted alternative or supplement to antibiotics, an approach now in clinical trials for infections unresponsive to conventional treatment. In agriculture, understanding filterable, obligate-parasitic plant viruses (the same tobacco mosaic virus system that founded the field) underlies crop-screening and quarantine programs that prevent viral epidemics in staple crops. In biotechnology, disabled viral vectors (adenovirus, lentivirus, adeno-associated virus) are engineered to deliver therapeutic genes into human cells for gene therapy, exploiting the very entry and genome-delivery machinery this article describes. And in conservation and ecology, marine phage dynamics drive the "viral shunt," recycling an estimated 20% or more of photosynthetically fixed carbon in ocean surface waters back into the microbial loop every day, making viruses a major invisible engine of global biogeochemical cycles.

Ask the tutor: How does the Baltimore Classification System use a virus's genome type to predict its entire replication strategy?