Bacteriophages and Host Recognition

A bacteriophage is a virus whose only natural hosts are prokaryotes — bacteria or archaea. Like all viruses, a phage is an obligate intracellular parasite: a protein (sometimes lipid-containing) shell called a capsid enclosing a genome of DNA or RNA, single- or double-stranded, that carries no independent metabolism and cannot reproduce outside a living cell. Phages show enormous morphological diversity, but the classical, still-dominant morphotype pictured in every textbook is the tailed phage: an icosahedral (or prolate) head storing double-stranded DNA, connected through a portal vertex to a tail that ends in a receptor-binding apparatus. Other major groups include cubic (polyhedral, tailless) phages, filamentous phages, and pleomorphic or spindle-shaped phages typical of archaeal viruses; these differences in capsid architecture were introduced under Virus Structure and Symmetry and recur here as the chassis on which host-recognition machinery is mounted. The concept this article is built around is adsorption: the process by which a phage particle attaches to a susceptible bacterial cell prior to genome delivery. Adsorption is receptor-mediated, meaning it depends on a specific molecular complementarity between a viral attachment protein and a defined chemical structure on the host envelope — not a generic "stickiness." The bacterial molecule involved is called the phage receptor, and it is virtually always a molecule the bacterium needs for its own physiology (a nutrient transporter, a component of the cell wall, a flagellum, a pilus) rather than something evolved for the virus's benefit. This is a crucial conceptual point: the bacterium cannot simply discard the receptor, because doing so costs fitness, which is why receptor loss is only one of several imperfect resistance strategies bacteria use against phage predation.

Host range — the set of bacterial species or strains a given phage can infect — is determined primarily, though not exclusively, by receptor compatibility at the adsorption step, together with subsequent intracellular barriers such as restriction-modification systems and CRISPR-Cas immunity (topics that belong properly to The Lytic Cycle and The Lysogenic Cycle). Because receptors vary chemically between species, and even between strains of one species (different capsular polysaccharides, different O-antigen sugars on lipopolysaccharide), phage host ranges are often narrow — frequently limited to one species or even a handful of strains within a species. This narrowness is both the frustration and the promise of phage biology: frustrating for anyone trying to find a single phage that treats an infection regardless of strain, promising for anyone who wants a therapeutic that leaves the rest of the microbiome untouched.

Anatomy of the tailed phage

The tailed phage is built from a small number of structural modules, each with a distinct job. The head (capsid) is an icosahedral protein shell, typically 50–200 nm across, that packages the double-stranded DNA genome under enormous internal pressure — measured directly by atomic force microscopy to reach tens of atmospheres, comparable to a car tire, in phages such as λ and T7. The portal is a dodecameric protein ring at one capsid vertex through which DNA is pumped in during assembly and later exits during infection. The tail connects to the portal; in Myoviridae-type phages (e.g., T4) it is long, rigid, and contractile, sheathed around an inner tube; in Siphoviridae-type phages (e.g., λ) it is long and flexible, non-contractile; in Podoviridae-type phages (e.g., T7) it is short, almost stub-like. At the distal tip sits the baseplate, a multiprotein complex bearing tail fibers and/or tail spikes — the actual receptor-recognition organelles. T4, the best-studied tailed phage, illustrates the division of labor concretely. Its baseplate carries six long tail fibers (built largely from the gene product 37 protein) that make the first, reversible contact with the host, and six short tail fibers that, once triggered, irreversibly lock onto the cell surface. The long tail fibers behave almost like antennae, scanning for the receptor lipopolysaccharide and outer membrane protein OmpC on Escherichia coli. Binding of the long fibers signals the baseplate to change conformation from a dome shape to a star shape, which in turn triggers contraction of the outer sheath, driving the inner tube through the outer membrane like a molecular syringe.

Step-by-step adsorption

Step 1 — Diffusion and random collision. Phage particles have no active motility; they move by Brownian diffusion in the surrounding fluid. Collision with a bacterium is therefore a matter of chance, governed by phage concentration, bacterial concentration, and diffusion coefficients. The rate of successful collisions can be described by a second-order rate law: dP/dt = −k · P · B where P is the free phage concentration, B is the bacterial concentration, and k is the adsorption rate constant (units of mL·min⁻¹, since concentrations are per mL). Measured values of k typically fall in the range of 10⁻⁹ to 10⁻⁸ mL/min per bacterium for well-adapted phage–host pairs, meaning a phage in a dense bacterial culture (~10⁸ cells/mL) adsorbs with a half-time of only a few minutes.

Step 2 — Reversible attachment. Initial contact between a tail fiber and a receptor is weak and reversible — the phage can detach and continue diffusing if the encounter is unproductive or if it strikes a non-permissive site. This stage is analogous to the initial, low-affinity "rolling" contact in other receptor-ligand systems: it allows the phage to sample many surface sites without committing.

Step 3 — Irreversible attachment. If the correct receptor is engaged with the correct geometry, a conformational change locks the tail structure onto the cell surface. In T4, this is the long-fiber-triggered baseplate reorganization described above; in many Siphoviridae, it is a conformational cascade in the tail tip complex (e.g., the gp9/gpJ machinery of λ, which engages the maltose transporter LamB). Irreversible attachment is essentially a point of no return: detachment now would mean the phage's genome-delivery apparatus has already begun to act.

Step 4 — Genome ejection. The irreversibly bound tail creates or opens a channel across the outer membrane, peptidoglycan layer, and inner membrane. Internal capsid pressure, combined in some phages with host-cell metabolic energy and phage or host proteins pulling DNA through the channel, drives the linear (or circular, depending on the phage) genome out of the capsid and into the host cytoplasm, while the empty protein coat remains outside. This "genome-in, coat-out" separation is the single most important physical fact about tailed-phage infection, and it is precisely what the Hershey–Chase experiment (below) was designed to reveal.

Receptors: the chemistry of specificity

The receptor determines everything about host range, and different phages have evolved to read entirely different molecular "words" on the bacterial surface:

Because many of these molecules are essential for nutrient uptake (LamB imports maltose), motility (flagella), or conjugation (pili), the bacterium faces a genuine cost-benefit tradeoff in mutating them away to escape phage predation — a tension central to the coevolutionary arms race between phages and bacteria, and a preview of the resistance mechanisms discussed alongside The Lysogenic Cycle.

From binding to commitment: a kinetic view

Adsorption's two-phase character (reversible then irreversible) can be visualized as a dose-response-like curve of infectivity over time. Immediately after mixing phage and bacteria, the fraction of free (unadsorbed) phage falls exponentially, and the shape of that decay — plotted as the classic "adsorption curve" — is one of the first quantities measured for any newly isolated phage, because the adsorption rate constant k is a defining biological parameter almost as fundamental as burst size or latent period (both covered under The Lytic Cycle).

Twort, d'Herelle, and the discovery of a particulate agent

The existence of bacteriophages was demonstrated independently by Frederick Twort (1915) and Félix d'Herelle (1917). Twort observed that colonies of micrococci on agar sometimes developed transparent, "glassy" degenerating patches that could be transferred to fresh colonies and would keep spreading, passing through bacteria-retaining filters — the classic operational definition of a filterable agent, the same criterion Beijerinck and Ivanovsky had earlier used to define plant viruses. D'Herelle, working independently on dysentery-causing Shigella, showed that filtrates from recovering patients' stool could clear turbid bacterial cultures, producing sharply defined circular clearings — plaques — when spotted onto a lawn of bacteria on solid agar. Critically, d'Herelle showed that the agent multiplied: a small amount added to a bacterial culture yielded far more infectious material afterward, proof that it was a self-replicating biological entity and not a static chemical toxin. He coined the term "bacteriophage" (bacteria-eater) and correctly proposed that these agents attacked bacteria with strict specificity — some filtrates cleared one Shigella strain but not another, an early empirical hint of the receptor specificity later explained mechanistically.

The one-step growth experiment: Ellis and Delbrück (1939)

Max Delbrück and Emory Ellis designed the one-step growth curve experiment specifically to quantify the life cycle of a phage as a population-synchronized process, and adsorption was the experimental lever they used to start the clock. They mixed phage with a large excess of susceptible bacteria for a short period, allowing essentially all free phage particles to adsorb; they then diluted the mixture enormously, so that any newly released phage would be too dilute to immediately reinfect a fresh cell. This trick — synchronizing infection by adsorption, then preventing a second round by dilution — allowed them to plot the number of infectious phage particles (measured as plaque-forming units) over time and observe a latent period with no free phage, followed by a sudden burst releasing many new virions per infected cell (the burst size). This experiment established that phage reproduction is not gradual budding but an all-or-nothing intracellular event, and it depended entirely on adsorption being fast, essentially complete, and treatable as a discrete "time zero" — direct experimental proof that adsorption is a genuine, measurable, rate-limited biological step rather than an instantaneous or negligible one.

The Hershey–Chase blender experiment (1952)

Alfred Hershey and Martha Chase used the physical separability of adsorbed phage — an intact protein coat sitting outside the cell, with genetic material delivered inside — to answer a question left open since Avery, MacLeod, and McCarty's 1944 transformation work: is DNA or protein the hereditary material? They labeled two separate batches of T2 phage: one with radioactive phosphorus (³²P), which is present in DNA's phosphate backbone but essentially absent from protein, and another with radioactive sulfur (³⁵S), present in the sulfur-containing amino acids of protein but absent from DNA. Each labeled phage stock was allowed to adsorb to E. coli for a few minutes — a step that depended entirely on the adsorption mechanism described above, using T2's tail fibers to bind the bacterial surface. The mixture was then agitated violently in a Waring blender, a shearing force sufficient to strip the empty phage coats off the bacterial surface without lysing the cells or disrupting the intracellular infection already underway. The results were unambiguous:

Centrifugation separated the blended mixture into a bacterial pellet and a supernatant containing detached phage "ghosts." The ³²P-labeled DNA stayed largely with the bacteria, while the ³⁵S-labeled protein was mostly released into the supernatant with the coats. Furthermore, infected bacteria still produced a normal burst of progeny phage even after the empty protein coats were sheared away, proving that the DNA alone carried the information sufficient to redirect the cell and produce new phage. This proved that DNA, not protein, is the hereditary material — and it worked as an experiment only because adsorption physically and mechanically partitions the phage into an external component (the coat) and an internal one (the genome), exactly as described in the mechanism above.

Modern structural confirmation

Cryo-electron microscopy and X-ray crystallography since the 1990s and 2000s have directly visualized baseplate conformational changes (dome-to-star transitions in T4), the atomic structure of the λ tail-tip complex bound to LamB, and single-molecule force measurements of DNA ejection, converting the classical genetic and biochemical inferences above into structurally resolved mechanisms — a case study in how molecular biology confirms and refines, rather than overturns, foundational 20th-century experiments.

Example 1 — Predicting host range from receptor genotype. A clinical microbiology lab isolates a novel lytic phage, "φ-K7," from hospital wastewater and wants to know its likely host range before testing it against a panel of 40 Klebsiella pneumoniae clinical isolates. Structural analysis shows φ-K7's tail carries a large tail-spike protein with depolymerase activity, similar to phages known to degrade specific capsular polysaccharide (K-antigen) types. Reasoning: Because the receptor is a capsular polysaccharide rather than a conserved outer-membrane protein, host range is expected to track capsular (K) serotype rather than species membership as a whole. K. pneumoniae strains are classified into more than 80 distinct K-types based on chemically distinct capsule structures. A depolymerase-tailed phage typically recognizes only one or a few closely related K-types, because the depolymerase active site must fit a specific sugar linkage pattern. Prediction and expected result: Among 40 clinical isolates spanning many K-types, we would predict φ-K7 to productively infect (plaque-form on) only the small subset sharing the matching capsular chemistry — commonly under 20% of a serotype-diverse panel — even though all 40 isolates are the same bacterial species. This is exactly the pattern typically observed for capsule-specific phages, and it is the central practical challenge in phage therapy: a single phage isolate is rarely a "one-size-fits-all" antibacterial, and clinical use generally requires a cocktail of phages with complementary receptor specificities, or a preliminary host-range assay against the patient's specific isolate.

Example 2 — Interpreting an adsorption-rate experiment quantitatively. A student mixes phage T4 (initial titer 2 × 10⁶ PFU/mL) with a growing culture of E. coli at a concentration of 4 × 10⁸ cells/mL. Samples are removed at intervals, bacteria and adsorbed phage are removed by low-speed centrifugation, and the supernatant (containing only free, unadsorbed phage) is titered by the plaque assay. The data:

Step 1 — Confirm exponential decay. Successive ratios: 8.0×10⁵ / 2.0×10⁶ = 0.40; 3.2×10⁵ / 8.0×10⁵ = 0.40; 1.3×10⁵/3.2×10⁵ ≈ 0.41. The roughly constant ratio per 2-minute interval confirms first-order (exponential) decay in free phage, consistent with the second-order rate law dP/dt = −k·P·B treated as pseudo-first-order because B (bacteria) is essentially constant over this short interval (bacterial doubling time, ~20–30 min, is much longer than the adsorption timescale). Step 2 — Extract the rate constant. Using P(t) = P₀·e^(−kBt), and taking the 0–4 min interval: ln(3.2×10⁵ / 2.0×10⁶) = −kB(4 min), so −1.83 = −kB(4), giving kB ≈ 0.458 min⁻¹. Dividing by B = 4×10⁸ cells/mL: k ≈ 0.458 / (4×10⁸) ≈ 1.1 × 10⁻⁹ mL/min This falls squarely within the textbook range for a well-adapted phage–host pair (10⁻⁹–10⁻⁸ mL/min), telling the student that T4 adsorbs efficiently and rapidly to this E. coli strain — useful confirmation before proceeding to a one-step growth curve to measure latent period and burst size.

Félix d'Herelle was a largely self-taught, itinerant microbiologist — Canadian-born, based for much of his career at the Pasteur Institute in Paris — whose career defies the stereotype of the credentialed academic scientist. Working on an outbreak of hemorrhagic dysentery among French troops in 1917, he noticed that filtered stool samples, passed through a filter fine enough to exclude bacteria, contained something that produced sharply bounded clear zones ("taches vierges," virgin spots, later called plaques) when spread on a bacterial lawn. He correctly inferred that these clearings represented localized destruction by a self-replicating, transmissible agent — a live "microbe of microbes" — rather than a static toxin, and he devoted much of the rest of his career to characterizing bacteriophages and pioneering their therapeutic use against bacterial disease, decades before antibiotics existed and, in the former Soviet republic of Georgia, continuing in specialized clinical use to this day.

Phage typing and epidemiology. Because phage host range tracks surface receptor chemistry so finely, panels of characterized phages are used to "type" bacterial isolates — a technique long predating genomic sequencing and still used for rapid outbreak tracking of pathogens such as Salmonella and Staphylococcus aureus.

Phage therapy against multidrug-resistant bacteria. As antibiotic resistance spreads, phages are being revisited as precision antibacterials. Because adsorption specificity confines a phage's action to particular strains, phage therapy can in principle spare beneficial commensal microbiota — but this same specificity means therapeutic phages must be matched to a patient's infecting strain, usually via host-range testing or engineered cocktails, exactly as reasoned through in Worked Example 1.

Engineering receptor-binding proteins. Synthetic biologists now swap tail-fiber or tail-spike genes between phages, or engineer them directly, to redirect a phage's host range toward a desired pathogen — for example, broadening or narrowing specificity for Klebsiella or Pseudomonas clinical strains, or building phages that specifically target antibiotic-resistant lineages.

Phage display and biotechnology. Filamentous phages such as M13, which adsorb to the tip of the bacterial F-pilus, are engineered to display foreign peptides or antibody fragments fused to their coat proteins, a technology (phage display) that won the 2018 Nobel Prize in Chemistry and underlies the discovery of many therapeutic antibodies.

Agriculture and food safety. Phage cocktails targeting receptor-defined serotypes of Salmonella, E. coli O157:H7, and Listeria are used as biocontrol agents on produce and in food processing, precisely because their narrow host range avoids disrupting beneficial bacteria in the food or gut microbiome.

Basic ecology and the microbiome. Because adsorption specificity constrains phage-host interactions so tightly, phage predation shapes microbial community structure globally, driving the "kill-the-winner" dynamics thought to maintain diversity in oceans, soils, and the human gut — a topic that connects directly forward to the population-level consequences of The Lytic Cycle and The Lysogenic Cycle.

Ask the tutor: How does a phage's receptor choice constrain or enable its ability to establish lysogeny versus commit to the lytic cycle?