Influenza A virus is built around a segmented genome wrapped in a borrowed membrane, and nearly every feature of its life cycle exploits that segmented design. Eight separate RNA strands let the virus swap genes between strains, dodge antibodies through rapid mutation, and replicate with a speed that keeps it circulating worldwide year after year. Understanding how the particle is built, how it enters a cell, how it copies itself, and how it undermines the immune response reveals why influenza remains one of the most persistent viral threats despite decades of vaccination.
Anatomy of the Viral Particle
An influenza A virion is roughly spherical to filamentous, enclosed in a lipid envelope stolen from the host cell it last budded from. Studding that envelope are two major surface glycoproteins: hemagglutinin (HA) and neuraminidase (NA). A smaller number of M2 ion channel proteins sit in the membrane as well. Beneath the envelope lies a shell of matrix protein (M1), and inside that shell are the eight ribonucleoprotein (RNP) complexes that carry the virus’s genetic information.
Each of the eight RNA segments is single-stranded and negative-sense, meaning it cannot serve directly as a message for making protein. Instead, each segment is coated with many copies of nucleoprotein (NP) and capped at one end by the viral RNA polymerase, a three-subunit enzyme made of PB1, PB2, and PA. Together, the RNA, NP coat, and polymerase form one RNP complex, the fundamental unit that the virus uses both to copy its genome and to produce messenger RNA.
1PubMed Central. Structure of influenza virus ribonucleoprotein complexes and their packaging into virionsHemagglutinin is by far the most abundant surface protein and the primary target of neutralizing antibodies. It is a trimeric spike whose “head” domain binds the virus to host-cell receptors, and whose “stem” domain later drives membrane fusion. Crystal structures of HA from highly pathogenic H5N1 strains show that its receptor-binding pocket has a preference for a particular type of sugar linkage on the cell surface, a detail that governs which species the virus can infect efficiently.
2PubMed Central. Human (α2→6) and avian (α2→3) sialylated receptors of influenza A virus show distinct conformations and dynamics in solutionNeuraminidase, the other major surface spike, is an enzyme. Its job comes at the end of the replication cycle: it clips sialic acid residues off the host cell surface so that newly made virions can detach and spread. The active site of neuraminidase recognizes sialic acid through a pocket of highly conserved amino acid residues, which is exactly why neuraminidase inhibitor drugs can target many different influenza strains at once.
3PubMed Central. The 2.2 A resolution crystal structure of influenza B neuraminidase and its complex with sialic acidReceptor Binding and the Species Barrier
Infection begins when hemagglutinin latches onto sialic acid residues decorating proteins and lipids on the surface of a host cell. The specific sugar linkage matters enormously. Human influenza viruses preferentially bind sialic acid attached in one configuration, while avian influenza viruses prefer a different one.
2PubMed Central. Human (α2→6) and avian (α2→3) sialylated receptors of influenza A virus show distinct conformations and dynamics in solution For years, the presumed absence of the avian-preferred linkage in human airways was thought to form an effective species barrier, limiting the ability of bird flu viruses to infect people. More detailed tissue mapping has since shown that human airways do carry some of the avian-type receptors, complicating that neat separation.
4PubMed Central. Sialic acid tissue distribution and influenza virus tropismThe receptor linkage preference is not an all-or-nothing lock and key. In solution, the two types of sugar receptors adopt distinct shapes and flexibilities, which influence how tightly and in what orientation hemagglutinin grabs them. A single amino acid change in the HA binding pocket can shift the preference enough to allow a bird virus to gain a foothold in mammalian airways, a switch that has preceded every known influenza pandemic.
Endosomal Acidification and Membrane Fusion
Once bound, the virus is pulled into the cell inside a membrane-wrapped bubble called an endosome. As the endosome matures, proton pumps in its wall lower the pH inside. This drop in acidity is the trigger hemagglutinin has been waiting for. The low pH kicks off a dramatic series of shape changes in the HA stem region: the central helices unwind and extend, the fusion peptide (a short hydrophobic stretch previously tucked away) springs out and embeds itself in the endosomal membrane.
5PubMed Central. pH-Dependent Mechanisms of Influenza Infection Mediated by HemagglutininStructural studies capturing intermediate states of this transition show that at around pH 5.2, the stem undergoes a rotation of about 9.5 degrees and shifts roughly 15 angstroms. The central helices straighten into an extended form that physically cannot accommodate the fusion peptide in its original resting position, making the conformational change essentially irreversible.
6PLOS Pathogens. Structural intermediates in the low pH-induced transition of influenza hemagglutinin The end result is that the viral membrane and the endosomal membrane are pulled together and fuse, dumping the RNP complexes into the cell’s cytoplasm.
Simultaneously, protons flow through the M2 ion channel into the interior of the virus particle. This acidifies the space between the envelope and the RNP complexes, causing the matrix protein M1 to change shape and release its grip on the RNPs. Without this step, the RNPs would remain tethered to M1 even after membrane fusion and would never reach the nucleus.
7PubMed Central. Stepwise priming by acidic pH and a high K+ concentration is required for efficient uncoating of influenza A virus cores after penetrationNuclear Replication
Here influenza diverges from most RNA viruses. Rather than setting up shop in the cytoplasm, influenza RNPs must travel into the cell’s nucleus to copy and transcribe their genomes. The nucleoprotein carries two nuclear localization signals that interact with the host’s own import machinery to shuttle each RNP through the nuclear pore. Blocking either signal reduces RNP import; blocking both nearly eliminates it.
8PubMed Central. Nuclear import of influenza A viral ribonucleoprotein complexes is mediated by two nuclear localization sequences on viral nucleoprotein This dependence on host nuclear import factors is also a point where host species compatibility comes into play. The virus’s polymerase subunits must similarly be imported and assembled inside the nucleus, and differences in host import pathways can restrict which species a given strain can replicate in efficiently.
9PubMed. Nuclear import of the influenza A virus transcriptional machineryOnce inside the nucleus, the viral polymerase performs two distinct tasks. For transcription, it snatches short capped RNA fragments from host messenger RNAs (a trick called “cap snatching”) and uses them as primers to produce viral mRNAs that the cell’s own ribosomes can translate. For genome replication, it makes a full-length positive-sense copy of each segment (called cRNA), then uses that as a template to produce new negative-sense viral RNA. The polymerase uses different initiation strategies on its two templates: on the cRNA promoter it starts internally and then realigns to the template’s end, while on the vRNA promoter it starts at the very first position.
10PubMed Central. Different de novo initiation strategies are used by influenza virus RNA polymerase on its cRNA and viral RNA promoters during viral RNA replicationA critical host factor in this process is ANP32, a family of proteins that the viral polymerase co-opts as a cofactor. Bird and mammalian versions of ANP32 differ, and influenza polymerases need specific adaptations to use mammalian ANP32 efficiently. The best-known such adaptation is a single amino acid change, PB2-E627K, which strongly favors use of mammalian ANP32B and is frequently seen in human-adapted strains. Alternative adaptations like PB2-D701N show no such bias toward ANP32B and appear more often in viruses isolated from pigs, dogs, and horses, where ANP32A is the preferred cofactor.
11PubMed Central. Mammalian ANP32A and ANP32B Proteins Drive Differential Polymerase Adaptations in Avian Influenza VirusAssembly, Export, and Budding
After new RNPs have been assembled in the nucleus, they must be exported back to the cytoplasm. The nuclear export protein (NEP) recruits M1 onto the RNPs, and the resulting complex is ferried out through nuclear pores. NEP’s C-terminal domain is essential for bridging M1 to the RNPs, while its N-terminal domain appears to destabilize the complex after export, preventing premature re-import.
12PubMed Central. The nuclear export protein of H5N1 influenza A viruses recruits Matrix 1 (M1) protein to the viral ribonucleoprotein to mediate nuclear exportAt the cell surface, HA plays a coordinating role beyond just being packaged into new particles. It helps concentrate other viral components, including bundled sets of all eight RNP segments, at specialized membrane regions called lipid rafts. Without adequate HA, the accumulation of NP, viral polymerases, NA, and M1 at lipid rafts is delayed and virion production drops, even though the eight segments are still bundled correctly on their own.
13PubMed Central. Influenza A Virus Hemagglutinin is Required for the Assembly of Viral Components Including Bundled vRNPs at the Lipid Raft Neuraminidase then cleaves sialic acid from the cell surface so that the freshly budded virion does not simply stick to the cell it just left. This is the step that neuraminidase inhibitor drugs like oseltamivir block, trapping new virions at the surface and limiting spread.
Suppressing the Innate Immune Response
Influenza does not just outrun the immune system. It actively disarms the cell’s earliest defenses. The most versatile saboteur is NS1, a small non-structural protein encoded on segment eight. NS1 targets RIG-I, a sensor that cells use to detect viral RNA in the cytoplasm. Normally, RIG-I detection triggers a signaling cascade that leads to production of type I interferons, the proteins that put neighboring cells on antiviral alert. NS1 short-circuits this by blocking TRIM25, an enzyme that must attach ubiquitin chains to RIG-I for the signal to propagate.
14PubMed Central. Influenza A virus NS1 targets the ubiquitin ligase TRIM25 to evade recognition by the host viral RNA sensor RIG-I A specific pair of amino acids in NS1 (at positions 96 and 97) mediates this interaction; mutant viruses lacking that interaction lose their ability to suppress interferon and become far less virulent in animal models.
NS1 also blocks a second ubiquitin ligase called Riplet, and this interaction shows species specificity. NS1 proteins from human-adapted viruses bind human Riplet and suppress its activity, while NS1 from avian or swine viruses generally cannot, hinting at one mechanism by which human-adapted strains are better at evading our particular brand of innate immunity.
15PLOS Pathogens. Species-Specific Inhibition of RIG-I Ubiquitination and IFN Induction by the Influenza A Virus NS1 Protein Beyond RIG-I, NS1 also suppresses noncanonical NF-κB signaling in lung cells, further dampening inflammatory gene expression.
16PubMed Central. The NS1 protein of influenza A virus blocks RIG-I-mediated activation of the noncanonical NF-κB pathway and p52/RelB-dependent gene expression in lung epithelial cellsA second arm of immune suppression involves shutting down the host cell’s own protein production. The virus deploys multiple proteins to prevent the formation of stress granules, cytoplasmic clusters that normally help cells stall viral replication. NS1 inactivates the kinase PKR to prevent a translational brake from being engaged. Nucleoprotein independently inhibits stress granule formation through a separate pathway. And PA-X, a protein produced from a ribosomal frameshift on the PA gene, aggressively degrades host mRNAs, draining the cytoplasm of the raw material cells need to mount an antiviral response.
17PLoS Pathogens. Influenza A Virus Host Shutoff Disables Antiviral Stress-Induced Translation Arrest PA-X and the accessory protein PB1-F2 can cooperate to enhance viral replication and increase virulence in mice, particularly during the early phase of infection.
18PubMed Central. Effects of the PA-X and PB1-F2 Proteins on the Virulence of the 2009 Pandemic H1N1 Influenza A Virus in MiceAntigenic Drift, Reassortment, and Pandemic Potential
Even when the immune system does respond, influenza has two major strategies for outpacing antibody recognition over time. The first is antigenic drift: the viral polymerase is error-prone, and mutations steadily accumulate in the head region of HA, the primary target of neutralizing antibodies. Because the strongest antibodies tend to focus on a small number of sites at the HA head, even a few amino acid substitutions there can allow a drifted strain to escape population immunity.
19PubMed Central. Antibody Immunodominance: The Key to Understanding Influenza Virus Antigenic DriftThe second strategy is antigenic shift, a wholesale exchange of genome segments when two different influenza strains infect the same cell simultaneously. Because the genome is split across eight separate RNAs, the progeny virions can contain a mix of segments from both parent strains. This reassortment can produce viruses carrying an entirely new HA or NA subtype against which the human population has little pre-existing immunity, the classic recipe for a pandemic. Reassortment frequency is not random; particular strains consistently push it higher or lower regardless of which other strain is present, making it an emergent property of the specific combination of co-infecting viruses.
20PubMed Central. Influenza A virus reassortment is strain dependentPre-existing immunity itself shapes which reassortant viruses succeed. Strain-specific neutralizing antibodies can selectively suppress one parent virus in a co-infection, giving reassortants carrying the other parent’s surface proteins a survival advantage and effectively promoting antigenic shift. Broadly cross-reactive antibodies, by contrast, do not produce this biased selection.
21PubMed Central. Subtype-specific neutralizing antibodies promote antigenic shift during influenza virus co-infectionImmunological Imprinting and Why First Infections Matter
Your very first influenza infection leaves a lasting stamp on how your immune system responds to every subsequent encounter. This phenomenon, sometimes called “original antigenic sin,” means that the antibody response tends to be biased toward the HA subtype you first met in childhood. Work in both ferrets and human cohorts has shown that infections with antigenically distinct subtypes later in life boost antibodies against the HA stalk region shaped by that first exposure, but those boosted stalk antibodies often bind poorly to the virus that actually triggered the boost.
22PubMed Central. Original antigenic sin priming of influenza virus hemagglutinin stalk antibodies One implication is that people remain susceptible to novel subtypes partly because their immune systems keep reinforcing an outdated blueprint.
This imprinting has complicated efforts to design a universal influenza vaccine. The conserved HA stalk, less variable than the rapidly mutating head, has been a top candidate for broadly protective vaccines. But if individuals’ stalk antibody repertoires are already shaped by their first childhood infection, the same stalk-targeting vaccine might produce very different responses in people born in different decades who were imprinted by different subtypes.
23PubMed Central. From Original Antigenic Sin to the Universal Influenza Virus VaccineAntiviral Drugs and Resistance
Current antivirals target three distinct steps of the life cycle discussed above. The adamantanes (amantadine and rimantadine) block the M2 ion channel, preventing the acidification of the virion interior needed for uncoating. Neuraminidase inhibitors (oseltamivir, zanamivir) block viral release from infected cells. And newer cap-dependent endonuclease inhibitors like baloxavir marboxil disrupt the cap-snatching step the polymerase uses to start transcription.
Resistance has been a recurring problem. High-level adamantane resistance emerged quickly and persists in virtually all circulating influenza A strains even without ongoing drug pressure.
24PubMed Central. Influenza and antiviral resistance: an overview Reassortment can spread resistance between lineages: in canine H3N2 influenza, a single amino acid substitution in M2 acquired through reassortment with pandemic H1N1 was enough to confer amantadine resistance while leaving neuraminidase inhibitor susceptibility intact.
25PubMed Central. Acquisition of amantadine resistance via M gene reassortment in canine H3N2 influenza virus and elucidation of the resistance mechanism Neuraminidase inhibitor resistance has stayed low overall, but it does arise, most often in people with weakened immune systems on prolonged oseltamivir courses. Resistance to baloxavir can also emerge during treatment, a reminder that monotherapy against a rapidly mutating virus is always a gamble.
Broadly neutralizing antibodies that target the conserved HA stalk have been identified across multiple influenza A subtypes and represent a potential route toward therapies and vaccines less vulnerable to drift and shift.
26PubMed Central. Broadly neutralizing antibodies against influenza virus and prospects for universal therapies Crystal structures have revealed at least three highly conserved sites on HA that these antibodies recognize, providing templates for next-generation vaccine design.
Defective Viral Genomes and Disease Severity
Not every RNA copy the polymerase makes is functional. Errors during replication frequently generate truncated or rearranged segments called defective viral genomes (DVGs). These broken segments cannot produce a viable virus on their own but can still trigger innate immune sensors when they accumulate inside infected cells. DVGs effectively act as an internal alarm, stimulating interferon and proinflammatory cytokine production.
The balance between functional and defective genomes turns out to matter for how sick someone gets. In a study comparing influenza patients with mild versus severe or fatal outcomes, viruses from severe cases carried roughly ten-fold fewer DVGs than those from mild cases.
27PLOS Pathogens. Reduced accumulation of defective viral genomes contributes to severe outcome in influenza virus infected patients Mouse experiments confirmed the mechanism: infection with virus stocks rich in DVGs triggered an early interferon response that limited viral replication and prevented severe disease, while stocks with few DVGs replicated to high titers and caused rapid weight loss, elevated inflammation, and higher mortality.
28PubMed Central. Levels of Influenza A Virus Defective Viral Genomes Determine Pathogenesis in the BALB/c Mouse Model DVG abundance may eventually serve as a virulence marker, and there is growing interest in whether synthetic DVGs could be developed as therapeutic agents to jumpstart the immune response during severe infections.