Proton Channel Activity and Uncoating: The M2 protein of influenza A is a pH-gated proton channel (a viroporin) embedded in the viral envelope[1]. Upon entry of the virion into an endosome, the endosomal acidification (pH \~5–6) activates M2’s channel, allowing protons to flow into the virion interior[2]. This influx of protons acidifies the core, causing the dissociation of the matrix protein M1 from the viral ribonucleoproteins (vRNPs) and enabling uncoating of the viral genome[3]. M2’s ion selectivity is extraordinarily high for protons (10^6–10^7-fold preference over other cations)[4], chiefly due to a histidine residue at position 37 (His37) that serves as a proton sensor/filter in the channel[5]. The channel conducts protons inward when external pH drops (as in endosomes) but does not efficiently conduct protons outward, ensuring unidirectional acidification of the virion core[2]. This proton-channel function of M2 is essential for productive infection, as blocking M2 (e.g. by adamantane drugs) prevents uncoating and halts replication[6].
Timing of Expression and Regulation: M2 is encoded by the M gene segment (segment 7) via a spliced mRNA. The unspliced transcript encodes M1, whereas a +1 spliced reading frame produces M2 (a 97-amino-acid protein)[7][8]. M2 is thus expressed later and in lower abundance relative to M1, due to regulated splicing efficiency. The M1:M2 protein ratio is high early in infection, ensuring abundant M1 for virion structure, while M2 levels rise when needed for assembly and budding[9][10]. Notably, influenza can tolerate some flexibility in M2 expression: certain strains (e.g. lab-adapted A/WSN/33) encode an alternate splicing product called “M42” that can functionally replace M2, illustrating that the virus ensures a proton channel is available even if encoded in a different form[7]. In human H3N2 viruses, a conserved single-nucleotide variant in the M segment (C55T) reduces M2 splicing, resulting in lower M2 protein levels; intriguingly, H3N2 viruses have adapted to remain pathogenic despite this, whereas H1N1 viruses are more attenuated by low M2 levels[10][11]. This suggests subtype-specific optimization of M2 requirements.
Interaction with Viral and Host Factors: Although M2 is a minor component of the virion (only 20–60 copies per particle), it plays multiple supporting roles beyond uncoating:
Assembly and Budding: M2 is crucial for virus assembly at the plasma membrane. It strongly interacts with M1, the matrix protein lining the inner virion shell. Fluorescence microscopy studies in infected cells show that M2 actively recruits M1 to the membrane, initiating assembly of virion buds[12]. This interaction ensures M1 accumulates at the correct site and can later bind to vRNPs. Following recruitment, M2’s amphipathic helix in its cytoplasmic tail (see Structure-Function) induces membrane curvature and scission during the final step of budding[13]. M2 localizes to the budding neck (at the edge of lipid rafts where hemagglutinin (HA) and neuraminidase cluster) and its amphipathic helix inserts into the inner leaflet, deforming the membrane to pinch off the virion[13]. This M2-mediated membrane scission is unique to influenza (an ESCRT-independent budding mechanism) and is essential for efficient virus release.
HA Maturation: Newly made HA protein of many avian and pathogenic strains is cleaved in the trans-Golgi network and could trigger prematurely in low pH. M2 is expressed in the Golgi and pumps protons out of Golgi vesicles, raising the luminal pH[14]. This prevents the premature conformational activation of HA during transport to the cell surface[15]. In highly pathogenic H5/H7 viruses (which have HA that is cleaved by furin in Golgi), M2’s role in buffering Golgi pH is critical to keep HA in its native, prefusion state until the virion buds[15]. Notably, the proton gating properties of M2 in different strains appear to have co-evolved with the pH stability of HA; viruses with more acid-sensitive HAs require more active M2 gating to protect HA, whereas those with more acid-stable HAs can tolerate differences in M2 activity[16].
Host Cell Modulation: M2 can modulate host processes. The ion channel activity of M2 in the Golgi and endosomes can inadvertently activate the NLRP3 inflammasome, triggering pro-inflammatory cytokine responses[17]. Additionally, M2 has a conserved LC3-interacting region (LIR) in its cytosolic tail that engages the host autophagy machinery. M2 interacts with LC3 (a marker on autophagosome membranes), and this interaction was shown to redistribute autophagosomes to the plasma membrane where virus buds[18]. M2-mediated relocalization of autophagosome membranes may supply additional membrane for virion budding and also delays the autophagosome-lysosome fusion (stalling autophagic degradation)[19]. This function is beneficial for the virus, as mutations in the M2 LIR motif dramatically reduce virus production[20]. A host protease (caspase-3) can cleave M2 and disrupt the M2–LC3 interaction, suggesting the cell may attempt to counteract this M2 function to restore autophagy and restrain virus release[21]. M2 is also targeted by certain host restriction factors; for example, the E3 ubiquitin ligase MARCH8 was recently shown to redirect M2 to lysosomes for degradation, thereby inhibiting influenza virion release[22] (a host defense strategy).
Summary of M2’s Roles: In summary, M2 is a multifunctional protein that is involved in virus entry, assembly, and interactions with host cell pathways. Key functions include:
Proton channel activity for entry/uncoating: M2 tetramers acidify the virion interior in endosomes, enabling dissociation of M1 from vRNPs[3]. This function is vital for genome release and initiates infection.
Stabilizing HA during virus egress: By equilibrating pH in the trans-Golgi, M2 prevents premature activation of HA during transport in the infected cell[15]. This ensures newly assembled virions have fusion-competent HA.
Driving virion assembly and release: M2 is essential for budding; it recruits M1 to the membrane for assembly[12] and its amphipathic helix induces membrane scission to release progeny virions[13].
Engaging host cellular processes: M2 modulates host pathways to the virus’s advantage – it can activate immune sensors (inflammasome) inadvertently[17], but also subverts autophagy via its LC3-interaction to promote viral budding and evade degradation[18]. These interactions underline M2’s role in balancing host responses and efficient virus production.
Overall Architecture: Influenza A M2 is a small integral membrane protein of 97 amino acids, comprising three domains: an N-terminal ectodomain (M2e, \~24 residues) exposed on the virion surface, a single transmembrane (TM) domain (\~19 residues) that spans the viral envelope, and a C-terminal cytosolic tail (\~54 residues)[8]. M2 exists as a homotetramer in the membrane, with four identical subunits assembling to form a pore. The tetramer is stabilized in part by disulfide bonds: each M2 monomer contains two conserved cysteines (Cys17 and Cys19) in the ectodomain that form disulfide linkages (within or between subunits), helping maintain the quaternary structure[23]. The ectodomain is otherwise very short but is antigenically conserved and forms a loop on the virion surface. The transmembrane helices of the four subunits pack together to create the channel, and the C-terminal tails extend inside the virion and host cell.
Transmembrane Channel and Key Residues: The TM domain of M2 is the functional core of the proton channel. In the tetramer, the four TM helices are tilted \~25° within the membrane and line a central aqueous pore[24]. Several pore-lining residues (Val27, Ala30, Ser31, Gly34) form the narrow channel lumen[24]. Importantly, His37 and Trp41 in the TM domain are absolutely critical for M2’s proton conductance mechanism[25]. These residues cluster within the channel and act as a pH-sensitive gate:
His37 (the “Histidine Box”): A ring of four histidines (one from each subunit) sits roughly mid-pore and acts as the proton sensor/selectivity filter[5][26]. At neutral pH, the histidine imidazole rings are mostly unprotonated, and the channel is closed. As external pH drops (≤6.5), the His37 residues become protonated (each histidine can accept a proton). Protonation causes electrostatic repulsion among the positively charged imidazolium rings, pushing the helices apart slightly[27]. This conformational change opens the pore to water and disrupts a critical hydrogen bond between Trp41 and Asp44, which leads to opening the channel gate[3]. His37 thereby senses acidity and also participates in proton relay: it can bind a proton from the outside and, through tautomerization and flipping of its imidazole, pass the proton into the interior water network (the “proton relay model”)[28][29]. Solid-state NMR studies show that the tetrameric His37 cluster undergoes stepwise protonation (with pK_a values \~8.2, 6.3, etc.), and that the channel conducts protons most efficiently when the His37 cluster is in a doubly- or triply-protonated state (+2 to +3 overall charge)[30]. This corresponds to the pH range of endosomes (\~5–6) where proton influx is maximal.
Trp41 (the Gate): Just below His37 in the channel is a ring of four tryptophan residues (Trp41), which function as a physical gate or valve. In the closed state (high pH), the indole side chains of Trp41 block the pore, preventing proton flow or leakage[31][32]. Upon low pH activation, once His37 becomes protonated and pushes the helices apart, the Trp41 “gate” swings open (the Trp side chain rotamer changes), allowing water and protons to traverse the pore[3]. Trp41 thus ensures that when external pH is neutral (virion outside host cell or post-uncoating), the channel remains shut, preventing protons from flowing in the wrong direction or dissipating the proton gradient[33]. Mutagenesis confirms these roles: replacing His37 with other residues abolishes proton selectivity, and replacing Trp41 with smaller residues (Ala, Cys, Phe) yields channels that cannot close properly at high pH[25]. In essence, His37 is the pH-activated switch, and Trp41 is the check-valve, together conferring M2’s unique gating and selectivity properties[26].
Several high-resolution structures have elucidated how these residues function. X-ray crystallography at 1.1 Å resolution (of an M2 TM peptide) revealed a cluster of ordered water molecules inside the pore that form hydrogen-bonded “water wires” extending \~17 Å from the channel entrance down to the His37 level[5]. These waters are positioned to shuttle protons via Grotthuss mechanism. The pore-lining carbonyl groups of the helices further help stabilize hydronium ions through second-shell interactions with bridging waters[34]. Notably, the structures at high vs. low pH show that water becomes more disordered and mobile as pH drops, consistent with the channel opening to a conductive state[35]. Molecular dynamics simulations support a “proton relay” mechanism: as each His37 in the tetrad gets protonated in turn, the network of hydrogen bonds in the water wire reorients, allowing protons to effectively move inward but not back out[36][2]. This dynamic hydration and reorientation likely enforce directionality of proton transport (from the acidic endosome into the virion)[36]. Alternative models, like the earlier “water wire” hypothesis, also envisioned a continuous chain of water facilitating proton hop without requiring histidine sidechain reorientation[28], but current evidence favors a hybrid mechanism where His37 binds and releases protons (relay) assisted by a transient water chain.
Structural Dynamics: The conformational states of the M2 channel can be summarized as closed (at high pH) and open (at low pH). At pH >7.5, all His37 are uncharged and the four Trp41 sidechains form a tight hydrophobic plug, excluding water from the pore and “dehydrating” the His37 region[32]. The helices are closely packed. As the environment acidifies to pH \~6.5 or below, stepwise protonation of His37 causes the helices to splay apart slightly (particularly the C-terminal half of the TM segment)[37][38]. At +2 to +3 charge on the histidine tetrad, the Trp41 gate is destabilized (its hydrogen bond with Asp44 breaks)[3] and the gate “flips” open, creating a continuous aqueous path into the virion[3][39]. The N-terminal end of the channel (around Val27) may constrict slightly in the open state (a “Val27 valve” effect) as the lower gate opens, but this likely helps prevent backflow[39]. At maximal protonation (+4, very low pH \~5), the electrostatic repulsion is greatest and the channel is fully open/hydrated, allowing protons to flood in[40]. After proton release inside, the histidines deprotonate and the channel returns to the closed conformation, ready for another cycle[40]. Importantly, this gating mechanism is highly conserved among influenza A M2 proteins, as discussed in the Evolution section.
Cytoplasmic Tail and Function: The M2 cytosolic tail (residues 47–97) contains two notable elements: an amphipathic helix (approximately residues 47–62, immediately following the TM domain) and a relatively unstructured acidic region (including residues 70–77 which bind M1, and the far C-terminus which contains the LC3-interacting motif)[41]. The amphipathic helix (AH) is highly conserved and is a key functional domain for virus assembly. It lies along the inner leaflet of the plasma membrane when M2 is inserted, and because one face of the helix is hydrophobic and the opposite face hydrophilic, it can insert shallowly into the membrane. This in-plane insertion induces curvature in the membrane. Experimental studies showed that the M2 AH is required for efficient budding: it causes membrane deformation in a cholesterol-dependent manner, helping to pinch off the budding virion[13]. Mutations disrupting the hydrophobic face of this helix or depleting membrane cholesterol both interfere with M2’s ability to deform the membrane, resulting in failure of virion scission[13]. Thus, structurally, the amphipathic helix acts like a lever or wedge in the membrane to facilitate the final membrane fission event in virus release. Farther down the tail, residues 70–77 are important for binding the M1 protein and possibly for localizing M2 to lipid raft microdomains (this region overlaps with a putative caveolin-binding motif)[41]. The extreme C-end (around residues 90–96) contains a “YXXØ” LC3-interacting region (where Ø is a bulky hydrophobic residue) – specifically a sequence that allows M2 to bind autophagosome protein LC3[18]. This short linear motif is conserved in almost all influenza A M2 sequences and is critical for M2’s role in redirecting autophagosomes, as described earlier[18]. Structurally, this suggests that M2’s tail, though mostly flexible, has evolved to host protein-interaction motifs (like the LIR and a potential endocytosis motif YXXΦ) that contribute to virus-host interactions.
Notable Structural Features and Their Functional Significance: To summarize the structure-function relationship of M2, below are key structural features and how they relate to M2’s activity:
Homotetrameric assembly: M2 forms a tetramer of four helices, creating a central pore[1]. The tetramer is stabilized by disulfide bonds in the ectodomain (Cys17–Cys19) and helix-helix packing in the membrane[23]. This oligomerization is essential – a monomer cannot form a channel. The fourfold symmetry allows cooperative gating (e.g., multiple His37 protonation states).
Transmembrane helix (residues \~25–46): Each subunit’s TM helix lines the channel; key pore-facing residues (V27, A30, S31, G34) shape the channel size and drug-binding site[24]. The helices tilt to form a narrow pore that is opened by small conformational changes. This helix also contains Ser31, the site where adamantane drugs (amantadine/rimantadine) bind; these drugs physically occlude the pore near S31 in drug-sensitive strains, plugging proton flow[42].
His37 and Trp41 (pH sensor and gate): These two residues within the TM domain are indispensable for function. His37 confers proton selectivity and pH activation – only protons can efficiently protonate the His and trigger channel opening[26]. Trp41 is the physical gate that prevents ion leakage; it’s closed at high pH and opens at low pH[33]. Together they ensure M2 opens only under the right conditions (acidic exterior) and maintain the proton gradient until needed.
Dynamic water network: The channel interior is not a static tube but is filled with water molecules whose arrangement depends on pH. At low pH (channel active), a continuous “water wire” connects the outside to the His37 cluster, enabling proton translocation via hydrogen-bonded hopping[5]. At high pH, this network is broken (Trp41 gate closed and His37 uncharged), so water is sparse in the pore and protons cannot traverse[35]. The structural studies capturing these states have validated the models of how M2 shuttles protons in a directional manner[35][43].
Cytoplasmic amphipathic helix: The α-helix spanning roughly residues 47–62 lies along the inner membrane surface and is highly conserved[44]. Its structure (amphipathic nature) is directly tied to function – by embedding its hydrophobic face into the lipid bilayer, it induces curvature. This structural element is crucial for membrane scission during budding[13]. Viruses with mutations in this region show abnormal virion morphology and reduced release, underscoring that the proper structure (amphipathic character and orientation) is needed for function[13].
Interaction motifs in the tail: Although largely unstructured, the tail’s conserved motifs (for M1 binding, caveolin binding, LC3 interaction) hint at a modular functionality. For instance, maintaining an accessible LC3-interacting region suggests the tail’s structure (perhaps transient helix or loop) can present this motif to host proteins[18]. Likewise, the ability of the tail to bind M1 is linked to specific residues (70–77) that likely form a recognition site when M1 is nearby[41]. In summary, the primary structure of the tail encodes interaction sites that are used during virus assembly and host modulation – even without a fixed secondary or tertiary structure, these sequences are conserved for functional binding.
Multiple structural biology approaches (solid-state NMR, X-ray crystallography, solution NMR, and cryo-EM) have collectively built this understanding of M2. Early NMR studies in lipid bilayers confirmed that M2 forms a tetrameric helix bundle and identified the histidine and tryptophan roles in channel gating[25]. Crystal structures of the M2 TM domain, both at high pH and low pH, visualized the positions of waters and sidechain conformational changes (e.g., different rotamers of Ser31, His37, Trp41 in closed vs open states)[45]. More recently, cryo-EM of intact influenza virions has observed M2, but because M2 is sparse in virions, high-resolution virion maps mainly show the larger HA/NA spikes[16]. Instead, high-res studies focus on M2 reconstituted in membrane mimetics. Nevertheless, all methods agree on the core structure: a tetrameric, pH-gated proton channel. Figure 1 (below) illustrates the M2 tetramer in the membrane and highlights key residues (H37, W41) and the amphipathic helix orientation. (Figure not shown due to text format.)
Sequence Conservation: The M2 protein is remarkably conserved across influenza A viruses, reflecting the strong functional constraints on this ion channel. Overall, M2 amino acid sequences from different subtypes typically show \~90–95% identity to one another[46][47]. In particular, the N-terminal ectodomain (M2e) is highly conserved. The first 9 residues of M2e are almost absolutely invariant among all influenza A isolates – this includes human, swine, equine, avian strains, and even the divergent bat influenza viruses[48]. For example, virtually all A strains share the N-terminal motif SLLTEVETP (positions 2–9 of M2). The remainder of M2e (residues 10–24) is slightly more variable but still conserves key residues like Arg12, Trp15, and the Cys17-Cys19 pair (which form the disulfide bond)[49]. The conservation of cysteines and other M2e residues indicates their importance for structure (disulfide stability) and possibly an essential epitope – indeed M2e is exposed on infected cell surfaces and has been proposed as a “universal” vaccine target because of its sequence conservation[23][49]. Moving to the TM domain, critical residues (Ser31, His37, Trp41, Asp44, etc.) are almost universally conserved across strains, since they govern channel function. For instance, His37 and Trp41 are present in essentially all influenza A M2 sequences (including avian and swine variants) because any change there typically abolishes ion channel activity[25]. The cytoplasmic tail of M2 is also largely conserved, especially the amphipathic helix region and the LC3-interacting motif. Alanine-scanning studies have shown that hydrophobic residues in the amphipathic helix (Leu50, Leu51, etc.) are invariant and required for budding function[50]. The LC3-interaction motif in the tail (such as a YxxL sequence around residues 90–93) is likewise conserved in nearly all strains[18]. This high degree of conservation implies strong purifying selection on M2 – changes are generally not tolerated unless they preserve the protein’s structure and function.
Evolutionary Pressures and Mutational Variability: Although M2 is conserved, certain sites in M2 have undergone notable mutations due to specific selective pressures:
Drug Selection (Amantadine Resistance): M2 was the target of the antiviral drugs amantadine and rimantadine, which block the channel. Widespread use of these drugs (especially in the 1990s–2000s) led to the emergence of specific resistance mutations in M2. The most prevalent is Ser31Asn (S31N) in the transmembrane domain. This single mutation prevents amantadine from binding effectively, and it rose in frequency under drug pressure[42]. By the mid-2000s, S31N became nearly fixed in human H3N2 and H1N1 strains globally, such that >95% of circulating human influenza A viruses were amantadine-resistant[51]. Other resistance mutations include V27A, L26F, A30T, and G34E, which also alter the drug-binding pocket[42]. Notably, these mutations were selected despite slight fitness costs, because they confer a strong survival advantage in the presence of drug. For example, S31N slightly reduces channel conductance but still maintains sufficient proton flux for the virus[25]. This suggests M2 can tolerate some changes at these pore-lining positions as long as the essential gating residues remain intact. The dominance of S31N in recent decades is a prime example of positive selection acting on M2.
Host Adaptation and Splicing: There is evidence that M2 evolution is also shaped by host-specific factors. As mentioned, human H3N2 viruses harbor a mutation (C55T in M segment RNA) that lowers M2 expression by altering splicing efficiency[10]. This variant became fixed in H3N2 after the 1968 pandemic. It likely conferred an advantage (or at least did not hinder H3N2) in human hosts, perhaps because H3N2’s other proteins or host interactions compensate for lower M2. In contrast, H1N1 lineages maintained a cytosine at that position (higher M2 levels) and are more dependent on robust M2 production[52]. This divergence suggests different influenza A lineages have evolved distinct optimal M1:M2 ratios for fitness in their host environment. Another host adaptation involves amantadine use in poultry: in avian H5N1 outbreaks, extensive (and often unregulated) amantadine usage led to many H5N1 isolates carrying S31N or other M2 mutations by the mid-2000s[53]. For instance, H5N1 viruses from Southeast Asia frequently contained L26I and S31N in M2, indicating independent emergence of drug resistance in birds[53]. These examples show how human vs. avian usage of antivirals and differences in host biology can drive M2 sequence changes.
Immune Pressure: Under normal infection, M2 is a minor antigen (compared to HA and NA). However, the extracellular M2e can be targeted by the immune system in some contexts. Because M2e is conserved, antibodies to M2e tend to be cross-reactive; still, the virus can accumulate a few substitutions in M2e under immune pressure. Positions 10–24 of M2e, while variable, sometimes change in response to passaging in immunized animals or during escape from M2e-targeted vaccines[49]. Notably, M2e has no glycosylation in most strains, so it lacks the “glycan shield” that HA has. A few human strains (some H3N2 isolates) have been reported with an extra glycosylation site in M2e (e.g., substitution that creates an NXT motif), but this is rare. Overall, natural infection induces only a weak anti-M2e antibody response[54], so immune-driven evolution of M2 is limited compared to HA. The main conserved immunogenic elements (like the first 9 amino acids of M2e) have stayed the same for decades[48].
Conservation of Mechanism Across Strains: Despite sequence differences, the fundamental mechanism of M2 as a proton channel is conserved across all influenza A strains. This includes avian, swine, equine, and human viruses, as well as the unconventional bat influenza viruses. Studies on bat influenza A subtypes H17N10 and H18N11 (discovered in bats in 2012) show that their M2 proteins, although genetically divergent, still function as proton channels and possess the hallmark features: a tetrameric TM domain with the critical His37 residue and the ability to conduct protons in a low-pH dependent manner[55]. The bat M2 proteins do have “atypical” features – for example, they have a longer ectodomain and some unique substitutions. Interestingly, bat M2 channels have a lower conductance, which may be an adaptation to avoid proton overload in bat cells (possibly reducing cytopathic effects)[56]. Nevertheless, the proton selectivity and gating (histidine-dependent) are retained, and when bat M2 is expressed in influenza virions, it can functionally replace classical M2[55]. This underscores that the viroporin role of M2 is universally conserved: any influenza A virus needs a mechanism to acidify the virion interior, and M2 provides that. Even influenza B and C viruses have analogous proteins (BM2 and CM2) with similar proton channel function (though they share little sequence homology with AM2)[57][58], indicating convergent evolution to achieve the same end.
On the other hand, where evolution has leeway, it has produced some variations: e.g., the length of the M2 cytoplasmic tail can vary slightly (some strains have a 98-aa M2 due to an extra residue), and the presence of the alternate M42 protein in some lab strains shows the genetic plasticity to encode the channel function differently[59]. But these changes do not alter the fundamental mechanism – they either maintain a functional channel or the virus is not viable.
To illustrate differences and conserved features of M2 across various influenza A virus groups, Table 1 provides a comparison of representative strains:
Table 1. Notable M2 Features across Different Influenza A Strains and Lineages
| Influenza A Group / Strain | M2 Notable Features |
|---|---|
| Seasonal Human (H1N1, H3N2) | M2 sequence is highly conserved; the N-terminal 9 amino acids are identical in virtually all human isolates[48]. Since the mid-2000s, most human strains carry the S31N mutation in M2, conferring amantadine resistance[51]. Critical channel residues (His37, Trp41, Asp44) remain invariant. M2 proton channel activity and functions are consistent across seasonal strains. |
| Avian Influenza (e.g. H5N1) | M2 protein is very similar to that of human viruses (first 9 residues SLLTEVETP are the same[48]). Many avian strains were naturally amantadine-sensitive (Ser31) until drug use in poultry selected for resistant mutants – e.g. H5N1 viruses from Asia often acquired S31N (with L26I) by 2005[53]. Aside from drug-related changes, avian M2 retains conserved features and can function interchangeably with human M2 in reassortants. Mechanistically, avian M2 channels behave the same, although pairing with avian HA may influence optimal gating (co-evolution with HA fusion pH[16]). |
| 2009 Pandemic H1N1 (pdmH1N1) | This strain’s M2 came from a swine lineage. The initial 2009 isolates had Ser31 (drug-sensitive), since swine influenza was not under amantadine pressure. Within a couple of years of human circulation, S31N mutants of pdmH1N1 became dominant, echoing the pattern of older seasonal strains. The pdmH1N1 M2 is otherwise typical in sequence and function, and it fully substitutes for previous human M2 in viral fitness. Notably, pdmH1N1 introduced the M segment from swine, but it proved functionally compatible in humans – highlighting the conservative nature of M2 function across hosts. |
| Equine & Other Lineages | Equine influenza A (H3N8 and H7N7) also have highly conserved M2 sequences, with the same key residues. Amantadine resistance was observed less frequently in these, but experimentally, the same mutations (S31N, V27A) confer resistance in equine M2 as well[60]. Canine and other mammalian-adapted influenza viruses (derived from avian strains) similarly show minimal divergence in M2 aside from sporadic drug-resistance mutations. This underlines that M2’s function is conserved across species-adapted lineages, with any necessary host-specific adaptations occurring outside the core channel region. |
| Bat Influenza (H17N10, H18N11) | M2 from bat influenza A viruses is genetically divergent (only \~50–60% identity to conventional M2) but preserves the essential motifs – for example, it still has a histidine at the position equivalent to 37, required for proton gating[61]. Bat M2 proteins form functional ion channels and show acid-activated proton conductance like other M2, albeit with some atypical properties[56]. One study noted bat M2 has a reduced proton transport rate, hypothesized to avoid excessive ion flux that could trigger host stress in bats[56]. Structurally, bat M2 proteins have a longer ectodomain but still form tetramers. Thus, even in this divergent lineage, the mechanism (pH-gated proton channel) is conserved, affirming its fundamental role for influenza A viruses. |
| Lab-Adapted Strains (WSN/33) | Some laboratory strains demonstrate the virus’s evolutionary flexibility with M2. A prime example is the M42 protein found in A/WSN/1933 (H1N1). Due to alternative splicing, WSN can produce a protein (sometimes called M42) that is essentially M2 with a different N-terminus. M42 can functionally replace M2’s ion channel role[59]. This shows that while the protein sequence may vary, influenza will maintain an M2-like function. Any attenuation of M2 (through mutation or truncation) in lab strains is often compensated by second-site mutations or alternate protein production, again underscoring that the proton channel function must be preserved for viability. |
Conserved Mechanism: In conclusion, the mechanism of action of M2 is conserved across all influenza A strains. Sequence analysis reveals strong purifying selection on the M2 gene, with only a few sites tolerating change (mostly under drug pressure or in the ectodomain under immune pressure). The proton channel activity – controlled by the His37 sensor and Trp41 gate – operates the same way in H1, H3, H5, and other subtypes, as well as in unique variants like bat influenza. Evolution has fine-tuned M2 in certain contexts (e.g. modulating how active the channel is in concert with HA stability[16]), but it has not reinvented the core function. Even radical innovations, such as the emergence of the M42 protein in a lab strain or the existence of distinct BM2 and CM2 in influenza B/C, highlight the same principle: a tetrameric proton channel is indispensable for the influenza viral life cycle. Thus, across the vast diversity of influenza A viruses, M2 remains a highly conserved molecular machine, reflecting both its functional importance and the limited evolutionary pathways that can maintain its critical role.
Sources:
Lamb, R.A. & Choppin, P.W. (1985). Identification of a second protein (M2) encoded by RNA segment 7 of influenza A virus[62]. Cell, 40(3), 627–633.
Pinto, L.H., Holsinger, L.J. & Lamb, R.A. (1992). Influenza virus M2 protein has ion channel activity[63]. Cell, 69(3), 517–528.
Cady, S.D. et al. (2009). Structure and function of the influenza A M2 proton channel[64]. Biochemistry, 48(4), 7356–7364.
Thomaston, J.L. et al. (2015). High-resolution structures of the M2 channel from influenza A reveal dynamic proton conduction pathways[5][35]. PNAS, 112(46), 14260–14265.
Rossman, J.S. et al. (2010). Influenza virus M2 protein mediates ESCRT-independent membrane scission[13]. Cell, 142(6), 902–913.
Iwatsuki-Horimoto, K. et al. (2006). Mutation in M2 per se does not cause attenuation of H5N1 influenza A viruses in chickens[53]. J. Virol., 80(24), 12702–12707.
Petrich, A. et al. (2021). Influenza A M2 recruits M1 to the plasma membrane (fluorescence microscopy study)[65]. Biophys J., 120(24), 5478–5490.
Deng, L. et al. (2015). M2e-based universal influenza A vaccines (review of M2 structure/function)[49][18]. Vaccines, 3(1), 105–136.
Mtambo, S.E. et al. (2021). Influenza viruses: structure, function, and inhibition of the M2 ion channel (review)[66][27]. Molecules, 26(4), 880.
Zhong, W. et al. (2005). Mechanism of amantadine resistance in influenza virus M2 channel (S31N prevalence)[51]. J. Virol., 79(10), 6715–6722.
[1] [8] [41] [57] [58] M2 proton channel - Wikipedia
https://en.wikipedia.org/wiki/M2_proton_channel
[2] [3] [4] [6] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [39] [40] [42] [43] [60] [62] [63] [64] [66] Influenza Viruses: Harnessing the Crucial Role of the M2 Ion-Channel and Neuraminidase toward Inhibitor Design
https://www.mdpi.com/1420-3049/26/4/880
[5] [34] [35] [36] [37] [38] [45] High-resolution structures of the M2 channel from influenza A virus reveal dynamic pathways for proton stabilization and transduction - PubMed
https://pubmed.ncbi.nlm.nih.gov/26578770/
[7] [9] [10] [11] [52] [59] Targeting influenza A virus by splicing inhibitor herboxidiene reveals the importance of subtype-specific signatures around splice sites | Journal of Biomedical Science | Full Text
https://jbiomedsci.biomedcentral.com/articles/10.1186/s12929-023-00897-4
[12] [65] Influenza A M2 recruits M1 to the plasma membrane: A fluorescence fluctuation microscopy study - PubMed
https://pubmed.ncbi.nlm.nih.gov/34808098/
[13] [14] [15] [16] [17] [18] [19] [20] [21] [23] [44] [48] [49] [50] [54] M2e-Based Universal Influenza A Vaccines
https://www.mdpi.com/2076-393X/3/1/105
[22] MARCH8 inhibits influenza A virus infection by targeting viral M2 ...
https://www.nature.com/articles/s41467-021-24724-2
[46] A broad protection provided by matrix protein 2 (M2) of avian ...
https://www.sciencedirect.com/science/article/abs/pii/S0264410X15006970
[47] ConSurf color-coded multiple sequence alignment of M2 family ...
[51] Investigation of a recent rise of dual amantadine-resistance ...
https://bmcgenomdata.biomedcentral.com/articles/10.1186/1471-2156-16-S2-S3
[53] Susceptibility of highly pathogenic A(H5N1) avian influenza viruses ...
https://pubmed.ncbi.nlm.nih.gov/17112602/
[55] Bat Influenza M2 Shows Functions Similar to Those of Classical ...
https://pmc.ncbi.nlm.nih.gov/articles/PMC12195650/
[56] The M2 proteins of bat influenza A viruses reveal atypical features ...
https://journals.asm.org/doi/abs/10.1128/jvi.00388-23
[61] The M2 proteins of bat influenza A viruses reveal atypical features ...