Keywords

Introduction

Crossing the membrane is a significant physical barrier for viruses, and different strategies are employed to facilitate this process. The newly synthesized components of influenza A virus (IAV) in the infected host cell will come together and organize laterally in a region of the plasma membrane called the “budozone” (Fig. 16.1) (Leser and Lamb 2017). After this assembly occurs, the budozone becomes the site from which budding occurs. Importantly, the lipid composition in this region differs markedly from the average stoichiometric ratios found in the host cell as it is enriched in cholesterol and certain lipids (e.g., saturated phospholipids and sphingolipids), which help drive the assembly of viral components at the phase boundary of lipid microdomains and alter membrane biomechanics (Fig. 16.1). Some of the viral proteins (e.g., hemagglutinin (HA) and nucleoprotein) appear to concentrate spontaneously in the cholesterol-rich membrane environment whereas matrix protein 2 (M2) association with the budozone materializes in a fashion dependent on interactions with other proteins (matrix protein 1 (M1) and HA in the cluster (Leser and Lamb 2017). Notably, the lateral enrichment of certain lipid components persists through the entire process of budding and virus release, making the emerging viral envelope likewise enriched in cholesterol and sphingolipids (Lenard and Compans 1974; Nayak and Barman 2002). Lateral heterogeneity of lipids and (host as well as viral) proteins enable thermodynamically or dynamically stabilized microdomains of functional relevance to exist under the control of cellular regulatory processes (Mouritsen and Jorgensen 1992; Mouritsen 2016). The popular term membrane/lipid “raft” is frequently used to signify such structures of the membrane. Rafts appear to be an essential platform for viral and host components to interact with one another.

Fig. 16.1
A schematic of the membrane of an infected cell. It has raised rod-shaped lipid rafts with round and square-shaped H A and N A on their surface. These lipid rafts are surrounded by multiple tiny, dot-shaped M 2.

“Overview of the budding of influenza viruses, showing the coalescence of HA and NA containing lipid rafts (shown in yellow), the formation of a filamentous virion and membrane scission caused by M2 clustered at the neck of the budding virus.” Reprinted from Rossman and Lamb (2011), Copyright 2011, with permission from Elsevier

Full size image

Viruses are compact and small by necessity. Their genetic material will typically contain a minimal set of genes encoding proteins strictly needed for functions they cannot hijack from the host cell. Therefore, the scission of enveloped viruses, the process in which the budding virus cuts the membrane in order to release from the host, is normally performed by using the endosomal sorting complex required for transport (ESCRT) machinery belonging to the host (Votteler and Sundquist 2013). Compelling evidence has led to the proposal that IAV is an exception to this principle. Microscopy data show that M2 is localized at the budding neck of the virus. Additionally, evidence from Nuclear magnetic resonance (NMR) and Electron paramagnetic resonance (EPR) spectroscopy, as well as computational studies, supports this hypothesis at the molecular level, although there are still many unresolved aspects.

This chapter will describe the roles and interactions between two of the key biomolecular components, cholesterol and M2, of the IAV-infected host cell lipid environment during budding and scission. Cholesterol is important in facilitating infection, replication, and budding of IAV (Sun and Whittaker 2003). M2 is important because of its evident role in facilitating the scission of the membrane to complete budding (Rossman et al. 2010b) in addition to its essential role as a proton channel during viral entry (Lamb et al. 1994; Helenius 1992; Pinto et al. 1992). We will set the frame by sketching out the key players and steps of IAV budding and scission (see below) and work toward a contemporary hypothesis for the roles of cholesterol and M2 in budding and scission based on recent discoveries. The essential biophysical techniques used to uncover the knowledge we have in this area are listed in the following section.

Methods

Experimental Techniques

Microscopic techniques are used to visualize objects. Cells, microbes, and viruses are typically studied using light, fluorescence, or electron microscopy (Harry et al. 1995; Swanson et al. 1969; Goldsmith and Miller 2009; Scaffidi et al. 2021). Light microscopy illuminates the sample with visible light and uses arrangements of optical lenses to magnify the image. Fluorescence microscopy uses a high-intensity light source to excite a molecular tag (the fluorophore) attached to components of the sample (Fig. 16.2). The tag reemits light on a different wavelength once the excitation relaxes to a low-energy state, allowing for image magnification and effective resolution beyond what is possible with conventional light microscopy. This technique is also especially suited for highlighting certain components in order to see where they localize, including specific organelles of a cell or even individual proteins. In electron microscopy, on the other hand, the sample is exposed to a beam of accelerated electrons. The electron beam has a wavelength of around five orders of magnitude shorter than the photons of visible light and can therefore resolve much smaller structures. Often the technique of immunogold labeling is used in conjunction with electron microscopy to determine the localization of specific components. This labeling works by attaching gold particles to a secondary antibody that will bind to a primary antibody, which itself attaches to the target component (e.g., protein) (Slot and Geuze 1985; Leser et al. 1996). The high electron density of the gold particle gives rise to high levels of electron scattering, making them appear as distinct dark spots in the image. However, in contrast to light microscopy techniques, imaging by electron microscopy requires fixation and positive staining of the sample, preventing the examination of living samples (Leser et al. 1996).

Fig. 16.2
5 schematics and 1 fluorescence scan image. c and d. Two schematics of an electroformation chamber. The labeled parts are copper tape, vacuum grease, spacer, and internal solution, respectively. e. In the fluorescence scan image, the vesicles appear circular in shape and are densely packed.

Vesicle formation “(a) Deposition of lipid droplets onto the electrode surface. (b) Evaporation of organic solvent under vacuum. (c) Construction of the electroformation chamber. (d) Electroformation chamber filled with an internal solution and connected to an alternating current function generator. (e) An image of fluorescently labeled giant unilamellar vesicles (GUVs) obtained using fluorescence microscopy. The scale bar denotes 50 μm.” Reprinted from Membranes, Vol 11:11, Zvonimir Boban, Ivan Mardešić, Witold Karol Subczynski, Marija Raguz, Giant Unilamellar Vesicle Electroformation: What to Use, What to Avoid, and How to Quantify the Results, Article 860, Copyright 2021, under CC BY 4.0 license from MDPI (Boban et al. 2021)

Full size image

In spectroscopic methods such as NMR and EPR, the goal is to measure quantities that will reveal properties of the sample by radiating it, for instance, with a strong magnetic field. The magnet excites the behavior of the atomic nuclei (in NMR) on the electron (in EPR). The results are quantifiable spectral lines or energies, which will often allow for making inferences about correlations or distances between components of the sample, for example, the distance between two amino acid residues of a protein (Klug and Feix 2008; Marion 2013; Fan and Lane 2016).

Theoretical Modeling and Simulation

The structural resolving of biomolecular components by experimental means such as X-ray crystallography, NMR, or cryo-electron microscopy enables interrogation of the molecules and their interplay using computational methods. Molecular dynamics (MD) simulation, affectionately called “the computational microscope,” is now routinely employed due to a relatively low entry barrier, both monetary and technical, and because the method enables investigation with unparalleled spatiotemporal resolution. MD simulations are used to study the behavior of complex molecular systems over time. A minimal, well-defined system called the simulation “box” containing all desired components (proteins, nucleic acids, lipids, ions, water, etc.) is constructed (Fig. 16.3). The interactions between all atoms in the system are described by physics-based potentials collectively known as the “force field” and the time evolution of the system can be simulated by integrating Newton’s equations of motion for the atoms in the system subject to the force field. From these data, a wide range of properties can be computed, including structural properties (e.g., distribution of distances and angles between atoms, and the organization of molecules), thermodynamic properties (e.g., temperature, entropy, and heat capacity), and kinetic properties (e.g., rates of reactions and transformations) (Allen and Tildesley 1989; Frenkel et al. 2002). For an accessible account of the more detailed aspects of the MD technique to nonexperts, we can refer to our recent review article (Ohkubo and Madsen 2021). In general, modern force fields demonstrate satisfactory accuracy, and there is a growing recognition of limitations in certain situations, such as those pertaining to cholesterol in lipid mixtures (Javanainen et al. 2023).

Fig. 16.3
An illustrative schematic presents a workflow for molecular dynamics simulation. The steps of the workflow are starting, preparation of simulation box, run, and analyze.

The general workflow of molecular dynamics simulation. A model structure of the M2 protein is selected together with the initial system setup configuration. We show a ribbon diagram of M2 resolved by solid-state NMR (PDB ID: 2L0J) (Sharma et al. 2010) embedded in a phospholipid bilayer consisting of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) and cholesterol. Water and ions fill the remainder of the box. The system is assembled, for example, using the CHARMM-GUI tool (Jo et al. 2008; Wu et al. 2014). From here, Newton’s equations of motion are solved, propagating the dynamics until the system’s properties stabilize over time. The resulting data are analyzed for the desired properties. Visualizations of the protein structure and simulation box components are created with VMD (Humphrey et al. 1996). Graph data graph is plotted with Matplotlib (Hunter 2007). The Fig. is created, in part, using BioRender (https://BioRender.com)

Full size image

In certain studies, “coarse-grained” computational techniques are employed, which refers to methods where the individual biomolecular constituents of the simulation are represented at a lower level of detail than the atomistic level. These techniques can range from intermediate resolution methods (e.g., the Martini force field (Marrink et al. 2007)) to highly coarse-grained techniques (e.g., elastic network models (Madsen et al. 2017; Grime and Madsen 2019)) and are used to accelerate computational speed, enabling the exploration of larger systems over longer time scales. However, it is important to note that using such methods can exacerbate inaccuracies in the system (Ingolfsson et al. 2014; Jarin et al. 2021; Pak et al. 2019). Comprehensive validation to ensure that the results are reliable is required for any application of theoretical modeling and simulation.

Biomolecules and Interactions Driving Budding and Scission

IAV proteins and RNA interact intricately with each other and with components of the host cell in and around the plasma membrane in a complex, multistep process to assemble and bud-off progeny viruses (Lakadamyali et al. 2004; Noda and Kawaoka 2010). The key players in this process are the viral protein HA, neuraminidase (NA), M1, M2, the ribonucleoprotein (RNP) complex, and the arena of interaction at the heterogeneous plasma membrane environment. In this section, we review the fundamentals of budding and scission of IAV.

Initiation of Budding

Viral proteins HA and NA are transported to the plasma membrane where they cluster with cholesterol, sphingolipids, and saturated phospholipids, forming the budozone, a heterogeneous lipid domain on the plasma membrane where the virus assembles (Fig. 16.4A) (Leser and Lamb 2017). The budozone region serves as a scaffold for the recruitment of other viral proteins and the viral RNA. It is thought that this initial clustering and lipid domain formation initiates the process of viral budding by creating a lipid phase boundary, resulting in a local change in membrane curvature (Fig. 16.4A) (Wang et al. 2012). This change in curvature leads to bud formation, which subsequently grows and matures into a new virus particle in a process driven by the mutual interactions between the viral proteins (in particular, M1 polymerization), host cell proteins, and the plasma membrane.

Fig. 16.4
3 illustrations. a and b. An inverted u-shaped bud of the virus has multiple, spoon-shaped H A on its surface. Many helical cytoplasmic tails hang from the groove of the bud.

“Model of influenza virus budding. (a) The initiation of virus budding caused by clustering of HA (shown in red) and NA (shown in orange) in lipid raft domains. M1 (shown in purple) is seen binding to the cytoplasmic tails of HA and NA and serves as a docking site for the vRNPs (shown in yellow). (b) Elongation of the budding virion caused by polymerization of the M1 protein, resulting in a polarized localization of the vRNPs. M2 (shown in blue) is recruited to the periphery of the budding virus though interactions with M1. (c) Membrane scission caused by the insertion of the M2 amphipathic helix at the lipid phase boundary, altering membrane curvature at the neck of the budding virus and leading to release of the budding virus.” Reprinted from Rossman and Lamb (2011), Copyright 2011, with permission from Elsevier

Full size image

Virion Assembly

The cytoplasmic tails of clustered HA and NA, coupled with the lipid environment of the budozone, attract the internal structural components of the virus, in particular, M1 and the RNP complex (Fig. 16.4B). Polymerization of M1 at the budozone is thought to be responsible for the formation and elongation of the nascent virus, which can form spherical or filamentous virion morphologies depending on specific sequence motifs within certain viral proteins (Bourmakina and Garcia-Sastre 2005; Elleman and Barclay 2004; Roberts et al. 1998). The resulting virions contain HA and NA on their surface, a helical layer of polymerized M1 protein underlying the membrane and an internal collection of RNP complexes, each packaging a different segment of the viral genome (Fig. 16.4B) (Waterson et al. 1963; Nermut and Frank 1971; Schulze 1972). The presence of these different virus morphologies can have significant implications for viral budding, entry, host interactions, and may affect the severity of the disease caused by the virus (Taubenberger and Morens 2008; Kalil and Thomas 2019). Despite the formation of fully assembled virus particles, the virions remain attached to the plasma membrane by a small membrane neck that requires scission for their complete release (Fig. 16.4C) (McCown and Pekosz 2006; Rossman et al. 2010b).

Scission of the Budding Neck

The process of IAV membrane scission is thought to involve the action of the viral protein M2, which promotes the formation of curved lipid bilayer structures that can lead to membrane scission (Fig. 16.4C) (Rossman et al. 2010b; Schmidt et al. 2013; Martyna et al. 2017; Wang and Hong 2015). The mechanical forces necessary for scission may be contributed by cholesterol (Hubert et al. 2020; Anderson et al. 2021), which fittingly is also required for the stability of the virus envelope (Lenard and Compans 1974; Nayak and Barman 2002; Bajimaya et al. 2017). During the initial process of viral budding, the M2 protein is conspicuously excluded from the core of the budozone and may not be actively involved in virion formation (Leser and Lamb 2017). The first two steps of budding can even occur in viruses where the M2 protein is deleted or otherwise dysfunctional (Rossman et al. 2010b; Herneisen et al. 2017; McCown and Pekosz 2006). Intriguingly, the process comes to a complete stall in the absence of M2, leaving behind the uncut constricted budding neck from which a secondary budding can occur at the same site. This gives rise to a characteristic phenotype when several unsuccessful budding attempts occur in succession and a “beads-on-a-string” morphology emerges (Fig. 16.5, top panel) (McCown and Pekosz 2006; Rossman et al. 2010b). As virion assembly progresses, the planar budozone becomes folded and incorporated into the emerging virus. The M2 protein, located at the periphery of the budozone is then drawn to the constricted membrane neck of the budding virus (Rossman et al. 2010b). There is evidence that interactions between the M2 amphipathic helix M2(AH), which is a highly conserved region of the protein, and the plasma membrane can alter membrane curvature at the virion neck and are responsible, at least in part, for the role that M2 directly plays in membrane scission (Rossman et al. 2010b; Roberts et al. 2013; Schmidt et al. 2013; Rossman et al. 2010a; Martyna et al. 2017; Wang and Hong 2015; Andreas et al. 2015). During the process of scission, some M2 protein gets incorporated into the newly released virion, generating an infectious virus capable of starting the next round of infection (Fig. 16.5, top panel).

Fig. 16.5
8 micrographs and 5 confocal images. a. In a micrograph labeled A slash Udorn slash 72, the virus appears filamentous in shape over the surface of the cell. b. The virus is bud-shaped in the micrograph labeled A slash Udorn slash 72 M 2.

“The M2 Amphipathic Helix is Necessary for Membrane Scission and Virion Release. MDCK cells were infected with an MOI of 3 pfu/cell of (a) A/Udorn/72” or (b) “A/Udorn/72 M2(AH-Mut) for 18 hr and thin sections were analyzed by electron microscopy. The scale bars indicate 100 nm.” (c) “GUVs electroformed with 30 or 0.5 molar % of cholesterol, with 0.5 mg/ml of lucifer yellow (shown in blue) added to the resuspension buffer, were treated with 10 mM of the indicated peptide and imaged within 1 hr.” Reprinted from Cell, Vol 142:6, Jeremy S. Rossman, Xianghong Jing, George P. Leser, Robert A. Lamb, Influenza Virus M2 Protein Mediates ESCRT-Independent Membrane Scission, Pages 902–913, Copyright 2010, with permission from Cell Press (Rossman et al. 2010b)

Full size image

A Contemporary Hypothesis on the Roles of Cholesterol and M2

M2 protein can form clusters in the plasma membrane and viral envelope (Sutherland et al. 2022; Madsen et al. 2018; Paulino et al. 2019; Elkins et al. 2017). Clustering is thought to be driven by hydrophobic interaction between M2 transmembrane (TM) domains, as well as the specific arrangements of the amino acids within the protein (Bao et al. 2022). The clustering process is believed to be highly dependent on lipid composition (Sutherland et al. 2022; Elkins et al. 2017; Elkins et al. 2018) and can be influenced by the microenvironment or the action of other proteins (Leser and Lamb 2017). Curvature sorting plays a role in membrane geometries where the curvature is commensurate with the wedge-shaped M2 (Madsen et al. 2018; Martyna et al. 2016; Ho et al. 2016). Proximity among multiple M2 proteins is thought to play a role in the functioning of the protein, with clustering thought to occur in a stepwise, hierarchical fashion (Fig. 16.6) (Sutherland et al. 2022; Chen et al. 2008).

Fig. 16.6
A schematic of a simulation box, 2 illustrations, and 1 micrograph. c. A micrograph exhibits an irregular diffused patch of infected M D C K cells.

“(a) M2 in a planar bilayer that is exposed to a constant, external compressive stress. (Top) (x, y)-plane projected area per lipid in the simulation box. Three time blocks are defined for further analysis, I, II, and III. (Bottom) Top view of protein tracer lines show how proteins gradually come together in the three time blocks (I, II, and III). (b) Final top view and side view snapshots after the membrane significantly deviates from planarity at t = 20 Mτ. Red arrows indicate the analogy with the in vitro system. (c) MDCK cells were infected with 3 pfu per cell of A/Udorn/72 for 18 h before fixation, immunogold labeling of M2, thin sectioning, and analysis by electron microscopy. (Scale bar: 100 nm.)” Reprinted from PNAS, Vol 115:37, Jesper J. Madsen, John M. A. Grime, Jeremy S. Rossman, Gregory A. Voth, Entropic forces drive clustering and spatial localization of influenza A M2 during viral budding, Pages E8595-E8603, with permission from PNAS (Madsen et al. 2018)

Full size image

At the molecular level, M2 protein interacts with cholesterol in several ways to facilitate the budding and release of IAV particles. They are both integrated into the plasma membrane and can therefore interact by direct contact. Cholesterol molecules are embedded within the bilayer membrane, though typically asymmetrically distributed among the two leaflets in a nonstatic fashion that is subject to regulation by cellular processes (Wood et al. 2011). The functional consequences of cholesterol asymmetry likely include changes in membrane fluidity, alterations in lipid domains, lateral and transbilayer (“flip-flop”) diffusion, lipid packing, and perhaps even the function of some proteins (Wood et al. 2011). The M2 protein is seen to associate with cholesterol-enriched rafts in the plasma membrane primarily through association with other proteins such as M1 (Leser and Lamb 2017). Since the transmembrane domain of M2 traverses the membrane, it is tempting to speculate that cholesterol molecules can get in direct contact with this region of the M2 protein. Support for this idea comes from sequence analysis of M2, which reveals a cholesterol recognition/interaction amino acid consensus (CRAC) sequence motif thought to be involved with cholesterol recognition in other proteins, as well as the observation that M2 co-purifies with cholesterol (Schroeder et al. 2005). However, extensive mutagenesis studies have shown that the association between cholesterol and M2, and the proper functioning of M2 in facilitating virus replication, are not affected when the CRAC motif is deleted (Thaa et al. 2012; Stewart et al. 2010; Thaa et al. 2014; Thaa et al. 2011). In fact, some strains of IAV do not contain a M2 CRAC motif at all. We must therefore conclude that the CRAC motif is of no major functional importance in this context. Instead, the way that cholesterol and the M2 protein interact is likely through their ability to participate in lipid rafts or through the affinity of the M2 TM domain with cholesterol, positioning it to facilitate scission (Fig. 16.6) (Pan et al. 2019; Kolokouris et al. 2021; Rossman and Lamb 2010; Elkins et al. 2017). The presence of cholesterol in the plasma membrane might further have an impact on the structure and orientation of the M2 protein (Ekanayake et al. 2016; Wright et al. 2022; Claridge et al. 2013). Cholesterol has been shown to increase the insertion depth of the M2 protein in the lipid bilayer, bringing it closer to the hydrophobic core of the bilayer (Elkins et al. 2017; Martyna et al. 2020).

The key to elucidating the interplay between cholesterol and M2 in budding and scission will be to understand how their individual effects on membrane fluidity and mechanical properties add up when they colocalize in the budding neck. Is the sum greater than the parts, and how does the presence of M2 in low copy number (14 to 68 molecules per virion (Zebedee and Lamb 1988)) tip the balance to facilitate membrane scission? There is evidence that lipid phase separation phenomena and membrane scission are closely related (Roux et al. 2005; Ryu et al. 2014). Lipid phase separation refers to the spontaneous segregation of different species in the plasma membrane into distinct domains, which can form spontaneously due to differences in lipid composition or physical properties such as size, shape, and charge. We remark that this is distinct from the occurrence of lipid rafts, which are dynamic in nature and extend over much smaller spatial and temporal scales (and therefore cannot be resolved by microscopic methods) (Rajendran and Simons 2005; Lozano et al. 2016). The phase separation process results in the formation of nonuniform lipid structures, including liquid-ordered and liquid-disordered domains (Heberle and Feigenson 2011). The interface between the distinct phases is the focal point of mechanical forces such as tension, shear, and bending stresses potentially causing membrane scission (Liu et al. 2006). A variety of factors may contribute to membrane scission: Changes in lipid composition, to which both cholesterol and M2 protein contribute, changes in mechanical stress, likely caused by lipid phase boundary dynamics, and the action of additional specific proteins. Lipid raft structures have also been proposed to play a role in scission (Schroeder et al. 2005; Barman and Nayak 2007).

The AH domain within M2 acts as a membrane-sculpting tool (Huang et al. 2015). An AH is characterized by a Janus-faced physical/chemical surface consisting of both hydrophobic and hydrophilic amino acids residues concentrated on opposing sides of the α-helix structure. This arrangement allows the AH to insert into the plasma membrane right at the hydrophilic/hydrophobic interface region so that the hydrophobic residues interact with the lipid tails while the hydrophilic residues interact with the aqueous environment. The M2AH can facilitate budding and scission by inducing changes, primarily in composition, curvature, and stability, that help generate the appropriate lipid environment for budding and membrane scission to occur in model membrane systems, even in the absence of the other M2 domains (Fig. 16.5, lower panel) (Rossman et al. 2010b; Roberts et al. 2013; Martyna et al. 2017; Martyna et al. 2020). The effect appears most dependent on the chemical/physical properties of the AH domain and not dependent on the conserved amino acid motif per se, as one might have been tempted to assume (Hu et al. 2020). Furthermore, the conformation of the M2 protein is pH-dependent and can be altered when there are changes in pH or lipid composition, particularly when cholesterol is present, an effect that is likely facilitated by the protein’s AH domain (Torabifard et al. 2020; Liao et al. 2013; Liao et al. 2015; Kim et al. 2015; Thomaston et al. 2013; Saotome et al. 2015; Martyna et al. 2020; Kwon et al. 2015; Kyaw et al. 2023). While proteins, in general, can adapt conformations to suit different lipid environments, the opposite may be an important additional factor in some situations (Bacle et al. 2021).

Several other proteins and cellular processes have been shown to play a role in the processes of budding and membrane scission mediated by cholesterol and the M2 protein of IAV (Mao et al. 2022; Bruce et al. 2010; Han et al. 2021; Zhu et al. 2017; Wohlgemuth et al. 2018; Beale et al. 2014; Bhowmick et al. 2017; Vahey and Fletcher 2019). It is important to note that the exact roles of these factors are not fully understood, and additional research is needed to reveal the complex network of interactions that drive these processes.

Conclusion

The sophisticated interplay between cholesterol and the viral M2 protein plays a crucial role in multiple stages of the influenza virus life cycle. Cholesterol is essential for the formation of membrane raft microdomains that serve as virus assembly sites and contribute to the mechanical forces necessary for membrane scission during the last stage of virus budding. Furthermore, cholesterol helps to maintain the stability of the virus envelope. M2 protein also plays a critical role in influenza virus budding by promoting the formation of curved lipid bilayer structures that facilitate the process of membrane scission. The interaction between cholesterol and the M2 protein affects the local lipid composition, fluidity, and stability of the membrane, thus contributing to the efficient and proper release of newly formed influenza viruses from virus-infected cells.

In closing, investigation of the intricacies of the mechanistic steps of the influenza virus life cycle is critical because of the ongoing threat of influenza epidemics and pandemics. The knowledge gained from studying the interplay between cholesterol, the viral M2 protein and other associated components in influenza virus budding can contribute to the development of novel antiviral strategies and strengthen our efforts to combat viral infections and protect public health.