Influenza viruses contain two palmitoylated (S-acylated) proteins: the major spike protein HA (haemagglutinin) and the proton-channel M2. The present review describes the fundamental biochemistry of palmitoylation of HA: the location of palmitoylation sites and the fatty acid species bound to HA. Finally, the functional consequences of palmitoylation of HA and M2 are discussed regarding association with membrane rafts, entry of viruses into target cells by HA-mediated membrane fusion as well as the release of newly assembled virus particles from infected cells.
- haemagglutinin (HA)
- influenza virus
- membrane fusion
- virus taxonomy
Palmitoylated proteins of influenza virus
Influenza viruses are enveloped viruses found in the Orthomyxoviridae family. Their membrane is lined from beneath by the matrix protein M1, which in turn envelopes the viral genome. In influenza A and B viruses there are two viral spikes embedded in the envelope: HA (haemagglutinin), which catalyses virus entry by binding to sialic acid moieties present on the host cell surface and by performing fusion of viral with endosomal membranes and NA (neuraminidase), which is required for the release of virus particles by removing potential receptors from infected cells. In the influenza C virus, all three activities (receptor-binding and -destroying, and membrane fusion) are combined in one spike, which is designated HEF (HA-esterase fusion glycoprotein). Virus particles also contain minor amounts of a proton channel, which is called M2 in influenza A virus and BM2 and CM2 in influenza B and C virus.
HA and HEF, as well as M2 and CM2, are palmitoylated at cytoplasmic and transmembrane cysteine residues, whereas the other viral proteins lack any lipid modifications. This review describes the biochemistry of HA acylation and how the modification affects targeting of HA and M2 to rafts and (probably as a consequence) virus budding and virus entry. Several recent reviews cover related topics, such as palmitoylation of other viral proteins , raft association of influenza virus proteins , budding of influenza virus particles  and HA-catalysed membrane fusion .
Fatty acid species bound to HA and HEF
HA (as seen in Figure 1) and HEF are trimeric type I transmembrane glycoproteins with an N-terminal signal peptide, a large ectodomain, a single TMR (transmembrane region) and a short cytoplasmic tail. HA from influenza A is highly variable; 17 antigenic subtypes have been isolated so far, whereas HA from influenza B virus and especially HEF are less variable. HA from influenza A virus is acylated at three cysteine residues, two are located in the cytoplasmic domain and one at the end of the TMR, whereas a cysteine present in some subtypes in the middle of the TMR is not acylated [5–7]. H11 subtype HA contains an additional cytoplasmic cysteine, which is also used as acylation site  (Table 1). HA from influenza B virus is palmitoylated at two cysteine residues in its cytoplasmic tail and HEF, which contains a very short cytoplasmic tail, is palmitoylated at a single cysteine at the end of the TMR [9,10].
It has been shown early on that [3H]palmitic acid, which was used to identify acylated proteins, can be converted into other fatty acid species of different chain length, these are then attached to the acylprotein. However, it was not known for proteins having multiple acylation sites whether each site contains the same proportion of fatty acids or whether specific cysteine residues are acylated predominantly with a particular carbon chain. Advancements in MS have recently allowed for quantitative analysis of fatty acid species linked to individual acylation sites (reviewed in ). The increase in mass of an acylated protein compared with the unmodified protein is rather small, therefore only fragments of viral spikes can be accurately measured. Purified virus particles are digested with proteases, which remove the ectodomain from the spikes, and the hydrophobic membrane-anchoring fragments are extracted with chloroform/methanol and directly analysed by MS. Tandem-MS sequencing of peptides can be used to both prove their amino acid sequence and identify the acylated cysteine residues .
These studies revealed a striking difference between HA of influenza B virus that contains 97% palmitate and HEF of influenza C virus that is predominantly (88%) stearoylated (Table 1), confirming previous data obtained by less sophisticated methods . HAs of the influenza A virus contain a mixture of palmitate and stearate. MS analysis of recombinant viruses with deletions of individual cysteine residues, as well as tandem-MS sequencing, revealed the surprising result that stearate is exclusively attached to the cysteine positioned in the TMR of HA .
Since this initial observation , 40 HAs from 14 subtypes have been analysed by MS [8,16]. The percentage of stearate in all HA variants differs from 35% (meaning that each TMR cysteine contains stearate) to 12% (indicating that only one out of three TMR cysteine residues is stearoylated). Interestingly, HAs that are present in virus strains isolated from humans contain less stearate compared with HAs isolated from mammals or birds. One reason might be that other viral proteins, especially M1, which is much less variable than HA, but contains host-specific amino acid substitutions, affect acylation of HA, but we could not detect variability in the stearate content of HA if internal proteins were exchanged between viruses [16a].
What might be the main molecular signal that determines preferential attachment of stearate to a cysteine residue, its sequence context or its location relative to the membrane span? At present, either bioinformatic sequence comparisons have not revealed obvious amino acid peculiarities between all analysed HA sequences that might be responsible for the variations in the amount of HA-bound stearate or they are located outside of the considered C-terminal area. An analysis of H9 subtype HA, which contains only cytoplasmic cysteine residues and very little stearate (4%), suggests the effect of location (Table 1). However, individual amino acids in the vicinity of the TMR cysteine also affect fatty acid selection, since HAs containing the same spacing of cysteine residues differ in the amount of attached stearate. For example, only one amino acid difference, a conservative exchange in the middle of the TMR (highlighted in Table 1), exists between the anchoring segments of HA from an equine and human H3, but 30% compared with 18% of stearate was attached to HA of those strains. An even larger difference is observed between human H1 HA (12%) and avian H5 HA (29%), although their cytoplasmic tails and the four amino acids upstream of the TMR cysteine are identical. In addition, two H1 HAs from duck and human strains possessing the same amino acid sequence within the whole anchoring segment showed a considerable variation in their stearate content (18% compared with 12%, Table 1). Alternatively, since the amount of HA [relative to NP (nucleoprotein) or M1] present in virus particles varies largely between virus strains, it might be that overexpression of HA might saturate the stearoylation machinery of the cell if HA is expressed in very high amounts and then the cysteine residues in the TMR become palmitoylated.
The specific attachment of stearate leads to speculation on the enzymology of acylation of HA. Members of the DHHC family, polytopic membrane proteins containing a DHHC (Asp-His-His-Cys) motif within one of their cytoplasmic domains, were shown to palmitoylate cellular proteins , but a DHHC protein that acylates influenza HA (or other viral proteins) has not been identified. Alternatively, S-acylation of proteins can occur via a non-enzymatic mechanism. Some cellular proteins are palmitoylated, at least in vitro, at authentic sites in the absence of any enzyme source [18,19]. In addition, peptides composed of unnatural β- or D-amino acids are palmitoylated on microinjection into cells indicating that the hallmarks of classical enzyme reactions, substrate recognition and specificity do not exist .
The site-specific attachment of stearate to HA argues against a purely non-enzymatic reaction mechanism. Acylation without an enzyme would not show any preference for a particular fatty acid, but should reflect the concentration of individual acyl-CoAs present in the membrane where acylation occurs. Since HEF and HA are both acylated at a late endoplasmic reticulum or early Golgi compartment [10,21], it is more likely that individual enzymes exist that differ in their acyl-CoA specificities, as recently demonstrated for DHHC 2 and 3 . One can imagine that the active site of a DHHC protein with a preference for stearoyl-CoA might penetrate deeper into the membrane to attach stearate to a transmembrane cysteine compared with an enzyme with specificity for palmitoyl-CoA.
Further MS analysis revealed that other viral glycoproteins, such as F of Newcastle disease virus and E1 of Semliki Forest virus, are also acylated with stearate at a transmembrane cysteine . Thus site-specific attachment of stearate is a common feature of viral spike proteins and, since viruses hijack the cellular acylation machinery, it also probably occurs on surface receptors of the cell.
Function of S-acylation of HA
HA is a highly variable molecule with very low amino acid conservation (≈20%) through all subtypes. However, comparison of all HA sequences present in the NCBI (National Center for Biotechnology Information) database (≈17000) showed that each molecule contains three, some even four, cysteine residues located in the cytoplasmic domain and at the cytosol-facing end of the TMR. Thus, assuming these cysteine residues (and the attached fatty acids) would not play an essential role for the life cycle of influenza viruses, it is hard to conceive why they have not been exchanged by similar amino acids during evolution of influenza viruses. Indeed, it was impossible to generate influenza virus mutants with two or three HA palmitoylation sites removed, implying that this modification is essential for virus growth [24–26].
S-acylation of HA as raft-targeting signal
Palmitoylation is one of the best characterized signals for association of proteins with rafts, heterogeneous lipid-assemblies enriched in cholesterol and sphingolipids . Raft association of a protein is often deduced from the partitioning of a protein into DRMs (detergent-resistant membranes). Using such assays, it has been shown that deletion of palmitoylation sites from HA reduces their partition into DRMs [24,25].
We used FRET (fluorescence resonance energy transfer) to show that HA, fused at its cytoplasmic tail to CFP (cyan fluorescent protein), clusters with an established marker for inner leaflet rafts, double-acylated yellow fluorescent protein . Furthermore, an artificial HA-derived FRET probe, consisting of a signal peptide, a fluorescent protein and the transmembrane as well as cytoplasmic domain of HA , clusters with a glycolipid-anchored protein, an established marker for rafts of the outer leaflet. For both HA constructs clustering was significantly reduced when all three palmitoylation sites were removed from HA. Removing the palmitoylation sites from HA slightly increased its mobility in the plasma membrane. However, HA diffuses much slower in comparison with the double acylated raft-marker indicating that association of HA with rafts is a dynamic process .
The concept that association of HA with rafts concentrates the protein in the plasma membrane might explain how protein palmitoylation might influence both assembly of virus particles as well as the membrane fusion activity of HA. Expression of HA at the plasma membrane organizes a large raft domain [30,31], that is thought to provide a platform for virus assembly and budding , enriching the viral components to facilitate their interactions. Likewise, the membrane fusion activity of HA critically depends on its surface density. Hence, clustering of HA in rafts might yield a concentration high enough to support fusion.
Interestingly, HEF of influenza C virus is completely soluble in cold detergent . Thus S-acylation is not sufficient to cause raft-localization of a viral glycoprotein. Even the attachment of several fatty acids does not necessarily cause raft-localization of a spike protein. The E1/E2 heterotrimer of Semliki Forest virus contains 15 covalently linked acyl chains, six of them being stearate, but the spike protein is nevertheless not associated with DRMs .
Involvement of S-acylation of HA in virus assembly
HA seems to be a key player for influenza virus assembly and budding; it organizes the viral budozone, which is a large raft domain [30,31]. If HA is expressed from a plasmid, VLPs (virus-like particles) are released that are morphologically very similar to authentic virions. Co-expression of M1 leads to its inclusion into VLPs, but it is much less efficient if the cytoplasmic tail or the acylation sites were deleted from HA . In the context of virus infection, viruses lacking this part of HA were found to have severe defects in genome packaging as well as irregular morphology . Direct evidence suggesting that palmitoylation of HA is involved in virus release was obtained for H3 subtype HA . Virus particles containing HA with deleted palmitoylation sites revealed defects in replication and incorporated reduced amounts of the internal components NP and M1. Removal of the palmitoylated cytoplasmic cysteine residues had a greater effect on budding compared with exchange of the stearoylated cysteine in the TMR, but the effect was opposite when association with DRMs was analysed. Importantly, exchange of the M1 protein by that of a different influenza virus restored assembly of viruses with non-palmitoylated HA.
Involvement of S-acylation in HA-catalysed membrane fusion
Upon virus entry and exposure to low pH levels, HA undergoes a conformational change that catalyses the fusion of viral and endosomal membranes . Membrane fusion assumingly proceeds via a hemifusion stage, where only the outer leaflets of the two fusing membranes are connected. Then a fusion pore in the membrane opens, flickers and ultimately dilates. The hemifusion stage is characterized by mixing lipids between the two membranes, solutes of the two compartments are not exchanged before fusion is complete.
Lipid mixing was not impaired in cells expressing non-acylated HA from all influenza virus strains examined suggesting that palmitoylation is not required for hemifusion. In accordance, glycolipid-anchored HA catalyses hemifusion, but not full fusion. This indicates that the TMR and the cytoplasmic tail are required for this process .
Inconsistent data have been published concerning the effect of acylation site removal on the ability of HA to cause full fusion. It was reported that non-acylated H2 and H3 subtype HA mediate cell–cell fusion or show unperturbed transfer of aqueous dyes into HA-expressing cells [6,24]. In contrast, non-acylated HA mutants from the H1, H7, another HA from H2 subtype and from influenza B virus show impaired fusion pore or syncytium formation [5,9,25,37]. Usually, removal of cytoplasmic palmitoylation sites, especially the most C-terminal one, which is surrounded by conserved hydrophobic residues, had a more severe effect on fusion compared with stearoylation of the transmembrane cysteine.
S-acylation of M2 and its function in virus budding
M2 (purple structure in Figure 1), is a tetrameric proton channel activated by acidic pH levels, the action of which is important for genome unpacking during virus entry. In each monomer, the first 24 amino acids form the unglycosylated ectodomain, the following 19 residues are the TMR, the remaining 54 residues build up the cytoplasmic tail. The sequence immediately following the TMR shapes a membrane-parallel amphiphilic helix. A cysteine residue in the helix is post-translationally acylated, primarily with palmitate and probably in the same compartment as HA [38,39]. In addition, the helix contains CRAC (cholesterol recognition/interaction amino acid consensus) motifs . CM2 from influenza C virus is also palmitoylated within a (predicted) amphiphilic region  that, however, does not contain cholesterol-binding motifs. BM2 from the influenza B virus is neither palmitoylated nor does bioinformatic analysis (using http://heliquest.ipmc.cnrs.fr/) predict an amphiphilic helix in the cytoplasmic tail.
Previously, M2 has been implicated in the ultimate step in virus budding: the scission of the virus particle from the plasma membrane. It was hypothesized that S-acylation in concert with binding to cholesterol target M2 to the edge of the viral budding site , a large coalesced raft phase that is organized by HA [30,31]. Complete immersion of the protein in the more ordered, hence thicker, raft domains is thought to be prevented by the relatively short TMR of M2. Thereby M2 becomes located in an ideal position to mediate the scission of virus particles, supposedly triggered by the amphiphilic helix in the cytoplasmic tail of M2, which is inserted into the membrane in a wedge-like manner to induce curvature [42,43].
We recently demonstrated that the cytoplasmic tail of M2 binds to membranes, both in vitro and in cells, and that the membrane-binding properties are modulated by exchange of the acylation site and of two tyrosine residues of the CRAC motif, which mediate cholesterol binding . Thus one biochemical requirement of this model, association of the amphiphilic helix with membranes to (possibly) induce curvature is fulfilled.
Mixed results were obtained when raft localization of M2 was tested experimentally. M2 expressed in transfected cells is not associated with DRMs  and M2 does not interact with the double-acylated marker for inner leaflet rafts in FRET experiments . However, in similar FRET experiments M2 associates with HA , which is present in a large raft domain. Likewise, using immuno-EM (electron microscopy), M2 has been localized to the base of budding filamentous virus particles . In addition, preparation of giant plasma membrane vesicles from cells expressing M2–GFP (green fluorescent protein) showed that the protein is partly present in the coalesced raft phase. Raft-targeting of M2 does not require the CRAC motifs, but is dependent on palmitoylation, similar to cellular proteins that were tested with this model system [27,44]. Although localization of M2 to the edge of a raft could not be directly demonstrated in any experiments, the results suggest that M2 has features of both a raft-associated and a non-raft-associated protein.
Does palmitoylation and cholesterol-binding affect the various activities of M2? Loss of the palmitoylation site does not influence the ion channel activity of M2 . More surprisingly, acylation does not affect the production of virus particles either. In cell culture, recombinant viruses where the acylated cysteine in M2 (or in CM2 of influenza C virus) has been replaced grow as similarly well as the corresponding wild-type virus [41,47], even if acylation is deleted simultaneously with the cholesterol-binding CRAC motif . Moreover, 15% of sequences of natural virus strains present in the NCBI database lack the acylation site in M2 . Intriguingly, however, attenuation of virus infectivity is observed on infection of mice with virus containing non-acylated  or CRAC-disrupted  M2. This indicates that there is a more complex influence of M2 in the context of the infected host that is not accounted for in cell culture experiments.
The joint work carried out in the laboratory of M.V., M.V.S. and L.V.K. is funded by the German Research Foundation [grant number Ve 141/10] and by the Russian Foundation for Basic Research [grant numbers 10-04-91333 and 12-04-01695]. Other work in the laboratory of M.V. is funded by the DFG (Priority Programme 1175, Collaborative Research Centre 740) and by 7th Framework Programme of the European Commission (Marie Curie Initial Training Network ‘Virus-Entry’).
Regulation of Protein Trafficking and Function by Palmitoylation: A Biochemical Society Focused Meeting held at St Anne's College, Oxford, U.K., 23–25 August 2012. Organized and Edited by Luke Chamberlain (University of Strathclyde, U.K.) and Tony Magee (Imperial College London, U.K.).
Abbreviations: CRAC, cholesterol recognition/interaction amino acid consensus; DRM, detergent-resistant membrane; FRET, fluorescence resonance energy transfer; HA, haemagglutinin; HEF, HA-esterase fusion glycoprotein; NA, neuraminidase; NCBI, National Center for Biotechnology Information; NP, nucleoprotein; TMR, transmembrane region; VLP, virus-like particle
- © The Authors Journal compilation © 2013 Biochemical Society