Amphipathic Helices of Cellular Proteins Can Replace the Helix in M2 of Influenza A Virus with Only Small Effects on Virus Replication.

M2 of influenza virus functions as a proton channel during virus entry. In addition, an amphipathic helix in its cytoplasmic tail plays a role during budding. It targets M2 to the assembly site where it inserts into the inner membrane leaflet to induce curvature that causes virus scission. Since vesicularization of membranes can be performed by a variety of amphiphilic peptides, we used reverse genetics to investigate whether the peptides can substitute for M2's helix. Virus could not be generated if M2's helix was deleted or replaced by a peptide predicted not to form an amphiphilic helix. In contrast, viruses could be rescued if the M2 helix was exchanged by helices known to induce membrane curvature. Infectious virus titers were marginally reduced if M2 contains the helix of the amphipathic lipid packing sensor from the Epsin N-terminal homology domain or the nonnatural membrane inducer RW16. Transmission electron microscopy of infected cells did not reveal unequivocal evidence that virus budding or membrane scission was disturbed in any of the mutants. Instead, individual virus mutants exhibit other defects in M2, such as reduced surface expression, incorporation into virus particles, and ion channel activity. The protein composition and specific infectivity were also altered for mutant virions. We conclude that the presence of an amphiphilic helix in M2 is essential for virus replication but that other helices can replace its basic (curvature-inducing) function.IMPORTANCE Influenza virus is unique among enveloped viruses since it does not rely on the cellular ESCRT machinery for budding. Instead, viruses encode their own scission machine, the M2 protein. M2 is targeted to the edge of the viral assembly site, where it inserts an amphiphilic helix into the membrane to induce curvature. Cellular proteins utilize a similar mechanism for scission of vesicles. We show that the helix of M2 can be replaced by helices from cellular proteins with only small effects on virus replication. No evidence was obtained that budding is disturbed, but individual mutants exhibit other defects in M2 that explain the reduced virus titers. In contrast, no virus could be generated if the helix of M2 is deleted or replaced by irrelevant sequences. These experiments support the concept that M2 requires an amphiphilic helix to induce membrane curvature, but its biophysical properties are more important than the amino acid sequence.

Two Cytoplasmic Acylation Sites and an Adjacent Hydrophobic Residue, but No Other Conserved Amino Acids in the Cytoplasmic Tail of HA from Influenza A Virus Are Crucial for Virus Replication.

Recruitment of the matrix protein M1 to the assembly site of the influenza virus is thought to be mediated by interactions with the cytoplasmic tail of hemagglutinin (HA). Based on a comprehensive sequence comparison of all sequences present in the database, we analyzed the effect of mutating conserved residues in the cytosol-facing part of the transmembrane region and cytoplasmic tail of HA (A/WSN/33 (H1N1) strain) on virus replication and morphology of virions. Removal of the two cytoplasmic acylation sites and substitution of a neighboring isoleucine by glutamine prevented rescue of infectious virions. In contrast, a conservative exchange of the same isoleucine, non-conservative exchanges of glycine and glutamine, deletion of the acylation site at the end of the transmembrane region and shifting it into the tail did not affect virus morphology and had only subtle effects on virus growth and on the incorporation of M1 and Ribo-Nucleoprotein Particles (RNPs). Thus, assuming that essential amino acids are conserved between HA subtypes we suggest that, besides the two cytoplasmic acylation sites (including adjacent hydrophobic residues), no other amino acids in the cytoplasmic tail of HA are indispensable for virus assembly and budding.

S-acylation of influenza virus proteins: Are enzymes for fatty acid attachment promising drug targets?

Covalent attachment of saturated fatty acids (palmitate and stearate) to hemagglutinin (HA) of influenza virus is a protein modification essential for viral replication. The enzymes catalysing acylation of viral proteins have not been identified, but likely candidates that acylate cellular substrates are members of a protein family that contain a DHHC (Asp-His-His-Cys) cysteine-rich domain. Since 23 DHHC-proteins with distinct, only partly overlapping substrate specificities are present in humans, only a few of them might acylate HA in airway cells of the lung. We argue here that these DHHC-proteins might be promising drug targets since their blockade should result in suppression of viral replication, while acylation of cellular proteins will not be (or very little) compromised.

Acylation and cholesterol binding are not required for targeting of influenza A virus M2 protein to the hemagglutinin-defined budozone.

Influenza virus assembles in the budozone, a cholesterol-/sphingolipid-enriched ("raft") domain at the apical plasma membrane, organized by hemagglutinin (HA). The viral protein M2 localizes to the budozone edge for virus particle scission. This was proposed to depend on acylation and cholesterol binding. We show that M2-GFP without these motifs is still transported apically in polarized cells. Employing FRET, we determined that clustering between HA and M2 is reduced upon disruption of HA's raft-association features (acylation, transmembranous VIL motif), but remains unchanged with M2 lacking acylation and/or cholesterol-binding sites. The motifs are thus irrelevant for M2 targeting in cells.

A bipartite S unit of an ECF-type cobalt transporter.

ECF-class transporters comprise abundant importers for micronutrients such as vitamins and transition-metal ions, and for intermediates of salvage pathways in bacteria and archaea. They are composed of ABC ATPases (A units), a conserved transmembrane protein (T unit) and a substrate-specific transmembrane protein (S unit or core transporter). Here we analyzed the function of an ECF-type Co(2+) transporter (CbiMNQO) and, in particular, the derived bipartite S unit CbiMN. CbiMN was characterized as the minimal unit that functions as a Co(2+) transporter. Neither the solitary CbiM nor a tripartite CbiMQO complex was active, indicating an essential role for CbiN. CbiN was loosely bound in CbiMNQO and CbiMN complexes, and did not copurify with its partners. Generating a contiguous reading frame resulted in a Cbi(MN) fusion protein that displayed Co(2+)-transport activity and interacted with CbiQO in vivo. Sixteen variants of Cbi(MN) with modifications in the strongly conserved N-terminal stretch of ten amino-acid residues were constructed and analyzed for transport activity. The results indicate that the length and sequence of this region are critical for functioning of the core transporter. Specifically, they point to essential roles of His2 and the distance of His2 to the amino group of the peptide chain in metal recognition.