|Home | About | Journals | Submit | Contact Us | Français|
The influenza C virus CM2 protein and a chimeric influenza A virus M2 protein (MCM) containing the CM2 transmembrane domain were assessed for their ability to functionally replace the M2 protein. While all three proteins could alter cytosolic pH to various degrees when expressed from cDNA, only M2 and MCM could at least partially restore infectious virus production to M2-deficient influenza A viruses. The data suggest that while the CM2 ion channel activity is similar to that of M2, sequences in the extracellular and/or cytoplasmic domains play important roles in infectious virus production.
The influenza A virus M2 protein is a homotetrameric, type III integral membrane protein that functions at several stages of the viral life cycle (24). The proton-selective ion channel activity of M2 (4, 18) mediates the release of viral ribonucleoprotein complexes (vRNPs) from matrix (M1) and viral membranes, allowing migration of vRNPs to the nucleus (9, 15, 23), and also raises the pH of the exocytic pathway, thereby preventing premature low-pH-induced conformational changes in hemagglutinin (HA) (5, 26). The cytoplasmic tail of M2 is required for efficient genome packaging into budding virus particles (8, 16, 17). M2 has also been proposed to mediate membrane scission; however, in the absence of M2, infected and transfected cells release virus particles and virus-like particles, respectively (2, 3, 8, 14, 25).
The influenza C virus CM2 protein is an integral membrane protein generated by proteolytic cleavage of an internal signal peptide of the p42 protein (10, 13, 22). CM2 forms disulfide-linked homotetramers (12, 21) and has been shown to conduct Cl− ions (11) or perhaps protons (19). CM2 has been shown to be involved in the release of vRNPs during virus uncoating and in packaging of vRNPs during virus assembly (7). CM2 is also able to raise the pH of the exocytic pathway; however, the role of this activity during influenza C virus replication is currently unclear (1). This study sought to investigate whether full-length CM2 or a chimeric M2 protein containing the CM2 transmembrane domain could functionally substitute for the influenza A virus M2 protein.
CM2 can be expressed efficiently from cDNA using the vaccinia virus-bacteriophage T7 polymerase expression system (21), but expression levels are lower from plasmids that rely on host cell transcription machinery. Efficient plasmid-based expression of CM2 was achieved by deleting the p42 internal signal peptide from the pCAGGS expression plasmid (20), most likely because of the elimination of an mRNA splice donor site present in the RNA corresponding to the signal peptide (Fig. 1A). The glycosylation site was eliminated by mutating Thr at residue 37 to Ala, and epitope tags for the antibodies 3F10 (antihemagglutinin epitope) and 14C2 (anti-M2 epitope) were added at the carboxy terminus in order to facilitate detection of the protein (Fig. 1A). In order to determine if the ion channel activity of CM2 can substitute for that of M2, a construct was generated with the extracellular and cytoplasmic tail of M2 and the transmembrane domain (TM) of CM2 (MCM) (Fig. 1A).
Because some reports show that M2 and CM2 have differential ion channel activities and yet both are able to raise the pH of the secretory pathway (1, 5, 6, 11, 19), the ion channel activities of the two proteins were compared in an assay that measures pH-dependent changes in enhanced yellow fluorescent protein (eYFP) fluorescence (8). Whereas low-pH treatment of 293T cells cotransfected with eYFP and an empty vector yielded no change in mean fluorescence intensity (MFI) over time, low-pH treatment of cells cotransfected with eYFP and M2 from A/Udorn/72 (M2Ud) resulted in an ~40% decrease in MFI (Fig. 1B and C). Low-pH treatment of cells cotransfected with eYFP and CM2 resulted in only an ~25% decrease in MFI (Fig. 1D), indicating that the CM2 protein was able to modulate the cytoplasmic pH of transfected cells but not to the same extent as the M2 protein. The MCM protein induced a reduction in eYFP fluorescence which resembled that of CM2 (Fig. 1E), suggesting that substitution of the CM2 TM for that of M2 conferred ion channel activity that resembled that of CM2. The levels of expression of M2, CM2, and MCM in transfected 293T cells were comparable (Fig. 1C to E), indicating that the decreased pH changes induced by CM2 and MCM compared to M2 were not attributable to differential protein levels but most likely reflected altered ion channel activities.
In order to determine if CM2 and MCM were able to complement an M2-null influenza A virus, clonal MDCK cell lines stably expressing the proteins were generated. Total expression of CM2 and that of MCM are shown relative to cell lines expressing M2Ud or M2 from A/WSN/33 with a mutation conveying amantadine resistance (M2WSN N31S) (Fig. 2A) (27, 28). The expression level of MCM was similar to that of M2, while the CM2 protein expression level was only 2% of the levels of M2. However, even trace amounts of M2 protein can functionally complement an M2-null influenza A virus (17), so the levels of CM2 expression are expected to be sufficient for the complementation assay. MCM was able to form disulfide-linked oligomers, similar to the M2 and CM2 proteins (Fig. 2B). In order to determine if constitutive expression of CM2 or MCM prevented replication of influenza A virus, titers of recombinant A/Udorn/72 (H3N2, rUd) and recombinant A/WSN/33 (H1N1, rWSN) were determined on the cell lines (Fig. 2C). Both rUd and rWSN had similar titers on all cell lines tested, suggesting that expression of CM2 or MCM does not inhibit influenza A virus infection. To assess the ability of CM2 and MCM to substitute for M2, titers of the recombinant M2-null viruses rUd M2Stop and rWSN M2Stop (8, 27) were determined on the cell lines (Fig. 2C). Both rUd M2Stop and rWSN M2Stop showed decreased titers on CM2-expressing cells compared to M2-expressing cells, indicating that CM2 cannot functionally replace M2. In comparing MCM- to M2-expressing cells, rUd M2Stop had similar titers while rWSN M2Stop showed a decrease in titer, demonstrating that the MCM protein can complement rUd M2Stop but not rWSN M2Stop, which suggests that the CM2 ion channel activity can functionally substitute for that of M2 in some influenza A virus strains. However, given the low level of CM2 expression, it remains to be determined if overexpression of the CM2 protein to levels much higher than those needed for M2 complementation can partially complement an M2-null influenza virus.
The ability of MCM to substitute for M2 was also tested in multistep growth curves (Fig. 2D and E). Neither rUd M2Stop nor rWSN M2Stop was able to grow in CM2-expressing cells despite the fact that the level of CM2 expression was higher than that needed to complement these viruses with M2 protein (8, 16, 17). MCM-expressing cells did support growth of both viruses, but the growth of rWSN M2Stop was more attenuated than that of rUd M2Stop. Since CM2 expression could not complement while MCM expression could partially complement two different M2-null viruses, we conclude that that the ion channel activity of CM2 can replace that of M2; however, the ectodomain and/or the cytoplasmic tail of M2 is required during influenza A virus replication.
M2 mutations that cause strain-specific defects in growth of either rUd or rWSN have previously mapped to the M1 protein (2, 8, 16). The ability of MCM expression to complement rUd M-WSN M2Stop (expressing 7 segments from Ud, M1 from WSN, and no M2) growth was comparable to its ability to complement rUd M2Stop (Fig. 2D and F). Likewise, the ability of MCM expression to complement rWSN M-Ud M2Stop (expressing 7 segments from WSN, M1 from Ud, and no M2) growth was comparable to its ability to complement rWSN M2Stop (Fig. 2E and G). Together, these data indicate that the decreased ability of MCM to complement rWSN M2Stop maps to viral sequences outside the M1 protein.
The inability of CM2 to complement M2-null viruses could be due to a defect in particle budding. Virus particles from high-multiplicity-of-infection (MOI) infections were concentrated by ultracentrifugation through 35% sucrose and analyzed by Western blotting and 50% tissue culture infective dose (TCID50) assays (Fig. 3). As previously published, rUd M2Stop virus particles from MDCK cells are released but are defective in the incorporation of full-length NP (NPa) and have lower infectious virus titers relative to total HA (Fig. 3A to C) (8). rWSN M2Stop virus particles grown on MDCK cells also have decreased titers relative to total HA but increased release and cleavage of HA (Fig. 3D and E). Both M2-null viruses grown on CM2-expressing cells showed phenotypes similar to the phenotypes of those grown on MDCK cells (Fig. 3). However, rUd M2Stop virus particles grown on MCM-expressing cells incorporated MCM and did not show the decreased incorporation of NPa observed in virus particles isolated from MDCK and CM2-expressing cells (Fig. 3A to C). rWSN M2Stop virus particles grown on MCM-expressing cells also had structural protein incorporation similar to that of particles from M2-expressing cells. The ratio of titer to HA (a measure of the infectivity of the particles) of viruses grown on MCM-expressing cells was intermediate between that of those grown on MDCK or CM2-expressing cells and that of those grown on M2-expressing cells (Fig. 3C and E). Taken together, these data demonstrate that expression of MCM, but not CM2, is able to complement the defect of NPa incorporation present in viruses grown on MDCK cells but that the particles from MCM-expressing cells still have a defect in infectivity which is more apparent in rWSN M2Stop than in rUd M2Stop.
Negative-stain electron microscopy was performed on the purified rUd M2Stop virus particles in order to assess any morphological differences in virus particles grown on the various cell lines. rUd M2Stop particles grown on MDCK or CM2-expressing cells had a smaller average diameter and higher particle/TCID50 ratios than did those grown on M2-expressing cells (Fig. 4). Virus particles isolated from MCM-expressing cells were of a size similar to the size of those grown on M2-expressing cells but had slightly higher particle/TCID50 ratios (Fig. 4F). All cell lines produced comparable numbers of particles (Fig. 4F). These data indicate that the infectivity of the virus particles correlated with the size of the particles and that MCM, but not CM2, could partially complement the rUd M2Stop virus.
These studies suggest that influenza A virus produced in the absence of M2 or in the presence of CM2 has a defect in the incorporation of NPa (and most likely viral RNA), smaller particles, an increased ratio of total to infectious particles, and absence of growth in multistep growth curves. The level of CM2 expression was higher than the level of M2 expression needed for complementation; however, we cannot rule out the fact that higher levels of CM2 expression might lead to some degree of complementation or that the addition of C-terminal epitope tags may be interfering with CM2 protein function. Virus particles produced in the presence of MCM incorporate a normal level of NPa and are of normal size but have a slightly reduced infectivity. We hypothesize that the defect in infectivity of virus particles grown on MCM-expressing cells is due to the differential ion channel activity of the MCM protein versus that of the M2 protein. The ion channel activity (Fig. 1) shows that MCM does not lower the pH of the cytoplasm as efficiently as does M2, most likely leading to a decreased ability to release vRNPs during virus entry.
We acknowledge Michael Delannoy from the Institute for Basic Biomedical Sciences Microscope Facility at the Johns Hopkins University School of Medicine for assistance with the electron microscopy and the members of the Pekosz laboratory for helpful discussions and suggestions.
This study was supported by the Eliasberg and Marjorie Gilbert Foundations.
Published ahead of print 14 September 2011