Influenza virus M2 protein inhibits epithelial sodium channels by increasing reactive oxygen species

Abstract

The mechanisms by which replicating influenza viruses decrease the expression and function of amiloride-sensitive epithelial sodium channels (ENaCs) have not been elucidated. We show that expression of M2, a transmembrane influenza protein, decreases ENaC membrane levels and amiloride-sensitive currents in both Xenopus oocytes, injected with human α-, β-, and γ-ENaCs, and human airway cells (H441 and A549), which express native ENaCs. Deletion of a 10-aa region within the M2 C terminus prevented 70% of this effect. The M2 ENaC down-regulation occurred at normal pH and was prevented by MG-132, a proteasome and lysosome inhibitor. M2 had no effect on Liddle ENaCs, which have decreased affinity for Nedd4-2. H441 and A549 cells transfected with M2 showed higher levels of reactive oxygen species, as shown by the activation of redox-sensitive dyes. Pretreatment with glutathione ester, which increases intracellular reduced thiol concentrations, or protein kinase C (PKC) inhibitors prevented the deleterious effects of M2 on ENaCs. The data suggest that M2 protein increases steady-state concentrations of reactive oxygen intermediates that simulate PKC and decrease ENaCs by enhancing endocytosis and its subsequent destruction by the proteasome. These novel findings suggest a mechanism for the influenza-induced rhinorrhea and life-threatening alveolar edema in humans.—Lazrak, A., Iles, K. E., Liu, G. Noah, D. L., Noah, J. W., Matalon, S. Influenza virus M2 protein inhibits epithelial sodium channels by increasing reactive oxygen species.

Influenza is a contagious respiratory virus responsible for an estimated 36,000 deaths annually in the United States alone, with a potential 10-fold increase in epidemic and pandemic scenarios. The virion of influenza particles contains a matrix protein (M1) and 3 transmembrane proteins [hemagglutinin (HA), neuraminidase, and a 97-aa transmembrane protein (M2)], with an extracellular N and an intracellular C terminus that form a homotetrameric proton (H⁺) channel, activated at acidic pH. M2 is sparsely distributed throughout the viral lipid envelope, with only 16–20 channels/virion (1).

Beneath the matrix coat is the ribonucleocapsid, which includes the vRNA genome, nucleoprotein, nuclear export protein, and the 3 viral polymerase subunits (1). During infection, viral attachment occurs via binding of HA to sialic acid residues on cell surfaces. Subsequently, viruses fuse with the endosomal membrane and are taken up by the endosomal pathway (2). In the acidic environment of the endosome, histidine 37 in the M2 pore becomes charged, causing tryptophan 41 to rotate, allowing protons to enter and acidify the virion interior. These events promote uncoating of the M1 protein and allow transport of the viral ribonucleoproteins to the cell nucleus (3), where the vRNA transcription and replication occur (1). Inhibition of the M2 channel by amantadine prevents uncoating and viral replication (4). The M2 channel remains an ideal target for antiviral compounds, which potentially may inhibit influenza virus replication at its early stages, thus protecting the host from some of its deleterious effects (such as rhinorrhea and pulmonary edema). However, some circulating human strains (and mouse-adapted strains, A/WSN/33 and A/PR/8/34) are amantadine-resistant.

Lung epithelial cells actively transport Na⁺ ions from the alveolar to the interstitial sides: Na⁺ ions enter the airway and distal lung epithelial cells (type I, type II, and Clara cells), down their electrochemical gradient, through amiloride-sensitive epithelial Na⁺ channels (ENaCs) and cation channels (5,6,7,8), and are transported across the basolateral membrane by the Na⁺,K⁺-ATPase (9). ENaC, a member of the DEG/ENaC channel superfamily, is a heterotrimer composed of ≥3 transmembrane subunits (α, β, and γ), which are expressed in unequal proportions in respiratory epithelia (10). The movement of Na⁺ ions through apical ion channels is the rate-limiting step in transepithelial Na⁺ transport, offering >90% of the resistance to transcellular Na⁺ transport, either in alveolar type (AT) I or ATII cells (11). Active Na⁺ transport limits the degree of alveolar edema after damage to the alveolar epithelium; patients with acute lung injury who are still able to concentrate alveolar protein (as a result of active Na⁺ reabsorption) have a better prognosis than those who cannot (12, 13).

Previous studies have shown that during attachment of inactivated influenza viruses to lung epithelial cells, the binding of HA to cell surface sialic acid residues initiates a series of events leading to activation of protein kinase C (PKC), which in turn down-regulates the activity of ENaCs in mouse tracheal and rat ATII cells (14, 15). In addition, intratracheal instillation of replication-deficient influenza A/PR/8/34 (PR8, H1N1) into the lungs of anesthetized ventilated rats decreased their ability to remove intratracheally instilled saline within a 2-h period (15).

However, both of these studies focused on events occurring during attachment of nonreplicating viruses to epithelial cells. Nonreplicating viruses infect a small number of cells, and the signal transduction pathways activated during attachment are likely to have little effect on the development of lung pathological changes, which usually appear a number of days postinfection. Presently, the effects of key influenza viral proteins on ENaC levels and activity have not been investigated, and a specific mechanism for influenza-related regulation of ENaCs has not been determined. These mechanisms may be crucial in our understanding of how replicating respiratory viruses decrease ENaC activity, resulting in rhinorrhea and pulmonary edema. Herein we focused our attention on the role of M2 because of the importance of this protein in regulating viral infection. Quantitative analysis of M2 channel localization has indicated that during viral replication, the M2 ion channel is inserted into the apical membrane of host cells in preparation for viral assembly and that it does not associate with the other viral proteins in lipid raft microdomains (16). Thus, the M2 channel is likely to either interact directly with cellular apical membrane proteins or initiate signal transduction cascades affecting ENaC as well as other ion channel functions.

To assess whether M2 contributes to ENaC down-regulation, we microinjected amantadine-sensitive and -insensitive influenza M2 and human ENaC cRNAs into Xenopus oocytes and transfected human airway (H441) and alveolar-like (A549) epithelial cells with M2 cDNAs, and measured proton (H⁺) and amiloride-sensitive Na⁺ (ENaC) currents, as well as the total and membrane α-, β-, and γ-ENaCs at various intervals after injection or transfection. Because our data indicated that expression of M2 (but not of M1, a matrix influenza protein) decreased ENaC levels and activity, we truncated the M2 C terminus to identify the M2 regions responsible for these effects. Finally, we tested whether M2 decreases amiloride-sensitive currents by physically interacting (i.e., colocalizing) with ENaCs or by increasing steady-state levels of reactive oxygen intermediates and activating protein kinase C (PKC) isoforms.

MATERIALS AND METHODS

cRNA RNA synthesis

Human ENaCs

The pSport plasmid (Gibco Life Technologies, Inc., Gaithersburg, MD, USA), containing human (h) α-, β-, or γ-ENaCs, was linearized by overnight incubation with NotΙ (Promega, Madison, WI, USA). Sense RNA was in vitro transcribed from purified plasmid DNA using T7 polymerase according to the manufacturer’s instructions (Ambion Applied Biosystems, Austin, TX, USA). The integrity of the cRNA was verified by denaturing gel electrophoresis through 1% agarose-formaldehyde gel as described previously (17, 18).

hENaC cDNAs with Liddle mutations

hENaC-α595x, hENaC-β566x, hENaC-βS520K, and hENaC-γ575x (provided by Dr. Peter Snyder, University of Iowa, Iowa City, IA, USA) were subcloned into a pCDNA3.1 vector as templates for in vitro transcription as described previously (18).

M2 and M1 cRNA transcription

The wild-type M1 and M2 and truncated M2 gene sequences from influenza strain A/Udorn/72 and A/WSN/33 were derived from reverse transcription of the respective viral mRNAs isolated from MDCK cells infected at a low multiplicity of infection for 6 h (19). Strain-specific primers flanking the M2 coding region and containing a 5′ T7 phage RNA polymerase promoter and a 3′ stop codon were used in a PCR reaction to amplify the M2 cDNA sequences, which were then cloned into the BamH1 site of a pUC19 vector for in vitro transcription. All of the constructs were verified by DNA sequencing. ^m7GpppG-capped cRNAs were transcribed using the mMessage mMachine® High Yield Capped RNA transcription kit (Applied Biosystems, Foster City, CA, USA), followed by a poly(A) tailing reaction [E-PAP poly(A) tailing kit; Ambion, Austin, TX, USA], according to the manufacturer’s protocol. The integrity of the cRNA was verified by denaturing gel electrophoresis through 1% agarose-formaldehyde gel. The cRNA was purified by phenol/chloroform extraction, precipitated in ethanol twice, and used for microinjection of oocytes. cRNAs were dissolved in RNase-free water, and the concentrations were determined spectrophotometrically.

Preparation of the M2 plasmid

We constructed plasmids with a dual expression cassette containing either the green fluorescent protein (GFP) or the red fluorescence protein (RFP) gene behind a rat EF1 promoter in one cassette and a LacZ gene behind a human mouse EF1 promoter in a second cassette. pVM2U (7469 bp) was produced from pVITRO1 GFP-LacZ (Invivogen, San Diego, CA, USA). The new plasmid contains the sequence for the A/Udorn/72 M2 protein inserted in place of the LacZ gene using restriction sites BspHI and NheI. In brief, primers Udorn M2 5′ pVITRO (5′-CGCGCGCGTCATGAGCCTTCTGACCGAGGTCG-3′) and Udorn M2 3′ pVITRO (5′-CGCGCGCGGCTAGCTCAGGTTTATTACTCCAGCTCTATGCTGAC-3′) were used to isolate the M2 cDNA sequence from pJNM2U in a single round of PCR. A double restriction digest was performed on the PCR product using BspH1 and NheI. An equivalent digest was performed on 5 μg of pVITRO1 GFP-LacZ to remove the LacZ gene from the parent vector. The vector (−LacZ) and M2 PCR product were gel-purified, ligated, amplified, screened by restriction digest, and selected. The correct sequence, orientation, and frame of the M2 insert were confirmed by sequencing. pIRES-M2 was produced from pIRES-DsRed2 (Clontech, Mountain View, CA, USA). The new plasmid contained the sequence for the A/Udorn/72 M2 protein inserted behind a cytomegalovirus promoter using restriction sites XhoI and EcoRI, in the same manner as described above.

Construction of truncated M2 cRNAs

The wild-type M2 gene sequence from influenza strain A/Udorn/72 was cloned into the BamH1 site of pUC19 for cRNA transcription, as described above. This clone was then used to generate the truncated M2 template. The truncated M2 sequences were generated via a single PCR step using a 5′ primer containing a phage T7 RNA polymerase promoter attached to the 5′ end of the M2 gene sequence and a 3′ primer containing a stop codon, which truncates the protein at aa 52, 62, 67, 72, 77, 82, 87, and 92, for a total of 8 different M2 truncations, as performed by Tobler et al. (20). The PCR product was restricted, gel-purified, and ligated into an empty pUC19 vector. The plasmids were amplified and purified, then digested with BamH1, and used as the cRNA transcription template. Again, ^m7GpppG-capped cRNAs were transcribed using the mMessage mMachine kit, followed by a poly(A) tailing reaction [E-PAP poly(A) tailing kit]. The cRNA was purified by phenol-chloroform extraction, double ethanol-precipitated, and used for microinjection of oocytes. C-terminally truncated M2 gene sequences were cloned, and the respective cRNAs were transcribed and purified in the same manner.

Microinjection of oocytes with cRNAs

Oocytes were obtained from anesthetized adult female Xenopus laevis toads by standard techniques as described previously (17, 18). In brief, 25 ng of total hENaC cRNA (8.3 ng each of α-, β- and γ-hENaCs) in a 50-nl volume was injected, followed shortly after by either 1) 25 ng of M2 or M1 cRNAs in 50 nl or 2) 50 nl of RNase-free water, using a Nanoliter 2000 microinjector (World Precision Instruments, Sarasota, FL, USA). Oocytes were then incubated for 48 h in oocyte culture medium at 18°C, at which time they were used for biophysical measurements (see below).

Detection of whole-cell and single-channel Na⁺ currents

Membrane currents were measured using the 2-electrode voltage-clamp technique (18). The oocytes were held in a small groove in an experimental chamber of 1 ml volume at room temperature (21°C). The chamber was filled with ND96 solution containing 96 mM NaCl, 2.7 mM KCl, 1 mM MgCl₂, 2 mM CaCl₂, and 10 mM HEPES at pH 7.6 (osmolarity 200–220 mosol). Oocytes were impaled with two 3 M KCl-filled electrodes, with resistances of 0.5–1.5 MΩ. A TEV 200 voltage-clamp amplifier (Dagan Corp., Minneapolis, MN, USA) was used to voltage clamp oocytes. Two reference electrodes were connected to the bath. The membrane potential was held at −40 mV. Current-voltage (I-V) relationships were obtained by stepping the membrane potential from −140 to +60 mV in 20-mV increments for 600 ms. Amiloride-sensitive currents were calculated as the difference currents before and after application of 10 μM amiloride, which inhibited >90% of the inward Na⁺ currents. Sampling protocols were generated by pCLAMP 8.0 (Axon Instruments, Union City, CA, USA). Currents were sampled at the rate of 1 kHz, were filtered at 500 Hz, and were simultaneously stored electronically and displayed in real time on a chart recorder.

Cell-attached patch clamp: single-channel currents

The vitelline membranes of oocytes expressing α-, β-, and γ-hENaCs or α-, β-, and γ-hENaCs and M2 were surgically removed. Oocytes were then transferred to the recording chamber containing 0.5 ml of ND96. Pipettes were made from a thick-walled capillary glass (LG16; Dagan) using a two-stage puller (PC-10; Narishige, Tokyo, Japan). Pipette resistance values were 9.7 ± 0.9 MΩ (mean±se; n>20) when filled with ND96. The ground electrode (Ag-AgCl) was connected to the bath using a 3% agar bridge (1 M KCl). In all cases, single-channel currents were recorded using an Axopatch 200B (Axon Instruments) at a membrane potential of −100 mV for at least 10 min. The resting potential of oocytes expressing ENaCs was near 0 mV (18). Signals were digitized at 5 kHz (DigiData 1322 A, 16 bit; Molecular Devices, Sunnyvale, CA, USA), filtered at 1 kHz, and acquired with an IBM-compatible computer using pCLAMP 9.2 software (Axon Instruments).

Detection of ENaC protein in Xenopus oocyte plasma membranes

Xenopus oocyte plasma membranes were purified as described previously (21), with minor modifications. Briefly, oocytes were washed twice in Barth solution (90 mM NaCl, 3 mM KCl, 0.82 mM MgSO₄, and 5 mM HEPES, pH 7.6) and homogenized in 0.8 ml of the same solution supplemented with protease inhibitor cocktail (Sigma-Aldrich Corp., St. Louis. MO, USA). Homogenates were centrifuged at 250 g for 10 min at 4°C, and the supernatant was centrifuged at 16,000 g for 20 min at 4°C to pellet down total membranes. Pellets were lysed in 1% Triton X-100 lysis buffer (150 mM NaCl; 50 mM Tris-Cl, pH 7.4; 2 mM EDTA; 1% Triton X-100; and protease inhibitor cocktail), and the lysates were centrifuged at 16,000 g for 20 min at 4°C. The supernatants were collected as plasma membrane extracts. Equal amounts of whole oocyte or plasma membrane proteins were separated by 8% SDS-polyacrylamide gel. After transfer onto PVDF membrane, the blots were probed for γ-ENaCs using rabbit antibodies to α- and β-ENaCs (Abcam, Cambridge, MA, USA) followed by horseradish peroxidase (HRP)-conjugated protein G. The level of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was also determined to demonstrate equal loading of the extracts and the purity of the plasma membrane extracts.

Detection of ENaCs on plasma membranes by biotinylation

Eighty oocytes per group, injected with either α-, β-, or γ-hENaCs only or α-, β-, or γ-hENaCs and M2, were washed in PBS. Surface biotinylation was performed according to the instructions provided by the biotinylation kit manufacturer (Pierce Biotechnology, Rockford, IL, USA). After elution from the neutravidin column, the volume was adjusted to 50 μl, and the samples were loaded onto an 8% SDS-polyacrylamide gel for Western blotting with rabbit antibody against γ-ENaCs (22). Oocytes were injected with cRNAs of α-, β-, and γ-hENaC subunits alone or along with M2 cRNA and incubated for 24 h. Then 200 control oocytes, oocytes expressing ENaCs, or oocytes expressing ENaCs plus M2 were washed 3 times with ice-cold PBS and resuspended in 1 ml of ice-cold lysis buffer [50 mM Tris-buffered saline (pH 7.5)-1% Triton X-100-containing protease inhibitors] and homogenized. The homogenates were centrifuged at 16,000 g for 10 min. Solubilized proteins were recovered in the supernatant, and 300 μl of each extract was immunoprecipitated with 2 μg each of rabbit anti-γ-ENaC antibody overnight at 4°C. The immunoprecipitates were captured with protein A/G beads (Pierce Biotechnology), washed with same buffer 3 times, released from the beads by heating at 95°C for 5 min in Laemmli sample buffer, and processed for Western blotting. Each of the immunoprecipitated samples (equal volume) was run on an 8% SDS-polyacrylamide gel. After transfer onto PVDF membrane, the blots were probed for γ-ENaCs using a rabbit antibody to the γ-ENaC (Abcam) followed by HRP-conjugated protein G.

Cell culture and transfection

H441 cells were obtained from American Type Culture Collection (Manassas, VA, USA) and grown in RPMI 1640 medium (Mediatech, Manassas, VA, USA) supplemented with 1% l-glutamine, 10% FBS, penicillin (100 U/ml), and streptomycin (100 μg/ml), as described previously (23, 24). Cells were seeded in 75-cm² flasks, incubated in a humidified atmosphere of 5% CO₂-95% O₂ at 37°C until they reached confluence, and passaged weekly. For consistency, only cells between passages 57 and 73 were used in our studies. Cells were lifted from the flasks using 0.05% trypsin and 0.53 mM EDTA (Mediatech), seeded at a density of 1 × 10⁶ cells/ml on round coverslips (10 mm in diameter; Fisher Scientific, Pittsburgh, PA, USA), and grown in RPMI 1640 medium supplemented with 200 nM dexamethasone to facilitate sodium channel differentiation (25). The medium was changed every other day. H441 cells at 70% confluence were transfected with 2 μg of plasmid cDNA (suspended in 10 μl of Polyfect transfection agent) containing either the GFP gene alone or GFP + M2.

Measurement of whole-cell currents in H441 cells

Forty hours later, GFP-containing cells (5–10% of total) were identified by the presence of green fluorescence when exposed to UV light and patched in the whole-cell mode (in the absence and presence of amiloride). Whole-cell currents were elicited by applying the step-pulse protocol from −100 to +100 mV in 10-mV increments for 500 ms from a holding potential of −40 mV. Data analysis was performed as described previously (18, 26). Amiloride-sensitive currents (I_ENaC) were calculated by subtracting remnant steady-state currents at a given voltage after perfusion with amiloride (10 μM) from their corresponding values just before amiloride perfusion. I-V relationships were constructed by averaging the current values between 400 and 500 ms from the start of voltage steps with Clampfit (Axon Instruments) and plotted using Origin software (OriginLab, Northhampton, MA, USA). Cells were perfused with an external solution containing amiloride (2 μM), and amiloride sensitive I-V relationships were calculated as described above.

Cell immunostaining for M2 and ENaC detection

H441 cells were seeded on glass coverslips at low density (30–50%). Twenty-four hours later, or when they had reached 60–70% confluence, they were transfected with M2 plasmid cDNA. Twenty-four hours later, they were fixed with 3% formaldehyde for 45 min at room temperature. Cells were rinsed with PBS and then treated with 0.5% Triton X-100 (in PBS) for 3 min at room temperature to permeabilize the cells and partially denature proteins to reveal antigenic sites. Cells were then rinsed 4 times with PBS to remove Triton X-100. To block nonspecific antigenic sites, cells were treated with a blocking buffer (PBS supplemented with 5% goat serum) for 30 min. At this time, the cells were incubated with a mouse anti-M2 antibody diluted in the blocking buffer to 10 μg/ml (Calbiochem, San Diego, CA, USA) for 1 h. They were then washed 4 times in PBS and incubated for 5 min in the blocking buffer and then with the first secondary antibody, goat anti-mouse IgG conjugated to Alexa Fluor 488 (Molecular Probes, Eugene, OR, USA), for 1 h. H441 cells were washed once with PBS, incubated with the blocking buffer for 5 min, and incubated with an anti-α-ENaC rabbit IgG antibody for 1 h, diluted at 10 μg/ml (Abcam) in the blocking buffer. The cells were washed as above and incubated for an additional 1 h with a donkey anti-rabbit IgG antibody (Abcam) conjugated to Texas Red diluted in the blocking buffer at 20 μg/ml. The cells were rinsed once with PBS and counterstained with Hoechst 33258 (20 μg/ml) for 4 min to visualize the nuclei by their blue color. They were then mounted onto glass slides using 0.2% n-propyl gallate (Sigma-Aldrich Corp.) in 1:9 (v/v) PBS-glycerol. Confocal microscopy imaging was performed on a Leica DMIRBE inverted epifluorescence/Nomarski microscope outfitted with Leica TCS NT SP1 laser confocal optics (Leica, Exton, PA, USA). The system was equipped with UV, argon/krypton, and helium/neon lasers for the imaging of a wide range of blue, green, red, and far-red fluorochromes. Precise control of fluorochrome excitation and emission was regulated, respectively, by an acousto-optical tunable filter and a TCS SP1 prism spectrophotometer. Spectral separation of fluorochromes was guaranteed by sequential scanning with single lasers and detection channels for each fluorochrome imaged and by the use of tight band-pass emission windows, which do not overlap with the emission spectra of the other fluorochromes imaged. En face optical sections (XY plane) through the z axis were generated using a stage galvanometer. During image acquisition, the image format size was set to 1024 × 1024 pixels for high resolution and the pinhole aperture to 150 μm for a high level of haze removal. The acquired images were prepared using Adobe Photoshop (Adobe Systems, San Jose, CA, USA).

Whole-cell reactive oxygen species (ROS) imaging

Steady-state levels of ROS were visualized using two different redox-sensitive molecules. In the first set, A549 cells were seeded on Lab-Tek II chamber slides and cultured and transfected with M2-RFP plasmids as described above. Slides were washed 3 times with PBS and then were incubated with 1 ml/well of 5-(and-6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate (DCF-DA) (Invitrogen, Valencia, CA, USA) reagent in PBS/Ca²⁺/Mg²⁺ for 30 min at 37°C. At 25 min after DCF-DA incubation nuclei were counterstained with Hoechst 33342 for 5 min. The cells were then washed 3 times with PBS/Ca²⁺/Mg²⁺ and imaged immediately. When the reduced fluorescein moiety is oxidized by cellular ROS, it emits a bright green light at an excitation/emission wavelength of 495/529 nm. Cells were imaged using fluorescence microscopy with the appropriate filters for the nuclei (Hoechst, blue) and ROS (DCF-DA, green). Transfected cells were identified by the presence of RFP. In the second set of experiments, A549 cells were incubated with 500 μl of MitoSOX Red (5 μM) (Invitrogen) in PBS/Ca²⁺/Mg²⁺ for 10 min at 37°C. MitoSOX Red targets the mitochondria, and when oxidized by superoxide it fluoresces red at an excitation/emission wavelength of 510/580 nm. The cells were washed twice with PBS/Ca²⁺/Mg²⁺ to minimize background emissions, and they were imaged immediately using a fluorescence microscope equipped with the appropriate filters. In this set of experiments, the transfected cells were identified by the presence of GFP. All images werre taken with a dry objective at ×40.

RESULTS

M2 decreases whole-cell ENaC currents in both oocytes injected with hENaC cRNAs and human airway cells

In our first series of experiments, we coinjected oocytes with M2 and α-, β-, and γ-hENaC cRNAs and transfected H441 and A549 cells with plasmids with dual expression cassettes containing either the GFP or RFP genes behind a rat EF1 promoter in one cassette and the A/Udorn/72 M2 in another. Epithelial cells transfected with M2 (5–10% of H441 and 10–15% of A549 cells) were identified by the expression of either green or red fluorescence and were patched in the whole-cell mode. Significant levels of proton currents (I_M2) were seen across oocytes (Fig. 1A, B), injected with M2 cRNAs of Udorn (amantadine-sensitive) and WSN (amantadine-insensitive) strains, respectively, and H441 cells (Fig. 1C), transfected with Udorn M2 cDNA, when extracellular pH was decreased to 5.5. These levels are similar to those reported previously (27). The proton currents generated by the Udorn but not by the WSN strains were inhibited by 100 μM amantadine in both oocytes and H441 cells (Fig. 1A, C, respectively) in agreement with previous reports (4). No proton currents were detected at physiological pH (7.6 for oocytes and 7.4 for airway cells). Human airway cells infected with M2 did not become apoptotic as shown by the lack of staining with annexin V (a marker of apoptosis) as described previously (24).

Figure 1.

M2 inhibits amiloride-sensitive Na⁺ currents at neutral pH of Xenopus oocytes and H441 cells. A, B) I-V relationships of I_M2 recorded from Xenopus oocytes 48 h after microinjection with Udorn (A) and WSN (B) M2 cRNAs (25 ng), using the 2-electrode voltage-clamp technique, at extracellular pH 7.6 or 5.5, as described in Materials and Methods. I-V relationships were obtained before and after perfusion with amantadine (100 μM). C) I_M2 in H441 cells transfected with a plasmid containing the Udorn M2 and GFP genes. Currents were measured in H441 expressing green fluorescence 48 h post-transfection at bath pH values of either 7.4 or 5.5; the proton current was totally inhibited by 100 μM amantadine. D) I-V relationships of I_ENaC from oocytes coinjected with M2 (25 ng) and α-, β-, and γ-hENaC cRNAs (8.4 ng each) 48 h after injection. Difference currents before and after perfusion with amiloride (10 μM) at a given potential are shown. E) Total currents across H441 cells transfected with plasmids containing either GFP alone or GFP-M2 before and after addition of amiloride (10 μM) in the bath solution (pH 7.4). Cells were held at −40 mV, and membrane potentials were changed from −100 to +100 mV in 10-mV increments for 500 ms. F) I-V relationships of I_ENaC of nontransfected H441 cells and cells transfected with plasmids containing either GFP or GFP + M2. In all cases (except E) values are means ± se; n = number of oocytes or H441 cells.

Oocytes injected with α-, β-, and γ-hENaCs expressed significant amounts of Na⁺ currents, 90% of which were inhibited by 10 μM amiloride (I_Na at −100 mV: control −3675±610; amiloride −343±134 nA, mean±se; n=10; P<0.001). In H441 cells, which express all three ENaC subunits, amiloride inhibited ∼75% of Na⁺ currents (I_Na at −100 mV: control −428±31; amiloride −86±8 pA, mean±se; n=5; P<0.001). M2 expression in oocytes and H441 inhibited both whole-cell Na⁺ and I_ENaC by 90 and 75% from their corresponding control values at pH 7.6 (oocytes) and 7.4 (H441 cells), respectively (Fig. 1D–F). An additional, but smaller, inhibition was observed at pH 5.5 in Xenopus oocytes (Fig. 1D). Significant inhibition of I_ENaC of Xenopus oocytes was also observed when oocytes expressing α-, β-, and γ-hENaCs were injected with M2 cRNA 24 h later (Fig. 2A) at pH 7.6. These data indicate that M2 inhibition of ENaC occurs even when this channel is not activated by acidic pH. In contrast, injection of oocytes with M1 (a matrix influenza protein that does not form channels) and α-, β-, and γ-hENaC cRNAs increased I_ENaC (Fig. 2B). Thus, the M2-induced decrease of amiloride-sensitive currents in Xenopus oocytes was not due to the concomitant expression of a second protein.

Figure 2.

A) M2 inhibits ENaC. Xenopus oocytes were microinjected with α-, β-, and γ-hENaC cRNAs (8.4 ng each). Twenty-four hours later, they were microinjected with either Udorn M2 cRNA (25 ng) or an equal volume of RNA-free sterile water. Whole-cell I_ENaC was measured 24 h after M2 injection. B) M1 does not inhibit ENaCs. Xenopus oocytes were microinjected with either α-, β-, and γ-hENaC cRNAs (8.4 ng each) plus sterile water or α-, β-, and γ-hENaC cRNAs plus Udorn M1 cRNA (25 ng). Whole-cell currents were recorded 48 h later. Values are means ± se; n = number of oocytes.

M2 decreases ENaC single-channel activity and surface protein levels in oocytes injected with ENaC cRNAs

We recorded ENaC single-channel activity in oocytes patched in the cell-attached mode and also measured α-, β-, and γ-ENaC levels in their cytoplasmic and membrane fractions. M2 expression decreased the number of active channels in the membrane patches of oocytes expressing both M2 and ENaCs compared with ENaCs alone (Fig. 3A, B). NP_o (the product of open channels times their open probability) values (mean±se; n=number of measurements) of ENaCs alone vs. ENaCs + M2 were 0.6 ± 0.1 (n=5) vs. 0.04 ± 0.009, n = 5 (P<0.01). The unitary channel conductance (4 pS) was unaffected. In agreement with the observed decrease of whole-cell currents and single-channel activity in ENaCs expressing Xenopus oocytes, M2 expression also decreased total and membrane levels of α- and β-ENaCs (Fig. 3C) and plasma membrane levels of γ-ENaCs (isolated by biotinylation) (Fig. 3D). These data indicate that the observed decrease of NP_o may be due to a decrease in both the number and open probability of ENaC channels at the plasma membrane. No colocalization among M2 and ENaCs could be identified (Fig. 4), indicating that M2 down-regulation of ENaCs was not due to their physical interaction.

Figure 3.

M2 reduces single-channel ENaC activity and P_o. A, B)Typical records (A) and corresponding amplitude histograms (B) of single channels from cell-attached patches of Xenopus oocytes 48 h after microinjection with α-, β-, and γ-hENaCs alone or ENaCs + M2. Amplitude histograms were constructed from records of at least 10 min. C) Western blots of total [whole-cell extract (WCE)] and membrane α- and β-hENaC levels from Xenopus oocytes microinjected with either ENaCs or ENaCs + M2. GAPDH levels indicate uniform loading of lanes. D) Western blots of γ-ENaC in biotinylated proteins from oocytes microinjected with either α-, β-, and γ-hENaCs or ENaCs + M2. Biotinylated proteins were precipitated with streptavidin-agarose, separated by SDS-PAGE, followed by Western blotting with anti-γ-ENaC antibodies, as described in Materials and Methods. Similar levels of nonspecific bands indicate uniform loading of each lane. Typical records were repeated 3 times with identical results. We were unable to detect α- and β-hENaC in biotinylated proteins, despite the fact that we could detect them in total membrane fractions (C).

Figure 4.

M2 and ENaC do not colocalize in transfected H441 cells. H411 cells were seeded on glass coverslips 24 h before transfection with M2 plasmids. Twenty hours later they were fixed, stained, and imaged with confocal microscopy, as mentioned in Materials and Methods. Images shown are at the level of the plasma membranes. A) Control H441 cells expressing endogenous ENaC were immunostained with a primary antibody to α-ENaC (rabbit) followed by a secondary antibody (donkey anti-rabbit) coupled to Texas Red (red). B) Same image as A with cell nucleus staining. C) H441 cells transfected with M2 cDNA plasmid showing α-ENaCs (red). D) Same cells as in C, immunostained with a primary antibody to M2 (mouse) followed by a secondary antibody (goat anti-mouse) coupled to Alexa Fluor 488 (green). E, F) Merged images of C and D without (E) and with (F) nucleus staining (blue). Note that there is no immunocolocalization of α-ENaC and M2 protein. Typical images were reproduced 9 times with the same results.

Identification of M2 regions responsible for decreased ENaC activity

We generated 7 truncations of the M2 C terminus (Fig. 5A) and tested their ability to generate proton currents at pH 5.5 (Fig. 5A, right panel) and to inhibit ENaCs at neutral pH in Xenopus oocytes (Fig. 5A, left panel). We first truncated a large portion of the M2 C terminus [from aa 52 to 97 (M2-52)]; the remaining segment is known to have a moderate cell surface expression compared with that of the full-length protein (20). We then microinjected Xenopus oocytes with M2-52 cRNA alone or M2-52 cRNA and α-, β-, and γ-hENaCs and recorded proton currents at pH 5.5 and I_ENaC at pH 7.6, 48 h later. M2-52 cRNAs expressed no significant levels of proton currents, most likely because of its inability to insert properly in the plasma membrane (20) and did not inhibit I_ENaC (Fig. 5A, left and right panels, respectively). Conversely, M2 cRNAs of M2 lacking aa 62–97 (M2-62) generated near-normal proton currents at pH 5.5 and inhibited 65% of I_ENaC at pH 7.4 (Fig. 5A, left and right panels, respectively). Thus, the 10 aa (53–62: RFFEHGLKRG) play a critical role in M2 function and ENaC inhibition. We also identified an additional 5-aa region (77–82: KEQQS), responsible for the remaining 35% inhibition of ENaCs (Fig. 5A). Our data indicate that there are ≥2 regions of the M2 C terminus that play a major role in the function of M2 and inhibition of ENaCs.

Figure 5.

A) Identification of regions of M2 responsible for decreased ENaC activity. Amino acid sequence of the A/Udorn/1972 M2 protein and the points of truncation for each construct are shown. The 3 protein domains are indicated above the sequence. B) Xenopus oocytes were coinjected with α-, β-, and γ-hENaC cRNAs and sterile water or with α-, β-, and γ-hENaC cRNAs plus either the full-length M2 or truncated M2 cRNAs. Forty-eight hours later, I_ENaC (left panel) and I_M2 (right panel) were recorded. Values were obtained at a membrane potential of −100 mV and are means ± 1 se; n = ≥21. *P < 0.01 vs. control; ^#P < 0.01 vs. corresponding ENaC or M2. B) MG-132 reverses the M2 effect on ENaCs. Oocytes were injected with α-, β-, and γ-hENaC or α-, β-, and γ-hENaC and M2 cRNAs. In some cases, MG-132 (40 nM) was added in the medium immediately after injection. Whole-cell currents were recorded 48 h later, and I_ENaC was calculated as described in Materials and Methods. Values are means ± se; n = number of measurements. C) M2 does not down-regulate ENaCs with Liddle mutations. Xenopus oocytes were injected with α-, β-, and γ-ENaC cRNAs with the Liddle mutation alone or Liddle ENaC and M2 cRNAs. I_ENaC was recorded as described in Materials and Methods. Values are means ± 1 se; n = number of measurements.

ENaC activity of Liddle mutants is not altered by M2

To test the hypothesis that M2 enhances ENaC internalization and degradation by enhancing ubiquitination, we coinjected oocytes with full-length M2 and ENaC cRNAs and incubated them with an inhibitor of both the proteasome and lysosome systems (40 nM MG-132) for 48 h. As shown in Fig. 5B, MG-132 completely prevented the inhibition of I_ENaC by M2, suggesting that significantly higher levels of ENaCs are degraded by the proteasome and lysosome systems in M2-injected oocytes. We have shown previously that similar results are obtained when epithelial cells are incubated with much higher doses of MG-132 for shorter periods of time (4 μM for 2 h) (18). To document that M2 decreased ENaC levels by enhancing ENaC ubiquitination, we coinjected Xenopus oocytes with M2 and ENaCs bearing Liddle mutations (a gift from Dr. Peter Snyder, University of Iowa). Baseline Liddle I_ENaC currents were significantly higher than those for the wild type and were not affected by M2 (Fig. 5C). In aggregate, these data show that M2 enhances ENaC ubiquitination, internalization, and degradation either by the proteasome or lysosome systems.

Transfection of human airway cells with M2 increases steady-state levels of reactive oxygen intermediates and PKC activation

Infection of human airway cells by viruses has been shown to activate NF-κβ increasing steady-state levels of reactive oxygen and nitrogen species (24). Because M2 acts as a proton channel, it may alter membrane potentials across a number of organelles, including the mitochondria, leading to the production of partially reduced reactive oxygen and nitrogen species. A549 cells were transfected using the PolyFect transfection reagent with a plasmid containing RFP alone (Fig. 6A–C) or a plasmid containing RFP-M2 (Fig. 6D–F) as described in Materials and Methods. Forty-eight hours later, cells were incubated with DCF-DA, which when oxidized by ROS (such as superoxide, hydrogen peroxide, hydroxyl radicals, or peroxynitrous acid) emits green fluorescence. Cells transfected with M2-RFP cDNA exhibited significant levels of green fluorescence, which colocalized with red fluorescence (due to M2 expression), indicating that transfection with M2 up-regulates ROS levels in cells (Fig. 6F). In contrast, minute green fluorescence was seen in cells transfected with RFP alone (Fig. 6C). Two independent observers graded the levels of green fluorescence using a scale from 0 (none) to 4 (most) in cells transfected with plasmids containing RFP alone or RFP + M2. Results were as follows (mean±se; n=number of measurements): RFP = 1 ± 0.4; RFP + M2 = 3.3 ± 0.33 (n=10 each; P=0.0035 by Mann-Whitney analysis). Similar results were obtained when A549 cells were transfected with M2-GFP and then incubated with MitoSOX Red, which readily permeates cells and mainly targets the mitochondria (GFP= 0.75±0.4; GFP+M2=2.9±0.4; n=8 each; P = 0.0085 by Mann-Whitney analysis) (Fig. 7). These data indicate that at least a portion of the reactive species is derived from the mitochondria. It should be noted that transfection of A549 cells with M2 decreased their I_ENAC levels to the same extent as in H441 cells. We opted to conduct these studies in A549 cells instead of H441 cells because of their higher transfection rate.

Figure 6.

Transfection of A549 cells with M2 increases intracellular levels of ROS. A549 cells were transfected with a RFP-containing plasmid (A–C) or a plasmid containing both RFP and M2 (D–F), as described in Materials and Methods. Forty-eight hours later, the A549 cells were incubated with DCF-DA following the manufacturer’s instructions. Nuclei were counterstained with Hoechst dye (blue). A, D) Red fluorescence, indicating cells transfected with plasmids. B, E) Green fluorescence, indicating increased production of reactive species. C, F) Merged images of red and green fluorescence. Note yellow-orange color in cells transfected with RFP-M2 plasmids (F) but not with RFP alone (C). Typical records reproduced repeatedly with 6 different preparations.

Figure 7.

Transfection of A549 cells with M2 increases mitochondrial levels of reactive species. A549 cells were transfected with a GFP-containing plasmid (A–C) or a plasmid containing both GFP and M2 (D–F). Forty-eight hours later, the A549 cells were incubated with the MitoSOX reagent following the manufacturer’s instructions. Nuclei were counterstained with Hoechst (blue). A, D) Green fluorescence, indicating cells transfected with plasmids. B, E) Red fluorescence, indicating increased production of reactive species, most likely by mitochondria. C, F) Merged images of red and green fluorescence. Note yellow-orange color in cells transfected with RFP-M2 plasmids (F) but not with RFP alone (C). Typical records were reproduced repeatedly with 6 different preparations.

To ascertain whether increased levels of reactive species were responsible for the decrease in I_ENaC, we incubated either Xenopus oocytes or A549 cells with glutathione (GSH) ester (1 mM), a cell-permeable form of GSH, immediately after microinjection with α-, β-, and γ-hENaC and M2 cRNAs or transfection with M2 plasmids, respectively. On entering the cytoplasm the ester is cleaved by esterases, thus confining the GSH in the cytoplasm. The medium was removed 24 h later, new GSH-ester was added, and I_ENaC were measured 48 h after injection or transfection. GSH totally reversed the M2 down-regulation of I_ENaC in both oocytes and A549 cells (Fig. 8A, B).

Figure 8.

Mitigation of M2 ENaC down-regulation by reactive species scavengers and PKC inhibitors. I_ENaC at −100 mV. Values are means ± se; numbers of measurements are shown next to each bar. A) Oocytes microinjected with α-, β-, and γ-hENaC or ENaC + M2 cRNAs; 500 μM GSH ester was added in the medium, and the solution was changed every 12 h. B) A549 cells transfected with M2 plasmids containing either GFP alone or M2 and GFP genes; 500 μM GSH ester was added in the medium, as described above. C, D) H441 cells transfected with plasmids containing either GFP or GFP plus M2 genes; Gö6796 (250 nM; a specific inhibitor of PKCα/β₁ isoforms) (C) or pseudosubstrate inhibitor, myristoylated (1 μM; a specific PKCζ inhibitor) (D) was added into the medium 24 h later for 24 h at which time I_ENaC was recorded. *P < 0.01 compared with the corresponding control (either ENaCs for oocytes or GFP for cells). ^#P < 0.01 vs. value to immediate left; Student’s t test.

These findings show that M2 increases levels of ROS, which are responsible for ENaC down-regulation and are also known to activate PKC, which affects ENaC function. To test the hypothesis that PKC is involved in the M2 down-regulation of I_ENaC in human airway cells, we transfected H441 cells with M2-GFP plasmids and 24 h later incubated them with Gö6796 (250 nM; a specific inhibitor of PKCα/β₁ isoforms) or a PKCζ inhibitor (1 μM; pseudosubstrate inhibitor, myristoylated) for an additional 24 h. As shown in Fig. 8C, D, inhibition of PKCα/β₁ but not of PKCζ significantly increased baseline I_ENaC. Both inhibitors, however, totally prevented the M2 down-regulation of I_ENaC. Thus, activation of a number of PKC isoforms (α, β₁, and ζ) plays an essential role in the M2 down-regulation of ENaCs.

DISCUSSION

In this study we show for the first time that an influenza virus type A protein, M2, down-regulates the function of a major epithelial ion channel, ENaC, both in oocytes microinjected with human ENaCs and in human airway cells expressing α-, β-, and γ-ENaCs and amiloride-sensitive currents, by increasing levels of reactive species and activating specific isoforms of PKC. These findings both complement and extend previous works in this field. In particular, the previous reports showing that adherence of HA of replication-deficient influenza viruses to the sialic acid residues on both airway and alveolar epithelial cells initiates a sequence of events leading to activation of PKC, which in turn down-regulated ENaC function in vitro and clearance of intratracheally instilled fluid across the distal lung spaces of rats (14, 15, 28). However, events occurring during viral attachment are likely to be transient and affect only a small fraction of cells. Our data showing that M2, a protein that plays a seminal role in viral replication and is known to translocate to the apical membranes of epithelial cells, offer considerable new insights as to the mechanisms by which replicating flu viruses may alter the function of an important ion channel (ENaC) that plays a decisive role in fluid homeostasis across the lung (9). Furthermore, decreased ENaC activity after transfection of human cells with M2 did not result from nonspecific cellular injury or apoptosis.

The influenza M2 channel is a homotetramer proton channel (1) expressed at the apical membranes of infected epithelial cells. The selectivity of the M2 channel to various cations has been addressed in many previous studies. For example, Shimbo et al. (27) reported that the permeability of the M2 channel to protons exceeds its permeability to other cations (such as Na⁺, Li⁺, NH₄⁺, and Rb⁺) by 5 orders of magnitude. The selectivity of M2 to protons depends on the histidine residue on its transmembrane domain (His-37), and mutant forms of the M2 channel in which His37 is replaced by other amino acids are more permeable to Na⁺, and K⁺ than wild-type M2 (29). Furthermore, H⁺ but not hydronium (H₃O⁺) ions cross through M2 (30,31,32).

With the exception of a single report showing M2 interference with influenza HA trafficking from the trans-Golgi network to the cell surface (33), there are no known reports of M2 affecting host cell protein function. During viral replication new proteins are synthesized, and M2 is sorted to the apical membrane of epithelial cells (4) where the functional ENaC is also located. Direct physical interaction among M2 and ENaC subunits may have created complexes that were targeted for destruction by either the proteasome or the lysosome systems. However, confocal microscopy studies in human airway cells expressing native ENaCs and transfected with M2 cDNAs failed to show significant colocalization among these proteins. In contrast, the M2-induced decrease of amiloride-sensitive currents in Xenopus oocytes expressing ENaCs was prevented by incubating them with MG-132 shortly after injection; in addition M2 had no effect on Liddle ENaCs. These data indicate that M2 modulates ENaC function through indirect, signal transduction mechanisms. Specifically, our data suggest that M2 increases steady-state levels of ROS and activates PKC, which modify ENaC, enhancing its propensity to be ubiquitinated and destroyed by the proteasome system.

ENaC is formed by the assembly of ≥3 subunits (α, β, and γ) (34) with an additional subunit (δ) also being present in lung epithelial cells (22). ENaCs are characterized by 4–6 pS conductance and long open and closing times and are inhibited by submicromolar concentrations of the K⁺-sparing diuretic amiloride(10, 34). ENaCs are constitutively open and do not require additional activation (10). Instead, ENaC activity is regulated by changes of the open probability P_o and the number N of channels at the apical plasma membranes. In the long term, factors that influence both ENaC mRNA and protein levels potentially modulate Na⁺ transport in the lung. In the short term, the number of channels at the cell membrane is controlled by mechanisms that involve both exocytosis of newly synthesized channels and their removal from the apical membrane by endocytosis (35). The half-life of ENaCs in mammalian cell membranes is short (<1 h).

ENaC is ubiquitinated in vivo on the α and γ but not on the β subunit (36). Neural precursor cell-expressed developmentally down-regulated protein 4 (Nedd4-2) is the ubiquitin ligase required for ubiquitination of ENaCs (37). Nedd4-2 contains 3 WW domains that bind to the PPXY regions of ENaC subunits (38, 39). Polyubiquitinated ENaCs are either degraded by the proteasome, whereas monoubiquitinated ENaCs are degraded in the lysosome (35). The coexpression of ENaC subunits and Nedd4-2 in Xenopus oocytes leads to the reduction of ENaC half-life at the cell surface and a simultaneous decrease in the amiloride-sensitive currents (40,41,42). In the lungs, Nedd4-2 is expressed in the epithelial cells with the same pattern as ENaCs (43).

The interaction between ENaCs and Nedd4-2 is disrupted in Liddle syndrome, a hereditable form of salt-sensitive hypertension (44). Specifically, truncations of the ENaC COOH terminus or mutations in the conserved COOH-terminal PY motifs abolish the interaction between ENaC and Nedd4-2, resulting in a longer ENaC half-life at the plasma membrane and lower levels of internalization and degradation. These lead to increased amiloride-sensitive currents and increased salt absorption across epithelial cells (45,46,47).

Reactive oxygen and nitrogen species can alter ENaC function by either post-translational modifications [such as oxidation, nitration, nitrosylation, and glutathionylation of critical residues (48, 49) or activation of signal transduction pathways (24, 50, 51)]. For example, we have shown that peroxynitrite generated by SIN-1 decreases amiloride-sensitive currents in Xenopus oocytes microinjected with α-, β-, and γ-ENaCs (52). Furthermore, a point mutation in the extracellular loop of the α-ENaC (Y283A) largely mitigated to some extent the detrimental effects of peroxynitrite on the amiloride-sensitive currents (53). Reactive species have been shown to activate PKC that acts by inhibiting ENaCs. For example, inhibition of PKC rapidly increased P_o and the appearance of new channels in patches of A6 (54), in lung ATII cells (15), and in Xenopus oocytes heterologously expressing α-, β- and γ-ENaCs (55). In contrast, stimulation of PKC inhibited whole-cell Na⁺ currents in Xenopus oocytes microinjected with hENaC cRNAs (55). Likewise, PKC activation decreased expression of both β- and γ-ENaC levels in A6 cells, the activity of single channels, and transepithelial Na⁺ reabsorption (56). Reactive species may be originating from activation of various membrane generators or from the mitochondria, as suggested by increased MitoSOX fluorescence, a probe that concentrates mainly in the mitochondria. A previous study reported that M2 depolarized the mitochondrial membrane potential of transfected HEK cells (57).

There are several hypotheses put forth to explain the mechanisms of ENaC down-regulation by PKC. In A6 cells, PKC has been shown to activate the mitogen-activated protein (MAP) kinase kinase kinase Raf-1, and the MAP kinase kinases MAPK/ERK (MEK) 1 and 2. Activation of MEK 1 and 2 enhances phosphorylation of β- and γ-, but not α-ENaCs (58). This phosphorylation event facilitates binding of Nedd4-2 to ENaCs, which promotes ENaC internalization and removal from the cell surface (37). Infection of lung epithelial cells with inactivated influenza virus A/PR/8/34 (PR8; H1N1) inhibited ENaC function and decreased Na⁺ transport across airway murine distal epithelial cells, and these events were prevented by pretreatment with PKC inhibitors (15). Recently we reported that injection of severe acute respiratory syndrome (SARS) proteins in ENaCs expressing oocytes also down-regulated ENaCs, and this effect was partially ameliorated by inhibition of PKC (17). Others have shown that activation of PKCζ plays a critical role in the inhibition of Na⁺,K⁺-ATPase during hypoxia (59). Recently, Wolk et al. (60) proposed that a down-regulation of alveolar fluid clearance in BALB/c mice infected with live A/WSN/33 virus resulted from stimulation of P2Y receptors by UTP/UDP, released in the alveolar lining fluid of infected lungs. However, the UTP-dependent inhibition of Na⁺ absorption may be downstream of PKC activation (61). PKC activation was also involved in the down-regulation of ENaCs by SARS S and E proteins (17).

In summary, data presented herein provide new insights into the mechanisms by which replicating influenza viruses may damage ENaCs, which may lead to rhinorrhea and pulmonary edema due to decreased alveolar fluid absorption (9). In a very interesting and highly relevant study, Hoffmann et al. (62) recently reported that agents that activate PKC or decrease ENaC activity increase influenza viral replication. Thus, there is a teleological advantage for the influenza virus to activate PKC and decrease ENaC activity and expression. Lastly, these data suggest that the influenza M2 protein functions as a virulence factor by affecting host protein function and may suggest new therapeutic strategies for altering the progression of the viral infection. It should also be noted that influenza infection in a human is a characterized by the progressive infection of lung tissue, not the simultaneous infection of the entire lung. Most cases of seasonal influenza infection never progress to complete infection of the lung. Our experiments deal with single cell measurements using one component of the influenza virus and should be taken in that context. Our results must be considered significant in patients, in whom influenza infection has progressed into a higher percentage of infected lung tissue and is complicated by pneumonia.