Research ArticleImmunology

Mitochondrial Membrane Potential Is Required for MAVS-Mediated Antiviral Signaling

See allHide authors and affiliations

Science Signaling  01 Feb 2011:
Vol. 4, Issue 158, pp. ra7
DOI: 10.1126/scisignal.2001147


Mitochondria, dynamic organelles that undergo cycles of fusion and fission, are the powerhouses of eukaryotic cells and are also involved in cellular innate antiviral immunity in mammals. Mitochondrial antiviral immunity depends on activation of the cytoplasmic retinoic acid–inducible gene I (RIG-I)–like receptor (RLR) signaling pathway and the participation of a mitochondrial outer membrane adaptor protein called MAVS (mitochondrial antiviral signaling). We found that cells that lack the ability to undergo mitochondrial fusion as a result of targeted deletion of both mitofusin 1 (Mfn1) and mitofusin 2 (Mfn2) exhibited impaired induction of interferons and proinflammatory cytokines in response to viral infection, resulting in increased viral replication. In contrast, cells with null mutations in either Mfn1 or Mfn2 retained their RLR-induced antiviral responses. We also found that a reduced mitochondrial membrane potential (ΔΨm) correlated with the reduced antiviral response. The dissipation in ΔΨm did not affect the activation of the transcription factor interferon regulatory factor 3 downstream of MAVS, which suggests that ΔΨm and MAVS are coupled at the same stage in the RLR signaling pathway. Our results provide evidence that the physiological function of mitochondria plays a key role in innate antiviral immunity.


Innate immunity against viral infection involves the activation of multiple signaling steps that culminate in the rapid production of the type I interferons (IFN-α and IFN-β) and other proinflammatory cytokines that provide a first line of defense against viruses (1). Germline-encoded pattern recognition receptors (PRRs) of the host recognize broadly conserved viral components known as pathogen-associated molecular patterns (PAMPs), including genomic DNA, double-stranded RNA (dsRNA), and 5′-triphosphated single-stranded RNA. Viral dsRNA is detected by two classes of PRRs, endosomal Toll-like receptor 3 (TLR3) and the cytoplasmic retinoic acid–inducible gene I (RIG-I)–like receptors (RLRs) RIG-I and melanoma differentiation–associated gene 5 (MDA-5), which initiate distinct signaling pathways (1, 2). Although the stimulation of TLR3 and RLRs by PAMPs activates their intracellular signaling cascades through different downstream effectors, they converge at the point of the activation of the transcription factors, nuclear factor κB (NF-κB) and interferon regulatory factor 3 (IRF-3), which lead to the induction of genes encoding antiviral compounds that promote the subsequent development of adaptive antiviral immunity (24).

The mitochondrion, a double-membraned organelle that contains its own genome, mitochondrial DNA (mtDNA), is the powerhouse of eukaryotic cells and generates adenosine triphosphate (ATP) through oxidative phosphorylation (5). In addition to their pivotal roles in energy production, apoptosis (6), calcium homeostasis (7), and their ability to exhibit high mobility (continual cycles of fusion and fission) in cells (8, 9), mitochondria also act as platforms for innate antiviral immunity in mammals (10, 11). Mitochondrial antiviral immunity depends on activation of the RLR signaling pathway and the participation of the mitochondrial outer membrane protein, mitochondrial antiviral signaling [MAVS (10), also known as IPS-1 (12), VISA (13), and Cardif (14)]. Antiviral immune responses are not induced when MAVS is not properly localized at the mitochondria (10), when MAVS is proteolytically cleaved from the mitochondrial membrane by hepatitis C virus protease (NS3/4A) (14, 15), or when cellular MAVS is absent (16, 17); these findings suggest that MAVS plays a central role in mitochondrial innate immunity.

We reported that the mitochondrial outer membrane guanosine triphosphatase (GTPase) mitofusin 2 (Mfn2), a mediator of mitochondrial fusion, interacts with MAVS through its heptad repeat region to inhibit antiviral immunity (18). In addition, Mfn1, the sequence of which has ~60% homology to Mfn2, regulates MAVS-mediated activation of IRF-3 in a different manner (19). One aspect of these findings is that both Mfn1 and Mfn2 associate with MAVS, which results in the regulation of signal transduction, thus implying that mitochondrial dynamics may play a major role in the antiviral signaling pathway. Here, we examined cells lacking both mitofusins and found that such cells were defective in MAVS-mediated antiviral signaling. Fusion-defective cells exhibited widespread loss of mitochondrial membrane potential (ΔΨm), and we found that cells with abnormal ΔΨm had defective antiviral innate immune responses. These findings underscore the importance of ΔΨm, a basic function of the organelle, with respect to innate immunity.


Cells lacking Mfns are defective in antiviral innate immune responses

On the basis of previous studies (18, 19), we reasoned that mitochondrial fusion could play a major role in the antiviral signaling pathway. To address this issue, we examined the role of MAVS-dependent signaling in mouse embryo fibroblasts (MEFs) with null mutations in both Mfn1 and Mfn2 [Mfns double mutant (Mfns-dm)], which have a complete lack of mitochondrial fusion (20, 21). We infected wild-type and Mfns-dm MEFs with either Sendai virus (SeV), a negative-stranded RNA virus of the Paramyxoviridae family, or encephalomyocarditis virus (EMCV), a positive-stranded RNA virus of the Picornavirus family, and examined the effects of mitofusins on the expression of the genes encoding IFN-β and the proinflammatory cytokine interleukin-6 (IL-6) by measuring the amounts of their mRNAs (Fig. 1A and fig. S1A). Virus-induced expression of endogenous IFN-β and IL-6 was severely impaired in the Mfns-dm MEFs relative to that in wild-type cells, and the production of endogenous IFN-β and IL-6 proteins was similarly reduced (Fig. 1B); however, substantial induction of IL-6 occurred when these cells were stimulated by lipopolysaccharide (LPS) (fig. S1B). In contrast, MEFs containing either Mfn1 or Mfn2 alone (Mfn2−/− and Mfn1−/− cells, respectively) had substantial antiviral responses (Fig. 1B and fig. S1C), with Mfn2−/− cells responding to a greater extent than did Mfn1−/− cells, consistent with our previous findings (18). Furthermore, Mfns-dm MEFs were much more permissive than were wild-type cells to infection by a recombinant vesicular stomatitis virus [VSVΔG*-G (22)], and the number of cells expressing green fluorescent protein (GFP) was substantially increased relative to that of wild-type cells (Fig. 1C). These results indicate that cells lacking both Mfn proteins had defects in their antiviral innate immune responses and that the presence of either Mfn homolog could be relevant to the mitochondrial antiviral response.

Fig. 1

Defective antiviral response in Mfns-dm cells. (A) Wild-type (WT) or Mfns-dm MEFs were infected with either SeV [16 hemagglutinin (HA) units/ml] or EMCV [at a multiplicity of infection (MOI) of 3] for 16 and 8 hours, respectively, and total RNAs from the cells were analyzed by reverse transcription PCR (RT-PCR) for the expression of IFN-β, IL-6, MAVS, and GAPDH (as a loading control). (B) WT and Mfns-dm MEFs were infected with SeV (16 HA units/ml) for 20 hours, and the amounts of IFN-β (left) and IL-6 (right) produced by the cells were measured by ELISA. As a control experiment, Mfn single mutant (Mfn1−/− or Mfn2−/−) MEFs were also evaluated under the same conditions. All data shown represent means ± SD (n = 3 experiments). **P < 0.01; ***P < 0.001; N.D., not detected. (C) Fluorescence microscopy and flow cytometric analysis of WT and Mfns-dm MEFs that were infected with VSVΔG*-G (a recombinant VSV expressing GFP) at an MOI of 0.5 for 24 hours [fluorescence-activated cell sorting (FACS) experiment; MOI of 1]. (D) WT and Mfns-dm MEFs were transfected with plasmid encoding Su9-eGFP, and the cells were stained with MitoTracker Red, which is sensitive to ΔΨm. Scale bar, 10 μm.

ΔΨm has a fundamental role in RLR-induced antiviral responses

Having identified that Mfns-dm MEFs were defective in antiviral signaling, we hypothesized that some mitochondrial defect existed in these cells that ultimately led to a dysfunction in cellular signaling, and that ΔΨm could be a candidate, because widespread heterogeneity in ΔΨm in Mfns-dm MEFs is known (21, 23). As previously reported (21, 24), we also confirmed that overexpression of mitochondrially targeted enhanced GFP (Su9-eGFP) in wild-type MEFs was uniformly detected with MitoTracker Red, a dye that is sensitive to ΔΨm, whereas most Mfns-dm MEFs displayed nonuniform staining (Fig. 1D). Many mitochondria were marked by GFP, but not by MitoTracker Red, because GFP expression was independent of ΔΨm, thus indicating widespread loss of ΔΨm.

To examine whether ΔΨm was essential for the mitochondrial antiviral response, we treated human embryonic kidney (HEK) 293 cells and wild-type MEFs with carbonyl cyanide m-chlorophenylhydrazone (CCCP) (fig. S2), a protonophore that dissipates ΔΨm by increasing membrane permeability to protons, and we assessed their immune responses to viral infection. The delivery of polyI:C (polyinosine-polycytidylic acid), a synthetic analog of viral dsRNA (25, 26), into HEK 293 cells by transfection stimulated the expression of IFN-α and IFN-β (Fig. 2A). However, treatment of HEK 293 cells with CCCP, but not oligomycin B, which is an inhibitor of the mitochondrial F1F0-ATPase (adenosine triphosphatase), suppressed the expression of endogenous IFN-α and IFN-β (Fig. 2A) and similarly reduced the production of endogenous IFN-β protein, as determined by enzyme-linked immunosorbent assay (ELISA) (Fig. 2B). Consistent with these findings, we found that the expression of IFN-β in response to infection with EMCV was also inhibited in CCCP-treated MEFs (Fig. 2C). CCCP inhibited the MAVS-dependent production of IFN-β in a dose-dependent manner (Fig. 2D and fig. S3), and removal of CCCP by washing the cells fully restored IFN-β production (Fig. 2D). Recovery of normal ΔΨm is observable by the regeneration of large isoforms of OPA1, a mitochondrial protein that is a target for ΔΨm-dependent proteolysis (27) (Fig. 2D).

Fig. 2

Loss of ΔΨm abolishes MAVS-mediated antiviral signaling. (A) HEK 293 cells treated with dimethyl sulfoxide (DMSO; control), CCCP (40 μM), or oligomycin B (10 μM) were transfected with polyI:C (10 μg), and the total RNAs isolated from each sample were analyzed by RT-PCR for the expression of IFN-β, IFN-α, MAVS, and GAPDH (as a control). (B) Similar to (A), except that the culture supernatants from HEK 293 cells were harvested 16 hours after transfection and were analyzed by ELISA to measure the amount of IFN-β produced. (C) Similar to (A), except that WT MEFs were infected with EMCV at an MOI of 3 for 8 hours. (D) HEK 293 cells treated with increasing amounts (10, 20, and 40 μM) of CCCP were transfected with polyI:C, and the IFN-β produced was analyzed by ELISA as described in (B). The far right lane (wash) indicates that CCCP (40 μM) was washed out after 9 hours. (E) We compared the activation of endogenous IRF-3 (as reflected by its dimerization and phosphorylation) by the delivery of polyI:C into HEK 293 cells that were untreated or were treated with CCCP. Cell lysates were resolved by electrophoresis under native (top panel) or denaturing conditions (middle and bottom panels) and then analyzed by Western blotting (IB) with antibodies for the indicated proteins. (F) WT MEFs treated with CCCP (40 μM) or rotenone (1 μM) were infected with SeV (16 HA units/ml) for 20 hours, and the amounts of IFN-β (left) and IL-6 (right) produced were measured by ELISA. All data shown in (B), (D), and (F) represent means ± SD (n = 3 experiments). **P < 0.01; ***P < 0.001; N.D., not detected.

In addition, we found that transfecting HEK 293 cells with polyI:C caused the dimerization and phosphorylation of endogenous IRF-3, two hallmarks of IRF-3 activation that were impaired when the cells were treated with CCCP (Fig. 2E). To verify that the observed inhibitory effects were not due to the induction of apoptosis under these experimental conditions, we confirmed that the CCCP-treated cells remained viable by testing them for the release of cytochrome c (fig. S4A) or the activation of caspase-3 (fig. S4B). Furthermore, the functionality of the cells treated with CCCP was unaffected when the cells were stimulated either with extracellular polyI:C (fig. S5A) or with LPS (fig. S5B), compounds that activate TLR3 and TLR4, respectively (28). The inhibition of mitochondrial ATP synthesis was not attributable to MAVS-dependent antiviral signaling, because rotenone, a specific inhibitor of mitochondrial complex I (COXI), had no inhibitory effect on the production of IFN-β and IL-6 (Fig. 2F); indeed, it may have increased the production of these cytokines because of the production of reactive oxygen species (ROS) (29). Consistent with this finding, the mitochondrial F1F0-ATPase inhibitor oligomycin B did not affect antiviral signaling (Fig. 2B and fig. S6). Together, these observations suggest that the extent of the dissipation in ΔΨm correlated with the extent of the defect in the RLR-induced antiviral response.

The increased abundance of uncoupling protein–2 leads to proton leakage from mitochondria and inactivation of MAVS-dependent signaling

The previous experiments demonstrated that the maintenance of normal ΔΨm was essential for the mitochondrial antiviral response. We therefore reevaluated the correlation between ΔΨm and the immune response by increasing the abundance of the mitochondrial inner membrane protein, uncoupling protein–2 (UCP-2), which induces mitochondrial proton leakage (3032). We constructed Myc-tagged full-length UCP-2 [UCP-2(full)] and its truncated mutant [UCP-2(1–200)], which contains only amino acid residues 1 to 200, as a nonfunctional control and confirmed that only full-length UCP-2 remained predominantly localized to the mitochondria in MEFs (Fig. 3A). Consistent with its subcellular localization, an increase in the abundance of UCP-2(full), but not UCP-2(1–200), had a moderate effect on the ΔΨm-sensitive color shift detected by MitoProbe JC-1 (Fig. 3B), a cationic dye that indicates mitochondrial depolarization through a reduction in its red-green fluorescence ratio. This result prompted us to explore the functional role of UCP-2 in the regulation of the antiviral response. Transfection of HEK 293 cells with a plasmid encoding Myc-tagged UCP-2(full) dose-dependently suppressed the MAVS- and RIG-I(1–250)–mediated activation of IFN-β and NF-κB reporter constructs, although this inhibitory activity of UCP-2 was not observed with its truncated form, UCP-2(1–200) (Fig. 3C and fig. S7). These results further highlight the importance of ΔΨm in the regulation of the antiviral signaling pathway.

Fig. 3

The increased abundance of UCP-2 inhibits MAVS-mediated activation of IFN-β and NF-κB reporters. (A) Subcellular localization of UCP-2(full) and its truncated mutant, UCP-2(1–200). The indicated constructs (Myc-tagged) were expressed in WT MEFs, and immunofluorescence against the Myc epitope (red) was used to identify cells that contained the proteins and to determine subcellular localization. Mitochondria in the same cells were identified with an antibody against mtHsp70 (green). Scale bar, 10 μm. (B) HEK 293 cells were transfected with 500 ng of plasmid encoding either Myc-tagged UCP-2(full) or UCP-2(1–200). Twenty-four hours after transfection, the cells were stained with JC-1 and analyzed by flow cytometry. The presence of Myc-tagged proteins was confirmed by Western blotting analysis with a monoclonal antibody against Myc (9E10). (C) HEK 293 cells were cotransfected with 50 ng of plasmids encoding either MAVS or RIG-I(1–250) together with increasing amounts (20, 50, and 100 ng) of plasmids encoding UCP-2(full) or UCP-2(1–200) in addition to the IFN-β luciferase reporter plasmid. The transfected cells were analyzed 24 hours later for IFN-β–dependent luciferase activity. All data shown in (C) represent means ± SD (n = 3 experiments). *P < 0.05; **P < 0.01.

Mechanistic investigation of the mitochondrial signalosome

In its quiescent state, endogenous MAVS ordinarily forms a stable higher-order complex on the outer mitochondrial membrane, and we previously proposed that rearrangement of MAVS-containing fractions from a higher- to a lower-order complex after activation of MAVS would lead to propagation of the downstream antiviral response (18). In the course of examining the transient state of the MAVS complex, we found that CCCP-induced dissipation of ΔΨm had no effect on either the localization of MAVS in the cell (Fig. 4A) or the amount of MAVS that existed in the nonproductive, higher-order complex (~600 kD) on the outer mitochondrial membrane (Fig. 4B). However, it is likely that a small amount of the lower-order MAVS complex fraction, which migrated at ~200 kD on a gel, was slightly reduced when the cells were treated with CCCP (Fig. 4B; compare first and third panels). Indeed, the reduction was greater than a factor of 3 when we quantified the bands by densitometry (Fig. 4C). Collectively, these results raise the possibility that loss of ΔΨm might prevent the structural rearrangement of the MAVS complex, which may exist as a readily available pool on the mitochondrial membrane to provide a quicker response upon viral infection. CCCP-induced dissipation of ΔΨm had no effect on the activation of endogenous IRF-3 in response to the increased abundance of TANK-binding kinase 1 (TBK-1), a kinase that targets IRF-3 and acts downstream of MAVS (Fig. 4D) (1, 2). The most likely explanation is that ΔΨm is involved at the same stage as MAVS in the RLR-mediated signaling pathway and that it couples with the MAVS complex to invoke an immune response from the mitochondrial signalosome (Fig. 4E) (33).

Fig. 4

The MAVS complex and ΔΨm. (A) Immunofluorescence microscopy of WT MEFs treated with DMSO or CCCP (40 μM). Cells were visualized with both endogenous proteins MAVS (red) and mtHsp70 (green). Scale bar, 10 μm. (B) Gel filtration elution profile of endogenous MAVS and OPA-1 (to confirm the dissipation of ΔΨm) extracted from the mitochondrial fraction of HEK 293 cells that had been treated with either DMSO or CCCP (40 μM). The positions corresponding to the elution of standard markers molecular mass and the void volume are indicated, and fractions were analyzed by Western blotting with antibodies against human MAVS and OPA-1. (C) Densitometric analysis of the MAVS complex in (B). DMSO-treated, higher-order complex and inputs are set at 100%, and data shown represent means ± SD (n = 3 experiments). (D) HEK 293 cells were transfected with plasmids encoding RIG-I(1–250) or TBK-1, and the activation of IRF-3 in the absence or presence of CCCP (40 μM) was compared as described for Fig. 2E. (E) Model of mitochondrial function that facilitates innate antiviral immunity. We propose that “normal” ΔΨm, which can be sustained by low-level fusion of mitochondria, is capable of supporting MAVS signaling (black arrow), whereas “abnormal” ΔΨm impairs MAVS signaling (red arrow). OM, mitochondrial outer membrane; IM, mitochondrial inner membrane; CARD, caspase recruitment domain.


Cellular innate immune responses to infection by RNA viruses result in the activation of multiple signaling steps, which ultimately regulate the production of type I IFNs and other proinflammatory cytokines, and mitochondria play an important role in this pathway through MAVS. Although the essential mediators that activate or inactivate MAVS-mediated signaling are well known, the roles of mitochondrial functions such as ATP synthesis or ΔΨm in the pathway remain poorly understood. Our findings reveal that dissipation of ΔΨm results in defective cellular antiviral immune responses.

Mitochondria are dynamic organelles that undergo cycles of homotypic fusion and fission events that are believed to play an important role in cellular functions (8, 9). Evidence suggests that mitochondrial dynamics also regulate antiviral signaling (19), which is consistent with our previous findings that Mfn2, a mediator of mitochondrial fusion, inhibits antiviral responses (18). Here, we showed that Mfns-dm MEFs, which completely lack mitochondrial fusion (20, 21), were defective in antiviral innate immune responses. Mfn1- and Mfn2-null MEFs contain predominantly fragmented mitochondria but readily have fused mitochondria with short rods, which are sufficient to escape major cellular dysfunction, as well as widespread heterogeneity of ΔΨm (21). Antiviral responses of these singly deficient cells were unaffected, or even increased, relative to that of wild-type cells (Fig. 1B and fig. S1C), which suggests that a single Mfn is sufficient to restore the defects observed in Mfns-dm MEFs. The most likely explanation for this is that residual mitochondrial fusion is involved in antiviral responses. Collectively, these findings suggest that an inadequate ΔΨm affects the mitochondrial antiviral response.

Our data from experiments in which cells were treated with the protonophore CCCP or in which we increased the abundance of UCP-2, a protein that causes mitochondrial proton leakage (3032), showed that the extent of dissipation of ΔΨm correlated with the extent of the defect in RLR-induced antiviral responses. The recovery from immunodeficiency that occurred through the removal of CCCP (Fig. 2D) and the ineffectiveness of CCCP in disrupting TBK-1 (Fig. 4D), an effector that acts downstream of MAVS, in the activation of IRF-3 emphasized the importance of ΔΨm as an essential factor for MAVS-mediated mitochondrial immune function. Note that the lack of MAVS-mediated antiviral signaling was not induced by inhibiting ATP synthesis, the most prominent functional role of mitochondria. In addition, it is interesting to compare our data with previous findings that showed that mice lacking UCP2 are resistant to infection by Toxoplasma gondii, a protozoan parasite that causes toxoplasmosis, in contrast to the lethality observed in infected wild-type littermates, which is a result of their increased generation of ROS (34). Future studies aimed at evaluating the role of UCP-2 in the differential susceptibility of these mice to pathogen infection may prove interesting.

Mitochondria are energy-generating systems that also appear to act as a platform for the first line of antiviral defense. MAVS is also localized at the membranes of peroxisomes, which are metabolic organelles, and peroxisomal MAVS is involved in the early induction of IFN-stimulated genes until the point at which mitochondrial MAVS induces a sustained antiviral response (35). Mitochondria and peroxisomes are interdependent in cellular metabolism and share several proteins, especially dynamics-related proteins (36). Therefore, it would not be surprising if the dissipation in ΔΨm led to perturbed MAVS signaling from peroxisomes as well as from mitochondria, although we expected to perturb the function of MAVS that was associated specifically with mitochondria. It will be critical to further understand how these organelles collectively fulfill an indispensable role in antiviral immunity and the basic physiology of mitochondria and how ΔΨm may be involved, for example, by performing experiments with cells deficient in peroxins. Together, our results provide a framework for understanding how the physiological function of mitochondria plays a role in innate antiviral immunity.

Materials and Methods


CCCP, rotenone, and Escherichia coli O111:B4 LPS were purchased from Sigma-Aldrich. Oligomycin B was obtained from Bioaustralis, MitoTracker Red CMXRos was purchased from Invitrogen, the MitoProbe JC-1 assay kit was from Molecular Probes/Invitrogen, and polyI:C was from InvivoGen. ELISA kits for human and mouse IFN-β were supplied by Kamakura Techno-Science Inc. and PBL Biomedical Laboratories, respectively, and an ELISA kit for mouse IL-6 was obtained from R&D Systems. All other reagents were of biochemical research grade.


The rabbit polyclonal antibodies against human and murine MAVS were described previously (18). Polyclonal antibodies against IRF-3, Myc (A-14), and cytochrome c (H-104) and monoclonal antibody against β-actin were purchased from Santa Cruz Biotechnology, and the rabbit monoclonal antibody against phosphorylated IRF-3 (pIRF-3) Ser396 (4D4G) and polyclonal antibody against caspase-3 were from Cell Signaling. Monoclonal antibody against Myc (9E10) was obtained from Covance. Monoclonal antibodies against OPA1 and mitochondrial heat shock protein 70 (mtHSP70) were supplied by BD Biosciences and Affinity BioReagents, respectively. The Alexa Fluor 488–conjugated monoclonal antibody against mouse immunoglobulin G (IgG) and the Alexa Fluor 568–conjugated polyclonal antibody against rabbit IgG were obtained from Molecular Probes/Invitrogen.

Cell lines and viruses

HEK 293 cells and MEF cell lines (wild type, Mfn1−/−, and Mfn2−/−), which were provided by D. Chan (Howard Hughes Medical Institute, Caltech, Pasadena, CA), were maintained in Dulbecco’s modified Eagle’s medium (DMEM, GIBCO BRL) supplemented with 1% l-glutamine, 1% penicillin-streptomycin, and 10% bovine calf serum and at 5% CO2 and 37°C. The Mfns-dm MEFs were maintained in a rich medium (DMEM supplemented with 10% fetal calf serum) as described previously (20, 21). SeV Cantell strain was purchased from the American Type Culture Collection, and VSVΔG*-G and EMCV were described previously (18, 37).


The plasmids encoding Myc-tagged human MAVS, hRIG-I(1–250), and hTBK-1 were described previously (18). To construct human UCP-2 containing a Myc epitope tag, we used a HEK 293 complementary DNA (cDNA) library as a template to amplify the gene encoding UCP-2. Using the oligonucleotides TK571 (see Supplementary Materials) and TK572, we subcloned an amplified fragment containing 5′ Not I and 3′ Eco RV sites into the pcDNA3.1 (−) plasmid (Invitrogen) containing an N-terminal 3×Myc epitope tag (18).

Reverse transcription polymerase chain reaction

Total RNA was isolated from HEK 293 cells and MEFs with TRIzol reagent (Invitrogen), and 2 μg of total RNA was reverse-transcribed with M-MLV RT (Wako Pure Chemical Industries) to generate cDNAs. Polymerase chain reaction (PCR) assays were performed with Ex Taq DNA polymerase (TaKaRa). The following primers were used to amplify cDNAs: human IFN-β, TK552/TK553; human IFN-α, TK556/TK557; human MAVS, TK350/TK354; human glyceraldehyde phosphate dehydrogenase (GAPDH), TK550/551; murine IFN-β, TK535/TK536; murine IL-6, TK554/TK555; murine MAVS, TK300/TK305; and murine GAPDH, TK537/538.

Analytical size exclusion chromatography

Size exclusion chromatography of mitochondrial extracts was performed on a Superdex-200 HR-10/30 column (GE Healthcare) as described previously (18), with a slight modification. In brief, the HEK 293 cells had their medium exchanged with medium containing 40 μM CCCP for 3 hours before the assay. The cells were then washed once with cold phosphate-buffered saline (PBS) (pH 7.2), scraped off their dishes, and lysed in 1 ml of homogenization buffer [20 mM Hepes (pH 7.5), 70 mM sucrose, and 220 mM mannitol] by 30 strokes in a Dounce homogenizer. The homogenate was centrifuged at 800g for 5 min, and the resulting supernatant was further centrifuged at 10,000g for 10 min at 4°C to precipitate the crude mitochondrial fraction. After the pellet was washed once with homogenization buffer, the mitochondrial extracts were solubilized with lysis buffer [50 mM tris-HCl (pH 7.2), 200 mM NaCl, 10% glycerol, and 1% digitonin] followed by centrifugation at 12,000g for 5 min. The supernatant of mitochondrial extracts was loaded onto the column equilibrated with 50 mM tris-HCl (pH 7.2) containing 200 mM NaCl, 10% glycerol, and 0.1% NP-40, and chromatography was performed at a flow rate of 0.3 ml/min at room temperature. The fractions were resolved by 8% SDS–polyacrylamide gel electrophoresis (SDS-PAGE) followed by Western blotting analysis with antibody against either human MAVS or OPA1. Scans of Western blots were analyzed with Fujifilm Multi Gauge Ver3.X.

Luciferase assays

HEK 293 cells were plated in 24-well plates (at 2 × 105 cells per well). The following day, the cells were cotransfected with 100 ng of a luciferase reporter plasmid (p125luc or pELAM), 2.5 ng of the Renilla luciferase internal control vector phRL-TK (Promega), and each of the indicated vectors with Lipofectamine 2000 reagent (Invitrogen). Empty vector [pcDNA3.1(−)] was used to maintain equivalent amounts of DNA in each well. Cells were harvested 24 hours after transfection and analyzed by a dual-luciferase reporter assay on the GloMax 20/20n luminometer (Promega). Each experiment was repeated at least three times.

Immunofluorescence microscopy

MEFs were plated on coverslips in 12-well plates (at 1.5 × 105 cells per well). Twenty-four hours after transfection, cells were fixed with 3.7% formaldehyde for 10 min, permeabilized with 0.1% Triton X-100 in PBS (pH 7.2), and blocked with 5% bovine calf serum. Myc-tagged constructs were detected with antibody against Myc (A-14) and Alexa Fluor 568–conjugated secondary antibody, and mitochondria were detected with antibody against mtHSP70 and Alexa Fluor 488–conjugated secondary antibody. ΔΨm was monitored as described previously (21, 24). Cells were imaged with an Olympus BX-FLA fluorescence microscope.

Flow cytometry

ΔΨm was analyzed with a BD FACSCalibur flow cytometer (BD Biosciences) with the cationic dye JC-1 (Molecular Probes). Cells (~1 × 106 cells/ml), treated with CCCP or oligomycin B for 16 hours, were washed once with PBS (pH 7.2) and harvested into a centrifuge tube. Cells were then resuspended in 500 μl of PBS (pH 7.2) containing 2 μM JC-1 and incubated at 37°C for 30 min. Cells were washed once with PBS (pH 7.2), and flow cytometric analysis was performed after staining. To analyze ΔΨm of HEK 293 cells expressing UCP-2, we transfected cells with plasmid encoding UCP-2 with Lipofectamine 2000 and performed flow cytometric analysis as described above.

Native PAGE

Native PAGE experiments were performed as described previously (38).


Measurements of species-specific production of IFN-β and IL-6 were performed as described previously (18).

Supplementary Materials

Fig. S1. Suppression of IFN-β production in Mfns-dm MEFs in response to polyI:C.

Fig. S2. HEK 293 cells and wild-type MEFs stained with MitoProbe JC-1.

Fig. S3. CCCP inhibits the MDA-5–dependent production of IFN-β in a dose-dependent manner.

Fig. S4. Analysis of cytochrome c release and caspase-3 activation in wild-type MEFs.

Fig. S5. CCCP-treated cells are unaffected by extracellular polyI:C or LPS.

Fig. S6. CCCP inhibits antiviral signaling in HEK 293 cells.

Fig. S7. UCP-2 inhibits MAVS-mediated activation of the NF-κB reporter.

References and Notes

  1. Acknowledgments: We are grateful to D. Chan (Howard Hughes Medical Institute and California Institute of Technology, Pasadena, CA) for providing the MEF cell lines (Mfn1-null, Mfn2-null, and Mfns-dm) and the Su9-eGFP expression plasmid and for his helpful discussions. We are also grateful to Y. Michikawa (National Institute of Radiological Sciences, Japan) for valuable comments on the study and N. Bashiruddin (Kyushu University, Japan) for advice on densitometric analysis. We thank R. Koga for technical assistance with the flow cytometric analysis and Y. Fuchigami for technical assistance with DNA sequencing and molecular biology. The EMCV was provided by T. Fujita and M. Yoneyama (Kyoto University, Japan). We thank T. Seya (Hokkaido University, Japan) for the human TLR3 expression plasmid and T. Taniguchi (University of Tokyo, Japan) for providing the p125luc reporter plasmid. Funding: This work was supported by grants-in-aid for Young Scientists (B) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan (20770123); Kyushu University Interdisciplinary Programs in Education and Projects in Research Development (P&P type D; 20301); Uehara Memorial Foundation; and Takeda Science Foundation to T.K. Author contributions: T.K. and K.Y. performed the experiments; Y.Y. and S.-i.K. provided viruses and E. coli O111:B4 LPS, respectively, and commented on the manuscript; T.K. and K.Y. analyzed the data and interpreted experimental results; and T.K. designed the experiments and wrote the paper. Competing interests: The authors declare that they have no competing interests.
View Abstract

Stay Connected to Science Signaling

Navigate This Article