The Atg5 Atg12 conjugate associates with innate antiviral immune responses

Abstract

Autophagy is an essential process for physiological homeostasis, but its role in viral infection is only beginning to be elucidated. We show here that the Atg5-Atg12 conjugate, a key regulator of the autophagic process, plays an important role in innate antiviral immune responses. Atg5-deficient mouse embryonic fibroblasts (MEFs) were resistant to vesicular stomatitis virus replication, which was largely due to hyperproduction of type I interferons in response to immunostimulatory RNA (isRNA), such as virus-derived, double-stranded, or 5'-phosphorylated RNA. Similar hyperresponse to isRNA was also observed in Atg7-deficient MEFs, in which Atg5-Atg12 conjugation is impaired. Overexpression of Atg5 or Atg12 resulted in Atg5-Atg12 conjugate formation and suppression of isRNA-mediated signaling. Molecular interaction studies indicated that the Atg5-Atg12 conjugate negatively regulates the type I IFN production pathway by direct association with the retinoic acid-inducible gene I (RIG-I) and IFN-beta promoter stimulator 1 (IPS-1) through the caspase recruitment domains (CARDs). Thus, in contrast to its role in promoting the bactericidal process, a component of the autophagic machinery appears to block innate antiviral immune responses, thereby contributing to RNA virus replication in host cells.

Fig. 1.

VSV facilitates autophagy for efficient replication. (A) WT or Atg5 KO MEFs were transfected with the LC3-GFP expression plasmid and were infected with or without VSV at moi = 1.0, and the LC3 signal was analyzed 8 h after infection by fluorescent deconvolution microscopy. (Scale bar, 10 μm.) (B) WT or Atg5 KO MEFs were infected with VSV at moi = 1.0, and cell lysates were prepared at 0, 4, or 8 h after infection. Samples were analyzed by immunoblotting to compare the levels of two different forms of LC3 [LC3-I (18 kDa) and LC3-II (16 kDa)]. (C) WT (open bars) or Atg5 KO (filled bars) MEFs were infected with VSV or HSV at moi = 0.1 or 1.0. Twenty-four hours after infection, the cells and culture supernatants were recovered, and the levels of viral titer were examined by the plaque assay. The graph shows the mean ± SD. *, P < 0.05 by Student's t test. (D and E) WT (open bars) or Atg5 KO MEFs (filled bars) were infected with VSV or HSV at moi = 1.0. (D) Total RNA was isolated 0, 4, and 8 h after stimulation, mRNA was reverse-transcribed, and then the levels of IFN-α4, IFN-β, or IP-10 cDNA were quantified by real-time PCR. (E) The nuclear extracts were isolated, and 12-μg samples were subjected to SDS/PAGE and immunoblotting analysis for phosphorylated IRF-3 or Sp1. (F) WT (open bars) or Atg5 KO MEFs (filled bars) were infected with VSV at moi = 1.0 in the presence of anti-IFN-α- and IFN-β-neutralizing IgG or control rabbit IgG. Twenty-four hours after infection, the culture supernatants were recovered, and the levels of viral titer were examined by the plaque assay. The graph shows the mean ± SD. *, P < 0.05 by Student's t test.

Fig. 2.

Hyperproduction of dsRNA-mediated type I IFNs in Atg5 KO MEF and the Atg5–Atg12 conjugate negatively regulates IPS-1-mediated promoter activation of type I IFNs. (A–D) WT (open bars) or Atg5 KO (filled bars) MEFs were transfected with 10 μg/ml poly(I:C). Total RNA was isolated 0, 2, 4, and 8 h after stimulation, mRNA was reverse-transcribed, and then the levels of targeted cDNAs [IFN-α4 (A), IFN-β (B), IL-6 (C), IP-10 (D)] were quantified by real-time PCR. (E) HEK293 cells were transiently transfected with 25 ng of Renilla luciferase reporter plasmid, 25 ng of firefly luciferase reporter plasmid for NF-κB, IFN-α4, or IFN-β promoter (Prom), 50 ng of expression plasmid for FLAG hRIG-I ΔC plus 0, 50, 100, or 150 ng of HA hAtg5, HA hAtg12, or HA hAtg5-K130R expression plasmid or empty vector to give a constant 250 ng of DNA per transfection. Forty-eight hours after transfection, cells were lysed and luciferase activity was measured by using a luminometer. The firefly luciferase activity was normalized to Renilla luciferase activity in each sample. The same samples were subjected to immunoblot analysis targeting the FLAG or HA epitope. The graph shows the mean ± SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001 by Student's t test.

Fig. 3.

The Atg5–Atg12 conjugate associates with IPS-1, RIG-I, and MDA5. (A) HEK293 cells were transiently transfected with expression plasmids for HA hAtg5 and HA hAtg12 plus FLAG-tagged hRIG-I, hMDA5, hIPS-1, or GFP in either full-length or truncated form. Forty-eight hours after transfection, cell lysates were prepared, immunoprecipitated with anti-FLAG antibody, and then subjected to immunoblotting analysis using anti-FLAG or anti-HA antibody. (B) HEK293 cells were incubated with or without VSV at moi = 1.0. The cell lysates were immunoprecipitated with anti-Atg5 or control antibody and were then subjected to immunoblotting analysis using anti-Atg5 or MAVS (IPS-1) antibody. (C) HeLa cells were treated with Mitotracker reagent, fixed, and incubated with goat anti-APG5 (Atg5) and rabbit anti-MAVS (IPS-1) antibodies. The cells were then washed, treated with Alexa 488-conjugated anti-goat IgG and Alexa 405-conjugated anti-rabbit IgG antibodies, and then analyzed under a confocal microscope. Fluorescent signals were depicted as black and white in “Atg5,” “IPS-1,” and “Mitotracker” views. In the “Merge” view, Atg5, IPS-1, or Mitotracker was represented in blue, green, or red, respectively. (D) FRET signal was examined after CFP-IPS-1 plus YFP-Mit, CFP-IPS-1 plus YFP-KDEL, CFP-IPS-1 plus YFP-Atg5, or CFP-IPS-1 T54A plus YFP-Atg5 were expressed in HeLa cells. (Left and Center) Cells were imaged for CFP (donor) (Left) and YFP (acceptor) (Center). The acceptor bleaching was completed within 20 sec. (Right) FRET efficiency based on the YFP bleaching. The ratio of the increase in CFP fluorescence signal to its signal after bleaching is represented as reported in ref. . Three to five independent experiments gave similar results.

Fig. 4.

The Atg5–Atg12 conjugate associates with IPS-1 and RIG-I through the CARDs. (A) A schematic diagram of IPS-1 and truncated or site-directed mutants. (B) HEK293 cells were transiently transfected with expression plasmids for HA hAtg5 and HA hAtg12 plus FLAG-tagged hIPS-1 in either full-length or truncated form. (C) HEK293 cells were transfected with expression plasmids for FLAG hRIG-I ΔC (0.75 μg) and HA hIPS-1 (0.75 μg) in the presence of 0, 1, or 4 μg of that for HA hAtg5 and HA hAtg12 or biotinylated actin related protein complex p34 (ARPC-2). Forty-eight hours after transfection, cell lysates were prepared, immunoprecipitated with anti-FLAG antibody, and then subjected to immunoblotting analysis using anti-FLAG antibody, anti-HA antibody, or streptavidin-HRP. Three to five independent experiments gave similar results.

Fig. 5.

A scheme of the molecular association among RIG-I, IPS-1, and the Atg5–Atg12 conjugate during RNA virus infection.