CaMKII-dependent non-canonical RIG-I pathway promotes influenza virus propagation in the acute-phase of infection

Abstract

Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) is one of hundreds of host-cell factors involved in the propagation of type A influenza virus (IAV), although its mechanism of action is unknown. Here, we identified CaMKII inhibitory peptide M3 by targeting its kinase domain using affinity-based screening of a tailored random peptide library. M3 inhibited IAV cytopathicity and propagation in cells by specifically inhibiting the acute-phase activation of retinoic acid-inducible gene I (RIG-I), which is uniquely regulated by CaMKII. Downstream of the RIG-I pathway activated TBK1 and then IRF3, which induced small but sufficient amounts of transcripts of the genes for IFN α/β to provide the capped 5'-ends that were used preferentially as primers to synthesize viral mRNAs by the cap-snatching mechanism. Importantly, knockout of RIG-I in cells almost completely inhibited the expression of IFN mRNAs and subsequent viral NP mRNA early in infection (up to 6 h after infection), which then protected cells from cytopathicity 24 h after infection. Thus, CaMKII-dependent acute-phase activation of RIG-I promoted IAV propagation, whereas the canonical RIG-I pathway stimulated antiviral activity by inducing large amounts of mRNA for IFNs and then for antiviral proteins later in infection. Co-administration of M3 with IAV infection rescued mice from the lethality and greatly reduced proinflammatory cytokine mRNA expression in the lung, indicating that M3 is highly effective against IAV in vivo. Thus, regulation of the CaMKII-dependent non-canonical RIG-I pathway may provide a novel host-factor-directed antiviral therapy.IMPORTANCEThe recent emergence of IAV strains resistant to commonly used therapeutic agents that target viral proteins has exacerbated the need for innovative strategies. Here, we originally identified CaMKII-inhibitory peptide M3, which efficiently inhibits IAV-lethality in vitro and in vivo. M3 specifically inhibited the acute-phase activation of RIG-I, which is a novel pathway to promote IAV propagation. Thus, this pathway acts in an opposite manner compared with the canonical RIG-I pathway, which plays essential roles in antiviral innate immune response later in infection. The CaMKII-dependent non-canonical RIG-I pathway can be a promising and novel drug target for the treatment of infections.

Keywords: CaMKII; RIG-I; cap-snatching; influenza virus; peptide library screening.

Fig 1

Multivalent peptide library screen identified CaMKII inhibitory peptides. (a) The tetravalent peptide screening library comprised tetravalent compounds with four randomized peptides of sequence. Met-Ala-X-X-X-Leu-X-X-X-X-Ala-X-X-X-Ala-Lys-Lys-Lys-U (U; amino hexanoic acid), where X indicates any amino acid except Cys. Ala was introduced at position 0, the phosphorylation site of substrates, to prevent the library from serving as substrates (upper panel). For compounds bound to CaMKII-KD, values in parentheses next to amino acids indicate the relative selectivity and amino acids with strong selectivity are in bold. The first and second designations (middle panel) refer to the two screens, and monovalent and tetravalent peptides with the six CaMKII-KD binding motifs were synthesized as candidate compounds (lower panel). (b) The inhibitory effects of monovalent peptides (upper left panel) or tetravalent peptides (upper right panel) on the kinase activity of CaMKII-KD with autocamtide-2 peptide as the phosphorylation substrate. Data are presented as a percentage of the control value without peptide (n = 3–7, mean ± SEM). IC₅₀ values of peptides are shown (lower panel). (c) Kinetics of the binding of monovalent peptide to immobilized CaMKII-KD was analyzed using the BIAcore system with an arbitrary resonance unit (RU) to indicate peptide binding (n = 3, mean ± SEM). (d) The inhibitory effects of monovalent peptides on the production of TNFα.RAW264.7 cells were incubated with each peptide for 30 min and then treated with LPS for 24 h. TNFα production in the culture medium was determined by ELISA. Data are presented as a percentage of the control value without peptide (n = 3–12, mean ± SEM). **P < 0.01; ***P < 0.005 (compared with the control by ANOVA followed by one-sided Dunnett’s test).

Fig 2

M3 inhibits IAV propagation in MDCK cells. (a) The effects of monovalent peptides on the cytopathicity induced by infection. MDCK cells were treated with peptides for 30 min and then infected with IAV strain PR8 at MOI = 20 (left panel) for 24 h or MOI = 0.001 (right panel) for 40 h. Data are presented as a percentage of the control value without infection (left panel; KN-93: n = 3, mM3: n = 4, M3: n = 10, M6: n = 11, virus alone: n = 16, right panel; n = 3, mean ± SEM). *P < 0.05; **P < 0.01; ***P < 0.001 (compared with the non-treated control cells by ANOVA followed by one-sided Dunnett’s test). (b) The effect of M3 on virus propagation. MDCK cells were incubated with M3 for 30 min and then infected with IAV strain PR8 at MOI = 0.2 for 16 h. The virus titer in the supernatant was determined by a plaque assay. Data are presented as the fold increase over the initial virus titer (60,000 pfu/mL) (n = 3–6, mean ± SEM). *P < 0.05; **P < 0.001 (compared with no M3 treatment by ANOVA followed by one-sided Dunnett’s test). (c) The effects of M3 on the cytopathicity induced by infection of other strains. MDCK cells were treated with M3 for 30 min and then infected with IAV strain Tokyo/UTHP013/2016 (H1N1 pdm, MOI = 100), Aichi/2/1968 (H3N2, MOI = 1), or IBV strain Wisconsin/01/2010 (Yamagata lineage, MOI = 1) for 24 h. Data are presented as a percentage of the control value without infection (n = 3, mean ± SEM). *P < 0.05; **P < 0.001 (compared with no M3 treatment by ANOVA followed by one-sided Dunnett’s test). (d) The effect of M3 on virus propagation after IAV infection (left panel). MDCK cells transfected with FLAG-tagged hCaMKIIor mock-transfected cells were incubated with or without 3 µM M3 for 30 min prior to infection with IAV strain PR8 at MOI = 0.2 for 16 h. The virus titer in the supernatant was determined by a plaque assay. Data are presented as the fold increase over the initial virus titer (n = 3, mean ± SEM). ***P < 0.001 (compared with no M3 treatment by Student’s t-test). The effect of M3 on activation of CaMKII induced by IAV infection (right panel). MDCK cells transfected with FLAG-tagged hCaMKII or mock-transfected cells were incubated with 3 µM M3 for 30 min prior to infection with IAV strain PR8 at MOI = 10 for 1 h. The cell lysates were analyzed by western blot using a phosphorylated Ser286-specific antibody or anti-FLAG antibody. Data are representative of the results from three independent experiments. (e) The binding of M3 to intracellular CaMKII. MDCK cells were treated with 3 µM biotinylated M3 for 90 min at 37°C. Cell lysates were incubated with avidin-agarose beads, and coprecipitating proteins were analyzed by western blot. PD, pull down.

Fig 3

M3 inhibits viral RNA and protein synthesis to reduce virus propagation. (a) The effects of M3 on viral protein synthesis. MDCK cells were incubated with or without 3 µM M3 for 30 min and then infected with IAV strain PR8 at MOI = 1 for the indicated time. Viral proteins in the lysate were analyzed by western blot, and the intensity of each band was quantitated and presented as the amount of each viral protein relative to β-actin (right panels). Data are representative of the results of three independent experiments (left panel) and are presented as a ratio compared to the value at 16 h post-infection without M3 (n = 3, mean ± SEM). *P < 0.05 (compared with no M3 treatment by Student’s t-test). (b) The effects of M3 on the expression of NP RNA. MDCK cells were incubated with or without 3 µM M3 for 30 min and then infected with IAV strain PR8 at MOI = 1 for the indicated time. The relative amounts of NP total RNA (left panel), vRNA, mRNA, and cRNA (right panels) were determined by reverse transcriptase-quantitative PCR (RT-qPCR) using GAPDH as the reference gene. Data are presented as a fold increase over the average RNA level at 1 h (left panel) or 3 h (right panels) post-infection without M3 (n = 3, mean ± SEM). *P < 0.05 (compared with no M3 treatment by Student’s t-test). (c) The effect of M3 treatment at various time points on the cytopathicity of IAV. MDCK cells were infected with strain PR8 at MOI = 20 for 1 h and then incubated for 23 h. The treatment of 3 µM M3 was performed as shown in the schematic diagram (left panel). Data are presented as a percentage of the control value without infection (right panel, n = 4, mean ± SEM). **P < 0.01 (compared with no M3 treatment by ANOVA followed by one-sided Dunnett’s test). n.s., not significant. hpi, hour post-infection.

Fig 4

M3 specifically inhibits cap-snatching early in infection. (a, b) The effects of M3 on the expression of mRNAs for IFNα or β (a) or Mx1 (b). MDCK cells were incubated with or without 3 µM M3 for 30 min and then infected with IAV strain PR8 at MOI = 1 for the indicated time. Data up to 9 h after infection are enlarged (a, lower panels). The relative amount of each mRNA was determined by RT-qPCR using GAPDH as the reference gene. Data are presented as the fold increase over the average RNA level before infection (n = 3, mean ± SEM). *P < 0.05 (compared with no inhibitor by Student’s t-test). (c) The effect of BXA on the expression of IFNα and β mRNAs.MDCK cells were incubated with or without 100 nM BXA for 30 min and then infected with IAV strain PR8 at MOI = 1 for the indicated time. Data for 3 h after infection are enlarged (left panels). The relative amount of each mRNA was determined by RT-qPCR using GAPDH as the reference gene. Data are presented as the fold increase over the average RNA level before infection (n = 3, mean ± SEM). **P < 0.01; ***P < 0.005 (compared with no inhibitor by Student’s t-test). The effect of M3 on the enhanced expression of IFNβ mRNA by BXA (right panel). MDCK cells were incubated with 100 nM BXA and/or 3 µM M3 for 30 min and then infected with IAV for 3 h. Data are presented as the fold increase over the average RNA level before infection (n = 3, mean ± SEM). *P < 0.05; **P < 0.01 (by Tukey’s test). (d) The effect of overexpression of kinase-inactive Flag-hCaMKII-K43R on the expression of IFNβ mRNA. MDCK cells transfected with Flag-hCaMKII-K43R or mock-transfected cells were infected with IAV strain PR8 at MOI = 1 for 6 h. The cell lysates were analyzed by western blot using anti-FLAG antibody. The relative amount of IFNβ mRNA was determined by RT-qPCR using GAPDH as the reference gene, and the fold increase over the average RNA level before infection was determined. Data are presented as a percentage of the control value (n = 3, mean ± SEM). **P < 0.01 (compared with mock by Student’s t-test).

Fig 5

M3 inhibits a non-canonical RIG-I pathway that acts early in infection to promote virus propagation. (a, b) The effects of M3 on the activation of IRF3 (a) or TBK1 (b) induced by the infection. A549 cells were incubated with or without 3 µM M3 for 30 min, and then infected with IAV strain PR8 at MOI = 1 for the indicated time. The cell lysates were analyzed by western blot, and the data are representative of the results from three independent experiments. (c) The effects of the TBK1 inhibitor GSK8612 on the cytopathicity induced by infection. MDCK cells were infected with IAV strain PR8 at MOI = 10 for 1 h and then incubated for 23 h with 10 µM GSK8612 treatment during the time periods shown in the upper panels. Data are presented as the percentage of the control value without infection (n = 4, mean ± SEM). *P < 0.05; **P < 0.01 (compared with no GSK8612 treatment by ANOVA followed by one-sided Dunnett’s test). (d) The effects of the RIG-I KO on the IFNβ mRNA expression. The expression of RIG-I in MDCK-derived RIG-I KO clones was determined by western blot. The upper and lower bands represent full-length RIG-I and its spliced variant, respectively (left panel). Cells of parental MDCK or both MDCK-derived RIG-I KO clones were infected with IAV strain PR8 at MOI = 1 for the indicated time (right panel). Data for 3 h after infection are shown in the lower panel. The relative amount of IFNβ mRNA was measured by RT-qPCR using GAPDH as the reference gene, with the amount at 9 h post-infection in parental cells equal to 100 (n = 3, mean ± SEM). ***P < 0.005 (compared with the control cells by ANOVA followed by one-sided Dunnett’s test). (e) The effect of the RIG-I KO on NP mRNA expression. Cells of MDCK or MDCK-derived RIG-I KO clone #1 were infected with IAV strain PR8 at MOI = 1 for the indicated time. The relative amount of NP mRNA was measured by RT-qPCR using GAPDH as the reference gene. Data are presented as the fold increase over the average RNA level before infection (n = 3, mean ± SEM). *P < 0.05 (compared with no inhibitor by Student’s t-test). (f) The effects of the RIG-I KO on the cytopathicity induced by infection. Cells of parental MDCK or the two MDCK-derived RIG-I KO clones were infected with IAV strain PR8 at MOI = 1 for 24 h. Data are presented as a percentage of the control value without infection (n = 3, mean ± SEM). *P < 0.05; **P < 0.01; ***P < 0.001 (compared with parental MDCK by ANOVA followed by one-sided Dunnett’s test).

Fig 6

M3 rescues mice from the lethality of IAV infection. (a) The effects of M3 on the lethality of IAV infection. Female BALB/c mice were infected intranasally with 2,000 plaque-forming units (pfu) of IAV strain PR8 with or without 0.5, 1.25, or 2.5 mg/kg M3, or 0.55 mg/kg KN-93. The body weight of each mouse is presented as a percentage of the value on day 0 (left panels), and the survival rate of each group is shown in the right panel. Each group contained 5–9 mice. **P < 0.01 (by log-rank test). (b, c) The effects of M3 on the expression of viral and host mRNAs in the lung after infection. Female BALB/c mice were infected intranasally with 2,000 pfu of IAV strain PR8 with or without 1.25 mg/kg M3. The relative amounts of viral NP mRNA (b) and various proinflammatory cytokine mRNAs (c) in lungs harvested 3 days after infection were determined by RT-qPCR using GAPDH as the reference gene. Data are presented as the fold increase over the average RNA level of the control mice (n = 5, mean ± SEM). *P < 0.05; **P < 0.01 (by Mann–Whitney U test).

Fig 7

M3 inhibits CaMKII-dependent activation of the non-canonical RIG-I pathway that enhances IAV propagation. At an early stage of IAV infection, CaMKII activates a non-canonical RIG-I pathway to activate TBK1 and IRF3 and then induces small but sufficient amounts of IFN α/β mRNAs whose capped 5′-ends are used preferentially to promote viral mRNA transcription through the cap-snatching mechanism. This pathway acts early in infection in an opposite manner compared with the canonical RIG-I pathway, which functions later in infection to induce high levels of mRNA for IFN α/β and then for antiviral proteins. M3 efficiently suppresses virus propagation by inhibiting the CaMKII-depenent activation of the non-canonical RIG-I pathway at the early stage of infection.