Identification of cytochrome c oxidase subunit 4 isoform 1 as a positive regulator of influenza virus replication

Abstract

Human infection with highly pathogenic H5N1 influenza virus causes severe respiratory diseases. Currently, the drugs against H5N1 are limited to virus-targeted inhibitors. However, drug resistance caused by these inhibitors is becoming a serious threat to global public health. An alternative strategy to reduce the resistance risk is to develop antiviral drugs targeting host cell proteins. In this study, we demonstrated that cytochrome c oxidase subunit 4 isoform 1 (COX41) of host cell plays an important role in H5N1 infection. Overexpression of COX41 promoted viral replication, which was inhibited by silencing or knockout the expression of COX41 in the host cell. The ribonucleoproteins (RNPs) of H5N1 were retained in the cell nucleus after knockout cellular COX41. Strikingly, inhibition of cellular COX41 by lycorine, a small-molecule compound isolated from Amaryllidaceae plants, reduced the levels of COX41-induced ROS and phosphorylation of extracellular signal-regulated kinase (ERK) in cells, thus resulting in the blockage of nuclear export of vRNP and inhibition of viral replication. In H5N1-infected mice that were treated with lycorine, we observed a reduction of viral titers and inhibition of pathological changes in the lung and trachea tissues. Importantly, no resistant virus was generated after culturing the virus with the continuous treatment of lycorine. Collectively, these findings suggest that COX41 is a positive regulator of H5N1 replication and might serve as an alternative target for anti-influenza drug development.

Keywords: COX41; anti-influenza; highly pathogenic H5N1 influenza virus; lycorine; viral ribonucleoproteins export.

Figure 1

Cytochrome c oxidase subunit 4 isoform 1 (COX41) was involved in the replication of H5N1. (A) Scheme illustrating the proteomic analysis of Madin Darby Canine Kidney (MDCK) cells in responses to H5N1 infection. MDCK cells were infected with H5N1 (MOI, 1) and detected at 12 h. The differentially expressed proteins were presented as a heat map. (B) Expression of COX41 was upregulated by H5N1 infection. MDCK cells were infected with H5N1 for 6 h (MOI, 5), 12 h (MOI, 5), and 24 h (MOI, 0.01), respectively. The level of COX41 was assessed by Western blot and quantitatively analyzed using ImageJ software. Data are mean ± s.d. (n = 2 ~ 3). ^**p < 0.01; ^***p < 0.001. (C) Overexpression of COX41 in cells increased the viral titers. HEK293T cells were transfected with plasmids expressing FLAG-tagged COX41 and then infected with H5N1 (MOI, 0.01). Viral titers were assessed at 72 h postinfection by TCID₅₀ assay. Data are mean ± s.d. (n = 3). ^*p < 0.05. (D) Detection of COX41 mRNA level in HEK293T cells transfected with COX41 siRNAs targeting COX41 genes at 24 h.p.i. Data are mean ± s.d. (n = 3). (E) Effects of COX41 siRNAs on viral titers as determined by TCID₅₀ assays. HEK 293 T cells were transfected with siRNAs and then infected with H5N1. The viral titers in cells were determined at 48 h.p.i. Data are mean ± s.d. (n = 3). ^*p < 0.05; ^**p < 0.01; ^***p < 0.001.

Figure 2

Cytochrome c oxidase subunit 4 isoform 1 (COX41) was involved in the nuclear export of vRNP. (A) Generation of COX41-defective HEK 293T cells. (B) The levels of COX41 in COX41-defective HEK 293T cells were determined by western blot assay. (C) Growth kinetics of the H5N1 (MOI, 0.01) in wild-type (WT) cells, COX41-KO-1 cells, and COX41-KO-1 cells with transfection of Flag-COX41. The supernatants in HEK293T cells were collected at 12, 24, 36, 48, and 72 h postinfection and subjected to viral titer determination by TCID₅₀ assays. Data are mean ± s.d. (n = 3). ^***p < 0.001. (D) WT and KO-1 cells were infected with H5N1 (MOI, 5 for 12 h.p.i. and MOI = 0.01 for 36 h.p.i.) after 1 h of adsorption. After infection, subcellular localization of NP was detected using a confocal microscope. Mean fluorescence intensity associated with NP expression level was analyzed by Image J software. Data are mean ± s.d. (n = 3 ~ 5). ^*p < 0.05; ^***p < 0.001.

Figure 3

The expression of COX41 was inhibited by lycorine treatment in vitro. (A) Chemical structure of lycorine. (B) Quantitative proteomic analysis of H5N1-infected cells in responses to lycorine treatment. MDCK cells were infected with H5N1(MOI, 1) in the presence of lycorine (0.52 μM) for 12 h. The upregulated proteins in H5N1-infected cells but downregulated proteins after lycorine treatment are presented as a heat map. (C) Western blot and quantitative analysis of COX41 levels in H5N1-infected MDCK cells. MDCK cells were infected with H5N1 (MOI, 0.01) and treated with DMSO or lycorine (0.52 μM). The level of COX41 in cells was determined at 24 h after the viral infection. Data are mean ± s.d. (n = 2 ~ 3). ^*p < 0.05, ^**p < 0.01, ^***p < 0.001 (D) Western blot analysis and quantitative analysis the expression of COX41 in H5N1-infected cells in the presence of DMSO or different concentrations of lycorine. Data are mean ± s.d. (n = 2 ~ 3). ^***p < 0.001. (E) Lycorine suppressed the expression of COX41. Plasmids expressing FLAG-tagged COX41 were transfected into HEK 293 T cells and incubated for 4 h, followed by treatment of lycorine (0.52 μM). The cell lysates were harvested at 6, 12, 24, 36, and 48 h after transfection and analyzed by western blot assay. Data are mean ± s.d. (n = 2 ~ 3). ^***p < 0.001.

Figure 4

Lycorine inhibited the expression of COX41 through the ROS-ERK1/2 axis. (A) Lycorine suppressed the production of H5N1-induced ROS in wild-type cells or COX41-KO-1 cells. Catalase (CAT, 1 mg/ml). Data are mean ± s.d. (n = 3). (B) Lycorine did not affect PI3K/Akt signaling in MDCK cells that were treated with lycorine (0.52 μM) at the indicated time points after the viral infection. (C) Western blot and quantitative analysis of the inhibitory effect of lycorine on phosphorylation levels of ERK. MDCK cells were infected with H5N1 in the absence or presence of lycorine (0.52 μM). The protein levels were detected at indicated time points postinfection. Data are mean ± s.d. (n = 3). ^*p < 0.05; ^**p < 0.01; ^***p < 0.01. (D) ROS increased phosphorylation levels of ERK. H₂O₂ (1 mm, a ROS activator) and CAT (1 mg ml⁻¹, a ROS inhibitor) were used to test the phosphorylation of ERK. Data are mean ± s.d. (n = 3). ^*p < 0.05; ^**p < 0.01; ^***p < 0.01.

Figure 5

Anti-influenza effects of lycorine on the survival rate and lung index, pathological damage, and viral titers in lungs of mice. (A) In vivo schedule. The BALB/c mice (n = 11) underwent i.p. injection with 5 mg kg⁻¹ day⁻¹ of lycorine at 24 h before infection with 2LD₅₀ virus and administered for 16 consecutive days. The mice from each group were randomly selected and euthanized on day 5 or 14. (B) Lycorine treatment increased the survival rate of the H5N1-infected mice. (C) Lycorine attenuated the lung indices, which were stimulated by H5N1 infection. Lung index = lung weight (g)/body weight (g) × 100. (D) Representative H&E staining images of lung and trachea tissues are shown as 400× magnification. (E) Viral titers in lung tissues of mice as determined by TCID₅₀ assay. Data are mean ± s.d. (n = 5). ^*p < 0.05; ^**p < 0.01; ^***p < 0.001.

Figure 6

Protective effects of lycorine on H5N1 infection in vivo and in vitro. COX41 activates the ERK signaling and promotes the nuclear export of vRNP. Lycorine inhibits the nuclear export of vRNP and viral replication through suppressing the expression of COX41.