Human enterovirus 71 protein interaction network prompts antiviral drug repositioning

Abstract

As a predominant cause of human hand, foot, and mouth disease, enterovirus 71 (EV71) infection may lead to serious diseases and result in severe consequences that threaten public health and cause widespread panic. Although the systematic identification of physical interactions between viral proteins and host proteins provides initial information for the recognition of the cellular mechanism involved in viral infection and the development of new therapies, EV71-host protein interactions have not been explored. Here, we identified interactions between EV71 proteins and host cellular proteins and confirmed the functional relationships of EV71-interacting proteins (EIPs) with virus proliferation and infection by integrating a human protein interaction network and by functional annotation. We found that most EIPs had known interactions with other viruses. We also predicted ATP6V0C as a broad-spectrum essential host factor and validated its essentiality for EV71 infection in vitro. EIPs and their interacting proteins were more likely to be targets of anti-inflammatory and neurological drugs, indicating their potential to serve as host-oriented antiviral targets. Thus, we used a connectivity map to find drugs that inhibited EIP expression. We predicted tanespimycin as a candidate and demonstrated its antiviral efficiency in vitro. These findings provide the first systematic identification of EV71-host protein interactions, an analysis of EIP protein characteristics and a demonstration of their value in developing host-oriented antiviral therapies.

Figure 1. The flow chart of this study.

The EV71-Human Protein Interaction Network was constructed by integrate data from public database and Y2H Screening, and used for functional annotation and drug prediction. The white blocks represent data resources.

Figure 2. Overview of validated EV71-human protein-protein interactions.

(a) Validated interactions between EV71 proteins and human proteins; (b) Number of human proteins that interacted with each EV71 protein; (c) A representative result of a yeast re-transformation assay; (d) Venn Diagram for 29 EIPs to show the most common annotations of these proteins.

Figure 3. The EAP interaction network and the topological analysis of EIPs in the human protein interactome.

(a) The protein interaction subnetworks between EAPs, with orange nodes representing EIPs and blue nodes representing the human proteins that interact with them. The node sizes are directly proportional to the node’s degrees in this subnetwork; (b,c) The degree (b) and betweenness centrality (c) distributions of all proteins (red points) and EIPs (blue points) in the human protein-protein interaction network. P(k) represents the probability of a node having a degree of k, P(b) represents the probability of a node having a betweenness centrality of (b). The dashed lines represents the average degrees of all proteins (red) and EIPs (blue); (d) The average degree, betweenness centrality and shortest path for the EIPs and all proteins.

Figure 4. Overlap analysis of EIPs and other virus-interacting proteins.

(a) Blue edges represent physical interactions between the viral proteins and EIPs, and orange and green edges represent EIP VTP and EHF interactions to the corresponding viruses; (b) The statistical analysis of other virus-interacting EIP numbers with the corresponding interaction numbers marked in parentheses; (c) Co-localization of ATP6V0C and 3A in the co-transfected RD cells; (d) The interaction between ATP6V0C and 3A was confirmed by co-immunoprecipitation; (e) Inhibition of EV71 replication by bafilomycin A1. RD cells were infected with the EV71 virus in the presence of varying concentrations of bafilomycin A1 (0, 6.25, 12.5, 25, 50 and 100 nmol/L) for 1 h. After washing, the cells were cultured in fresh growth mediumfor an additional 12 h. The cell cultures were subjected to plaque assays. Compared with virus control, *P < 0.05, **P < 0.01, ±s, n = 3; (f) Cytotoxicity of bafilomycin A1. Percent viability of RD cells was determined normalized to the absorbance at 450 nm of a drug-free culture; (g) Inhibition of EV71 propagation by ATP6V0C siRNA; (h) Increased EV71 propagation by ATP6V0C. RD cells were transfected with ATP6V0C siRNA or pCMV-myc-ATP6V0C and subsequently infected with EV71. 24 h later, the virus RNAs in cell cultures were isolated and subjected to real time RT-PCR. Virus growth is shown as the average percentage relative to the control cells infected with EV71. *P < 0.05, ±s, n = 3.

Figure 5. The anti-EV71 activity of drugs identified as significantly inhibiting EIPs.

(a) Relative viability of EV71-infected cells in response to eleven drugs. RD cells were seeded into 96-well plates and infected with EV71 at an MOI of 1 for 1 hr. Then, the cells were treated with small-molecule drugs at final concentrations of 10 μM and 20 μM. Cell viability was detected using the CCK8 assay at 48 hr post-infection, and signals from the mock-treated reaction were set as 100%; (b) The anti-EV71 activity of tanespimycin in the CPE assay. RD cells were seeded into 96-well plates and infected with EV71 at an MOI of 1 for 1 hr. Then, the cells were treated with various doses of tanespimycin. Tanespimycin-mediated cell viability was detected using the CCK8 assay at 48 hr post-infection; (c) The anti-EV71 activity of tanespimycin based on the detection of viral RNA using real-time RT-PCR. RD cells were seeded into 96-well plates and infected with EV71 at an MOI of 1 for 1 hr. Then, the cells were treated with various doses of tanespimycin. After incubation for 24 h and 48 h, the cell culture supernatants were harvested. Viral RNA was isolated and subjected to real-time RT-PCR; (d) Tanespimycin cytotoxicity. RD cells were treated with various tanespimycin concentrations for 48 h. Cell viability was tested using a Cell Counting Kit-8 assay. All tests were performed independently three times. The results are expressed as the mean ± SD of three samples.