Host modulators of H1N1 cytopathogenicity

Abstract

Influenza A virus infects 5-20% of the population annually, resulting in ~35,000 deaths and significant morbidity. Current treatments include vaccines and drugs that target viral proteins. However, both of these approaches have limitations, as vaccines require yearly development and the rapid evolution of viral proteins gives rise to drug resistance. In consequence additional intervention strategies, that target host factors required for the viral life cycle, are under investigation. Here we employed arrayed whole-genome siRNA screening strategies to identify cell-autonomous molecular components that are subverted to support H1N1 influenza A virus infection of human bronchial epithelial cells. Integration across relevant public data sets exposed druggable gene products required for epithelial cell infection or required for viral proteins to deflect host cell suicide checkpoint activation. Pharmacological inhibition of representative targets, RGGT and CHEK1, resulted in significant protection against infection of human epithelial cells by the A/WS/33 virus. In addition, chemical inhibition of RGGT partially protected against H5N1 and the 2009 H1N1 pandemic strain. The observations reported here thus contribute to an expanding body of studies directed at decoding vulnerabilities in the command and control networks specified by influenza virulence factors.

Figure 1. Identification of Host Modulators of Influenza Infection.

(A) HBEC30 were infected with A/WSN/33/H1/N1 (WSN) at an MOI of 5 and examined for accumulation of viral proteins by immunoblot at the indicated time-points post-infection. (B) Cells treated as in B were immunostained for detection of viral protein accumulation at single cell resolution. Top panels labeled WSN show anti-influenza A staining and bottom panels labeled Hoescht show nuclear staining with Hoescht. (C) Parallel cultures were also examined for consequences on cell viability over a 72-hour time-course. (D) The rank-ordered Z-score distribution from each of 21,125 siRNA pools targeting the annotated human genome is shown. Dashed lines indicate 3 standard deviations above (red) and below (green) the mean of the distribution.

Figure 2. Characterization of Viral Replication Response.

(A) A panel of 33 siRNAs was assayed for viral protein accumulation and infectious particle production. HBEC30s were transfected with siRNA and infected with WSN at an MOI of 5. For primary infection, cells were fixed at indicated time points and viral protein was detected by immunostaining of viral proteins. Supernatants from infected cells were collected at 24 hours post infection and used for secondary infection of MDCK cells with viral protein detection by immunostaining (right column, 2⁰). Resistors are shown in the top panel and sensitizers in the bottom. (B) HBEC30 were transfected with siRNA targeting IFITM3 or control siRNA and infected with WSN at an MOI of 0.1. Viral protein was detected at 12 hours post infection by immunostaining with anti-influenza antibodies (WSN panels) (Meridian Life Science, Inc, Cat# B65141G). (C) Cells treated in B were counted and the percent of infected cells was quantified. (D) Cells treated as in B were incubated 48 hours post infection and cell viability was measured. (E) Supernatants from WSN infected HBEC30s were collected 24 hours post infection and used for secondary infection of MDCK cells with viral protein detection by immunostaining. (F) HBEC30s were transfected with indicated siRNAs and infected with WSN lacking the viral protein NS1, cell viability was measured 48 hours post infection. (G) Cells treated as in F were fixed at 24 hours post infection and immunostained for viral protein for calculation of percentage of infected cells. (H) Supernatants from cells in G were used for secondary infection in MDCK cells and viral protein was detected by immunostaining. (P values; * <0.05, ** <0.01, *** <0.0001).

Figure 3. Functional Classification of Candidate Hits.

(A) siRNA screen results from this study were compared with data from two published screens for cell cycle modulators and the overlap is shown. (B) Intersection of hits from this study with those examining host modulators of HIV infection (Table 4 in Supporting Information S1). (C) Cell viability data was queried against four published screens using viral replication as the end-point assay. Candidate hits were binned into functional classes based upon perturbation of viral cytopathogenicity together with viral replication. (D) Two pharmacologically addressable Netwalk subnetworks are shown.

Figure 4. Viral Inhibition by SB218078.

(A) HBEC30s were treated with SB218078 at indicated concentrations and cell viability was measured after 48 hours. (B) HBEC30s were treated as in A and infected with WSN at an MOI of 5 followed by immunostaining at indicated time points. Top panels labeled WSN show anti-influenza A staining and bottom panels labeled Hoescht show nuclear staining with Hoescht. (C) Fluorescence intensity was measured and quantified from B. (D) Percentage of infected cells from B. (E) A549 cells were pretreated with 218078 and infected with WSN at an MOI of 5. Viral protein was detected by immunostaining. Top panels show anti-influenza A staining (WSN) and bottom panels show nuclear staining (Hoescht). (F) Quantification of percent of infected cells in E. (G) HBEC3-KT cells were pretreated with SB218078, infected with WSN at an MOI of 5 and immunostained for detection of viral protein. The percentage of infected cells was quantified. (P values; * <0.05, ** <0.01, *** <0.0001).

Figure 5. Viral Inhibition by 3-IPEHPC.

(A) HBEC30s were pretreated with 3-IPEHPC or buffer and infected with WSN at an MOI of 5. Viral protein was detected by immunostaining. Top panels show anti-influenza A staining (WSN) and bottom panels show nuclear staining (Hoescht). (B) Overall fluorescence intensity of cells in A was quantified. (C) Quantification of percent of infected cells in A. (D and E) A549 cells were pretreated with 3-IPEHPC and infected with either avian H5/N1 or the recent H1/N1 pandemic strain. Lysates from infected cells were collected 24 hours post infection and viral protein was detected by immunoblot. (P values; * <0.05, ** <0.01, *** <0.0001).

Figure 6. miRNA Screen.

(A) HBEC30s were transfected with miRNA mimics and screened using conditions identical to the siRNA screen. Z-Scores were calculated for individual oligos and plotted according to rank order. Dashed lines indicate 2 standard deviations above (red) and below (green) the mean of the distribution. (B) HBEC30s were transfected with selected miRNA mimics, infected with WSN and cell viability phenotype was measured 48 hours post infection. (C) Cells treated as in B were fixed and immunostained for viral protein 12 hours post infection. (D) A549 cells were transfected with miRNA mimics and infected with pandemic H1/N1. Cell lysates were collected 24 hours post infection and viral proteins were detected by immunoblotting. (E) Network analysis of miRNA predicted targets. Node behavior in siRNA screens is indicated. Edges indicate physical or functional interactions among nodes. (P values; * <0.05, ** <0.01, *** <0.0001).