Identification of RNA helicase A as a new host factor in the replication cycle of foot-and-mouth disease virus

Abstract

Foot-and-mouth disease virus (FMDV), as with other RNA viruses, recruits various host cell factors to assist in the translation and replication of the virus genome. In this study, we investigated the role of RNA helicase A (RHA) in the life cycle of FMDV. Immunofluorescent microscopy (IFM) showed a change in the subcellular distribution of RHA from the nucleus to the cytoplasm in FMDV-infected cells as infection progressed. Unlike nuclear RHA, the RHA detected in the cytoplasm reacted with an antibody that recognizes only the nonmethylated form of RHA. In contrast to alterations in the subcellular distribution of nuclear factors observed during infection with the related cardioviruses, cytoplasmic accumulation of RHA did not require the activity of the FMDV leader protein. Using IFM, we have found cytoplasmic RHA in proximity to the viral 2C and 3A proteins, which promotes the assembly of the replication complexes, as well as cellular poly(A) binding protein (PABP). Coimmunoprecipitation assays confirmed that these proteins are complexed with RHA. We have also identified a novel interaction between RHA and the S fragment in the FMDV 5' nontranslated region. Moreover, a reduction in the expression of RHA, using RHA-specific small interfering RNA constructs, inhibited FMDV replication. These results indicate that RHA plays an essential role in the replication of FMDV and potentially other picornaviruses through ribonucleoprotein complex formation at the 5' end of the genome and by interactions with 2C, 3A, and PABP.

FIG. 1.

FMDV infection alters RHA subcellular localization. (A) Uninfected and FMDV-infected (A₂₄ Cruzeiro) LFBK cells were probed with rabbit anti-RHA followed by goat anti-rabbit-AF488 (green) with nuclear material being stained with DAPI (blue). Samples included uninfected control cells and infected cells at 2, 4, and 6 hpi. (B) LFBK cells infected with BEV-1 (top) or FMDV A₂₄ Cruzeiro (bottom) at 1, 3, and 5 hpi were probed with rabbit anti-RHA, followed by goat anti-rabbit-AF488 (green); DAPI staining is not shown.

FIG. 2.

Nonmethylated RHA accumulates with infection. (A) Uninfected and FMDV-infected LFBK cells were probed with rabbit anti-RHA (designated RHA) followed by goat anti-rabbit-AF488 (green). Alternatively, the samples were probed with a mouse RHA antibody that recognizes only the demethylated form of RHA (designated DM-RHA) followed by goat anti-mouse-AF568 (red). Nuclear material was stained with DAPI (blue). (B) LFBK cells were or were not pretreated with the MDA cocktail and infected or not infected with FMDV in the continued presence or absence of MDA. Infected cells were examined at 4 hpi. Samples were then probed with mouse anti-DM-RHA followed by goat anti-mouse-AF568 (red) and with rabbit anti-3A followed by goat anti-rabbit-AF488 (green). Nuclear material was stained with DAPI (blue). (C) LFBK cells were treated with 0, 0.1, 0.5, and 1 mM MDA cocktail and then infected or not with FMDV in the continued presence or absence of MDA. Cellular lysates were prepared from harvested uninfected and infected cells and were examined by Western blot probing with antibodies to FMDV 3A, 3C^pro, 3D^pol, and tubulin. (D) The samples harvested in panel C were also examined by plaque assay (see Materials and Methods). The counted plaques were used to calculate the virus titer, and the values were plotted onto a bar graph using Microsoft Excel.

FIG. 3.

RHA redistribution is not triggered by FMDV leader proteinase (L^pro). LFBK cells were infected with either the WT A₂₄ Cruzeiro strain of FMDV or a “leaderless” derivative (LL) lacking the coding region for L^pro. Uninfected and infected cells were probed simultaneously with rabbit anti-L^pro (indicated panels) and mouse anti-DM-RHA (indicated panels) followed by goat anti-rabbit-AF488 (green) and goat anti-mouse-AF568 (red).

FIG. 4.

Cytoplasmic RHA overlaps with viral and cellular components of the FMDV replication complex. Uninfected and FMDV-infected LFBK cells at 4 hpi were simultaneously probed with mouse anti-DM-RHA and rabbit anti-2C (A), mouse anti-DM-RHA and rabbit anti-3A (B), rabbit anti-RHA and mouse anti-3D^pol (C), rabbit anti-RHA and mouse anti-PCBP2 (D), or rabbit anti-RHA and mouse anti-PABP (E) followed by goat anti-rabbit-AF488 (green) and goat anti-mouse-AF568 (red). Nuclear material was stained with DAPI (blue) (A to E).

FIG. 5.

RHA coprecipitates with FMDV 2C and 3A and cellular PABP. Uninfected LFBK cells or cells infected with FMDV at an MOI of 10 were harvested at 2, 3, and 4 hpi, and cells infected with FMDV at an MOI of 10⁻³ were harvested at 24 hpi. Lysates were immunoprecipitated with anti-PCBP2-, anti-PABP-, anti-2C-, anti-3A-, anti-3C^pro-, and anti-3D^pol-coupled agarose beads (Seize X protein G beads). Bound protein was eluted using a low-pH solution (pH 2.5), and the collected samples were analyzed by Western blot probing with anti-RHA. For each immunoprecipitation reaction, two consecutive eluates are represented in the figure for each time point examined.

FIG. 6.

RHA interacts with the S fragment of the FMDV 5′ NTR. (A) Eluates from the immunoprecipitation reactions described in the legend to Fig. 5 were mixed with ³²P-labeled positive-sense single-stranded RNA corresponding to the FMDV S fragment. Samples were tested for protein-RNA interaction using the filter binding assay (see Materials and Methods). (B) Purified RHA1 was mixed separately with ³²P-labeled S fragment, cre, and 3′ NTR and evaluated for protein-RNA interaction as described for panel A. (C) Purified 3C^pro and 3D^pol were separately mixed with ³²P-labeled S fragment and evaluated for protein-RNA interaction as described for panel A.

FIG. 7.

Knockdown of RHA expression reduces FMDV titer. LFBK cells were transfected and incubated with or without nonspecific siRNA constructs or constructs directed against RHA for 72 h and infected with FMDV. (A) The concentration of endogenous RHA and virus 3D^pol in each sample was evaluated by Western blot probing with anti-RHA and anti-3D^pol. The blot was also probed with antitubulin to confirm equal loading between lanes. (B) The Western blot from panel A was scanned, and the relative intensity of the detected bands was quantified using the ImageJ software. Quantities determined for each indicated protein were plotted side by side in a bar graph using Microsoft Excel. (C) Prior to virus infection, cells transfected or not with the various siRNA constructs were evaluated for cytotoxicity using the XTT assay. The absorbances obtained at 450 nm were plotted using Microsoft Excel. Two independent experiments are shown for each condition. (D) Samples were serially diluted and applied to confluent monolayers of BHK-21 cells for a virus titer assay. Calculated virus titers were plotted in logarithmic scale using Microsoft Excel. NC, negative control.