Dual roles for an arginine-rich motif in specific genome recognition and localization of viral coat protein to RNA replication sites in flock house virus-infected cells

Abstract

Assembly of many RNA viruses entails the encapsidation of multiple genome segments into a single virion, and underlying mechanisms for this process are still poorly understood. In the case of the nodavirus Flock House virus (FHV), a bipartite positive-strand RNA genome consisting of RNA1 and RNA2 is copackaged into progeny virions. In this study, we investigated whether the specific packaging of FHV RNA is dependent on an arginine-rich motif (ARM) located in the N terminus of the coat protein. Our results demonstrate that the replacement of all arginine residues within this motif with alanines rendered the resultant coat protein unable to package RNA1, suggesting that the ARM represents an important determinant for the encapsidation of this genome segment. In contrast, replacement of all arginines with lysines had no effect on RNA1 packaging. Interestingly, confocal microscopic analysis demonstrated that the RNA1 packaging-deficient mutant did not localize to mitochondrial sites of FHV RNA replication as efficiently as wild-type coat protein. In addition, gain-of-function analyses showed that the ARM by itself was sufficient to target green fluorescent protein to RNA replication sites. These data suggest that the packaging of RNA1 is dependent on trafficking of coat protein to mitochondria, the presumed site of FHV assembly, and that this trafficking requires a high density of positive charge in the N terminus. Our results are compatible with a model in which recognition of RNA1 and RNA2 for encapsidation occurs sequentially and in distinct cellular microenvironments.

FIG. 1.

Schematic representation of coat protein mutants containing modifications in the ARM spanning residues 32 to 50. The two regions shaded in gray have been shown previously to be important for packaging of either RNA2 (residues 1 to 31) or both FHV RNAs (residues 381 to 407) (29, 45). Three phenylalanine residues within the latter region have been identified as important contributors to FHV RNA recognition, presumably via base-stacking interactions with the viral RNA (45). Mutations in the FHV ARM included substitutions of alanine or lysine for arginine (mutants 1 to 7) and replacement with the ARM of the HIV-1 Rev protein (mutant 8).

FIG. 2.

Effect of mutations in ARM on RNA replication, coat protein synthesis, and virus particle yield. (A) FHV RNA and proteins were generated in Drosophila DL-1 cells following the transfection/infection protocol described. At 48 hpi, total cellular RNA was extracted, electrophoresed through an agarose gel, and stained with ethidium bromide (top panel). Virions were partially purified from the cells and run on a sodium dodecyl sulfate-polyacrylamide gel, followed by Coomassie staining (middle panel). Samples representing mutant 5 and mutant 6 were also subjected to immunoblot analysis with anti-FHV antiserum (bottom panel). DI-RNA2 designates an RNA2-derived DI RNA while α and β designate coat precursor protein α and cleavage product β, respectively. (B) For mutants 5 and 6, analyses of FHV RNA and coat protein synthesis were also carried out at 12 h posttransfection of Drosophila DL-1 cells. An ethidium bromide-stained agarose gel of total RNA (left panel) and immunoblot analysis of cell lysates (right panel) are shown. The α-tubulin immunoblot serves as a loading control.

FIG. 3.

Effect of mutations in ARM on RNA packaging. RNA was extracted from gradient-purified virions assembled from the indicated mutant coat proteins and subjected to electrophoresis in nondenaturing agarose gels, followed by ethidium bromide staining (A to C and F). For mutant 6, RNA extracted from particles sedimenting in three neighboring gradient fractions (fr) is shown in panel C. (D) Electron micrograph of negatively stained purified mutant 6 particles. (E) Northern blot analysis of RNA extracted from mutant 6 particles using negative-sense RNA1 and RNA2 probes. Equivalent amounts of this RNA were loaded into each lane. DI-RNA1 and DI-RNA2 designate RNA1- and RNA2-derived DI RNAs, respectively, and arrows point to smaller RNA1- and RNA2-derived RNA species detected within the packaged RNA. RNA extracted from authentic FHV particles (wt) was included as a control.

FIG. 4.

Subcellular distribution of coat protein and RdRp during the course of infection. Drosophila S2 cells were infected with FHV at an MOI of 5 and processed for immunofluorescence staining at 6, 9, and 15 hpi as described in Materials and Methods. Polyclonal rabbit anti-FHV and monoclonal mouse anti-FHV RdRp antibodies were used to stain coat protein and RdRp, respectively, and DAPI was used as a nuclear stain. Z-series projections of several optical sections are shown. A more powerful laser setting was required to detect coat protein at 6 hpi (A) than at 9 and 15 hpi (E and I) because this protein was present at low intracellular levels during the initial stages of infection. Yellow pixels represent subcellular areas where coat protein and RdRp colocalize (D, H, and L), and arrows point to coat protein aggregates visualized at late stages of infection (I).

FIG. 5.

Comparison between distribution patterns of coat protein and ER, actin, and microtubule networks. Drosophila S2 cells were infected with FHV at an MOI of 5, and at 9 hpi the cells were fixed, permeabilized, stained with polyclonal anti-FHV for the detection of coat protein, and counterstained with either monoclonal anti-KDEL (ER marker), Alexa Fluor 555-phalloidin (actin marker), or monoclonal anti-α-tubulin (microtubule marker). Z-series projections of several optical sections are shown. Yellow pixels in merged images are indicative of colocalization between green and red fluorescence signals.

FIG. 6.

Redistribution of mitochondria and ER in FHV-infected Drosophila S2 cells. (A) Confocal immunofluorescence image of perinuclear area of infected cell labeled for detection of mitochondria (MitoTracker Red) and RdRp (monoclonal anti-FHV RdRp). The green-magenta color combination utilized for this pseudo-colored image illustrates that fluorescence signals for RdRp form rings around those for MitoTracker (indicated by white arrows). Scale bar, 5 μm. (B) Comparison between patterns of mitochondrial distribution in uninfected and infected cells. Mitochondria were detected using MitoTracker Red, and cellular limits (white lines) were determined by differential interference contrast microscopy. (C) Comparison between ER distribution patterns in uninfected cells and infected cells at 9 and 15 hpi. The ER was detected using monoclonal anti-KDEL antibodies. A single optical section is shown in panel A while Z-series projections of several optical sections are shown in panels B and C.

FIG. 7.

Subcellular localization of coat protein mutants. Drosophila S2 cells were cotransfected with wt RNA1 and in vitro RNA transcripts encoding wt and mutated versions of coat protein, which included RNA packaging-deficient mutants Δ31, mutant 6, and Δγ381, as well as mutants 7 and 8. At 12 h posttransfection, the cells were stained with rabbit anti-FHV for the detection of coat protein and monoclonal mouse anti-FHV RdRp. (A) Representative confocal immunofluorescence images for wt, Δ31, mutant 6, and Δγ381. Z-series projections of several optical sections are shown. Subcellular regions of immunofluorescence colocalization are yellow. (B) Graphic representation of the percentage of red pixels (RdRp) colocalizing with green pixels (coat protein) for wt and mutant coat proteins. The error bars represent the standard error of the mean of at least five independently analyzed cells. *, P = 0.005 compared to the wt.

FIG. 8.

Subcellular localization eGFP fusion proteins containing tags derived from the N terminus of coat protein. (A) Schematic representation of eGFP fusion constructs. The eGFP reading frame (green) in the DI RNA used for the generation of these constructs is indicated. “Mito target” in mutant eGFP-M designates the mitochondrion-targeting signal of FHV RdRp. (B) Representative confocal fluorescence images for eGFP fusion studies. Drosophila S2 cells were cotransfected with wt RNA1- and RNA2-derived DI RNAs encoding eGFP or N-terminally tagged versions of this protein. At 12 h posttransfection, the cells were stained with monoclonal mouse anti-FHV RdRp for the detection of RdRp at mitochondria. The images represent single optical sections. Subcellular regions of fluorescence colocalization are yellow. (C) Graphic representation of the percentage of red pixels (RdRp) colocalizing with green pixels (eGFP) for eGFP proteins containing tags derived from the N terminus of the coat protein. The error bars represent the standard error of the mean of at least seven independently analyzed cells.