The cellular chaperone heat shock protein 90 facilitates Flock House virus RNA replication in Drosophila cells

Abstract

The assembly of viral RNA replication complexes on intracellular membranes represents a critical step in the life cycle of positive-strand RNA viruses. We investigated the role of the cellular chaperone heat shock protein 90 (Hsp90) in viral RNA replication complex assembly and function using Flock House virus (FHV), an alphanodavirus whose RNA-dependent RNA polymerase, protein A, is essential for viral RNA replication complex assembly on mitochondrial outer membranes. The Hsp90 chaperone complex transports cellular mitochondrial proteins to the outer mitochondrial membrane import receptors, and thus we hypothesized that Hsp90 may also facilitate FHV RNA replication complex assembly or function. Treatment of FHV-infected Drosophila S2 cells with the Hsp90-specific inhibitor geldanamycin or radicicol potently suppressed the production of infectious virions and the accumulation of protein A and genomic, subgenomic, and template viral RNA. In contrast, geldanamycin did not inhibit the activity of preformed FHV RNA replication complexes. Hsp90 inhibitors also suppressed viral RNA and protein A accumulation in S2 cells expressing an FHV RNA replicon. Furthermore, Hsp90 inhibition with either geldanamycin or RNAi-mediated chaperone downregulation suppressed protein A accumulation in the absence of viral RNA replication. These results identify Hsp90 as a host factor involved in FHV RNA replication and suggest that FHV uses established cellular chaperone pathways to assemble its RNA replication complexes on intracellular membranes.

FIG. 1.

Effect of Hsp90 and fatty acid synthetase inhibitors on S2 cell growth, viability, and infectious virion production. (A) Uninfected S2 cells were incubated with vehicle only (circles), 50 μM cerulenin (triangles), 5 μM geldanamycin (squares), or 5 μM radicicol (diamonds), and cell viability (closed symbols, left y axis) and cell number (open symbols, right y axis) were determined by trypan blue exclusion with a hemocytometer. Individual data points represent the mean values from three experiments. (B) S2 cell were infected with FHV at an MOI = 10 and simultaneously treated with vehicle only (None), 50 μM cerulenin (Cer), 5 μM geldanamycin (GA), or 5 μM radicicol (Rad). Infectious virions were harvested at 12 h postinfection, purified by pelleting through sucrose gradients to remove residual inhibitors, and quantitated by an immunofluorescence-based assay. Results represent the means ± standard errors of the means (SEM) from three experiments and are expressed as severalfold increases over the calculated virus input at an MOI = 10 to account for inoculum virion recovery. FHV virion production in infected S2 cells is extremely vigorous, with an estimated burst size of approximately 250,000 particles per cell (47).

FIG. 2.

Hsp90 inhibition suppresses FHV RNA replication in infected cells. (A) S2 cells were infected with FHV and treated with 50 μM cerulenin (C), 5 μM geldanamycin (G), or 5 μM radicicol (R) at 0 h postinfection (lanes 3 to 5) or 4 h postinfection (lanes 6 and 7) and harvested at 12 h postinfection. We used cerulenin as a positive control, as this fatty acid synthetase inhibitor has been previously shown to suppress positive-strand RNA virus replication (16, 36). Total protein or RNA from an equal number of cells was separated by electrophoresis, and protein A (upper blot) was detected by immunoblotting, while (+)RNA1 and (+)RNA3 (third blot) or (−)RNA1 (fourth blot) was detected by Northern blotting. The Hsp60 immunoblot and rRNA band detected by ethidium bromide staining are shown as loading controls. (B) Quantitative data for FHV protein A, genomic (+)RNA1, template (−)RNA1, and subgenomic (+)RNA3 accumulation are relative to the control (Fig. 2A, lane 2) and represent the means ± SEM from three experiments. (C) Titration of geldanamycin (GA) effect on FHV protein A and RNA accumulation. Cells were infected and treated with increasing GA concentrations at the time of infection, and viral protein A and RNA accumulation were analyzed at 12 h postinfection as described above. Individual data points represent the mean values from three experiments.

FIG. 3.

Hsp90 inhibition does not suppress FHV RdRp activity in vitro or in cells. (A) Membrane fractions from control S2 cells (lane 1) or FHV-infected S2 cells that contain a crude replication complex (CRC) (lanes 2 to 6) were incubated with ³H-labeled UTP and unlabeled ribonucleotides and vehicle only (lane 2), 5 mM EDTA (lane 3), or decreasing concentrations of geldanamycin (GA; lanes 4 to 6), and the reaction products were separated by nondenaturing agarose gel electrophoresis and detected by fluorography. The positions of RNA molecular weight (MW) markers transcribed in vitro are shown on the left. The major radiolabeled bands corresponding to probable FHV dsRNA1 (∼6,200 nucleotides) and dsRNA2 (∼2,800 nucleotides), the predominant FHV RdRp in vitro products (56), were quantitated by densitometry, and the numbers represent the mean values from two experiments relative to untreated controls (lane 2). (B) Total RNA from control (lane 1) or FHV-infected (lanes 3 to 5) S2 cells incubated with ³H-labeled uridine for 2 h in the presence of actinomycin D and vehicle only (lane 3), 5 mM EDTA (lane 4), or 5 μM geldanamycin (GA; lane 5) was separated by denaturing agarose gel electrophoresis and analyzed by fluorography. The positions of all three FHV RNA species are shown on the right. ³H-labeled RNA from S2 cells expressing an FHV RNA1 replicon (lane 2) (Fig. 4A) is shown as reference for the identification of RNA1 and RNA3 bands from infected cells, and the rRNA band detected by ethidium bromide staining is shown as a loading control. There is a faint background radiolabeled band visible in both mock-infected cells (lane 1) and cells expressing an FHV RNA1 replicon (lane 2) that comigrates with RNA2. Results are from one representative experiment of two.

FIG. 4.

Hsp90 inhibition suppresses FHV RNA replication in the absence of infectious virions. (A) Schematic of pS2F1, an MT promoter-driven plasmid that expresses an FHV RNA1 replicon. Authentic viral 5′ and 3′ RNA termini were generated in cells through precise placement of the MT promoter start site and a hepatitis δ ribozyme (Rz), respectively. Drosophila RNA polymerase II-transcribed FHV RNA1 can serve as a template for both replication and translation into protein A, and thus viral RNA replication and subgenomic (sg) RNA3 synthesis become autonomous and MT promoter independent after Cu²⁺ induction. (B) S2 cells transiently transfected with empty vector (lane 1) or pS2F1 (lanes 2 to 5) were induced with 0.5 mM Cu²⁺ in the presence of 50 μM cerulenin (C; lane 3), 5 μM geldanamycin (G; lane 4), or 5 μM radicicol (R; lane 5) 24 h after transfection. To control for transfection efficiency, we transfected cells with pS2F1 as a single culture and subsequently aliquoted cells into 12-well tissue culture plates for induction and treatment with individual inhibitors. Total protein and RNA were harvested from an equal number of cells at 12 h postinduction and analyzed by immunoblotting and Northern blotting. The numbers represent the levels of protein A and (+)RNA3 accumulation relative to control (lane 2) and are the mean values from three experiments. The potent suppression of protein A accumulation by the fatty acid synthetase inhibitor cerulenin in pS2F1-transfected cells (lane 3) was in contrast to its moderate suppression of protein A accumulation in FHV-infected cells (Fig. 2A, lane 3) and may reflect a differential effect of cerulenin on the translation of FHV RNA1 derived from cytosolic virions versus plasmid-derived nuclear transcripts. (C) Metallothionein (MT) promoter- or glucocorticoid receptor responsive (GRE) promoter-driven luciferase (LUC) or β-galactosidase (LacZ) activity in transiently transfected S2 cells in the absence (black bars) or presence (gray bars) of 5 μM geldanamycin (GA). To control for transfection efficiency, cells were transfected with individual reporter plasmids as a single culture as described above. Results are the means ± SEM from three experiments and are expressed as percentages of control reporter gene activity, where activity in the absence of geldanamycin was set at 100%.

FIG. 5.

Hsp90 inhibition suppresses FHV protein A accumulation in the absence of viral RNA replication. (A) S2 cells stably transfected with pS2FA (upper two blots) or pS2LacZ (bottom blot) were induced with 0.5 mM Cu²⁺ (lanes 2 to 7) in the absence (lane 2) or presence (lanes 3 to 7) of decreasing geldanamycin concentrations (GA conc.). The relative height of the triangle represents geldanamycin concentrations, which are 10-fold dilutions between adjacent lanes. Total protein from an equivalent number of cells was harvested at 12 h postinduction and analyzed for protein A (Ptn A) or β-galactosidase (β-Gal) accumulation by anti-HA immunoblotting. The Hsp60 immunoblot is shown as a loading control. A background band detected by the polyclonal anti-HA antibody is present that comigrates with β-galactosidase (120 kDa) but above protein A (112 kDa). (B) S2 cells transfected with pS2FA (lanes 2 to 5) were induced in the presence of 1 μM GA (lane 3) or GA plus the proteasome inhibitor lactacystin (Lact; lanes 4 and 5), harvested at 12 h postinduction, and immunoblotted for protein A (upper blot) or Hsp60 (lower blot). The numbers represent the mean protein A levels from two or three experiments relative to control (lane 2).

FIG. 6.

RNAi-mediated suppression of Hsp83 in Drosophila cells reduces protein A accumulation. (A) S2 cells stably transfected with pS2LacZ were treated with 10 μg lacZ (lane 3) or Hsp83 (lane 4) dsRNA. After 48 h, cells were induced with 0.5 mM Cu²⁺ in the presence (lane 2) or absence (lanes 1, 3, and 4) of 1 μM geldanamycin (GA) and harvested at 12 h postinduction, and lysates from an equal number of cells were analyzed for β-galactosidase (β-Gal), Hsp83, and Hsp60 accumulation by immunoblotting. (B) S2 cells stably transfected with pS2FA were treated with 10 μg lacZ dsRNA (lane 3) or decreasing amounts of Hsp83 dsRNA (lanes 4 to 7), induced, harvested, and analyzed as described above. The relative height of the triangle represents the amount of Hsp83 dsRNA added to cultures, which are twofold dilutions between adjacent lanes. (C) Quantitative data for Hsp83 (black bars) and protein A (gray bars) accumulation in pS2FA-transfected cells. Results are the mean values from two or three experiments relative to untreated controls (Fig. 6B, lane 1).