A critical role of a cellular membrane traffic protein in poliovirus RNA replication

Abstract

Replication of many RNA viruses is accompanied by extensive remodeling of intracellular membranes. In poliovirus-infected cells, ER and Golgi stacks disappear, while new clusters of vesicle-like structures form sites for viral RNA synthesis. Virus replication is inhibited by brefeldin A (BFA), implicating some components(s) of the cellular secretory pathway in virus growth. Formation of characteristic vesicles induced by expression of viral proteins was not inhibited by BFA, but they were functionally deficient. GBF1, a guanine nucleotide exchange factor for the small cellular GTPases, Arf, is responsible for the sensitivity of virus infection to BFA, and is required for virus replication. Knockdown of GBF1 expression inhibited virus replication, which was rescued by catalytically active protein with an intact N-terminal sequence. We identified a mutation in GBF1 that allows growth of poliovirus in the presence of BFA. Interaction between GBF1 and viral protein 3A determined the outcome of infection in the presence of BFA.

Figure 1. Inhibition of poliovirus growth by BFA depends on the host cell.

A. Schematic map of the poliovirus genome. B. Poliovirus was propagated on Vero or BER-40 cells for 6 h in the presence of 2 µg/ml BFA. Virus yields were measured by plaque assay on HeLa cell monolayers. Dilutions of the infected cell suspensions are indicated.

Figure 2. GBF1 sequence determines BFA resistance of cells and virus.

A. Sequencing chromatograms of GBF1 Sec7 domain from Vero and BER-40 cells. Results from analyses with three different primers are shown. B. GBF1 A795E mutant confers BFA resistance to HeLa cells. Cells were transfected with either a control vector, vector expressing wild type YFP-GBF1 fusion or vector expressing YFP-GBF1 A795E mutant fusion, and subsequent cell growth in the presence of BFA was measured by a luminescent cell viability assay. C. Expression of GBF1 A795E mutant rescues protein secretion in the presence of BFA. Cells were co-transfected with pCMV-Gluc vector expressing secreted Gaussia luciferase and with either control plasmid or with vectors expressing wild type GBF1-YFP or GBF1 A795E-YFP fusions. The amount of secreted protein observed in each sample without BFA was defined as 100%. D. GBF1 A795E mutant rescues replication of poliovirus in the presence of BFA. A polio Renilla luciferase replicon was introduced in HeLa cells previously transfected with vectors expressing wt YFP-GBF1 fusion, YFP-GBF1 A795E mutant or with an empty vector. BFA was added where indicated at 1 µg/ml concentration at the time of replicon transfection, and polio RNA replication was measured by luciferase assay. Expression of the GBF1 proteins was measured by Western blot; calnexin staining was used as a loading control (panel D, right).

Figure 3. Requirement of active GBF1 for poliovirus replication in the presence of BFA.

A. HeLa cells were transfected with a plasmid expressing active YFP-GBF1 fusion, inactive YFP-GBF1 E794K fusion or control vector one day prior to transfection with a polio replicon containing the Renilla luciferase gene, and incubated in the presence of 1 µg/ml of BFA where indicated. Luciferase activity was assayed as a measure of polio RNA replication. The Western blot on the right shows expression of GBF1 derivatives. Actin staining served as a loading control. B. HeLa cells were treated with either GBF1 or control non-specific siRNA for 3 days prior to transfection with the polio Renilla replicon. The Western blot shows the level of knock down of GBF1 protein. The same membrane was stripped and probed again with anti-BIG2 antibodies as a loading control. C. HeLa cells were transfected with either a control vector or plasmid pArf1 Q71L-CFP, expressing a constitutively activated form of Arf1-CFP fusion. The next day the cells were transfected with a polio replicon containing the Renilla luciferase gene, and incubated in the presence of 2 µg/ml of BFA where indicated. Luciferase activity was assayed as a measure of polio RNA replication. The Western blot on the right shows expression of Arf1 Q71L-CFP fusion. Actin staining serves as a loading control.

Figure 4. Synthesis of poliovirus protein 3A stimulates GBF1-dependent Arf activation in vitro.

RNA coding for poliovirus 3A was translated in HeLa cell S10 extracts. The samples contained 80 µg/ml BFA where indicated (or the corresponding amount of solvent DMSO in other samples). The membranes were collected by centrifugation and assessed by western blot with anti-GBF1, anti-Arf and anti-COPI antibodies. Translation efficiency was monitored by labeling an aliquot of the translation reaction with ³⁵S methionine (lower panel).

Figure 5. Mutation in poliovirus 3A protein increases viral sensitivity to BFA and decreases recruitment of GBF1 to membranes.

A. HeLa cells were transfected with either wild type or 3A-2 mutant polio Renilla replicons and incubated in the presence of the indicated amounts of BFA. Luciferase was assayed as a measure of polio RNA replication. The inset shows a direct comparison of the wild type and 3A-2 replicons in the absence of BFA. B. HeLa cells were transfected with a plasmid expressing YFP-GBF1 A795E, a BFA-resistant mutant of GBF1, or a control vector. The next day the cells were transfected with either wild type or 3A-2 mutant polio Renilla replicon RNA and incubated with or without 1 µg/ml of BFA while RNA replication was monitored. The inset shows a direct comparison of the wild type and 3A-2 replicons in the cells transfected with an empty vector in the absence of BFA. C. HeLa cells were treated with either GBF1 or control non-specific siRNA for 2 days prior to transfection with the polio Renilla wild type or 3A-2 mutant replicons. Luciferase was assayed as a measure of polio RNA replication. D. RNA coding for either wild type 3A or 3A-2 mutant was translated in HeLa cell S10 extracts. The membranes were collected by centrifugation and assessed by western blot with anti-GBF1 antibodies. The same membrane was stripped and probed again with anti-calnexin antibodies as a loading control. Translation efficiency was monitored by labeling an aliquot of the translation reaction with ³⁵S methionine (lower panel).

Figure 6. The N-terminal region of GBF1 is important for rescue of polio replication from BFA inhibition and for localization to sites of viral RNA synthesis.

A. HeLa cells were transfected with vectors expressing YFP-GBF1 fusions with BFA-resistance mutation A795E with either the wild type N-terminus or with a deletion of 37 N-terminal amino-acids (Δ37). The next day the cells were transfected with polio Renilla replicon RNA and incubated in the presence or absence of 1 µg/ml BFA while monitoring viral RNA synthesis. B. HeLa cells were transfected with plasmids expressing Venus-GBF1 fusions with either wild type N-terminus (panels A–D) or with a deletion of 37 N-terminal amino-acids (Δ37) (panels E–H). The next day cells were infected with poliovirus at a multiplicity of 10 PFU/cells (panels A–C and E–G) or mock-infected (panels D and H), incubated for 4 h, fixed and processed for immunofluorescent staining with anti-polio 3A antibodies (red). Nuclear chromatin is stained with Hoechst 33342.

Figure 7. Polio-induced vesicles form in the presence of BFA, but fail to induce Arf1 relocalization.

A. Polio non-structural proteins were expressed from non-replicating RNA. The cells were incubated in the presence or absence of 2 µg/ml of BFA. An equivalent amount of DMSO solvent was added to the incubation medium in the sample without BFA. After 4.5 h incubation, cells were fixed and processed for electron microscopy. B. Poliovirus non-structural proteins were expressed from non-replicating RNA in HeLa cells expressing Arf1-EGFP fusion. The cells were incubated in the presence of 2 µg/ml of BFA where indicated (panels E–H). An equivalent amount of DMSO solvent was added to the incubation medium in samples without BFA (panels A–D). To observe the localization of Arf1-EGFP fusion protein in cells without expression of polio proteins, an empty plasmid was substituted for the polio cDNA-bearing plasmid (panels D and H).

Figure 8. Vesicles formed in the presence of BFA do not support viral RNA replication.

A. Polio Renilla replicon RNA was synthesized in cells in the presence of 2 mM guanidine-HCl and 2 µg/ml BFA as indicated. These conditions allowed for RNA translation and formation of vesicles, but prevented RNA synthesis. After 4.5 h the medium was changed to remove guanidine-HCl and allow RNA synthesis, and Renilla luciferase activity was monitored as a measure of viral RNA replication. BFA and guanidine were present at both steps in the control sample. B. The same experiment was performed with a replicon containing a deletion in the polymerase gene.