Host cell virus entry mediated by Australian bat lyssavirus G envelope glycoprotein occurs through a clathrin-mediated endocytic pathway that requires actin and Rab5

Abstract

Background: Australian bat lyssavirus (ABLV), a rhabdovirus of the genus Lyssavirus which circulates in both pteropid fruit bats and insectivorous bats in mainland Australia, has caused three fatal human infections, the most recent in February 2013, manifested as acute neurological disease indistinguishable from clinical rabies. Rhabdoviruses infect host cells through receptor-mediated endocytosis and subsequent pH-dependent fusion mediated by their single envelope glycoprotein (G), but the specific host factors and pathways involved in ABLV entry have not been determined.

Methods: ABLV internalization into HEK293T cells was examined using maxGFP-encoding recombinant vesicular stomatitis viruses (rVSV) that express ABLV G glycoproteins. A combination of chemical and molecular approaches was used to investigate the contribution of different endocytic pathways to ABLV entry. Dominant negative Rab GTPases were used to identify the endosomal compartment utilized by ABLV to gain entry into the host cell cytosol.

Results: Here we show that ABLV G-mediated entry into HEK293T cells was significantly inhibited by the dynamin-specific inhibitor dynasore, chlorpromazine, a drug that blocks clathrin-mediated endocytosis, and the actin depolymerizing drug latrunculin B. Over expression of dominant negative mutants of Eps15 and Rab5 also significantly reduced ABLV G-mediated entry into HEK293T cells. Chemical inhibitors of caveolae-dependent endocytosis and macropinocytosis and dominant negative mutants of Rab7 and Rab11 had no effect on ABLV entry.

Conclusions: The predominant pathway utilized by ABLV for internalization into HEK293T cells is clathrin-and actin-dependent. The requirement of Rab5 for productive infection indicates that ABLV G-mediated fusion occurs within the early endosome compartment.

Figure 1

Chemical inhibition of dynamin inhibits ABLV G-mediated viral entry into HEK293T cells. (A) HEK293T cell monolayers were pretreated with dynasore diluted in OptiMEM® for 30 min at 37°C and were then infected with max-GFP encoding rVSV reporter viruses (MOI = 1). Cells were harvested 20 hrs post infection, fixed with 2% paraformaldehyde and then analyzed for GFP expression (indicative of productive infection). GFP positive cells were counted with a Nexcelom Vision automated cell counter with fluorescence detection and the percent of virus-infected cells was calculated by dividing the number of GFP positive cells by the total number of cells counted. Under these experimental conditions, a MOI of 1 yielded 60-70% virus-infected cells in untreated cells. Drug was maintained for the entire course of infection and its effect on cell viability (B) was determined by trypan blue staining. Reporter viruses that express VSV G were included as a positive control to assess dynasore activity. Results are expressed as percent virus-infected cells relative to that of untreated controls and represent 3 independent experiments; error bars are standard error of the mean (SEM).

Figure 2

Chemical inhibition of CME inhibits ABLV G-mediated viral entry into HEK293T cells. (A) HEK293T monolayers were pretreated with chlorpromazine diluted in OptiMEM® for 30 min at 37°C. Cells were then infected with rVSV (MOI = 3) for 8 hrs and then analyzed as described in Figure 1. Under these conditions, a MOI of 3 yielded 50-60% virus-infected cells in untreated cells. Drug was maintained for the entire course of infection and its effect on cell viability (B) was determined by trypan blue staining. The rVSV-VSV G reporter virus was included as positive control. Results are expressed as percent virus-infected cells relative to that of controls and represent 3 independent experiments; error bars are SEM.

Figure 3

Inhibition of CME with a DN form of Eps15 significantly reduces ABLV G-mediated viral entry into HEK293T cells. HEK293T cells transfected with an eGFP control plasmid or eGFP-tagged Eps15 mutants were infected with VSVΔG-ABLV G* or –VSV G*-RFP pseudoviruses at a MOI = 2 for 20 hrs and then analyzed for RFP expression as described in Figure 1. Under these conditions, a MOI of 2 yielded 25-30% virus-infected cells among cells transfected with the eGFP control plasmid. DIIIΔ2, a mutant of Eps15 that does not affect clathrin coated pit formation; EH29, DN Eps15 mutant. (A) Representative microscopic fields of cells transfected with each Eps15 construct and infected with VSVΔG-ABLVp G*- RFP pseudovirus. Similar results were obtained for ABLVs G-expressing pseudovirus (not shown). (B) Quantitation of virus-infected cells. A minimum of 1000 cells were counted for each sample per experiment. VSVΔG- VSV G* was included as positive control. Results are expressed as percent virus-infected cells relative to that of controls and represent 3 independent experiments; error bars are SEM. Significance of effect on virus infection was determined using Student’s t-test. **, p < 0.0005; *, p < 0.005. (C) Cholera toxin B (CTX-B) subunit uptake into HEK293T cells is not inhibited by EH29. To verify the specificity of EH29 to inhibit CME, 18 hrs post transfection HEK293T cell monolayers grown on 12 mm coverslips were incubated with Alexa Fluor 594-labeled CTX-B (10 μg/ml) for 1 hr at 37°C. Cells were then washed twice with PBS, fixed and imaged. Images were taken by confocal microscopy with a mid z-section shown. Nuclei were stained with DAPI (4′,6-diamidino-2-phenylindole, dihydrochloride).

Figure 4

CavME endocytosis and macropinocytosis are not required for ABLV G-mediated viral entry. (A) Chemical inhibition of CavME. HEK293T cell monolayers were pretreated with filipin diluted in OptiMEM® for 1 hr at 37°C. Cells were then infected with rVSV (MOI = 1) for 20 hrs and analyzed as described in Figure 1. Drug was maintained for the entire course of infection. (B) Cholera toxin B (CTX-B) subunit uptake is inhibited by filipin. Following pretreatment with filipin as described in (A), HEK293T cell monolayers grown on 12 mm coverslips were incubated with Alexa Fluor 488-labeled CTX-B (10 μg/ml) for 1 hr at 37°C. Cells were then washed twice with PBS, fixed and imaged. Images were taken by confocal microscopy with a mid z-section shown. Nuclei were stained with DAPI, (4′,6-diamidino-2-phenylindole, dihydrochloride). (C) Chemical inhibition of macropinocytosis. HEK293T cell monolayers were pretreated with EIPA diluted in OptiMEM® for 1 hr at 37°C. Cells were then infected with rVSV (MOI = 1) for 20 hrs and analyzed as described in Figure 1. rVSV encoding VSV G (MOI = 1) or EboGP (MOI = 15) were included as negative and positive controls, respectively, to assess EIPA activity. Drug was maintained for the entire course of infection. For (A) and (C) results are expressed as percent virus-infected cells relative to that of untreated controls and represent 3 independent experiments; error bars are SEM. Under these experimental conditions, the chosen MOIs yielded 60-70% virus-infected cells in untreated controls. EIPA, 5-(N-ethyl-N-isopropyl) amiloride.

Figure 5

Actin is required for ABLV G-mediated viral entry. (A) HEK293T cell monolayers were pretreated with latrunculin B (LatB) diluted in OptiMEM® for 1 hr at 37°C. Cells were then infected with max-GFP encoding rVSV reporter viruses (MOI = 3). Under these conditions, a MOI of 3 yielded at least 50% virus-infected cells in untreated controls. Cells were harvested 8 hrs post infection and analyzed as described in Figure 1. Drug was maintained for the entire course of infection and its effect on cell viability (B) was determined by trypan blue staining. Reporter viruses that express VSV G were included as a positive control to assess LatB activity. Results are expressed as percent virus-infected cells relative to that of untreated controls and represent 3 independent experiments; error bars are standard error of the mean (SEM).

Figure 6

Effect of dominant negative Rab GTPases on ABLV G-mediated viral entry. HEK293T cells expressing dsRed-tagged WT and DN Rab5, Rab7, and Rab11were infected with maxGFP encoding rVSV that express ABLVp G, ABLVs G, or VSV G at a MOI = 3 for 8 hrs or with rVSV that expresses EboGP at a MOI = 15 for 20 hrs and then analyzed as described in Figure 1. Infection of all cells was assessed. Under these conditions, the chosen MOIs yielded 40-50% virus-infected cells in controls. Results are expressed as percent virus-infected cells relative to that of WT Rab controls and represent 3 independent experiments; error bars are SEM. (A) Rab5. (B) Rab7. (C) Rab11. **, p < 0.0001; *, p < 0.005.