Bunyavirus requirement for endosomal K+ reveals new roles of cellular ion channels during infection

Abstract

In order to multiply and cause disease a virus must transport its genome from outside the cell into the cytosol, most commonly achieved through the endocytic network. Endosomes transport virus particles to specific cellular destinations and viruses exploit the changing environment of maturing endocytic vesicles as triggers to mediate genome release. Previously we demonstrated that several bunyaviruses, which comprise the largest family of negative sense RNA viruses, require the activity of cellular potassium (K+) channels to cause productive infection. Specifically, we demonstrated a surprising role for K+ channels during virus endosomal trafficking. In this study, we have used the prototype bunyavirus, Bunyamwera virus (BUNV), as a tool to understand why K+ channels are required for progression of these viruses through the endocytic network. We report three major findings: First, the production of a dual fluorescently labelled bunyavirus to visualize virus trafficking in live cells. Second, we show that BUNV traffics through endosomes containing high [K+] and that these K+ ions influence the infectivity of virions. Third, we show that K+ channel inhibition can alter the distribution of K+ across the endosomal system and arrest virus trafficking in endosomes. These data suggest high endosomal [K+] is a critical cue that is required for virus infection, and is controlled by cellular K+ channels resident within the endosome network. This highlights cellular K+ channels as druggable targets to impede virus entry, infection and disease.

Fig 1. K⁺ ions at pH 6.3 expedite BUNV infection.

(A) Schematic protocol of BUNV priming at 37°C with buffers of varying pH and salt concentrations for 2 hrs. Buffers were subsequently diluted out with media and A549 cells were infected with treated virions for 18 hrs prior to cell lysis. Western blot analysis was performed on cell lysates using BUNV-N protein as a marker of BUNV infection and GAPDH as a loading control. (B) Cells were infected with BUNV that had been pre-treated with buffers at pH 7.3, 6.3 or 5.3 (no salt). Cells were lysed and analysed by western blot, as in A (n = 3). (C) BUNV virions were treated with pH 7.3, 6.3 and 5.3 buffers with and without 140 mM (i) KCl or (ii) NaCl. Cells were infected with treated BUNV diluted in media and lysates were analysed by western blot, as in A (n = 3). (iii) Levels of BUNV-N at pH 6.3 under the indicated conditions were quantified by densitometry (n = 3). All error bars indicate mean ± SEM. *Significant difference from control (P < 0.05). (D) As with C, using a pH 6.3 buffer with and without 140 mM K₂SO₄ (n = 3). (E) As with C, using a pH 6.3–6.7 buffer with and without the indicated concentrations of KCl (n = 3). (F) A549 cells were infected with WT or pH 6.3/high KCl treated BUNV for 1 hr and non-internalized virions removed. Infection was allowed to proceed and cells lysed at 3 hr intervals from 12–24 hpi. Western blot analysis (as in A) was carried out to compare WT and pH 6.3/High KCl treated BUNV-N production (n = 3). (G) BUNV virions were treated with a pH 6.3 buffer with and without 140 mM KCl, as in A. A549 cells were treated 30 min prior to infection with 5 mM TEA and cells infected with the treated virions (with and without KCl) in the presence or absence of TEA throughout infection. Cells were lysed 18 hpi and BUNV-N production assessed as in A (n = 3). (H) BUNV virions were treated at pH 6.3 in the presence of 140 mM KCl with or without TEA 10 mM in the priming buffer. Buffers and drug were subsequently diluted out with media and A549 cells infected for 18hrs and BUNV-N levels assessed as in A (n = 3).

Fig 2. Production of SYTO82/DiD-BUNV to monitor virus trafficking.

(A) Time course of A549 cells infected with BUNV. Cells were lysed at 3 hr intervals post-infection. Western blot analysis of BUNV-N protein and GAPDH (loading control) are shown (n = 3). (B) A549 cells were infected with BUNV supernatants collected from infected A549 cells at the indicated 3 hr intervals. Infected cells were fixed at 18 hpi and BUNV-N protein was labelled using anti-BUNV-N antibodies alongside Alexa-Fluor 594 nm secondary antibodies. Widefield images taken on the IncuCyte Zoom are shown (n = 3). Scale bar = 200μm (C) Schematic representation of BUNV labelling. A549 cells were infected with BUNV for 18 hrs, then SYTO82 dye was added to label the viral RNA segments until virus supernatants were collected at 24 hrs. Virus supernatants were concentrated, the BUNV envelope labelled with DiDvbt and SYTO82/DiD-BUNV was purified on a 10–30% iodixanol gradient. 1 ml fractions were collected (n = 3). (D) Fractions from SYTO82/DiD-BUNV purification were used to infect A549 cells for 18 hrs. Western blot analysis for BUNV-N was performed to confirm virus infectivity. (E) Cells were infected as in D, fixed, and stained with anti-BUNV-N and Alexa Fluor-488 antibodies. Widefield images were taken on the IncuCyte Zoom (n = 3). Scale bar = 200 μM. (F) A549 cells were infected with SYTO82/DiD-BUNV for 2 hrs and fixed. SYTO82 (em._max 560 nm) and DiDvbt (em._max 665 nm) fluorescent signals were imaged alongside DAPI in fixed cells. (G) Cells were infected with SYTO82/DiD-BUNV for 1 hr at 4°C, then heated to 37°C and infection was allowed to proceed until fixing at 2 hrs, 4 hrs, 8 hrs or 12 hrs. Biotinylated EGF-488 (2 μg/ml) was added for 15 min at 37°C prior to fixing to act as a cell marker. Confocal images were taken for n>80 cells for each time point and the EGF-488 fluorescence channel was removed in the representative images (Scale bar = 10 μM). (H) NH₄Cl (10 μM) was added at the indicated timepoints post-BUNV infection and BUNV-N expression assessed by western blot analysis 24 hours post-infection. CTL = no drug included during the timecourse (n = 3).

Fig 3. BUNV traffics through endosomes containing K⁺ ions.

(A) AG4 (10 μM) was added to A549 cells for 40 min to allow endosomal uptake, alongside (B)(i) Texas Red labelled EGF (2 μg/ml) or Magic Red cathepsin B dye. Non-internalised dyes were subsequently removed and live cells were imaged. Representative images are shown (n≥100 cells). Scale bar = 10 μM. (ii) Total numbers of AG4 positive puncta were counted per cell and % of colocalised AG4 puncta with each marker calculated in ≥100 cells. (C) AG4 (10 μM) was added to A549 cells for 40 min at either 37°C or 4°C and live cells were imaged as in A. Scale bar = 10 μM. (D) Schematic representation of AG4 uptake into endocytic vesicles and increased fluorescence with passage through early endosomes (EE) into late endosomes (LE) identifying K⁺-rich regions, identifiable using Texas Red labelled EGF. AG4 fluorescence decreases with passage into lysosomes (L). (E) A549 cells were infected with labelled-BUNV in the presence of AG4 (10μM) to allow virus penetration into cells and live cells were imaged 2 hrs or 8 hrs post-infection. Images are representative of ≥ 50 cells. (F)(i) A549 cells were infected with SYTO82/DiD-BUNV and EGF-488 (2 μg/ml) for 1 hour at 4°C and cells warmed to 37°C for the indicated timepoints. Images were taken of live cells at the indicated time points post-warming and are representative of ≥60 cells. Scale bar = 10 μM. (ii) Cells were transfected with Rab7-GFP and infected as in F(i) 24 hours post transfection. Images are representative of ≥ 40 cells. Scale bar = 10 μM. G(i) as in F(i) but cells were infected in the presence of 488-labelled Tf (25 μg/mL) or (ii) cells transfected with Rab11-GFP.

Fig 4. K⁺ channel modulation can impede normal K⁺ accumulation across the endocytic network.

(A) A549 cells were treated for 30 min with 10 mM TEA or left untreated. AG4 (10 μM) was then added in the presence or absence of drug for 40 min. Dye was removed and TEA was re-added onto cells. Fluorescence intensities were quantified using IncuCyte ZOOM imaging and analysis software, and data normalised to untreated (unt) controls over three independent cell populations. NS–no significant difference between no-drug and TEA treated controls (p≥0.05). Scale bar = 10 μM. (B) (i) Cells were treated with 10 mM TEA (or left untreated) and AG4 (10 μm) added as in A, with the addition of Magic Red during the 40 min incubation with AG4. Representative images are shown (n≥60 cells). Scale bar = 10 μM. (ii) Total number of AG4 positive puncta were counted per cell ± TEA and the % of colocalised puncta presented. n≥60 cells, (* = p≤0.05). Scale bar = 10 μM. (C) (i) Cells were treated with 10 mM TEA or left untreated, and AG4 (10 μM) added as in A, with the addition of the pH indicator pHrodo red dextran (10 μg/ml) during the 40 min incubation with AG4. Representative images are shown (Scale bar = 10 μM) and the % of co-localised puncta presented in (ii) n≥60 cells (* = p≤0.05). D Fluorescence intensity of Magic Red was quantified using IncuCyte ZOOM imaging and analysis software and data normalised to untreated controls over three independent cell populations. NS–no significant difference between untreated and TEA treated cells (p≥0.05). Representative images are also shown.

Fig 5. K⁺ channel modulation arrests BUNV trafficking in endosomes.

(A) Cells were treated with TEA (10 mM) for 30 min (or left untreated) and infected with SYTO82/DiD-BUNV for a further 4 hrs in the presence/absence of TEA. EGF-488 was added for the final 15 min of infection and cells were fixed 4 hpi. Confocal images were taken and the EGF-488 fluorescence channel removed in the representative images showing only SYTO82 and DiDvbt (n≥40). Scale bar = 10 μM. (B)(i) Cells treated with TEA (10 mM) or left untreated as in A, were infected with SYTO82/DiD-BUNV and fixed at 2, 4 or 8 hpi. EGF-488 (2 μg/ml) was added for 15 min prior to fixation as in A, with the representative images showing only SYTO82 and DiDvbt channels (n>65 cells). Scale bar = 10 μM. (ii) As in (i) but cells were treated with Qd (200 μM) and fixed 8 hpi (n>65 cells). (iii) The number of SYTO82/DiD-BUNV virions per cell were quantified using images from (i) and (ii) for n>65 cells and normalised to the untreated (no-drug) control. (C) A549 cells were infected with SYTO82/DiD-BUNV for 1 hour at 4°C and treated with cytopainter to label lysosomes. Cells were warmed to 37°C for 1 hr, virus/dye removed by washing and cells incubated for up to 8 hpi. Representative live cell images are shown (≥80 cells). Scale bar = 10 μM. (ii) The number of SYTO82/DiD-BUNV virions co-localising with cytopainter positive puncta were calculated and the % of co-localised puncta presented in (ii) (* = p≤0.05).