The phosphatidylinositol-3-phosphate 5-kinase inhibitor apilimod blocks filoviral entry and infection

Abstract

Phosphatidylinositol-3-phosphate 5-kinase (PIKfyve) is a lipid kinase involved in endosome maturation that emerged from a haploid genetic screen as being required for Ebola virus (EBOV) infection. Here we analyzed the effects of apilimod, a PIKfyve inhibitor that was reported to be well tolerated in humans in phase 2 clinical trials, for its effects on entry and infection of EBOV and Marburg virus (MARV). We first found that apilimod blocks infections by EBOV and MARV in Huh 7, Vero E6 and primary human macrophage cells, with notable potency in the macrophages (IC50, 10 nM). We next observed that similar doses of apilimod block EBOV-glycoprotein-virus like particle (VLP) entry and transcription-replication competent VLP infection, suggesting that the primary mode of action of apilimod is as an entry inhibitor, preventing release of the viral genome into the cytoplasm to initiate replication. After providing evidence that the anti-EBOV action of apilimod is via PIKfyve, we showed that it blocks trafficking of EBOV VLPs to endolysosomes containing Niemann-Pick C1 (NPC1), the intracellular receptor for EBOV. Concurrently apilimod caused VLPs to accumulate in early endosome antigen 1-positive endosomes. We did not detect any effects of apilimod on bulk endosome acidification, on the activity of cathepsins B and L, or on cholesterol export from endolysosomes. Hence by antagonizing PIKfyve, apilimod appears to block EBOV trafficking to its site of fusion and entry into the cytoplasm. Given the drug's observed anti-filoviral activity, relatively unexplored mechanism of entry inhibition, and reported tolerability in humans, we propose that apilimod be further explored as part of a therapeutic regimen to treat filoviral infections.

Fig 1. Apilimod blocks EBOV infection of Huh 7 and Vero E6 and human primary macrophages.

Huh 7 and Vero E6 cells and human monocyte-derived macrophages (hMDM) were treated for 1 h with apilimod (A) or toremifene citrate (B). Two-fold dilutions of the drugs were tested in an 8-point dose-response curve. Then cells were infected at a multiplicity of infection (MOI) of 0.5 for 48 h. Antiviral activity is shown in blue and cytotoxicity is shown in red. The experiment was run on duplicate plates with triplicate wells per dose (mean ± SD; n = 3). The experiment was repeated on 2–4 days. A single representative graph is shown.

Fig 2. Apilimod blocks MARV infection of Huh 7 and Vero E6 cells and human primary macrophages.

Huh 7 and Vero E6 cells and human monocyte-derived macrophages (hMDM) were treated for 1 h with apilimod (A) or toremifene citrate (B). Two-fold dilutions of the drugs were tested in an 8-point dose-response curve. Then cells were infected at a multiplicity of infection (MOI) of 0.5 for 48 h. Antiviral activity is shown in blue and cytotoxicity is shown in red. The experiment was run on duplicate plates with triplicate wells per dose (mean ± SD; n = 3). The experiment was repeated on 2–4 days. A single representative graph is shown.

Fig 3. Apilimod inhibits VLP entry and trVLP infection with similar potency.

(A) HEK 293T/17 cells were pretreated with the indicated concentration of apilimod or DMSO (0 μM) for 1 h at 37°C. VLPs were bound to the cells by spinfection in the presence of the indicated concentration of apilimod or DMSO. After incubation at 37°C (3 h), VLP entry was assayed. To assess cell viability, parallel 293T/17 cells were treated as for entry, but without VLPs or CCF2 loading. After 3 h at 37°C, cell viability was determined. (B) Cells (HEK 293T/17) pretreated as in (A) were infected with trVLPs (in the presence of the indicated concentration of apilimod) for 48 h at 37°C and infection by trVLPs was then assayed. To measure cell viability, parallel HEK 293T/17 cells were pretreated as above and then mock infected (± the indicated concentration of apilimod) for 48 h at 37°C followed by determination of cell viability. Antiviral activity is shown in blue and cytotoxicity is shown in red. (C) Comparison of normalized VLP entry (blue bars) and trVLP infection (red bars) inhibition; data are from (A) and (B). Data are averages of triplicate samples ± SD. Similar results for parallel tests of VLP entry and trVLP inhibition were observed in two additional experiments. Cell viability was tested in one of these experiments, and similar results were obtained. (D) BSC-1 cells were treated with apilimod and VLP entry was assayed as in (A). Data are averages of triplicate samples ± SD. Similar results were observed in an additional experiment.

Fig 4. Apilimod inhibits EBOV GP-mediated entry in a PIKfyve-dependent manner.

HEK 293T/17 cells in 6-well plates were transiently transfected with plasmids encoding GFP-PIKfyve or (as control) GFP. After 18 hr, the cells were reseeded into 96-well plates. After a further 18 hr in culture, the cells were pretreated with the indicated concentration of apilimod for 1 hr at 37°C. MLV-luc-GPΔmucin pseudovirions were then added to the cells in the presence of the indicated concentration of apilimod (or DMSO) and the cells incubated for 48 hr at 37°C, at which point infection was assayed as described in the Methods section. Data indicate averages ± SD of triplicate samples; * indicates p < 0.05. Similar results were obtained in a repeat experiment.

Fig 5. Apilimod inhibits trafficking of EBOV-GP VLPs to NPC1⁺ endolysosomes.

(A-D) BSC-1 cells were pretreated with DMSO (Mock), 20 μM nocodazole, or 0.2 μM apilimod for 1 h at 37°C. VLPs were then bound to cells by spinfection in the presence of the indicated drug, and the cells were then washed and incubated in the presence of the indicated drug for 90 min at 37°C. The cells were then washed, fixed, permeabilized, stained, and analyzed for VLP colocalization with NPC1. (A) Average Manders colocalization coefficients (± SD) from 2 experiments (n = 45 fields each treatment). Each data point (blue dot, triangle or square) represents the Manders colocalization coefficient for 1 image field. (B-D) Representative micrographs of cells treated with (B) DMSO, (C) 20 μM nocodazole, or (D) 0.2 μM apilimod. Scale bars, 10 μm (primary images) and 1μm (insets). ***p<0.001. (E) VLP colocalization with NPC1 was monitored as above except that samples were fixed at 0, 30, 60 or 90 min. Data are the Manders colocalization coefficients from 24–29 microscope fields per sample. Values are averages ± SEM. Apilimod-treated samples were statistically different from mock (***p < 0.0001) at 60 and 90 min.

Fig 6. Apilimod causes accumulation of EBOV-GP VLPs in early endosomes.

(A-C) BSC-1 cells were pretreated with DMSO (Mock) or 0.2 μM apilimod for 1 h at 37°C. Samples were analyzed for VLP colocalization with EEA1 at t = 90 min post initiation of VLP internalization. (A) Average colocalization coefficients (± SD) from 2 experiments (n = 40 and 42 fields for mock- and apilimod-treated cells, respectively). Each data point (blue dot, triangle or square) represents the Manders colocalization coefficient for 1 image field. (B and C) Representative micrographs of cells treated with (B) DMSO (Mock) or (C) 0.2 μM apilimod. Yellow arrows indicate areas of VLP colocalization with EEA1, and red arrows indicate VLPs that have not colocalized with EEA1. The white arrow (C, left panel) indicates a VLP that is within an enlarged EEA1⁺ endosome. Note that this VLP would not score as colocalized with EEA1. Hence the value for VLP colocalization with EEA1 in apilimod-treated cells is likely an underestimate. Scale bars, 10 μm (primary images) and 1μm (insets). ***p<0.001.

Fig 7. Apilimod has no effect on endosome acidification.

BSC-1 cells grown overnight in 35 mm glass bottom dishes (MatTek) were treated for 3 h at 37°C with (A) DMSO (mock), (B) 25 nM bafilomycin (Baf), or (C) 0.2 μM apilimod. Acridine Orange (6.6 μg/mL) was added, and the cells were further incubated at 37°C for 20 min. The cells were then washed and imaged. Scale bars, 10 μm. Images are representative of all observed fields (30 for Baf and 35 for mock- and apilimod-treated cells, respectively).

Fig 8. Apilimod has no effect on cathepsin B+L activity.

Confluent dishes of BSC-1 cells were treated with the indicated concentration of apilimod, DMSO (mock), or 10 μM EST for 1 h at 37°C. The cells were then washed, lysed, and the pH adjusted to 5.0 in reaction buffer. Total cathepsin B+L activity was then assayed using the substrate Z-Phe-Arg-7-amido-4-methylcoumarin. Values represent average fluorescence units ± SD of triplicate samples.

Fig 9. Apilimod does not induce cholesterol accumulation.

BSC-1 cells were grown overnight on 35 mm glass bottom dishes (MatTek). The medium on the cells was replaced with serum-free medium containing 0.05 μM TopFluor Cholesterol (Avanti) plus (A) DMSO (mock), (B) 5 μM U18666A, or (C) 0.2 μM apilimod. The cells were incubated for 18 h at 37°C, washed once with PBS, and imaged in cell imaging medium. Images are representative of all observed fields (25 per treatment).