key: cord-0002681-odl5p8i4
authors: Nelson, Elizabeth A.; Dyall, Julie; Hoenen, Thomas; Barnes, Alyson B.; Zhou, Huanying; Liang, Janie Y.; Michelotti, Julia; Dewey, William H.; DeWald, Lisa Evans; Bennett, Richard S.; Morris, Patrick J.; Guha, Rajarshi; Klumpp-Thomas, Carleen; McKnight, Crystal; Chen, Yu-Chi; Xu, Xin; Wang, Amy; Hughes, Emma; Martin, Scott; Thomas, Craig; Jahrling, Peter B.; Hensley, Lisa E.; Olinger, Gene G.; White, Judith M.
title: The phosphatidylinositol-3-phosphate 5-kinase inhibitor apilimod blocks filoviral entry and infection
date: 2017-04-12
journal: PLoS Negl Trop Dis
DOI: 10.1371/journal.pntd.0005540
sha: 9018234c987db8e1aec9ca3cae1806b75e4835ad
doc_id: 2681
cord_uid: odl5p8i4

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 (IC(50), 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.

a1111111111 a1111111111 a1111111111 a1111111111 a1111111111

The epidemic of Ebola virus disease (EVD) that raged through Western Africa between 2013 and 2016 was the most severe filovirus disease epidemic in recorded history [1, 2] . While several promising therapeutic antibodies [3] [4] [5] [6] [7] [8] [9] [10] [11] and novel small molecules [12] [13] [14] [15] [16] [17] [18] [19] remain in development, no therapeutic is yet approved to treat patients with EVD. In the continuing pursuit of an anti-Ebola virus (EBOV) therapeutic, one strategy is to identify approved drugs that show anti-EBOV activity [20] [21] [22] [23] [24] [25] [26] [27] [28] , with the goal of repurposing them for an anti-EBOV therapeutic, either alone or as part of a multi-component regimen [29] [30] [31] [32] [33] [34] .

Most of the approved drugs that have been identified as blocking EBOV infection inhibit the entry phase of the viral lifecycle [19] [20] [21] [22] [23] [24] [25] 27, 28] . Cell entry by EBOV is a complex process [35, 36] entailing virus binding to cell surface attachment factors, internalization by macropinocytosis, processing by endosomal proteases, and transport to endolysosomes containing Niemann-Pick C1 (NPC1) [14, 37] , the intracellular receptor for EBOV [38] . Finally, EBOV fuses with the limiting membrane of NPC1 + endolysosomes [39] [40] [41] , liberating its genome and associated proteins into the cytoplasm to begin replication.

The essential role of NPC1 in EBOV entry and infection was powerfully illuminated in a haploid genetic screen [37] . The same screen revealed other gene products critical for EBOV entry [42, 43] including many involved in endosome and lysosome biogenesis and maturation. One of the latter proteins was phosphatidylinositol-3-phosphate 5-kinase (PIKfyve) [37] , a lipid kinase that phosphorylates phosphatidylinositol-3-phosphate (PI3P) to generate phosphatidylinositol-3,5-bisphosphate (PI(3,5)P 2 ). PIKfyve and PI(3,5)P 2 are known to be critical for endosome maturation [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] .

Apilimod is a small molecule that binds to and inhibits the phosphotransferase activity of PIKfyve [54] . The drug was developed as a suppressor of interleukin 12 and 23 production [55] , and was tested in phase 2 clinical trials for treatment of Crohn's disease [56, 57] , psoriasis [58] , and rheumatoid arthritis [59] . Although no clinical benefit has yet been reported, apilimod is deemed to be well tolerated in humans. We chose to test whether apilimod could inhibit infections by EBOV and Marburg virus (MARV) for three reasons. The first was that apilimod binds [54] to the EBOV entry factor PIKfyve [37] . The second was because apilimod emerged from a blinded screen of 35 drugs (S1 Fig; S1 Table) , which were selected as potential inhibitors based upon hypotheses of drugable targets and from theoretical considerations of pathways possibly involved in the EBOV life cycle. The third reason was because apilimod is well tolerated in humans. We find that apilimod inhibits infection by both EBOV and MARV, being notably effective in primary human macrophages, which are initial targets of filoviral infection [60, 61] . Mechanistic studies revealed that apilimod blocks EBOV entry into the cell cytoplasm by working through PIKfyve and that its effect is to block viral particle trafficking to NPC1 + endolysosomes, the site of EBOV fusion [39] [40] [41] . Hence we propose that apilimod be further explored as part of a cocktail of small molecules to combat EVD.

Vero E6 (African green monkey kidney; ATCC 1586) cells were obtained from the American Type Culture Collection (Manassas, VA). Huh 7 (human hepatocellular carcinoma) cells were obtained from Dr. Hideki Ebihara (National Institute of Allergy and Infectious Diseases (NIAID), Rocky Mountain Laboratories, Hamilton, MT). Peripheral blood mononuclear cells (PBMCs) were prepared from human whole blood (Biological Specialty Corporation; Colmar, PA; Cat # 3100-03-04) and human monocyte-derived macrophages (hMDM) were generated from peripheral blood mononuclear cells at the Integrated Research Facility (IRF) immunology core laboratories as described previously [62, 63] . hMDM were characterized by flow cytometric analysis for expression of major macrophage markers, including human leukocyte antigen-D related, CD11b, CD14, CD163, and CD86, to confirm that the hMDM population was mature and highly purified [63] . HEK 293T/17 (Human embryonic kidney; ATCC CRL-11268 via University of Virginia Tissue Culture Facility) and BSC-1 (Grivet monkey kidney; gift from Dr. Xiaowei Zhuang, Harvard University, Cambridge, MA) cells were maintained in growth medium: high glucose Dulbecco's Modified Eagle Medium (DMEM) supplemented with 1% L-glutamine, 1% sodium pyruvate, and 1% antibiotic/antimycotic, all from Gibco Life Technologies (Carlsbad, CA), and either 10% supplemented calf serum (SCS; Hyclone, GE Healthcare Bio-Sciences, Pittsburgh, PA) for HEK 293T/17 cells or 10% fetal bovine serum (FBS, Seradigm, VWR International, Radnor, PA) or 10% cosmic calf serum (CCS, Hyclone) for BSC-1 cells.

Toremifene citrate (CAS 89778-27-8) was purchased from Sigma-Aldrich (St. Louis, MO; Cat# T7204-25MG) and apilimod (CAS 541550-19-0) was purchased from Axon MedChem (Groningen, NL; Cat# 1369). Nocodazole (CAS 31430-18-9) was purchased from Sigma-Aldrich (St. Louis, MO; Cat # M1404-2MG). [3] and MARV was detected with a mouse antibody against the MARV VP40 protein (Cat# IBT 0203-012, IBT Bioservices, Rockville, MD) for 1-2 h at 37˚C. Cells were stained with a secondary antibody, anti-mouse IgG, peroxidase labeled antibody (Cat# 074-1802, KPL Inc., Gaithersburg, MD). Luminescence was detected using Pico chemiluminescent Substrate (Thermo Fisher Scientific Inc., Rockford, IL) and an Infinite M1000 Pro plate reader (Tecan, Morrisville NC).

For quantitation of drug toxicity, 1 black opaque cell plate for each cell type was mock infected (no virus) and treated with drug dilutions under the same conditions as the infected cells. After 48 h, cell viability was measured using the CellTiter Glo Luminescent Cell Viability Assay kit according to the manufacturer's protocol (Promega, Madison, WI). Luminescence was read on an Infinite M1000 Pro plate reader.

Following background subtraction, inhibition was measured as percent relative to untreated infected cells. Non-linear regression analysis was performed, and IC 50s were calculated from fitted curves (log [agonist] vs response [variable slope] with constraint to remain above 0; GraphPad Software, La Jolla, CA). Error bars of dose-response curves represent the standard deviation of three replicates.

Entry reporter viral-like particles (VLPs) bearing GP from the Yambuku-Mayinga isolate of EBOV were prepared as described previously [24, 25, 41] . In brief, HEK 293T/17 cells (~80% confluent) were transfected with cDNAs encoding EBOV GP, VP40, mCherry-VP40, and βlactamase-VP40 (βlam-VP40). The cell medium was collected 24 and 48 h post-transfection and cleared of debris. VLPs in the cleared medium were then pelleted through a 20% sucrose cushion by centrifugation, resuspended in HM buffer (20 mM HEPES, 20 mM MES, 130 mM NaCl, pH 7.4), and repelleted. The final VLP pellet was resuspended (1:100 starting volume of medium) in 10% sucrose-HM. The total protein concentration of the VLPs was determined by bicinchoninic acid (BCA) assay. All entry-reporter VLP preparations were assessed by western blot analyses (for the presence of GP as well as EBOV VP40) and titered on HEK 293T/17 cells to confirm entry competency.

The VLP entry assay scores the ability of βlam-VP40 (from incoming entry reporter VLPs) to cleave a βlam substrate preloaded into the target cell cytoplasm; this only occurs if the VLP fuses with an endosome. The assay was performed as described previously [24, 25, 41] . In brief, 30,000 HEK 293T/17 cells or BSC-1 cells were seeded per well in a clear 96-well plate. 18-24 h post seeding, the cells (~80%-90% confluent) were treated with the indicated concentration of apilimod (Axon MedChem; DMSO for mock) diluted in Opti-MEM I (OMEM, Gibco Life Technologies, Thermo Fisher Scientific) for 1 h at 37˚C in a 5% CO 2 incubator. VLPs diluted in OMEM (with DMSO or the same concentration of apilimod) were bound to the cells by spinfection (250× g) for 1 h at 4˚C. After 3 h in a 37˚C, 5% CO 2 incubator, the βlam substrate CCF2-AM (Life Technologies, via ThermoFisher Scientific, Waltham, MA, USA) was loaded into the cells using 20 or 5 mM Probenecid (MP Biomedicals via ThermoFisher Scientific, Waltham, MA, USA), for BSC-1 or HEK 293T/17 cells, respectively. The cells were incubated overnight at RT and then fixed and analyzed by flow cytometry.

To measure corresponding cell viability, 3 x 10 4 HEK 293T/17 cells, seeded and grown as above but in 96-well opaque white plates were treated as above for VLP entry, but without addition of VLPs or CCF2-AM. Following overnight incubation at RT (as above), the medium was removed and replaced with 50 μL of fresh medium per well. Fifty microliters (per well) of CellTiter-Glo 2.0 (Promega, Madison WI, USA) was then added. After shaking for 2 min at RT at 575 rpm on a Jitterbug orbital shaker (Boekel Scientific, via ThermoFisher Scientific, Waltham, MA, USA), the plate was incubated at RT for 10 min, after which the luminescent signal was detected using a BioTek Synergy HT plate reader (BioTek, Winooski, VT, USA).

Transcription/replication-competent virus-like particles (trVLPs) were prepared as described [25, 64, 65] . In brief, HEK 293T/17 cells were seeded in six well plates and transfected 24 h later (when~50% confluent) using TransIT-LT1 (Mirus, Madison, WI, USA) with pCAGGS-NP, pCAGGS-VP35, pCAGGS-VP30, pCAGGS-L, a tetracistronic minigenome plasmid, and pCAGGS-T7 polymerase. The minigenome plasmid encodes Renilla luciferase, as well as the matrix protein VP40, the nucleocapsid associated protein VP24, and the GP from EBOV. 24 h post transfection, the medium in each well was replaced with 4 mL fresh growth medium containing 5% FBS. 96 h after transfection, the medium (containing trVLPs harboring the Renilla luciferase-containing minigenome) was harvested, pooled, and cleared of cellular debris by centrifugation for 5 min at 800× g and used for trVLP assays as described below.

The trVLP assay measures the ability of trVLPs containing a Renilla luciferase-encoding tetracistronic EBOV minigenome to infect target cells pretransfected with plasmids encoding proteins to enhance trVLP entry (the adhesion factor Tim-1) and (other plasmids) to support replication of the minigenome. If trVLPs enter target cells, the minigenome is replicated and transcribed, leading to Renilla luciferase reporter activity [64, 65] . In brief: Cells were pretreated with apilimod (Axon MedChem; DMSO for mock) as described above. The pretreatment solution was then removed and replaced with 100 μL trVLPs diluted to 200 μL in growth medium containing 10% SCS and the indicated concentration(s) of apilimod (DMSO for mock). The cells were then incubated for 48 h at 37˚C in a 5% CO 2 incubator, after which the medium was replaced with 40 μL of fresh growth medium containing 10% SCS. 40 μL of Renil-laGlo substrate (Promega, Madison, WI, USA) was then added to each well and the plate immediately analyzed on a GloMax plate reader (Promega, Madison, WI, USA).

To assess cell viability in corresponding samples without trVLPs, the pretreatment solution was removed and replaced with 200 μL fresh growth medium containing 10% SCS and the indicated concentrations of apilimod (DMSO for mock). The cells were then incubated for 48 h at 37˚C in a 5% CO 2 incubator, after which the medium was replaced with 40 μL of fresh growth medium containing 10% SCS. 40μL of CellTiter-Glo 2.0 (Promega) was then added to each well and the plate placed on a Jitterbug orbital shaker (575 rpm) for 2 min at RT. The plate was then incubated at RT for 10 min, after which the luminescent signal was detected using a Synergy HT (BioTek, Winooski, VT, USA) plate reader.

BSC-1 cells were seeded in 35mm glass bottom dishes (MatTek, Ashland, MA) that were coated with 20 μg/mL fibronectin (Sigma-Aldrich, St. Louis, MO, USA). The next day, when the cells were 90-100% confluent, the cells were treated with the indicated drug at the indicated concentration, diluted in growth medium containing 10% cosmic calf serum, for 3 h at 37˚C in a 5% CO 2 incubator. Acridine Orange (Life Technologies, Thermofisher Scientific, Waltham, MA, USA) was added directly to each dish to reach a final concentration of 6.6 μg/mL. The cells were incubated at 37˚C in a 5% CO 2 incubator for 20 min and then were washed 3 times with phosphate buffered saline (PBS), 5 min per wash. Cell imaging medium [Live cell imaging solution (Molecular Probes, Cat# A14291DJ, Thermo Fisher Scientific, Waltham, MA) containing 10% FBS and 4.5 g/L glucose] was added to the dishes and images were taken using a Nikon C1 laser scanning confocal unit attached to a Nikon Eclipse TE2000-E microscope with a 100X, 1.45-numerical-aperature (NA) Plan Apochromat objective (Nikon, Melville, NY).

BSC-1 cells were seeded in 35mm glass bottom dishes (MatTek, Ashland, MD, USA) that had been coated with 20 μg/mL fibronectin (Sigma-Aldrich). The next day, when the cells were 90-100% confluent, the cells were treated with the indicated drug at the indicated concentration plus 0.05 μM TopFluor Cholesterol (Avanti Polar Lipids, Alabaster, AL), diluted in serum-free growth medium, for 18 h at 37˚C in a 5% CO 2 incubator. Following incubation, the cells were gently rinsed once with PBS and cell imaging medium (Live cell imaging solution (Molecular Probes) containing 10% FBS and 4.5 g/L glucose) was added to the dishes. The cells were incubated at 37˚C in a 5% CO 2 incubator for 30 min. Images were then taken using a 60X /1.45 numerical aperture (NA) Nikon Plan Apo total internal reflection fluorescence oil immersion objective attached to a Nikon Eclipse TE2000-E microscope equipped with a Yokogawa CSU 10 spinning-disk confocal unit, a 512-by-512 Hamamatsu 9100c-13 EM-BT camera, a motorized stage maintained at 37˚C, and a Nikon Perfect Focus system.

VLP trafficking experiments were performed in BSC-1 cells essentially as described previously [41] with the following minor modifications. Cells were pretreated with apilimod or nocodazole (indicated concentrations) diluted in OMEM for 1h at 37˚C prior to VLP addition. VLPs at 0.5 μg/well were bound to the cells by spinfection (250 x g) for 1 h at 4˚C. After incubation at 37˚C (CO 2 incubator) for the indicated times, the cells were fixed and washed. Next, primary antibodies (1:1000 rabbit α-NPC1, (Abcam) or 1:1000 mouse α-early endosome antigen 1 (EEA1), BD Biosciences, San Jose, CA) were added for 45 min at room temperature (RT) and, following washing, secondary antibodies (1:1500 α-mouse or α-rabbit AlexaFluor 488, Life Technologies, Thermo Fisher Scientific) were added for 30 min at RT. The cells were washed and the coverslips were mounted overnight on glass slides using ProLong Gold Antifade reagent (Life Technologies, Thermo Fisher Scientific). The coverslips were then sealed and images were taken using a Nikon C1 laser scanning confocal unit attached to a Nikon Eclipse TE2000-E microscope with a 100X, 1.45-NA Plan Apochromat objective. Colocalization of VLPs (red, mCherry-VP40) and endosomal markers (green, NPC1 or EEA1) was assessed as Manders coefficients. Statistics were analyzed using GraphPad Prism 7. Normality of the data was assessed using the D'Agostino & Pearson normality test. Significance of normally distributed data was determined by T-test, and significance of non-normally distributed data was determined by Mann-Whitney test.

Cathepsin B+L activity was assayed as described previously [20, 21, 24] 

Thirty-five drugs obtained from the National Center for Advancing Translational Sciences (NCATS) were dissolved in DMSO at 500 μM. Drugs were diluted in DMEM (Life Technologies, Thermo Fisher Scientific) supplemented with 2 mM L-Glutamine (Q; Life Technologies, Thermo Fisher Scientific) and 100 U/ml penicillin and 100 μg/ml streptomycin (PS; Life Technologies, Thermo Fisher Scientific). Drugs were added to confluent Vero E6 cells. Drugs and cells were then incubated at 37˚C and 5% CO 2 in a humidified incubator in 96-well plates for final concentrations of 10, 1, or 0.1 μM in a final volume of 100 μl DMEM/PS/Q with 2% FBS (Life Technologies, Thermo Fisher Scientific). Cells were returned to the incubator for 2 h.

For efficacy studies, 50 μl DMEM/PS/Q containing 1x10 3 TCID 50 of recombinant EBOV expressing firefly luciferase from an additional transcriptional unit (rgEBOV-luc2, Genbank Accession number KF990214.1) [66] was added to the cells. At 48 h post-inoculation the supernatant was removed and 100 μl GloLysis buffer (Promega) was added to the cells and incubated for 10 min at RT. Afterwards, 40 μl lysate was added to 40 μl BrightGlo reagent (Promega) in white opaque 96 well plates, and reporter activity was measured using a GloMax luminometer.

For cytotoxicity studies, 50 μl of DMEM/PS/Q without virus was added to the cells following the 2 h pre-incubation with drugs, and cells were returned to the incubator. At 48 h, 100 μl of supernatant was removed, and 50 μl of CellTiterGlo reagent (Promega) was added to the cells. Cells were incubated for 2 min on an orbital shaker at 60 RPM, and then for an additional 10 min without shaking at RT. Supernatants were transferred to white opaque 96-well plates, and reporter activity was measured using a GloMax luminometer (Promega). Ribavirin at final concentrations of 1 mg/ml, 100 μg/ml, and 10 μg/ml, as well as DMSO at concentrations corresponding to the DMSO concentrations found in the drug dilutions served as controls. All experiments involving infectious rgEBOV-luc2 were performed in the maximum containment laboratory of the Rocky Mountain Laboratories, National Institutes of Health, Hamilton, MT, following approved protocols.

Production of MLV-luciferase (luc) pseudovirus particles HEK 293T/17 cells were seeded at a density of 3 x 10 6 cells per 10 cm plate. The next day, when the cells were approximately 60% confluent, the media above the cells was replaced with 6mL OMEM and the cells were transfected with 2.4 μg pTG-luc, 1.2 μg pCMV-MLVgag-pol, 1.2 μg pGPΔmucin (encoding Ebola GP deleted for its mucin domain), and 1.2 μg of MLV-gag-βlam diluted to 300 μL in OMEM (per plate), using 18 μL Lipofectamine 2000 (Invitrogen, Thermo-Fisher Scientific, Walthan, MA) diluted to 300 μL in OMEM (per plate). 4h post transfection, 6 mL of antibiotic-free growth medium containing 10% SCS was added to each plate, and the cells were incubated for 48 h at 37˚C in a CO 2 incubator. Cell medium containing pseudovirus was then collected, pooled, and cleared of cellular debris by centrifugation at 250 x g for 7 min. The clarified supernatant containing pseudovirus was then passed through an 0.45 μm filter and the pseudoviruses were concentrated 100-fold by high-speed centrifugation through a 25% sucrose cushion in HM buffer (20mM HEPES, 20mM MES, 130mM NaCl, pH7.4) for 75 min at 103,745 x g. The final pseudovirus pellet was resuspended in growth medium (100-fold concentrated from harvest supernatant).

HEK 293T/17 cells were seeded at a density of 5 x 10 5 cells per well in 6-well plates. When the cells were~50% confluent (~18-24 h post seeding), they were transfected with plasmids encoding GFP-PIKfyve or pEGFP-Cl using TransIT LT1 transfection reagent (Mirus, Madison, WI) following the manufacturer's instructions. 18 h post transfection, the cells were re-seeded in 96 well opaque white plates (BD Falcon, ThermoFisher Scientific, Waltham, MA) at a density of 3 x 10 4 cells per well. Transfection was confirmed by fluorescence microscopy. 18 h post re-seeding, the cells were pretreated for 1 h at 37˚C with apilimod. MLV-luciferase particles pseudotyped with EBOV GPΔmucin were added to the cells in the presence of apilimod, and infection was allowed to proceed for 48 h at 37˚C. The cells were then washed once with PBS and overlaid with 50μL PBS. Luciferase activity was then immediately assayed by adding 50μL of Britelite plus (Perkin Elmer, Waltham, MA) and reading on a Glomax plate reader (Promega, Madison, WI, USA) following the manufacturer's instructions.

We first tested whether apilimod blocks EBOV infection in cell cultures. Apilimod blocked EBOV infection of Huh 7 (liver) cells, Vero E6 (kidney) cells, and primary human monocytederived macrophages (hMDMs) (Fig 1) with 6-to 247-fold higher activity than the positive control, toremifene citrate [20, 22] . Apilimod also blocked MARV infection of the same cell types (Fig 2) with 38-to 1160-fold higher activity than the positive control. Apilimod was notably potent (IC 50 , 10 nM) against both filoviruses in hMDMs (Figs 1 and 2, Table 1 ). While similar potency (IC 50 , 15-25 nM) was seen in Vero E6 cells, apilimod was~10-fold less potent (IC 50 , 140 nM) in Huh 7 cells (Fig 1, Fig 2 and Table 1 ).

To begin to probe the mechanism by which apilimod blocks EBOV infection, we directly compared dose-response profiles for blocking EBOV entry and replication using entry reporter VLPs [24] and trVLPs [64] , respectively. Both sets of VLPs bore the GP from the Mayinga isolate of EBOV. Apilimod blocked EBOV particle entry ( Fig 3A) and replication ( Fig 3B) with similar dose-response profiles (Fig 3C) . This finding suggested that apilimod blocks the entry phase of the filoviral lifecycle.

Since apilimod targets PIKfyve [54] , since PIKfyve is required for EBOV entry and infection [37] , and since apilimod blocks EBOV entry and infection (Figs 1-3) , we reasoned that apilimod blocks EBOV entry and infection by targeting PIKfyve. To test this hypothesis we overexpressed PIKfyve (GFP-PIKfyve) and compared the dose response needed for apilimod to block EBOV-GP mediated pseudovirus infection. As predicted, and as seen in Fig 4, higher doses of apilimod were needed to achieve similar levels of inhibition of EBOV GP-mediated infection in GFP-PIKfyve vs. GFP expressing cells. This supports our proposal that apilimod blocks EBOV entry and infection through a PIKfyve-dependent pathway.

Recent work has shown that EBOV traffics deep in the endocytic pathway, to NPC1 + endolysosomes, for fusion and entry [39] [40] [41] . Therefore we asked whether apilimod prevents EBOV VLPs from reaching NPC1 + endolysosomes. We used BSC-1 cells for these experiments as they are more suitable (flatter and more adherent) for immunofluorescence analysis than the HEK 293T/17 cells used in previous experiments (Figs 3 and 4) . We first demonstrated that apilimod blocks EBOV VLP entry into BSC-1 cells with the same approximate dose-dependency as its effects in HEK 293T/17 cells (Fig 3D) . Given that, we next asked if apilimod blocks trafficking of EBOV GP VLPs to NPC1 + endolysosomes in BSC-1 cells. As seen in Fig 5, this was, indeed, the case. Apilimod blocked EBOV VLP trafficking to NPC1 + endolysosomes to a similar extent as nocodazole, a microtubule destabilizer that is known to block traffic between early and late endosomes [67] .

The findings presented in Fig 5A-5D were obtained after allowing VLPs pre-bound to the cell surface to enter cells for 90 min at 37˚C. This time point was chosen based on our extensive prior analysis of the time courses of EBOV VLP co-localization with NPC1 + endolysosomes and entry into BSC-1 cells [41] . To assure that apilimod did not accelerate VLP trafficking to NPC1 + endolysosomes, we analyzed co-localization of VLPs at various times up to 90 min in cells treated or not treated with apilimod. As seen in Fig 5E, at no point during this time course were VLPs seen to associate with NPC1 + endolysosomes in apilimod-treated cells, supporting our conclusion that apilimod blocks trafficking of EBOV particles to NPC1 + endolysosomes.

Concomitant with decreased trafficking of EBOV GP VLPs to NPC1 + endolysosomes, apilimod caused EBOV VLPs to accumulate in EEA1 + endosomes (Fig 6) . In apilimod-treated cells, EEA1 + endosomes appeared larger than those in mock-treated cells, consistent with previous reports showing enlarged endosomes in cells genetically deficient for PIKfyve or treated with PIKfyve inhibitors [44, 45, 50, 53] .

The findings presented in Figs 5 and 6 indicate that the primary mechanism by which apilimod blocks EBOV entry and infection (Figs 1-4) is by blocking virus transport from early (EEA1 + ) endosomes to their site of fusion in NPC1 + endolysosomes. To further test this model, we asked whether apilimod affects other attributes of the endosomal pathway needed for EBOV entry, either endosome acidification or the activity of cathepsin B and L [42, 43] . At a concentration that strongly blocked EBOV entry and infection, apilimod had no detectable effect on endosome acidification (Fig 7) . Bafilomycin, an inhibitor of EBOV infection that blocks endosome acidification, was used as a positive control. Apilimod also had no apparent direct effect on the activity of cathepsin B and L (Fig 8) , in contrast to EST, a known inhibitor of cathepsin B, H, and L. Several cationic amphiphilic drugs such as U18666A that block EBOV entry and infection [20] [21] [22] 24, 25, 68] induce cholesterol accumulation in endolysosomes [20, 24] . In contrast, apilimod did not cause a detectable increase in cholesterol levels in endolysosomes (Fig 9) . Hence apilimod appears to block filoviral entry and infection by inhibiting virus particle trafficking to NPC1 + endolysosomes, the portal for entry of the EBOV genome into the host cell cytoplasm [39] [40] [41] .

As tragically demonstrated by the recent epidemic of EVD in Western Africa (2013-2016), a pressing need remains to develop therapeutics to treat patients infected with filoviruses [1] . While novel monoclonal antibody and small molecule therapeutics are in the pipeline [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] , a parallel approach is to consider repositioning an approved drug or a drug that has proven safe in phase 2 clinical trials to treat EVD. In this study we tested the potential utility of apilimod as an anti-filoviral agent. We evaluated this drug for three reasons: (a) apilimod directly targets PIKfyve [54] , a known EBOV entry factor [37] ; (b) apilimod emerged from a blinded screen of 35 drugs targeting cell signaling pathways (S1 Fig and S1 Table) ; and (c) apilimod was reported to be well tolerated in humans in several phase 2 clinical trials [56] [57] [58] [59] . After demonstrating that apilimod has activity against both EBOV and MARV in several cell types, notably in human macrophages (Figs 1 and 2, Table 1 ), we demonstrated that, by working through PIKfyve, its primary mode of action is to block trafficking of EBOV particles to NPC1 + endolysosomes (Figs 3-9) , the site of EBOV fusion and entry [39] [40] [41] .

What is PIKfyve and how might it participate in EBOV entry?

EBOV journeys deep into the cellular endosomal system, entering the cytoplasm through endolysosomes that are positive for NPC1 and two-pore channel 2 (TPC2) [39] . In addition to NPC1, its intracellular receptor [14, 37, 38] , EBOV requires multiple factors involved in endosome and lysosome biogenesis and maturation for entry [37] . One of the latter factors is PIKfyve [37] , which phosphorylates PI3P to generate PI(3,5)P 2. Here we have shown that apilimod, which binds to PIKfyve [54] , blocks EBOV entry and infection in a PIKfyve-dependent manner.

The inhibitory effect of apilimod on EBOV entry is likely due to a defect in the maturation of endolysosomes, as extensive evidence indicates the importance of PIKfyve and PI(3,5)P 2 in this process [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] . Although the exact mechanism by which PIKfyve and PI(3,5)P 2 orchestrate endosome maturation is not known, several mechanisms have been postulated. Considered in these mechanisms are the observations that two Ca ++ channels found in (endo) lysosomes-transient receptor potential cation channel, mucolipin 1 (TRPML1) [50, 69] and TPC2 [70] [71] [72] -are downstream effectors of PIKfyve and PI(3,5)P 2 . Through its action on TRPML1, PIKfyve has been shown to regulate the fission and consequent remodeling and maturation that reduces the size of macropinosomes containing endocytosed material from the cell surface and exterior [50] . In addition, the TPC2 channel has been reported to be activated by PI(3,5)P 2 [71, 72] . Intriguingly, both macropinocytosis [35, 73, 74] and TPC2 [75] are involved in EBOV entry and infection. Although we do not yet know all of the endolysosomal factors needed to trigger EBOV GP for fusion [36, 39, 43, 76] , it appears clear that proper endosomal maturation is required. These findings are consistent with the mounting evidence for a role of PIKfyve in EBOV entry and our observation that the PIKfyve inhibitor, apilimod, blocks transport of EBOV particles to NPC1 + endolysosomes. It is therefore likely that interconnected effects of apilimod on PIKfyve [37, 54] , TPC2 [75] , and endolysosome maturation culminate in its blockade of EBOV entry and infection.

Several approved drugs that function as EBOV entry inhibitors (e.g., clomiphene, toremifene, and sertraline) block EBOV entry into the cytoplasm after EBOV particles have been delivered to NPC1 + endolysosomes [20, 21, 24, 25] . Hence they likely interfere with some aspect of the virus-endolysosome membrane fusion process, per se. Other approved drugs, including chloroquine, niclosamide, atovaquone, amodiaquine and quinacrine [21] [22] [23] , block endosomal acidification. Hence these drugs likely interfere with the processing of EBOV GP by acid-optimal endosomal cathepsins [42, 43] and/or low pH-induced conformational changes required for fusion activity of cleaved GP [77, 78] .

In contrast to these mechanisms, our findings indicate that apilimod blocks EBOV entry by blocking particle delivery into NPC1 + endolysosomes. The only other approved drugs that we know of with anti-EBOV activity that are expected to have this mode of action are microtubule-disrupting agents, including colchicine, nocodazole, vinblastine, and vinorelbine [21] [22] [23] . Hence the mode of action of apilimod as an anti-filoviral agent is novel. Rather than blocking EBOV trafficking to NPC1 + endosomes by interfering with microtubules, apilimod blocks EBOV trafficking by inhibiting PIKfyve.

Our findings indicate that apilimod has similar anti-viral activity against EBOV and MARV, consistent with the need for NPC1 in endolysosomes for the entry of these and other filoviruses [37, 79, 80] . We therefore consider it likely that apilimod, a host-directed small molecule, will have broad or even pan-filoviral activity. Furthermore, since many other viruses, socalled late penetrating viruses [36, 81] , traffic beyond early endosomes for entry, it is possible that apilimod will block entry and infection by members of other virus families.

Apilimod is an investigational drug. Although it has been tested in phase 2 clinical trials for the treatment of Crohn's disease, psoriasis, and rheumatoid arthritis, the drug has not yet been approved for any indication. Nonetheless apilimod was well-tolerated in humans in the reported phase 2 trials [56] [57] [58] [59] . We found that intraperitoneal delivery of 10 mg/kg of apilimod to mice resulted in a C max of 2.53 μM. This is well above the IC 50 for apilimod inhibition of EBOV infection in the three cell lines tested,~250 times greater than the IC 50 in hMDMs (10 nM), initial major targets of filoviral infections [60, 61] . We therefore consider it plausible that apilimod be used in the treatment of EVD. And, while apilimod may not function as a single agent, it may perform well as a component in an anti-filoviral small molecule cocktail. In summary, we introduce apilimod, a small molecule PIKfyve inhibitor that has proven safe in phase 2 clinical trials, as a potential anti-filoviral agent.

Vero E6 cells were pretreated for 2 h with compounds at the indicated concentrations. Cells were then infected with 1000 TCID50 of a reporter-expressing Ebola virus in presence of the compounds, and 2 d later reporter activity (here shown on a log 10 scale) was measured. In parallel, cell viability was determined in drug-treated but non-infected cells using a commercial cell viability assay, with viability in untreated cells set to 100%. As positive control, ribavirin was used at concentrations of 1 mg/ml, 100 μg/ml, and 10 μg/ml, and DMSO served as negative control. Mock indicates non-infected, untreated cells. Mean and standard deviation of 3 biological replicates are shown. (TIF) S1 Table. Compounds blindly screened in the rgEBOV-luc2 infection assay (S1 Fig). Indicated are the screen code, NCGC number, name, and attributed mechanism of action for each compound. During the screen, apilimod (NCGC00263093-01) was coded D03; it is identified as apilimod in S1 Fig.  ( 

Ebolavirus Evolution: Past and Present

Reversion of advanced Ebola virus disease in nonhuman primates with ZMapp

Structural and molecular basis for Ebola virus neutralization by protective human antibodies

Protective monotherapy against lethal Ebola virus infection by a potently neutralizing antibody

Isolation of potent neutralizing antibodies from a survivor of the 2014 Ebola virus outbreak

Discovery of an antibody for pan-ebolavirus therapy

Bispecific Antibody Affords Complete Post-Exposure Protection of Mice from Both Ebola (Zaire) and Sudan Viruses

Protection against filovirus diseases by a novel broad-spectrum nucleoside analogue BCX4430

Therapeutic efficacy of the small molecule GS-5734 against Ebola virus in rhesus monkeys

Small molecule inhibitors reveal Niemann-Pick C1 is essential for Ebola virus infection

Cysteine Cathepsin Inhibitors as Anti-Ebola Agents

Ebola virus entry into host cells: identifying therapeutic strategies. Curr Clin Microbiol Rep

Chemical and Structural Aspects of Ebola Virus Entry Inhibitors

Filovirus proteins for antiviral drug discovery: A structure/ function analysis of surface glycoproteins and virus entry

Novel Small Molecule Entry Inhibitors of Ebola Virus

FDA-Approved Selective Estrogen Receptor Modulators Inhibit Ebola Virus Infection

A screen of approved drugs and molecular probes identifies therapeutics with anti-Ebola virus activity

Identification of 53 compounds that block Ebola virus-like particle entry via a repurposing screen of approved drugs. Emerg Microbes Infect

A Systematic Screen of FDA-Approved Drugs for Inhibitors of Biological Threat Agents

Multiple cationic amphiphiles induce a niemann-pick C phenotype and inhibit ebola virus entry and infection

Clomiphene and Its Isomers Block Ebola Virus Particle Entry and Infection with Similar Potency: Potential Therapeutic Implications

Can you teach old drugs new tricks?

Inhibition of Ebola and Marburg Virus Entry by G Protein-Coupled Receptor Antagonists

Filoviruses utilize glycosaminoglycans for their attachment to target cells

The Quad Pill, a Once-Daily Combination Therapy for HIV Infection

Emerging treatments for chronic hepatitis C

A Rapid Screening Assay Identifies Monotherapy with Interferon-ß and Combination Therapies with Nucleoside Analogs as Effective Inhibitors of Ebola Virus

High-throughput combinatorial screening identifies drugs that cooperate with ibrutinib to kill activated B-cell-like diffuse large B-cell lymphoma cells

Synergistic drug combinations tend to improve therapeutically relevant selectivity

Chemical combinations elucidate pathway interactions and regulation relevant to Hepatitis C replication

Ebola virus entry: a curious and complex series of events

Fusion of Enveloped Viruses in Endosomes

Ebola virus entry requires the cholesterol transporter Niemann-Pick C1

Ebola virus entry requires the host-programmed recognition of an intracellular receptor

Ebolavirus Glycoprotein Directs Fusion through NPC1+ Endolysosomes

Direct Visualization of Ebola Virus Fusion Triggering in the Endocytic Pathway

Ebola Virus and Severe Acute Respiratory Syndrome Coronavirus Display Late Cell Entry Kinetics: Evidence that Transport to NPC1+ Endolysosomes Is a Rate-Defining Step

Endosomal proteolysis of the Ebola virus glycoprotein is necessary for infection

Role of Endosomal Cathepsins in Entry Mediated by the Ebola Virus Glycoprotein

Phosphatidylinositol 3,5-bisphosphate and Fab1p/PIKfyve underPPIn endo-lysosome function

Mammalian cell morphology and endocytic membrane homeostasis require enzymatically active phosphoinositide 5-kinase PIKfyve

YM201636, an inhibitor of retroviral budding and PIKfyve-catalyzed PtdIns(3,5)P2 synthesis, halts glucose entry by insulin in adipocytes

A selective PIKfyve inhibitor blocks PtdIns(3,5)P(2) production and disrupts endomembrane transport and retroviral budding

The mammalian phosphatidylinositol 3-phosphate 5-kinase (PIKfyve) regulates endosome-to-TGN retrograde transport

PIKfyve: Partners, significance, debates and paradoxes

PIKfyve Regulates Vacuole Maturation and Nutrient Recovery following Engulfment

PI(3,5)P(2) controls membrane trafficking by direct activation of mucolipin Ca(2+) release channels in the endolysosome

PIKfyve inhibition interferes with phagosome and endosome maturation in macrophages

Vacuole size control: regulation of PtdIns(3,5)P2 levels by the vacuole-associated Vac14-Fig4 complex, a PtdIns(3,5)P2-specific phosphatase

PIKfyve, a class III PI kinase, is the target of the small molecular IL-12/IL-23 inhibitor apilimod and a player in Toll-like receptor signaling

Selective abrogation of Th1 response by STA-5326, a potent IL-12/IL-23 inhibitor

double-blind, placebo-controlled trial of the oral interleukin-12/23 inhibitor apilimod mesylate for treatment of active Crohn's disease

A phase 1/2A trial of STA 5326, an oral interleukin-12/23 inhibitor, in patients with active moderate to severe Crohn's disease

Apilimod inhibits the production of IL-12 and IL-23 and reduces dendritic cell infiltration in psoriasis

Brief report: a phase IIa, randomized, double-blind, placebo-controlled trial of apilimod mesylate, an interleukin-12/interleukin-23 inhibitor, in patients with rheumatoid arthritis. Arthritis Rheum

Analysis of Ebola Virus Entry Into Macrophages

Ebola virus exploits a monocyte differentiation program to promote its entry

Characterization of Yellow Fever Virus Infection of Human and Non-human Primate Antigen Presenting Cells and Their Interaction with CD4+ T Cells

Evaluation of the activity of lamivudine and zidovudine against Ebola virus

A Novel Life Cycle Modeling System for Ebola Virus Shows a Genome Length-Dependent Role of VP24 in Virus Infectivity

Modeling The Lifecycle Of Ebola Virus Under Biosafety Level 2 Conditions With Virus-like Particles Containing Tetracistronic Minigenomes

A novel Ebola virus expressing luciferase allows for rapid and quantitative testing of antivirals

Virus entry at a glance

Niemann-pick C1 is essential for ebolavirus replication and pathogenesis in vivo

Up-regulation of lysosomal TRPML1 channels is essential for lysosomal adaptation to nutrient starvation

NAADP mobilizes calcium from acidic organelles through two-pore channels

TPC proteins are phosphoinositide-activated sodium-selective ion channels in endosomes and lysosomes

Convergent regulation of the lysosomal two-pore channel-2 by Mg 2+ , NAADP, PI(3,5)P 2 and multiple protein kinases

Ebolavirus is internalized into host cells via macropinocytosis in a viral glycoprotein-dependent manner

Cellular entry of ebola virus involves uptake by a macropinocytosis-like mechanism and subsequent trafficking through early and late endosomes

Ebola virus. Twopore channels control Ebola virus host cell entry and are drug targets for disease treatment

A new player in the puzzle of filovirus entry

Structure and function of the complete internal fusion loop from Ebolavirus glycoprotein 2

Designed protein mimics of the Ebola virus glycoprotein GP2 α-helical bundle: stability and pH effects

Structural basis for Marburg virus neutralization by a cross-reactive human antibody

The Glycoproteins of All Filovirus Species Use the Same Host Factors for Entry into Bat and Human Cells but Entry Efficiency Is Species Dependent

Late-penetrating viruses

performed this work as an employee of Loveless Commercial Contracting, Inc. a subcontractor to BMI. JM, WHD, and GGO Jr. performed this work as employees of MRI Global, a subcontractor to BMI. Research at the NCATS was supported by its intramural program. Additional funding was provided by RO1 AI114776 (JMW), DOD Contract #: W81XWH-12-P-0007 (JMW) and, in part, by the intramural research program of the NIAID (TH), National Institutes of Health. The findings and conclusions in this report are those of the authors and do not necessarily reflect the views or policies of the US Department of Health and Human Services or of the institutions and companies affiliated with the authors. These funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

We thank the IRF Cell Culture staff and IRF Viral Repository team for preparing cells and virus, respectively. We thank the IRF Drug Screen team for maintaining the high-quality performance of the screening assays used. In addition, we acknowledge Laura Bollinger and Jiro Wada at the IRF for technical writing services and figure preparation, respectively. We thank Dr. Hideki Ebihara (National Institute of Allergy and Infectious Diseases (NIAID), Rocky Mountain Laboratories, Hamilton, MT) for Huh 7 cells. We thank Dr. Xiaowei Zhuang, Harvard University, Cambridge MA for the gift of BSC-1 cells and the following for gifts of plasmid DNAs: Dr. Gary Whittaker (Cornell University, Ithaca NY) and Dr. Jean Dubuisson (University of Lille, France) for pTG-luc and pCMV-MLVgag-pol; Dr. Pietro de Camilli (Yale