We quickly applied a BSL-2 Ebola virus entry assay for a repurposing screen of 600 selected drugs. This assay identifies compounds that block Ebola VLP entry into host cells mediated by Ebola virus surface GP. Although the active compounds found to block Ebola VLP entry need to be further tested in wild-type Ebola virus infection assays and in animal models to confirm their antiviral activity, our results represent a proof-of-principle investigation supporting the suitability of this assay for rapid screening of Ebola antiviral compounds that inhibit viral entry. Based on our screen of 600 FDA-approved drugs, we quickly identified 23 compounds that block Ebola VLP entry, with the most potent compound being the microtubule inhibitor Vinblastine (IC 50 =48 nM). We also found an additional 30 active FDA-approved compounds with an IC 50 <10 µM and SI>10 in the extended screening of 2216 additional approved or investigational drugs in our collection. We decided to publish all our data from this drug repurposing screen so that other researchers may quickly access all the accumulated data in this ongoing emergency situation caused by the current unprecedented Ebola virus outbreak. Although our laboratories do not have direct access to BSL-4 facilities for the further confirmation of the activity of these compounds in Ebola virus infection assays and animal models, we hope that our data will be used by other researchers who do possess the capability of performing these experiments.

The 23 drugs identified with inhibitory effects on Ebola VLP entry can be grouped into six categories based on their approved indications. The group with the most potent anti-Ebola virus entry activity consists of microtubule inhibitors, including Vinblastine (IC 50 =48 nM), Vinorelbine (IC 50 =66 nM), Vincristine (IC 50 =140 nM), Colchicine (IC 50 =238 nM), Nocodazole (IC 50 =402 nM), Mebendazole (IC 50 =3.4 µM), and Albendazole (IC 50 =4.8 µM). Microtubules have previously been implicated in Ebola virus VP40 and GP functions.13,14,15 Nocodazole's activity in blocking Ebola virus entry has been described, and the mechanism of action is believed to be related to microtubule depolymerization, leading to the inhibition of viral entry.13

A second group comprises the estrogen receptor modulators Toremifene (IC 50 =0.56 µM), Tamoxifen (IC 50 =0.73 µM), and Raloxifene (IC 50 =1.8 µM). A 2013 report indicated that Toremifene and another estrogen receptor modulator, Clomiphene, effectively block Ebola virus infection independent of the estrogen receptor and were effective in a mouse infection model.16 As previously reported, HeLa cells are estrogen receptor-negative,17,18,19 and the effect of estrogen receptor modulators in blocking Ebola virus entry is believed to be unrelated to the classical estrogen receptors.16 The identification of estrogen receptor modulators and microtubule inhibitors in our screen validates the BSL-2 Ebola virus entry assay for the identification of virus entry inhibitors.

A third group of compounds has antihistamine and anticholinergic activities and includes Clemastine (IC 50 =1.1 µM), Maprotiline (IC 50 =2.4 µM), and Benztropine (IC 50 =2.6 µM). A fourth group contains the antipsychotic/antidepressant drugs Clomipramine (IC 50 =4.9 µM), Thiothixene (IC 50 =1.9 µM), and Trifluoperazine (IC 50 =4.4 µM). The possible mechanism of action of these drugs in inhibiting Ebola VLP entry remains unknown.

A fifth group includes pump/channel blockers such as Digoxin (IC 50 =0.76 µM), Dronedarone (IC 50 =2.2 µM), and Propafenone (IC 50 =6.2 µM). The activity of Digoxin has been found in multiple drug repurposing screens against many targets, including cancer and HIV.20,21,22 Digoxin was reported to suppress HIV-1 structural protein synthesis by altering viral RNA processing23 and to inhibit HIV-1 gene expression via α-1 subunit of the Na+/K+-ATPase inhibition.20 However, Digoxin shows cytotoxicity at a compound concentration similar to that effective in blocking Ebola VLP entry. Dronedarone, a multichannel inhibitor indicated for the treatment of cardiac arrhythmia, inhibits class I sodium (at rapid pacing rates) and class IV calcium channels and exhibits non-competitive anti-adrenergic activity.10 Propafenone is also an orally active antiarrhythmic drug that blocks sodium channels and has beta-adrenoreceptor antagonist and weak calcium antagonist activities.24 The potencies of this group of compounds against Ebola virus entry are relatively weak.

The last four drugs are anticancer drugs and antibiotics. Sunitinib is a multi-kinase inhibitor.25 Daunomycin, similar to Doxorubicin, intercalates DNA, inhibiting DNA biosynthesis. DNA intercalators have also been previously found to induce the production of the antiviral cytokine interferon.11 Azithromycin and Clarithromycin are macrolide antibiotics that block bacterial protein synthesis. It remains to be determined how these drugs affect Ebola VLP entry.

During the manuscript submission and review process, we screened an additional 2216 approved drugs from our NCGC pharmaceutical collection,7 which also contains experimental and animal drugs. We identified and confirmed 30 additional active compounds selected using criteria of an IC 50 <10 µM with a maximal inhibition >50% and an SI>10-fold in the Ebola virus entry assay (Table 2). Table 2 also shows three additional active compounds previously shown to inhibit Ebola virus infection.10,11,12 In addition, we identified 95 active compounds with IC 50 values of 10 to 30 µM or an SI<10-fold or without FDA approval for human use (Supplementary Table S1). At present, we cannot exclude that a few of our reported drugs inhibit reporter beta-lactamase activity instead of inhibiting Ebola VLP-entry. Anti-Ebola virus activities of 16 approved drugs that were tested in the Ebola virus infection assays and/or animal models have been previously reported in the literature.12,15,16,26,27,28,29,30,31 We found that 11 of these 16 compounds are active in our Ebola VLP entry assay (Tables 1 and 2): Nocodazole (IC 50 =0.4 µM), Toremifene (0.55 µM), Tamoxifen (0.76 µM), Raloxifene 1.84 (1.53 µM), Cepharanthine (1.53 µM), Clomiphene (1.72 µM), Dronedarone (2.2 µM), Amodiaquine (4.43 µM), Imipramine (13.7 µM), Chloroquine (15.3 µM), and Nilotinib (15.3 µM).

Although more experiments will be needed to fully understand the possible use of any of these compounds against Ebola virus infection, here we describe a surrogate Ebola virus entry assay that can be used for HTS in a BSL-2 facility for the rapid screening of extensive compound collections. The identification of several drugs previously known to inhibit Ebola virus entry in our repurposing screen further supports the usefulness of our approach. All the screening data obtained in this Ebola VLP entry assay are deposited in the PubChem database for open access. Using the Ebola VLP entry assay described here, we will continue to screen additional experimental drugs in our collection that are currently in clinical trials, and we will also expand this compound screening using the entire NCATS diverse collection of approximately 400 000 compounds. We hope that our screening efforts will result in the identification of novel lead compounds for the development of drugs to treat Ebola virus infections.