Virtual screen for repurposing approved and experimental drugs for candidate inhibitors of EBOLA virus infection

The ongoing Ebola virus epidemic has presented numerous challenges with respect to control and treatment because there are no approved drugs or vaccines for the Ebola virus disease (EVD). Herein is proposed simple theoretical criterion for fast virtual screening of molecular libraries for candidate inhibitors of Ebola virus infection. We performed a repurposing screen of 6438 drugs from DrugBank using this criterion and selected 267 approved and 382 experimental drugs as candidates for treatment of EVD including 15 anti-malarial drugs and 32 antibiotics. An open source Web server allowing screening of molecular libraries for candidate drugs for treatment of EVD was also established.

This article is included in the Ebola Collection The current Ebola virus outbreak is one of the largest outbreaks of its kind in history and the first in West Africa. By January 14, 2015, a total of 21296 probable and confirmed cases, including 8429 deaths from Ebola virus disease (EVD), had been reported from five countries in West Africa -Guinea, Liberia, Nigeria, Senegal, and Sierra Leone (http://apps.who.int/iris/bitstream/10665/ 148237/2/roadmapsitrep_14Jan2015_eng.pdf?ua=1). EVD with a high case-fatality rate of 40% and with currently no approved vaccine or therapy, represents a major public health threat.
In response to the current Ebola virus outbreak, the international community has urged for accelerated development of drugs against EVD but also has endorsed the clinical use of unregistered treatments for Ebola 1 . Conventional time and the money consuming approach of drug development (> 10 years; > 2 billions $) does not meet the current urgent need for anti-Ebola drugs. Repurposing or repositioning of existing drugs could overcome some of these obstacles and help in the rapid discovery and development of therapeutics for EVD, although this approach does not negate the need for some preclinical studies and clinical trials for validation of the proposed indications. Recently, results of two large repurposing screenings of Food and Drug Administration (FDA)approved drugs have been reported. In the first study, Madrid and co-workers performed in vitro and in vivo (in mice) screening of 1012 FDA-approved drugs and selected 24 candidate entry inhibitors for Ebola virus 2 . In the second study, 53 inhibitors of Ebola virus infection with IC 50 < 10 µM and selectivity index SI > 10-fold have been identified by in vitro screening of 2816 FDA-approved drugs 3 . In the same study, an additional 95 drugs which are active against Ebola virus infection with IC 50 > 10 µM and SI <10-fold were also reported.
Although in vitro and in vivo screening for repurposing/repositioning of existing drugs could significantly accelerate discovery of new drugs these approaches are time-consuming and costly for screening of large drug libraries. Recently, we proposed a novel approach for in silico screening of molecular libraries for drug candidates 4-8 . This approach, which uses the average quasi valence number (AQVN) and the electron-ion interaction potential (EIIP), parameters determining long-range interaction between biological molecules, might hold a key to overcoming some of these obstacles in experimental screening by significantly reducing the number of compounds which should be in vitro and in vivo tested 9 . Drug repurposing screen to identify active compounds that block Ebola entry Specific recognition and targeting between interacting biological molecules at distances > 5Å are determined by the average AQVN and the EIIP 10 , which are derived from the general model pseudopotential 11,12 . These parameters for organic molecules are determined by the following simple equations 10 : Where Z* is the average quasi-valence number (AQVN) determined by where Z i is the valence number of the i-th atomic component, n i is the number of atoms of the i-th component, m is the number of atomic components in the molecule, and N is the total number of atoms. EIIP values calculated according to equation 1 and equation 2 are expressed in Rydberg units (Ry).
Among 3300 currently used molecular descriptors, AQVN and EIIP represent the unique physical properties which characterize the long-range interactions between biological molecules 10 . Small molecules with similar AQVN and EIIP values interact with the common therapeutic target, which allow establish criterions for virtual screening of molecular libraries for compounds with similar therapeutic properties 4-9 . Here we develop the EIIP/AQVN-based criterion for virtual screening of molecular libraries for candidate drugs against Ebola virus infection.

Results and discussion
Previously   (Table 1). Madrid and co-workers selected 24 drugs by in vitro screening of 1012 FDA-approved drugs, which are effective against Ebola virus infection 2 . They also showed that among these compounds, four antimalarial drugs (chloroquine, hydroxychloroquine, amodiaquine and aminoquinoline-13) also are effective against Ebola virus infection in vivo 2 . Among 53 compounds which effectively inhibit Ebola virus infection in vitro, which Kouznetsova and co-workers selected from 2816 approved drugs, are also three antimalarial drugs (mefloquione, chloroquine, amodiaquine) 3 . It was also suggested that application of chloroquine for prevention of virus transmission should be considered because this compound significantly inhibits Ebola virus infection 13 . Our analysis showed that 15 of 22 approved ant-malarial drugs (http://en.wikipedia. org/wiki/Antimalarial_medication) are located in EVIIS (Table 3). Six 2-alkylquinolines have been also included in this study. This chemical series is promising as some derivatives exhibited antiviral activity such as 2PQ, and 2QQ 32,33 antimalarial activity such as 2PQ and 2PentQ2 34 , antileishmanial activity such as 2PQ 35,36 and neurotrophin-like activity on dopaminergic neurons such as 2QI15 37 . These compounds exhibit some advantages in regard to their chemical synthesis with few steps and good yields as well as their chemical stability in tropical conditions of storage. Their combined effects against virus and Leishmania parasites suggested they could be an advantage for the treatment of Leishmania/HIV co-infections and they were considered as attractive enough to enter the pipeline of DNDi on 2010.
All these data strongly suggest that this class of drugs should be further investigated as a promising source of therapeutics for treatment of EVD. Anti-malarial drugs with dual activity should be of special interest because malaria represents the highest health-related disease in African countries with EVD.
Among 3828 FDA-approved drugs screened for anti-Ebola activity were six antibiotics which inhibit Ebola virus infection (azthromycin, erythromycin, spiramycin, dirithromycin, maduramicin, charitromycin) 2,3 . All these antibiotics are within EVIIS and four of them are in cEVIIS. Analysis of 184 approved antibiotics (Dataset 4) showed that only 32 (17.4%) have AQVN and EIIP values in EVIIS, and that 11 of them are located within cEVIIS. Previously we reported domains of AQVN and EIIP which characterize different classes of antibiotics (Table 4) 6 . According to these data, among antibiotics some macrolides, pleuromutilins and aminoglycosides have the highest chance for inhibition of Ebola virus infection. Of note is that five of six antibiotics with experimentally proved activity against Ebola virus infection (azthromycin, erythromycin, spiramycin, dirithromycin, charitromycin) are macrolides. Antibiotics representing candidate Ebola virus infection inhibitors selected by EIIP/AQVN criterion are given in Table 5.  Previous, we determined AQVN and EIIP domains characterizing different classes of anti-HIV drugs 4-9 . As can be seen in Table 6, the EIIP/AQVN domain of CCR5 HIV entry inhibitors is within EVIIS, and domains of CXCR4 HIV entry inhibitors and HIV protease inhibitors partially overlaps EVIIS. The EIIP/AQVN domains of other classes of anti-HIV agents are located outside EVIIS. This indicates that some HIV entry inhibitors and HIV protease inhibitors could also be effective drugs against Ebola virus infection.
In conclusion, the presented results show that the EIIP/AQVN criterion can be used as an efficient filter in virtual screening of molecular libraries for candidate inhibitors of Ebola virus infection. Approved (Dataset 2) and experimental drugs (Dataset 3), anti-malarial drugs (Table 3) and antibiotics (Table 5) selected by this criterion Author contributions Conceived and designed the study: VV SG NV. Developed the analysis tools: VP. Analyzed the data: VV SG NV DRB PPML BF DPC. Wrote the paper: VV DRB PPML.

Competing interests
No competing interests were disclosed.

Grant information
This work was supported by the Ministry of Education, Science and Technological Development of the Republic of Serbia (Grant no. 173001).

Patrick Butaye
School of Veterinary Medicine, Ross University, Basseterre, Saint Kitts and Nevis This manuscript deals with the analysis of molecules for their activity against Ebola Virus (EBV). in silico They started from a reference library of compounds who have previously demonstrated and/or in vitro in activity against EBV and analyzed these compounds by the determination of their EIIP and AQVN. vivo With these data, they scanned a larger collection of compounds with unknown activity against EBV and selected possible candidates to test for their activity and/or . This is a straight forward in vitro in vivo manuscript however it may be better structured. The part of the M&M dealing with the EIIP and AQVN is more appropriate to go into the introduction since this is background information of the calculation. For clarity of the manuscript is also better to separate the results section as it is difficult to follow now. Start with the analysis of the compounds with known activity, the two datasets, and then proceed with results from the unknown dataset. Then in the discussion, the different products of interest may be evaluated. This will largely increase the readability. Upon the antibiotics, it would be good to elaborate a bit on how they work on EBV, since common sense tells that antibiotics do not work on viruses. A better explanation on how these products may interact and inhibit/kill EBV would also be good.

I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.
No competing interests were disclosed. Competing Interests: