In silico analysis suggests repurposing of ibuprofen for prevention and treatment of EBOLA virus disease

Introduction

The recent Ebola virus outbreak in West Africa caused (as of April 8, 2015) a total of 25,556 confirmed cases, including 10,587 deaths (World Health Organization, Ebola data and statistics). Although reports of new, confirmed cases of Ebola seemed to decrease in April of 2015 to approximately 30 cases per week, medical and aid organizations are warning that the current crisis is not over. Furthermore, there remains a risk for future epidemics. Therefore, there is an urgent need to develop new preventive and therapeutic approaches that can be rapidly utilized. New drugs for the treatment of EVD should be safe, efficacious, easy to manufacture and inexpensive in order to be successfully deployed in African countries. In contrast to the significant progress recently achieved in development of an effective Ebola virus vaccine¹, therapeutic options are still limited². The main obstacle represents identification of an appropriate therapeutic target, which has been largely hampered by the time and money-consuming development of new drugs, especially for neglected diseases. Furthermore, the registration of newly identified drugs, like favipiravir, for potential usage in EVD patients takes a relatively long period of time, even though it has been shown to be effective in non-human primate models. In order to avoid these obstacles, several approaches for repurposing of approved drugs for the treatment of EVD patients have been proposed in literature^3–7.

Recently, we proposed a relatively simple theoretical criterion for the fast virtual screening of molecular libraries for candidate inhibitors of Ebola virus infection⁸. Using this criterion, which is based on calculation of the average quasi-valence number (AQVN) and the electron-ion interaction potential (EIIP) - parameters determining long-range interaction between biological molecules^9–13 - we selected 267 approved and 382 experimental drugs as candidates for treatment of EVD. Further detailed analysis of these drugs, including molecular docking, revealed ibuprofen as an inexpensive, widely accessible and minimally toxic candidate for potential prevention and treatment of EVD. The molecular mechanism underlying the possible protective effect of ibuprofen against EVD is suggested.

Material and methods

Results and discussion

We screened the CEDD library¹⁵ for optimal candidates for prevention and treatment of EVD. To achieve this end, we used the following criteria: a) high efficacy of binding of drug to the Ebola virus glycoprotein GP1; b) low toxicity and accessibility (nonprescription drugs); and c) low cost. By mining of CEDD using these criteria, we selected ibuprofen as the best candidate. Ibuprofen is a nonsteroidal anti-inflammatory drug (NSAID). Its mode of action, as for other NSAIDs, is not completely understood, but may be related to prostaglandin synthetase inhibition. It achieves this effect on prostaglandin synthesis by inhibiting cyclooxygenase (COX I and II isotypes), an enzyme that is present in at least one isoform most of the tissues of the body and which is responsible for production of not only prostaglandins but also prostacyclins and tromboxane.

Generally, ibuprofen is minimally toxic drug. A meta-analysis of eight placebo-controlled studies showed that application of non-prescription doses of ibuprofen (800–1200 mg/day) over 10 days caused no significant adverse events¹⁹. Similar results were obtained in a prospective study of ibuprofen use in healthy volunteers taking the maximum permitted, non-prescription dose of ibuprofen (1200 mg/day) for 10 days²⁰. It was demonstrated that administration of ibuprofen in children is safe. A randomized, community-based study on the safety of ibuprofen in febrile children aged 2 years old (5 or 10 mg/kg) showed that the risk of hospitalization for gastrointestinal bleeding associated with this drug was 17 per 100,000²¹. It was also shown that fecal blood loss associated with application of prescription doses of ibuprofen (2400 mg/day) did not exceed the normal range²². Even in the case of post-surgical usage for pain control in children undergoing tonsillectomy, ibuprofen administration is considered to be safe and does not increase risk of hemorrhages²³. Accordingly, we believe that it would be crucial to better understand the potential risk and benefits of ibuprofen usage in experimental models of EVD. Taken together, these data indicate that ibuprofen could be safely used in nonprescription doses in EVD patients, with potential antiviral effects as well as to alleviate the symptoms.

To assess specificity of ibuprofen as a potential inhibitor of the Ebola virus infection, distribution of all approved NSAIDs in AQVN/EIIP space was analyzed. Results presented in (Table 1 and Figure 1) show that only ibuprofen and its isomer dexibiprofen are located within the domain of the AQVN/EIIP space. Interestingly, most of the experimentally verified inhibitors of the Ebola virus infection are located in the same space (e.g. chloroquine, amodiaquine, brincidofovir, etc)⁸. This suggests that ibuprofen and dexibuprofen are the only drugs from the NSAID group that potentially have anti-Ebolavirus effects, which should be tested both in vitro and in vivo.

Figure 1. Distribution of nonsteroidal anti-inflamatory drugs in the AQVN/EIIP space.

Domain of inhibitors of Ebola virus infection⁸ is marked in red.

A previous in silico study suggested that Elastin Microfibril Interface Located Proteins (EMILINs) are involved in interaction between GP1 and endothelial extracellular matrix (ECM)¹⁶. Docking of ibuprofen to the GP1 model gave conformations which bound to EMILINs binding domain on GP1¹⁶. The binding site of ibuprofen on GP1 spans the edge of this region and consists of Thr 338, Ser 340, Gln 344 and Ala 415¹⁶. The intermolecular interactions between ibuprofen and binding site amino-acids include hydrogen bonds of carboxyl group with Thr 338, Ser 340 and Gln 334. Additional stabilization is provided through aromatic and hydrophobic interaction with Ala 415. This binding site is placed between two loops, which provide the possibility of stabilizing a particular conformation, and therefore possibly blocking receptors. The appropriately high binding energy of -9.0 kcal/mol favors this assumption. The binding conformation is presented in Figure 2. At this stage it can be hypothesized that ibuprofen prevents interaction between Ebola virus and ECM by blocking the interaction between GP1 and EMILIN. There are some literature data that support our current hypothesis. EMILIN-1 is a glycoprotein expressed in the vascular tree that binds to the TGF-β1 precursor and prevents its processing by cellular protease furin²⁴. It was shown that Emilin-1 knockout mice display increased TGF-β1 signaling in the walls of their blood vessels, leading to peripheral vasoconstriction and arterial hypertension²⁵. These matrix-dependent changes in the vascular hemodynamics caused by TGF-β1 and EMILIN-1 are important because they ultimately affect the cardiovascular morbidity and mortality rates. Recently, it was shown that activation of the TGF-β1 signaling pathway by Ebola virus plays an important role in pathogenesis of EVD²⁶. These findings suggest the possibility that binding of GP1 to EMILIN-1 prevents its interaction with TGF-β1, which results in activation of TGF-β1 signaling pathway. Binding of ibuprofen to GP1 could prevent GP1/EMILIN-1 interaction allowing EMILIN-1 to keep control of TGF-β1 signaling pathway.

Figure 2. Ibuprofen docked to GP1 with marked amino-acid residues.

Green dotted lines: hydrogen bonds; grey: hydrophobic interactions.

In conclusion, presented results should encourage further investigation of ibuprofen and ibuprofen-inspired drugs as inexpensive, low-toxic and wide-accessible candidates for prevention and its usage in the treatment of EVD.

Data availability

F1000Research: Dataset 2. Approved and experimental drugs selected as candidate for treatment of EVD, 10.5256/f1000research.6110.d42877¹⁵