A Case for Montelukast in COVID-19: "The use of Computational Docking to estimate the effects of Montelukast on potential viral main protease catalytic site"

Salman Mansoor (  salmanmansoor.dr@gmail.com ) Department of Neurology, Sligo University Hospital, Sligo, Ireland https://orcid.org/0000-0001-55155972 Shoab Saadat Department of Nephrology, Mid-essex Hospitals, NHS Trust, United Kingdom Aitzaz Amin Department of Medicine, Sligo University Hospital, Sligo, Ireland Imran Ali Department of Rheumatology, Sligo University Hospital, Sligo, Ireland Muhammad Tauseef Ghaffar Department of Medicine, Sligo University Hospital, Sligo, Ireland Usman Amin Department of Rheumatology, Waterford University Hospital, Waterford, Ireland Mohsin Mukhtar Department of Endocrinology, St.Luke’s Regional Hospital, Kilkenny, Ireland


Introduction
Montelukast has been used widely in treatment of Asthmatic patients since 1998 in the United States 1 .
The mechanism of action of Montelukast has been due to its antagonistic effects on the leukotriene D4 receptors which relaxes smooth muscles in lungs and reduces in ammation 1 .

COVID-19 Pandemic:
World was taken by surprise when a novel Coronavirus related illness broke out in late 2019 in Wuhan province of China and later was declared as a Pandemic due to its global impact by World Health Organization (WHO) 2 . Since then the illness has impacted the world which have been unprecedented with over a million people who have been affected. The mortality rate was reported to be 4% in China but it is observed to vary in different countries 3 .
Immune Mechanism: The current literature in relation to the immunological response in SARS-CoV-2 is scarce. However, studies done in the past on Coronoviruses (CoVs) in particular SARS-CoV and MERS-CoV do bridge this knowledge gap. The host immune system after the viral entry recognizes this by pattern recognition receptors (PRRs) that include Toll like receptors (TLR), C-type lectin-like receptors, RIG-1 like receptors (RLR) AND NOD-like receptor (NLR) 4 . There are different pathways through which the virus induces expression of in ammatory markers that limits spread and accelerates macrophage phagocytosis of the viral antigens 3,4 . The viral N-Protein of SARS-CoV is linked to protect the virus against this immune assault 4 . CD4+T and CD8+ T cells play an important role in the host adaptive immune response whereby CD4+T cells stimulate the B-cells to in producing antibodies while CD8+T cells directly kill the virusinfected cells. The free radical injury to various organs is as a result of the hyper-exaggerated immune response specially lungs and may result in multi-organ failure and death 5 . This immune cascade is also termed as "cytokine storm".
The main difference from autopsy ndings from other related diseases with cytokine storm was destruction of lymphoid tissue in SARS-COV-2 6 . Low CD4+T cells and CD8+T cells on immunohistochemical staining were observed 6 . The characteristic nding in the lungs showed diffuse alveolar damage and in ltration of monocytes and macrophages 6 .
Cytokine Storm: Cytokine storm was the term rst used in 1993 in an article on graft-versus host disease but since 2005 after avian H5N1 in uenza it has been widely used 7 . It has also been reported with infections related to cytomegalovirus, Epstein-Barr virus-associated hemophagocytic lymphohistiocytosis, group A streptococcus, in uenza virus, variola virus, and severe acute respiratory syndrome coronavirus (SARS-CoV) and SARS-CoV-2 8-12 .
The SARS-cov-2 iS known to activate the innate and adaptive immune response by binding to ACE2 receptor 13 . The underlying systemic response in cytokine storm results in release of TNFα, IL-1β, IL-2, IL-6, IFNα, IFNβ, IFNγ, and MCP-1 which result in production of free radicals from the immune cells causing hyper-permeability, end-organ damage and acute respiratory distress syndrome (ARDS) 14 .

Montelukast Mechanism of Action:
Montelukast is proposed to potentiate its effects through two probable mechanisms: 1. Its effect in preventing the cytokine storm by modulating the immune response Montelukast acts on cysteinyl leukotriene 1 (CysLT1) receptor inhibiting tracheal contraction by inhibiting leukotriene D4 and E4 15,16 . Leukotrienes are proin ammatory cytokines which that are produced as a result of 5-Lipoxygenase pathway (5-LO) pathway 17 . The receptors for Leukotrienes are found in airways which bring about changes in relation to in ammation and spasm. Montelukast binds to these receptor sites to inhibit the action of proin ammatory cytokines 18,19,20 .
In a septic shock model Montelukast was found to reduce the mortality by reducing the levels of TNF-α and IL-6 levels 21 . It increased the levels of lipid peroxidase and myeloperoxidase activity in lungs, liver and kidney 21 . This in turn protected against multiple organ failure secondary to systemic in ammatory response 21 .
The high dose administration of Montelukast was also found in asthmatic patients to mediate the levels of proin ammatory cytokines IL-6, TNFα and MCP-1 through the inhibition of NF-B activation 22 .
2. The direct effect on the virus which is explored using the computational docking.
Montelukast Structure: The molecular formula for Montelukast is C 35 H 36 ClNO 3 S it belongs the quinolone group of compounds, monocarboxylic acid and an aliphatic sulphide as shown in Figure. Computational Docking on potential viral main protease inhibitor site: Background: During the current SARS-CoV-2 pandemic molecular docking techniques have taken a pivotal role in generating potential treatments. In order to nd the most suitable drug for a given receptor, we usually use a computation method called docking by virtue of which we get to position a given drug in a given receptor site and calculate the binding a nities between the drug and the receptor pocket.

Methodology
The drug is initially prepared by adding the necessary hydrogen ions and electrostatic charges before placing it in the receptor pocket. A similar preparation is performed on the receptor pocket as well. Once prepared, the computer presents the drug to the pocket in several different 3 dimensional conformations and orientations using a probability based method (Monte Carlo procedure) and for each conformation, the binding a nities between the drug and receptor pocket are calculated using a selected force-eld rules. The drug conformation with the highest binding a nity (or the lowest mfScore) is selected as the best ligand conformation.
In this study, we used molecular docking to estimate the direct effects of Montelukast on the main protease (Mpro) inhibitor site of the SARS-CoV-2. While other studies have been performed on the homology models, we obtained the Mpro crystalized structure, A-chain (304 amino acid residues) from protein data bank (PDB code 5REK) for this analysis 24 . This protein is essential for processing of polyproteins translated from the main viral RNA and acts on at least 11 cleavage sites on large polyprotein 1AB inhibiting its activity is most likely to block the viral replication 25 .
We then used computational methods (Internal Coordinate Mechanics pro version 3.7) to identify the most suitable drug binding pockets. Based on a transformation of the Lennard-jones potential by convolution with a gaussian kernel of a certain size, a grid map of binding potentials and construction of equipotential surfaces along the maps was carried out subsequently identifying the position and size of potential ligand binding pockets 26,27 .
There were a total of three candidate pockets along the protein surface which were identi ed. The selected pocket had a volume of 284 Å 3 and a surface area of 347 Å 2 . The hydrophobicity ratio score was 0.43 with 1 being the maximum. By using biased probability Monte Carlo (BPMC) procedure which randomly selects a ligand conformation in the internal coordinate space and makes a step to a new random position independent of the previous one. The selected pocket happened to contain the most important residues (H41 and C145) involved in the active catalytic process of Mpro 28 .

Results
Montelukast was docked on to the selected pocket. The lowest potential of mean force score (mfScore) came out to be -71.68 with a hydrophobicity score of -7.9. Montelukast in docked position was neighboured by T24, T24, T26, S46, G143, H41, E166, Q189, T25,P168, N142, C145, M165, L27, T45 in the ascending order of distance from the ligand. Montelukast was seen to be making multiple hydrogen bonds with the neighbouring residues (T24, T24, T26, S46) with the closest bond with T24 (Distance= 1.71 angstrom). Important nding was its hydrogen bond with H41 and hydrophobic interactions with C145 as these residues for important members of the active catalytics site 28 . Figure.  Montelukast docked into the receptor pocket of main protease (The Closest hydrogen bound between Montelukast and nearby residue has a distance of 1.71 A)