biomolecules Article Computational Identification of Potential Anti-Inflammatory Natural Compounds Targeting the p38 Mitogen-Activated Protein Kinase (MAPK): Implications for COVID-19-Induced Cytokine Storm Seth O. Asiedu 1 , Samuel K. Kwofie 2,3,* , Emmanuel Broni 2 and Michael D. Wilson 1,4 1 Department of Parasitology, Noguchi Memorial Institute for Medical Research, College of Health Sciences, University of Ghana, Legon, Accra P.O. Box LG 581, Ghana; soasiedu@noguchi.ug.edu.gh (S.O.A); MWilson@noguchi.ug.edu.gh (M.D.W) 2 Department of Biomedical Engineering, School of Engineering Sciences, College of Basic and Applied Sciences, University of Ghana, Legon, Accra P.O. Box LG 77, Ghana; ebroni002@st.ug.edu.gh 3 West African Centre for Cell Biology of Infectious Pathogens, Department of Biochemistry, Cell and Molecular Biology, College of Basic and Applied Sciences, University of Ghana, Accra P.O. Box LG 54, Ghana 4 Department of Medicine, Loyola University Medical Center, Maywood, IL 60153, USA * Correspondence: skkwofie@ug.edu.gh   Abstract: Severely ill coronavirus disease 2019 (COVID-19) patients show elevated concentrations Citation: Asiedu, S.O.; Kwofie, S.K.; Broni, E.; Wilson, M.D. of pro-inflammatory cytokines, a situation commonly known as a cytokine storm. The p38 MAPK Computational Identification of receptor is considered a plausible therapeutic target because of its involvement in the platelet Potential Anti-Inflammatory Natural activation processes leading to inflammation. This study aimed to identify potential natural product- Compounds Targeting the p38 derived inhibitory molecules against the p38α MAPK receptor to mitigate the eliciting of pro- Mitogen-Activated Protein Kinase inflammatory cytokines using computational techniques. The 3D X-ray structure of the receptor (MAPK): Implications for with PDB ID 3ZS5 was energy minimized using GROMACS and used for molecular docking via COVID-19-Induced Cytokine Storm. AutoDock Vina. The molecular docking was validated with an acceptable area under the curve Biomolecules 2021, 11, 653. (AUC) of 0.704, which was computed from the receiver operating characteristic (ROC) curve. A https://doi.org/10.3390/ compendium of 38,271 natural products originating from Africa and China together with eleven biom11050653 known p38 MAPK inhibitors were screened against the receptor. Four potential lead compounds ZINC1691180, ZINC5519433, ZINC4520996 and ZINC5733756 were identified. The compounds Academic Editor: José L. Medina-Franco formed strong intermolecular bonds with critical residues Val38, Ala51, Lys53, Thr106, Leu108, Met109 and Phe169. Additionally, they exhibited appreciably low binding energies which were Received: 27 February 2021 corroborated via molecular mechanics Poisson–Boltzmann surface area (MM-PBSA) calculations. Accepted: 26 April 2021 The compounds were also predicted to have plausible pharmacological profiles with insignificant Published: 29 April 2021 toxicity. The molecules were also predicted to be anti-inflammatory, kinase inhibitors, antiviral, platelet aggregation inhibitors, and immunosuppressive, with probable activity (Pa) greater than Publisher’s Note: MDPI stays neutral probable inactivity (Pi). ZINC5733756 is structurally similar to estradiol with a Tanimoto coefficient with regard to jurisdictional claims in value of 0.73, which exhibits anti-inflammatory activity by targeting the activation of Nrf2. Similarly, published maps and institutional affil- ZINC1691180 has been reported to elicit anti-inflammatory activity in vitro. The compounds may iations. serve as scaffolds for the design of potential biotherapeutic molecules against the cytokine storm associated with COVID-19. Keywords: COVID-19; coronavirus; p38 MAPK; cytokine storm; anti-inflammatory compounds; Copyright: © 2021 by the authors. natural products; molecular dynamics simulation; molecular docking Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons 1. Introduction Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ Coronavirus disease 2019 (COVID-19) is caused by severe acute respiratory syndrome 4.0/). coronavirus 2 (SARS-CoV-2) [1,2]. It is a positive-sense single-stranded RNA virus [3] Biomolecules 2021, 11, 653. https://doi.org/10.3390/biom11050653 https://www.mdpi.com/journal/biomolecules Biomolecules 2021, 11, 653 2 of 25 belonging to the Coronaviridae family. They are uniquely enveloped non-segmented viruses with large surface spike proteins observed as corona (crown-like) projections using elec- tron microscopy [4]. The virus is transmitted through droplets, contacts, fecal-oral, blood, mother-to-child, airborne, fomite, and animals to humans [5–8]. COVID-19 symptoms include fever, dry cough, runny nose, sore throat, loss of taste and smell, and difficulty in breathing [9]. During critical stages, complications observed include acute respira- tory distress syndrome (ARDS), pneumonia, septic shock, arrhythmia, and disseminated intravascular coagulation [9–11]. Several studies on these complications have reported elevated levels of plasma cytokines [12], a condition referred to as cytokine storm [13]. The p38 mitogen-activated protein kinase (MAPK) is critical in COVID-19 cytokine storms [14,15]. In the human host, angiotensin II (AngII), a pro-inflammatory peptide hormone, mediates its effects through p38 MAPK activation [14,16]. Angiotensin-converting enzyme 2 (ACE2) converts (Ang II) into angiotensin 1–7 (Ang 1–7), which binds to the Mas receptor. This counterbalances the pro-inflammatory effects of Ang II by decreasing the activation of p38 MAPK [17]. Upon cell entry, SARS-CoV-2 binds and downregulates ACE2 [18,19]. The loss of ACE2 activity upon viral entry allows for unbridled inflammation [14]. The p38 MAPK is implicated in the propagation of the SARS-CoV-2 lifecycle [14] and this results from enhanced replication due to increased MAPK activity [15]. Moreover, SARS-CoV, a close neighbor of the SARS- CoV-2 virus, expresses a protein that directly upregulates p38 MAPK in vitro [20] and this process seems to be a pathogenic step in the lifecycle of many RNA respiratory viruses [21]. A study of the effect of p38 MAPK inhibitors on SARS-CoV infected mice reported an 80% increase in survival after treatment [22]. The p38 MAPK is a plausible anti-inflammatory drug target for COVID-19 patients and its inhibitors have been trialed for the treatment of other ailments [23–25]. Anti-inflammatory compounds including dexamethasone are used as treatment op- tions in COVID-19 patients [26], since no FDA-approved drugs specific for the treatment and prevention of COVID-19 exist. There are over 2300 clinical trials per the Global Coron- avirus COVID-19 Clinical Trial Tracker (https://www.covid-trials.org/ (accessed on 28 September 2020)) [27] on novel vaccines, drugs, and repurposed compounds including p38 MAPK inhibitor Losmapimod. The use of computational methods has been suggested as a viable alternative for the faster and cheaper discovery of COVID-19 therapeutic molecules [28–33]. Natural products are used as COVID-19 remedies [34–37]. This is due to their structural and chemical diversity which serves as a potent source of novel drug-like compounds [38,39]. Freely available cheminformatics-based natural compound databases such as the Traditional Chinese Database [40], African Natural Product Database [41], and the North African Natural Product Database [42] collectively contain over 30,000 unique natural products. Consequently, the work aims at finding potential natural product-derived inhibitory molecules of p38 MAPK with the propensity to ameliorate the cytokine storm in severely ill COVID-19 patients. This study utilized computational techniques including molecu- lar docking, dynamics simulations, and characterization of binding mechanisms using Molecular mechanics Poisson-Boltzmann surface area (MM-PBSA) calculations. It further predicted the pharmacological profiles and biological activity of drug-like and lead-like anti-inflammatory molecules as potential inhibitors of p38 MAPK. 2. Materials and Methods 2.1. Protein Structure Retrieval and Processing The experimentally solved 3D structure of p38α MAPK was retrieved from the Protein Data Bank [43,44] with PDB ID: 3ZS5 [45]. Missing residues at 172–183 were initially added via SwissModel [46]. PyMOL v 4.0.0 [47,48] was used for preprocessing the 3ZS5 protein structure by removing water molecules and other ligands. Molecular dynamics (MD) simulation was performed using the Groningen Machine for Chemical Simulations GROMACS version 2018 [49]. Optimized Potentials for Liquid Simulations (OPLS)/All Biomolecules 2021, 11, 653 3 of 25 Atom (AA) force field was used to generate the protein topology and position restrain files [50,51]. Periodic Boundary Conditions (PBC) were applied to the structure with the protein centered 1 nm from the edge of a cubic box to monitor the movement of all particles and avoid edge effects on the surface atoms [52]. The system was solvated with SPC water [53,54] and neutralized with six Na atoms. The structure was energy minimized using the steepest descent algorithm at 50,000 steps. 2.2. Screening Library of Compounds A library of 38,271 natural product compounds comprising 2277 compounds from the North African Natural Product Database [42], 833 compounds from the African Nat- ural Product Database [41], and 35161 compounds from the Chinese Natural Product Database [40] was retrieved. The compounds were filtered using DataWarrior [55] with molecular weights between 250 g/mol and 350 g/mol [56,57]. Further, compounds pre- dicted as mutagenic, causing reproductive effects, irritant, and tumorigenic were eliminated via DataWarrior. Known p38 MAPK inhibitors used as controls for the molecular dock- ing were Losmapimod, ARRY-797, Dilmapimod, Doramapimod, Pamapimod, PH-797804, SB 202190, SB 203580, Talmapimod, VX-702, and Skepinone-L with PubChem compound IDs 11552706, 46883775, 10297982, 156422, 16220188, 22049997, 5169, 176155, 9871074, 10341154, and 45279963, respectively. 2.3. Validation of Docking Protocol The easyROC [58] was used to generate the receiver operating characteristic (ROC) curve. The ROC curve was used to validate the ability of Autodock Vina to discriminate between active and decoy compounds. The SMILES of the eleven inhibitors served as inputs for the generation of 50 decoys each via Directory of Useful Decoys (DUD-E) [59]. Each decoy has similar physicochemical properties to a known inhibitor but is chemi- cally distinct [59]. A total of 561 compounds comprising 11 actives and 550 decoys were screened against the p38 MAPK structure. The 11 actives comprised the known inhibitors of p38 MAPK. 2.4. Virtual Screening of Ligands AutoDock Vina [60] was used to screen the library against the p38α MAPK protein structure. The pre-filtered library was imported into OpenBabel [61] and minimized using the Universal Force Field (Uff) for 200 steps and optimized using the conjugate gradient. A grid box of dimensions (78.35, 47.84, 49.86) Å and center (9.97, 30.65, 20.29) Å was used for docking with a default exhaustiveness of eight. 2.5. Pharmacological Profiling The pharmacokinetic and physicochemical profiles of the compounds were predicted via SwissADME [62] with the Simplified Molecular Input Line-Entry System (SMILES) of the compounds as inputs. 2.6. Elucidation of the Protein-Ligand Interactions LigPlot+ [63] was used to generate the 2D protein-ligand interactions with default settings. The best poses of the hits were saved in “.pdb” file formats and then visualized using PyMOL. 2.7. Prediction of Biological Activities of Hit Compounds The biological activities of the hits were predicted using the Bayesian-based Predic- tion of Activity Spectra for Substances (PASS) [64]. The SMILES files of the compounds were used to perform structural similarity searches for antiviral and anti-inflammatory compounds with a DrugBank [65] similarity threshold of 0.7. Biomolecules 2021, 11, x FOR PEER REVIEW 4 of 26 2.7. Prediction of Biological Activities of Hit Compounds The biological activities of the hits were predicted using the Bayesian-based Predic- tion of Activity Spectra for Substances (PASS) [64]. The SMILES files of the compounds were used to perform structural similarity searches for antiviral and anti-inflammatory Biomolecules 2021, 11, 653 compounds with a DrugBank [65] similarity threshold of 0.7. 4 of 25 2.8. Molecular Dynamics Simulation of Protein-Ligand Complexes 2.8. MMoDle sciumlaurlDatyionnams iocfs tShiem purlaottieoinn-olfigParontdei nco-LmigpalnedxeCso wmeprleex peserformed using GROMACS 2020 vMerDsiosinm [u49la].t iTohnes 3oZf Sth5e pprorotetienin t-olpigoalnogdiecos mwperleex geesnwerearteedp eursfionrgm tehde uGsRinOgMGORSO9M6 4A3Ca1S f2o0rc2e0 fvieerldsi o[6n6[]4 i9n] .GTRhOe 3MZAS5CpS,r owtehienretoaps otlhoeg ileigsawnder efilgeesn wereartee dgeunseinragtethde vGiaR tOhMe PORSO96D4R3Ga1 [f6o7r]c. eEfiaeclhd c[o66m]pinleGx RwOaMs sAoClvSa,twedh ewreiaths twheatleigr amndolfiecleuslewse irne ag ecnuebriact ebdoxv ioaft hsiezPe R5O.0D nRmG a[n67d] . nEeaucthracliozmedp wlexithw saisx sNoalv iaotnesd. Ewnitehrgwy amteirnimmoizleactiuolne sofin eaachcu cboimc pbloexx owfassi zceon5d.0ucntmed afnodr 5n0e,0u0t0ra sltiezpeds uwsiinthg stihxeN staeeiopness.t dEensecregnyt malginoirmithizmat.i oTnheo lfigeaacnhdsc owmerpel erxeswtraasinceodn dbeufcotreed thfoer c5a0n,o0n00icastl eepnssuemsinbgle,t haendst etehpeens tthdee sicseonthtearlmgoarl-iitshomba. rTich eenlsigemanbdles. wEeqrueilriebsrtartaiionne dofb eefaocrhe ctohme pcalenxo wniacsa lpeenrsfoermmbeled, faonrd 10th0e pns tahpeieiscoet ahnerdm fianla-ils MobDar sicimenuslaetmiobnl ew. aEsq cuoilnibdruacttieodn foofre 1a0c0h ncso mwipthle xtimwea sstpeeprsf oorfm 2e fds fuonrd1e0r0 PpMs Eap. iTehcee afrnede fibinnadliMngD ensiemrguielast iwonerwe acosmcopnudteudct eudsinfogr G1_00mnmspwbistah [t6im8]e. Tstheep sbionfd2infsgu fnredee renPeMrgEy. Tchonetfrriebeubtiionnd ipnegr erneesridgiuees wwaerse ccaolcmuplautteedd uussiinngg MGM_m-PmBpSAbs aan[d68 t]h. eT ohuetbpiuntd pinlogtsf rgeeeneenraertegdy wcointhtr Rib. ution per residue was calculated using MM-PBSA and the output plots generated with R. 33. .RReessuultlsts aanndd DDisisccuussssioionn 33.1.1. .PPrrooteteinin SStrtruuccttuurree RReettrrieievvaal laanndd AAnnaalylyssisis TThhee PPDDBB ddaattaabbaassee ccoonnttaaiinnss sseevveerraal lssoollvveedd pp3388αα MAPPKK ssttrruuccttuurreess. .TThhee pp3388αα iiss tthhee mmoosst teexxpprreesssseedd isisooffoorrmm ooff tthhee pp3388 MAPPKK [[6699]].. TThhee sseelleecctteedd pprrootteeiinn ssttrruuccttuurree wiitthh PDBB IIDD: :33ZZSS55 wwaass ssoolvlveedd uussiningg xx--rraayy ddiiffffrraaccttiioonn aatt aa hhiigghh rreessoolluuttiioonn ooff 11..66 Å.. TThhiiss eennssuurreedd tthhee ddeettaailieledd rreepprreesseennttaattiioonn ooff tthhee waatteerr--meeddiiaatteedd hhyyddrrooggeenn--bboonndiinngg nneettwoorrkk [[4455]].. 33ZZSS55 hhaass aalslsoo bbeeiningg uusesded inin a aprperveivoiuosu csocmompuptuatiaotnioanl asltusdtuyd [y70[]7 0a]ndan idt isit biosubnodu ntod tthoe tihnehiibnihtoibr iStoBr 2S0B3528003 5[4850].[ 4T5h].eT shtreusctruurcet iusr ae misoanmomoneor mcoenrscisotninsigs toinf 3g6o2f r3e6s2idrueessid auneds iasn mdaids em uapd eofu tpwof dtowmoadionms caoinmspcroimsinpgri sNin- ganNd- Can-tderCm-tienraml idnoaml daoinmsa. iTnhs.e TNh-eteNrm-teinrmali ndaolmdaoimn aisin coismcopmospeods oedf laorfglealryg eβl-yshβe-esthse, ewtsh,iwle hthile Cth-teeCrm-tienraml idnoaml daoinm isa ienssiesnetsisaellnyt ihaellyicahle [l7ic1a]l. T[7h1e] .cTathaelyctaict aAlyTtPic bAinTdPinbgin sditien giss fioteuinsdf obuentwdebeent wtheee ntwthoe dtowmoadinosm (aFingsur(Fe i1g)u. rTeh1is). sTithei siss citheairsachtearriazcetde rbizye ad pbeyptaidpe pftlipd,e rfleispu,lrteinsug ltiinn gthine tphreespernetsaetniotant ionf towf tow hoyhdyrdogroegne-bno-bnodn dodnoonros rsini nththee AATPP ppoocckkeett, ,aa cclleeaarr ddiissttiinnccttiioonn ffrroom tthhee aacccceepttoorr--ddoonnoorr--aacccceeppttoorr ppaattteerrnn nnoorrmaalllyy pprreesseenntt iinn tthhee AATTPP ppoocckkeett ooff ootthheerr pprrootteeiinn kkiinnaasseess [[4455]].. OOnnee lalarrggee bbiinnddiinngg ccaavviittyy wiitthh aa ssuurrffaaccee aarreeaa ooff 11004400.3.30099 ÅÅ22 aanndd vvoolulummee 11336655.5.5522 ÅÅ3 3wwasa spprerdedicitcetedd vviaia CCAASSTTpp 33.0.0 [7[722].] . Figure 1. A surface representation of the structure of p38 MAPK (Chain A). The catalytic ATP binding pocket is colored green. The structure is a monomer with a binding cavity of 1365.552 Å3. 3.2. Docking Protocol Validation The docking protocol of AutoDock Vina [60] was validated using the ROC curve. Eleven inhibitors with their respective decoys were screened against the p38 MAPK struc- ture to generate a ROC curve. The ROC curve evaluates the performance of AutoDock Biomolecules 2021, 11, x FOR PEER REVIEW 5 of 26 Figure 1. A surface representation of the structure of p38 MAPK (Chain A). The catalytic ATP binding pocket is colored green. The structure is a monomer with a binding cavity of 1365.552 Å3. 3.2. Docking Protocol Validation The docking protocol of AutoDock Vina [60] was validated using the ROC curve. Biomolecules 2021, 11, 653 Eleven inhibitors with their respective decoys were screened against the p38 MAPK 5storfu2c5- ture to generate a ROC curve. The ROC curve evaluates the performance of AutoDock Vina to classify inhibitors from decoys [73]. An area under the curve (AUC) value of 1 is Vcionnasitdoecrleadss pifeyrfinechti,b witohrilsef raonm AdUeCco byeslo[7w3 ]0. .A5 nsiagrneiafiuesn dpeororth celacsusrifvieca(tAioUnC [)74v]a. lAuedodfit1ioins- caolnlys,i daenr AedUpCe rvfaelcut,ew ohf i0l.e7 atno 0A.8U iCs cboenloswide0r.5eds iagcnciefipetsapboleo, r0c.8la tsos i0fi.9c aisti ovner[y7 4g]o.oAdd, danitdio anbaollvy,e a0n.9A iUs eCxvcealllueento [f705.7,7t6o].0 T.8hies ocbotnasiindeedre AdUacCc ecpotmabplue,te0d.8 ftroom0.9 thisev RerOyCg ocuodrv, ea nwdaasb 0o.7v0e40 (.9Fiigs- euxrcee l2l)e,n wt h[7i5ch,7 6is] .cTonhseidoberteadin aecdceApUtaCblceo [m75p]u. Ttehdisf rinomdictahteesR tOhaCt cAuurvtoeDwoacsk 0V.7in0a4 r(eFaigsuonreab2l)y, wdhisitcihngisucisohnessi dbeertewdeaecnc eapcttaivbele a[n7d5 ]i.nTahcitsivien dciocmatpesouthnadtsA ouf tpo3D8o MckAVPinKa. rAeaustoonDaobclky Vdiisntain i-s gmuoisrhee ascbcuetrwateee tnhaanc tAivuetoaDndocikn a4c itniv seelceocmtinpgo aunntdi-sSAofRpS3-8CoMVA-2P MK.prAo upro co tmoDpoocukndVsin [7a7i]s. Mmoorree- ac Aouv cuerrtoD, a toA e uthtoaDnck Vino A ac u wk t oDaVs ein oack w4ains seemlepcting anti-SARS-CoV-2 M cmployed succleosysefudl lysutcocsecsrsefeunllcyo mtop oscurnedesn o mpoundsagcoaminpstopu3n [ 8d 7s7 ]. MMAagPa o Ki reover, ,nwsth ipc3h8 wMeArePfKou, wndhitcohi nwheirbeit fpou38ndM tAo PinKhiinbivt iptr3o8 [M70A].PK in vitro [70]. FFiigguurree2 2..E Evvaaluluaattininggt htheep peerfroformrmaannceceo of ft htheev viritrutualasl csrcereeneinnigngv ivaiaR OROCCcu cruvrev.eB. iBnidnidnigngen eenregrigesiefsr om tfhreomsc rteheen sicnrgeeonfiinngh oibf iitnohrsibaitnodrsd aencdoy dseacgoyaisn asgt athinestp t3h8eM p3A8P MKArePcKe prteocrepwtoer ewuesred utsoedg eton egreantertahte ctuhrev ceu. rAvne.A AUnC AoUfC0. 7o0f 40.w70a4s wobatsa ionbetdaifnreodm frtohme R thOeC RcOuCrv ceu. rve. 33.3.3.. PPrree--FFiilltteerriinnggo offL Liibbrraarryya anndd MooleleccuulalarrD DoocckkininggS Sttuuddieiess AA ttoottaall ooff 77228822 oouutt ooff tthhee3 388,2,27711 nnaattuurraall pprroodduuccttss weerree oobbttaaiinneedd aafftteerr pprree--fifilltteerriinngg.. TThheesseec coommppoouunnddssh haaddm mooleleccuulalarrw weeigighhtstsb betewtweenen2 52050g /gm/moolla anndd3 35500g g//mooll,, aannddw weerree pprreeddicicteteddt otob been noonn--ttooxxicic. .T Thhee ccoomppoouunnddss weerrees sccrreeeenneedda aggaaiinnssttt thhee ATTPP-b-bininddininggp poocckkeett ooffp p3388 MAPPKKa anndd weerreer raannkkeeddb baasseeddo onnt thheeiirrb biinnddiinngge enneerrggieiess.. AuuttooDoocckkV Vininaac coommbbinineess bbootthhe emppiriricicaalls sccoorrininggf fuunnccttioionnssa annddk knnoowwleleddggee-b-baasseeddp pootetenntitaialslst otoc coommppuutetet htheeb bininddiningg eenneerrggieiesso offc coomppoouunnddss[ [6600]].. ZIINCC3388332211663311,,a ac coomppoouunndda annnnoottaatteeddi innt thhee TTCM daattaabbaassee,, had the lowest binding energy of −12.4 kcal/mol, whereas ZINC5734567 had the highest binding energy of −2.5 kcal/mol. Among the known inhibitors, SB 202190 had the least binding energy of −11.0 kcal/mol, followed by SB 203580 with −10.9 kcal/mol. Losmapi- mod that is currently undergoing clinical trial had binding energy of −9.3 kcal/mol, whilst that of Dilmapimod was −7.9 kcal/mol. In this study, a threshold of −11.0 kcal/mol was used for selecting hits. Selected compounds had lower binding energies relative to the known inhibitors. A previous study used a threshold of −10.0 kcal/mol for p38 MAPK [78]. Another docking study on the potential inhibitors of p38 MAPK identified a promising Biomolecules 2021, 11, 653 6 of 25 compound with a binding energy of −11.1 kcal/mol [79]. A total of 42 compounds and two inhibitors (SB 202190 and SB 203580) were selected for downstream analysis. 3.4. Pharmacological Profiling of Hit Compounds Pharmacological profiling studies are important aspects of the drug development pipeline. The human intestinal absorption, permeability glycoprotein (P-gp) binding, and cytochrome P450 3A4 (CYP450) inhibition were considered [80]. The human intestinal absorption measures the likelihood of absorption for orally administered drugs into the bloodstream. Two of the hits, namely ZINC85550217 and ZINC85543655 were predicted to have low GI absorption. Additionally, 16 out of the 42 compounds including ZINC38321631 were predicted as P-gp substrates with likely decrease in drug bioavailability [81,82]. CYP3A4 is responsible for the metabolism of about 50% of all drugs [83]. The two standard inhibitors (SB 202190 and SB 203580) were predicted as CYP3A4 inhibitors. A total of 17 hits were predicted to be CYP3A4 inhibitors suggesting they may interfere with the metabolism of other drugs. After eliminating compounds predicted to have poor pharmacological profiles, 18 compounds were selected as promising hits (Supplementary Table S1) with none of them violating Lipinski’s rule of five for evaluating drug-likeness. 3.5. Visualization and 2-D Representation of Protein-Ligand Interactions Hydrogen bonding and hydrophobic contacts are critical in the stabilization of a ligand within the binding pocket of a receptor [84]. All 18 selected hit compounds were observed to bind firmly in the binding pocket of the receptor. ZINC95486106 had the lowest binding energy of −12.1 kcal/mol. It formed hydrogen bonding with Met109 (bond length of 3.03 Å) and hydrophobic contacts with Val30, Ala51, Val38, Leu171, Leu108, Gly170, Lys53, Tyr35 and Phe169 (Figure 3a). Eight other compounds comprising ZINC95913720, ZINC95919076, ZINC33832090, ZINC1691180, ZINC4520996, ZINC4215683, ZINC4023706 and ZINC70454959 formed at least one hydrogen bond with the residues of the binding pocket (Table 1). ZINC4215683 formed three hydrogen bonds with residues Glu71 and Lys53. It also formed hydrophobic contacts with Gly170, Tyr35, Phe169, Val38, Leu86, Leu104, Leu171, Thr106, Leu75 and Val105. Additionally, ZINC95913720 formed a hydro- gen bond with Lys53 and hydrophobic contacts with Val30, Ala51, Val38, Leu108, Gly170, Tyr35, Phe169, and Val38 (Figure 3b). ZINC95919076 formed a hydrogen bond with Lys53 and hydrophobic contacts with Val30, Ala51, Val38, Leu108, Gly170, Tyr35, Phe169, Val138, and Met109 (Figure 3d). ZINC1691180 also formed a hydrogen bond with Tyr35 (bond length of 3.17 Å) and hydrophobic contacts with Val30, Ala51, Val38, Phe169, Val38, Met109, Glu71, Leu75, Leu104, Thr106 and Gly31. ZINC4520996 formed hydrogen bonds with Tyr35 and hydrophobic contacts with Ala51, Val38, Val30, Lys53, Gly31, Glu71, Thr106, Phe169, Leu104 and Thr106. ZINC4023706 formed three hydrogen bonds with residues Tyr35 and Gly170, and hydrophobic contacts with Phe169, Leu104, Leu75, Glu71, Thr106, Val38, Ala51, Lys53, Val20 and Gly31. ZINC33832090 had binding energy of −11.8 kcal/mol and formed hydrophobic interactions with Ala51, Val38, Gly170, Tyr35 and Phe169 (Figure 3c). It also interacted with Lys53 via hydrogen bonding of length 3.18 Å (Table 1 and Figure 3c). The known inhibitors SB 202190 and SB 203580 formed only hydrophobic interactions with Tyr35, Val38, Ala51, Lys53, Leu104, Thr106, Leu108, Met109, Phe169, Gly170, and Leu171. Residues worth noting are Ala51, Val38 and Phe169, which interacted separately with 15, 17, and 18 molecules, respectively. These three residues can be considered critical in the stabilization of ligands in the binding pocket. These critical residues also interacted with both SB 202190 and SB 203580. Previous studies identified Lys53, Leu108 and Met109 as crucial residues [45,71,85,86], whilst another identified Phe169 as important [87]. Biomolecules 2021, 11, 653 7 of 25 Table 1. Summary of the protein-ligand interactions generated with LigPlot+ for the top 18 hits and 2 inhibitors. The binding energies, number of hydrogen bonds, hydrogen bond interacting residues with their respective bond lengths, and hydrophobic contacts of the compounds are shown. ZINC ID/Drug Binding Energy Number of Hydrogen Bond Hydrogen Bond Name (kcal/mol). Hydrogen Bonds Residues Length (Å) Hydrophobic Contacts Val30, Ala51, Val38, ZINC95486106 −12.1 1 Met109 3.03 Leu171, Leu108, Gly170, Lys53, Tyr35, Phe169 Val30, Ala51, Val38, ZINC95913720 −11.8 1 Lys53 3.26 Leu108, Gly170, Tyr35, Phe169 ZINC33832090 −11.8 1 Lys53 3.18 Ala51, Val38, Gly170,Tyr35, Phe169 Val30, Ala51, Val38, ZINC95919076 −11.7 1 Lys53 2.89 Leu108, Gly170, Tyr35, Phe169, Val138, Met109 Val30, Ala51, Val38, ZINC1691180 −11.6 1 Tyr35 3.17 Phe169, Glu71, Leu75, Leu104, Thr106, Gly31 Ala51, Val38, Gly170, ZINC5519433 −11.6 Lys53, Tyr35, Phe169,Leu75, Leu86, Ile84, Leu104, Val105, Thr106 Ala51, Val38, Val30, ZINC4520996 −11.6 1 Tyr35 3.16 Lys53, Gly31, Glu71,Thr106 Phe169, Leu104, Thr106 Ala51, Val38, Gly170, ZINC1531907 −11.6 Lys53, Tyr35, Phe169,Leu75, Leu86, Ile84, Leu104, Val105, Thr106 Ala51, Val38, Gly170, ZINC4098804 −11.6 Lys53, Tyr35, Phe169,Leu75, Leu86, Ile84, Leu104, Val105, Thr106 Ala51, Val30, Gly170, ZINC95919075 −11.5 Lys53, Tyr35, Phe169, Val38, Met109, Leu108 ZINC13302897 −11.4 Ala51, Gly170, Lys53,Tyr35, Phe169, Val38, Gly170, Tyr35, Phe169, ZINC4215683 −11.2 3 Glu71, Lys53 [2] 2.81, 3.12, 2.90 Val38, Leu86, Leu104,Leu171, Thr106, Leu75, Val105 Val30, Val38, Ala51, ZINC13302884 −11.2 Thr106, Leu104, Glu71,Leu75, Lys53, Phe169, Tyr35 Leu171, Val38, Gly170, ZINC13302890 −11.2 Tyr35, Lys53, Phe169, Ala51 Phe169, Leu104, Leu75, ZINC4023706 −11.1 3 Tyr35, Gly170 [2] 3.22, 3.18, 3.17 Glu71, Thr106, Val38,Ala51, Lys53, Val20, Gly31 Biomolecules 2021, 11, 653 8 of 25 Table 1. Cont. ZINC ID/Drug Binding Energy Number of Hydrogen Bond Hydrogen Bond Name (kcal/mol). Hydrogen Bonds Residues Length (Å) Hydrophobic Contacts Leu75, Ile84, Leu104, ZINC5733756 −11.1 Lys53, Phe169, Leu171,Tyr35, Val38, Leu171, Thr106 ZINC70454959 −11.1 1 Lys53 2.95 Tyr35, Gly170, Leu171,la51, Val138, Phe169 Tyr35, Val38, Ala51, ZINC85993836 −11.1 Lys53, Leu75, Ile84,Val105, Leu104, Thr106, Phe169, Gly170 Tyr35, Val38, Ala51, SB 202190 −11.0 Lys53, Leu104, Thr106,Leu108, Met109, Phe169, Gly170, Leu171 Tyr35, Val38, Ala51, SB 203580 −10.9 Lys53, Leu104, Thr106,Leu108, Met109, Val30, Phe169, Gly170, Leu171 3.6. Biological Activity Prediction The prediction of activity spectra for substances (PASS) [64] was employed to predict relevant biological activities of the hits. PASS utilizes structural-activity relationship to predict the potential biological activities and mechanisms. When probable activity (Pa) is greater than the probable inactivity (Pi) for a particular compound activity and Pa > 0.3, it is essential to test in vitro the predicted activity [57,88]. The biological activities that were considered included anti-inflammation, kinase inhibition, antiviral, platelet aggregation inhibition and immunosuppression. PASS predictions are usually corroborated experimentally [89,90]. ZINC5519433 was predicted as a platelet aggregation inhibitor (Pa = 0.509 and Pi = 0.008), and antiviral (rhinovirus) with Pa = 0.470 and Pi = 0.037. Platelet activation accelerates inflammation and can lead to disseminated intravascular coagulation, a condition observed in critically ill COVID-19 patients [91,92]. ZINC5733756 was predicted as antiviral (influenza) with Pa = 0.755 and Pi = 0.004, anti-inflammatory (Pa = 0.694 and Pi = 0.017), and antiviral (rhinovirus) with Pa = 0.599 and Pi = 0.006. ZINC4023706 was predicted to be antiviral (influenza) with Pa = 0.764 and Pi = 0.004, anti- inflammatory (Pa = 0.693 and Pi = 0.017), and antiviral (rhinovirus) with Pa = 0.554 and Pi = 0.011. ZINC95486106 was predicted as anti-inflammatory (Pa = 0.694 and Pi = 0.017), antiviral (influenza) with Pa = 0.716 and Pi = 0.005, Beta-adrenergic receptor kinase inhibitor (Pa = 0.823, Pi = 0.011), G-protein-coupled receptor (GPCR) kinase inhibitor (Pa = 0.823 and Pi = 0.011), and RNA-directed RNA polymerase inhibitor (Pa = 0.518 and Pi = 0.006). More so, ZINC1691180 and ZINC4520996 were revealed as anti-inflammatory (Pa = 0.645 and Pi = 0.024), antiviral (rhinovirus) with Pa = 0.525 and Pi = 0.017, and antiviral (influenza) with Pa = 0.748 and Pi = 0.004. ZINC1691180 has been shown to elicit anti-inflammatory response in mice [93]. 3.7. The Rationale for the Selection of Compounds A lead compound is a molecule that is likely to be of therapeutic relevance and may be modified to improve potency, selectivity, and pharmacological profiles [94]. After screening a library of 7282 pre-filtered compounds against the p38 MAPK structure, 42 compounds with binding energies ≤ −11.0 kcal/mol were selected for pharmacological profiling. A total of 18 compounds were selected from this lot as promising hits due to predicted good pharmacological profiles. The protein-ligand interactions of the promising hits were evaluated to identify residues critical for binding. Further biological activity predictions Biomolecules 2021, 11, x FOR PEER REVIEW 8 of 26 Lys53 Tyr35, Gly170, Leu171, la51, ZINC70454959 −11.1 1 2.95 Val138, Phe169 Tyr35, Val38, Ala51, Lys53, BZioImNoCle8cu5l9e9s32803216, 11, 653−11.1 Leu75, Ile84, Val105, Le9uo1f 0245, Thr106, Phe169, Gly170 Tyr35, Val38, Ala51, Lys53, SB 202190 −11.0 were per formed to identify compounds with relevant biologiLceaul1a0c4t,i vTihtire1s06a,n Ldemu1e0c8h, aMneist1m09s, of action. The multistage techniques informed the selectionPhoef16fi9v, eGclyo1m70p, oLueun1d7s1, namely ZINC5519433, ZINC5733756, ZINC95486106, ZINC1691180T,yar3n5d, ZIVNalC384,5 20A9l9a651(,T abLlyes253)., SB 203580 −10.9 The sele cted compounds were predicted a s having plausiLbelue1p04h,a Trhmr1a0c6o,l oLgeuic1a0l8,p Mroefit1le09s, including high gastrointestinal absorption, non-Pgp substrateVsa, la3n0d, Pnhoen1-6C9,Y GPl3yA1740,i nLheuib1i7t1o rs. Assessment of their predicted biological activities showed all five to have anti-inflammatory and antiviral propensities, with Pa > Pi and Pa > 0.3. (a) (b) Figure 3. Cont. Biomolecules 2021, 11, x FOR PEER REVIEW 9 of 26 Biomolecules 2021, 11, 653 10 of 25 (c) (d) Fiigurre 3.. DDoocckkiningg ppoosesse sanandd pprortoetieni-nl-igliagnadn dinitnetrearcatciotino nstustduidesie osf otfopto pfoufor uhriths itws iwthi tthet hloewloewst ebsitnbdiindgi negneerngeiersg i(eas) (Za)INZCIN95C498564180661,0 (6b,)( bZ)INZICN9C59915397123072, 0(c, )( cZ)IZNICN3C33833823029009, 0a,nadn d(d(d) )ZZININCC9595991199007766 aaggaaininsst tpp3388 MAPPK ssttrruuccttuurree.. TThhee bbiinnddiinngg ppoocckkeettss aarree rreepprreesseenntteedd aass ssuurrffaacceess aanndd tthhee lliiggaannddss aass sstticickkss. .InIn ththe eLLigiPgPlolto+t +rerperperseesnetnattaiotinosn, sth, teh leiglaignadnsd asrea rdeisdpislapylaedye ads apsuprpulrep lsetisctkisc,k sh,yhdyrdoprohpohboicb iccocnotanctatsc tasraer sehsohwown nasa srerded spspookke eaarcrcs,s ,aanndd tthhee hhyyddrrooggeenn bboonnddss wwiitthh tthheeiirr rreessppeeccttiivvee bboonndd lengths as green. lengths as green. 3.6. Biological Activity Prediction The prediction of activity spectra for substances (PASS) [64] was employed to predict relevant biological activities of the hits. PASS utilizes structural-activity relationship to predict the potential biological activities and mechanisms. When probable activity (Pa) is greater than the probable inactivity (Pi) for a particular compound activity and Pa > 0.3, it is essential to test in vitro the predicted activity [57,88]. The biological activities that were considered included anti-inflammation, kinase inhibition, antiviral, platelet aggregation Biomolecules 2021, 11, x FOR PEER REVIEW 11 of 26 Biomolecules 2021, 11, x FOR PEER REVIEW 11 of 26 The inhibitory constants (Ki) of the selected compounds and the known inhibitors wereT chaelc uinlhatiebdit o(Sryu pcpolnesmtaennttsa (rKy iT) aobfl eth Se2 )s eulseicntegd a cpormevpioouunsldys d aensdcr itbheed k mnoewthno din [h57ib,1it0o1r]s. Awlelr teh cea slceulelcatteedd c(Soumpppoluemndesn thaardy lToawbeler Sp2r)e duiscitnegd a K pi rveavliuoeuss ltyh adne saclrl itbheed k mnoetwhno din [h5i7b,1it0o1r]s. Auslel dth hee sreelienc t(eSdu pcpomlempoeunntadrsy hTaadb lleo wS2e)r, pinrdediciacttiendg K thi eviarl iunehsi bthitaonr ya lpl othteen ktinaol w[1n0 2i]n. hFiobrit tohres usesleedct heder ecoinm (pSouupnpdlesm, ZenINtaCry9 5T4a8b6l1e0 S62 w), ains dpirceadtiincgte tdh etoir hinahveib tihtoer ylo pwoetsetn Ktiai lv [a1l0u2e] .o Ffo 1r. 3th48e nseMle ctwedh icleo mZpIoNuCn5d7s3, 3Z7I5N56C 9h54a8d6 1t0h6e whaisg hperestd icKtei dv taol uhea voe ft h7e. 2l9o1w ensMt K. i ZvaINluCe 1o6f9 111.38408, ZnMIN Cw55h1il9e4 3Z3,I NanCd5 7Z3I3N7C55465 20h9a9d6 hthaed thhieg hsaemst e Kpir edviacltueed Kofi v7a.l2u9e1 ofn M3.1. 35Z InNMC (1S6u9p11p8le0-, Biomolecules 2021,m 11, x FOR PEER REVIEW 11 of 26 ZIeNnCta5r5y1 T94a3b3le, Sa2n)d. FZoIrN thCe4 5k2n0o9w96n hinahdi bthiteo rssa, mthee pprreeddiicctteedd KKii vvaalluuees orfa n3g.1e3d5 fnroMm ( 8S.u6p3 pnlMe- Biomolecules 2021,tm 1o1 e, 1xn .F6tOa2Rr PyµE METRa Rb(ESlVeuI ESpW2p )l.e Fmoer nthtaer kyn Toawbnle i nS2h)i.b Tithoers k, tnhoew pnr eidnihcitbeidto Krsi vwaeluree se rxapnegreimd efrnotma11l l o8yf .2 6d6 3e tne Mr- mto in1.e6d2 tµoM h a(vSue pICpThl50 mvalues ranging from 3.7 to 41.2 nM (Supplementary Table S2) [103–e inehnibtiatorryy Tcoanbstlaen Sts2 ()K. iT) hofe t hken soewlecnte din choimbpitoournsd sw aenrde theex pkneorwimn einhtiabliltyor sd eter- 1m0i7n]e. dA RtoR hYa-7vw9ee7 rIe Cw c ith the lowest IC50 value of 4.5 nM had binding energy of −9.5 kcal/mol, which is higher tThh a50elc vuinlaahltiuebdiet os(S urapnplgeimnegn tfarroy mTa b3le.7 S 2t)o u 4si1n.g2 an pMrev (ioSusplyp dleesmcreibnetda mrye thToadb [l5e7 ,S1021)] .[ 103– 107]. ARRY-7A9ll7 t hwea istnehl e ttchhteadt lo rfy constants (Ki) of the secoowm SB 20358 (−10.9 kcal l/emcteodl) c womitpho uthndes h aingdh tehset kInCow nv ainluhieb i4to1r50 .s2 nM. were calculated (Suppeposlute mnICdesn th aavrdya llTouawebe loer fSp 24r)e. d5ui scnitneMgd a Kh pia rvdeav lbiuoieunss dltyhi andneg sa celrln itbheeerd gk mynoe owthfno − di9n [.h5i7b ,k1it50 0co1ar]sl. / mol, uAslel dth hee sreelienc t(eSdu pcpolmempoeunntadrsy hTaadb lleo wS2e),which is higher than that of SB 20358 (−10.9r k pinrdediciacttiendg Kthi eviar liunehsi bthitaonr ya lpl othteen ktinaol w[1n0 2i]n. hFiobri ttohres Table 2. A list of selected cosuemsleepdct oheudn c cal/m eredoinsm (wpSouiutphnp dltehsm,e ZiernI N2taDCry9 s 5Tt4ra8ub6cl1et0u S6r2 ew),s a ians ndpdirce acdtoi o incmgt le)md w otonit hh/aI UvtheP etAh heC ilg onhwaemessett sKI.C i v50a vluael oufe 1 4.31.2 nM. their inhibitory potential [102]. For th48e nM while ZINC57337556 had t Table 2. A list of selected cosemlepctoeudn cdosm wpoituhn dths,e ZirI N2DC9 s5t4r8u6c1t0 h6e whaisg hperestd icKtei dv taol uhea voef th7e. 2l9o1w ensMt K. i ZvIaNluCe1 o6f9 111.38408, Ligand ID CommZnoMINn C/wI5U5h1iPl9eA4 3ZC3,I NanCda5 m7Z3Ie3N7 C55465 20h9a9d6 htuhaerde tshh aiegn hsdaem sct oe mKpirm edvoiancltu e/e2dI U DKoPfi SAv7taCr.l2uu 9nec1a tomufn re3Mes.1.. 35Z nINMC (1S6u9p11p8le0-, mentary Table S2). For the known inhibitors, the Ligand ID CommZINnC/I5U51P9A43C3, aNnda ZIN C4520996 had the same pprreeddiicctteedd KKii vvaalluuees oraf n3g.1e3d5 fnroMm ( 8S.u6p3 pnlMe- tmo e1n.6ta2r yµ MTa b(Sleu pS2p)l.e Fmoer nthtaer kyn Toawbnle iSn2h)i.b Tithoers k, nthoew pnr eidnihcitbe2idtDo Krsi S vwaterlurueec set rxuapnreegr iemd efrnotamll y8. 6d3e tneMr- mtoi n1.e6d2 tµoM h a(vSue pICpl5e0 mvaelnuteasr yr aTnagbilneg S f2r)o. mTh 3e. 7k ntoo w41n.2 i nnhMib i(tSourps pwleemree enxtapreyr imTaebnltea lSl2y) d[1et0e3r–- 1m0i7n]e. dA RtoR hYa-7v9e7 I wC it hv athluee lso wraensgt iInCg5 0f vroamlu e3 o.7f t4o.5 4 n1M.2 hnaMd (bSiunpdpinlegm enenertgaryy o Tf a−b9l.5e kSc2a) l[/m10o3l–, 50 Biomolecules 2021, 11, 653 w10h7i]c. hA iRs RhYig-h7e9r7 twhaitnh tthhaet loofw SeBs t2 0IC358 v (a−l1u0e.9 o kf c4a.5l/ mnMol )h wadit hb itnhdei nhgig ehneesrt gICy 5o0 fv −a9lu.5e k 4150 ca.l2/ mnM1o1l., of 25 ZINC5519433 Table 2. A liswZt ohufi sicehle oicsnt ehiding c hoAemr p tohuannd tsh wati tohf t ShBei r2 20D35 s8tr (u−c1t0u.r9e sk acnadl/ mcooml)m wonit h/I UthPeA hCi gnhaemset sI.C 50 value 41.2 nM. ZINC5519T4a3b3l e 2. ALilgisatnodf IsDTe laebclete 2d. Ac olimsZt opufo siuhelneoCcdnotesmidnw mc oAiomthn p/otIhUunePdiArs C2wD iNthsa ttmhreuei rc t2uDr setsruacntudrecso amndm coomnm/oInU /2IPUDAP SCAtCrnu nacamtmuereess. . Ligand ID Common/IUPAC Name 2D Structure Ligand ID Common/IUPAC Name 2D Structure ZINC5519433 Zuihonin A ZINC5519433 ZINC5519433 Zuihonin A Zuihonin A ZINC5733756 8,11,13-Abietatriene-3beta-ol ZINC5733756 8,11,13-Abietatriene-3beta-ol ZINC5733756 ZINC57337856,1 1,13-Abieta8tr,1ie1n,1e3--3Abbeietata-torliene-3beta-ol ZINC5733756 8,11,13-Abietatriene-3beta-ol (1S,2aS,2bR,4aS,5R,8aS,8bR,10aR)-1,5,8a- ZINC95486106 ((11SS,,22ataSrSi,2m,2bbeRtR,h4,ay4Sal,-5ShR,e5,xR8a,S8d,ae8Sbc,aR8h,b1yR0da,1Rro0)-ac1yR,5c),l-8o1a-,-5,8a- buta[a]phen(a1trnSi,tm2haeSrt,e2hnbyRel--,45a-Sc,a5Rrb,8oaSx,y8bliRc, 1a0caiRd)- 1,5,8a- ZINZCIN95C48965140866106 hexadetcraihmyedthyl-hexadZINC95486106 rocyclotbruimta[eac]aphhyendarnotchyrcelnoe-ethyl-hexadecahydr-o5c-yclo- buta[a]phen(ac1anSr,but2atbahoS[rxe,2bR,4a]pynhleiec-n5a-Sc,a5Rrb,8oaxS,y8bliR,10aR)-1,5,8a-ZINC95486106 trimethyal-nhtehxraedneec-5a-hcya c acid drbroocxyyclilco -acid acid buta[a]phenanthrene-5-carboxylic acid BiomolecuZleIsN 20C21,6 1911, x1 8FO0 R PEZEIRN CR1E6V9I1E1W80 Methyl dehMydetrhoyal dbeiehtyadtreo abietate 12 of 26 ZINC1691180 Methyl dehydroabietate ZINC1691180 ZINC1691180 Methyl dehMydetrhoyal bdeiehtyadtreo abietate Methyl (1S,4aR,10aS)-1,4a-dimethyl-7- Methyl (1S,4aR,10aS)-1,4a-dimethyl-7-propan-2- ZINZCIN45C240599260996 ypl-r2o,3p,4a,n9,-120-,y1l0-a2-,h3e,4x,a9h,1y0d,r1o0pah-ehneaxnathhryednreo-1p-he- nanthcraerbnoex-y1l-actaerboxylate 3.8. MZIoNleCcu5l5a1r9 D43y3nhamadicbs iSnidmiunlgateinoner ogfy Soelfe−cte1d1 .C6okmcaplo/umndosl (Table 1). It formed hydrophobic interaAct io10n0s wnsi thMrDes isdimueuslainticolun dwinags Lpyesr5f3o,rmTherd1 0o6na neidghPth es1tr6u9c,twurheisc hcohnasviestbinegen orfe pthoer tuedn- inbooutnhder psrtuotdeiiens, ttoheb ecocmritpicleaxl eins loifg athned fbivined sienlgec(tTeadb cleom1)p. oItuinsd7s1,% ansdtr uthcteu trwaloly ksnimowilnar pt3o8 mMaAcePliKg ninahnibviitaoDrsr SuBg B2a0n3k58[09 5a]n; da nSBat 2u0r2a1l 9co0.m Tphoisu wndast haaimt aetdte antu parteosbitnhge athceti vstartuiocntuoraf lp s3t8a- MbiAliPtyK aanndd cohnafsoerxmhaibtiiotendala cnhtai-ningflesa minm phatyosriyolaocgtiicvailt yco[n96d]i.tiPornesv [i1o0u8s].e Txpheer pimareanmtaeltesrtus devieasl- suhaotwede dwZerIeN tChe5 5r1o9o4t3 m3 eaasna sqkuinaares edeinvhiaibtiiotonr (RanMdSDse)l,e tchteiv readtoiucs- Jouf ngyNra-tteiormn i(nRagl),k ainndas tehse (JcNonKf)o[r9m7a].tiTohniaslc cohraronbgoersa wteistho utirmsteu. dies on the potential lead-likeness of ZINC5519433. Afurther lead optimization may likely increase its selectivity towards p38 MAPK [98]. ZTIhNeC R1M69S1D18 0is aan dmeZtIrNicC u4s5e2d0 9to96 awsseerses bthoeth stparbeidliitcyt eodf tao phraovteeina bstirnudcitnugree n[1e0r9g]y. Aofn −R1M1.S6Dk cpalol/t mofo lth(eT apb3l8e M1)A. PBKo-tlhigfaonrdm ceodmhpyldexroesg ethnabt otennddin tgo wcointhveTrygre3 5is, iwnditihcavtiavrey ionfg a lestnagbtlhes aanndd whyedllr-eoqpuhiolibbircatceodn tsaycststewmi th[11so0]m. eCcornitsiicdaelrrinesgi dtuhees p(T38a bMleA1)P.KZ-ZININCC15659119148303 hRaMs bSeDe nprloept,o ar tsetdeatdoye vrioskee toa n0t.i3-i8n nflmam wmaast oorbyseprrvoepde rdtuiersinvgia trheeg 1u lnasti nogf stihmeuplraotdiounc.t iIot nthoefn indflesacmenmdaetdo raynfda cmtoarinTtNainFαedi nanm aivceer[a9g3e]. RTMNSFDα omf e0d.3ia ntems t(hFiegaucrteiv 4aat)i.o Mn oorfepo3v8erM, tAhPe KRM[9S9D] aonfd thoeb sptr3u8c tMioAnPoKf -tZhiIsNpCa5th73w3a7y56w ciothmZpIlNexC g1r6a9d1u18a0llyc oruolsde bteo u0s.4e2fu nl min saervoeurnedC 1O5V nIDs -a1n9d quickly dropped to an average RMSD of 0.3 nm around 17 ns (Figure 4a). It maintained the average RMSD of 0.3 nm till about 62 ns, where a sharp rise to an average of 0.45nm was observed till the end of the 100 ns simulation period (Figure 4a). The p38 MAPK- ZINC4520996 complex experienced a rise in RMSD to 0.45 nm around 13 ns and main- tained stability till the end of the simulation time (Figure 4a). An initial rise to 0.40 nm during the first 6 ns was observed for the p38 MAPK-ZINC1691180 complex (Figure 4a). The RMSD dropped to about 0.32 nm around 20 ns and was maintained till the 100 ns period (Figure 4a). Additionally, the RMSD plot for p38 MAPK-ZINC95486106 revealed stability after 15 ns, with an average RMSD of 0.33 nm (Figure 4a). Considering the known inhibitors, average RMSDs of 0.35 nm and 0.38 nm were observed for the SB 203580 and SB 202190-p38 MAPK complexes, respectively. The p38 MAPK complexes of ZINC5519433, ZINC95486106, ZINC1691180, ZINC4520996, SB 203580 and SB 202190 had lower average RMSDs relative to the unbound protein (Figure 4a). This may suggest sta- bility induced by ligand binding. In a pharmacophore-guided study to identify p38 MAPK inhibitors, a 10 ns MD sim- ulation was performed on the most potent molecule (compound 48) and the protein com- plex [111]. It was observed that the RMSD ranged between 0.5 to 2.5 Å (0.05 to 0.25 nm) [111]. Other MD simulation studies on p38 MAPK-ligand complexes reported RMSD val- ues up to about 0.6 nm [112,113] consistent with the results presented herein. The Rg was measured to evaluate the compactness of the protein systems during the MD simulation [114]. The Rg of p38-MAPK-ZINC5519433 fell from 2.25 nm to 2.20 nm during the first 5 ns and it further rose and peaked at 2.26 nm around 9 ns. It then fluctuated over the re- maining time with an average Rg of 2.20 nm (Figure 4b). For the p38 MAPK-ZINC5733756 complex, an initial Rg of 2.2 nm was obtained and it then fluctuated around 2.15 nm until 60 ns, where it dropped to 2.02 nm. The p38 MAPK-ZINC4520996, p38 MAPK- ZINC1691180, p38 MAPK-ZINC95486106, p38 MAPK-SB 203580, and p38 MAPK-SB 202190 complexes were observed to be stable over the simulation time with average Rg of 2.12 nm, 2.13 nm, 2.13 nm, 2.14 nm, and 2.15 nm, respectively. These were close to the average Rg of 2.12 nm exhibited by the unbound protein. A stably folded protein main- tains a reasonable steady radius of gyration throughout the simulation [114]. Biomolecules 2021, 11, 653 12 of 25 cases. ZINC5733756 had binding energy of −11.1 kcal/mol due to strong intermolecular interaction with residues including Thr106 and Phe169 (Table 1). Moreover, ZINC5733756 is 73% structurally similar to estradiol, a hormone that exerts anti-inflammatory activities by targeting the activation of Nrf2 [100]. Furthermore, ZINC95486106 formed interactions with critical residues namely Lys53, Leu108, and Phe169 contributing to its binding energy of −12.1 kcal/mol. It was also predicted to be GPCR and Beta-adrenergic receptor kinase inhibitors via PASS with Pa > 0.8. The inhibitory constants (Ki) of the selected compounds and the known inhibitors were calculated (Supplementary Table S2) using a previously described method [57,101]. All the selected compounds had lower predicted Ki values than all the known inhibitors used herein (Supplementary Table S2), indicating their inhibitory potential [102]. For the selected compounds, ZINC95486106 was predicted to have the lowest Ki value of 1.348 nM while ZINC57337556 had the highest Ki value of 7.291 nM. ZINC1691180, ZINC5519433, and ZINC4520996 had the same predicted Ki value of 3.135 nM (Supplementary Table S2). For the known inhibitors, the predicted Ki values ranged from 8.63 nM to 1.62 µM (Supplementary Table S2). The known inhibitors were experimentally determined to have IC50 values ranging from 3.7 to 41.2 nM (Supplementary Table S2) [103–107]. ARRY-797 with the lowest IC50 value of 4.5 nM had binding energy of −9.5 kcal/mol, which is higher than that of SB 20358 (−10.9 kcal/mol) with the highest IC50 value 41.2 nM. 3.8. Molecular Dynamics Simulation of Selected Compounds A 100 ns MD simulation was performed on eight structures consisting of the unbound protein, the complexes of the five selected compounds, and the two known p38 MAPK inhibitors SB 203580 and SB 202190. This was aimed at probing the structural stability and conformational changes in physiological conditions [108]. The parameters evaluated were the root mean square deviation (RMSD), the radius of gyration (Rg), and the conformational changes with time. The RMSD is a metric used to assess the stability of a protein structure [109]. An RMSD plot of the p38 MAPK-ligand complexes that tend to converge is indicative of a stable and well-equilibrated system [110]. Considering the p38 MAPK-ZINC5519433 RMSD plot, a steady rise to 0.38 nm was observed during the 1 ns of simulation. It then descended and maintained an average RMSD of 0.3 nm (Figure 4a). Moreover, the RMSD of the p38 MAPK-ZINC5733756 complex gradually rose to 0.42 nm around 15 ns and quickly dropped to an average RMSD of 0.3 nm around 17 ns (Figure 4a). It maintained the average RMSD of 0.3 nm till about 62 ns, where a sharp rise to an average of 0.45nm was observed till the end of the 100 ns simulation period (Figure 4a). The p38 MAPK-ZINC4520996 complex experienced a rise in RMSD to 0.45 nm around 13 ns and maintained stability till the end of the simulation time (Figure 4a). An initial rise to 0.40 nm during the first 6 ns was observed for the p38 MAPK-ZINC1691180 complex (Figure 4a). The RMSD dropped to about 0.32 nm around 20 ns and was maintained till the 100 ns period (Figure 4a). Additionally, the RMSD plot for p38 MAPK-ZINC95486106 revealed stability after 15 ns, with an average RMSD of 0.33 nm (Figure 4a). Considering the known inhibitors, average RMSDs of 0.35 nm and 0.38 nm were observed for the SB 203580 and SB 202190-p38 MAPK complexes, respectively. The p38 MAPK complexes of ZINC5519433, ZINC95486106, ZINC1691180, ZINC4520996, SB 203580 and SB 202190 had lower average RMSDs relative to the unbound protein (Figure 4a). This may suggest stability induced by ligand binding. In a pharmacophore-guided study to identify p38 MAPK inhibitors, a 10 ns MD simulation was performed on the most potent molecule (compound 48) and the protein complex [111]. It was observed that the RMSD ranged between 0.5 to 2.5 Å (0.05 to 0.25 nm) [111]. Other MD simulation studies on p38 MAPK-ligand complexes reported RMSD values up to about 0.6 nm [112,113] consistent with the results presented herein. The Rg was measured to evaluate the compactness of the protein systems during the MD simulation [114]. The Rg of p38-MAPK-ZINC5519433 fell from 2.25 nm to 2.20 nm during the first 5 ns and it further rose and peaked at 2.26 nm around 9 ns. It then fluc- Biomolecules 2021, 11, 653 13 of 25 tuated over the remaining time with an average Rg of 2.20 nm (Figure 4b). For the p38 MAPK-ZINC5733756 complex, an initial Rg of 2.2 nm was obtained and it then fluctuated around 2.15 nm until 60 ns, where it dropped to 2.02 nm. The p38 MAPK-ZINC4520996, p38 MAPK-ZINC1691180, p38 MAPK-ZINC95486106, p38 MAPK-SB 203580, and p38 MAPK-SB 202190 complexes were observed to be stable over the simulation time with Biomolecules 2021, 11, x FOR PEER REVIEW 14 of 26 average Rg of 2.12 nm, 2.13 nm, 2.13 nm, 2.14 nm, and 2.15 nm, respectively. These were close to the average Rg of 2.12 nm exhibited by the unbound protein. A stably folded protein maintains a reasonable steady radius of gyration throughout the simulation [114]. (a) Figure 4. Cont. Biomolecules 2021, 11, x FOR PEER REVIEW 15 of 26 Biomolecules 2021, 11, 653 14 of 25 (b) FFiigguurree 44.. MMoolleeccuullaarr ddyynnaammiiccss ssiimmuulalattioionnss ggrraapphhss (a(a) )RRMMSSDD vveresrusus stitmime e(n(ns)s )anadnd (b()b )RRg gvevresrussu stimtime e(n(sn)s. )I.n I(na)( aa,nbd), (tbh)e, uthneb uonubnodupn3d8 pM38A MPKAPpKro pterionte(i3nZ (S35Z)S, 5Z),I NZICN1C69116198101,80Z, IZNICN4C542502909969,6Z, IZNINCC5551591493433,3,a nanddZ ZININCC55773333775566, , SSBB 220022119900,, aanndd SSBB 220033558800––pp3388 MMAAPPKK ccoommpplleexxeess aarree sshhoowwnn aass bbllaacckk,, rreedd,, ggrreeeenn,, bblluuee,, yyeellllooww,,b brroowwnn, ,a asshha annddp puurrpplele, ,r reessppeecctitviveelyly. . The distance between the centre of mass of a protein and an inhibitor influences the molecular interactions, which in turn play a crucial role in the structural stability of the protein-ligand complexes [115]. Snapshots at 25 ns intervals (time step = 0, 25, 50, 75, and 100 ns) were generated to elucidate the position and binding modes of the ligands during the 100 ns MD simulation period (Figure 5 and Supplementary Figure S1). In the snapshot analysis, all the ligands were observed to stably bind in the ATP binding pocket of the p38 MAPK protein throughout the 100 ns simulation time. Moreover, the number of hydrogen bonds over the simulation time with a distance cut-off of 0.35 nm was also generated (Figure 6). The 2D protein-ligand interaction profiles of the snapshot frames of the protein-ligand complexes at time intervals of 25 ns were also generated to study the interactions during the 100 ns simulation period (Supplementary Figures S2 and S3). For the p38 MAPK-ZINC1691180 complex, the hydrogen bond of length 3.17 Å wi th Tyr35 before MD (Supplemen(at)a ry Figure S2a) was lost at 25 ns and was not formed again throughout the simulation time. However, at 75 ns, ZINC1691180 formed a new hydrogen bond with Ser32 of bond length 2.96 Å, which was lost at 100 ns (Figure S3a). For the p38 MAPK-ZINC4520996 complex, the hydrogen bonding with Tyr35 pre-MD was not present at 25 ns (Supplementary Figures S2b and S3b). However, at 50 ns, ZINC4520996 formed a new hydrogen bond with Asp112 of bond length 3.01 Å (Supplementary Figure S3b). Biomolecules 2021, 11, x FOR PEER REVIEW 15 of 26 Biomolecules 2021, 11, 653 15 of 25 The hydrogen bond with Asp112 was lost at 75 ns and re-formed at 100 ns with a bond length of 3.2 Å (Figure S3b). For the p38 MAPK-ZINC5519433 complex, the ligand was observed to form a hydrogen bond at 25 ns with Gly170 of bond length of 3.15 Å (Supplementary Figure S3d). However, at 50, 75, and 100 ns, there were no hydrogen bonds observed. For the protein-ZINC95486106 complex, the hydrogen bond with Met109 (bond length of 3.03 Å) was lost throughout the simulation time (Supplementary Figures S2c and S3c). However, at 100 ns, two hydrogen bonds were observed with Ser32 of a bond length of 2.73 Å and Lys15 with a bond length of 3.06 Å (Supplementary Figure S3c). For the p38 MAPK-ZINC5733756 complex, ZINC5733756 did not form any hydrogen bond with the protein before MD (Supplementary Figure S2e). At 25 ns, ZINC5733756 was observed to form one hydrogen bond with Asp112 of length 2.86 Å. At 50 ns, it formed two hydrogen bonds with the p38 MAPK protein composed of one with Asp112 (bond length of 2.71 Å) and the other with Asn115 (bond length of 2.70 Å). Hydrogen bonds with Asp112 and Asn115 of lengths 2.65 and 3.14 Å were observed at 75 ns, respectively (Supplementary Figure S3e). At 100 ns, ZINC5733756 maintained only one hydrogen bond of length 2.70 Å with Asp112 (Supplementary Figure S3e). The multiple hydrogen bonds formed indicate the strong interactions between ZINC5733756 and the p38 MAPK protein throughout the simulation period and could influence the activity of the ligand [116]. 3.9. Evaluation of Selected Compounds via MM-PBSA Calculations MM-PBSA technique was used to estimate the free binding energies of the seven com- plexes. This technique addresses limitations with current scoring functions [117,118]. The binding free energy is a more reliable method of evaluating docking studies [77,117,119]. The other parameters computed included the van der Waal, electrostatic, polar, and non-polar solvation energies (Table 3) [120,121]. The selected molecules comprising ZINC5519433, ZINC95486106, ZINC5733756, ZINC1691180 and ZINC4520996 had pre- dicted average free binding energies of −185.122, 30.620, −61.726, −146.008, and −151.561 kJ/mol, respectively. SB 202190 and SB 203580 were predicted to have plau- sible free binding energies of −166.369 kJ/mol and −236.175 kJ/mol, respectively (Table 3). Since ZINC95486106 was the only selected compound with positive binding free energ y, it was eliminated as a potential lead. Optimization of the ligand structure in future studies can improve its affinity to th(be)p 38 MAPK protein. The poor binding free energy could Figure 4. Molecular dynamliacsr gseimlyublaetidounes gtoratphhesh (iag) hReMleScDtr voesrtsautisc teimneer (gnys)r ealnadti v(be) tRogt hveerostuhse trimcoem (npso).u Inn d(as)( Tanadb l(eb4),) and the unbound p38 MAPK prhoetenicne (3mZaSy5)l, iZmINit Cit1s6l9e1a1d80l,i kZeInNeCs4s5[25079]9.6I,n ZaINllCse5v51e9n43c3o,m anpdo uZnINdsC5a7s3s3e7s5se6,d S,Bth 2e02v1a9n0,d aenrdW aal SB 203580–p38 MAPK compfolerxceess acroe nshtroiwbunt aesd bflavcko,r raebdl,y gtroeetnh,e bfluree, ybeilnlodwin, gbreonwenr,g aisehs a[2n9d, 1p2u0r,p1l2e1, ]r.esHpoecwtievveleyr., the polar solvation energy contributed large positive energies to binding in all the complexes [122]. (a) Figure 5. Cont. Biomolecules 2021, 11, x FOR PEER REVIEW 16 of 26 Biomolecules 2021, 11, 653 16 of 25 (b) (c) (d) (e) FFiigguurree 55.. SSnnaappsshhoottss aatt 2255 nnss iinntteerrvvaallss ((((ttiimee sstteepp == 00,, 2255,, 5500,, 7755,, aanndd 110000 nnss)) ffoorr tthhee bbiinnddiinngg mooddeess ooff tthhee lliiggaanndd--pp3388 MAAPPKK ccoommpplleexxeess.. TThheec caartrotooonnr erpeprerseesnetnattaiotinonsh sohwosw(sa )(aZ)I NZICN1C69116198101,8(0b, )(ZbI)N ZCIN45C2405929069, 9(c6), Z(cI)N ZCI9N5C48965140866,1(0d6),Z (IdN) CZ5IN51C94535319a4n3d3 (e) ZINC5733756- p38 MAPK complexes. The ligands are represented as spheres and the protein as cartoons. All the ligands were observed to bind stably in the ATP binding pocket. Biomolecules 2021, 11, x FOR PEER REVIEW 17 of 26 Biomolecules 2021, 11, 653 17 of 25 and (e) ZINC5733756- p38 MAPK complexes. The ligands are represented as spheres and the protein as cartoons. All the ligands were observed to bind stably in the ATP binding pocket. FFiigguurree 66.. TThhe etototatal lnnuummbbere roof fhhyyddroroggeenn bboonndds sfoformrmeedd bbeetwtweeeenn ththee sseelelecctetedd ccoommppoouunndds saanndd ththee pproroteteinin sstrturucctuturere. . ZZIINCC11669911118800,, ZZIINCC44552200999966,, ZZIINCC9955448866110066, ,Z ZININCC5551591493433,3Z, IZNINC5C753733735765, 6S,B S2B0 22109201,9a0n, danSBd 2S0B3 528003-5p8308-pM3A8 PMKAcPomK pcloemxe-s palreexreesp arrees erenpterdesaesnbteldac aks, rbeldac,kg,r reedn,, gbrlueen, ,y bellluoew, ,yberlloown, barnodwans ha,nrde sapsehc,t riveseplye.ctively. Table 3. The energy3c.9o.n Etrvibaulutiaotniosnfo orf tShelepcrtoetde iCno-lmigpaonudncdosm vpiale MxeMs b-aPsBedSAon CtahlecuMlaMti-oPnBsS A computations. Electrostatic MM-PBVSaAn DteecrhWnaiqalue was Puosleadr stool veastitoimn ate the Nfroene- Pboinladring energies of the seven Name Energycomplexes. ThiEsn teercghynique addresseEsn leimrgiytations wSitohl vcautirornenEtn secrogrying fuBnicntidoinngs [E1n1e7r,g1y18]. (kJ/molT)he binding (fkreJ/em oeln)ergy is a m(okrJe/m roell)iable metho(dk J/omf oel)valuating do(ckkJi/nmgo ls)tudies ZINC5733756 −15.833 ± 1[67.74,21317,11−9]1. 1T8h.1e1 1ot±he7r7 .p7a65ramete4r7s. 7c8o7m±p2u8t.e1d17 includ−ed9 .t6h8e9 v±a8n.8 d5e6r Waal−, e9l5e.8c4tr6o±sta74ti.c9,3 0po- ZINC5519433 −2.575 ± 7l.a1r6,9 and no−n2-2p2o.6l8a5r ±so1lv0.a5t7i2on ene5r8g.i3e3s2 (±Ta1b5l.0e5 35) [120,−12118].1. 9T4h±e s0e.5l3e5cted m−o1l8e5c.u12le2s± co2m1.3p4r7is- ZINC95486106 75.738 ± 77in.2g9 0ZINC5−58169.044343,± Z5I0N.9C309548614096.6, 7Z2I±N7C55.973353756, Z−I8N.7C4716±9141.88406 and Z3IN0.6C2405±204929.765 5had ZINC1691180 −12.086 ±p5r.0e7dicted −a1v8e0r.5a9g3e± f1re7.e4 1b5indin6g3 .0e8n4e±rg1ie6s.3 7o1f −185−.11262.4, 1330±.602.807, 0−61.7−261,4 6−.010486.±00187,. 29a7nd ZINC4520996 2.277 ± 5.−416561.561 k−J/2m03o.6l,9 8re±sp1e7c.6ti6v5ely. SB66 2.803251±901 a1.n7d89 SB 203−51860.9 w76e±re 0p.7r3e8dicted− t1o5 1h.a56v1e ±pl2a2u.6s2ib2le SB 202190 −2.041 ± 3.990 −209.281 ± 15.503 63.370 ± 16.798 18.417 ± 0.895 −166.369 ± − ± free bindin−g energi±es of −166.369 kJ/mol and −236.175 kJ/mol, respectively (Table 3 1)9. .S3i5n5ce SB 203580 5.094 3.533 280.578 10.532 70.328 ± 22.075 −20.831 ± 1.546 −236.175 ± 26.555 ZINC95486106 was the only selected compound with positive binding free energy, it was eliminated as a potential lead. Optimization of the ligand structure in future studies can Table 4. A table showing tihme preor-vres iitdsu aeffeineirtgyy toco tnhter ipbu38ti oMnsAoPfKth percortiteiicna.l Trehseid puoeos rin btienradcitningg frweiet hentheergliyg acnodusl.dT lhaergely energies were calculated frobme dthueeM toM th-PeB hSiAghco emlepcutrtaotsitoantsi.c Tehneeregnyer rgeylavtailvuee stoa rtehpe roetshenerte cdoimn pkJo/umnodl.s (Table 4) and hence may limit its lead likeness [57]. In all seven compounds assessed, the van der Waal forces Residue ZINC5733756 ZcIoNnCtr5i5b1u9t4e3d3 favZorINabCly95 t4o8 6th10e6 freeZ bIiNnCdi1n6g91 e1n80ergiesZ I[N29C,1425020,192916]. HoSwBe2v02e1r9, 0the pSoBla2r 0s3o5l8v0a- Tyr35 −0.0810 tio−n4 e.0n9e6r9gy contrib−u0t.e0d88 l1arge positi−v1e. 9e1n9e6rgies to bi−n0d.i5n4g51 in all the− c1o.6m99p5lexes [−1232.0]8. 67 Val38 −3.8266 −8.A72d4d7itionally, −a 0p.7e0r2-5residue de−co7.m65p8o2sition of t−h4e. 7b4i7n1ding en−er1g1y.3 6w8a2s perf−or4m.33e4d1 to Ala51 −1.7431 ga−in3 .u58s0e9ful insigh−t − in0.t6o8 6im8 portant i−nt2e.7r0a6c6tions of ke−y4 r.0e0s1id2ues (Fig−u5r.e1 373 a9nd Su−pp2.l7e3m69en- Lys53 4.9439 4.3252 12.3692 8.4199 13.3187 12.708 14.7836 Thr106 −0.6584 tar−y6 F.1i2g2u4re S4). Res0i.d0u03e0s contributin0.g5 6b9i3nding free −en0.e1r4g8i3es greate−r 2t.h51a4n0 5 kJ/m−o6l .7o4r1 l7ess Leu108 −3.1269 th−an3 .−2578 k5J/mol are −w0o.7rt5h0y9 of consid−e4ra.0t6io40n as critica−l f7o.2r5 b90inding [5−7]3.. 8F1o5r4 the p3−8 0M.0A60P8K- Met109 −4.2382 ZI−N0C.3595319433 com−pl2e.1x9, 1c2ritical resi−d0u.2e7s8 P7he169 and− 1T.8h9r01006 contr−ib0u.4t8e4d2 favora−b0le.7 9e8n0er- Phe169 −5.8140 gie−s9 .o2f4 7−09.2470 kJ/−m1o.3l 2a8n5d −6.1224− k14J/.5m95o4l, to ligan−d6 b.8i0n0d1ing, res−p1e4c.t9i1v6e2ly (Ta−b1le0 .248 4a2nd Figure 7). Asp168 was found to contribute unfavorable energy of 7.1894 kJ/mol (Figure 7). Biomolecules 2021, 11, 653 18 of 25 Additionally, a per-residue decomposition of the binding energy was performed to gain useful insight into important interactions of key residues (Figure 7 and Supplementary Figure S4). Residues contributing binding free energies greater than 5 kJ/mol or less than −5 kJ/mol are worthy of consideration as critical for binding [57]. For the p38 MAPK-ZINC5519433 complex, critical residues Phe169 and Thr106 contributed favorable energies of −9.2470 kJ/mol and −6.1224 kJ/mol, to ligand binding, respec- tively (Table 4 and Figure 7). Asp168 was found to contribute unfavorable energy of 7.1894 kJ/mol (Figure 7). Moreover, only Phe169 contributed binding free energies of ±5.0 kJ/mol to ZINC5733756 binding (Table 4 and Supplementary Figure S4d). From the energy decomposition plot, Tyr30, Tyr38, Leu167, and Phe169 were predicted as critical residues that contributed favorably to ZINC1691180 binding (Table 4 and Supplementary Figure S4a). However, Lys53 and Asp112 contributed adversely towards ZINC1691180 binding with energies of 8.4199 and 9.7176 kJ/mol, respectively. For ZINC4520996 complex, Lys53, Leu167, Asp168, and Phe169 contributed energies of ±5.0 kJ/mol to binding (Table 4 and Supplementary Figure S4b). Altogether, Lys53 was found to contribute high repulsive positive energies in the p38 MAPK-ZINC5519433, ZINC1691180, ZINC4520996, SB 203580 and SB 202190 complexes (Table 4), unfavorable for ligand binding in the ATP binding pocket. Phe169 contributed favorable energies <−5.0 kJ/mol in all complexes examined except ZINC95486106, corroborating its critical role in ligand binding as suggested in previous studies [87,110]. Figure 7. Molecular mechanics Poisson–Boltzmann surface area (MM-PBSA) plot of binding free energy contribution per-residue of ZINC5519433-p38 MAPK complex. Fluctuations of the residues Tyr35, Val38, Ala51, Lys53, Thr106, Leu108, Met109, and Phe169 are colored red. Biomolecules 2021, 11, 653 19 of 25 4. Summary and Potential Implication of the Study on COVID-19-Induced Cytokine Storm The binding of SARS-CoV-2 to the human ACE2 and cell entry has been shown to downregulate ACE2 [18,19], which in turn leads to uncontrolled inflammation [14]. ACE2 is known to convert angiotensin-II to angiotensin (1–7), which binds to the MAS receptor (MASR) to promote vasodilation, vascular protection, anti-fibrosis, anti-proliferation, anti- inflammation, and anti-angiogenesis [123]. Angiotensin-II signals pro-inflammatory, pro- vasoconstrictive, and pro-thrombotic activity through p38 MAPK activation, which is counter-balanced by Angiotensin (1–7) downregulation of p38 MAPK activity [14]. The downregulation of ACE2 leads to poor conversion of angiotensin -II to angiotensin (1–7). Studies on the complications of SARS-CoV-2 infection have reported higher levels of plasma cytokines, a condition known as the cytokine storm [12,13,124]. SARS-CoV was reported to upregulate p38 MAPK activity via a viral protein [20] and SARS-CoV-2 has also been suggested to employ a similar mechanism [14]. The p38 MAPK has thus been identified as critical in the COVID-19 cytokine storms [14,15]. Therefore, this study sought to identify potential p38 MAPK inhibitors which could be explored as useful in ameliorating the cytokine storms in severely ill COVID-19 pa- tients. Pharmacoinformatics-based approaches including molecular docking and dy- namics studies predicted four potential lead compounds with good binding affinity to the p38 MAPK protein and negligible toxicity. The predicted biological activities of the selected compounds showed that they possessed anti-inflammatory, kinase in- hibition, antiviral, platelet aggregation inhibition, and immunosuppression activities. Dexamethasone, which is an anti-inflammatory compound is used as a treatment op- tion in COVID-19 patients [26]. Moreover, Losmapimod, which is a p38 MAPK in- hibitor, is currently undergoing clinical trials for use as a treatment option in severely ill COVID-19 patients (https://clinicaltrials.gov/ct2/show/NCT04511819 (accessed on 8 April 2021)). Herein, ZINC5519433 was predicted as a platelet aggregation inhibitor and anti-rhinovirus. ZINC5733756 was also predicted as an anti-inflammatory, anti-influenza, and anti-rhinovirus. Moreover, ZINC1691180 and ZINC4520996 were predicted to be anti-inflammatory and antiviral. Interestingly, ZINC1691180 has previously been shown to elicit an anti-inflammatory response in mice [93]. The effects of the predicted compounds as potential inhibitors of p38 MAPK to possibly reduce the cytokine storm can be investi- gated experimentally. The plethora of predicted activities and mechanisms serve as clues that warrant further evaluation of these compounds. The molecules can be optimized as well as used for fragment-based de novo drug design of potential novel biotherapeutics. These predicted compounds can help fuel the pace of searching for effective antivirals with anti-inflammatory activity. 5. Conclusions The study used multistage computational techniques to predict four natural product- derived molecules comprising ZINC5733756, ZINC1691180, ZINC4520996 and ZINC5519433, which have the potential to inhibit the p38 MAPK. The compounds were selected based on high binding affinity and plausible binding mechanisms with the p38 MAPK protein structure obtained through molecular dynamics simulations including MM-PBSA. Key residues Tyr38, Ala51 and Phe169 were corroborated as critical for binding, which could guide the design and selection of future p38 MAPK inhibitors. Furthermore, these drug-like compounds were predicted as anti-inflammatory, kinase inhibitors, antiviral, platelet ag- gregation inhibitors, and immunosuppressive. Since the activation of the p38 MAPK leads to hyperinflammation in severely ill COVID-19 patients, the potential of these molecules to attenuate the cytokine storm can be explored. Further in vitro and in vivo evaluations of the suggested molecules could be undertaken to corroborate the predicted inhibitory activity. The study can serve as a clue for the design of novel p8α MAPK selective inhibitors which may have therapeutic implications in COVID-19-induced cytokine storm. Biomolecules 2021, 11, 653 20 of 25 Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/biom11050653/s1, Table S1: Predicted pharmacokinetics of the 18 selected hits and two inhibitors. The profiles used were gastrointestinal (GI) absorption, cytochrome P450 (CYP) 3A4 and permeability glycoprotein (P-gp) substrates. Also included is the number of violations of Lipinski’s rule of five, Table S2: Predicted inhibitory constants (Ki) and binding energy of the selected compounds and those of the known inhibitors. Also, included are the experimentally determined inhibition constants of known inhibitors., Figure S1: Snapshots at 25 ns interval (time step = 0, 25, 50, 75 and 100 ns) for the binding modes of the (a) SB 202190, and (b) SB 203580 ligand-p38 MAPK complexes, Figure S2: 2D representation of the protein-ligand complexes before MD for (a) ZINC1691180, (b) ZINC4520996, (c) ZINC95486106, (d) ZINC5519433, and (e) ZINC5733756-p38 MAPK complexes, Figure S3: 2D representation of the protein-ligand complexes at 25 ns, 50 ns, 75 ns and 100 ns for (a) ZINC1691180, (b) ZINC4520996, (c) ZINC95486106, (d) ZINC5519433 and (e) ZINC5733756 (f) SB 202190 and (g) SB 203580- p38 MAPK complexes, Figure S4: Molecular mechanics Poisson-Boltzmann surface area (MM-PBSA) plot of binding free energy contribution per-residue for (a) ZINC1691180, (b) ZINC4520996, (c) ZINC95486106, (d) ZINC5733756, (e) SB 202190, and (f) SB 203580-p38 MAPK complexes. Fluctuations of the residues Tyr35, Val38, Ala51, Lys53, Thr106, Leu108, Met109 and Phe169 are colored red. Author Contributions: M.D.W., S.K.K., and S.O.A. conceptualized and designed the study. S.O.A. performed the computational analysis with contributions from S.K.K., E.B. and M.D.W. S.O.A. and S.K.K. co-wrote the first draft. All authors have read and agreed to the published version of the manuscript. Funding: The project is not funded. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Not applicable. Acknowledgments: We used the high performance-computing system (Zuputo) at the West African Centre for Cell Biology of Infectious Pathogens (WACCBIP), University of Ghana. We are grateful for the technical support given by Kweku Enninful and Bismark Dankwa of the Parasitology Department of Noguchi Memorial Institute for Medical Research, University of Ghana. Conflicts of Interest: The authors declare no conflict of interest. References 1. Fisher, D.; Heymann, D. Q&A: The novel coronavirus outbreak causing COVID-19. BMC Med. 2020, 18, 1–13. [CrossRef] 2. Zhou, P.; Yang, X.L.; Wang, X.G.; Hu, B.; Zhang, L.; Zhang, W.; Si, H.R.; Zhu, Y.; Li, B.; Huang, C.L.; et al. A pneumonia outbreak associated with a new coronavirus of probable bat origin. Nature 2020, 579, 270–273. [CrossRef] [PubMed] 3. Rabi, A.F.; Al Zoubi, S.M.; Kasasbeh, A.G.; Salameh, M.D.; Al-Nasser, D.A. SARS-CoV-2 and Coronavirus Disease 2019: What We Know So Far. Pathogens 2020, 9, 231. [CrossRef] [PubMed] 4. Woo, P.C.Y.; Huang, Y.; Lau, S.K.P.; Yuen, K.-Y. Coronavirus genomics and bioinformatics analysis. Viruses 2010, 2, 1804–1820. [CrossRef] [PubMed] 5. Davies, N.G.; Klepac, P.; Liu, Y.; Prem, K.; Jit, M.; Pearson, C.A.B.; Quilty, B.J.; Kucharski, A.J.; Gibbs, H.; Clifford, S.; et al. Age-dependent effects in the transmission and control of COVID-19 epidemics. Nat. Med. 2020, 26, 1205–1211. [CrossRef] 6. Chaw, L.; Koh, W.C.; Jamaludin, S.A.; Naing, L.; Alikhan, M.F.; Wong, J. SARS-CoV-2 transmission in different settings: Analysis of cases and close contacts from the Tablighi cluster in Brunei Darussalam. MedRxiv 2020. [CrossRef] 7. Ding, Z.; Qian, H.; Xu, B.; Huang, Y.; Miao, T.; Yen, H.-L.; Xiao, S.; Cui, L.; Wu, X.; Shao, W.; et al. Toilets dominate environmental detection of SARS-CoV-2 virus in a hospital. MedRxiv 2020. [CrossRef] 8. Oreshkova, N.; Molenaar, R.-J.; Vreman, S.; Harders, F.; Munnink, B.B.O.; Honing, R.W.H.; Gerhards, N.; Tolsma, P.; Bouwstra, R.; Sikkema, R.; et al. SARS-CoV2 infection in farmed mink, Netherlands, April 2020. Biorxiv 2020. [CrossRef] 9. Cascella, M.; Rajnik, M.; Cuomo, A.; Dulebohn, S.; Di Napoli, R. Features, Evaluation and Treatment Coronavirus (COVID-19); Statpearls: Treasure Island, LS, USA, 2020. 10. Zhou, F.; Yu, T.; Du, R.; Fan, G.; Liu, Y.; Liu, Z.; Xiang, J.; Wang, Y.; Song, B.; Gu, X.; et al. Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: A retrospective cohort study. Lancet 2020, 6736, 1–9. [CrossRef] 11. Ye, Q.; Wang, B.; Mao, J. The pathogenesis and treatment of the ‘Cytokine Storm” in COVID-19. J. Infect. 2020, 80, 607–613. [CrossRef] 12. Liu, Y.; Zhang, C.; Huang, F.; Yang, Y.; Wang, F.; Yuan, J.; Zhang, Z.; Qin, Y.; Li, X.; Zhao, D.; et al. Elevated plasma levels of selective cytokines in COVID-19 patients reflect viral load and lung injury. Natl. Sci. Rev. 2020, 7, 1003–1011. [CrossRef] Biomolecules 2021, 11, 653 21 of 25 13. Sinha, P.; Matthay, M.; Calfee, C. Is a “Cytokine Storm” Relevant to COVID-19? JAMA Intern. Med. 2020, 180, 1152–1154. [CrossRef] 14. Grimes, J.M.; Grimes, K.V. p38 MAPK inhibition: A promising therapeutic approach for COVID-19. J. Mol. Cell. Cardiol. 2020, 144, 63–65. [CrossRef] 15. Wehbe, Z.; Hammoud, S.; Soudani, N.; Zaraket, H.; El-Yazbi, A.; Eid, A.H. Molecular Insights Into SARS COV-2 Interaction with Cardiovascular Disease: Role of RAAS and MAPK Signaling. Front. Pharmacol. 2020, 11, 836. [CrossRef] 16. Park, J.K.; Fischer, R.; Dechend, R.; Shagdarsuren, E.; Gapeljuk, A.; Wellner, M.; Meiners, S.; Gratze, P.; Al-Saadi, N.; Feldt, S.; et al. p38 mitogen-activated protein kinase inhibition ameliorates angiotensin II-induced target organ damage. Hypertension 2007, 49, 481–489. [CrossRef] 17. Yu, X.; Cui, L.; Hou, F.; Liu, X.; Wang, Y.; Wen, Y.; Chi, C.; Li, C.; Liu, R.; Yin, C. Angiotensin-converting enzyme 2-angiotensin (1-7)- Mas axis prevents pancreatic acinar cell inflammatory response via inhibition of the p38 mitogen-activated protein kinase/nuclear factor-?B pathway. Int. J. Mol. Med. 2018, 41, 409–420. [CrossRef] 18. Hoffmann, M.; Kleine-Weber, H.; Schroeder, S.; Krüger, N.; Herrler, T.; Erichsen, S.; Schiergens, T.S.; Herrler, G.; Wu, N.-H.; Nitsche, A.; et al. SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor. Cell 2020, 2, 271–280. [CrossRef] 19. Verdecchia, P.; Cavallini, C.; Spanevello, A.; Angeli, F. The pivotal link between ACE2 deficiency and SARS-CoV-2 infection. Eur. J. Intern. Med. 2020, 76, 14–20. [CrossRef] 20. Kopecky-Bromberg, S.A.; Martinez-Sobrido, L.; Palese, P. 7a Protein of Severe Acute Respiratory Syndrome Coronavirus Inhibits Cellular Protein Synthesis and Activates p38 Mitogen-Activated Protein Kinase. J. Virol. 2006, 80, 785–793. [CrossRef] 21. Börgeling, Y.; Schmolke, M.; Viemann, D.; Nordhoff, C.; Roth, J.; Ludwig, S. Inhibition of p38 mitogen-activated protein kinase impairs influenza virus-induced primary and secondary host gene responses and protects mice from lethal H5N1 infection. J. Biol. Chem. 2014, 289, 13–27. [CrossRef] 22. Jimenez-Guardeño, J.M.; Nieto-Torres, J.L.; DeDiego, M.L.; Regla-Nava, J.A.; Fernandez-Delgado, R.; Castaño-Rodriguez, C.; Enjuanes, L. The PDZ-Binding Motif of Severe Acute Respiratory Syndrome Coronavirus Envelope Protein Is a Determinant of Viral Pathogenesis. PLoS Pathog. 2014, 10, e1004320. [CrossRef] 23. Yokota, T.; Wang, Y. P38 MAP kinases in the heart. Gene 2016, 575, 369–376. [CrossRef] 24. Lee, J.K.; Kim, N.J. Recent advances in the inhibition of p38 MAPK as a potential strategy for the treatment of Alzheimer’s disease. Molecules 2017, 22, 1287. [CrossRef] 25. Grant, S.K. Therapeutic Protein Kinase Inhibitors. Cell. Mol. Life Sci. 2009, 66, 1163–1177. [CrossRef] 26. Sharun, K.; Tiwari, R.; Dhama, J.; Dhama, K. Dexamethasone to combat cytokine storm in COVID-19: Clinical trials and preliminary evidence. Int. J. Surg. 2020, 82, 179–181. [CrossRef] 27. Thorlund, K.; Dron, L.; Park, J.; Hsu, G.; Forrest, J.I.; Mills, E.J. A real-time dashboard of clinical trials for COVID-19. Lancet Digit. Health 2020, 2, e286–e287. [CrossRef] 28. Ciliberto, G.; Cardone, L. Boosting the arsenal against COVID-19 through computational drug repurposing. Drug Discov. Today 2020, 25, 946–948. [CrossRef] 29. Kwofie, S.K.; Broni, E.; Asiedu, S.O.; Kwarko, G.B.; Dankwa, B.; Enninful, K.S.; Tiburu, E.K.; Wilson, M.D. Cheminformatics-Based Identification of Potential Novel Anti-SARS-CoV-2 Natural Compounds of African Origin. Molecules 2021, 26, 406. [CrossRef] 30. Tian, D.; Liu, Y.; Liang, C.; Xin, L.; Xie, X.; Zhang, D.; Wan, M.; Li, H.; Fu, X.; Liu, H.; et al. An update review of emerging small-molecule therapeutic options for COVID-19. Biomed. Pharmacother. 2021, 137, 111313. [CrossRef] 31. Gaudêncio, S.P.; Pereira, F. A Computer-Aided Drug Design Approach to Predict Marine Drug-Like Leads for SARS-CoV-2 Main Protease Inhibition. Mar. Drugs 2020, 18, 633. [CrossRef] 32. Shehroz, M.; Zaheer, T.; Hussain, T. Computer-aided drug design against spike glycoprotein of SARS-CoV-2 to aid COVID-19 treatment. Heliyon 2020, 6, e05278. [CrossRef] [PubMed] 33. Yadav, M.; Dhagat, S.; Eswari, J.S. Emerging strategies on in silico drug development against COVID-19: Challenges and opportunities. Eur. J. Pharm. Sci. 2020, 155, 105522. [CrossRef] [PubMed] 34. Boozari, M.; Hosseinzadeh, H. Natural products for COVID-19 prevention and treatment regarding to previous coronavirus infections and novel studies. Phyther. Res. 2021, 35, 864–876. [CrossRef] [PubMed] 35. Nugraha, R.V.; Ridwansyah, H.; Ghozali, M.; Khairani, A.F.; Atik, N. Traditional Herbal Medicine Candidates as Complementary Treatments for COVID-19: A Review of Their Mechanisms, Pros and Cons. Evid. Based Complement. Altern. Med. 2020, 2020, 2560645. [CrossRef] 36. Akindele, A.J.; Agunbiade, F.O.; Sofidiya, M.O.; Awodele, O.; Sowemimo, A.; Ade-Ademilua, O.; Akinleye, M.O.; Ishola, I.O.; Orabueze, I.; Salu, O.B.; et al. COVID-19 Pandemic: A Case for Phytomedicines. Nat. Prod. Commun. 2020, 15, 1934578X2094508. [CrossRef] 37. Huang, J.; Tao, G.; Liu, J.; Cai, J.; Huang, Z.; Chen, J. Current Prevention of COVID-19: Natural Products and Herbal Medicine. Front. Pharmacol. 2020, 11, 1635. [CrossRef] 38. Cragg, G.M.; Newman, D.J. Natural products: A continuing source of novel drug leads. Biochim. Biophys. Acta Gen. Subj. 2013, 1830, 3670–3695. [CrossRef] 39. Ji, H.-F.; Li, X.-J.; Zhang, H.-Y. Natural products and drug discovery. Can thousands of years of ancient medical knowledge lead us to new and powerful drug combinations in the fight against cancer and dementia? EMBO Rep. 2009, 10, 194–200. [CrossRef] Biomolecules 2021, 11, 653 22 of 25 40. Chen, C.Y.-C.C. TCM Database@Taiwan: The world’s largest traditional Chinese medicine database for drug screening In Silico. PLoS ONE 2011, 6, e15939. [CrossRef] 41. Ntie-Kang, F.; Zofou, D.; Babiaka, S.B.; Meudom, R.; Scharfe, M.; Lifongo, L.L.; Mbah, J.A.; Mbaze, L.M.; Sippl, W.; Efange, S.M.N. AfroDb: A Select Highly Potent and Diverse Natural Product Library from African Medicinal Plants. PLoS ONE 2013, 8, e78085. [CrossRef] 42. Ntie-Kang, F.; Telukunta, K.K.; Döring, K.; Simoben, C.V.; Moumbock, A.F.A.; Malange, Y.I.; Njume, L.E.; Yong, J.N.; Sippl, W.; Günther, S. NANPDB: A Resource for Natural Products from Northern African Sources. J. Nat. Prod. 2017, 80, 2067–2076. [CrossRef] 43. Berman, H.M.; Westbrook, J.; Feng, Z.; Gilliland, G.; Bhat, T.N.; Weissig, H.; Shindyalov, I.N.; Bourne, P.E. The protein data bank. Nucleic Acids Res. 2000, 28, 235–242. [CrossRef] 44. Burley, S.K.; Berman, H.M.; Kleywegt, G.J.; Markley, J.L.; Nakamura, H.; Velankar, S. Protein Data Bank (PDB): The Single Global Macromolecular Structure Archive. In Methods in Molecular Biology; Humana Press Inc.: New York, NY, USA, 2017; Volume 1607, pp. 627–641. [CrossRef] 45. Azevedo, R.; Van Zeeland, M.; Raaijmakers, H.; Kazemier, B.; De Vlieg, J.; Oubrie, A. X-ray structure of p38 bound to TAK-715: Comparison with three classic inhibitors. Acta Crystallogr. Sect. D Biol. Crystallogr. 2012, 68, 1041–1050. [CrossRef] 46. Waterhouse, A.; Bertoni, M.; Bienert, S.; Studer, G.; Tauriello, G.; Gumienny, R.; Heer, F.T.; De Beer, T.A.P.; Rempfer, C.; Bordoli, L.; et al. SWISS-MODEL: Homology modelling of protein structures and complexes. Nucleic Acids Res. 2018, 46, W296–W303. [CrossRef] 47. DeLano, W.L. PyMOL: An open-source molecular graphics tool. CCP4 Newsl. Protein Crystallogr. 2002, 40, 82–92. [CrossRef] 48. Yuan, S.; Chan, H.C.S.; Hu, Z. Using PyMOL as a platform for computational drug design. Wiley Interdiscip. Rev. Comput. Mol. Sci. 2017, 7, e1298. [CrossRef] 49. Abraham, M.J.; Murtola, T.; Schulz, R.; Páll, S.; Smith, J.C.; Hess, B.; Lindah, E. Gromacs: High performance molecular simulations through multi-level parallelism from laptops to supercomputers. SoftwareX 2015, 1–2, 19–25. [CrossRef] 50. Jorgensen, W.L.; Maxwell, D.S.; Tirado-Rives, J. Development and testing of the OPLS all-atom force field on conformational energetics and properties of organic liquids. J. Am. Chem. Soc. 1996, 118, 11225–11236. [CrossRef] 51. Robertson, M.J.; Tirado-Rives, J.; Jorgensen, W.L. Improved Peptide and Protein Torsional Energetics with the OPLS-AA Force Field. J. Chem. Theory Comput. 2015, 11, 3499–3509. [CrossRef] 52. Gajula, M.; Kumar, A.; Ijaq, J. Protocol for Molecular Dynamics Simulations of Proteins. Bio-Protocol 2016, 6, 1–11. [CrossRef] 53. Mark, P.; Nilsson, L. Structure and dynamics of the TIP3P, SPC, and SPC/E water models at 298 K. J. Phys. Chem. A 2001, 105, 9954–9960. [CrossRef] 54. Berweger, C.D.; van Gunsteren, W.F.; Müller-Plathe, F. Force field parametrization by weak coupling. Re-engineering SPC water. Chem. Phys. Lett. 1995, 232, 429–436. [CrossRef] 55. Sander, T.; Freyss, J.; Von Korff, M.; Rufener, C. DataWarrior: An open-source program for chemistry aware data visualization and analysis. J. Chem. Inf. Model. 2015, 55, 460–473. [CrossRef] 56. Doytchinova, I.; Atanasova, M.; Valkova, I.; Stavrakov, G.; Philipova, I.; Zhivkova, Z.; Zheleva-Dimitrova, D.; Konstantinov, S.; Dimitrov, I. Novel hits for acetylcholinesterase inhibition derived by docking-based screening on ZINC database. J. Enzyme Inhib. Med. Chem. 2018, 33, 768–776. [CrossRef] [PubMed] 57. Kwofie, S.K.; Broni, E.; Teye, J.; Quansah, E.; Issah, I.; Wilson, M.D.; Miller, W.A.; Tiburu, E.K.; Bonney, J.H.K.K. Pharmacoinformatics-based identification of potential bioactive compounds against Ebola virus protein VP24. Comput. Biol. Med. 2019, 113, 103414. [CrossRef] [PubMed] 58. Goksuluk, D.; Korkmaz, S.; Zararsiz, G.; Karaagaoglu, A.E. easyROC: An interactive web-tool for ROC curve analysis using R language environment. R J. 2016, 8, 213–230. [CrossRef] 59. Mysinger, M.M.; Carchia, M.; Irwin, J.J.; Shoichet, B.K. Directory of useful decoys, enhanced (DUD-E): Better ligands and decoys for better benchmarking. J. Med. Chem. 2012, 55, 6582–6594. [CrossRef] 60. Trott, O.; Olson, A.J. AutoDock Vina: Improving the Speed and Accuracy of Docking with a New Scoring Function, EfficientOpti- mization, and Multithreading. J. Comput. Chem. 2010, 31, 455–461. [CrossRef] 61. O’Boyle, N.M.; Banck, M.; James, C.A.; Morley, C.; Vandermeersch, T.; Hutchison, G.R. Open Babel: An open chemical toolbox. J. Cheminform. 2011, 3, 33. [CrossRef] 62. Daina, A.; Michielin, O.; Zoete, V.; Brooks, C.L.; Huang, R. SwissADME: A free web tool to evaluate pharmacokinetics, drug- likeness and medicinal chemistry friendliness of small molecules. Sci. Rep. 2017, 7, 42717. [CrossRef] 63. Laskowski, R.A.; Swindells, M.B. LigPlot+: Multiple ligand-protein interaction diagrams for drug discovery. J. Chem. Inf. Model. 2011, 51, 2778–2786. [CrossRef] 64. Parasuraman, S. Prediction of activity spectra for substances. J. Pharmacol. Pharmacother. 2011, 2, 52–53. [CrossRef] 65. Wishart, D.S.; Knox, C.; Guo, A.C.; Cheng, D.; Shrivastava, S.; Tzur, D.; Gautam, B.; Hassanali, M. DrugBank: A knowledgebase for drugs, drug actions and drug targets. Nucleic Acids Res. 2008, 36, D901–D906. [CrossRef] 66. van Gunsteren, W.F. Biomolecular Simulation: The GROMOS96 Manual and User Guide; Biomos: Zürich, Switzerland, 1996; ISBN 9783728124227. 67. Schüttelkopf, A.W.; van Aalten, D.M.F. PRODRG: A tool for high-throughput crystallography of protein–ligand complexes. Acta Crystallogr. Sect. D Biol. Crystallogr. 2004, 60, 1355–1363. [CrossRef] Biomolecules 2021, 11, 653 23 of 25 68. Kumari, R.; Kumar, R.; Consortium, O.S.D.D.; Lynn, A. g_mmpbsa—A GROMACS tool for MM-PBSA and its optimization for high-throughput binding energy calculations. J. Chem. Inf. Model. 2014, 54, 1951–1962. [CrossRef] 69. Cuenda, A.; Rousseau, S. p38 MAP-Kinases pathway regulation, function and role in human diseases. Biochim. Biophys. Acta—Mol. Cell Res. 2007, 1773, 1358–1375. [CrossRef] [PubMed] 70. Li, K.; Li, Z.; Tao, Y.; Wang, Q.; Lai, Y.; Wu, W.; Peng, S.; Guo, Z.; Huang, H. Discovering novel P38α inhibitors for the treatment of prostate cancer through virtual screening methods. Future Med. Chem. 2019, 11, 3125–3137. [CrossRef] 71. Wang, Z.; Harkins, P.C.; Ulevitch, R.J.; Han, J.; Cobb, M.H.; Goldsmith, E.J. The structure of mitogen-activated protein kinase p38 at 2.1-Å resolution. Proc. Natl. Acad. Sci. USA 1997, 94, 2327–2332. [CrossRef] [PubMed] 72. Tian, W.; Chen, C.; Lei, X.; Zhao, J.; Liang, J. CASTp 3.0: Computed atlas of surface topography of proteins. Nucleic Acids Res. 2018, 46, W363–W367. [CrossRef] 73. Empereur-mot, C.; Guillemain, H.; Latouche, A.; Zagury, J.-F.; Viallon, V.; Montes, M. Predictiveness curves in virtual screening. J. Cheminform. 2015, 7, 52. [CrossRef] 74. Wang, S.; Jiang, J.H.; Li, R.Y.; Deng, P. Docking-based virtual screening of TβR1 inhibitors: Evaluation of pose prediction and scoring functions. BMC Chem. 2020, 14, 52. [CrossRef] [PubMed] 75. Mandrekar, J.N. Receiver Operating Characteristic Curve in Diagnostic Test Assessment. J. Thorac. Oncol. 2010, 5, 1315–1316. [CrossRef] [PubMed] 76. Hsin, K.Y.; Matsuoka, Y.; Asai, Y.; Kamiyoshi, K.; Watanabe, T.; Kawaoka, Y.; Kitano, H. systemsDock: A web server for network pharmacology-based prediction and analysis. Nucleic Acids Res. 2016, 44, W507–W513. [CrossRef] [PubMed] 77. Ngo, S.T.; Tam, N.M.; Pham, M.Q.; Nguyen, T.H. Benchmark of Popular Free Energy Approaches Revealing the Inhibitors Binding to SARS-CoV-2 Mpro. J. Chem. Inf. Model. 2021. [CrossRef] 78. Mishra, A.; Dey, S. Molecular docking studies of a cyclic octapeptide-cyclosaplin from sandalwood. Biomolecules 2019, 9, 740. [CrossRef] 79. Santana, I.V.; Côrtes Filho, A.B.; Pinheiro, A.A.F.; dos Santos, É.G.; Lima, D.M.; Barreto, M.M.; Lima, E.R.; Valasques Junior, G.L. In silico screening of compounds from the Bahia semiarid region for identification of potential inhibitors of the p38 MAPK protein. Res. Soc. Dev. 2020, 9, e4439108723. [CrossRef] 80. Chandrasekaran, B.; Abed, S.N.; Al-Attraqchi, O.; Kuche, K.; Tekade, R.K. Computer-Aided Prediction of Pharmacokinetic (ADMET) Properties. In Dosage Form Design Parameters; Elsevier: Amsterdam, The Netherlands, 2018; Volume 2, pp. 731–755. ISBN 9780128144220. 81. Elmeliegy, M.; Vourvahis, M.; Guo, C.; Wang, D.D. Effect of P-glycoprotein (P-gp) Inducers on Exposure of P-gp Substrates: Review of Clinical Drug–Drug Interaction Studies. Clin. Pharmacokinet. 2020, 59, 699–714. [CrossRef] 82. Kwon, H.; Lionberger, R.A.; Yu, L.X. Impact of P-glycoprotein-mediated intestinal efflux kinetics on oral bioavailability of P-glycoprotein substrates. Mol. Pharm. 2004, 1, 455–465. [CrossRef] 83. Klein, K.; Zanger, U.M. Pharmacogenomics of Cytochrome P450 3A4: Recent Progress toward the “Missing Heritability” Problem. Front. Genet. 2013, 4, 12. [CrossRef] 84. Du, X.; Li, Y.; Xia, Y.-L.; Ai, S.-M.; Liang, J.; Sang, P.; Ji, X.-L.; Liu, S.-Q. Insights into Protein-Ligand Interactions: Mechanisms, Models, and Methods. Int. J. Mol. Sci. 2016, 17, 144. [CrossRef] 85. Wang, Z.; Canagarajah, B.J.; Boehm, J.C.; Kassisà, S.; Cobb, M.H.; Young, P.R.; Abdel-Meguid, S.; Adams, J.L.; Goldsmith, E.J. Structural basis of inhibitor selectivity in MAP kinases. Structure 1998, 6, 1117–1128. [CrossRef] 86. Tong, L.; Pav, S.; White, D.M.; Rogers, S.; Crane, K.M.; Cywin, C.L.; Brown, M.L.; Pargellis, C.A. A highly specific inhibitor of human p38 MAP kinase binds in the ATP pocket. Nat. Struct. Biol. 1997, 4, 311–316. [CrossRef] 87. Bukhtiyarova, M.; Karpusas, M.; Northrop, K.; Namboodiri, H.V.M.; Springman, E.B. Mutagenesis of p38α MAP kinase establishes key roles of Phe169 in function and structural dynamics and reveals a novel DFG-OUT state. Biochemistry 2007, 46, 5687–5696. [CrossRef] 88. Jamkhande, P.G.; Barde, S.R. Evaluation of anthelmintic activity and in silico PASS assisted prediction of Cordia dichotoma (Forst.) root extract. Anc. Sci. Life 2014, 34, 39–43. [CrossRef] 89. Patil, K.R.; Mohapatra, P.; Patel, H.M.; Goyal, S.N.; Ojha, S.; Kundu, C.N.; Patil, C.R. Pentacyclic triterpenoids inhibit ikkβmediated activation of nf-κb pathway: In silico and in vitro evidences. PLoS ONE 2015, 10, e0125709. [CrossRef] 90. Emon, N.U.; Jahan, I.; Sayeed, M.A. Investigation of antinociceptive, anti-inflammatory and thrombolytic activity of Caesalpinia digyna (Rottl.) leaves by experimental and computational approaches. Adv. Tradit. Med. 2020, 20, 451–459. [CrossRef] 91. Morrell, C.N.; Aggrey, A.A.; Chapman, L.M.; Modjeski, K.L. Emerging roles for platelets as immune and inflammatory cells. Blood 2014, 123, 2759–2767. [CrossRef] 92. Levi, M.; Thachil, J. Coronavirus Disease 2019 Coagulopathy: Disseminated Intravascular Coagulation and Thrombotic Microangiopathy—Either, Neither, or Both. In Seminars in Thrombosis and Hemostasis; Thieme Medical Publishers: New York, NY, USA, 2020. [CrossRef] 93. Yoshioka, H.; Mizuno, Y.; Yamaguchi, T.; Ichimaru, Y.; Takeya, K.; Hitotsuyanagi, Y.; Nonogaki, T.; Aoyagi, Y. Methyl dehydroabi- etate counters high fat diet-induced insulin resistance and hepatic steatosis by modulating peroxisome proliferator-activated receptor signaling in mice. Biomed. Pharmacother. 2018, 99, 214–219. [CrossRef] 94. Hughes, J.P.; Rees, S.S.; Kalindjian, S.B.; Philpott, K.L. Principles of early drug discovery. Br. J. Pharmacol. 2011, 162, 1239–1249. [CrossRef] Biomolecules 2021, 11, 653 24 of 25 95. Wishart, D.S.; Feunang, Y.D.; Guo, A.C.; Lo, E.J.; Marcu, A.; Grant, J.R.; Sajed, T.; Johnson, D.; Li, C.; Sayeeda, Z.; et al. DrugBank 5.0: A major update to the DrugBank database for 2018. Nucleic Acids Res. 2018, 46, D1074–D1082. [CrossRef] 96. Ma, J.; Hwang, Y.K.; Cho, W.H.; Han, S.H.; Hwang, J.K.; Han, J.S. Macelignan attenuates activations of mitogen-activated protein kinases and nuclear factor kappa B induced by lipopolysaccharide in microglial cells. Biol. Pharm. Bull. 2009, 32, 1085–1090. [CrossRef] [PubMed] 97. Kaoud, T.S.; Park, H.; Mitra, S.; Yan, C.; Tseng, C.C.; Shi, Y.; Jose, J.; Taliaferro, J.M.; Lee, K.; Ren, P.; et al. Manipulating JNK signaling with (-)-zuonin A. ACS Chem. Biol. 2012, 7, 1873–1883. [CrossRef] [PubMed] 98. Maruca, A.; Ambrosio, F.A.; Lupia, A.; Romeo, I.; Rocca, R.; Moraca, F.; Talarico, C.; Bagetta, D.; Catalano, R.; Costa, G.; et al. Computer-based techniques for lead identification and optimization i: Basics. Phys. Sci. Rev. 2019, 4. [CrossRef] 99. Zhou, F.H.; Foster, B.K.; Zhou, X.-F.; Cowin, A.J.; Xian, C.J. TNF-alpha mediates p38 MAP kinase activation and negatively regulates bone formation at the injured growth plate in rats. J. Bone Miner. Res. Off. J. Am. Soc. Bone Miner. Res. 2006, 21, 1075–1088. [CrossRef] [PubMed] 100. Song, C.-H.; Kim, N.; Kim, D.-H.; Lee, H.-N.; Surh, Y.-J. 17-β estradiol exerts anti-inflammatory effects through activation of Nrf2 in mouse embryonic fibroblasts. PLoS ONE 2019, 14, e0221650. [CrossRef] [PubMed] 101. Islam, M.A.; Pillay, T.S. Identification of promising anti-DNA gyrase antibacterial compounds using de novo design, molecular docking and molecular dynamics studies. J. Biomol. Struct. Dyn. 2019, 6, 1798–1809. [CrossRef] 102. Reynolds, C.H.; Reynolds, R.C. Group Additivity in Ligand Binding Affinity: An Alternative Approach to Ligand Efficiency. J. Chem. Inf. Model. 2017, 57, 3086–3093. [CrossRef] 103. Kumar, S.; McDonnell, P.C.; Gum, R.J.; Hand, A.T.; Lee, J.C.; Young, P.R. Novel homologues of CSBP/p38 MAP kinase: Activation, substrate specificity and sensitivity to inhibition by pyridinyl imidazoles. Biochem. Biophys. Res. Commun. 1997, 235, 533–538. [CrossRef] 104. Fehr, S.; Unger, A.; Schaeffeler, E.; Herrmann, S.; Laufer, S.; Schwab, M.; Albrecht, W. Impact of p38 MAP kinase inhibitors on LPS-induced release of TNF-α in whole blood and primary cells from different species. Cell. Physiol. Biochem. 2015, 36, 2237–2249. [CrossRef] 105. Koeberle, S.C.; Romir, J.; Fischer, S.; Koeberle, A.; Schattel, V.; Albrecht, W.; Grütter, C.; Werz, O.; Rauh, D.; Stehle, T.; et al. Skepinone-L is a selective p38 mitogen-activated protein kinase inhibitor. Nat. Chem. Biol. 2012, 8, 141–143. [CrossRef] 106. Goldstein, D.M.; Kuglstatter, A.; Lou, Y.; Soth, M.J. Selective p38α inhibitors clinically evaluated for the treatment of chronic inflammatory disorders. J. Med. Chem. 2010, 53, 2345–2353. [CrossRef] 107. Hope, H.R.; Anderson, G.D.; Burnette, B.L.; Compton, R.P.; Devraj, R.V.; Hirsch, J.L.; Keith, R.H.; Li, X.; Mbalaviele, G.; Messing, D.M.; et al. Anti-inflammatory properties of a novel N-phenyl pyridinone inhibitor of p38 mitogen-activated protein kinase: Preclinical-to-clinical translation. J. Pharmacol. Exp. Ther. 2009, 331, 882–895. [CrossRef] 108. Mazumder, M.; Ponnan, P.; Das, U.; Gourinath, S.; Khan, H.A.; Yang, J.; Sakharkar, M.K. Investigations on Binding Pattern of Kinase Inhibitors with PPAR γ: Molecular Docking, Molecular Dynamic Simulations, and Free Energy Calculation Studies. PPAR Res. 2017, 2017, 1–11. [CrossRef] 109. Musyoka, T.M.; Kanzi, A.M.; Lobb, K.A.; Tastan Bishop, Ö. Structure Based Docking and Molecular Dynamic Studies of Plasmodial Cysteine Proteases against a South African Natural Compound and its Analogs. Sci. Rep. 2016, 6, 1–12. [CrossRef] 110. Yang, Y.; Shen, Y.; Liu, H.; Yao, X. Molecular dynamics simulation and free energy calculation studies of the binding mechanism of allosteric inhibitors with p38α MAP kinase. J. Chem. Inf. Model. 2011, 51, 3235–3246. [CrossRef] 111. Khan, M.F.; Verma, G.; Alam, P.; Akhter, M.; Bakht, M.A.; Hasan, S.M.; Shaquiquzzaman, M.; Alam, M.M. Dibenzepinones, dibenzoxepines and benzosuberones based p38α MAP kinase inhibitors: Their pharmacophore modelling, 3D-QSAR and docking studies. Comput. Biol. Med. 2019, 110, 175–185. [CrossRef] 112. Suplatov, D.; Kopylov, K.; Sharapova, Y.; Švedas, V. Human p38α mitogen-activated protein kinase in the Asp168-Phe169-Gly170- in (DFG-in) state can bind allosteric inhibitor Doramapimod. J. Biomol. Struct. Dyn. 2019, 37, 2049–2060. [CrossRef] 113. Tabassum, H.; Ahmad, I.Z. Molecular Docking and Dynamics Simulation Analysis of Thymoquinone and Thymol Com- pounds from Nigella sativa L. that Inhibits P38 Protein: Probable Remedies for Hepatocellular Carcinoma. Med. Chem. 2019, 16, 350–357. [CrossRef] 114. Lobanov, M.I.; Bogatyreva, N.S.; Galzitskaia, O. V Radius of gyration is indicator of compactness of protein structure. Mol. Biol. 2008, 42, 701–706. [CrossRef] 115. Boateng, R.A.; Tastan Bishop, Ö.; Musyoka, T.M. Characterisation of plasmodial transketolases and identification of potential inhibitors: An in silico study. Malar. J. 2020, 19, 1–19. [CrossRef] 116. Wen, C.C.; Kuo, Y.H.; Jan, J.T.; Liang, P.H.; Wang, S.Y.; Liu, H.G.; Lee, C.K.; Chang, S.T.; Kuo, C.J.; Lee, S.S.; et al. Specific plant terpenoids and lignoids possess potent antiviral activities against severe acute respiratory syndrome coronavirus. J. Med. Chem. 2007, 50, 4087–4095. [CrossRef] [PubMed] 117. Wang, C.; Greene, D.; Xiao, L.; Qi, R.; Luo, R. Recent Developments and Applications of the MMPBSA Method. Front. Mol. Biosci. 2018, 4, 87. [CrossRef] [PubMed] 118. Gupta, S.; Lynn, A.M.; Gupta, V. Standardization of virtual-screening and post-processing protocols relevant to in-silico drug discovery. 3 Biotech 2018, 8, 1–7. [CrossRef] [PubMed] Biomolecules 2021, 11, 653 25 of 25 119. Macchiagodena, M.; Pagliai, M.; Karrenbrock, M.; Guarnieri, G.; Iannone, F.; Procacci, P. Virtual Double-System Single-Box: A Nonequilibrium Alchemical Technique for Absolute Binding Free Energy Calculations: Application to Ligands of the SARS-CoV-2 Main Protease. J. Chem. Theory Comput. 2020, 16, 7160–7172. [CrossRef] 120. Deng, N.J.; Zhang, P.; Cieplak, P.; Lai, L. Elucidating the energetics of entropically driven protein-ligand association: Calculations of absolute binding free energy and entropy. J. Phys. Chem. B 2011, 115, 11902–11910. [CrossRef] 121. Campanera, J.M.; Pouplana, R. MMPBSA decomposition of the binding energy throughout a molecular dynamics simulation of amyloid-beta (Aß10-35) aggregation. Molecules 2010, 15, 2730–2748. [CrossRef] 122. Martins, L.C.; Torres, P.H.M.; de Oliveira, R.B.; Pascutti, P.G.; Cino, E.A.; Ferreira, R.S. Investigation of the binding mode of a novel cruzain inhibitor by docking, molecular dynamics, ab initio and MM/PBSA calculations. J. Comput. Aided. Mol. Des. 2018, 32, 591–605. [CrossRef] 123. Medina-Enríquez, M.M.; Lopez-León, S.; Carlos-Escalante, J.A.; Aponte-Torres, Z.; Cuapio, A.; Wegman-Ostrosky, T. ACE2: The molecular doorway to SARS-CoV-2. Cell Biosci. 2020, 10, 1–17. [CrossRef] 124. Mendoza-Pinto, C.; García-Carrasco, M.; Munguía Realpozo, P.; Méndez-Martínez, S. Therapeutic options for the management of severe COVID-19: A rheumatology perspective. Reumatol. Clín. 2020. [CrossRef]