Dinamični model encima MurD iz E. coli in in silico načrtovanje novih inhibitorjev : doktorska disertacija

Perdih, Andrej (author), Šolmajer, Tomaž (mentor)

V zadnjih letih opažamo precejšen porast identificiranih patoloških bakterijskih sevov, ki so odporni na večino znanih protibakterijskih zdravilnih učinkovin. Tako zaskrbljujoče stanje kaže na nujnost pospešenega iskanja novih protibakterijskih učinkovin v vlogi učinkovitega orožja za boj proti patološkim prokariontskim organizmom. Med mnogimi že znanimi in na novo identificiranimi potencialnimi tarčami so encimi, ki sodelujejo v biosintezi bakterijskega peptidoglikana, že dlje časa zanimive tarče zaradi možnosti doseganja selektivne toksičnosti. Sinteza peptidnega dela monomerne enote peptidoglikana poteka intracelularno preko sukcesivnega dodajanja aminokislin na začetni UDP-prekurzor (UDP-MurNAc), ki ga katalizirajo visoko specifični encimi, imenovani Mur sintetaze (MurC, MurD, MurE in MurF). Med njimi encim MurD (UDP-N-acetilmuramil-L-Ala:D-Glu ligaza) katalizira reakcijo tvorbe peptidne vezi med D-glutamatom in uridindifosfatnim (UDP) prekurzorjem UDP-MurNAc-L-Ala (UMA), ki je sočasno sklopljena s hidrolizo energijske molekule ATP od ADP in prostega fosfata Pi.V doktorskem delu smo z metodami molekulskega modeliranja postavili kompleksen dinamični model encima MurD iz bakterijske vrste E. coli, ki povezuje do sedaj znana eksperimentalna spoznanja o tej protibakterijski tarči, in z in silico metodami strukturno podprtega načrtovanja zdravilnih učinkovin identificirali nove inhibitorje Mur ligaz. Z uporabo tarčne molekulske dinamike (TMD) smo, izhajajoč iz objavljenih eksperimentalno določenih odprtih in zaprtih struktur encima, postavili dinamične modele vezave obeh substratov (ATP in UMA) v njuni aktivni mesti in analizirali zapiranje C-terminalne domene MurD. Poleg reproduciranega zaporedja vezave substratov, so TMD simulacije pokazale, da v obratu C-terminalne domene sodeluje več aminokislin. Med njimi še posebej izstopata aminokislinska preostanka Pro300 in Arg302. Simulacije »off-path« tehnike (OPS), ki so razširitev že uveljavljene metode replik (RPATh), smo uporabili za določitev energetike zapiranja C-terminalne domene, ki smo jo dobili s pomočjo TMD simulacij. Rezultati so pokazali, da je prehod iz odprte MurD strukture 1EEH v zaprto 2UAG energijsko bistveno zahtevnejši od prehoda, izhajajočega iz odprte 1E0D strukture. Gibanja Cterminalne domene, ki so omejena na ravnino centralne in N-terminalne domene, so energijsko bistveno manj zahtevna kot gibanje iz te ravnine. S hibridnimi kvantnomehansko/molekulskomehanskimi (QM/MM) metodami smo študirali tri možne mehanizme nastanka tetraedričnega intermediata encima MurD, ki pogosto predstavlja začetno strukturo pri razvoju MurD inhibitorjev. Reakcijske poti med optimiziranimi začetnimi strukturami, konstruiranimi na osnovi dostopnih eksperimentalnih podatkov in tetraedričnimi intermediati smo določili s QM/MM različico metode replik (RPATh). Dobljeni molekularni modeli encimske reakcije so v skladu z ekspermentalnim reakcijskim vrstnim redom in kažejo, da v reakciji najprej reagira substrat UMA z ATP molekulo v acilfosfatni intermediat in da se D-Glutaminska kislina najverjetneje pred reakcijo z acil-fosfatnim intermediatom deprotonira. Simulacije proste energije vezave, izvedene z metodo linearne interakcijske energije (LIE) v seriji Nsulfonilnih derivatov glutaminske kisiline, so reproducirale eksperimentalno določene vrednosti DGvezave. Rezultati simulacij kažejo, da najpomembnejšo vlogo pri vezavi teh inhibitorjev igrajo van der Waalsove interakcije. Energijska analiza prispevkov posameznih delov inhibitorjev (glutaminska kislina, naftalenski del, sulfonamidni del in lipofilni rep) k celotni vezavi je kvantificirala pomen posameznih strukturnih fragmentov, ki sestavljajo ta razred spojin. Izhajajoč iz dostopnih strukturnih podatkov za encima MurD in MurE (oba iz E. coli) smo razvili protokol virtualnega rešetanja (VS), ki je s kombinacijo trodimenzionalnih farmakoforov in molekulskega sidranja služil preiskovanju velikih knjižnic virtualnih molekul. Identificirali smo nov razred dualnih inhibitorjev encimov MurD in MurE, ki kot možen nadomestek za glutaminsko kislino, vsebujejo benzen 1,3- dikarboksilatni fragment.

• Language: Slovenian
• Keywords: farmacevtska kemija, protibakterijske učinkovine, bakterijska rezistenca, tarče, bakterije, celična stena, peptidoglikan, biosinteza, encimi Mur, ligaze Mur, MurD, encimski inhibitorji, računalniško podprto načrtovanje, molekularno modeliranje, dinamični modeli, Escherichia coli, disertacije
• Work type: Dissertation
• Typology: 2.08 - Doctoral Dissertation
• Organization: FFA - Faculty of Pharmacy
• Place of publishing: Ljubljana
• Publisher: [A. Perdih]
• Year: 2009
• Number of pages: 166 str.
• PID: 20.500.12556/RUL-127048
• UDC: 615.281.9:004.4(043.3)
• COBISS.SI-ID: 245586688
• Publication date in RUL: 14.05.2021

Secondary language

• Language: English
• Title: Dynamical model of the E. coli MurD enzyme and in silico design of novel inhibitors

Abstract

The increasing incidence of bacterial resistance to most available antibiotics has made the discovery of novel efficacious antibacterial agents urgent. In this process, previously unexploited targets are being considered. As an essential bacterial component unique to prokaryotic cells, peptidoglycan is traditionally an optimal target with respect to selective toxicity. Four ADP-forming bacterial ligases – MurC, MurD, MurE, and MurF – are involved in the intracellular phase of peptidoglycan assembly, catalyzing the synthesis of a peptide moiety by consecutive addition of the amino acids to the starting UDP-precursor (UDP-MurNAc). Among them the MurD (UDP-N-acetylmuramoyl-L-alanine:Dglutamate ligase) catalyses a highly specific incorporation of the D-glutamate into the cytoplasmic intermediate UDP-N-acetyl-muramoyl-L-alanine (UMA) utilizing ATP hydrolysis to ADP and Pi. In this work by using various molecular modeling methodologies a complex dynamical model for the MurD enzyme from the E. coli bacterial species was derived at the atomic level, taking into consideration all the available experimental observations. Furthermore, in silico structure-based drug design approach was utilized leading to the identification of novel inhibitors of Mur ligase family. Targeted nanosecond molecular dynamics (TMD) simulations were performed in order to examine the substrate (ATP and UMA) binding process and gain insight into structural changes that occur during the conformational closure of the MurD C-terminal domain into the active conformation. The experimentally determined binding order of the substrates was reproduced by TMD simulations. The key interactions essential for the conformational transitions (amino acid residues: Pro300 in Ar302) and substrate binding were identified. Off-path simulation (OPS) technique, an extension of the established Replica path method (RPATh), was initiated to evaluate the energy pathway of the two TMD-generated C-terminal domain closing pathways. The study established much higher energy demands if the C-terminal domain closing process commenced from the open structure in which this domain is located out-of plane with respect to the N-terminal and central domains (open structure 1EEH) in comparison to the open structure in which the conformational movement is confined to this plane (open structure 1E0D). A hybrid quantum mechanical/molecular mechanical (QM/MM) molecular modeling approach was utilized, to evaluate three possible reaction pathways leading to the tetrahedral intermediate formation – a frequent drug design starting point. Geometries of the starting structures based on crystallographic experimental data and tetrahedral intermediates were carefully examined together with a role of crucial amino acids and water molecules. The QM/MM replica path method (RPATh) was used to generate the reaction pathways between the starting structures and the corresponding tetrahedral reaction intermediates, producing reaction pathways which were in agreement with in a sequential kinetic mechanism. The D-Glu moiety most likely enters the enzyme reaction in its deprotonated form. Binding free energies were calculated for a series of MurD N-sulphonyl-glutamic acid inhibitors using the Linear Interaction Energy (LIE) method. Analysis of interaction energy revealed non-polar van der Waals interactions as the main driving force for the binding of these inhibitors, and excellent agreement with the experimental free energies was obtained. Analysis of fragment contribution to binding free energies for selected inhibitor moieties (glutamic acid, sulphone amide group, naphthalene moiety and lipophilic tail) in this structural class substantiated the insight into the source of inhibitory activity. Based on the available structural data for the MurD and MurE enzymes (both from E. coli) a virtual screening campaign was performed, combining three-dimensional structure-based pharmacophores and molecular docking calculations, resulting in the identification of a novel class of glutamic acid surrogates - benzene 1,3-dicarboxylic acid derivatives possessing dual MurD and MurE inhibitory activity.