Multidimensional computational modeling of Potent BACE1 (β-Secretase) inhibitors towards Alzheimer’s disease treatment.

Abstract

Alzheimer’s disease (AD), as a progressive multifactorial neurodegenerative abnormality of the brain, is often connected with loss or death of neurons as its primary pathogenesis. Another kind of dementia is associated with memory loss and unstable and irrational behaviors, especially among the elderly above 60 years. In South Africa, there are over four million people above the age of 60 years, with an approximation of one hundred and eighty-seven thousand living with dementia. The two distinguishing features (hallmarks) of AD are neurofibrillary tangles and β-amyloid plaques. The β-amyloid plaques result when amyloid precursor protein (APP) is cleaved by β-amyloid precursor protein cleaving enzyme1 (BACE1), otherwise known as β-secretase. Since 1999 the first BACE1 was discovered, it has become a major interest in attempting to develop drugs for the inhibition or reduction of the β-amyloid aggregates in the brain. Reducing or inhibiting the accumulation of β-amyloid has long been the target in the design of drugs for AD treatment. Having a good knowledge of the characteristic properties (BACE1) would assist in the design of potent selective BACE1 inhibitors with fewer or no side effects. Hitherto, only five drugs have been approved by the Food and Drug Administration (FDA) for the remediation of Alzheimer’s disease, and none of the approved drugs targets BACE1. In about twenty years of its discovery, several past and ongoing studies have focused on BACE1 therapeutic roles as a target in managing AD. Several attempts have previously beenmade in designing some small drugmolecules capable of good BACE1 inhibition. Some of the initially discovered BACE1 inhibitors include verubecestat, lanabecestat, atabecestat, and umibecestat (CNP-520). Although these inhibitors significantly lowered β-amyloid plaques in persons having neurological Alzheimer’s at its clinical trials (phase 3), they were suddenly terminated for some health concerns. The termination contributed to the reasons why there are insufficient BACE-targeted drugs for AD treatment. Lately, a novel potent, orally effective, and highly selective AM-6494 BACE1 inhibitor was discovered. This novel BACE1 inhibitor exhibited no fur coloration and common skin alteration, as observed with some initial BACE1 inhibitors. AM-6494 with an IC50 value of 0.4 nM in vivo is presently selected and at the preclinical phase trials. Before this study, the inhibition properties of this novel BACE1 inhibitor at the atomistic and molecular level of BACE1 inhibition remained very unclear. The first manuscript (chapter two) is a literature review on Alzheimer's disease and β-secretase inhibition: An update focusing on computer-aided inhibitor design. We provide an introductory background of the subject with a brief discussion on Alzheimer’s pathology. The review features computational methods involved in designing BACE1 inhibitors including the discontinued drugs. Using the topical keywords BACE1, inhibitor design, and computational/theoretical study in theWeb of Science and Scopus database, we retrieved over 49 relevant articles. The search years are from 2010 and 2020, with analysis conducted from May 2020 to March 2021. Our second manuscript (chapter three) reviewed BACE1 exosite-binding antibody and allosteric inhibition as an alternative therapeutic development. We studied BACE1 biological functions, the pathogenesis of the associated diseases, and the enzymatic properties of the APP site cleavage. We suggested an extensive application of advanced computational simulations in the investigation of anti-BACE1 body and allosteric exosites. It is believed that this investigation will further help in reducing the associated challenges with designing BACE1 inhibitors while exploring the opportunities in the design of allosteric antibodies. The review also revealed that some molecules exhibited dual binding sites at the active site and allosteric site. As a result, we recommend an extensive investigation of the binding free energy beyond molecular docking (such as advanced molecular dynamic simulations) as this promises to reveal the actual binding site for the compounds under investigation. Chapter four contains the detailed computational science techniques which cover the application of the vitally essential methods of molecular mechanics (MM), quantum mechanics (QM), hybrid of QM/MM, basis sets, and other computational instruments employed in this study. In the third manuscript (chapter five), we carried out computational simulations of AM-6494 and CNP- 520.CNP520 was one of the earliest BACE1 drugs that were terminated, chosen in this study forcomparative reasons. This simulation was to elucidate and understand the binding affinities of these two inhibitors at the atomistic level. We explored the quantum mechanics (QM) density functional theory (DFT) and hybrid QM/MM of Our Own N-layered Integrated molecular Orbital and Molecular Mechanics (ONIOM) in these simulations. These computational approaches helped in predicting the electronic properties of AM-6494 and CNP-520, including their binding energies when in complex with BACE1. Considering the debates on which protonated forms of Asp 32 and Asp 288 gives a more favorable binding energy, we analysed the two forms which involved the protonation and un-protonation of Asp 32 and Asp 228.The ONIOM protonated model calculation gave binding free energy of -33.463 kcal/mol (CNP-520)and 62.849 kcal/mol (AM-6494) while the binding free energy of -59.758 kcal/mol was observed for the unprotonated AM-6494 model. These results show the protonated model as a more favourable binding free energy when compared with the un-protonation AM-6494 model. Further thermochemistry processes coupled with molecular interaction plots indicate that AM-6494 has better inhibition properties thanCNP-520.However, it was observed that the protonation and the un-protonation of Asp 32 and Asp 228 modelscould adequately illustrate the interatomic binding of the ligands-BACE1 complex. To further explicate the binding mechanism, conformational and structural dynamism of AM-6494 relative to CNP-520 in complex with BACE1, we carried out advanced computational simulations in the fourth manuscript (chapter six). The extensive application of accelerated molecular dynamics simulations, as well as principal component analysis, were involved. From the results, AM-6494 further exhibited higher binding affinity with van der Waals as the predominant contributing energy relative to CNP-520. Furthermore, conformational analysis of the β-hairpin (flap) within the BACE1 active site exhibited efficient closed flap conformations in complex withAM-6494 relative to CNP-520, whichmostly alternated between closed and semi-open conformational dynamics. These observations further elucidate that AM- 6494 shows higher inhibitory potential towards BACE1. The catalytic dyad (Asp32/228), Tyr14, Leu30, Tyr71, and Gly230 constitute essential residues in both AM-6494 potencies CNP-520 at the BACE1 binding interface. The results from these extensive computational simulations and analysis undoubtedly elucidate AM-6494 higher inhibition potentials that will further help develop new molecules with improved potency and selectivity for BACE1. Besides, grasping the comprehensive molecular mechanisms of the selected inhibitors would also help in fundamental pharmacophore investigation when designing BACE1 inhibitors. Finally, the implementation of computational techniques in the designing of BACE1 inhibitors has been quite interesting. Nevertheless, the designing of potent BACE1 inhibitors through the computational application of the QM method such as the density functional theory (DFT), MM, and a hybrid QM/MM method should be extensively explored. We highly recommend that experimentalists should always collaborate with computational chemists to save time and other resources. ISIZULU ABSTRACT Iqoqa Isifo se-Alzheimer (AD), njengoba siqhubeka siyinhlanganisela yezimbangela ze- neurodegenerative engajwayelekile ebuchosheni, isikhathi esiningi kuxhumana nokulahleka noma ukufa kwama-neurons njengongqaphambili we-pathogenesis. Kungolunye uhlobo lwedementia oluhambisana nokulahlekelwa ukukhumbula kanyenokuxenga kanye nokuphanjanelwa ingqondo, ikakhulukazi kubantu abadala esebeneminyaka engaphezulu kuka-60. ENingizimu Afrikha, kunabantu abangaphezulu kwezigidi ezine abangephezulu kweminyaka ewu-60, ngokuhlawumbisela nje abayinkulungwane namashumi ayisishayangolombili nesikhombisa baphila nedemetia. Zimbili izimpawu ezihlukanisekayo ze-AD ziba-ama-neurofibrillary tangles kanye ne-B-amyloid plaques. I-B-amyloid plaques ingumphumela ngesikhathi i-amyloid eyiprotheni egijimayo iqhwakele oketshezini i-enzyme1 (BACEI), ngale kwalokho yaziwanjenge B-secretase. Kusukela ngo 1999 i-BAC1 yatholakala, isiphenduke ungqaphambili emizamweni yokwakha isidakamizwa sokwehlisa i-B-amyloid ngokwezinga lengqondo. Ngokunciphisa ukwanda kwe-B-amyloid isiphenduke okuqondiwe mayelana nokuqopha isidakamizwa ukuze kwelashwe i-AD. Ukuba nolwazi oluhle oluthinta isici sezakhi ze-BACE1 kuzosiza ekubazeni amandla akhethiwe i-BACE1 ukuvimbela imiphumela engaqondiwe. Kuze kube manje mihlanu imithi esiphasisiwe ngabezokuphatha ukudla kanye nezidakamizwa (FDA) ukwelapha isifo se-Alzheimer kanye nokuthi azikho kulezi eziphasisiwe izidakamizwa ebhekana ngqo ne-BACE1. Emva kokuba selitholakele lapho nje eminyakeni engu 20, sekunezinye esikhathini esedlule kanye nezifundo ezisaqhubeka zigxile ngokubheka kakhulu iqhaza lokwelapha i-BAC1 njengokuqondiswe ekungameleni u-AD. Imizamo eminingana yenziwa esikhathini esedlule ukuqopha uketshezi lwezidakamizwa olukwazi ukuvimba kahle i-BACE1. i-B-amyloid plaques kumuntu one-neurological ye-Alzheimer’s kumzamo (isigaba 3), kwabuye kwanqanyulwa ngenxa yokukhathazeka ngokwezempilo. Ukunqanyulwa kwanikela kuzizathu zokusilele kwezidakamizwa okuqondene nokulashwa kwe-AD. Kamuva, i-novel enamandla, ngisho ngawo umlomo kanye neyakhethwa ngezinga eliphezulu i-AM-6494 BACE1 evikelayo yatholakala. Le noveli i-BACE1 evimbayo yabukisa hhayi ukushintsha kombala woboya kanye nokushintsha kwesikhumba okujwayelekile, njengoba kubukwa nezivimbo zokuqala ze-BACE1. I-AM-6494 ne-IC50 enobumqoka buka 0.4nM kuyo i-vivo ekhethwa ngokwamanje kanye nesigaba sembulambethe yemizamo. Ngaphambi kwalesi sifundo, izakhi zesivimbela zale noveli i-BACE1zivimba ngokwe-atomistic kanye neqophelo le-molecular ye-B ACE1evimbayo kusale nje kungacacile. Umqulu wokuqala (isahluko sesibili) ukubuyekezwa kwesifo se-Alzheimer’s kanye no-B-secretase ovimbayo: ezikhumbuzayo ezigxile ngokusizwa yikhompuyutha eyisivimbo ngokwakhiwa. Sethula isendlalelo sesifundo kanye nengxoxo kafushane nezimbangela nemiphumela ye-Alzheimer. Ukubukezwa kwezimpawu zendlela zobukhompuyutha kufaka ekuqopheni isivimbo se-BACE1 nokuqhutshekiswa kwesidakamizwa. Ngokusebenzisa ofeleba begama BACE1, kusho ukwakha isivimbo, kanye nesifundo senjulalwazi kulwembu lobuchwepheshe kanye ne-Scopus sesizindalwazi. Sathola amaphepha acwaningiwe anokuhlobana angaphezulu kuka 49. Unyaka wokuthungatha usukela ku2010 kuya ku2020, nohlaziyo lwenziwa kusukela kuNhlaba 2020 kuya kuNdasa 2021. Umqulu wethu wesibili (isahluko sesithathu) sabuyekeza i-BACE ehlanganisa i-exosite antibody kanye ne-allosteric yokuthuthukisa ukwelashwa. Sakufunda ukusebenza kwesayensi yokuphila ye-BACE1, i-pathogenesis ehambisana nezifo kanye nezakhi zama-enzymatic esizinda sokuhlukana se-APP. Saphakamisa ukufakwa okunzulu nokucokeme kokulinganisa ngobuchwepheshe bekhompuyutha ekuphenyeni ama-anti-BACE1 omzimba kanye ne-allosteric ye-exosites. Kuyakholeka ukuthi uphenyo luzoqhubeka nokusiza ekwehliseni izinselelo ezihambisana nokwakha isithiyo se-BACE1 ngesikhathi kuhlolwa amathuba okwakheka kwe-allosteric yama-antibodies. Ubuyekezo luphinde lwaveza uketshezi olubukisa isizinda sokuhlanganisa kabili kusizinda esikhuthele kanye nesizinda se-allosteric. Umphumela, kube ukwenza isincomo mayelana nocwaningo olunzulu oluzohlanganisa umfutho okhululekile odlulele

ku-molecular docking (njengesicokeme se-molecular yokuhlukahlukana kokulinganisa) njengoba lokhu kuthembisa ukuveza isiza esibopha ngempela ama-compounds angaphansi Isahluko sesine siqukethe imininingwane ngamaqhinga e-computational sayensi efaka isicelo esibalulekile sezindlela ezibalulekile ze-molecular mechanics (MM), i-quantum mechanics (QM), i-hybrid ye-QM/MM, ngesisekelo samasethi kanye namanye amathuluzi ekhompuyutha akhethwa kulesi sifundo. Kumqulu wesithathu (isahluko sesihlanu), siqhube isilinganiso se-computational ye-AM-6494 kanye CNP-520.I-CNP-520 kwakungenye yezidakamizwa zokuqala zeBACE1 ezashatshalaliswa, zakhethwa kulesisifundo ngezizathu zokuqhathanisa. Ukulinganisa kwakuchaza kanye nokuqonda ukusondelana ngokuhlanganiswa kwezithiyo ezimbili kusigaba se-atomistic. Kwahlolwa i-quatum mechanics (QM) yesisindo yokusebenza kwenjulalwazi (DFT) kanye ne-hybrid QM/MM yokwethu okuno-N oluwugqinsi lwe-molecular Orbital kanye ne-Molecular Mechanics (ONIOM) kulolu linganiso. Lezi zindlelakwenza ze-computational zasiza ekuqageleni kwezakhiwo zama-electronic e-AM-6494 kanye CNP-520, kungena namandla okuhlanganisa ngesikhathi kuba lukhuni ne-BACE1. Ngokucabanga izinkulumo mpikiswano mayelana nokuma kwe-protonated ye-Asp32 kanye Asp288 kunika ukuvumelana namandla okuhlanganisa, nokuhlaziya izimo ezimbili ezifaka i-protonation kanye ne-unprotonation ye-Asp32 kanye Asp228. I-ONIOM ye-protonated yomfanekiso wokubala wanikeza amandla akhululekile okuhlanganisa -33,463kcal/mol (NP-520) kanye 62.849 kcal /mol kwavela i-unprotonate ye-AM6494. Imiphumela itshengisa ukuthi i-protonated iyisifanekiso njengoba kuyisona esivumela ukuhlanganiswa ngokukhululeka ngesikhathi lapho bekuqhathanisa ne-unprotonation yomfanekiso u-AM-649. Kuqhutshelwa phambili nemisebenzi ye-thermochemistry kuhlangana nokudlelana ne-molecular plots kutshengisa ukuthi i-AM-649 inezakhiwo ezinhle zokuvimba kune CNP-520. Yize kunjalo kwabonakala ukuthi i-protonation kanye ne-unprotonation ye-Asp32 kanye neyomfanekiso owu- Asp228 bekungatshengisa ngokwenele ukuhlanganisa ngokwe-interatomic yama-ligands EBACE1 ebilukhuni. be-BACE1 ngokwedlulele isilinganiso se-computational. Ukwenza ngokujulile kuphangiswa isilinganiso se-molecular ngokuhlukana, kwakakwa nohlaziyo olusemqoka lwezingxenyana. Imiphumela ye-AM-6494 yaqhubeka yatshengisa ukusondelana kokuhlanganiswayo no-van der Waals njengohamba phambili ekunikeleni amandla ahlobene ne-CNP-520. Ukuvuma kohlaziyo lwe-B-hairpin ngaphakathi ku-BACE1 kutshengiswa esizeni esiphilayo esivala ngendlela umnyakazo wokuvuma kobunkimbinkimbi be-AM-6494 ehlobene neCNP-520, ngokuvamile eshitshashintshayo phakathi kwevalekile kanye nezishaya sakuvuleka kokuvuma okunhlobonhlobo. Lokhu kuhlolwa kuqhubeke kwachazwa ngokuthi i-AM-6494 itshengisa ukuvimba okukhulu nokunethemba mayelana ne-BACE1. Isikhuthazizinguquko se-dyad (Asp32/228), Tyr14, Leu 30, Tyr 71, kanye ne-Gly230 kwakha izinsalela ezibalulekile nxazombili kuAM-6494ne-potencies yeCNP-520 kuBACE1 nesixhumanisi esihlanganisayo. Imiphumela ivela kulama-computational anzulu ayisilinganiso kanye nohlaziyo olucacisa ngokungangabazi i-AM-6494 enesivimbelo esiphakeme esingakwazi ukuqhubeka nokusiza intuthuko yama-molecules amasha anamandla athuthukile kanye nakhethelwe i-BACE1. Ngaphandle kwalokhu, ukucosha izinkambiso ezibanzi ze-moleculor mayelana nezivimbo ezikhethiwe kuzosiza mayelana nophenyo olubalulekile lwe- pharmacophore ngesikhathi kuqoshwa izivimbo se-BACE1. Ekugcineni, ukwenziwa kwe-computational ngokwamacebo ekubazeni izivimbo ze-BACE1 kube into ehlaba umxhwele. Nokho ukubaza izivimbo ezinamandla ze-BACE1 ngokusebenzisa i-computational yendlela ye-QM njengenjulalwazi yesisindo esisebenzayo (DFT), MM, kanye nendlela ye-hybrid QM/MM kufanele iphenywe kanzulu. Sincoma kakhulu ukuthi ongoti abenza izibonisi kufanele njalo bahlangane nama-computational chemists ukonga isikhathi kanye nezinye izinsiza.