Scaffold Morphing and In Silico Design of Potential BACE-1 (β-Secretase) Inhibitors: A Hope for a Newer Dawn in Anti-Alzheimer Therapeutics

Alzheimer’s disease (AD) is the prime cause of 65–80% of dementia cases and is caused by plaque and tangle deposition in the brain neurons leading to brain cell degeneration. β-secretase (BACE-1) is a key enzyme responsible for depositing extracellular plaques made of β-amyloid protein. Therefore, efforts are being applied to develop novel BACE-1 enzyme inhibitors to halt plaque build-up. In our study, we analyzed some Elenbecestat analogues (a BACE-1 inhibitor currently in clinical trials) using a structure-based drug design and scaffold morphing approach to achieve a superior therapeutic profile, followed by in silico studies, including molecular docking and pharmacokinetics methodologies. Among all the designed compounds, SB306 and SB12 showed good interactions with the catalytic dyad motifs (Asp228 and Asp32) of the BACE-1 enzyme with drug-likeliness properties and a high degree of thermodynamic stability confirmed by the molecular dynamic and stability of the simulated system indicating the inhibitory nature of the SB306 and SB12 on BACE 1.


Introduction
According to the World Health Organization, they estimate that in 2022, Alzheimer's disease and associated dementias will be the seventh major cause of death globally. It is currently the paramount reason for the suffering of about 50 million patients (https:// www.who.int/ (accessed on 6 June 2023)). The estimates point towards the probable increase in cases to around 13.8 million by 2060 [1,2]. The disease is disastrous in terms of mortality and morbidity and is characterized by deficits in the cognitive and motor abilities of the patients and dementia. It results due to neuron degeneration in the cortical and hippocampal regions of the brain, which control learning, memory, and cognitive abilities [3]. Alzheimer's disease (AD) is associated with the formation of senile plaques and neurofibrillary tangles in the brain, which causes dementia and memory loss [4][5][6]. The primary hallmark of Alzheimer's disease is the accumulation of β-amyloid proteins in the synapses of the cortical neurons and the formation of tau tangles inside the neurons due to the twisting of tau protein fibers. These accumulations lead to the neurons' death was drawn in the software search bar, and analogues were developed using the four inbuilt protocols in the software. The analogues possessing improved pharmacokinetic, physicochemical, and pharmacodynamic properties were developed and sorted in the increasing order of their synthetic accessibility. After sorting out, 2003 molecules were found to possess a score ≤ 4.5; after further eliminating analogues with similar structures and properties, 1880 proceeded for further in silico (absorption, distribution, metabolism, excretion) ADME evaluation.

Scaffold Morphing through Bioisosteric Replacement
The chemical structure of elenbecestat was analyzed, and keeping the nucleus of the molecule pyrazine-2-carboxamide constant, multiple sites were switched using their bioisosteres by using the MolOpt web server (Table 1) [18]. The molecule elenbecestat was drawn in the software search bar, and analogues were developed using the four inbuilt protocols in the software. The analogues possessing improved pharmacokinetic, physicochemical, and pharmacodynamic properties were developed and sorted in the increasing order of their synthetic accessibility. After sorting out, 2003 molecules were found to possess a score ≤ 4.5; after further eliminating analogues with similar structures and properties, 1880 proceeded for further in silico (absorption, distribution, metabolism, excretion) ADME evaluation. The chemical structure of elenbecestat was analyzed, and keeping the nucleus of the molecule pyrazine-2-carboxamide constant, multiple sites were switched using their bioisosteres by using the MolOpt web server (Table 1) [18]. The molecule elenbecestat was drawn in the software search bar, and analogues were developed using the four inbuilt protocols in the software. The analogues possessing improved pharmacokinetic, physicochemical, and pharmacodynamic properties were developed and sorted in the increasing order of their synthetic accessibility. After sorting out, 2003 molecules were found to possess a score ≤ 4.5; after further eliminating analogues with similar structures and properties, 1880 proceeded for further in silico (absorption, distribution, metabolism, excretion) ADME evaluation.

Scaffold Morphing through Bioisosteric Replacement
The chemical structure of elenbecestat was analyzed, and keeping the nucleus of the molecule pyrazine-2-carboxamide constant, multiple sites were switched using their bioisosteres by using the MolOpt web server (Table 1) [18]. The molecule elenbecestat was drawn in the software search bar, and analogues were developed using the four inbuilt protocols in the software. The analogues possessing improved pharmacokinetic, physicochemical, and pharmacodynamic properties were developed and sorted in the increasing order of their synthetic accessibility. After sorting out, 2003 molecules were found to possess a score ≤ 4.5; after further eliminating analogues with similar structures and properties, 1880 proceeded for further in silico (absorption, distribution, metabolism, excretion) ADME evaluation.

Scaffold Morphing through Bioisosteric Replacement
The chemical structure of elenbecestat was analyzed, and keeping the nucleus of the molecule pyrazine-2-carboxamide constant, multiple sites were switched using their bioisosteres by using the MolOpt web server (Table 1) [18]. The molecule elenbecestat was drawn in the software search bar, and analogues were developed using the four inbuilt protocols in the software. The analogues possessing improved pharmacokinetic, physicochemical, and pharmacodynamic properties were developed and sorted in the increasing order of their synthetic accessibility. After sorting out, 2003 molecules were found to possess a score ≤ 4.5; after further eliminating analogues with similar structures and properties, 1880 proceeded for further in silico (absorption, distribution, metabolism, excretion) ADME evaluation.

Scaffold Morphing through Bioisosteric Replacement
The chemical structure of elenbecestat was analyzed, and keeping the nucleus of the molecule pyrazine-2-carboxamide constant, multiple sites were switched using their bioisosteres by using the MolOpt web server (Table 1) [18]. The molecule elenbecestat was drawn in the software search bar, and analogues were developed using the four inbuilt protocols in the software. The analogues possessing improved pharmacokinetic, physicochemical, and pharmacodynamic properties were developed and sorted in the increasing order of their synthetic accessibility. After sorting out, 2003 molecules were found to possess a score ≤ 4.5; after further eliminating analogues with similar structures and properties, 1880 proceeded for further in silico (absorption, distribution, metabolism, excretion) ADME evaluation.

Scaffold Morphing through Bioisosteric Replacement
The chemical structure of elenbecestat was analyzed, and keeping the nucleus of the molecule pyrazine-2-carboxamide constant, multiple sites were switched using their bioisosteres by using the MolOpt web server (Table 1) [18]. The molecule elenbecestat was drawn in the software search bar, and analogues were developed using the four inbuilt protocols in the software. The analogues possessing improved pharmacokinetic, physicochemical, and pharmacodynamic properties were developed and sorted in the increasing order of their synthetic accessibility. After sorting out, 2003 molecules were found to possess a score ≤ 4.5; after further eliminating analogues with similar structures and properties, 1880 proceeded for further in silico (absorption, distribution, metabolism, excretion) ADME evaluation.

Scaffold Morphing through Bioisosteric Replacement
The chemical structure of elenbecestat was analyzed, and keeping the nucleus of the molecule pyrazine-2-carboxamide constant, multiple sites were switched using their bioisosteres by using the MolOpt web server (Table 1) [18]. The molecule elenbecestat was drawn in the software search bar, and analogues were developed using the four inbuilt protocols in the software. The analogues possessing improved pharmacokinetic, physicochemical, and pharmacodynamic properties were developed and sorted in the increasing order of their synthetic accessibility. After sorting out, 2003 molecules were found to possess a score ≤ 4.5; after further eliminating analogues with similar structures and properties, 1880 proceeded for further in silico (absorption, distribution, metabolism, excretion) ADME evaluation.

In silico Pharmacokinetic Studies
The ADME properties of 1880 molecules were evaluated in this step to assess their drug-likeliness and pharmacokinetic properties. The drug-likeliness properties of the analogues were assessed by their conformation to Lipinski's rule of five (molecular weight < 500; QPlogPo/w < 5, H-B donors ≤ 5, and H-bond acceptors ≤ 10). Since the drug targets Alzheimer's disease and would work as a BACE-1 inhibitor, the analogues were filtered for BBB permeability. Of all the analogues evaluated, only 50 molecules were selected as they possessed both drug-likeliness properties and BBB permeability ( Table 2). These 50 compounds exhibited an acceptable range of physicochemical and pharmacokinetic parameters and high GI absorption. Hence, the predicted ADME properties illustrated all the sorted analogues as good drug candidates, which proceeded further for molecular docking analysis.

In silico Pharmacokinetic Studies
The ADME properties of 1880 molecules were evaluated in this step to assess their drug-likeliness and pharmacokinetic properties. The drug-likeliness properties of the analogues were assessed by their conformation to Lipinski's rule of five (molecular weight < 500; QPlogPo/w < 5, H-B donors ≤ 5, and H-bond acceptors ≤ 10). Since the drug targets Alzheimer's disease and would work as a BACE-1 inhibitor, the analogues were filtered for BBB permeability. Of all the analogues evaluated, only 50 molecules were selected as they possessed both drug-likeliness properties and BBB permeability ( Table 2). These 50 compounds exhibited an acceptable range of physicochemical and pharmacokinetic parameters and high GI absorption. Hence, the predicted ADME properties illustrated all the sorted analogues as good drug candidates, which proceeded further for molecular docking analysis.

In Silico Pharmacokinetic Studies
The ADME properties of 1880 molecules were evaluated in this step to assess their drug-likeliness and pharmacokinetic properties. The drug-likeliness properties of the analogues were assessed by their conformation to Lipinski's rule of five (molecular weight < 500; QPlogP o/w < 5, H-B donors ≤ 5, and H-bond acceptors ≤ 10). Since the drug targets Alzheimer's disease and would work as a BACE-1 inhibitor, the analogues were filtered for BBB permeability. Of all the analogues evaluated, only 50 molecules were selected as they possessed both drug-likeliness properties and BBB permeability ( Table 2). These 50 compounds exhibited an acceptable range of physicochemical and pharmacokinetic parameters and high GI absorption. Hence, the predicted ADME properties illustrated all the sorted analogues as good drug candidates, which proceeded further for molecular docking analysis.

Molecular Docking
Analysis of molecular docking interactions of the selected analogues is given in Table 3. Out of the 51 molecules docked (50 analogues and elenbecestat), most of the formulated analogues showed interactions with both of the catalytic dyad residues (Asp32 and Asp228) of the BACE-1 protein via van der Waal interactions, pi-anion, halogen fluorine interactions, or carbon-hydrogen bonding. These molecules also showed significant interaction with other important amino acid residues of the BACE-1 protein like Trp115, Gly34, Gly230, Thr231, Thr232, Val332, Lys224, Arg235, and Tyr71. Table 3. Docking score and key interactions of designed elenbecestat analogs with BACE-1.

S. No.
Compound ID Docking Score (kcal/mol) Interactions  Only one molecule out of all these SB282 showed no interactions with any of the catalytic dyad residues. Of all the docked molecules, the top two molecules with the maximum docking score or minimum binding energy were SB306 and SB12 (Figure 1). Only one molecule out of all these SB282 showed no interactions with any of the catalytic dyad residues. Of all the docked molecules, the top two molecules with th maximum docking score or minimum binding energy were SB306 and SB12 (Figure 1).  The SB306 has the most efficient binding with the BACE-1 protein with CDOCKER interaction energy (−44.3515 kcal/mol). The SB306 analogue had a Hexahydro-1H-pyrrolizine ring substituent in place of a 5-Methyl-4a,5,7,7a-tetrahydro-4H-furo [3,4-d] [1,3] thiazine-2amine ring at the fluorophenyl ring in elenbecestat, the rest of the structure was similar. The SB306 analogue showed carbon-hydrogen bond interaction with the ASP32 residue through the hydrogen of the pyrrolizine ring and van der Waals interaction with the ASP228. The fluorine atoms in the structure show maximum interactions with the amino acid residues of the binding site. The two fluorine atoms of the difluoromethyl substitution at the pyrazine ring showed halogen fluorine bonding with the THR231, SER229, and GLY230 and carbon-hydrogen bonding with the GLY13. The methyl group showed alkyl interaction with VAL166 and ALA335. The pyrazine ring showed amide-pi stacked interaction with the GLN12 and pi-alkyl interaction with LEU30. The hydrogen of the pyrazine ring shows carbon-hydrogen bonding with SER229. Another selected analogue SB12 had the second-best docking score, with CDOCKER interaction energy being −43.3585 kcal/mol. The analogue had a 2-oxo-2-pyrrolidin-1-yl ethyl ring substituent in place of the 5-Methyl-4a,5,7,7a-tetrahydro-4H-furo [3,4-d] [1,3]thiazine-2-amine ring at the fluorophenyl ring in elenbecestat. It showed van der Waals interaction with ASP32, conventional hydrogen bonding, and carbon-hydrogen bond interaction with ASP228. The ASP228 had conventional hydrogen bonding interaction with the amide substituent of the carboxamide moiety and carbon-hydrogen bond interaction with the hydrogen of the pyrazine ring. The hydrogens of the pyrrolidine ring formed carbon-hydrogen bonding with GLY230 residue, and the two fluorine atoms of the difluoromethyl substitution showed a significant number of interactions like conventional hydrogen bonding and carbon-hydrogen bonding with the LYS224 and THR329. The methyl group showed alkyl interactions with ILE226 and TYR198. The pyrrolidine ring showed pi-alkyl interactions with LEU30. The rest of the neighboring amino acids like ILE118, SER35, ARG235, VAL332, THR231, THR232, VAL166, GLY11, GLY13, ILE110, and TRP115 showed van der Waals interactions with the molecule. The elenbecestat had the CDOCKER interaction energy of −42.8232 kcal/mol, which proves the newer analogues to be promising candidates for further research in BACE-1 inhibition. The 2D and 3D views of interactions of SB306 and SB12 are given in Figures 2 and 3, respectively.
Molecules 2023, 28, x FOR PEER REVIEW 11 of 20 dro-1H-pyrrolizine ring substituent in place of a 5-Methyl-4a,5,7,7a-tetrahydro-4H-furo [3,4-d] [1,3] thiazine-2-amine ring at the fluorophenyl ring in elenbecestat, the rest of the structure was similar. The SB306 analogue showed carbon-hydrogen bond interaction with the ASP32 residue through the hydrogen of the pyrrolizine ring and van der Waals interaction with the ASP228. The fluorine atoms in the structure show maximum interactions with the amino acid residues of the binding site. The two fluorine atoms of the difluoromethyl substitution at the pyrazine ring showed halogen fluorine bonding with the THR231, SER229, and GLY230 and carbon-hydrogen bonding with the GLY13. The methyl group showed alkyl interaction with VAL166 and ALA335. The pyrazine ring showed amide-pi stacked interaction with the GLN12 and pi-alkyl interaction with LEU30. The hydrogen of the pyrazine ring shows carbon-hydrogen bonding with SER229. Another selected analogue SB12 had the second-best docking score, with CDOCKER interaction energy being −43.3585 kcal/mol. The analogue had a 2-oxo-2-pyrrolidin-1-yl ethyl ring substituent in place of the 5-Methyl-4a,5,7,7a-tetrahydro-4H-furo [3,4-d] [1,3] thiazine-2-amine ring at the fluorophenyl ring in elenbecestat. It showed van der Waals interaction with ASP32, conventional hydrogen bonding, and carbon-hydrogen bond interaction with ASP228. The ASP228 had conventional hydrogen bonding interaction with the amide substituent of the carboxamide moiety and carbon-hydrogen bond interaction with the hydrogen of the pyrazine ring. The hydrogens of the pyrrolidine ring formed carbon-hydrogen bonding with GLY230 residue, and the two fluorine atoms of the difluoromethyl substitution showed a significant number of interactions like conventional hydrogen bonding and carbon-hydrogen bonding with the LYS224 and THR329. The methyl group showed alkyl interactions with ILE226 and TYR198. The pyrrolidine ring showed pi-alkyl interactions with LEU30. The rest of the neighboring amino acids like ILE118, SER35, ARG235, VAL332, THR231, THR232, VAL166, GLY11, GLY13, ILE110, and TRP115 showed van der Waals interactions with the molecule. The elenbecestat had the CDOCKER interaction energy of −42.8232 kcal/mol, which proves the newer analogues to be promising candidates for further research in BACE-1 inhibition. The 2D and 3D views of interactions of SB306 and SB12 are given in Figures 2 and 3, respectively.

Molecular Dynamics Simulations
An MD simulation study of the best two compounds was performed to analyze their thermodynamic stability and dynamic behavior of the ligand-protein complex and to study the effect on the conformational alterations induced by ligand binding with the active pocket of BACE-1 [19]. After MD, although the important interactions were conserved, some additional interactions were also observed. The protein-ligand complex of BACE-1 and SB306 were analyzed, and the compound SB306 showed hydrogen bonding interactions with the ASP32, GLY11, π-π interactions with TYR14. Strong hydrogen bonding interactions were also observed with the VAL31 (Figure 4A). These interactions are summarized in the protein-ligand contacts plot ( Figure 4B). After MD simulations, the root mean square deviation (RMSD) was calculated for the SB306 compound concerning protein and plotted against time (ns) ( Figure 4C).

Molecular Dynamics Simulations
An MD simulation study of the best two compounds was performed to analyze their thermodynamic stability and dynamic behavior of the ligand-protein complex and to study the effect on the conformational alterations induced by ligand binding with the active pocket of BACE-1 [19]. After MD, although the important interactions were conserved, some additional interactions were also observed. The protein-ligand complex of BACE-1 and SB306 were analyzed, and the compound SB306 showed hydrogen bonding interactions with the ASP32, GLY11, π-π interactions with TYR14. Strong hydrogen bonding interactions were also observed with the VAL31 (Figure 4A). These interactions are summarized in the protein-ligand contacts plot ( Figure 4B). After MD simulations, the root mean square deviation (RMSD) was calculated for the SB306 compound concerning protein and plotted against time (ns) ( Figure 4C).  The protein-ligand complex of BACE-1 and SB12 were also analyzed, and the compound showed strong hydrogen bonding interactions with ASP32 and VAL31. The compound showed strong π-π interactions with TYR14 and hydrogen bonding interactions with GLY11 ( Figure 5A). These interactions are summarized in the protein-ligand contacts plot ( Figure 5B). After the MD studies, the RMSD graph of the SB12 ligand concerning BACE-1 was plotted against time (ns) ( Figure 5C). The protein-ligand complex of BACE-1 and SB12 were also analyzed, and the compound showed strong hydrogen bonding interactions with ASP32 and VAL31. The compound showed strong π-π interactions with TYR14 and hydrogen bonding interactions with GLY11 ( Figure 5A). These interactions are summarized in the protein-ligand contacts plot ( Figure 5B). After the MD studies, the RMSD graph of the SB12 ligand concerning BACE-1 was plotted against time (ns) ( Figure 5C).

Clinical Trials Screening
The clinical trials database (clinicaltrials.gov.in) was screened for small molecule BACE inhibitors. A total of 7 molecules were found, listed in Table 4. Most of the molecules currently undergoing clinical trials reduced concentration of Aβ in CSF, but their trials were terminated due to liver toxicity and severe adverse effects. Elenbecestat (E2609), which is currently undergoing phase III clinical trials, was found to be effective in reducing the Aβ levels in CSF and plasma, and also resulted in a decreased rate of decline in cognitive abilities, and also no hepatic toxicity was found to occur, but some side effects are associated with the molecule [12,20].  The plots of RMSD of the compounds SB306 and SB12 showed that docked complexes were quite stable throughout, with minor fluctuations in the range of 1.2-2.0 Å. The stability of the simulated system indicates the inhibitory nature of the SB306 and SB12 on BACE 1. Compound SB306 has shown almost no or a very small vibration throughout the simulation and remains stable within the macromolecular cavity, while the ligand SB12 has shown a couple of moves during the initial 20 ns time within the binding cavity to achieve the most stable conformation. Both the compounds SB306 and SB12 have shown a high degree of stability within the macromolecular cavity throughout the simulation period, which was expected to initiate the therapeutic response.

Clinical Trials Screening
The clinical trials database (clinicaltrials.gov.in) was screened for small molecule BACE inhibitors. A total of 7 molecules were found, listed in Table 4. Most of the molecules currently undergoing clinical trials reduced concentration of Aβ in CSF, but their trials were terminated due to liver toxicity and severe adverse effects. Elenbecestat (E2609), which is currently undergoing phase III clinical trials, was found to be effective in reducing the Aβ levels in CSF and plasma, and also resulted in a decreased rate of decline in cognitive abilities, and also no hepatic toxicity was found to occur, but some side effects are associated with the molecule [12,20].

Scaffold Morphing
This technique for designing drugs refers to modifying structural features to ameliorate the synthetic feasibility, potency, and drug-likeliness properties. The method provides the approach to enhancing the overall therapeutic profile of the chosen molecule [21,22]. The scaffold morphing works on the principle of bio-isosteric replacement, replacing each functional group or segment of the molecule with their bioisosteres and boosting the effectiveness in addition to pharmacokinetic profiling of the compound [21,22]. In this research study, the scaffold morphing of elenbecestat was performed using the freely available web server MolOpt [18]. The three molecule sites were identified, and various analogues were generated by replacing multiple molecule segments using 4 inbuilt transformation rules: AI generative model, data mining, data mining (fast), and similarity comparison. The generated elenbecestat analogues were sorted based on synthetic accessibility. The range of synthetic accessibility lies from very easy (1) to very difficult (10) [23]. A cut-off of 4.5 was employed to screen the formed analogues, and the 1880 analogues proceeded for in silico pharmacokinetic studies ( Figure 6).

In silico Pharmacokinetic Predictions
The pharmacokinetic properties of the selected 1880 molecules were evaluated using the online free web server swissADME tool (http://www.swissadme.ch (accessed on 25 May 2023)). Various molecular attributes were examined, including physicochemical properties, water solubility, pharmacokinetics, lipophilicity, and drug likeliness properties comprising (Ro5) Lipinski's rule [23,24]. Since the drug being evaluated is for Alzheimer's disease, the BBB permeability is also considered along with other parameters for

In Silico Pharmacokinetic Predictions
The pharmacokinetic properties of the selected 1880 molecules were evaluated using the online free web server swissADME tool (http://www.swissadme.ch (accessed on 25 May 2023)). Various molecular attributes were examined, including physicochemical properties, water solubility, pharmacokinetics, lipophilicity, and drug likeliness properties comprising (Ro5) Lipinski's rule [23,24]. Since the drug being evaluated is for Alzheimer's disease, the BBB permeability is also considered along with other parameters for evaluating the BACE inhibitor analogues [25][26][27]. Only the analogues exhibiting BBB permeability were proceeded for further molecular docking studies. Out of 1880 molecules examined, only 50 molecules were found to possess drug-likeliness properties and BBB permeability.

Molecular Docking Studies
To explore the binding affinity of the formed and sorted elenbecestat analogues with the BACE-1 enzyme, docking was carried out using Biovia Discovery Studio software [24,28]. Molecular docking of the formulated analogues and elenbecestat molecule was performed using the crystal structure of BACE-1 protein possessing 2Å, and an Rvalue of 0.223 was obtained using (Research Collaboratory for Structural Bioinformatics) RCSB-protein data bank (PDB) (https://www.rcsb.org/ (accessed on 25 May 2023)), the PDB ID: 6OD6 [29][30][31]. The protein was prepared using the 'macromolecule module' of the Biovia discovery studio software. The ligands were prepared using the Discovery Studio software's 'small molecule module.' The docking was proceeded by the Dock ligands 'CDOCKER' protocol, and the docked poses' molecular interactions was noted down [19]. In docking experiments, the grid box was generated with the X, Y, and Z axes coordinates of −36.937, −45.388, and 19.516, respectively, for 6OD6. A maximum of 10 binding modes were allowed to be generated for each ligand during the docking execution program [32][33][34]. Before docking the ligands, the docking protocol was validated by redocking the co-crystallized ligand (ligand ID: M7D) attached to the BACE-1 protein. Similar validated docking protocols for the analysis of the prepared ligand were followed. The interactions of the docked ligands with the catalytic aspartate (Asp) dyad (Asp32 and Asp228) residues present in the BACE-1 protein were observed. The interactions, docking score, and binding energy for each docked ligand were noted.

Molecular Dynamic Simulations
The designed analogues having good binding energies against the BACE-1 enzyme in molecular docking were further assessed for their thermodynamic stability concerning time by executing molecular dynamics (MD) simulation for a duration of 100 ns by using the molecular dynamics module of Desmond software by Schrodinger [35][36][37][38]. The simulations aided in stabilizing the protein-ligand complex by minimizing the system's energy by attaining the most preferred conformation after complexation concerning time. The interaction persisting between the complex ligand and the target macromolecule during the 100 ns duration of the simulation was analyzed by studying its simulation-interaction diagram [37,[39][40][41][42][43][44]. This analysis leads to assessing important interactions for ligand binding that can be utilized to assess ligand's affinity further. The system was first built using the TIP3P solvent model with an orthorhombic box shape; the input system's ionic state was adjusted using the salt solution of 0.15 M. The simulation was carried out using the NPT ensemble and a time step of 1.0 fs; the constant temperature of 310 K was adjusted by using the Nose-Hoover Chain method as the thermostat and pressure of 1.01325 bar using Martyn-Tobias-Klein as the barostat.

Conclusions and Future Remarks
A scaffold morphing and a structure-based drug designing approach were successfully utilized to identify putative analogues of elenbecestat against BACE 1. Initially, the bioisosteric replacement was carried out on various suggested sites of elenbecestat to generate a library of its analogues. From a library of more than 10,000 analogues, 2003 were selected based on synthetic feasibility and submitted for in silico ADME prediction to confer the drug-like properties of these analogues. Using a molecular docking approach, the 50 analogues were further evaluated for binding affinity with the target protein. Though all the compounds showed good interactions with the target protein (BACE 1), compounds SB306 and SB12 were the most active molecules, suggesting a plausible binding mode with the active or catalytic site of the enzyme, which is responsible for the BACE 1 activity. To analyze the stability and dynamic behavior of the ligands-protein complexes and to study the effect on the conformational alterations induced by binding of the ligand with the active pocket of BACE 1, the molecular dynamic simulation study of the top two compounds (SB306 and SB12) was performed. The plots of RMSD of both the compounds showed stability of the complexes all through with minor fluctuations in the range of 1.2-2.0 Å. The stability of the simulated system indicated the inhibitory nature of the SB306 and SB12 on BACE 1. Based on these results, it can be suggested that a slight structural modification in the elenbecestat may improve its therapeutic profile. The protocol adopted in this study may be used as a framework in the future for developing novel small molecules for the treatment of Alzheimer's disease and suggested that the designed new compounds could be further investigated for pharmacological development in AD therapy.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented in this study are available within the article.