Biaryl scaffold-focused virtual screening for anti-aggregatory and neuroprotective effects in Alzheimer’s disease

Alzheimer disease (AD) is the primary cause of dementia worldwide. Currently, more than 35 million people are suffering from this disease around the globe. By the year 2050, the diseases burden is excepted to raise four times i.e. almost 1 out of 85 persons will be suffering from AD [1]. The major pathological hallmarks of AD include widespread neuronal and synaptic loss, excessive presence of astrocytes, and aggregation of multiple proteinaceous deposits for instance β-amyloid plaques and neurofibrillary tangles (NFT) [2]. The number of hypotheses are proposed along the years to describe the root cause of AD such as the production of β-amyloid, cholinergic hypothesis, excitotoxicity and oxidative stress hypothesis [3] as summarized in Fig. 1. Senile plaques are the main and distinguished neurological feature of the AD directly related to its onset and progression [4]. The production of amyloid beta (Aβ) takes place by the proteolytic cleavage of beta-secretase protein on amyloid precursor protein (APP) whereas in AD, a pathogenic mutations affects the protease cleavage sites in APP and aid its cleavage [5]. The Aβ are mainly divided into two isoforms, based upon the length of amino acids, Aβ of 40 amino acid residues (Aβ40) and Aβ of 42 amino acid residues (Aβ42) are the two isoforms. Although Aβ42 differ in only 2 amino acids, it is much more neurotoxic and aggregates faster as compared to Aβ40. In the cerebrospinal fluid (CSF) presence of Aβ42 is a well-known biomarker of AD, and is used both in AD research and increasingly in clinical practice [6]. The increased level of Aβ42 as compare to Aβ40 has been generally considered to play a critical role. The increased Aβ42/Aβ40 ratio is closely related to presenilin mutations correlating to early onset of AD [7]. Although Aβ40 is several-fold more abundant in the brain Aβ42 is the major and sometimes exclusive component in amyloid plaques, due to its more aggregation prone nature [7].

AD is also characterized by the cholinergic deficit in the affected brain. The acetylcholine-releasing neurons especially there cell bodies which lied in basal forebrain degrades selectively in AD affecting cognitive functions and memory as these neurons are vital in the normal functioning of cerebral cortex and related structures. In AD there is a modification and alteration in polymorphism of acetylcholinesterase (AChE) in brain [8]. An increased amount of AChE levels around the Aβ plaques and NFT is commonly reported feature of AD. The current therapy of AD is mainly based on the use of AChE (AChE-I) inhibitors. The effect of these AChE-I is modest and transient due to up-regulation of AChE activity following chronic AChE-I therapy [9].

Another mechanism by which AD can develop is by excessive presence of the reactive oxygen species (ROS) in the mitochondria. This rise in ROS is due to aging or stress and if the antioxidant system of the body fails to cope with this condition AD may develop. The role of oxidative stress in AD is evident from the fact that brain of these patients shows a substantial oxidative damage [10]. Monoamine oxidase (MOA), is also involved in AD by increasing ROS in brain. MAO and its isoenzymes i.e. MAO-A and MAO-B are liable for the catalysis of biogenic amines, like serotonin, dopamine and norepinephrine so its inhibition results in an augmented level of neurotransmitters in the CNS [11].

Similarly, the key excitatory neurotransmitter glutamate which is involved in synaptic plasticity and learning, is also linked to AD. Dysfunction in the glutamate system is found to be linked with Aβ peptide linked oxidative stress. Oxidative stress leads to reduced activity of glutamate synthase resulting in significantly higher level of glutamate in brain [12]. The over-activation of N-methyl-D-aspartate receptor (NMDA) receptor due to excessive glutamate leads to the continuous calcium ions (Ca²+) influx into the nerve cells, generating a slow excitotoxicity at post synaptic level ultimately leading to a gradual neurodegenerative effect in AD patients [13]. Thus, NMDA receptor antagonists could be advantageous therapeutically in the management of AD.

Inhibition of these targets individually with currently approved or developing drugs has been proved relatively unsuccessful at reversing the progression of AD. A likely solution lies in a multi-pharmacological approach to altered activities of numerous of these targets at the same time, particularly those associated with the progression of the disease. Such multiple target drugs developed for AD have targeted two or more of known targets (cholinesterases, BACE1, MAO, NMDA) or have disease progression retarding properties, such as metal chelation, reduce oxidative stress or have anti-inflammatory potential, or can prevent Aβ or tau aggregation [14]. Ligands for drug targets combinations should be assessed against disease progression to define best possible combinations.

Molecular docking; a computational technique, is used for the estimation of the binding affinity between two molecules like the protein–protein and ligands-protein [15]. Virtual screening or computer-aided drug design (CADD) combined with wet lab techniques contributes towards the development of new drug molecules [16]. CADD is particularly useful especially in three major areas: (1) selection of most suitable compounds from large libraries of possibly actives compounds (2) Addition of an appropriate functional groups in the lead compounds for making it more suitable for new drug development (3) By adding pharmacophore features, designing the new molecules from a target structure. In light of current study the interactions of selected compounds with the various targets of AD was determined, using docking and in silico absorption, distribution metabolism and excretion (ADME) pharmacokinetics studies [15].

Molecular docking is an important technique. It is actually not a standalone technique but works best if treated as a supplementary technique along with other in silico methods as well as in vitro and in vivo experiments [28]. While there is a dire need for new pharmaceutical research in this field, in silico drug analysis is an effective and promising tool for discovering therapeutic utility of both new and already existing drugs [29]. There are number of examples of repurposed drugs which were discovered by the in silico approach and are now being used in many diseases [23] including AD [30].

The main focus of the present study is to signify importance of docking analysis and identify compounds containing biaryl scaffold for the management of AD. About, 107 compounds were screened using docking. Among screened compounds significantly interacting compounds with selected protein targets of AD were identified.

Some previous studies showed the possible potential of marketed antipsychotic drugs against various targets associated with AD by using docking approach [23]. Previously conventional structure-based docking method for the identification of drug molecules for BACE were largely futile [31]. By using in silico computer-aided drug design approaches, the activity of biaryl compounds against AD targets BACE1 and BACE2 was also determined and interestingly it was found that the fused-ring compounds are in general more active than the biaryl-based ligands [32].

In present study 4 targets of AD were chosen. The results of docking were firstly validated. As a general rule, the success of the docking scoring function is validated if the conformation of the bound ligand in the crystal structure resembles the conformation of the docked ligand [33].

One of the selected targets was AChe. An enzyme Ache catalyzes the metabolism of acetylcholine and some other choline esters neurotransmitters. Number of well-known drugs interact with acetylcholinesterase [34]. The human AChE’s active site is 20 Å deep. The active site comprises of catalytic site of AChE (Glu334,Ser 203 and His447), acyl-binding pocket (Phe297and Phe295), oxyanion hole (Ala204, Gly120 and Gly121), quaternary ammonium binding locus (Trp86) and finally, PAS (Tyr341, Trp286, Tyr124, Tyr72 and Asp74), which groups at the active site gorge’s entry [35].

Similarly, monoamine oxidase (MAO) an enzyme majorly involves in the oxidation of numerous vital monoamine hormones and neurotransmitters such as adrenaline, noradrenaline, dopamine, and serotonin [36]. There is a hydrophobic cavity in MAO-A which has a volume of ~ 400 Å. The structure of MAO-A consist of one larger cavity or a bipartite cavity depending upon the conformation of Phe208, but in this case it fails to work as gating residue. The MAO-A has conserved active site residues which comprise of a pair of Tyr of the “aromatic sandwich” and a Lys-hydrogen bonded to the N(5) position of the Flavin i.e. Lys305 in enzyme [37]. There are additional non-conserved active site residues primarily Ile180 and Asn181in MAO-A. The Phe208–Ile335 in MAO-A is the main factor in controlling the differential inhibitor and substrate specificities of these enzymes [38]. In MAO-A, the least docking score was − 10.5 with ZINC000000390492 and in the case of known inhibitor, Hermine, the docking score was − 7.5.

The inhibition of BACE1 through the development of selective and potent inhibitors has been in a limelight in the quest for treatment of AD. BACE an enzyme playing vital role in the proteolytic cleavage of APP is an important target of AD management [39]. The BACE-1 ligand ZINC000002010548 exhibited highest docking score of—8.7 and known inhibitor LY2886721 had docking score of − 9.1 in the previous study [40]

Another target, for protein 1PBQ an ion channel protein and glutamate receptor found in nerve cells, has vital interacting residues with 5,7-Dichlorokynurenic acid (DCKA); a selective NMDA antagonist. DCKA: The co-crystallized ligand interacts with Pro124, Ser180, Arg131 and Thr126. 1PBQ’s hydrophobic pocket has following amino acid residues: Phe92, Phe16, Phe250 and Trp223 [41].

Pharmacokinetic properties of drugs is a significant parameter in drug selection and to determine its utility as a clinically beneficial agent. Docking models have been simulated as a useful alternative to wet lab research procedures especially at initial stages were chemical structures are numerous but resources are scarce [42]. Hence, in present study the physio-chemical factors were determine to assess the ADME properties of the drugs. Lipinski’s rule of five required that the drug must have molecular weight of 500 Da or less, donorHB ≤ 5, accptHB ≤ 10 and octanol–water partition coefficient log P < 5. If drug follow these rules then it’s considered as an orally active drug. The molecules which fails to follow more than one of above stated rules may face difficulty with bioavailability. In current study none of the compounds is violating Lipinski rule of 5. Therefore, in silico computational analysis has been able to identify some encouraging compounds which may prove to be useful in the management of AD. Further research is needed in this regard.

AD is a complex disease involving many different pathways and drug targets, for instance, AChE, BACE-1, MAO-A and NMDA. In addition to these targets, anti-inflammatory and antioxidant drugs are also proved to be beneficial. Several molecules belonging to different chemical classes have already been developed against these individual targets to relieve the symptoms of this ailment but a multi-target approach is required. In this context the present study explored the potential of biaryl scaffold to inhibit these multiple targets. Of all the compounds screened, biaryl sulphonamides were found to be the top candidate for the cholinergic (AChE), beta-secretase cleavage enzyme (BACE-1), monoaminergic (MAO-A) and glutamatergic system (NMDA). Further analogues can also be computationally designed and tested against these druggable targets. Hence, virtual screening can successfully identify auspicious compounds which might be worthwhile therapeutically in AD.