Molecular docking and dynamic simulation studies of α4β2 and α7 nicotinic acetylcholine receptors with tobacco smoke constituents nicotine, NNK and NNN

Abstract Smoking constitutes a major global health problem. As it triggers various health hazards including cancers, cardiac and pulmonary illness, it is imperative to understand the mechanism of action of various smoke constituents on our cellular processes. Various in vitro studies have compiled the affinity of cigarette smoke constituents on various nicotinic acetylcholine receptors (nAChRs). But the nature of the intermolecular interactions contributing to this affinity and the key amino acids in the receptor active sites involved in this are not investigated so far. Here, we are examining the interaction of α7nAChR and α4β2nAChR on nicotine, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) and N-nitrosornicotine (NNN), the physiologically significant constituents in smoke, through molecular docking and dynamics simulations study. The docking of α4β2nAChR structure with the ligands nicotine, NNK and NNN yielded docking scores of −41.45 kcal/mol, −59.28 kcal/mol and −54.60 kcal/mol, respectively, and that of α7nAChR receptor molecule with the ligands yielded docking scores of −59.54 kcal/mol, −71.06 kcal/mol and −70.86 kcal/mol, respectively. The study showed that NNK exhibited the highest affinity with the ligands which was confirmed by dynamics simulation. But higher stability of interactions as surmised from Molecular dynamics simulations was found for nicotine with α4β2nAChR and NNN with α7nAChR. The findings validate the in vitro studies comparing the affinities of these compounds. The study will be useful in formulating effective nAChR agonists to treat neurological disorders and antagonists for smoke deaddiction and improve health standards. Communicated by Ramaswamy H. Sarma


Introduction
Cigarette smoking is an ill habit that burdens public health socially and economically. It is a modifiable risk factor for cardiovascular diseases, pulmonary illnesses and certain cancers. Cigarette smoking is the most widespread mode of tobacco consumption. According to the latest WHO data, there are 1.3 billion tobacco users which include smoke and smokeless users. Annually, 7 million deaths occur due to direct tobacco usage in any form and 1.2 million deaths due to secondhand smoke exposure. In the United States, smoking-related cancers are the third highest contributor to mortality trailing cardiovascular diseases and pulmonary illnesses (Benowitz, 2010;USDH, 2014). According to GLOBOCAN 2012 report, in India, 1 million new cancer cases and 7, 00,000 cancerrelated deaths were reported (Mallath, 2014).
Cigarette smoke consists of more than 7000 chemicals of which more than 60 are carcinogens as reported by the International Agency for Research on Cancer (Hecht, 2006). Nicotine is the principal noncarcinogenic component of tobacco that acts as a stimulant and renders an addictive property to tobacco (USDH, 2014;Benowitz, 2008). Apart from nicotine, cigarette smoke contains polycyclic aromatic hydrocarbons like benzo[a]pyrene, nitrosamine derivatives of nicotine like NNK(4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone) and NNN (N-nitrosornicotine) among others. The carcinogenic effects of NNK and NNN are mediated through DNA adduct formation and subsequent mutation (Sturla et al., 2005 andHecht, 1998). Various Cytochrome P 450 oxidases in the body transform NNK and NNN into DNA-reactive metabolites that form DNA adducts resulting in mutations or chromosomal aberrations (Xue et al., 2014;Hecht, 1999).
Nicotinic acetylcholine receptors (nAChRs) are universally expressed in the plasma membrane of mammalian cells. Historically, they were thought to be present in the nervous system, known as neuronal nAChRs, and muscles, known as muscle nAChRs. nAChRs are gated ion channels that allow the passage of cations upon conformational change brought about by the binding of its physiological agonist, acetylcholine (Lindstrom et al., 1996). Neuronal nAChRs comprises either 5 identical subunits of a7, a8, or a9 subunits or a combination of a2, a6, and a10 subunits with either of b2 or b4 subunits (Jensen et al., 2005;Dani & Bertrand, 2007;Liu, 2014). Nicotine and nitrosamines like NNK and NNN mimic acetylcholine and results in the opening of ion channels triggering an influx of various cations (Na þ , K þ and Ca 2þ ) resulting in the release of neurotransmitters, angiogenic and neurotropic factors. Apart from this, membrane depolarization caused by the influx of cations will trigger the opening of voltage-gated Ca 2þ channels. As Ca 2þ is an intracellular second messenger, the rise in [Ca 2þ ] activates other intracellular signaling pathways like the mitogen-activated protein kinase (MAPK), protein kinase C (PKC) and vascular endothelial growth factor (VEGF) that regulates cell proliferation, metastasis, angiogenesis and apoptosis (Schuller, 2008;Kunzelmann, 2005;Roderick & Cook, 2008).
a7nAChR and a4b2nAChR are evolutionarily the oldest and well-investigated members of the nAChR receptor family (Le Novere & Changeux, 1995). Among all the nAChRs, a7nAChR is reported to be more selective to Ca 2þ ions than other members of the group (Gopalakrishnan et al., 1995;Lindstrom et al., 1996). It was also observed that nicotine binds with higher affinity to a4b2nAChR than a7nAChR. This leads to protracted desensitization of the receptor. In cancers, a7nAChR regulates the pathways leading to tumor progression, and a4b2nAChR controls inhibitory pathways (Kawai & Berg, 2001). Thus, extended abrogation of a4b2nAChR activity and a stimulatory a7nAChR activity in the presence of nicotine renders a tumor progressive cellular environment (Benowitz, 2008). Empirically, it was also observed that NNK exhibited 1300-fold affinity for a7nAChR receptor and NNN displayed 5000-fold affinity for heteromeric a-b nAChRs than nicotine. This alludes to the pathological significance of tobacco nitrosamines in nAChR stimulation in cancers that were previously unappreciated (Arredondo et al., 2006;Schuller & Orloff, 1998). The therapeutic effects of various nAChR agonists were studied in animal models for Alzheimer's disease and other neurological disorders like Schizophrenia. Other neuroprotective effects from agonists were found useful against attention-deficit hyperactivity disorders, depression, and cognitive dysfunction (Teaktong et al., 2004;Yu et al., 2005 andTaly et al., 2009). Various nAChR antagonists, during in vivo and in vitro studies, were reported to be potential anticancer drugs having anti-angiogenic and antiproliferative activity (Holladay et al., 1997;Sandall et al., 2003).
In order to develop potential nAChR agonists or antagonists for therapeutic purposes, it is important to screen compounds for comparative affinity and toxicity studies before proceeding with in vitro and in vivo studies. Molecular docking is an established in silico method for predicting the stability of the interaction between a molecule of interest with a receptor/protein under investigation. The docked poses of the ligand with the receptor that offers the lowest binding energy (docking score) are used for comparing the binding affinities of various ligands with receptor/protein under investigation screening for potential drug targets. They also help us understand the intermolecular interactions that contribute to the binding affinities and investigate the conformational state of protein/receptor that permits the same. Molecular dynamics simulation predicts key protein-ligand interactions in a dynamic environment where interaction is modeled with flexible protein and ligand structures. Intermolecular interactions studies were carried out using RMSD, RMSF and 3D interaction diagram.
The binding affinity of various compounds including nicotine with a7nAChR-LBD structure (Ligand binding domain) was investigated using a competitive docking model (Ng et al., 2018). A homology model of a4b2nAChR was constructed, and molecular docking and dynamic simulations were carried out to screen for potential tobacco smoke constituents binding the receptor (Mao et al., 2016). The effect of cigarette smoke constituents on the central nervous system (CNS) was investigated by molecular docking and molecular dynamics simulations of NNK and NNN with enzymes acetylcholinesterase and butyrylcholinesterase that showed smoke-induced altered molecular mechanisms (Jamal & Alharbi, 2022).
A molecular docking model that predicts the intermolecular interactions that contribute to the greater affinity for the nitrosamines for a7and a4b2nAChR's were not yet investigated. To address this gap in knowledge, we decided to perform molecular docking of a7and a4b2nAChR's structures with nitrosamines NNK and NNN, along with nicotine. Highresolution X-ray crystallography structure data of human a4b2nAChR from a recent study were publicly available (Morales-Perez et al., 2016). The complete structure of human a7nAChR is not elucidated till date. The above-mentioned crystal structure of a4b2nAChR along with a surrogate crystal structure for human a7nAChR receptor derived from Lymnaea stagnalis was used for our studies. To optimize these interaction studies, we also conducted molecular dynamics simulations studies to investigate the stability of interactions. This study may also contribute to the screening of better nAChR agonists and antagonists through understanding the potential molecular interactions contributing to the higher affinity of NNK and NNN for the receptors in the investigation.

Protein and ligand preparation
The 3D crystal structure of human a4b2nAChR bound with nicotine (PDB ID: 5KXI) was retrieved from the Protein Data Bank (http://www.rcsb.org/) (Morales-Perez et al., 2016). A surrogate crystal structure for human a7nAChR receptor, an acetylcholine binding protein (AChBP) structure derived from Lymnaea stagnalis (PDB ID: 5J5I), was also fetched for investigation (Kaczanowska et al., 2017). Protein preparation was carried out using the protein preparation wizard of Discovery Studio. Water molecules were removed from the structure, and hydrogen atoms, charged atoms and missing residues were added. Smart minimizer-based minimizations were carried out to optimize atoms. The 2D structures of the ligands were downloaded from Pubchem database, and 3D structure preparation and refinement were done using Ligprep wizard of Discovery Studio (Sastry et al., 2013).

Molecular docking
Molecular docking was performed using iGEMDOCK v2.1 for predicting the binding energy and probable conformations. Key amino acid residues in the ligand-binding sites were obtained from PDB. Empirical scoring function and generic evolutionary methods were used to predict most favorable conformers. The binding affinity of docked complexes was evaluated based on binding energy (kCal/mol). Interactions were enumerated as hydrogen bonding and van der Waal's interactions (Shaikh et al., 2022).

Molecular Dynamics simulation
Molecular dynamics simulations were carried out using Desmond package of Schr€ odinger (Schr€ odinger, 2019 and Shivakumar et al., 2010). Protein-ligand complex simulations were run using OPLS2005 force field parameters. Simple point charge water model was used to build a system, emulating water molecules. Orthorhombic boundary conditions at a distance of 10 A o beyond atoms in the complex (Jorgensen et al., 1983;Jorgensen et al., 1996). Na þ and Cl À ions were added to the system to neutralize the charges in the complexes. Simulations were run using Nose-Hoover method with NPT ensemble at 300 K temperature and pressure of 1 bar over a period of 100 ns (Reddy et al., 2015;Basha et al., 2014;and Nos� e, 1984). Best conformations were picked based on interactions and properties like RMSD and RMSF, and the timeline of interactions was investigated.

Molecular docking
The ligand-protein interaction of cigarette smoke constituents with a7 and a4b2 nicotinic acetylcholine receptors were investigated by molecular docking using iGEMDOCK. This software uses the tool GEMDOCK (Generic Evolutionary Method for Molecular DOCKing) for screening the ligand and predicting probable docking conformations with the lowest binding energy. Binding energy contribution for Hydrogen bonding and van der Waal's interactions were separately enumerated along with the total binding energy for each ligand and are provided in Table 1. The 3D docked positions of the ligands with the receptors (PDB ID; 5KXI and 5J5I) are depicted in Figures 1 and 2.
The docking of a4b2nAChR structure with the ligands nicotine, NNK, and NNN yielded docking scores of À 41.45 kcal/mol, À 59.28 kcal/mol, and À 54.60 kcal/mol, respectively. A comparison of binding free energy between the 3 ligands showed that NNK showed relatively smaller binding energy and higher affinity with a4b2nAChR structure. The figures show that all the ligands were occupying similar binding sites, that is, between the two side subunits as reported before. The molecular interactions conferring stability to the complexes include hydrogen bond, van der Waals, and hydrophobic interactions.
The a4b2nAChR formed hydrogen bonding interactions with both NNK and NNN. NNK forms 2 hydrogen bonds with T 150 . NNN forms 2 hydrogen bonds with S 136 . Nicotine could not develop any hydrogen bonds with any amino acid residues of a4b2nAChR.
The docking of a7nAChR molecule with the ligands nicotine, NNK and NNN yielded docking scores of À 59.54 kcal/ mol, À 71.06 kcal/mol and À 70.86 kcal/mol respectively. Both NNK and NNN exhibited similar docking scores and were higher than that of nicotine. Nicotine formed one hydrogen bond with S 70 . NNK developed 3 hydrogen bonds with N 66 ,H 69 and S 70. NNN could not develop any hydrogen bonds with any amino acid residues of a7nAChR. The main contributor of hydrogen bonds in NNK include the side chain carbonyl oxygen and nitro groups. These side chain groups are derived from the opening of the pyrrolidine ring structure in nicotine (Schuller & Orloff, 1998). The pyridine ring of nicotine is the contributor to hydrogen bond with a7nAChR. The pyrrolidine ring structure in NNN forms a hydrogen bond with a4b2nAChR. The additional side chain that offers  more room for hydrogen bonding and hydrophobic interactions may be contributing to higher stability during the docking of NNK with the receptors.

Molecular dynamics simulation
Molecular dynamics simulations were carried out to investigate the conformational stability of intermolecular interactions of the individual receptor-ligand complexes. The time-dependent changes in the structural flexibility and modifications in the interaction were assessed over a period of 100 ns using Desmond Maestro. Time-dependent trajectories of specific parameters like RMSD, RMSF, protein-ligand interactions, contact type bar diagrams, and timeline of interactions plots were used to investigate the extent of structural and conformational changes in each of them.

RMSD analysis
The conformational deviations of the receptor backbone (N, C a , and C bonds) and the stability of the ligand's interaction in the molecular dynamics simulation trajectory is mapped as a protein-ligand RMSD plot. The Protein RMSD plot assesses the structural conformations of protein when it is dynamically interacting with the ligand. Ligand RMSD plot indicates the stability of ligand interactions with a receptor/ protein binding domain. The RMSD plot of MD simulations for a4b2nAChR (PDB ID 5KXI) with the ligands nicotine, NNK, and NNN are shown in Figure 3 (a), (c), and (e), respectively. Observing the protein RMSD across the MD trajectory when bound to the ligands nicotine, NNK and NNN, it is evident that within 2 ns, the complexes were experiencing low oscillations and that across the entire simulation time frame of 100 ns the fluctuations noticed were within the acceptable range of 1-3 A o .
For nicotine, highly stable interactions with marginal oscillations in protein RMSD values were observed after 40 ns of simulation. For NNK and NNN, such stable interactions were observed after 2 ns and 20 ns of MD simulation course, respectively. The ligand RMSD plot for nicotine was stable for 50 ns, and the course of interactions for the rest of the period showed fluctuations higher than corresponding protein RMSD values. NNN ligand RMSD plot had shown a sharp spike in fluctuation at 50 ns of MD simulations and then stabilized for the rest of the time. For NNK, the oscillations in ligand RMSD values were observed at 40 ns and after 75 ns, rapid fluctuations were observed. The rapid fluctuations in Ligand RMSD for NNK might be an indication of ligand dissociation from the receptor active site. These dissociations are more evident after 75 ns where interactions are unstable. Oscillations in NNN at 50 ns also indicated the dissociation of ligand, but another cycle of binding ensued and the interactions were stable throughout the rest of the simulation period.
The RMSD plot of MD simulations for a7nAChR receptor (PDB ID 5J5I) with the ligands nicotine, NNK and NNN are shown in Figure 4 (a), (c) and (e), respectively. The protein RMSD plot of nicotine, upon initiation of MD simulation, showed a rise in RMSD value from 1.2 A o up to 5.4 A o over a period of 60 ns, and non-significant oscillations in RMSD values were observed for the rest of the simulation course. For NNK, protein RMSD values equilibrated at 2.80 A o by 20 ns of simulation time and were stabilized. Protein RMSD value of NNN, after initial equilibration at 2.5 A o , showed a nonsignificant rise at 30 ns and then fall and equilibrated to 3.5 A o at 50 ns which persisted till the end of simulation time. The ligand RMSD value of nicotine was presented with erratic oscillations with no visible stabilization observed across the whole simulation period. For NNK, fluctuations in ligand RMSD value were observed at 18 ns and 75 ns. But for the rest of the simulation period, an equilibrium ligand RMSD value in the range of 1.2-1.6 A o was observed. For NNN, after an initial rapid oscillation, by 20 ns, stabilization of ligand RMSD value was observed. Here, rapid swing in the Ligand RMSD for nicotine indicates an unstable interaction between the receptor active site and ligand.

RMSF analysis
The RMSF values characterize the localized deviation of a molecule from its original position as a function of time while it interacts with the ligands. Relative conformational flexibility of various segments of protein including secondary structures were identified. The RMSF plot of MD simulations for a4b2nAChR (PDB ID 5KXI) with the ligands nicotine, NNK and NNN are shown in Figure 3 (b), (d) and (f), respectively. The C terminal region of the receptor structure was found to fluctuate in comparison with the N terminal region. Vigorous fluctuations involving loop regions were observed during interactions with all three ligands. The RMSF plot of MD simulations for a7nAChR (PDB ID 5J5I) with the ligands nicotine, NNK and NNN are shown in Figure 4(b), (d) and (f), respectively. RMSF plots were similar for all 3 ligands. The N terminal alpha helices displayed rigid structures that stabilized the RMSF values, compared to loop structures at the end that showed oscillations.

Intermolecular interaction studies
The molecular interactions between ligands and amino acid residues of the receptors upon MD simulations were investigated. 2D ligand interaction with receptor amino acid residues that lasted for more than 30% of MD simulation time was represented in the 2D interaction diagrams. Receptor-ligand contact subtypes (H-bonds, Hydrophobic, Ionic and water bridges) are plotted as bar charts showing a normalized fraction of the time of interaction with individual amino acid residues across the simulation time of 100 ns.
The receptor-ligand 2D interaction diagrams and bar diagram depicting temporally significant contacts for a4b2nAChR receptor (PDB ID 5KXI) with the ligands nicotine, NNK, and NNN are given in Figure 5. The 2D interaction diagram of nicotine shows hydrogen bond interaction with a subunit S 134 for more than 30% of simulation time. NNK and NNN were found to have lost the interactions and were observed to be moved away from the receptor binding site compared to nicotine. When we analyze the bar chart showing time-dependent interactions and timeline of contacts made, we observe that for nicotine, as expected, S 134 offered significant contacts for about 70% of the interaction interval with the ligand. For NNK, a subunit N 54 that offers water bridge connections and b subunit E 177 with hydrogen bond interactions were awarding stable interactions but were less frequent than for nicotine. For NNN, the normalized temporal interaction fraction was much less, with b subunit H 46 with hydrophobic contacts and E 177 with hydrogen bond interactions were predominant with interaction frequency less than 30%.
The receptor-ligand 2D molecular interaction diagrams and bar diagram depicting temporally significant contacts for a7nAChR receptor (PDB ID 5J5I) with the ligands nicotine, NNK and NNN are given in Figure 6. The 2D interaction diagram of nicotine was found to be devoid of interactions and were found to have dissociated from the receptor docking site. For NNN, hydrophobic interactions were observed with W 143 , with an aqueous solvation layer around the atoms for 30% of the simulation time. The residues W 143 , Y 185 and Y 89 offered interactions, predominantly hydrogen bonds, for 70%, 60% and 55% of simulation interval, respectively, with the ligand NNN. The bar diagram and timeline of interactions show that nicotine exhibited short-term insignificant interactions with the residues in the receptor, for up to 3% of interaction fraction only. For NNK, the bar diagram shows that M 114 contributed hydrophobic, water bridge and hydrogen bond interactions for 80% of the simulation time. Other residues were offering interactions for less than 40% of the simulation time. The timeline interaction diagrams in Supplementary figures S5 and S6 is showing the individual contribution of key residues over the entire simulation time of 100 ns. Also shown in the panel on top of each figures are the time-dependent enumeration of the number of contacts that complements the result interpretation given above.
Observations show that for a4b2nAChR, nicotine and NNK were offered interactions by receptor active site for more than 70% of simulation time compared to NNN. The larger number of amino acid residues of a4b2nAChR with which NNK and NNN interacted proves that the ligands were dissociating from the receptor active site and redocking. This complements the ligand RMSD results given in Figure 3. The amount of time for which Water Bridge interactions were observed with both the ligands as shown in Figure 5 also points to this assumption. a7nAChR active site formed stable interactions with NNK and NNN but not with nicotine. We observed that with a4b2nAChR, nicotine, NNN and NNK are developing intermolecular interactions with similar amino acid residues in the receptor active site, even if the relative interaction fractions were different as shown in Figure 5. With a7nAChR, we could not observe any similarity in key residues contributing to the interaction ( Figure 6). Ligands, NNK and NNN forged stable interactions with a disparate set of residues, and nicotine rendered very weak intermolecular interactions during the timeline of simulations and interacted with a wide array of residues alluding to the minimal stable binding. For nicotine, the interaction time fraction was very less and that display of large number of amino acid residues with which interactions were reported showed that the ligand was dissociating from the receptor active site and that it was non-committal in binding and hydrated in an aqueous environment. Intermolecular interactions were stable for NNK and NNN compared to nicotine. We compared the key molecular interactions of the ligand nicotine with receptors in our investigation with that of the previously reported a4b2 nAChR homology model done through molecular docking simulations (Mao et al., 2016). The interaction of nicotine with a7nAChR homology model using a competitive docking model and Glide-based docking was also compared with our study results (Ng et al., 2018). Surprisingly, we could not observe any similarity in the key residues that offered intermolecular interactions.

Protein secondary structure elements and ligand properties
Secondary structure elements (a-helices and b-strands) in the receptors were examined over the entire simulation time across the whole protein. SSE distribution by residue index and SSE composition over the simulation time for a4b2 and a7nAChR's with the ligands are shown in Supplementary figures S1 and S2. For a4b2 receptor, we observed that alternating a-helices and b-strands were presented during the entire simulation time with all three ligands. For a7nAChR, upon ligand binding, we could observe predominant b-strand structures interlude small patches of residues devoid of such structures. Ligand torsion profiles provide a concise view of the evolution of the conformational torsion of the rotatable bonds in the ligands during the course of the simulation. The torsion profile gives insight into rotatable bonds and their conformational strain to maintain protein-bound conformations. The radial plots represent the conformation of torsion radiating outward from the center during the simulation period. The bar plots represent potential energy in kCal/mol corresponding to the torsion angles for each rotational bond and their probability densities. The ligand torsion profiles for NNK and NNN with a4b2 and a7nAChR receptors are separately shown in Supplementary figure S3. During simulation, it was noted that both NNK and NNN possessed rotatable bonds. Nicotine contains no rotatable bond. NNK had 7 rotatable bonds and NNN had 3 rotatable bonds. One bond each among NNK and NNN was not showing torsional potential energy during the simulation time and lacking conformational strain. More conformational strains were observed in bonds associated with intermolecular interactions, like C-C bonds near carbonyl and hydroxyl groups in NNK and hydroxyl group of NNN. Ligand properties including ligand RMSD, Radius of Gyration (rGyr), number of intramolecular hydrogen bonds (IntraHB), Molecular Surface area (MolSA), Solvent Surface area (SASA) and Polar Surface area (PSA) were also assessed for all the ligands with a4b2 and a7nAChRs and are separately provided in Supplementary figure S4.

Conclusion
Both a4b2 and a7nAChRs are physiologically significant members of nicotinic acetylcholine receptors associated with nicotine addiction. The role of nicotine and nitrosamines in eliciting physiological responses like proliferation, angiogenesis and metastasis in cancer cells via these receptors is also well established. The binding interactions for nitrosamine constituents and nicotine with a4b2 and a7nAChRs were investigated using molecular docking and further characterized using MD simulations. The binding energy studies with molecular docking showed that NNK and nicotine could render stable binding with both the receptor active sites as inferred from lower binding energy for the same. The pyrrolidine ring structure present in nicotine is replaced by a side chain that offered more hydrogen bonds is the reason for better docking performance by NNK. The pyridine ring structure and nitro group in NNN contributed to hydrogen bonds with a7nAChR. During MD simulation analysis, RMSD analysis showed that nicotine and NNK contributed stable ligand and protein RMSD values during simulation time inferring stable ligand-receptor binding in comparison with NNN. But, both Ligand RMSD and intermolecular interactions showed that NNK and NNN dissociated and re-docked to the receptor showing lesser stability. With a7nAChR, nicotine was also found to be dissociated from the binding site, rendering less stable interaction. From all the studies, we could see that NNK offered better binding affinity but offered lesser binding stability with a4b2 nAChR. NNN could offer good binding energy and good stability of ligand RMSD and intermolecular interactions with a7nAChR.We compared the key molecular interactions of nicotine with an a4b2 nAChR homology model done through MD simulations and that of a7nAChR homology model using a competitive docking model and Glide-based docking that were previously reported. Surprisingly, we could not observe any similarity in the key residues that offered intermolecular interactions.
In vitro studies that were reported earlier showed that NNN had 5000 times more affinity toward a4b2 nAChR compared to nicotine and NNK had 1300 times more affinity toward a7nAChR than with nicotine. The current study concurs with the higher affinity of NNK for both the receptors and the higher affinity of NNN with a7nAChR. Concluding, the study will render a better understanding of the receptor-ligand interactions among the physiologically most significant tobacco constituents like NNK and NNN. More investigations need to be carried out to study the interactions of nitrosamines and nicotine with nAChRs. The information from ligand active site contributors of hydrogen bonding will be useful in screening for better active groups in potential nicotine agonists and antagonists for pharmacophore modeling also. These interaction studies would also help us to perform in silico screening for developing novel a7and a4b2 nAChRs agonists for the treatment of neurological and psychiatric disorders as well as antagonists with smoke de-addiction and potential antiproliferative properties.

Disclosure statement
No potential conflict of interest was reported by the authors