Search sphere This sphere includes the acetylcholine

Search sphere 1: This sphere includes the JQ1 binding site between the α–γ subunits. The center point of the search area was located at x=35.26, y=77.67, z=137.74 and the initial search radius was set to 27Å which assures that the corresponding area outside the channel pore of the two subunits was covered. After visual inspection all ligand poses that were detected were located close to the search center point and a refined docking with a radius of 12Å was conducted to see if it would yield other results. However, the found binding poses were identical to those found with the large search radius of 27Å. This was done for the five search areas search sphere 1, 2, 3, 4 and 5 covering the outside surface of the extracellular domain. Search sphere 2: It includes the second acetylcholine binding site between the α–δ subunits. The center point of the search area was located at x=76.87, y=40.73, z=135.82 and the search radius was set to 27Å. Search sphere 3: x=44.87, y=48.93, y=132.44, located in the cavity between the γ–C loop and the α subunits. Search sphere 4: x=66.44, y=89.73, z=135.92, located in the cavity between the δ–C loop and the β subunit. Search sphere 5: x=91.29, y=68.90, z=136.43, located in the cavity between the β–C loop and the α subunit. Search sphere 6: The center point of search was located in the middle of the channel pore at x=62.05, y=64.48, z=146.26 and the radius of search was set to 27Å. It covers the upper part of the channel pore from the extracellular entrance of the pore down to overlap with search sphere 7 for about 10Å. Search sphere 7: The center point of search was located in the middle of the channel pore at x=64.31, y=65.37, z=110.53 and the radius of search was set to 26Å. It covers in particular the central part of the inner channel reaching from the middle of the extracellular domain down into the transmembrane region for approx. 10Å. Search sphere 8: The center point of search was located in the middle of the channel pore and almost in the middle of the transmembrane helices region at x=63.26, y=64.29, z=95.76. The radius of search was set to 20Å. It covers the channel pore inside the transmembrane area down to the intracellular exit of the channel (Fig. 2).
Results The virtual screening yielded the highest GoldScore (Jones et al., 1997) values for MB327 in two binding sites inside the nAChR channel (Table 1), binding site MB327-1 in search sphere 6 (GoldScore=29.7) in the extracellular domain and binding site MB327-2 in search sphere 7 (GoldScore=29.9) at the start of transmembrane helices (Fig. 3). In both sites, MB327-1 and MB327-2, the GoldScores were almost identical making a differentiation based on the docking calculations impossible. But for both sites the GoldScore is higher than in the orthosteric binding sites ACh-1 between α–γ subunits (26.0) and ACh-2 between α and δ (24.6) indicating that a binding event of MB327 is more likely to occur in one of the channel binding sites (MB327-1 or MB327-2) than in the orthosteric binding sites. This is in good agreement with the findings of Niessen et al. (Niessen et al., 2011, Niessen et al., 2013) who demonstrated that MB327 doesn’t show any interaction with the acetylcholine binding site of Torpedo nAChR. For nicotine the GoldScores are also unambiguous. The highest scores (35.0 and 34.8, search spheres 1 and 2 in Table 1) are obtained in the orthosteric binding sites ACh-1 and ACh-2 which was expected since nicotine is known to bind to orthosteric binding sites. Acetylcholine (23.9) and carbachol (21.9) show the highest GoldScore values for the ACh-2 binding site between the α–δ subunits (search sphere 2, Table 1). The next best GoldScore is calculated inside search sphere 6 inside the channel pore (22.2 for acetylcholine and 21.8 for carbachol) while the second orthosteric site, ACh-1, scores only third (20.7 for acetylcholine and 20.5 for carbachol, search sphere 1, Table 1). However, search sphere 6 covers the complete inner surface of the channel pore including all the subunits and acetylcholine and carbachol don’t dock into the MB327-1 binding site in search sphere 6 between the γ–α subunits. Both compounds are docking between the α–δ subunits in another putative binding site. The best GoldScore for epibatidine is calculated in binding site ACh-1 (42.8). Although the score of epibatidine in ACh-2 (41.6) is only slightly smaller, epibatidine shows a rather high GoldScore of 42.7 in the search sphere 4 between the δ–β subunits. That the two orthosteric binding sites are not calculated with the highest scoring for acetylcholine, carbachol and epibatidine might be closely related to the nAChR 3D structure used in the present study which represents the closed conformation. This closed conformation of the nAChR may significantly differ from the open protein conformation which might be more relevant for ligand binding and hence give rise to more significant differences of docking scores for the orthosteric binding sites as compared to other possible sites in the nAChR protein. Considering conformational changes of the protein backbone and of the amino acid side chains for the docking procedure would have required very time intensive molecular dynamics calculations. Hence, to limit computation time to a tolerable level no such studies including also protein flexibility could be performed.