OR51E2 is a receptor for propionic and acetic acids, and its agonists tend to smell vinegar-like. Other short chain fatty acids (SCFAs) with chains of up to 5 carbon atoms activate this protein progressively less strongly with increasing chain length, and SCFAs of 6 carbons or more are non-agonists and have odors that are more cheesy, sweaty, or milky instead of acetic. OR51E2 is also sensitive to alpha and beta ionones, and in fact individuals with a mutated form of another receptor (OR5A1) report an acetic odor for β-ionone instead of a pleasant floral odor, correlating to the mutant 5A1 protein’s insensitivity to that aroma compound. Here we present a molecular docking pose for beta ionone in OR51E2, using PrimaryDock in “soft pocket” mode.
The command to generate this dock was:
bin/primarydock primarydock.config --prot pdbs/OR51/OR51E2.8f76.h.pdb --lig sdf/beta-ionone.sdf --pose 25 --appendprot --soft TMR2 TMR4 TMR5 TMR6 TMR7…where OR51E2.8f76.h.pdb is a version of the OR51E2.8f76.pdb file included with PrimaryDock, that had been hydrogenated using the HYDRO command of the Pepteditor utility before its use in PrimaryDock. This model is in turn based on the cryogenic electron microscopy model of OR51E2 bound to propionate RCSB 8f76. The “soft dock” parameter has the effect of moving the transmembrane helices during docking in order to minimize clashes and maximize intermolecular binding. This is pose #2 of a total of 14, most of which place the keto group of the ligand in a similar position to this pose.
The resulting pose has good coordinate bindings to His1043.33 (-18 kJ/mol), Phe1554.57 (-3.24 kJ/mol), Leu1584.60 (-3.65 kJ/mol), Gly1985.39 (-3.11 kJ/mol), Ile2025.43: (-8.87 kJ/mol), Leu2546.51 (-9.26 kJ/mol), Leu2576.54 (-5.06 kJ/mol), and Ser2586.55 (-6.75 kJ/mol). Residues 104, 155, 158, 198, 202, and 258 are also involved in propionate binding. The binding energy to Arg2626.59 is only -1.62 kJ/mol, suggesting that this residue is less important for ionones than for carboxylic acids, and that ionones’ coordination to TMR6 is based more on hydrophobic forces vs. the ionic bond formed by propionate. The placement of the ligand suggests that methyl ionone gamma would be less likely to activate OR51E2, since the added methyl group would “push against” His104 and decrease the binding energy to that residue. Methyl ionone gamma has a cleaner odor, more muguet and less acetic compared to alpha and beta ionones.
The result of the soft dock setting was to move TMR5 0.63 Å outward from protein center, TMR6 0.95 Å outward, and TMR7 0.45 Å outward. TMR4, meanwhile, moved inward 0.77 Å. There were also other horizontal, vertical, and rotational motions of the TMR helices, all of which are included in the .dock file. The expansion of helices V through VII appears to be necessary to accommodate the bulky 6-membered ring of ionone.
The results suggest that certain residues may be influential in determining ligand agonism for a given olfactory receptor, besides a general requirement for strong binding to TMR3 and/or TMR4 in addition to the active state position of TMR6. It can be noted that Asp3.32 is believed to be important for hTAAR ligand binding, being conserved in all six human TAARs, while Asp5.42 is predicted to be involved in binding of TAAR6 and TAAR8 to the diamines cadaverine and putrescine. The fact that both residues are directly adjacent to residues identified for OR51E2 agonists implies that these are active binding zones for olfactory GPCRs. Additionally, all four positions 3.32, 3.33, 5.42, and 5.43 were identified as ligand binding site residues by Man et al (2004).