Supplementary Information Structural Assessment of Agonist Efficacy in the -Opioid Receptor: Morphine and Fentanyl Elicit Different Activation Patterns Adrián Ricarte1,2,3, James A. R. Dalton1,2,3* and Jesús Giraldo1,2,3* 1 Laboratory of Molecular Neuropharmacology and Bioinformatics, Unitat de Bioestadística and Institut de Neurociències, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain 2 Unitat de Neurociència Traslacional, Parc Taulí Hospital Universitari, Institut d’Investigació i Innovació Parc Taulí (I3PT), Institut de Neurociències, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain 3 Instituto de Salud Carlos III, Centro de Investigación Biomédica en Red de Salud Mental, CIBERSAM, 28029 Madrid, Spain *Corresponding authors: James A. R. Dalton, E-mail: James.Dalton@uab.es Jesús Giraldo, E-mail: Jesus.Giraldo@uab.es Figure S1. Stability of BU-72 bound in activated hMOR model during control MD simulations. Plots showing stability in terms of RMSD of a) ligand, and TM domain with respect to: b) initial receptor conformation or c) fully-active crystal structure. Shades of blue represent replicas #1 and #2 of BU-72-bound activated hMOR state. Straight-lines indicate crystallized activated or inactive conformations with respect to fully-active crystal structure. Figure S2. Morphine, fentanyl and naltrexone conformational clustering and RMSD in MD simulations. a) Morphine, fentanyl and naltrexone poses observed over respective trajectories of each replica starting from activated or inactive receptor states with bound fentanyl, morphine or naltrexone, classified into 5 different clusters, using a cutoff of 2.0 Å between groups, with cluster 1 predominant in each case (fentanyl, morphine or naltrexone clusters #1-5 in: i) red, dark blue, black, yellow and light green, ii) light blue, orange, purple, salmon and dark green, or in iii) mauve, sky blue, pale green, khaki, brown; respectively). b) Plots showing ligand physical stability in terms of RMSD for replicas #1-3 of bound fentanyl, morphine or naltrexone (shades of green, red or purple, top, mid or bottom rows, respectively) starting from activated or inactive state (left or right columns, respectively). Figure S3. Proportion of residues in close contact with morphine, fentanyl and naltrexone in MD simulations. Residues of hMOR orthosteric pocket closely contacted (<3.5 Å, with at least two trajectories with a contact frequency >20.0%, independently of the initial state) by fentanyl, morphine or naltrexone in replicas #1-3 (shades of green, red or purple, respectively) of MD simulations starting from activated or inactive state. Table S1. Average and SD of metrics: average protein RMSD from i) initial state or ii) with respect to fully active crystal, average iii) TM3 upward, iv) TM5 inward, v) TM6 outward or vi) NPxxY inward movement with respect to protein center (normalized by inactive crystal structure), or vii) average distance between Arg1673.50-Thr2816.34 of each replica #1-3 starting from activated or inactive receptor state with either fentanyl, morphine or naltrexone bound. Metric (threshold) Replicas Fentanyl bound Morphine bound Naltrexone bound Average protein RMSD from initial state (±3.5 Å) Replica #1 3.2 Å (0.2 SD) 3.5 Å (0.3 SD) 3.8 Å (0.2 SD) Starting from activated hMOR Replica #2 3.5 Å (0.4 SD) 3.1 Å (0.5 SD) 4.1 Å (0.3 SD) Replica #3 3.3 Å (0.2 SD) 3.1 Å (0.2 SD) 3.8 Å (0.3 SD) Replica #1 3.1 Å (0.2 SD) 3.2 Å (0.3 SD) 3.0 Å (0.2 SD) Starting from inactive hMOR Replica #2 2.9 Å (0.2 SD) 2.4 Å (0.2 SD) 2.5 Å (0.2 SD) Replica #3 2.8 Å (0.3 SD) 2.9 Å (0.2 SD) 3.0 Å (0.3 SD) Average protein RMSD from fully active crystal (±3.0 Å) Replica #1 2.9 Å (0.2 SD) 2.4 Å (0.2 SD) 2.8 Å (0.2 SD) Starting from activated hMOR Replica #2 2.9 Å (0.3 SD) 2.3 Å (0.2 SD) 3.4 Å (0.4 SD) Replica #3 2.4 Å (0.2 SD) 2.4 Å (0.4 SD) 2.9 Å (0.2 SD) Replica #1 4.3 Å (0.2 SD) 4.0 Å (0.3 SD) 3.6 Å (0.2 SD) Starting from inactive hMOR Replica #2 3.8 Å (0.5 SD) 3.5 Å (0.2 SD) 3.4 Å (0.2 SD) Replica #3 3.7 Å (0.3 SD) 4.0 Å (0.2 SD) 3.8 Å (0.3 SD) Average TM3 upward movement (+0.3 Å) Replica #1 +0.4 Å (0.2 SD) +0.2 Å (0.2 SD) +0.4 Å (0.2 SD) Starting from activated hMOR Replica #2 +0.5 Å (0.3 SD) +0.3 Å (0.2 SD) -0.5 Å (0.3 SD) Replica #3 +0.6 Å (0.2 SD) +0.2 Å (0.3 SD) +0.1 Å (0.2 SD) Replica #1 +0.1 Å (0.2 SD) +0.3 Å (0.2 SD) +0.3 Å (0.2 SD) Starting from inactive hMOR Replica #2 -0.1 Å (0.2 SD) +0.6 Å (0.2 SD) +0.6 Å (0.2 SD) Replica #3 -0.4 Å (0.3 SD) +0.2 Å (0.2 SD) +0.1 Å (0.2 SD) Average TM5 inward movement (-0.8 Å) Replica #1 -1.7 Å (0.5 SD) -1.5 Å (0.4 SD) -0.3 Å (0.4 SD) Starting from activated hMOR Replica #2 -1.4 Å (0.4 SD) -1.3 Å (0.4 SD) +1.8 Å (0.9 SD) Replica #3 -1.3 Å (0.2 SD) -1.2 Å (0.6 SD) +0.2 Å (0.6 SD) Replica #1 +0.9 Å (0.4 SD) -0.5 Å (0.9 SD) +0.8 Å (0.4 SD) Starting from inactive hMOR Replica #2 +0.8 Å (0.4 SD) -1.5 Å (0.8 SD) -0.4 Å (0.9 SD) Replica #3 +1.7 Å (0.4 SD) +0.7 Å (0.5 SD) +0.4 Å (0.5 SD) Average TM6 outward movement (+1.7 Å) Replica #1 +3.3 Å (0.8 SD) +3.4 Å (0.6 SD) +0.7 Å (0.5 SD) Starting from activated hMOR Replica #2 +3.7 Å (0.5 SD) +3.6 Å (0.5 SD) +1.2 Å (0.7 SD) Replica #3 +3.4 Å (0.5 SD) +3.3 Å (0.6 SD) +0.6 Å (0.5 SD) Replica #1 +0.8 Å (0.4 SD) +0.5 Å (0.5 SD) +1.0 Å (0.4 SD) Starting from inactive hMOR Replica #2 +1.0 Å (0.5 SD) +0.5 Å (0.4 SD) +0.4 Å (0.5 SD) Replica #3 +1.6 Å (0.7 SD) +0.3 Å (0.4 SD) +0.6 Å (0.5 SD) Average NPxxY inward movement (-0.8 SD) Replica #1 -2.9 Å (0.7 SD) -2.6 Å (0.6 SD) +0.0 Å (0.5 SD) Starting from activated hMOR Replica #2 -2.3 Å (0.3 SD) -2.4 Å (0.3 SD) -2.4 Å (0.6 SD) Replica #3 -2.4 Å (0.5 SD) -2.8 Å (0.5 SD) -0.2 Å (0.8 SD) Replica #1 -1.3 Å (0.3 SD) -1.5 Å (0.3 SD) +0.2 Å (0.3 SD) Starting from inactive hMOR Replica #2 -1.1 Å (0.3 SD) -0.0 Å (0.3 SD) -0.1 Å (0.3 SD) Replica #3 -1.9 Å (0.7 SD) +0.3 Å (0.4 SD) +0.4 Å (0.4 SD) Average Arg-Thr distance (±10 Å) Replica #1 12.2 Å (0.7 SD) 12.5 Å (0.8 SD) 7.4 Å (0.5 SD) Starting from activated hMOR Replica #2 13.6 Å (0.6 SD) 13.4 Å (0.7 SD) 6.3 Å (0.6 SD) Replica #3 13.3 Å (0.6 SD) 12.7 Å (1.0 SD) 6.3 Å (0.6 SD) Replica #1 6.2 Å (0.5 SD) 6.6 Å (0.6 SD) 6.2 Å (0.4 SD) Starting from inactive hMOR Replica #2 6.1 Å (0.3 SD) 6.6 Å (0.4 SD) 6.3 Å (0.4 SD) Replica #3 8.0 Å (1.1 SD) 6.0 Å (0.2 SD) 6.1 Å (0.3 SD) Figure S4. Entire TMD receptor conformational change. TM domain RMSD with respect to: a) initial receptor conformation or b) fully-active crystal. Green, red or purple shades represent replicas #1-3 with bound fentanyl, morphine or naltrexone, respectively, in MD simulations starting from activated or inactive state (left or right columns, respectively). Straight-lines indicate 3.5 Å threshold of conformational change with respect to fully-active crystal or initial state of activated and inactive crystal structures. Figure S5. TM helix movement with respect to protein center in MD simulations I. As a function of time, and normalized by inactive crystal position29, a) TM3 z-offset vertical movement, and b) distance between protein center and intracellular tip of TM5, in replicas #1-3 with bound fentanyl, morphine or naltrexone (green, red or purple shades, respectively) starting from activated or inactive state (respective left or right columns). Flat-lines indicate respective metrics measured in fully-active, activated and inactive crystals. Figure S6. TM helix movement with respect to protein center in MD simulations II. As a function of time, and normalized by inactive crystal position29, distance between protein center and a) intracellular tip of TM6, or b) NPxxY motif, in replicas #1-3 with bound fentanyl, morphine or naltrexone (green, red or purple shades, respectively) starting from activated or inactive state (respective left or right columns). Flat-lines indicate respective metrics measured in fully-active, activated and inactive crystals. Figure S7. Intracellular Arg1673.50-Thr2816.34 distance in MD simulations. As a function of time, a) intracellular Arg1673.50-Thr2816.34 distance in MD simulations starting from activated or inactive state (left or right columns, respectively). Green, red or purple shades represent replicas #1- #3 with bound fentanyl, morphine or naltrexone, respectively. Straight lines indicate the thresholds dividing active- from inactive-like conformations. Figure S8. Average potential energy per replica of protein-ligand and sidechain-sidechain interactions in MD simulations. Different shades of grey show average potential energy and standard deviation of relevant interactions closely contacted by fentanyl or morphine in replicas #1- 3 starting from activated or inactive state, and the potential energy of Asp1493.32-Tyr3287.43 interaction. Figure S9. Ligand-Asp1493.32 and TM3-TM7 interaction in MD simulations. As a function of time, replicas #1-3 with bound fentanyl, morphine or naltrexone (shades of green, red or purple, respectively) starting from activated or inactive states of hMOR (left or right columns, respectively) showing a) interaction distance (<4.5 Å) between Asp1493.32 and each ligand amine group, or b) Asp1493.32 χ1 dihedral angle. Straight lines represent thresholds of interaction formation distance of 4.5 Å or Asp1493.32 χ1 dihedral angle of 115º, respectively. Figure S10. Met1533.36 conformational change in MD simulations. As a function of time, a) Asp1493.32-Tyr3287.43 distance fluctuation within interaction formation distance (<4.5 Å), and b) Met1533.36 χ1 dihedral angle, of replicas #1-3 with bound fentanyl, morphine or naltrexone (top, mid and bottom rows, green red or purple shades, respectively) starting from activated or inactive hMOR state (left or right columns, respectively). Straight lines represent thresholds of 4.5 Å or 90º for interaction formation distance or Met1533.36 χ1 dihedral angle, respectively. Figure S11. Gln1262.60 conformational change in MD simulations. Graphs showing as a function of time a) Gln1262.60 χ2 dihedral angle conformational change of replicas #1-3 starting form activated or inactive state (left or right columns, respectively) with bound fentanyl, morphine or naltrexone (green, red or purple shades, respectively). Straight lines represent threshold of 240º for Gln1262.60 χ2 dihedral angle. Figure S12. Lys2355.39 χ2 conformational change in MD simulations and interaction with ligand bound. Graphs showing as a function of time a) distance fluctuation between Lys2355.39 amino group and respective ligand oxygen acceptors within interaction formation distance (<4.5 Å), and c) Lys2355.39 χ2 dihedral angle conformational change, of replicas #1-3 starting form activated or inactive state (left or right columns, respectively) with bound fentanyl, morphine or naltrexone (green, red or purple shades, respectively). Straight lines represent thresholds of 4.5 Å or 240º for interaction formation distance or Lys2355.39 χ2 dihedral angle, respectively. Figure S13. Distribution of orthosteric pocket conformational changes (independent of agonist bound or starting receptor state) in MD simulations. Distribution of a) Asp3.32-Tyr7.43 interaction distance with respect to Gln1262.60 conformation, b) Lys2355.39 amine group-ligand oxygen interaction distance and Lys2355.39 conformation, or c) Asp3.32-Tyr7.43 H-bonding and Ala2425.46- Trp2956.48 interaction distance. Figure S14. Trp2956.48 conformation and interaction with Ala2425.46 in MD simulations. Graphs showing, as a function of time, a) Trp2956.48 χ2 dihedral angle, and b) Trp2956.48-Ala2425.46 distance, of replicas #1-3 with bound fentanyl, morphine or naltrexone (top, middle and bottom rows, colored with green, red or purple shades, respectively) starting from activated or inactive hMOR states (left or right columns, respectively). Figure S15. Statistical analyses to assess the association between orthosteric and intracellular metrics I. Chi2 statistical analysis of orthosteric metrics: a) Asp1493.32-Tyr7.43 H-bond, b) Tyr1503.33-ligand H-bond or c) Lys2355.39-ligand H-bond; comparing active- or inactive-like states of intracellular metrics, independently of the ligand bound or the starting state of the receptor: i) TM3 upward-, ii) TM5 inward-, iii) TM6 outward-, iv) NPxxY inward-movement, or v) Arg1673.50- Thr2816.34 distance. Figure S16. Statistical analyses to assess the association between orthosteric and intracellular metrics II. Chi2 statistical analysis of orthosteric metrics of a) Trp2956.48-Ala2425.46 H-bond or b) Met1533.36 gauche- conformation selection; or c) T-test of Trp2956.48-Ala2425.46 distance considering equal or unequal variances depending on the Levene test; comparing active- or inactive- like states of intracellular metrics, independently of the ligand bound or the starting state of the receptor: i) TM3 upward-, ii) TM5 inward-, iii) TM6 outward-, iv) NPxxY inward-movement, or v) Arg1673.50-Thr2816.34 distance. Figure S17. Fentanyl docking in active MOR crystal structures. Upward and lateral view of docked fentanyl (purple) in activated hMOR model (green) compared with docked fentanyl (salmon) in fully active Gi-bound mMOR crystal structure30 (blue).