Mobitec Gmbh | Date: 2007-09-11
Flourescent chemicals and biomolecule labeling kits comprised primarily of reagants for use in scientific research.
No statistical methods were used to predetermine sample size. Chemicals and reagents used in this study were purchased from commercial sources (Sigma, Tocris, Fisher scientific, ZINC database suppliers) or synthesized as outlined in the Supplementary Information. HEK293 (ATCC CRL-1573; 60113019; certified mycoplasma free and authentic by ATCC) and HEK293-T (HEK293T; ATCC CRL-11268; 59587035; certified mycoplasma free and authentic by ATCC) cells were from the ATCC and are well validated for signalling studies. Cells were also validated by analysis of short tandem repeat (STR) DNA profiles and these profiles showed 100% match at the STR database from ATCC. U2OS cells expressing human μOR were obtained as cryopreserved stocks from DiscoverX and were not further authenticated. The inactive-state μ-opioid receptor structure (PDB: 4DKL) was used as input for receptor preparation with DOCK Blaster (http://blaster.docking.org)44. Forty-five matching spheres were used based on a truncated version of the crystallized ligand. The covalent bond and linker region of the antagonist β-funaltrexamine were removed for sphere generation. The ligand sampling parameters were set with bin size, bin size overlap, and distance tolerances of 0.4 Å, 0.1 Å, and 1.5 Å, respectively, for both the matching spheres and for the docked molecules. Ligand poses were scored by summing the receptor-ligand electrostatics and van der Waals interaction energy corrected for ligand desolvation. Receptor atom partial chargers were used from the united atom AMBER force field except for Lys233 and Tyr326, where the dipole moment was increased as previously described43. Over 3 million commercially available molecules from the ZINC20 (http://zinc.docking.org) lead-like set were docked into the receptor using DOCK3.621 (http://dock.compbio.ucsf.edu). Among the top ranking 0.08% of molecules were inspected and 23 were selected for experimental testing in the primary screen. A resource to perform these docking studies is publicly available (http://blaster.docking.org). For a secondary screen, analogues of the top three hits from the primary screen (compounds 4, 5 and 7) with a similarity of greater than 0.7 (as defined in the ZINC search facility) were identified in the ZINC database. Additionally, substructure searches were performed using the scaffolds of each of these three compounds. The searches yielded 500 purchasable compounds, which were then docked as in the primary screen. Analogues were manually inspected for interactions and selected for further experimental testing. For a primary screen of selected molecules, binding to μOR was assessed by measuring competition against the radioligand 3H-diprenorphine (3H-DPN). Each compound was initially tested at 20 μM and was incubated with 3H-DPN at a concentration equal to the K (0.4 nM) of the radioligand in μOR containing Sf9 insect cell membranes. The reaction contained 40 fmol of μOR and was incubated in a buffer of 20 mM HEPES pH 7.5, 100 mM sodium chloride, and 0.1% bovine serum albumin for 1 h at 25 °C. To separate free from bound radioligand, reactions were rapidly filtered over Whatman GF/B filters with the aid of a Brandel harvester and 3H-DPN counts were measured by liquid scintillation. Compounds with more than 25% of 3H-DPN radioactivity were further tested in full dose–response to determine the affinity (K ) in HEK293 membranes. Subsequently, the 15 analogues were tested in full dose–response for affinity at the μOR and the κOR by the National Institutes of Mental Health Psychoactive Drug Screen Program (PDSP)45, as were the affinities of compounds 12, PZM21, and their stereoisomers at the μOR, δOR, κOR and nociception receptor. Radioligand depletion assays to test the irreversible binding of compound PZM29 were performed as described previously46. Human embryonic kidney 293 (HEK 293) cells were transiently transfected with μOR or the cysteine mutant μOR:N127C using the Mirus TransIT-293 transfection reagent (MoBiTec, Goettingen, Germany), grown for 48 h, harvested, and homogenates were prepared as described47. For radioligand depletion experiments, homogenates were preincubated in TRIS buffer (50 mM Tris at pH 7.4) at a protein concentration of 50–100 μg/ml or 70–120 μg/ml for μOR and μOR:N127C, respectively and the covalent ligand (at 5 μM) for different time intervals. Incubation was stopped by centrifugation and reversibly bound ligand was washed three times (resuspension in buffer for 30 min and subsequent centrifugation). Membranes were then used for radioligand binding experiments with 3H-diprenorphine (final concentration: 0.7 nM, specific activity: 30 Ci/mmol, purchased from Biotrend, Cologne, Germany) to determine specific binding at the μOR (B = 4,000–6,500 fmol/mg protein, K = 0.25–0.45 nM) and the μOR:N127C receptor (B = 1,300–6,000 fmol/mg protein, K = 0.18–0.25 nM), respectively as described48. Non-specific binding was determined in the presence of 10 μM naloxone. For data analysis, the radioactivity counts were normalized to values where 100% represents effect of buffer and 0% represents non-specific binding. Five independent experiments, each done in quadruplicate, were performed and the resulting values were calculated and pooled to a mean curve which is displayed. The [35S]-GTPγS binding assay was performed with membrane preparations from HEK 293 cells coexpressing the human μOR and the PTX insensitive G-protein subunits G or G 49. Cells were transiently transfected using the Mirus TransIT-293 transfection reagent (MoBiTec, Goettingen, Germany), grown for 48 h, harvested and homogenates were prepared as described47. The receptor expression level (B ) and K values were determined in saturation experiments with 3H-diprenorphine (specific activity: 30 Ci/mmol, purchased from Biotrend, Cologne, Germany) (B = 3,700 ± 980 fmol/mg protein, K = 0.30 ± 0.093 nM for μOR+G or B = 5,800 ± 2,000 fmol/mg, K = 0.46 ± 0.095 nM for μOR+G , respectively). The assay was carried out in 96-well plates with a final volume of 200 μl. In each well, 10 μM GDP, the compounds (0.1 pM to 100 μM final concentration) and the membranes (30 μg/ml final protein concentration) were incubated for 30 min at 37 °C in incubation buffer containing 20 mM HEPES, 10 mM MgCl • 6 H O and 70 mg/l saponin. After the addition of 0.1 nM [35S]-GTPγS (specific activity 1,250 Ci/mmol, PerkinElmer, Rodgau, Germany) incubation was continued at 37 °C for further 30 min or 75 min for μOR+G or μOR+G , respectively. Incubation was stopped by filtration through Whatman GF/B filters soaked with ice cold PBS. Bound radioactivity was measured by scintillation measurement as described previously48. Data analysis was performed by normalizing the radioactivity counts (ccpms) to values when 0% represents the non-stimulated receptor and 100% the maximum effect of morphine or DAMGO. Dose–response curves were calculated by nonlinear regression in GraphPad Prism 6.0. Mean values ± s.e.m. for EC and E values were derived from 3–12 individual experiments each done in triplicate. To measure μOR G -mediated cAMP inhibition, HEK-293T cells were co-transfected using calcium phosphate in a 1:1 ratio with human μOR and a split-luciferase based cAMP biosensor (pGloSensorTM-22F; Promega). For experiments including GRK2 co-expression, cells were transfected with 1 μg/15-cm dish of GRK2. After at least 24 h, transfected cells were washed with phosphate buffered saline (PBS) and trypsin was used to dissociate the cells. Cells were centrifuged, resuspended in plating media (1% dialysed FBS in DMEM), plated at a density of 15,000–20,000 cells per 40 μl per well in poly-lysine coated 384-well white clear bottom cell culture plates, and incubated at 37 °C with 5% CO overnight. For inactivation of pertussis-toxin (PTX) G experiments, cells were plated with 100 ng/ml final concentration PTX. The next day, drug dilutions were prepared in fresh assay buffer (20 mM HEPES, 1× HBSS, 0.1% bovine serum album (BSA), and 0.01% ascorbic acid, pH 7.4) at 3× drug concentration. Plates were decanted and 20 μl per well of drug buffer (20 mM HEPES, 1× HBSS, pH 7.4) was added to each well. Drug addition to 384-well plates was performed by FLIPR adding 10 μl of drug per well for a total volume of 30 μl. Plates were allowed to incubate for exactly 15 min in the dark at room temperature. To stimulate endogenous cAMP via β adrenergic-G activation, 10 μl of 4× isoproterenol (200 nM final concentration) diluted in drug buffer supplemented with GloSensor assay substrate was added per well. Cells were again incubated in the dark at room temperature for 15 min, and luminescence intensity was quantified using a Wallac TriLux microbeta (Perkin Elmer) luminescence counter. Data were normalized to DAMGO-induced cAMP inhibition and analysed using nonlinear regression in GraphPad Prism 6.0 (Graphpad Software Inc., San Diego, CA). Determination of functional activity of PZM21-29 for SAR studies was performed using a BRET-based cAMP accumulation assay50. HEK-293T cells were transiently co-transfected with pcDNA3L-His-CAMYEL42 (purchased from ATCC via LCG Standards, Wesel, Germany) and human μOR, achieving a cDNA ratio of 2:2 using Mirus TransIT-293 transfection reagent. 24 h post-transfection, cells were seeded into white half-area 96-well plates at 20 × 104 cells/well and grown overnight. On the following day, phenol-red-free medium was removed and replaced by PBS and cells were serum starved for 1 h before treatment. The assay was started by adding 10 μl coelenterazine h (Progmega, Mannheim, Germany) to each well to yield a final concentration of 5 μM. After 5 min incubation, compounds were added in PBS containing forskolin (final concentration 10 μM). Reads of the plates started 15 min after agonist addition. BRET readings were collected using a CLARIOstar plate reader (BMG LabTech, Ortenberg, Germany). Emission signals from Renilla Luciferase and YFP were measured simultaneously using a BRET1 filter set (475-30 nm/535-30 nm). BRET ratios (emission at 535-30 nm/emission at 475-30 nm) were calculated and dose–response curves were fitted by nonlinear regression using GraphPad Prism 6.0. Curves were normalized to basal BRET ratio obtained from dPBS and the maximum effect of morphine and DAMGO. Each curve is derived from three to five independent experiments each done in duplicate. Calcium release was measured using a FLIPRTETRA fluorescence imaging plate reader (Molecular Devices). Calcium release experiments were run in parallel to G Glosensor experiments with the same HEK-293T cells transfected with μOR, except cells for FLIPR were plated in poly-lysine coated 384-well black clear bottom cell culture plates. Cells were incubated at 37 °C with 5% CO overnight and next day media was decanted and replaced with Fluo-4 direct calcium dye (Life Technologies) made up in HBSS with 20 mM HEPES, pH 7.4. Dye was incubated for 1 h at 37 °C. Afterwards, cells were equilibrated to room temperature, and fluorescence in each well was read for the initial 10 s to establish a baseline. Afterwards,10 μl of drug (3×) was added per well and the maximum-fold increase in fluorescence was determined as fold-over-baseline. Drug solutions used for the FLIPR assay were exactly the same as used for G Glosensor experiments. To activate endogenous G -coupled receptors as a positive control for calcium release, TFLLR-NH (10 μM, PAR-1 selective agonist) was used. Internalization was measured using the eXpress DiscoveRx PathHunter GPCR internalization assay using split β-galactosidase complementation. In brief, cryopreserved U2OS cells expressing the human μOR were thawed rapidly and plated in supplied medium and 96-well culture plates. Next day, cells were stimulated with drugs (10×) and allowed to incubate for 90 min at 37 °C with 5% CO . Afterwards, substrate was added to cells and chemiluminescence was measured on a TriLux (Perkin Elmer) plate counter. Data were normalized to DAMGO and analysed using Graphpad Prism 6.0. β-Arrestin recruitment was measured by either the PathHunter enzyme complementation assay (DiscoveRx) or by previously described bioluminescence resonance energy transfer (BRET) methods51. Assays using DiscoveRx PathHunter eXpress OPRM1 CHO-K1 β-Arrestin GPCR Assays were conducted exactly as instructed by the manufacturer. Briefly, supplied cryopreserved cells were thawed and resuspended in the supplied medium, and plated in the furnished 96-well plates. Next day, 10× dilutions of agonist (prepared in HBSS and 20 mM HEPES, pH 7.4) were added to the cells and incubated for 90 min. Next, the detection reagents were reconstituted, mixed at the appropriate ratio, and added to the cells. After 60 min, luminescence per well was measured on a TriLux (Perkin-Elmer) plate counter. Data were normalized to DAMGO and analysed using the sigmoidal dose–response function built into GraphPad Prism 6.0. To measure μOR mediated β-arrestin recruitment by BRET in the presence or absence of GRK2 co-expression, HEK-293T cells were co-transfected in a 1:1:15 ratio with human μOR containing C-terminal renilla luciferase (RLuc8), GRK2, and venus-tagged N-terminal β-arrestin-2, respectively. In the case of experiments where GRK2 expression was varied, pcDNA3.1 was substituted for GRK2 to maintain the same concentration of DNA transfected. After at least 24 h, transfected cells were plated in poly-lysine coated 96-well white clear bottom cell culture plates in plating media at a density of 125,000–250,000 cells per 200 μl per well and incubated overnight. The next day, media was decanted and cells were washed twice with 60 μl of drug buffer and incubated at room temperature for at least 10 min before drug stimulation. 30 μl of drug (3×) was added per well and incubated for at least 30 min in the dark. Then, 10 μl of the RLuc substrate, coelenterazine H (Promega, 5 μM final concentration) was added per well, and plates were read for both luminescence at 485 nm and fluorescent eYFP emission at 530 nm for 1 s per well using a Mithras LB940 microplate reader. The ratio of eYFP/RLuc was calculated per well and the net BRET ratio was calculated by substracting the eYFP/RLuc per well from the eYFP/RLuc ratio without venus–arrestin present. Data were normalized to DAMGO-induced stimulation and analysed using nonlinear regression in GraphPad Prism 6.0. Multiple approaches have been described to quantitate ligand bias, including operational models, intrinsic relative activity models, and allosteric models31, 52. In the absence of GRK2, we observe no β-arrestin-2 recruitment for PZM21 and TRV130. This prevents a quantitative assessment of bias by the operational model. In the case where GRK2 is overexpressed, we observe arrestin recruitment for PZM21 and TRV130. In this case, we utilize the operational model to calculate ligand bias and display equiactive bias plots for comparison of ligand efficacy for distinct signalling pathways31, 53. The Glosensor G , DiscoverX PathHunter β-arrestin, or net BRET concentration response curves were fit to the Black-Leff operational model to determine transduction coefficients (τ/K ). Compound bias factors are expressed after normalization against the prototypical opioid agonist DAMGO used as a reference. Bias factors are expressed as the value of ΔΔlog[τ/K ]. To identify potential off-target activity of PZM21, we used the National Institutes of Mental Health Psychoactive Drug Screen Program. Compound PZM21 was first tested for activity against 320 non-olfactory GPCRs using the PRESTO-Tango GPCRome screening β-arrestin recruitment assay30. We used 10 μM PZM21 and activity at each receptor was measured in quadruplicate. Potential positive receptor hits were defined as those that increase the relative luminescence value twofold. Positive hits were subsequently re-tested in full dose–response mode to determine whether the luminescence signal titrates with increasing concentrations of PZM21. A number of false-positive hits were discounted by this approach. PZM21 inhibition of hERG channel was performed as described previously54 and neurotransmitter transporter assays were determined used the Molecular Devices Neurotransmitter Assay Kit (Molecular Devices). Adult male C57BL/6J (aged 3–5 months) obtained from Jackson Laboratories (Bar Harbour, Maine) were used to investigate behavioural responses, respiratory effects, and hyperlocomotion induced by PZM21 and compared with morphine or vehicle (0.9% sodium chloride). For μOR knockout animals, Oprm1−/− mice (B6.129S2-Oprm1tm1Kff/J) were obtained from Jackson Laboratories. All drugs were dissolved in vehicle and injected subcutaneously. Behavioural studies were conducted at the University of North Carolina and Stanford University following the National Institutes of Health’s guidelines for care and use of animals and with approved mouse protocols from the institutional animal care and use committees. Sample sizes (number of animals) were not predetermined by a statistical method and animals were assigned to groups randomly. Drug treatment groups were only blinded for measurement of affective versus reflexive analgesia; other experiments were not blinded to investigators. Predefined exclusion criteria were set for analgesia and conditioned preference experiments. No animals were excluded from statistical analysis. Statistical analyses were performed after first assessing the normality of distributions of data sets and Leven’s test was used to assess equality of variances. Analgesia-like responses in were measured as previously described55 using a hotplate analgesia meter with dimensions of 29.2 × 26.7 cm with mice restricted to a cylinder 8.9 cm in diameter and 15.2 cm high (IITC Life Sciences, Woodland Hills, California). Response was measured by recording the latency to lick, flutter, or splay hind paw(s), or an attempt to jump out of the apparatus at 55 °C, with a maximum cut-off time of 30 s. Once a response was observed or the cut-off time had elapsed, the subject was immediately removed from the hotplate and placed back in its home cage. The animals were acclimated to the hotplate, while cool, and a baseline analgesic response time was acquired several hours before drug treatment and testing. Mice were injected with either vehicle (n = 8), morphine (5 mg/kg, n = 8 or 10 mg/kg, n = 8), TRV130 (1.2 mg/kg, n = 9) or PZM21 (10 mg/kg, n = 8; 20 mg/kg, n = 11; or 40 mg/kg, n = 8). After injection of drug, the analgesic effect expressed as percentage maximum possible effect (%MPE) was measured at 15, 30, 60, 90 and 120 min after drug treatment. If animals did not display hind paw lick, splay, or flutter, they were removed from the trial. Additionally, if animals attempted to jump out of the plate or urinated on the hotplate they were removed from the trial. To assess analgesia by the tail-flick assay, a tail-flick analgesia meter (Columbus Instruments, Columbus, Ohio). Mice were gently immobilized with a cotton towel and the tail base was placed on a radiant light source emitting a constant temperature of 56 °C. The tail withdrawal latency was measured at similar time points as the hotplate assay after administration of vehicle (n = 8), morphine (5 mg/kg, n = 4; 10 mg/kg, n = 8) or PZM21 (10 mg/kg, n = 8; 20 mg/kg; n = 14). The cut-off time for the heat source was set at 10 s to avoid tissue damage. Analgesic response times were measured similar to the hotplate assay. Oprm1−/− and wild-type C57Bl/6J mice (male; 8–11 weeks) were acclimated to the testing environment and thermal-plate equipment for three non-consecutive days between 11:00 and 13:00 before any pharmacological studies. Acclimation was achieved by individually confining mice within an enclosed semi-transparent red plastic cylinder (10 cm depth × 15 cm height) on a raised metal-mesh rack (61 cm height) for 30 min, and then exposing each mouse to the thermal-plate equipment (non-heated; floor dimensions, 16.5 × 16.5 cm; Bioseb), while confined within a clear plastic chamber (16 cm length × 16 cm width × 30 cm height). Acclimation exposure to the thermal plate lasted for 30 s, and exposure was repeated after 30 min to mimic the test day conditions. The testing environment had an average ambient temperature of 22.6 °C and illumination of 309 lx from overhead fluorescence lighting. The same male experimenter (G.C.) was present throughout the entire duration of habituation and testing to exclude possible olfaction-induced alterations in sensory thresholds56. Cutaneous application of a noxious stimulus, or time spent on a hotplate apparatus can broadly elicit several distinct behavioural responses: 1) withdrawal reflexes: rapid reflexive retraction or digit splaying of the paw; 2) affective-motivational responses: directed licking and biting of the paw, and/or a motivational response characterized by jumping away from the heated floor plate. Paw withdrawal reflexes are classically measured in studies of hypersensitivity, and involve simple spinal cord and brainstem circuits57. In contrast, affective responses are complex, non-stereotyped behaviours requiring processing by limbic and cortical circuits in the brain, the appearance of which indicates the subject’s motivation and arousal to make the unpleasant sensation cease by licking the affected tissue, or seeking an escape route36, 57, 58, 59, 60, 61, 62, 63, 64. To distinguish between potential differential analgesic effects of PZM21, mice were placed on the heated apparatus (52.5 °C), and the latency to exhibition of the first sign of a hindpaw reflexive withdraw, and the first sign of an affective response was recorded. A maximum exposure cut-off of 30 s was set to reduce tissue damage. Mice were injected with either vehicle (n = 6), morphine (10 mg/kg, n = 10), or PZM21 (20 mg/kg, n = 13). After injection of drug, the analgesic effect on either reflex or attending responses was expressed as percentage maximum possible effect (%MPE), and was measured at −30 (baseline), 15, 30, 60, 90, 120, and 180 min relative to drug treatment. For studies comparing Oprm1−/− and wild-type C57Bl/6J mice, the analgesic response in the hotplate assay was measured 30 min after injection of vehicle (n = 5 for both genotypes), morphine (10 mg/kg, n = 5 for both genotypes) or PZM21 (20 mg/kg, n = 6 for Oprm1−/− and n = 5 for wild-type). Analgesia to formalin injection was carried out as described previously65. Mice were first habituated for 20 min to the testing environment which included a home cage without bedding, food, and water. After habituation, vehicle (n = 6), morphine (10 mg/kg, n = 7), or PZM21 (40 mg/kg, n = 7) was injected subcutaneously. This was followed by injection of 20 μl of 1% formalin in 0.9% saline under the skin of the dorsal surface of the right hindpaw. Animals were returned to their home cage and behavioural responses were recorded for one hour. Nociception was estimated by measuring the cumulative time spent by animals licking the formalin-injected paw. As opioids classically display two phases of analgesic action, nociceptive behaviour was measured during both the early phase (0 to 5 min) and the late phase (20 to 30 min). In Fig. 4, an asterisk indicates a significant difference between drug and vehicle (P < 0.05 calculated using a one-way ANOVA with Bonferroni correction). Respiration data was collected using a whole body plethysmography system (Buxco Electronics Inc., Wilmington, North Carolina) as described66. This method measures respiratory frequency, tidal volume, peak flows, inspiratory time, and expiratory time in conscious and unrestrained mice. Briefly, Buxco airflow transducers were attached to each plethysmography chamber and a constant flow rate was maintained for all chambers. Each chamber was calibrated to its attached transducer before the experiment. Animals were first habituated to the clear plexiglass chambers for 10 min. Respiratory parameters were recorded for 10 min to establish a baseline before injection of vehicle (n = 8), morphine (10 mg/kg, n = 8), TRV130 (1.2 mg/kg, n = 8) or PZM21 (40 mg/kg, n = 8). Respiratory parameters were then collected on unrestrained mice for 100 min post drug injection. To decrease respiratory variability induced by anxiety, mice were shielded from view of other animals and experimenter. In Fig. 4, an asterisk indicates a significant difference between drug and vehicle (P < 0.05 calculated using a repeated measures ANOVA with Bonferroni correction). To measure constipatory effects of morphine and PZM21, we assessed the total accumulated faecal boli as described6. Briefly, mice were injected with vehicle (n = 10), morphine (10 mg/kg, n = 16) or PZM21 (20 mg/kg, n = 16) and placed within a plexiglass chamber (5 cm × 8 cm × 8 cm) positioned on a mesh screen. Mice were maintained without food or water for 6 h. Faecal boli were collected underneath the mesh on a paper towel and the cumulative mass was measured every hour for six hours. In Fig. 4, an asterisk indicates a significant difference between drug and vehicle (P < 0.05 calculated using a repeated measures ANOVA with Bonferroni correction). A photocell-equipped automated open field chamber (40 cm × 40 cm × 30 cm; Versamax system, Accuscan Instruments) contained inside sound-attenuating boxes was used to assess locomotor activity. Baseline ambulation of freely moving mice was monitored over 30 min, followed by injection with vehicle (n = 7), morphine (10 mg/kg, n = 5) or PZM21 (20 mg/kg, n = 6). Locomotor activity was monitored for another 150 min. In Fig. 4, an asterisk indicates a significant difference between drug and vehicle (P < 0.05 calculated using a repeated measures ANOVA with Bonferroni correction). A three-chambered conditioned place preference apparatus (Med-Associates, St. Albans, Vermont) consisting of white or black chambers (16.8 × 12.7 × 12.7 cm each) with uniquely textured white mesh or black rod floors and separated by a neutral central chamber (7.2 × 12.7 × 12.7 cm) was used for conditioned place preference testing. On day 1 (preconditioning day), mice were placed in the central chamber and allowed to explore freely for 30 min. Time spent in each compartment was used to estimate baseline chamber preferences and mice showing specific chamber bias more than 70% were not studied further. On days 2–9 (conditioning days) mice were injected with either vehicle or drug and paired with either the white mesh or the black rod chambers. All mice received vehicle on days 2, 4, 6, 8 and drug on days 3, 5, 7, 9. On day 10 (test day), mice were again placed in the central chamber as on day 1 and allowed to explore freely for 30 min. Time spent in each chamber was expressed as percentage preference. Place preference was tested with morphine (10 mg/kg, n = 16), PZM21 (20 mg/kg, n = 8), or TRV130 (1.2 mg/kg, n = 7). In Fig. 4, an asterisk indicates a significant difference between vehicle and drug chambers (P < 0.05 by one-sample t-test with hypothetical value of 50) while NS indicates non-significance (P > 0.05). Drug induced catalepsy was measured in mice using the bar test67, which includes a horizontally placed 3-mm diameter wooden bar fixed 4 cm above the floor. Mice were habituated with the bar and the environment for 20 min before subcutaneous injection of either haloperidol (2 mg/kg, n = 8), morphine (10 mg/kg, n = 8), or PZM21 (20 mg/kg, n = 8). To measure catalepsy, both forepaws were gently placed on the bar and the length of time during which each mouse remained in the initial position was measured. The effect was measured at 15, 30 and 90 min after drug injection. Maximum cut-off time for each challenge was 90 s. Studies were performed by the Preclinical Therapeutics Core and the Drug Studies Unit at the University of California San Francisco. Ten mice were injected subcutaneously with 20 mg/kg of PZM21. At each time point, 1 ml of blood was collected from three mice and the serum concentration of PZM21 determined by liquid chromatography–mass spectrometry (LC/MS). Mice were subsequently sacrificed and entire brains were homogenized for determination of PZM21 concentrations by LC/MS. All studies were performed with approved mouse protocols from the institutional animal care and use committees. Metabolism experiments were performed as described previously68. In brief, pooled microsomes from male mouse liver (CD-1) were purchased (Sigma Aldrich) and stored at −75 °C until required. NADPH was purchased (Carl Roth) and stored at −8 °C. The incubation reactions were carried out in polyethylene caps (Eppendorf, 1.5 ml) at 37 °C. The incubation mixture contained PZM21 (80 μM) or positive controls (imipramine and rotigotine), pooled liver microsomes (0.5 mg of microsomal protein/ml of incubation mixture) and Tris-MgCl buffer (48 mM Tris, 4.8 mM MgCl , pH 7.4). The final incubation volume was 0.5 ml. Microsomal reactions were initiated by addition of 50 μl of enzyme cofactor solution NADPH (final concentration of 1 mM). At 0, 15, 30 and 60 min the enzymatic reactions were terminated by addition of 500 μl of ice-cold acetonitrile (containing 8 μM internal standard), and precipitated protein was removed by centrifugation (15,000 rcf for 3 min). The supernatant was analysed by HPLC/MS (binary solvent system, eluent acetonitrile in 0.1% aqueous formic acid, 10−40% acetonitrile in 8 min, 40−95% acetonitrile in 1 min, 95% acetonitrile for 1 min, flow rate of 0.3 ml/min). The experiments were repeated in three independent experiments. Parallel control incubations were conducted in the absence of cofactor solution to determine unspecific binding to matrix. Substrate remaining and metabolite formation was calculated as a mean value ± s.e.m. of three independent experiments by comparing AUC of metabolites and substrate after predetermined incubation time to AUC of substrate at time 0, estimating a similar ionization rate, corrected by a factor calculated from the AUC of internal standard at each time point. The stereochemically pure isomers of 12 and PZM21 were synthesized from corresponding (R)- and (S)-amino acid amides, which were either commercially available or readily prepared from the corresponding acid or ester (see Supplementary Information). The primary amino group was dimethylated using an excess of aqueous formaldehyde and sodium triacetoxyborohydride in aqueous acetonitrile. The carboxamides 16a,b were converted to primary amines by treatment with borane-tetrahydrofurane complex under reflux yielding the diamines 17a,b. Henry reaction of thiophene-3-carbaldehyde with nitroethane afforded the nitropropene derivative 18, which was converted into the racemic alkylamine 19. Activation with 4-nitrophenyl chloroformate yielded the carbamates 20, which were coupled with the enantiopure primary amines 17a,b to achieve diastereomeric mixtures of the corresponding ureas 12 and 21. HPLC separation using a semi-preparative Chiralpak AS-H column gave the overall eight pure stereoisomers of 12 and 21 including PZM21. To determine the absolute configuration of the final products and efficiently prepare PZM21, we synthesized enantiomerically enriched carbamate 20, coupled it with the corresponding primary amines. For enantiomeric enrichment, we performed chiral resolution of the racemic primary amine 19 via repetitive crystallization with di-p-anisoyl-(S)-tartaric acid. After triple crystallization, we obtained 19 enriched in dextrorotatory enantiomer ([α] 25 = +20.5°). The corresponding (R)-acetamide has been previously characterized as dextrorotatory ([α] 20 = +49.8°), so enantiomerically enriched 19 was treated with acetic anhydride and triethylamine, and the specific rotation of the product was measured. Based on the value of specific rotation of the resulting acetamide ([α] 21 = −46.6°), we assigned the absolute configuration of the major isomer to be (S). (S)-enriched 20 was used for synthesis of the final urea derivatives and absolute configuration of diastereomers in pairs was assigned based on the equality of retention time in chiral HPLC. A full description of the synthetic routes and analytical data of the compounds 12, PZM21 and its analogues PZM22-29 are presented in the Supplementary Information. PZM21 was docked to the inactive state μOR structure using DOCK3.6 (ref. 21) as described for the primary screen, with the exception that the 45 matching spheres used were generated based on the docked pose of compound 12. The resulting ligand-receptor complex was further optimized through minimization with the AMBER protein force field69 and the GAFF ligand force field supplemented with AM1-BCC charges. Docking of PZM21 and TRV130 to the active state μOR structure (PDB: 5C1M) was also performed with DOCK3.6 with parameters as described above. The amino terminus of the active state μOR, which forms a lid over the orthosteric binding site (residues Gly52–Met65) was removed before receptor preparation. Matching spheres were generated based on the pose of PZM21 in the inactive state. The resulting complexes were then minimized with AMBER. The pose of PZM21 in the active state μOR structure was further refined using Glide (Schrödinger) in XP mode. Molecular dynamics simulations were based on crystal structures of μOR in the inactive- and active-state conformation (PDB: 4DKL and 5CM1, respectively). In both cases, all non-receptor residues (T4 lysozyme in the inactive state and Nb39 in the active state) were removed. For the active state, amino-terminal residues were removed as in the docking studies. Initial coordinates of PZM21 were generated by molecular docking as described above. The receptor was simulated with two tautomers of His2976.52, either in the neutral Nδ or the Nε state. The μOR-PZM21 complex was embedded in a lipid bilayer consisting of dioleoylphosphatidylcholine (DOPC) molecules as described previously47. The charges of the inactive- and active-state simulation systems were neutralized by adding 11 and 14 chloride ions, respectively. To carry out MD simulations, the GROMACS package was used as described previously70. Briefly, the general AMBER force field (GAFF)71 was used for PZM21 and the lipids and the AMBER force field ff99SB72 for the receptor. Parameters for PZM21 were assigned using antechamber, and charges were calculated using Gaussian09 (Gaussian, Inc.) at the HF/6-31(d,p) level and the RESP procedure according to the literature73. During the simulations, PZM21 was protonated at its tertiary amine and simulated as a cation. The SPC/E water model74 was used, and the simulations were carried out at 310 K. Analysis of the trajectories was performed using GROMACS. Each simulation in a given condition was initiated from identical coordinates, but with initial atom velocities assigned independently and randomly. An overview of the simulation systems and their simulation times is shown in the Supplementary Information. Other than the in vivo studies, no statistical analysis was applied to in vitro or cell-based signalling assays. Sample size (number of assays for each compound or receptor) was predetermined to be in triplicate or quadruplicate for primary screening assays at a single concentration. For concentration–response assays, the sample size (number of assays for each compound at selected receptors) was also predetermined to be tested for a minimum of three assays, each in triplicate or quadruplicate. None of the functional assays were blinded to investigators.