AnimalsAnimal experiments were performed in accordance with the protocols approved by the Laboratory Animal Committee of Kissei Pharmaceutical Co., Ltd. or the Guidelines of the Institutional Animal Care and Use Committee of RIKEN, Yokohama Branch (2018-075). BALB/cA mice (CLEA Japan and Jackson Laboratory Japan), BALB/cCr mice (Japan SLC), C57BL/6JJcl mice (CLEA Japan), C.B-17 SCID mice (Jackson Laboratory Japan), Sprague-Dawley (SD) rats (Japan SLC and Jackson Laboratory Japan), Dark Agouti (DA) rats (Japan SLC) and Hartley guinea pigs (Japan SLC) were maintained in standard controlled conditions with a 12-h lighting cycle and access to chow and water ad libitum. Cynomolgus monkeys (Del Mundo Trading and SICONBREC Inc.) were maintained in standard controlled conditions with a 12-h lighting cycle and access to chow from 9:00 to 13:00 and water ad libitum, and experiments were performed by Ina Research Inc. according to protocols approved by the Institutional Animal Care and Use Committee. In total, seven male monkeys were used for the study.ChemistryKSI-6666, (R)-KSI-6666, (S)-KSI-6666, compounds 1, 2, 3, 4, and 5, the Merck S1PR1 agonist, W146, and KRP-203 were synthesized by Kissei Pharmaceutical Co., Ltd. Detailed synthetic schemes and synthetic procedures are presented in the Supplementary Methods in Supplementary Information file.Compound-induced reduction in blood leukocyte and lymphocyte count in ratsCompounds dissolved in the vehicle, which consisted of 5–10% dimethyl sulfoxide (Wako) and 5–10% Wellsolve (Celeste) mixed in distilled water (Otsuka), were orally administered to 6–16-week-old male SD rats. Blood samples were chronologically collected in EDTA-2K-containing blood collection tubes (BD) under anesthesia before and after compound administration. Leukocytes or lymphocytes in blood were counted by an automated hematology analyzer KX-21NV (Sysmex) or ProCyte Dx (IDEXX). In some experiments, plasma fractions were isolated and concentrations of compounds were measured by liquid chromatography-mass spectrometry (Q Exactive, Thermo Fisher Scientific). Detailed measurement procedures are presented in the Supplementary Methods in the Supplementary Information file.Compound-induced reduction in blood lymphocyte count in miceCompounds were dissolved in 10 mM Na2CO3 and 40% (2-hydroxypropyl)-β-cyclodextrin (Sigma-Aldrich) and intraperitoneally administrated to 8-week-old female BALB/cA mice (10 mg/kg). At 1, 2, and 4 h after the administration, whole blood was collected into tubes pretreated with EDTA-2K (Dojin). After the lysis of erythrocytes by using RBC Lysis Buffer, cell suspensions were stained with anti-mouse CD19 (BD 557399), CD4-PECy7 (BioLegend 100528), and CD8-APCCy7 (BioLegend 100714) antibodies after blocking by anti-CD16/32 antibodies (BioLegend 101302). Dead cells were excluded via 7-aminoactinomycin D (Wako) staining. Flow cytometric analysis was performed by using Canto II (BD), and data were analyzed by Flowjo software (BD).Compound-induced reduction in blood lymphocyte count in cynomolgus monkeysThree-to-five-year-old male cynomolgus monkeys were used. The test compound suspended in a vehicle consisting of 0.5% methylcellulose 400 solution (Wako) and 0.01 M hydrochloric acid (Wako) mixed in distilled water was orally administered to monkeys. Blood samples were chronologically collected in EDTA-2K-containing blood collection tubes while the monkey was awake, before and after compound administration. Lymphocytes in blood were counted using the hematology analyzer ADVIA120 (Siemens).Pharmacokinetic study in ratsCompounds dissolved in the vehicle, which consisted of 10% dimethyl sulfoxide and 10% Wellsolve mixed in saline (Otsuka), were intravenously administered to 6-week-old male SD rats. Blood samples were chronologically collected from awake rats after compound administration and heparinized. Plasma fractions were isolated and concentrations of compounds were measured by liquid chromatography-mass spectrometry (Q Exactive, Thermo Fisher Scientific). Detailed measurement procedures are presented in the Supplementary Methods in the Supplementary Information file. The pharmacokinetic parameters were calculated by non-compartmental analysis using WinNonlin 6.1 software (Pharsight).GTP binding assayMembranes were prepared from HEK293 cells stably transfected with human S1PR1. GTP binding assay was performed using a GTP Gi Binding Assay Kit (Cisbio) according to the manufacturer’s protocol. In the antagonist assay, serial dilutions of the test compound and 10 nM Merck S1PR1 agonist were co-incubated with 1.25 μg/well of cell membranes suspended in the stimulation buffer supplied by the manufacturer containing 50 mM MgCl2 and 2 µM GDP, followed by the addition of GTP Eu Cryptate (donor) and d2 antibody (acceptor). Membranes were incubated overnight at room temperature and the fluorescence signals from the donor (620 nm) and the acceptor (665 nm) were read using a PHERAstar FSX microplate reader (BMG Labtech). The fluorescence ratio (signal 665 nm/signal 629 nm × 104) was calculated. The half-maximal inhibitory concentration (IC50) values were calculated by nonlinear regression analysis of the concentration-response curve in GraphPad Prism 6 software. In assays to evaluate the effect of the test compound alone, membranes were treated with serial dilutions of the compound and were subjected to the same operations as noted above. In functional reversibility assays, serial dilutions of the test compound and serial dilutions of FTY720-P (Cayman) were co-incubated with membranes followed by the same operations as noted above. The fluorescence ratio obtained from the response to 10 µM FTY720-P alone was calculated as a 100% response (Emax = 100%), and that in the condition without FTY720-P or the test compound was defined as 0%. The maximum value of the response to FTY720-P in co-treatment with the test compound (calculated by nonlinear regression analysis of the concentration-response curve in GraphPad Prism 6) was regarded as the Emax value.Receptor internalization assayHEK293 cells (RIKEN BRC) transiently transfected with pBiT3.1-N in which the coding sequence of human S1PR1 was inserted (pBiT3.1-N-WT-S1PR1), resulting in the expression of S1PR1 with an HiBiT-tag at the N-terminus, were prepared using Lipofectamine 2000 (Thermo Fisher Scientific). In the functional reversibility assay, serial dilutions of the test compound were preincubated with cells in a culture medium consisting of Dulbecco’s Modified Eagle’s Medium (Wako) supplemented with 10% fetal bovine serum (Biological Industries) for 10 min at 37 °C, followed by the induction of receptor internalization by the addition of serial dilutions of FTY720-P or S1P (Cayman). Following 60-min incubation at 37 °C, cell surface levels of HiBiT-tagged human S1PR1 were monitored using a Nano Glo HiBiT Extracellular Detection System (Promega) according to the manufacturer’s protocol. The surface level of receptors, represented by the luminescence, was read by using a PHERAstar FSX microplate reader. The luminesce induced by 10 µM FTY720-P or 100 µM S1P alone was calculated as 100% response (Emax = 100%), and that in the condition without agonist or the test compound was defined as 0%. The maximum value of the response to FTY720-P or S1P following pretreatment of cells with the test compound (calculated by nonlinear regression analysis of the concentration-response curve in GraphPad Prism 6) was regarded as the Emax value. In functional reversibility assays with cells expressing mutant S1PR1, pBiT3.1-N with insertion of the mutated human S1PR1 sequence where the original codon for Met124 (ATG) was replaced by GTG or CTG, leading to the expression of M124V (pBiT3.1-N-M124V-S1PR1) or M124V (pBiT3.1-N-M124L-S1PR1), respectively, was transfected into cells. In assays to evaluate the effect of the test compound alone, cells in culture medium were treated with the compound for 60 min at 37 °C and the surface S1PR1 levels were evaluated as noted above. In assays to evaluate the antagonistic effect of the test compound, cells in the culture medium were treated with serial dilutions of the compound for 10 min at 37 °C, followed by the addition of 10 nM FTY720-P.Ca2+ mobilization assayCHO-K1 cells (ECACC) transiently transfected with human S1PR2 or S1PR3, or cells transiently co-transfected with human S1PR1, S1PR4, S1PR5, mouse S1PR1, or rat S1PR1 and human Gα15, were prepared using FuGENE HD (Promega). Ca2+ mobilization assay using a Fluo4 Direct Calcium Assay Kit (Thermo Fisher Scientific) was carried out according to the manufacturer’s protocol. Following the pretreatment of Fluo4-loaded cells with serial dilutions of the test compound in an assay buffer consisting of Hanks’ Balanced Salt Solution (Thermo Fisher Scientific) and 20 mM HEPES (Thermo Fisher Scientific) supplemented with 0.1% bovine serum albumin (Sigma-Aldrich) and 2.5 mM probenecid (Thermo Fisher Scientific) for 15 min, Ca2+ mobilization was induced by 10 nM Merck S1PR1 agonist for cells expressing S1PR1, or by 100 nM S1P for human S1PR2-, S1PR3-, S1PR4-, or S1PR5-transfected cells. The increase in cellular calcium level was evaluated using an FDSS 7000 plate reader (Hamamatsu Photonics) and the maximum fluorescent signal in each well was calculated. IC50 values were calculated by nonlinear regression analysis of the concentration-response curve in GraphPad Prism 6 software. In the evaluation of the dissociation half-life of the test compounds from S1PR1, the compound was incubated with cells expressing human S1PR1 and Gα15 in the culture medium consisting of Ham’s F-12 Nutrient Mixture (Thermo Fisher Scientific) supplemented with 10% fetal bovine serum for 60 min at 37 °C, followed by the extensive wash with phosphate-buffered saline (Wako). Following incubation for 1–24 h in a culture medium at 37 °C, a Ca2+ mobilization assay was carried out as noted above. Ca2+ mobilization was induced by a 10 nM Merck S1PR1 agonist. The ratio of the maximum fluorescent signal to the minimum after the addition of agonist in each well was calculated. The ratio obtained from the response to 10 nM Merck S1PR1 agonist was calculated as 100% and that in the condition without the agonist or the test compound was defined as 0% to evaluate the response by receptors spared from compound binding after washout. The dissociation half-life was calculated by nonlinear regression analysis (one phase exponential decay) of the incubation time–response curve in GraphPad Prism 6 software.cAMP assayHEK293 cells transiently transfected with pBiT3.1-N-WT-S1PR1 or pBiT3.1-N-M124V-S1PR1 were prepared using Lipofectamine 2000. Cells were pretreated with serial dilutions of the test compound in assay buffer consisting of Hanks’ Balanced Salt Solution, 20 mM HEPES, and 0.5 mM 3-isobutyl-1-methylxanthine (Wako) for 10 min, followed by the addition of serial dilutions of FTY720-P. Following incubation for 10 min, 5 µM forskolin (Wako) was added for 20 min. Cellular cAMP levels were examined using a LANCE Ultra cAMP Kit (PerkinElmer) according to the manufacturer’s protocol, and the time-resolved fluorescence resonance energy transfer (TR-FRET) emission (665 nm) was read by using a PHERAstar FSX microplate reader. All procedures were performed at room temperature. The TR-FRET fluorescence obtained from the response to 1 µM FTY720-P alone was calculated as a 100% response (Emax = 100%) and that in the condition without FTY720-P or the test compound was defined as 0%. The highest value of the response to FTY720-P in cells co-treated with the test compound (calculated by nonlinear regression analysis of the concentration-response curve in GraphPad Prism 6) was regarded as the Emax value.Effect on heart rate in guinea pigsSix-to-seven-week-old male Hartley guinea pigs anesthetized by intraperitoneal injection of 35 mg/kg pentobarbital sodium (Kyoritsu Seiyaku) and 2.5 mg/kg of butorphanol tartrate (Meiji Seika Pharma) were subjected to the insertion of an endotracheal tube via a tracheotomy. The thorax was incised, and the right common carotid artery was catheterized with a PE90 catheter (BD) filled with heparinized saline (50 IU/mL) connected to a blood pressure amplifier (Nihon Kohden) and heart rate counter (Nihon Kohden) via a DTXPlus Transducer (Merit Medical). Signals were monitored and recorded by using a PowerLab 8/35 instrument (AD Instruments) and LabChart software (AD Instruments). KSI-6666 (30 mg/kg) dissolved in dimethyl sulfoxide was intravenously administered to the animals, followed by an intravenous injection of 0.3 mg/kg fingolimod hydrochloride (LC Laboratories) dissolved in dimethyl sulfoxide 15 min after the administration of KSI-6666. The heart rate was recorded continuously during the study. The average of the heart rate counted for 1 min just before the administration of KSI-6666 (-15 min), just before and every 10 min after the administration of fingolimod (0, 10, 20, and 30 min) was calculated and used for data analysis.Chemotaxis assayTwelve-week-old female BALB/cA mice were sacrificed, and spleens were collected. Splenocytes were prepared after the removal of red blood cells using RBC Lysis Buffer (BioLegend). Cells (2 × 107 cells/mL) in assay medium consisting of phenol red-free RPMI-1640 (Thermo Fisher Scientific) supplemented with 0.1% fatty acid-free bovine serum albumin (Sigma-Aldrich) were incubated with serial dilutions of the test compound for 5 min, and 100 µL cell suspensions were transferred to Transwell inserts (Corning) loaded in 24-well culture plates (BD) filled with 600 µL assay medium per well and 20 nM S1P. Cells were incubated for 3 h at 37 °C in a 5% CO2 incubator, followed by removal of the inserts. Migrated cells in plates were stained with CyQuant (Thermo Fisher Scientific) for 1 h at 37 °C in the CO2 incubator and fluorescence intensity, correlated with cell number, was measured by using a PHERAstar FSX microplate reader.Naïve CD4+CD45RBhi T cell adoptive transfer colitis modelEight-to-nine-week-old female BALB/cA mice were sacrificed, and spleens were collected. Splenocytes were prepared after the removal of red blood cells using RBC Lysis Buffer. Naïve T cells were isolated by negative selection using a Mouse Naive T Cell CD4+/CD62L+/CD44low Column Kit (R&D Systems). Naïve CD4+ T cells were treated with fluorescein isothiocyanate-labeled (FITC) anti-mouse CD45RB antibody (BioLegend 103305), followed by washing. Cells treated with anti-FITC MicroBeads (Miltenyi Biotec) were magnetically separated via a MACS column (Miltenyi Biotec) and CD4+CD45RBhigh T cells were thus purified. Cell suspension (6 × 105 cell/mL) was intraperitoneally transferred into 8–10-week-old female C.B-17 SCID mice (500 µL/mouse). Two weeks after the transfer, oral administration of compounds suspended in the vehicle consisting of a 1:9 mixture of 0.5% methylcellulose 400 solution and distilled water twice daily was started therapeutically and lasted for 14–16 days. An equimolar amount of hydrochloric acid was mixed only in the suspension of KSI-6666. The day after the last administration, the DAI score, consisting of the sum (0–12 points) of the grades for three observations [weight loss (%; none: 0; 0–5: 1; 5–10: 2; 10–20: 3; >20: 4); stool consistency (normal: 0; loose stool: 2; diarrhea: 4); and hematochezia (absence: 0; presence: 4)] was evaluated. Mice were sacrificed and colons were collected, weighed, and further evaluations were performed. The histopathological assessment was performed using a modification of the previously described method50. The following scoring system, consisting of the sum (0–30 points) of the grades for several observations [decrease in goblet cells (0, 2, 4, 6); inflammatory cell infiltration (0, 2, 4, 6); mucosal thickening (0, 2, 4, 6); crypt score based on abnormal crypt architecture including distortion, branching, atrophy, and crypt loss (0, 3, 6, 9); and mucosal erosion and ulceration (0, 3)], were assessed by a veterinary pathologist in no double-blinded fashion. For mRNA expression analysis, total RNA was extracted from colons by using an RNeasy Mini Kit (QIAGEN). cDNA was reverse transcribed by using a PrimeScript RT Reagent Kit Perfect Real Time (TaKaRa Bio Inc.). Real-time PCR was performed using a 7500 Fast Real-Time PCR System (Applied Biosystems). SYBR Premix Ex Taq Perfect Real Time (TaKaRa Bio Inc.) for S100a9 with a primer set (MA058882) obtained from TaKaRa Bio Inc. and Premix Ex Taq Perfect Real Time (TaKaRa Bio Inc.) for 18S rRNA with TaqMan Ribosomal RNA Control Reagents (Applied Biosystems) were used for the reaction. The ΔΔCt method was applied to calculate the relative expression level of the target mRNA using the 18S rRNA gene as the reference. % increase in body weight was calculated by the body weight ratio between the day compound administration started and the end of the study.Oxazolone-induced colitis modelA total of 150 µL of 3% 4-ethoxymethylene-2-phenyl-2-oxazolin-5-one (oxazolone; Sigma-Aldrich) dissolved in 100% ethanol (Wako) was applied on shaved back skins of 9–10-week-old male BALB/cCr mice for sensitization. One week after sensitization, 100 µL of 1% oxazolone dissolved in 50% ethanol oxazolone was intrarectally administered to the mice. On the same day, test compounds suspended in the vehicle consisting of a 1:9 mixture of 0.5% methylcellulose 400 solution and distilled water were orally administered to mice twice (before and after sensitization). Mice were sacrificed 24 h after sensitization, and colons were collected and weighed. Homogenates of colons were prepared using TisseLyser (QIAGEN) and the amount of interleukin (IL)-4 protein was evaluated using a Mouse IL-4 Quantikine ELISA Kit (R&D Systems). Change in body weight was calculated by the difference between the day of intrarectal oxazolone administration and the end of the study.Experimental autoimmune encephalomyelitis (EAE) in ratsThe immunization emulsion consisted of syngeneic CNS antigen (one part brain to two parts of spinal cord) in phosphate-buffered saline (Thermo Fisher Scientific) emulsified 1:1 in complete Freund’s adjuvant (Thermo Fisher Scientific) was intradermally injected into two different sites on the dorsal base of the tail root of the 9-week-old female DA rat under anesthesia at 100 µL/site. On the next day of the immunization, oral administration of compounds suspended in the vehicle consisting of a 1:9 mixture of 0.5% methylcellulose 400 solution and distilled water twice daily was started prophylactically and lasted for 12 days. An equimolar amount of hydrochloric acid was mixed only in the suspension of KSI-6666. Rats were daily scored by the following criteria [0: normal appearance; 1: limp tail; 2: unilateral partial paralysis of hind limbs or front limbs; 3: bilateral complete paralysis of hind limbs or front limbs; 4: quadriplegia; 5: death] during the study period started from the day of immunization (day 0) and lasted until day 13. The area under curve (AUC) of the score over time was calculated by GraphPad Prism 6.Experimental autoimmune encephalomyelitis (EAE) in miceEAE was induced in 9-week-old female C57BL/6JJcl mice using an EAE induction kit (Hooke Laboratories) according to the manufacturer’s protocol. Briefly, anesthetized mice were immunized by the subcutaneous injection of myelin oligodendrocyte glycoprotein peptide (MOG35–55) at two different sites of the back (100 µL/site), followed by the intraperitoneal injection of 100 ng pertussis toxin (day 0). The pertussis toxin injection was repeated on the following day (day 1). Mice were daily weighed and scored by the following criteria [0: normal appearance; 0.5: weak tail; 1: limp tail; 1.5: limp tail + weak hind limbs; 2: unilateral partial paralysis of hind limbs; 2.5: unilateral complete paralysis of hind limbs or bilateral partial paralysis of hind limbs; 3: bilateral complete paralysis of hind limbs; 3.5: bilateral complete paralysis of hind limbs + hind limbs clustered on one side of the body; 4: bilateral complete paralysis of hind limbs + partial front limb paralysis; 5: moribundity or death] on day 2, day 4 and daily basis from day 6. The day after the first observation of clinical score 1 or above in each animal, oral administration of compounds dissolved in the vehicle, which consisted of 5% dimethyl sulfoxide and 5% ethanol mixed in PEG300 (Hampton Research), once daily was started therapeutically and lasted for 14 days. The final assessment was performed on the day after the final administration. The clinical score and body weight of the animal resulting in the death was recorded as 5 and the weight before death, respectively, for the remaining days of the experiment. Animals weighing less than 17 g on day 2 were excluded from the study.Molecular docking analysisOur approach to docking the (R)-KSI-6666 molecule to the potential binding sites in S1PR1 used two steps: (1) structure and ligand preparation for docking, and (2) docking of the (R)-KSI-6666 molecule. An initial structure of human S1PR1 was modeled from the X-ray structure of S1PR1 with the W146 molecule excluding the Escherichia virus T4 structure (PDB: 3V2Y)35 and refined for docking simulations using the Protein Preparation Wizard Script within Maestro51 (Schrödinger LLC, New York, NY, USA). For the (R)-KSI-6666 molecule, ionization and energy minimization were performed by the OPLS3 force field in the LigPrep Script in Maestro. A minimized structure of (R)-KSI-6666 was employed as the input structure for docking simulation. To account for both ligand and receptor flexibility in the first step, the Glide ‘Induced Fit Docking (IFD)’ protocol52 (Schrödinger LLC) was used, followed by iteratively combining rigid receptor docking using Glide and protein remodeling by side chain searching and minimization using Prime (Schrödinger LLC) techniques. We generated 20 initial orientations of (R)-KSI-6666 in a grid box defined by the center of the co-crystalized ligand (W146) using Glide docking [standard precision (SP) mode]. Next, we applied the soften-potential docking options, which involved scaling the van der Waals radii by 0.5 for receptor and (R)-KSI-6666 atoms. In the protein remodeling stage, all residues within a 5.0-Å radius of each initial docked (R)-KSI-6666 were refined using Prime. (R)-KSI-6666 was then redocked into the refined receptor structure using Glide in SP mode. Finally, the best pose for (R)-KSI-6666 was rescored according to GlideScore (SP mode). The Docking score of co-crystalized ligand W146 was calculated using the score-in-place procedure of Glide SP mode.Conventional molecular dynamics (cMD) and metadynamics (MetaD) simulationTwo types of MD simulations were performed. One is a cMD simulation to evaluate the stability of the docking pose and the other is a MetaD simulation to evaluate the ligand-binding persistency. Binding models of (R)-KSI-6666 and W146 to the structure of human S1PR1 were placed in a large palmitoyl-oleoyl-phosphatidylcholine (POPC) bilayer, and transferable intermolecular potential 3 point (TIP3P) water molecules were solvated with 0.15 M NaCl. After minimization and relaxation of the model, the productions of cMD and MetaD phases were performed for five independent 500-ns and ten independent 100-ns simulations, respectively, using Desmond version 2.3 (Schrödinger LLC) in the isothermal–isobaric (NPT) ensemble at 300 K and 1 bar using a Langevin thermostat. The OPLS3e force field53,54 was used for the simulations. The long-range electrostatic interactions were computed using the Smooth Particle Mesh Ewald method. All system setups were performed using Maestro. The MM-GBSA protocol (Schrödinger LLC) was used to calculate the binding free energy of ligands based on cMD trajectories, and the last 100 ns of 500-ns trajectory each was extracted and summed for the analysis. MetaD simulation is a widely used enhanced sampling method that allows the sampling of free-energy landscapes. In this simulation, we defined the biasing collective variables as the distance between the center of mass of the ligands and the ligand-binding residues (Tyr29, Lys34, Tyr98, Asn101, Ser105, Gly106, Thr109, Trp117, Arg120, Glu121, Met124, Phe125, Leu128, Ser192, Val194, Leu195, Pro196, Tyr198, Ile203, Cys206, Thr207, Phe210, Trp269, Leu272, Phe273, Leu276, Glu294, Leu297, Val298, Ala300 and Val301 for (R)-KSI-6666; Tyr29, Lys34, Tyr98, Asn101, Ser105, Gly106, Thr109, Trp117, Arg120, Glu121, Met124, Phe125, Leu128, Thr193, Val194, Leu195, Pro196, Ile203, Cys206, Thr207, Phe210, Trp269, Leu272, Phe273, Leu276, Glu294, Leu297, Ala300, and Val301 for W146). The initial Gaussian hill height was set at 0.3. The MetaD procedure for (R)-KSI-6666 in the M124V variant was identical to that for wild-type S1PR1. The model of M124V-S1PR1 was constructed using Maestro.Protein–ligand interaction fingerprint analysisPLIF analyses from MetaD trajectories were performed using the PLIF modules of MOE 2020.0901 (Chemical Computing Group Inc., Montreal, Quebec, Canada). Energy-based contacts, including side chain hydrogen bonds (donor or acceptor), backbone hydrogen bonds (donor or acceptor), solvent hydrogen bonds (donor or acceptor), ionic interactions, metal binding interactions, and π interactions, and surface contact fingerprints, were employed. After the PLIFs had been generated based on 1000 snapshot structures (every 1 ns in ten independent 100-ns simulations) for each ligand, the probability of each bit occurring for each residue for (R)-KSI-6666 compared with W146 was calculated as Qb.$${Q}_{b}=\frac{{C}_{b}}{{C}_{0}}\cdot 2\left({P}_{b}-\frac{1}{2}\right)$$Cb is the number of snapshots within the set that have the fingerprint bit set, and C0 is the total number of snapshots. Pb is the relative probability of the fingerprint bit entry over the snapshots for (R)-KSI-6666 compared with that for W146 if it contains the bit. Qb takes values from −1 to 1. A higher positive value indicates a higher probability of the bit occurring for (R)-KSI-6666.Quantification and statistical analysisGraphs are presented as mean ± s.e.m, as indicated in the figure legends. Three or more independent experiments were performed in each in vitro study. Three or more animals/groups were used in each in vivo study. Statistical analyses were performed by F-test, Student’s t-test (two-tailed), Welch’s t-test (two-tailed), one-way ANOVA, two-way ANOVA, Dunnett’s test, and Tukey’s test. p values and sample sizes can be found in the figure legends. No randomization or blinding was used.Reporting summaryFurther information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
