PAT
Category: Postdoc Degree Level:PHD Abstract ID#469
Molecular Determinants Governing Activation versus Inactivation of Serotonin (5-HT)2C Receptors by 4-phenyl-2dimethylaminotetralin Analogs 1
Y. Liu1, TC. Cordova-Sintjago2, CE. Canal1 and RG. Booth 1, 2
Center for Drug Discovery, Northeastern University, Boston, MA; 2Department of Medicinal Chemistry, College of Pharmacy, University of Florida, Gainesville, FL
RESULTS
ABSTRACT The ligand-receptor interactions necessary for activating or inactivating 5-HT2C receptors, novel targets for neuropsychiatric disorders, remain unclear. Depending on substituents at the meta (m) position of the pendant phenyl ring of (-)-trans-(2S,4R)-4-phenyl-N,Ndimethyl-1,2,3,4-tetrahydronaphthalene-2-amine (PATs), PATs exhibit either 5-HT2C agonist, e.g. (-)-trans-m-Br-PAT, or inverse agonist, e.g. (-)-trans-m-CF3-PAT, properties. Molecular modeling and docking studies were performed to identify putative amino acid residues of the 5-HT2C receptor binding pocket that are involved in the function of PATs. Subsequently, site-directed mutagenesis of these residues was performed, and the pharmacology of (-)-trans-m-Br-PAT and (-)-trans-m-CF3-PAT at 5-HT2C was assessed. S3.36A, T3.37A, F5.47A and W6.48A point mutants significantly decreased affinity (Ki) of ()-trans-m-Br-PAT and (-)-trans-m-CF3-PAT (P < 0.05 relative to wild-type 5-HT2C (WT)). Functional potencies to activate or inactivate 5-HT2C-Gq-mediated phospholipase C signaling were also significantly decreased (P < 0.05). Functional potencies of both (-)trans-m-Br-PAT and (-)-trans-m-CF3-PAT (P < 0.05), but not affinity were decreased at C7.45A, C7.45S, S7.46A, S5.43A, F6.44A, and M6.47A. The function of (-)-trans-m-CF3PAT was inverted from inverse agonist (WT) to agonist at T3.37A, S5.43A, F5.47A, W6.48A, N6.55A, and the S5.43A plus N6.55A double mutant receptors. Conversely, the function of (-)-trans-m-Br-PAT was inverted from agonist (WT) to inverse agonist at the N6.55L point-mutated receptor. Collectively, affinity data suggest similar binding modes between (-)-trans-m-Br-PAT and (-)-trans-m-CF3-PAT at 5-HT2C, and that interactions between substituents in the meta position and N6.55 play a vital role in determining the function of meta-substituted PATs at 5-HT2C. Results from these studies provide insights into the molecular switches necessary for PATs to activate or inactivate 5-HT2C. These findings guide rationale design of novel PATs that possess therapeutic potential for treating neuropsychiatric disorders.
Figure 1. 5-HT2C in silico model (A)
(B)
Figure 3. 5-HT2C amino acid residues T3.37, F5.47, and N6.55 are involved in the binding of (-)-trans-m-Br-PAT and (-)-trans-m-CF3-PAT
(C)
(A) N 6.55
S 5.43 T 3.37 F 5.47 W 6.48
S 7.46 C 7.45
M 6.47 5HT2B(4IB4) 5HT2C_004
Helix 1 Helix 3 Helix 5 Helix 7
Helix 2 Helix 4 Helix 6
F 6.44
A. Superimposition of 5-HT2B crystal structure (PDB: 4IB4) (purple) and 5-HT2C model (cyan), RMSD=1.13 Å. B. Cartoon representation of 5-HT2C model with 7-transmembrane domain color-coded C. Close-up view of predicted ligand binding (in B) showing the potential interacting residues in the orthosteric binding site (as gray spheres). Figures were generated using PyMOL.
Test Ligand
WT 5-HT2C
T3.37A
S5.43A
F5.47A
N6.55A
N6.55L
S5.43A N6.55A
(-)-trans-m-Br-PAT
25.8 ±3.9
209.5 ±37.4(**)
28.4 ±6.4
274.9 ±46.3(**)
3.7 ±1.1(**)
11.5 ±2
6.6 ±0.9(**)
22.2 ±8.9
(-)-trans-m-CF3-PAT
58.0 ±8.3
320.9 ±125(*)
53.7 ±18
290.1 ±86.1(*)
5.6 ±0.4(**)
13.8 ±4.2
26.4 ±4.8
43.4 ±5.8
5-HT
73.0 ±4.3
56060 ±8040
546.6 ±178.7
4696.5 ±955
464.8 ±87.6
Figure 2. PAT ligand docking poses at 5-HT2C model (A) ∆G = - 8.55 kcal/mol
(B) ∆G = - 8.32 kcal/mol
= X=Br or CF3
=
C7.45A S7.46A
W6.48A
S3.36A
11.6 ±3.7
716.8
120.2
37.6 ±11.5
1093.6 ±590.8
234.9 ±90.8
172.4
2509 ±16
804.6 1764.2 238 2840 ±387.4 ±1051.8 ±38.6 ±1677.3
Binding affinities of (-)-trans-m-Br-PAT (A) and (-)-trans-m-CF3-PAT (B) at WT and mutant 5-HT2C receptors. Ki (nM) values are mean± SEM of 1-5 experiments. *, P≤0.05, **, P≤0.005 compared to WT.
MATERIALS & METHODS Competition Binding Assay: Radioligand competitive displacement binding assays were performed in 96-well plates in assay systems of 250 μl as previously described (1). Human embryonic kidney (HEK) 293 cells were transiently transfected with cDNA for human WT pointedmutated 5-HT2C receptors as described. Membranes expressing each receptor were prepared and used in binding assays that measured ability of test ligand to displace 5-HT2C radioligand. [3H]mesulergine. IP-One Functional Assay: The ability of the ligands to affect WT and the point-mutated 5-HT2C receptors-mediated phospholipase C activity by measuring the level of inositol phosphate (IP)1formation in the presence of different concentrations of ligand. Cells were treated with different concentration of the test ligands at 37°C for 2h. After incubation, the reaction was terminated by adding 5 μl of the donor and acceptor fluorescent conjugates in lysis buffer. The plate was incubated at room temperature for 1 hour to reach equilibrium. Inositol phosphate 1 levels were assessed by the ratio of fluorescence emission at 665 nM and 620 nM (fluorescence resonance energy transfer, HTRF) using a Synergy H1 plate reader with an HTRF filter cube (BioTek, Winooski, VT) (2). Mutagenesis: Point mutations of 5-HT2C receptors were made by polymerase chain reactions using QuickChange II Site-directed mutagenesis kit as previously described(2). Computational Studies: Homology model of the human WT 5-HT2C GPCR was calculated using the Robetta server (http: //robetta.bakerlab.org) with human WT 5-HT2B GPCR crystal structure (PDB code 4IB4) as the template. The lowest RMSD model was selected for geometry validation using PROCHECK. Prior to ligand docking, PRODRG (Dundee, UK) was used to generate (-)trans-m-Br-PAT and (-)-trans-m-CF3-PAT ligand coordinates. Finally, AUTODOCK v4.2 (Scripps, CA) was used to perform in silico docking of (-)-trans-m-Br-PAT and (-)-trans-m-CF3-PAT into 5HT2C model using Lamarckian genetic algorithm with grid dimension of 60, 60, 80. Docking results were analyzed and visualized using LigPlot+ (Roman Laskowski, 2009) and PyMol (Delano Scientific LLC). Structures of Ligands:
(B)
Figure 4. 5-HT2C amino acid residue N6.55 is critically involved in determining the function of (-)-trans-m-Br-PAT and (-)-trans-m-CF3-PAT (A)
(B)
Schematic diagrams of 5-HT2C receptor with (-)-trans-m-Br-PAT (A) and (-)-trans-m-CF3-PAT (B) interactions showing the lowest free energy calculated as well as the potential hydrophobic and hydrogen bonding. Figures were generated using LigPlot+.
CONCLUSIONS Test Ligand
WT 5HT2C
T3.37A
S5.43A
F5.47A
N6.55A N6.55L
S5.43A C7.45A N6.55A
S7.46A
W6.48A S3.36A
1. Several mutations decrease the binding affinities of PATs (S3.36A, T3.37A, F5.47A, W6.48A) suggest these amino acids are likely forming direct interactions 42.1 61.8 119 2354.2 9.1 133.4 13.7 1238 2595 894.1 615.2 ±18.7 ±18 (**) ±906(**) ±2.4(**) ±39 ±1.3(*) ±52(**) ±711(**) ±134 with the ligands or making inter-helical interactions that are important for ligand (-)-trans-m-Br-PAT ±7.8 binding. 16.9 2464 516 8198 31.9 283 62.6 271.6 542 4659 ND 2. N6.55 is a molecular switch that determines the function of (-)-trans-m-Br-PAT and (-)-trans-m-CF3-PAT ±4.2 ±631(**) ±91.6 (**) ±3412 ±8.9 ±81.6 ±8.3 ±38(**) ±157.5(**) (-)-trans-m-CF3-PAT at 5-HT2C. S5.43, W6.48, T3.37, F5.47 and C7.45 are also 0.66 68.7 1.57 108.6 35.3 18.9 32.8 1.0 81.1 1.17 9.2 5-HT important for maintaining (-)-trans-m-CF3-PAT function as an inverse agonist. ±0.2 ±0.1 ±1.0 ±48 ±8.4 ±9.6 ±8.3 ±0.5 ±18.7 ±6.2 3. The molecular pathways of 5-HT2C activation and inactivation by PATs involve Functional activities of (-)-trans-m-Br-PAT (A) and (-)-trans-m-CF -PAT (B) at WT and mutant 5-HT2C receptors. EC50 or IC50 (nM) values are 3 amino acids in the periphery of the orthosteric binding pocket including C7.45, mean± SEM of 1-5 experiments. *, P≤0.05, **, P≤0.005 compared to WT. S7.46, F6.44, M6.47. Mutating these residues to alanine significantly decreases potencies of both (-)-trans-m-Br-PAT and (-)-trans-m-CF3-PAT without affecting the REFERENCES binding affinities. 1. Booth, R.G., Fang, L.J., Huang, Y., Wilczynski, A., Sivendran, S. (2009) Eur J Pharmacol, 615:1-9 4. Combining the power of computational chemistry and molecular pharmacology, 2. Canal, C.E., Cordova-Sintjago, T.C., Liu, Y., Kim, M. S., Morgan, D., Booth, R.G.(2013) J we identified the molecular determinants of 5-HT2C agonism vs inverse agonism Pharmacol Exp Ther, 347(3):705-16 of (-)-trans-m-Br-PAT and (-)-trans-m-CF3-PAT, respectively. The molecular mechanisms of pharmacological function of PAT compounds at 5-HT2C guide Funding Source (RG Booth, PI): National Institutes of Health Grants RO1-DA030989, RO1-DA023928, RO1-MH081193 rational design of novel 5-HT2C agonists to treat neuropsychiatric disorders. Patent protection (composition of matter and use): WO2010/129048, WO2009/061436, WO2008/156707, WO2008/154044.
Molecular Determinants Governing Activation versus Inactivation of Serotonin (5-HT)2C Receptors by 4-phenyl-2dimethylaminotetralin Analogs 1
Y. Liu1, TC. Cordova-Sintjago2, CE. Canal1 and RG. Booth 1, 2
Center for Drug Discovery, Northeastern University, Boston, MA; 2Department of Medicinal Chemistry, College of Pharmacy, University of Florida, Gainesville, FL
RESULTS
ABSTRACT The ligand-receptor interactions necessary for activating or inactivating 5-HT2C receptors, novel targets for neuropsychiatric disorders, remain unclear. Depending on substituents at the meta (m) position of the pendant phenyl ring of (-)-trans-(2S,4R)-4-phenyl-N,Ndimethyl-1,2,3,4-tetrahydronaphthalene-2-amine (PATs), PATs exhibit either 5-HT2C agonist, e.g. (-)-trans-m-Br-PAT, or inverse agonist, e.g. (-)-trans-m-CF3-PAT, properties. Molecular modeling and docking studies were performed to identify putative amino acid residues of the 5-HT2C receptor binding pocket that are involved in the function of PATs. Subsequently, site-directed mutagenesis of these residues was performed, and the pharmacology of (-)-trans-m-Br-PAT and (-)-trans-m-CF3-PAT at 5-HT2C was assessed. S3.36A, T3.37A, F5.47A and W6.48A point mutants significantly decreased affinity (Ki) of ()-trans-m-Br-PAT and (-)-trans-m-CF3-PAT (P < 0.05 relative to wild-type 5-HT2C (WT)). Functional potencies to activate or inactivate 5-HT2C-Gq-mediated phospholipase C signaling were also significantly decreased (P < 0.05). Functional potencies of both (-)trans-m-Br-PAT and (-)-trans-m-CF3-PAT (P < 0.05), but not affinity were decreased at C7.45A, C7.45S, S7.46A, S5.43A, F6.44A, and M6.47A. The function of (-)-trans-m-CF3PAT was inverted from inverse agonist (WT) to agonist at T3.37A, S5.43A, F5.47A, W6.48A, N6.55A, and the S5.43A plus N6.55A double mutant receptors. Conversely, the function of (-)-trans-m-Br-PAT was inverted from agonist (WT) to inverse agonist at the N6.55L point-mutated receptor. Collectively, affinity data suggest similar binding modes between (-)-trans-m-Br-PAT and (-)-trans-m-CF3-PAT at 5-HT2C, and that interactions between substituents in the meta position and N6.55 play a vital role in determining the function of meta-substituted PATs at 5-HT2C. Results from these studies provide insights into the molecular switches necessary for PATs to activate or inactivate 5-HT2C. These findings guide rationale design of novel PATs that possess therapeutic potential for treating neuropsychiatric disorders.
Figure 1. 5-HT2C in silico model (A)
(B)
Figure 3. 5-HT2C amino acid residues T3.37, F5.47, and N6.55 are involved in the binding of (-)-trans-m-Br-PAT and (-)-trans-m-CF3-PAT
(C)
(A) N 6.55
S 5.43 T 3.37 F 5.47 W 6.48
S 7.46 C 7.45
M 6.47 5HT2B(4IB4) 5HT2C_004
Helix 1 Helix 3 Helix 5 Helix 7
Helix 2 Helix 4 Helix 6
F 6.44
A. Superimposition of 5-HT2B crystal structure (PDB: 4IB4) (purple) and 5-HT2C model (cyan), RMSD=1.13 Å. B. Cartoon representation of 5-HT2C model with 7-transmembrane domain color-coded C. Close-up view of predicted ligand binding (in B) showing the potential interacting residues in the orthosteric binding site (as gray spheres). Figures were generated using PyMOL.
Test Ligand
WT 5-HT2C
T3.37A
S5.43A
F5.47A
N6.55A
N6.55L
S5.43A N6.55A
(-)-trans-m-Br-PAT
25.8 ±3.9
209.5 ±37.4(**)
28.4 ±6.4
274.9 ±46.3(**)
3.7 ±1.1(**)
11.5 ±2
6.6 ±0.9(**)
22.2 ±8.9
(-)-trans-m-CF3-PAT
58.0 ±8.3
320.9 ±125(*)
53.7 ±18
290.1 ±86.1(*)
5.6 ±0.4(**)
13.8 ±4.2
26.4 ±4.8
43.4 ±5.8
5-HT
73.0 ±4.3
56060 ±8040
546.6 ±178.7
4696.5 ±955
464.8 ±87.6
Figure 2. PAT ligand docking poses at 5-HT2C model (A) ∆G = - 8.55 kcal/mol
(B) ∆G = - 8.32 kcal/mol
= X=Br or CF3
=
C7.45A S7.46A
W6.48A
S3.36A
11.6 ±3.7
716.8
120.2
37.6 ±11.5
1093.6 ±590.8
234.9 ±90.8
172.4
2509 ±16
804.6 1764.2 238 2840 ±387.4 ±1051.8 ±38.6 ±1677.3
Binding affinities of (-)-trans-m-Br-PAT (A) and (-)-trans-m-CF3-PAT (B) at WT and mutant 5-HT2C receptors. Ki (nM) values are mean± SEM of 1-5 experiments. *, P≤0.05, **, P≤0.005 compared to WT.
MATERIALS & METHODS Competition Binding Assay: Radioligand competitive displacement binding assays were performed in 96-well plates in assay systems of 250 μl as previously described (1). Human embryonic kidney (HEK) 293 cells were transiently transfected with cDNA for human WT pointedmutated 5-HT2C receptors as described. Membranes expressing each receptor were prepared and used in binding assays that measured ability of test ligand to displace 5-HT2C radioligand. [3H]mesulergine. IP-One Functional Assay: The ability of the ligands to affect WT and the point-mutated 5-HT2C receptors-mediated phospholipase C activity by measuring the level of inositol phosphate (IP)1formation in the presence of different concentrations of ligand. Cells were treated with different concentration of the test ligands at 37°C for 2h. After incubation, the reaction was terminated by adding 5 μl of the donor and acceptor fluorescent conjugates in lysis buffer. The plate was incubated at room temperature for 1 hour to reach equilibrium. Inositol phosphate 1 levels were assessed by the ratio of fluorescence emission at 665 nM and 620 nM (fluorescence resonance energy transfer, HTRF) using a Synergy H1 plate reader with an HTRF filter cube (BioTek, Winooski, VT) (2). Mutagenesis: Point mutations of 5-HT2C receptors were made by polymerase chain reactions using QuickChange II Site-directed mutagenesis kit as previously described(2). Computational Studies: Homology model of the human WT 5-HT2C GPCR was calculated using the Robetta server (http: //robetta.bakerlab.org) with human WT 5-HT2B GPCR crystal structure (PDB code 4IB4) as the template. The lowest RMSD model was selected for geometry validation using PROCHECK. Prior to ligand docking, PRODRG (Dundee, UK) was used to generate (-)trans-m-Br-PAT and (-)-trans-m-CF3-PAT ligand coordinates. Finally, AUTODOCK v4.2 (Scripps, CA) was used to perform in silico docking of (-)-trans-m-Br-PAT and (-)-trans-m-CF3-PAT into 5HT2C model using Lamarckian genetic algorithm with grid dimension of 60, 60, 80. Docking results were analyzed and visualized using LigPlot+ (Roman Laskowski, 2009) and PyMol (Delano Scientific LLC). Structures of Ligands:
(B)
Figure 4. 5-HT2C amino acid residue N6.55 is critically involved in determining the function of (-)-trans-m-Br-PAT and (-)-trans-m-CF3-PAT (A)
(B)
Schematic diagrams of 5-HT2C receptor with (-)-trans-m-Br-PAT (A) and (-)-trans-m-CF3-PAT (B) interactions showing the lowest free energy calculated as well as the potential hydrophobic and hydrogen bonding. Figures were generated using LigPlot+.
CONCLUSIONS Test Ligand
WT 5HT2C
T3.37A
S5.43A
F5.47A
N6.55A N6.55L
S5.43A C7.45A N6.55A
S7.46A
W6.48A S3.36A
1. Several mutations decrease the binding affinities of PATs (S3.36A, T3.37A, F5.47A, W6.48A) suggest these amino acids are likely forming direct interactions 42.1 61.8 119 2354.2 9.1 133.4 13.7 1238 2595 894.1 615.2 ±18.7 ±18 (**) ±906(**) ±2.4(**) ±39 ±1.3(*) ±52(**) ±711(**) ±134 with the ligands or making inter-helical interactions that are important for ligand (-)-trans-m-Br-PAT ±7.8 binding. 16.9 2464 516 8198 31.9 283 62.6 271.6 542 4659 ND 2. N6.55 is a molecular switch that determines the function of (-)-trans-m-Br-PAT and (-)-trans-m-CF3-PAT ±4.2 ±631(**) ±91.6 (**) ±3412 ±8.9 ±81.6 ±8.3 ±38(**) ±157.5(**) (-)-trans-m-CF3-PAT at 5-HT2C. S5.43, W6.48, T3.37, F5.47 and C7.45 are also 0.66 68.7 1.57 108.6 35.3 18.9 32.8 1.0 81.1 1.17 9.2 5-HT important for maintaining (-)-trans-m-CF3-PAT function as an inverse agonist. ±0.2 ±0.1 ±1.0 ±48 ±8.4 ±9.6 ±8.3 ±0.5 ±18.7 ±6.2 3. The molecular pathways of 5-HT2C activation and inactivation by PATs involve Functional activities of (-)-trans-m-Br-PAT (A) and (-)-trans-m-CF -PAT (B) at WT and mutant 5-HT2C receptors. EC50 or IC50 (nM) values are 3 amino acids in the periphery of the orthosteric binding pocket including C7.45, mean± SEM of 1-5 experiments. *, P≤0.05, **, P≤0.005 compared to WT. S7.46, F6.44, M6.47. Mutating these residues to alanine significantly decreases potencies of both (-)-trans-m-Br-PAT and (-)-trans-m-CF3-PAT without affecting the REFERENCES binding affinities. 1. Booth, R.G., Fang, L.J., Huang, Y., Wilczynski, A., Sivendran, S. (2009) Eur J Pharmacol, 615:1-9 4. Combining the power of computational chemistry and molecular pharmacology, 2. Canal, C.E., Cordova-Sintjago, T.C., Liu, Y., Kim, M. S., Morgan, D., Booth, R.G.(2013) J we identified the molecular determinants of 5-HT2C agonism vs inverse agonism Pharmacol Exp Ther, 347(3):705-16 of (-)-trans-m-Br-PAT and (-)-trans-m-CF3-PAT, respectively. The molecular mechanisms of pharmacological function of PAT compounds at 5-HT2C guide Funding Source (RG Booth, PI): National Institutes of Health Grants RO1-DA030989, RO1-DA023928, RO1-MH081193 rational design of novel 5-HT2C agonists to treat neuropsychiatric disorders. Patent protection (composition of matter and use): WO2010/129048, WO2009/061436, WO2008/156707, WO2008/154044.
