*Corresponding Author *K. F.: *T. H.:

Transcription

1 Theoretical Analysis of Activity Cliffs among Benzofuranone Class Pim1 Inhibitors Using the Fragment Molecular Orbital Method with Molecular Mechanics Poisson-Boltzmann Surface Area (FMO+MM-PBSA) Approach Chiduru Watanabe,, Hirofumi Watanabe, Kaori Fukuzawa,*,, Lorien J. Parker,, # Yoshio Okiyama, Hitomi Yuki, Shigeyuki Yokoyama, Hirofumi Nakano, Shigenori Tanaka, and Teruki Honma*, RIKEN Center for Life Science Technologies, Suehiro-cho, Tsurumi-ku, Yokohama , Japan Institute of Industrial Science, The University of Tokyo, Komaba, Meguro-ku, Tokyo , Japan Department of Physical Chemistry, School of Pharmacy and Pharmaceutical Sciences, Hoshi University, Ebara, Shinagawa, Tokyo , Japan RIKEN Structural Biology Laboratory, Suehiro-cho, Tsurumi-ku, Yokohama , Japan # Department of Structural Biology, St. Vincent's Institute, 9 Princes St, Fitzroy, Victoria 3065, Australia Drug Discovery Initiative, The University of Tokyo, Hongo, Bunkyo-ku, Tokyo , Japan Graduate School of System Informatics, Kobe University, 1-1 Rokkodai, Nada-ku, Kobe , Japan *Corresponding Author *K. F.: k-fukuzawa@hoshi.ac.jp. *T. H.: honma.teruki@riken.jp 1

2 A. Initial hit of the benzofuranone class inhibitor Structure formulas of the initial hit compound (IC 50 = 410 nm) of the benzofuranone class inhibitor and Cpd 1 (IC 50 = 2 nm) 1 are illustrated in Figure S1. Figure S1. Structure formulas of the initial hit and Cpd 1 of the benzofuranone class Pim1 inhibitors. 2

3 B. The CH- interaction analysis The CH- interactions were determined with the CHPI program. 2 6 The criteria for typical and weak CH- interactions, which are indicated by purple and orange lines in Figure 1, are described in Figure S2. Typical CH- interactions Weak CH- interactions D max 3.05 Å 3.30 Å max º º º º Figure S2. Method for exploring XH/π contacts. A six-membered aromatic ring is shown as an illustrative example. (A) O: center of the π-plane. A 1 and A 2 : nearest and second-nearest sp 2 -atoms, respectively, to the hydrogen atom H. : dihedral angle defined by the A 1 OA 2 and HA 1 A 2 planes. : X-H-I angle. D pln : perpendicular distance between H and the π-plane (H/I). D atm : HA 1 distance. D lin : distance between H and the line A 1 -A 2 (H/J). (B) Regions to be searched. Region 1: zone where H is above the ring. Regions 2 and 3: zones where H is outside of region 1 but may interact with the π-ring. Unless otherwise noted, the program was run to search for short H/π contacts with the following conditions: D pln < D max (region 1); D lin < D max (region 2); D atm < D max (region 3); max < < max ; < max. D hpi : H/ distance (D pln for region 1, D lin for region 2, D atm for region 3). 3

4 C. Structure preparation The X-ray crystal structures of Pim1 and inhibitor complexes are shown in Figure S3 and its data collection and refinement statistics (PDB-ID: 5VUC, 5VUA, and 5VUB) are listed in Table S1. The hydrogen bond network between Pim1 and ligand is shown in Figures 1B and 3. The hydrogen bond lengths obtained from X-ray, MM-opt, and QM/MM-opt are summarized in Table S2. Stable structures of ligands in complex and with solvation are shown in Figure S4. We confirmed that the indole/azaindole rings were preferred to the flip configuration in the solvent by QM calculation at the HF/6-31G* level with the PCM model. The entropy of ligands in the case of stable structure with water is listed in Table S3. Figure S3. Protein surface of Pim1 (A) and X-ray crystal structures of Pim1 and inhibitor complexes (B). Carbon atoms colored gray, yellow, green, and orange represent Cpds 1, 3, 5, and 6 (PDB-ID: 5VUC, 3UMW, 5VUA, and 5VUB), respectively. 4

5 Table S1. Data collection and refinement statistics. Compound PDB code 5VUC 5VUA 5VUB Wavelength Resolution range (Å) ( ) ( ) ( ) Space group P6 5 P6 5 P6 5 Unit cell 98.3, 98.3, , 98.1, , 98.1, , 90, , 90, , 90, 120 Total reflections (5913) (4476) (5718) Unique reflections (2987) (2236) (2973) Multiplicity 2.0 (2.0) 2.0 (2.0) 1.9 (1.9) Completeness (%) 100 (100) 100 (100) 100 (100) Mean I/sigma(I) 18.9 (5.00) 17.3 (5.1) 18.4 (4.7) Wilson B-factor R-meas (0.23) (0.236) (0.296) Reflections used in refinement (2987) (2238) (2974) Reflections used for R-free 1509 (169) 1122 (128) 1482 (145) R-work 17.5 (21.0) 16.9 (20.8) 17.5 (20.8) R-free 19.9 (24.3) 19.8 (25.4) 19.7 (24.7) Number of non-hydrogen atoms macromolecules ligands Protein residues RMS(bonds) RMS(angles) Ramachandran favored (%) Ramachandran outliers (%) Average B-factor macromolecules ligands solvent *Statistics for the highest-resolution shell are shown in parentheses. 5

6 Table S2. Bond lengths of hydrogen bond network between ligand and amino acid residues for X-ray, MM-opt, and QM/MM-opt structures, respectively. Hydrogen bond distance between ligand and amino acid residues, r 1, r 2, r 3 and r 4, are shown in Figure 3. r 1 between Fragment (1) and Glu121 (Å) r 2 between Fragment (1) and Lys67 (Å) Cpd IC 50 (nm) X-ray MM-opt QM/MM-opt X-ray MM-opt QM/MM-opt r 3 between Fragment (2) and Asp128 (Å) r 4 between Fragment (2) and Glu171 (Å) Cpd IC 50 (nm) X-ray MM-opt QM/MM-opt X-ray MM-opt QM/MM-opt

7 Figure S4. Stable structures of ligands in complex (A) and with solvation (B). In the most stable structures in water, the indole/azaindole ring is flipped. Table S3. The entropy of the six compounds (kcal/mol) calculated by the MM method using the most stable structures in water (the indole/azaindole ring is flipped). Entropy based on the MM method at T = K (kcal/mol) Cpd Max-Min Translational Rotational Vibrational Total

8 D. MM-PBSA results bind The MM-based binding free energies, G MM, between Pim1 and each inhibitor in the general MM-PBSA scheme are listed in Table S4 for X-ray, MM-opt, and QM/MM-opt results, where vdw def sol,,, and are the non-bonding ele E MM E MM E MM G MM-PBSA electrostatic interaction, van der Waals interaction, ligand deformation, and solvation energies between protein and a ligand, respectively. Table S4. The energies (kcal/mol) obtained with the MM-PBSA calculation of Pim1 and each inhibitor complex for three types of complex (X-ray, MM-opt, and QM/MM-opt). Structure Cpd IC 50 (nm) ele vdw MM-PBSA calculation def sol bind G MM-PBSA G MM X-ray Structure Cpd IC 50 (nm) R with pic R 2 with pic ele vdw def sol bind G MM-PBSA G MM MM-Opt Structure Cpd IC 50 (nm) R with pic R 2 with pic ele vdw def sol bind G MM-PBSA G MM QM/MM-Opt R with pic R 2 with pic

9 E. NPA charge analysis The natural population analysis (NPA) of the Pim1 and inhibitor complex obtained by QM/MM-opt structures was performed at the HF/6-31G* level. Figure S5 shows the NPA charges of Cpds 1 6. Figure S5. NPA charge analysis of complex structures for Cpds 1 6. Red color indicates positive charge, and blue color means negative charge. Sum of NPA charge in the indole/azaindole ring outlined by the pink dotted oval is indicated by q indole. 9

10 F. PIEDA For detailed analysis at the residue level, the ES and DI components between ligand Fragment (1) and surrounding fragments of amino acid residues are shown in Tables S5 and S6, respectively. Values of amino acid residue fragments within a distance of 4.5 Å from a ligand were selected. Table S5 Electrostatic (ES) contribution of IFIE between Fragment (1) of ligands and selected amino acid residues evaluated by PIEDA. Fragment residue Formal fragment charge a Distance E ES of Fragment (1) (kcal/mol) b Average c max-min d R Lys67 S Arg122 M Ile104 S Phe49 S Arg122 S Leu120 S Leu44 S Val52 S Val126 S Asp186 M Ser46 M Gly45 M Asp186 S Leu174 S Pro123 M Ala65 S Ile185 S Leu120 M Gly47 M Phe187 M Ala65 M Glu89 S Ile185 M Glu121 M Val52 M

11 Ile104 M Leu44 M Glu121 S e Summation of the selected residues Summation of all residues a Distance between atoms of ligand and one of each fragment. b Average of E ES among Cpds 1 and 6. c Difference between maximum and minimum values of E ES among Cpds 1 and 6. d Correlation between E ES and pic 50. e Residues within 4.5Å from atoms of ligand fragment (1) to its nearest neighbor atoms. Table S6. Dispersion interaction (DI) contribution of the IFIE between Fragment (1) of ligands and selected amino acid residues evaluated by the PIEDA. Fragment residue Formal fragment charge a Distance E DI of Fragment (1) (kcal/mol) b Average c max-min d R Lys67 S Arg122 M Ile104 S Phe49 S Arg122 S Leu120 S Leu44 S Val52 S Val126 S Asp186 M Ser46 M Gly45 M Asp186 S Leu174 S Pro123 M Ala65 S Ile185 S Leu120 M

12 Gly47 M Phe187 M Ala65 M Glu89 S Ile185 M Glu121 M Val52 M Ile104 M Leu44 M Glu121 S e Summation of the selected residues Summation of all residues a Distance between atoms of ligand and one of each fragment. b Average of E DI among Cpds 1 and 6. c Difference between maximum and minimum values of E DI among Cpds 1 and 6. d Correlation between E DI and pic 50. e Residues within 4.5Å from atoms of ligand fragment (1) to its nearest neighbor atoms. 12

13 G. FMO+MM-PBSA binding energy with QM-based ligand deformation energy. We compared the FMO+MM-PBSA energy values (eq 7) with those by the following eq 13 including QM-based ligand deformation energy. bind int def solv G E E G. (7) FMO FMO MM MM PBSA G bind FMO E int FMO E def MO G solv MM PBSA. (13) The recalculated FMO+MM-PBSA energy values using eq 13 are listed in Table S7. bind The correlation between pic 50 and the recalculated G FMO (R 2 (QM/MM opt): 0.67) was not improved in comparison to that based on eq 7 (R 2 (QM/MM opt): 0.85) in this dataset. Table S7. The FMO+MM-PBSA energy values (kcal/mol) calculated by eq 13 including QM-based ligand deformation energy using three types of complex structures (X-ray, MM-opt, and QM/MM-opt). Structure Cpd IC 50 (nm) E int FMO E def MO G solv MM-PBSA G bind FMO X-ray R with -pic R 2 with -pic Structure Cpd IC 50 (nm) E int FMO E def MO G solv MM-PBSA G bind FMO MM-opt R with -pic R 2 with -pic Structure Cpd IC 50 (nm) E int FMO E def MO G solv MM-PBSA G bind FMO QM/MM-opt

14 R with -pic R 2 with -pic

15 References 1. Nakano, H.; Saito, N.; Parker, L. J.; Tada, Y.; Abe, M.; Tsuganezawa, K., Yokoyama, S.; Tanaka, A.; Kojima, H.; Okabe, T.; Nagano, T. Rational Evolution of a Novel Type of Potent and Selective Proviral Integration Site in Moloney Murine Leukemia Virus Kinase 1 (PIM1) Inhibitor from a Screening-Hit Compound. J. Med. Chem. 2012, 55, Umezawa, Y.; Nishio, M. CH/ Interactions as Demonstrated in the Crystal structure of Guanine-nucleotide Binding Proteins, Src homology-2 Domains and Human Growth Hormone in Complex with their Specific Ligands. Bioorg. Med. Chem. 1998, 6, Umezawa, Y.; Nishio, M. CH/ Interactions in the Crystal Structure of Class I MHC Antigens and Their Complexes with Peptides. Bioorg. Med. Chem. 1998, 6, Umezawa, Y.; Nishio, M. CH/ Interactions in the Crystal Structure of TATA-box Binding Protein/DNA Complexes. Bioorg. Med. Chem. 2000, 8, Umezawa, Y.; Nishio, M. CH/ Hydrogen Bonds as Evidenced in the Substrate Specificity of Acetylcholine Esterase. Bioorg. Biopolymers. 2005, 79, Nishio, M. The CH/ Hydrogen Bond in Chemistry. Conformation, Supramolecules, Optical Resolution and Interactions Involving Carbohydrates. Phys. Chem. Chem. Phys. 2011, 13,