DMD27615
Supplemental Information: Contrasting Influence of NADPH and a NADPH-regenerating System on the Metabolism of Carbonyl-Containing Compounds in Hepatic Microsomes.
Journal: Drug Metabolism and Disposition.
Christopher S. Mazur, John F. Kenneke, Michael-Rock Goldsmithand Cather Brown.
U.S. EPA, National Exposure Research Laboratory, Ecosystems Research Division,
960 College Station Rd., Athens, GA, USA 30605 (C.S.M., J.F.K.)
U.S. EPA, National Exposure Research Laboratory, Human Exposure and Atmospheric Sciences Division, Research Triangle Park, NC (M.R.G.)
Senior Service America, U.S. EPA, Athens, GA 30605 (C.B.)
Metabolism Profile Figure 1.
Saturation kinetic profile of triadimenol and cortisol formation as a function of NADPH concentration in rat hepatic microsomes.
Metabolism Profile Figure 2.
Statistical cofactor comparison of CYP-mediated propiconazole product formation in rat hepatic microsomes. NADPH-regenerating system (A); NADPH (B); G6P (C); G6P+NADP+ (D); G6P+NADPH (E). All systems contained MgCl2 with percent activity calculated relative to system A. Results are means +/- 95 % confidence interval for triplicate determinations
Computational Details
1. Ligand selection
The 11β-HSD1substrates (with variable rates) were identified in the literature (7-ketocholesterol, cortisone, menadione, 15-d-PGJ12, Oracin, ketoprofen). Using structural qualifiers of these known substrates for 11β-HSD1 we sought additional ligands that may be 11β-HSD1 substrates (contain substructure: SMILE: C(CC)(CC)=O, MW: 150-450, XLogP: 2-5, HBD:1-4, HBA: 1-3, ROT: 0-7, TPSA: 26-95, reference database EPA DSSTox) A total of 45 ligands were identified that satisfied these filters within DSSTox (
2. Qualifying a “substrate”
Upon docking, only chemicals that had a separation distance of less 3.9 Å between NADPH and carbonyl groups were considered substrates (See figure 1 – supporting information). This distance was determined by using cortisone docked distance as a metric. A total of 23 ligands of 45 filtered chemicals were identified, we list a subset of these including the natural ligand cortisone in the manuscript.
Supporting Information - FIGURE 1 : Protein - Ligand Interaction map of natural substrate (11β-HSD1) bound to active site and separation distance (3.95 Å) between target carbonyl (C=O) and NADPH (H-) annotated. Orientation obtained by docking cortisone into 2BEL crystal structure with AMBER99 force-field, London-dG score of -52.86 kJ/mol. See table for additional chemicals.
3. Docking Simulations Energetics and Pose Analysis: CYP-3A4 versus 11-β-hsd1
Comparing the pre-quilibrium process of triadimefon binding to (I) 11β-HSD1 (2IRW) or (II) CYP450-3A4 (2V0M) using either energetics basis obtained in MOE (kCal/mol) suggests ~0.6 kCal/mol (i.e. 2.5 kJ/mol) lower energy binding to the reductase system, and nominally more stability.
substrate / 11--HSD1 (2IRW) / CYP450-3A4 (2V0M)MOE (kCal/mol) / MOE (kCal/mol)
1S-Triadimefon / -5.3 / -4.7
1R-Triadimefon / -5.7 / -5.1
However, what is critical is that the lowest energy poses of triadimefon in 11β-HSD1 are catalytically competent, whereas CYP3A4 are not, despite being close in energy (~1 kJ) to a t-butyl presenting orientation. The lowest energy pose of triadimefon binds with ~2.5kJ/mol lower energy to 11β-HSD1 than CYP3A4, presents itself in a catalytically competent fashion amenable to nucleophilic attack, in which the reaction (carbonyl addition) has an intrinsically lower activation energy compared to the radical oxidation mechanism of CYP3A4. Furthermore, the cofactor hydride to carbonyl distance of triadimefon is 17% closer (2.56 Å) than that of the natural substrate cortisone (2.99 Å), whereas triadimefon bound to CYP3A4 has a t-butyl group presenting itself to the catalytic center at a distance of 9.7 Å (i.e. catalytically incompetent, shown on pages 3, 4 and 5 of supporting information).
Figures found on Supplemental Information - Page S-4 (triadimefon in CYP3A4), PageS-5 (triadimefon in Human 11β-HSD1), and Page S-6 (cortisone, triadimefon, NADPH overlayed in Human 11β-HSD1) were rendered in MOE, and 2D representations were rendered in Ligand Interaction Maps also in MOE 2008.10 (Molecular Operating Environment, Chemical Computing Group, Montreal Canada).
Supporting Information - FIGURE 2:Human CYP3A4 binding pocket (from 2V0M) with reaction center (Fe3+ in crimson) with triadimefon (ball and stick). A green line was drawn measuring 9.71 Å between the t-butyl hydrogen and the catalytic center, an exceedingly far distance for catalytic competency.
Supporting Information - FIGURE 3:The 2-D representation (Ligand interaction map) of the crystal structure 2V0M with bound Triadimefon with key contacts.
Supporting Information - FIGURE 4:Human 11β-HSD1 (2BEL) binding pocket with triadimefon (stick). The size of the pocket is snug and the distance from the triadimefon carbonyl to the NADPH reducing unit (green line) is 2.56 Å, apparently stabilized by H-bond interactions with SER170 and TYR183.
Supporting Information - FIGURE 5:The 2-D representation (Ligand interaction map) of the crystal structure 2BEL with bound Triadimefon with key contacts.
Supporting Information - FIGURE 6:Overlay of cortisone (yellow) and triadimefon (magenta) and NADPH (green) in Human 11β-HSD1 (2BEL) binding pocket. Note involvement of TYR183 and SER170 for carbonyl stabilization, apparently common to substrates studied here and potentially a critically requirement for delineating viable substrates from non-substrates, beyond the required presence of a carbonyl moiety.
Supporting Information - FIGURE 7:A 2D representation of cortisone and key contacts within active site of Human 11β-HSD1 (2BEL) binding pocket as shown in 3D in overlayed structure above.
Intraluminal Carbonyl Reduction in Hepatic Microsomes: Mazur, Kenneke, Goldsmith, Brown
Supplemental Information: Page S-1