Figure 1.
Location in CYP2D6*1 of the amino acid variants (the structure is shown in two views).
*34 has a single mutation at R296C (purple) on helix I and distal to the active site. *17-2 has the R296 mutation as well as T107I (blue) while *17-3 also has the S486T mutation (green), but distal to the active site. *53 has two mutations in SRS1 at F120I and A122S (orange) that are near the active site. Heme is shown in red.
Table 1.
Allelic Variants of CYP2D6 Analyzed in this Study∧.
Figure 2.
The structure of SCH 66712 consists of phenyl, imidazole, piperazine, and fluorinated heteroaromatic rings.
SCH 66712 is metabolized by CYP2D6 to four mono-oxygenation products. One product is formed by oxygenation of the phenyl ring (most likely on the para-carbon) and three products are formed by oxygenation at sites on the piperazine and/or heteroaromatic rings, as previously described [21].
Figure 3.
Changes in root mean square fluctuations (ΔRMSF) of backbone atoms of *34 (purple), *17-2 (blue), *17-3 (green), and *53 (orange) relative to *1 (corresponds to zero line) with (A) no ligand bound and (B) with SCH 66712 bound.
Positive values for ΔRMSF correlate to atoms of increased flexibility while negative values for ΔRMSF correlate with more rigid atoms compared to reference CYP2D6*1. The x-axis indicates amino acid position. Helices are indicated in blue (and labeled at the top of Panel A), β-sheets in green, and turns in pink. The F-G, C-D, and G-H loop regions show the greatest variability in flexibility between variants and *1 both with and without SCH 66712 bound. There was also a modest difference in flexibility in the meander loop region.
Figure 4.
Interaction of SCH 66712 with allelic variants.
(A) Initial docking poses of SCH 66712 with each CYP2D6 variant. SCH 66712 was docked using AutoDock Vina as described in the Methods. Other poses not shown include positioning of the heteroaromatic ring above the heme iron. Both positions, phenyl or heteroaromatic ring pointing toward heme, are consistent with known sites of metabolism by *1 [21]. (B) Representative snapshots of the most prominent SCH 66712 binding mode for each variant as determined by PCA from 1000 snapshots (PCA population kernels shown in Figure S12). Additional, minor, binding modes were also identified for *1 and *17-3 as indicated in Figure S13. Helix I is shown in the background. For each variant, the phenyl ring of SCH 66712 points toward the heme. In *1, *17-2, and *17-3 the heteroaromatic ring of SCH 66712 is pointing to a pocket between helices I and G. In *53 and *34, SCH 66712 is oriented with the heteroaromatic ring pointing to a pocket between helices F and E.
Figure 5.
Structural fluctuations of each CYP2D6 variant with SCH 66712 bound.
Rainbow color scheme indicates degree of fluctuation with blue indicating little fluctuation to red indicating large fluctuation. The size of the backbone strand is also indicative of fluctuation with large diameter indicative of fluctuations. All structures showed a rigid core surrounding the heme with the area of the greatest flexibility in the F-G loop and helix A′ regions on the distal side. *34 was the most flexible of the variants and *53 was the most rigid.
Figure 6.
Root square mean deviation of SCH 66712 within active site of each CYP2D6 variant shows system convergence (each variant is colored as indicated in
Figure 3). *53 and *1 reach equilibrium within 20 ns, *17-2 and *17-3 within 50 ns, and *34 does not converge after 100 ns.
Table 2.
Total free energy of SCH 66712 binding estimated by MM-PBSA.
Figure 7.
Major tunnels identified in CYP2D6 throughout the MD simulations all depicted in one frame and shown from two views.
Tunnels are shown using van der Waals representations. Pathways were determined from 1000 snapshots from the molecular dynamics simulations using the PyMOL plugin CAVER 3.0 and are depicted together on frame 500 of *1. The channels shown are 2c (blue), 2e (green), 2b (red), and Solvent (turquoise). Channel names are given using the nomenclature of Cojocaru et al. [46]. Heme is shown in red sticks. For clarity, other channels identified are not shown. Channel 2c exits between helix I and the B-C loop; channel 2e exits between the B-C and B-B′ loops; channel 2b exits between the B-B′ loop and β-1/β-3 sheets; and the solvent channel exits between helices I and F and sheet β-4.
Figure 8.
Time evolution of the bottleneck radius of channels 2b, 2c, 2e, and solvent in each CYP2D6 variant during simulations.
For analysis, 1000 snapshots of each variant were used. The snapshots were taken each 100 ps over a 100 ns simulation time. The color map ranges from 0.9 Å (red) to 1.5 Å or greater (green) bottlenecks. Grey indicates that no tunnel with bottleneck radius ≥0.9 Å was identified for the given pathway cluster. Channel 2b was the most open channel for the variants, but not *1. Furthermore, over the course of the simulation, 2b became more open for the variants. Channel 2c was the major channel for *1, but was not open frequently or wide for any of the variants. Channel 2e was most open in *34. The solvent channel was a major open channel for the variants, particularly *17-3, but was only open a few times for *1 over the course of the 100 ns simulation. CAVER 3.0 was used for channel identification as described in the methods.