Team:Patras/Model
Modeling
The purpose of visualization is insight, not pictures!
- All
- Introduction
- CYP2C19
- CYP2D6
- Conclusion
In this section, the dry lab calculations will be described. To begin with, molecular docking was performed with AutoDock 4.2. However, in order to prepare the ligands and the receptors before performing the molecular docking, more processing was done by other programs.
Ligand Preparation
Ligand preparation consists of two main actions, creating the structure and secondly minimizing the energy. For both creation and minimization of our ligands (Mephenytoin and Dextromethorphan), we used ChemDraw 3D. To minimize the energy, we performed MMFF94 minimization. After that, the ligand is ready to be docked.
Receptor Preparation
Ligand preparation consists of two main actions, creating the structure and secondly minimizing the energy. For both creation and minimization of our ligands (Mephenytoin and Dextromethorphan), we used ChemDraw 3D. To minimize the energy, we performed MMFF94 minimization. After that, the ligand is ready to be docked.
Docking
As we already mentioned, molecular docking was performed by AutoDock while Pymol2 was used as a visualizer for the 3D imaging of the homologous models.
CYP2C19
The wild type was docked to Mephenytoin in order to construct a standard model. That way, we could compare the interactions between the 11 variations and the docked wild type so that we could export a prediction of the variants’ activity.
Between the variants’ sequences and the sequences registered in PDB, there were some differences on account of the starting and ending amino acids. Be that as it may, some amino acids were not the same. This may happen because of the crystallization requirements. Either way, those changes made no difference between two of our variants (L15P and R26* stop gained) concerning the amino acid sequence and, therefore, their 3D structure. As a result, they were not used further as homologous models
The wild-type protein was screened with Mephenytoin, and the docked region is shown below.
From the literature, we found that there are important residues that interact with the ligand. Some of those residues are, Val113, Ile205, Ala297, Thr301, Leu366, and Phe476, with the last two of them being essential. Those residues form a hydrophobic binding cavity in which the Mephenytoin inserts to interact with the heme that is found just on the bottom of this region. The cavity’s function is to transfer the ligand closer to heme in order to be oxidized. The interaction between those residues and Mephenytoin was altered in most of the variants’ 3D structures.
Change in the 274 glutamic acid residue to aspartic led the drug closer to Phe476 but increased the distance to Ala297 and Leu366, which is essential to the interaction. In this homologous model, that change led to the decrease of the amino acid sequence by two amino acids.
Next, conversion of Asp286 to Asp286 led to no significant difference in the interaction. However, the change of a polar uncharged amino acid to a charged one close to the heme region may interfere with its function.
In the mutation D293G, the ligand’s conformation with the least energy was the one that is shown above. In this variant, the functional groups have totally changed their orientation, making Mephenytoin unable to interact properly with the important residues.
The variant T304A showed no significant differences in their structural conformation compared to the CYP2C19 wild type. However, assays that were performed showed that the activity decreased by almost 80%.
The I327V substitution found in the α-loop of the I helix was seen to lack aryl interactions between Mephenytoin and the residues V113 and L366. These changes, in turn, affect mephenytoin’s orientation leading to a decrease of almost 75%.
L380P variant presented a complete absence of pi-pi stack interaction between mephenytoin and Phe476. As it seems, the loss of this interaction is critical for the enzyme's activity since the metabolite cannot be calculated.
Change in position 413 converting the leucine to methionine also increases the distance between the ligand and the Phe476. However, Mephenytoin interacts with Phe476 with its ethyl-group resulting in a decreased enzymatic activity but not absent, found from the in vitro assays.
Finally, in the variant F487S, even though the structural conformation and binding groups of the enzymatic active site stay almost unchanged, the in vitro assays could not detect the drug’s metabolite.
CYP2D6
For the three-dimensional structural modeling of CYP2D6, we used the structure of PDB that was complexed with quinine (PDB code 4WNV). This complex was used due to quinine’s similar structure with dextromethorphan. Performing modeling using X-ray structures complexed with ligands that have a similar structure with the ligand of interest gives an advantage to rigid receptor dockings as the structure is already changed without demanding any computing power. Between the variants’ sequences and the sequences registered in PDB, there were some differences on account of the starting and ending amino acids. Be that as it may, some amino acids were not the same.
Again we performed molecular docking to the wild-type CYP2D6 protein to have the three-dimensional structure as a standard. CYP2D6 active site also has five essential residues, Phe120, Leu213, Glu216, Asp301, Val374, and Phe483. Glu216 acts as the first anchoring point through salt bridge formation with the ligand.
Next, Asp301 binds with Phe120 through hydrogen bonds locating the latter close to the ligand to interact through hydrophobic Pi-Pi interactions. Similarly, Phe483, Leu213, and Val374 interact with the ligand through alkyl or aryl groups, stabilizing it to the correct position.
The F112S mutation makes the enzyme less hydrophobic close to the hydrophobic active site. This change moves Dextromethorphan closer to Phe483, however, so that the ligand cannot properly interact with it. In addition, dextromethorphan moves further from Phe112, resulting in a possible decrease in the enzymatic activity.
Finally, mutation in position 476 converting Serine to Isoleucine increases the distance between Phe483 and dextromethorphan, leading to a weaker interaction between those two. Even though there are not many other important structural differences, in vitro assays, the metabolite could not be detected.
Conclusion
Molecular docking has proven to be essential in drug discovery. However, it reaches its full potential when combined with other experimental procedures. In our project, the use of molecular docking was in agreement with most of the in vitro assays to both CYP2C19 and CYP2D6 variants when interacting with mephenytoin and dextromethorphan, respectively.