Structure of Cytochrome P450 2C8 with no ligand bound. Surface plasmon resonance and isothermal titration calorimetry experiments using CYP2C8 and glucuronides.

Recombinant protein expression of CYP2C8 was carried out in DH5alpha E.coli cell line. Protein purification was performed using nickel affinity and ion-exchange chromatography. High quality of CYP2C8 protein was purified that was used for producing protein crystals necessary for X-ray crystallograpy, and the same protein was used to determine the binding affinity using surface plasmon resonance and isothermal titration calorimetry methods. Regarding the structural studies of CYP2C8 and glucuronide complex and the resulting output files from the crystallography software - CYP2C8_No_GemAG_lowResolution.mtz and .pdb files: The crystals of CYP2C8-gemfibrozil acyl-glucuronide complex diffracted to ~3.4 Å resolution, and the crystallographic data was collected and processed. The three-dimensional structure was solved at this resolution, however, there was no electron density found corresponding to the glucuronide in the active site or elsewhere in the structure. The structure was determined to be a ligand-free structure. Regarding the surface plasmon resonance (SPR) experiments with clopidogrel acyl-glucuronide and CYP2C8. The SPR experiments using the clopidogrel acyl-glucuronide dissolved in water at low stock concentration of 1 mM suggested that the dissociation constant (KD) with cytochrome P450 (CYP2C8) is between 3 to 25 uM. The experimental data looked clearer when this acyl glucuronide was dissoved in water at low stock concentration compared to when dimethyl sulfoxide (DMSO) was used as a solvent. The clopidogrel aycl-glucuronide is insoluble in water at higher stock concentration than 1 mM. It is soluble in DMSO, however, a lot of non-specific binding was observed using DMSO that significantly affected the binding data and analysis. Thus, dH2O was used to make low stock concentration, although not a best option to titrate using low stock concentrations. In addition, the gemfibrozil acyl-glucuronide, another mechanism inhibitor of CYP2C8 studied in this project was not soluble in water even at low concentration of 1 mM. Thus, it was difficult to obtain any data. Of note, a drug named losartan that was soluble in dH2O was used as a contrl to determine the binding with CYP2C8 using SPR and isothermal titration calorimetry (ITC). The results illustrated binding of losartan with CYP2C8 and yielded a good fit. The binding spectra of clopidogrel acyl-glucuronide with CYP2C8 was not as good a fit as losartan, it was still possible to estimate the binding affinity of this glucuronide to be between 3 to 25 uM range. To conclude, solubility and stability of clopidogrel acyl glucuronide and gemfibrozil acyl glucuronide had a significant impact in the experiments analyzing the binding with CYP2C8. These glucuronides are not soluble and stable in water to determine the binding constants effectively using SPR and ITC, and DMSO is not an ideal solvent for these techniques. Regarding the Isothermal Titration Calorimetry (ITC) experiments with clopidogrel acyl-glucuronide and CYP2C8. Solution studies using ITC were probed to determine the binding affinity of clopidogrel acyl-glucuronide with CYP2C8. The experiments were carried out by titrating CYP2C8 with clopidogrel acyl-glucuronide using the MicroCal ITC200 (GE Healthcare). Clopidogrel acyl-glucuronide was dissolved in the protein buffer (50 mM Potassium Phosphate pH 7.4, 500 mM NaCl, 1 mM EDTA and 0.2 mM TCEP) at a low stock concentration of 1 mM and the ITC experiments were performed at 25 degree Celsius temperature, 40 μM protein concentration and clopidogrel acyl glucuronide concentrations at 1, 2, and 4 μM. Losartan (2 μM) was used as a positive control. The ITC binding isotherms were integrated, normalized for concentration, and analyzed according to a one set of sites binding model to calculate the binding affinity. However, all the ITC studies done with CYP2C8 and clopidogrel acyl-glucuronide at three different concentrations of the glucuronide dissolved in the protein buffer did not show any binding (see the 1st slide of the power point .pptx file attached for ITC). ITC studies done with CYP2C8 and losartan as a positive control showed binding with a dissociation constant or KD value of 89.9 ± 13.1 μM (slide 2 of the power point attached). In conclusion, the control ligand losartan did show binding to CYP2C8, indicating that the experiments performed using the active CYP2C8 protein in the ITC buffer conditions were appropriate. The clopidogrel acyl glucuronide ligand did not exhibit any binding with ITC. Attempts to dissolve gemfibrozil acyl-glucuronide in water at low concentrations were unsuccessful and thus, it was difficult to pursue such titrations since DMSO is not a suitable solvent for ITC. Overall, the low solubility and stability of clopidogrel acyl glucuronide and gemfibrozil acyl glucuronide in water likely contributed impacting the binding of these ligands with CYP2C8.