Investigation of inhibition of human glucose 6-phosphate dehydrogenase by some 99mTc chelators by in silico and in vitro methods

Abstract

The inhibitory effects of methoxyisobutylisonitrile (MIBI), diethylene triamine pentaacetic acid (DTPA), dimercaptosuccinic acid (DMSA) and metilendifosfonat (MDP) on human erythrocyte glucose 6-phosphate dehydrogenase (hG6PD) activity were investigated. For this purpose, hG6PD was initially purified 557-fold at a yield of 51.43% using 2′,5′-adenosine diphosphate (ADP) sepharose 4B affinity gel chromatography. The in vitro effects of these chelators on hG6PD enzyme were studied. IC₅₀ values of MIBI, DTPA, DMSA and MDP were 0.056, 0.172, 0.274 and 0.175 mM, of hG6PD, respectively. It was detected in in vitro studies that the hG6PD enzyme is inhibited due to these radiopharmaceutical chelators. In addition to in vitro studies, in order to better understand the molecular mechanism of studied compounds, combined in silico approaches, including molecular docking and molecular dynamics (MD), simulations were successfully performed. MD simulations shed light on inhibition mechanisms of the individual inhibitors into the ligand-binding pocket of hG6PD. Essential amino acids for binding are also investigated using per-residue interaction analysis studies.

Keywords:

• Chelator
• human G6PD
• molecular modeling
• radiopharmaceutical

Introduction

Glucose-6-phosphate dehydrogenase (D-glucose-6-phosphate: NADP⁺oxidoreductase, EC 1.1.1.49; G6PD) is a critical enzyme in mammalian erythrocytes that catalyzes the first reaction and rate-limiting enzyme in the pentose phosphate pathway (PPP)1,2. The main physiological function of G6PD is the production of NADPH and ribose 5-phosphate, which are essential for reductive biosynthesis and nucleic acid synthesis (Figure 1)3. G6PD abnormalities affect as much as 7.5% of the global population, with a wide range of occurrence based on geographic distribution. These abnormalities occur less than 0.1% of certain Japanese populations to more than 35% in some African and European populations4. NADP⁺is reduced to NADPH by glutathione reductase (glutathione: NADP⁺oxidoreductase, EC 1.8.1.7; GR) in erythrocytes using glutathione as substrate. Glutathione prevents hemoglobin denaturation, preserves the integrity of red blood cell membrane sulfhydryl groups and detoxifies hydrogen peroxide and oxygen radicals in the erythrocytes5–7. A decrease in G6PD may result in deficiencies of NADPH and reduced glutathione in erythrocytes and a scarcity of reduced glutathione in erythrocytes causes their early hemolysis in the spleen8,9.

Figure 1. Pentose phosphate pathway (PPP) and glutathione coupling: Glucose-6-phosphate dehydrogenase (G6PD), 6-phosphogluconate dehydrogenase (6PGD), glutathione reductase (GR), glutathione (GSH), oxidized glutathione (GSSG), glutathione peroxidase (GPx).

Knowledge of the physiological and biochemical mechanisms that account for the myocardial localization of these agents is important for two reasons. First, an elucidation of the mechanisms of uptake, distribution and retention of these agents is important for the better understanding of the diagnostic information that each can provide. Second, understanding cationic fluxes in the heart can provide an important design strategy for new radiopharmaceuticals10. There are numerous radiopharmaceutical nuclides used in diagnostic nuclear medicine. However, at present, ^99mTc is used in about 90% of all scintigraphic examinations11. One of the most important diagnostic technique in nuclear medicine for cardiology is the examination of myocardal perfusion for the diagnosis of infarction and ischemia. Methoxyisobutyl isocyanide (MIBI) is the compound used in heart scintigraphy with ^99mTc11 and each MIBI acts as monodentally, so that six MIBI molecules are needed to form a hexa MIBI complex with single ^99mTc cations12. The complex will be cationic and be absorbed into the parenchymal cells of the heart myocardium13. Dynamic kidney scintigraphy, renography and determination of glomerular filtration rate (GFR) can be performed using ^99mTc-diethylene triaminepentaacetic acid (DTPA). This means that the physiological events in the kidneys can be studied. ^99mTc-DTPA is passed through the kidneys without any binding to kidney tissue. Kidney localization and conditions like infarction can be determined by the use of dimercaptosuccinic acid (DMSA)11. One of the most common examinations in nuclear medicine and oncology is the bone scan for metastases from prostate and breast cancers. The chelator used in bone scintigraphy is methylenediphosphonate (MDP)14. Many radiopharmaceuticals are now being used in therapies.

However, the effects of MIBI, DTPA, DMSA and MDP have not been analyzed on activities of hG6PD. Therefore, the in vitro effects of these radiopharmaceuticals on hG6PD purified from human erythrocytes were investigated in this study. Using the obtained IC₅₀ values, undesirable side effects on G6PD activity and body metabolism can be diminished in used therapies.

Materials and methods

Materials

2′,5′-Adenosine diphosphate (ADP) sepharose 4B was obtained from Pharmacia. NADP⁺, glucose-6-phosphate, protein assay reagent and chemicals for electrophoresis were purchased from Sigma. All other chemicals used were of analytical grade and were purchased from Merck. MIBI, DTPA, DMSA and MDP were obtained from Eczacıbaşı Monrol (Turkey).

Preparation of hemolysate

Fresh human blood collected in eppendorf tubes with ethylenediamine tetraacetic acid (EDTA) was centrifuged at 2500xg for 15 min and the plasma and leukocyte coat were removed. The packed of red cells were washed with KCl solution (0.16 M) three times and pelleted each time by centrifuged at 2500xg. The supernatants were discarded. The pelleted erythrocytes were hemolysed with 5 vol. of ice-cold water and centrifuged at 4 °C at 10 000×g for 30 min to remove the ghosts and any remaining intact cells6,8. Ammonium sulfate was slowly added and the fraction between 35% and 65% saturation was collected and centrifuged at 5000xg for 15 min. the pellet was dissolved in 50 mM phosphate buffer (pH 7.0), then dialyzed at 4 °C against 50 mM K-acetate/50 mM K-phosphate buffer (pH 7.0) for 2 h with two changes of buffer.9

Purification of hG6PD by affinity chromatography

Dialyzed sample was loaded onto a 2′,5′-ADP sepharose 4B affinity column and the gel was washed with 25 mL of 0.1 M K-acetate/0.1 M K-phosphate (pH 6.0), followed by 25 mL of 0.1 M K-acetate/0.1 M K-phosphate (pH 7.85) and finally with 0.1 M KCl/0.1 M K-phosphate (pH 7.85). Elution was carried out with an elution buffer (80 mM K-phosphate + 80 mM KCl + 0.5 mM NADP⁺ + 10 mM EDTA, pH 7.85) at a flow rate of 20 mL/h. Eluate fractions were collected in 2-mL tubes and the enzyme activity in each was calculated separately. All of the purification procedures were performed at 4 °C15. Quantitative protein determination was carried out with Bradford’s method using bovine serum albumin as a standard16. The purified enzyme was run on a SDS polyacryamide gel electrophoresis gel (0.1% SDS) according to Laemmli’s method17 using a 4% stacking gel and a 10% running gel.

Measurements of hG6PD activity

hG6PD activity assays were run at 25 °C according to Beutler’s methods, which depends on the reduction of NADP⁺by hG6PD, in the presence of glucose 6-phosphate. For this spectrophotometric measurement, the reaction medium contained 0.1 mM Tris-HCl (pH 8.5) with 0.2 mM NADP⁺ and 0.6 mM (G6P) in a total volume of 1 mL. One unit of enzyme (EU) activity was defined as the amount of enzyme that reduced 1 μmol NADP⁺per minute at pH 8.518.

Inhibitor study

MIBI, DTPA, DMSA and MDP were used as inhibitors. In the media with or without inhibitor, the substrate (G6P) concentrations were 0.012, 0.024, 0.036, 0.048 and 0.06 mM. Inhibitor (radiopharmaceutical chelator) solutions were added to the reaction medium, resulting in three different fixed concentrations of inhibitors in 1 mL of total reaction volume. In order to determine IC₅₀ values, activities were calculated with a 0.60 mM constant substrate (G6P) and different inhibitor concentrations. Cuvette activity in the absence of inhibitor was acknowledged to be 100%. Regression analysis graphs were drawn using inhibition% values by a statistical package on a computer. The inhibitor concentrations causing up to 50% inhibition (IC₅₀) were determined from the graphs2,19.

Atomistic in silico approaches

System setup

X-ray structure of human glucose-6-phosphate dehydrogenase complexed with glucose-6-phosphate was retrieved from the Protein Data Bank (PDB ID: 2BHL)20. The structure was repaired and prepared via following of consequent steps; building of missing side chains, assignment of bond orders, adding hydrogen atoms and keeping of the crystallized water molecules within 5 Å of the active site, assigned by the cocrystallized ligand in PBD file. The constructed system was assigned at the physiological pH (pH = 7.4) and then minimized using molecular mechanic (MM) methods with the force field of OPLS 200521. The obtained system from process above, were implemented in atomistic simulations, carried out in current calculations.

Ligand data set

Two-dimensional (2D) structures of the used compounds are sketched in Maestro molecular modeling package of the Schrodinger.22 Protonation states of studied compounds are determined and these molecules are optimized in the physiological pH using LigPrep module23. Conformational search is also performed for each compound to find the lowest energy conformer.

Molecular docking simulations

Docking protocol of Glide24 and also Prime module25 implemented in Schrodinger Suit were used to perform a flexible-receptor docking study. Induced fit docking (IFD) module of Schrodinger Suit26 employs Glide and Prime protocols to carry out this calculation. This approach help to predict interaction energy (docking score) of studied compounds and also conformations of small molecules at the binding pocket of the target structure. IFD consists of following consequent steps; (i) docking of ligand at the binding pocket of the protein, (ii) refining of atoms of amino acids (backbone and side-chain atoms) within 5 Å of docked ligand, (iii) redocking of ligand into the active site.

Molecular dynamics (MD) simulations

Classical molecular dynamics (MD) simulations were implemented to investigate the time-dependent conformational changes of complex structures and per-residue interaction energy analyses. HG6PD in complexes with DTPA, DMSA and MDP were considered, and the three complexes (initial conformers were retrieved from the outputs of IFD poses and a control system (ligand-free protein; apo state) were submitted in MD simulations. Four atomistic MD simulations were independently carried out using Desmond code27. All systems were solvated by SPC water model and then naturalized. The thickness of water layer was set to 10 Å. Before the production MD run, the systems were minimized with a maximum iteration of 2000 steps. Then, each of the systems were submitted in a 10 ns MD simulations. Temperature and pressure were assigned on 300 K and 1.01325 bar, respectively, using isothermal–isobaric (NPT) ensemble. Nose-Hoover chain and Martyna–Tobias–Kleinwas were implemented thermostat and barostat methods, respectively. Cutoff radios of 9 Å and long-range method of smooth Particle Mesh Ewald (PME) were used for Coulomb interactions.

Results and discussion

Worldwide, G6PD deficiency is frequently seen in African, Mediterranean, Middle East and far East nations and their lineages, with a frequency ranging from 5% to 40%28. The use of pamaquine in malaria therapy resulted in same severe side effects, including dark color urine and anemia in some patients. Along with these outward symptoms, G6PD deficiency was also determined in these individuals. Use of this and a number of other drugs cause the hemolysis that is connected with complications. These data showed that enzymes and related reactions were important for the maintenance of life29.

GSH is used in antioxidant defence mechanisms and its production requires that NADPH be synthesized in the pentose phosphate metabolic pathway in which G6PD and 6PGD participate19–30. Inhibitory effects of many drugs and other chemicals on G6PD enzyme activities in different animal species and human beings have been reported in many investigations. For example, thallium-201, thiamphenicol, amikacin, gentamycin, netilmycin, chloramine-T and CuSO₄ can inhibit erythrocyte G6PD activity2,9,15. In other studies, anesthetics inhibited hGR and carbonic anhydrase (CA) activity in the human erythrocyte31–34. Ciftci et al. investigated effects of some antibiotics on human erythrocyte G6PD and these antibiotic IC₅₀ values for 2.9 to 125 mM15.

As shown in Table 1, the specific activity of the purified hG6PD was 5.68 EU/mg proteins, which represented a yield of 51.43% and a purification of 557-fold. The IC₅₀ value for MIBI, DTPA, DMSA and MDP were 0.056, 0.172, 0.274 and 0.175 mM, of hG6PD, respectively (Table 2). The IC₅₀ graphics showed in Figures 2–5.

Figure 2. Activity% versus [MIBI] regression analysis graphs for hG6PD in the presence of seven different MIBI concentrations.

Figure 3. Activity% versus [DTPA] regression analysis graphs for hG6PD in the presence of seven different DTPA concentrations.

Figure 4. Activity% versus [DMSA] regression analysis graphs for hG6PD in the presence of six different DMSA concentrations.

Figure 5. Activity% versus [MDP] regression analysis graphs for hG6PD in the presence of six different MDP concentrations.

Table 1. Purification scheme of hG6PD by 2′,5′-ADP sepharose 4B affinity gel chromatography.

Purification step	Activity (EU/ml)	Total volume (ml)	Protein (mg/ml)	Total Protein (mg)	Total Activity (EU)	Specific Activity (EU/mg)	Yield (%)	Purification factor
Hemolysate	0.42	30	41.2	1236	12.6	0.0102	100	1
Ammonium sulfate precipitation (35–65)%	1.12	8	1.38	11.04	8.96	0.811	71.11	79.5
2′,5′-ADP Sepharose 4B affinity chromatography	2.16	3	0.38	1.14	6.48	5.68	51.43	557

Table 2. IC₅₀ and predicated binding energy values of hG6PD inhibitors.

Cmp	Structure	IC50 (mM)	IFD scores (kcal/mol)
MIBI		0.056	–
DTPA		0.172	−8.821
DMSA		0.274	−9.061
MDP		0.175	−8.214

Molecular modeling and simulations

Docking and atomistic MD simulations approved as informative approaches, profoundly provided to track and pursue the treatments of protein and small molecules35–45. Conformational stabilities and transitions, protein–ligand interactions (hydrogen bonding, hydrophobic interactions, salt bridges, etc.) and prediction of interaction energies of ligands, can be investigated using molecular-modeling applications. Table 2 shows predicted interaction energies of studied compounds DTPA, DMSA and MDP using IFD method (docking pose of MIBI compound could not obtained using IFD methods). Docking scores of DTPA, DMSA and MDP molecules are founds as −8.821, −9.061 and −8.214 kcal/mole, respectively.

In order to better monitor and understand the inhibition mechanism of each compound and how the inhibitors influence the structure of protein upon binding at the active site, the complexes, derived from IFD study were submitted in 10 ns MD simulations.

As seen in Figure 6 (upper panel), the flexibility degree of hG6PD (Cα atoms were included) upon binding to DTPA, DMSA and MDP were evaluated throughout the simulations. Ligand-free system revealed higher mobility compared to holo conformers. HG6PD in complexes with MDP and DTPA were fluctuating in stable conditions and converging after 4 ns. In addition, fluctuations of ligands at the active site based on their heavy atoms are investigated using root mean square deviation (RMSD) analysis in Figure 6 (lower panel). DMSA shows different RMSD profile throughout the simulations compared to other two studied compounds DTPA and MDP due to its unique inhibition mechanism.

Figure 6. RMSD evolution of Cα of hG6PD in complexes with MDP, DTPA and DMSA compounds, and control system (upper panel) and also heavy atoms of three ligands into the active site (lower panel) during 10 ns MD simulation.

Protein–ligand interaction analysis of MDP, DTPA and DMSA into the ligand-binding pocket of hG6PD throughout the MD simulations is shown in Figure 7. Hydrogen bonding, hydrophobic interactions and water bridges formed between ligands and target were considered and their interaction occupancies were calculated using derived trajectory frames from MD simulations. Three complexes were submitted in postprocessing analysis and the interaction fractions for each residue at the binding pocket are shown in Figure 7. In the case of DMSA, the profile suggests that His201 is the most popular amino acid, contributed in the binding interactions. This residue simultaneously establishes hydrogen bonding, ionic bond and water bridges with DMSA atoms. Lys171 is the other key amino acid, participating in forming of hydrogen bonding, water bridges and ionic bond interactions. A water molecule in the active site mediates in establishing of polar interactions by Asp258, as describing the important role of waters in stabilization of ligand. DTPA compound is large in size, and it is expected to have interactions with various amino acids. Lys360, Asp258, Arg246, Glu239 and His201 are the most important amino acids forming the active site of the complex. The interaction profile of MDP-bound complex reveals that Lys47, Glu239, Lys205, His201, Tyr202, Gln305 and Arg365 amino acids can significantly play into the ligand-protein interactions. In light of our interaction analysis, performed throughout MD simulation, we found that inhibitor-binding pocket of hG6PD is increasingly tending to accommodate the ligands via polar and charged connections, and the role of the hydrophobic interactions has not been pronounced. Figure 8 shows 3D ligand interaction diagrams of studied compounds.

Figure 7. Protein–ligand interaction analysis of MDP, DTPA and DMSA into the ligand-binding pocket of hG6PD throughout the MD simulations.

Figure 8. 3D ligand interaction diagrams of studied compounds.

Conclusions

The used of ^99mTc chelators (MIBI, DTPA, DMSA and MDP) may worsen the health of patients who already have a deficiency of G6PD enzyme. According to these data, these radiopharmaceutical chelators were found to be more efficient inhibitors to G6PD enzyme activity than were other drugs. Owing to the wide use of these compounds in both humans and animals, knowing the effect of these materials on G6PD activity is important. For this reason, if these materials are given to the patients with G6PD deficiency, their dosages should be very tightly monitored to prevent their side-effects on G6PG in red blood cells. Molecular simulations results revealed that polar interactions, not the hydrophobic interactions play important role at the binding of the studied compounds in this study.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.