Benzimidazoles: Selective inhibitors of Topoisomerase I with

Dec 26, 2018 - Copyright © 2018 American Chemical Society ... a new analog of Hoechst 33342,was observed as selective and differential inhibitor of H...
0 downloads 0 Views 7MB Size
Article Cite This: Biochemistry XXXX, XXX, XXX−XXX

pubs.acs.org/biochemistry

Benzimidazoles: Selective Inhibitors of Topoisomerase I with Differential Modes of Action Sandhya Bansal,† Souvik Sur, and Vibha Tandon* Special Centre for Molecular Medicine, Jawaharlal Nehru University, New Delhi 110067, India

Biochemistry Downloaded from pubs.acs.org by NEWCASTLE UNIV on 01/12/19. For personal use only.

S Supporting Information *

ABSTRACT: DNA topoisomerases are unique enzymes that alter the topology of DNA by cleavage and religation mechanisms. Small molecules such as camptothecins and noncamptothecins are reported to inhibit different classes of topoisomerases. Benzimidazole, 2-(3,4-dimethoxyphenyl)-5-[5-(4-methylpiperazin-1-yl)-1Hbenzimidazol-2-yl]-1H benzimidazole (DMA), a new analogue of Hoechst 33342, was observed as a selective and differential inhibitor of human and Escherichia coli DNA topoisomerase I. In this study, we have concluded that DMA and Hoechst 33342 have differential binding toward human and E. coli topoisomerase I. We also dissected the mechanism of differential binding, as DMA and Hoechst 33342 bind to human topoisomerase I with linear kinetics with reversible binding, whereas the same molecules bind to E. coli topoisomerase I in a nonlinear and irreversible manner, which contributes to higher affinity and comparatively good IC50 values toward E. coli topoisomerase I. Interestingly, DMA and Hoechst 33342 showed inhibition of mutant human topoisomerases I, i.e., A653P, N722S, and T729P, whereas these clinically relevant mutants are resistant to camptothecins.

DNA topoisomerases. The control of DNA structure by DNA topoisomerases can affect stress response and genomic stability, with implications for bacterial pathogenesis and cancer. Recently, the crystal structure of mycobacterial topoisomerase I was analyzed to understand the metal and DNA interactions at the active site of the enzyme.13 Human topoisomerase I has been subdivided into four distinct domains.14,15 The N-terminal 214 amino acids of the human enzyme are dispensable for relaxation activity in vitro and constitute a hydrophilic, unstructured, and highly protease sensitive region of the protein.16 Contained with the N-terminal domain are four nuclear localization signals and sites for interaction with other cellular proteins. The N-terminal domain is followed by a highly conserved, 421-amino acid core domain that contains all of the catalytic residues except the active site tyrosine.17 A protease sensitive and poorly conserved linker domain comprising 77 amino acids connects the core domain to the 53-amino acid C-terminal domain. An active form of the enzyme can be reconstituted by mixing together fragments corresponding approximately to the core domain (residues 175−659) and the C-terminal domain (residues 713−765), and thus, the linker is dispensable for relaxation activity in vitro. The active site Tyr723 is found within the C-terminal domain. Crystal structures of several forms of the human enzyme with

DNA topoisomerases are ubiquitous in nature and true magicians of the DNA world. These enzymes are termed topoisomerases because they are able to change the topology of DNA molecules without changing the underlying chemical structure of the DNA. DNA topoisomerases play important roles in basic cellular biology.1,2 DNA cleavage by all topoisomerases is accompanied by the formation of a transient phosphodiester bond between a tyrosine residue in the protein and one of the ends of the broken strand. DNA topology can be modified during the lifetime of the covalent intermediate, and the enzyme is released as the DNA is religated.3−5 The first description of a topoisomerase was published in 1971, that of Escherichia coli topoisomerase I, a monomer of 97 kDa6 that is encoded by the topA gene.7 It has a preference for binding at single-strand regions of DNA.8 E. coli topoisomerase I is composed of 865 amino acids that can be divided into three domains. The first 582 N-terminal amino acids correspond to a core “cleavage/strand passage” domain containing the active site tyrosine at position 319.9,10 Human DNA topoisomerase I is a member of the type 1B family because when it cleaves DNA it becomes covalently attached to the 3′ DNA end of the break site.11,12 Every bacterial pathogen has at least one type IA topoisomerase, providing a target for discovery of new antibiotics to combat multidrug resistant infections, including MDR- and XDR-TB. The study of the structure and mechanism of bacterial topoisomerase I provides the basic foundation for translational application of this enzyme as a novel antibacterial target. Drug discovery research extends to anticancer drugs targeting human © XXXX American Chemical Society

Received: October 16, 2018 Revised: December 22, 2018 Published: December 26, 2018 A

DOI: 10.1021/acs.biochem.8b01102 Biochemistry XXXX, XXX, XXX−XXX

Article

Biochemistry

Afterward, the protein and optimized structures of ligands were converted to required pdbqt format using AutodockTools 1.5.4.37 Autodock tools (ADT) were used to merge nonpolar hydrogens of protein and assign atomic charges. The nonpolar hydrogen of each compound was merged, and rotatable bonds were assigned. Grid maps were generated for each atom type using AUTOGRID. An active site box of 70 Å × 60 Å × 70 Å with a grid spacing of 0.375 Å was created and placed at the center of the protein. Docking calculations were carried out using the Lamarckian genetic algorithm. In docking calculations, we have chosen the time-consuming Lamarkian genetic algorithm (GA). One hundred iterations of GA with 2.5 million energy evaluations per run were carried out. The population size was set to 150, and a maximum of 27000 generations per run was carried out, followed by automatic clustering of poses. Fifty independent runs for the compound were performed with each topoisomerase structure. The resulting positions were clustered according to a root-mean-square criterion of 0.5 Å. We obtained two and three clusters of mode with Hoechst 33342 and DMA, respectively, with E. coli topoisomerase I and two clusters of mode with Hoechst 33342 and DMA both with human topoisomerase. The energy difference between the clusters was around 0.05−0.08 kcal mol−1. Finally, the best conformation of the ligand with the lowest binding energy was selected to analyze the probable interaction, i.e., topoisomerase and benzimidazoles irrespective of the population size of clusters. Molecular visualizations were performed in Discovery Studio.38 Plasmid DNA Relaxation Assay. DNA topoisomerase was assayed by decreasing the mobility of the relaxed isomers of supercoiled pBSKII(+) DNA in an agarose gel. The total volume of the reaction mixture was 25 μL, which contained 0.5 μg of pBSKII(+), 2 units of enzyme topoisomerase I (1 unit is defined as the amount of enzyme required to convert 50% of 0.5 μg of supercoiled DNA substrate to the relaxed form under standard assay conditions), and 1× relaxation buffer [25 mM Tris-HCl (pH 7.5), 5% glycerol, 50 mM KCl, 0.5 mM dithiothreitol, and 30 μg/mL bovine serum albumin]. Reactions were performed at 37 °C for 30 min and then terminated by adding 10 mM ethylenediaminetetraacetic acid (EDTA), 0.5% sodium dodecyl sulfate, 0.25 μg/mL bromophenol blue, and 15% glycerol to the mixture. The samples were electrophoresed in a horizontal 1% agarose gel in tris-borate/EDTA buffer [40 mM tris-acetate and 2 mM EDTA (pH 8.0)] at 1.5 V/cm for 14−16 h at room temperature. DMSO concentrations in each reaction mixture were maintained at 1% by the addition of serially diluted ligand stocks so as not to produce solvent-mediated inhibition of topoisomerase I. The gels were stained with ethidium bromide (5 μg/mL), destained in water, and photographed under ultraviolet illumination with an alpha imager 2200. Fluorescence Measurements. Equilibrium binding experiments were conducted using Fluoromax, Horiba (Jobin Yvon) using a 1 mL quartz cuvette with a 1 mm path length at 25 °C. Titrations were performed in pairs of two. All spectral titrations were performed at 25 °C in fluorescence buffer [20 mM Tris-HCl (pH 7.5), 50 mM NaCl, and 10 mM MgCl2]. Background emission (2%) was corrected by the subtraction of spectra of blank buffer and enzyme buffer samples from all pairs of experiments. A hyperbolic plot of an increasing concentration of DNA versus λmax emission wavelength was drawn, and dissociation constant Kd was determined from the slope of the plot for DMA, Hoechst 33342 and camptothecin with plasmid DNA, and all of the proteins (human topoisomerase I, E. coli topoisomerase I, core domain of human topoisomerase I,

both noncovalently and covalently bound DNA have been determined.1 These co-crystal structures represent the only examples to date of a topoisomerase containing bound DNA. Though crystals were grown with an N-terminally truncated active form of the protein missing the first 174 amino acids, Xray density was only interpretable beginning at residue 215, and thus, the entire N-terminal domain is missing from the structure.2 Camptothecin binds to and reversibly stabilizes the covalent topoisomerase I−DNA complex, which slows the religation phase of the enzyme’s catalytic cycle and prolongs the lifetime of the covalent protein−DNA complex.17 A recent crystal structure of the ternary complex between the active topo70 construct of human topoisomerase I covalently linked to DNA and topotecan revealed that the drug stacks into the DNA duplex and replaces the base pair adjacent to the single-strand site of cleavage.18,19 Topoisomerase I mutations that render the enzyme resistant to camptothecin can provide evidence for the interactions of the drug molecule with the enzyme. The important question of why structurally different drugs inhibit topoisomerase at different or the same concentrations remains. Several human topoisomerase I mutations, located either near or far from the active site, have been shown to render the enzyme resistant to the camptothecin family drug. Point mutations at residues 363 (Gly to Cys), 418 (Glu to Lys), 503 (Gly to Ser), 533 (Asp to Gly), 583 (Asp to Gly), 653 (Asp to Pro), 717 (Gly to Val), 718 (Thr to Ala), 722 (Asn to Ser), and 729 (Thr to Ala or Pro) result in the production of camptothecin resistant enzymes.20−29 Indolocarbazoles,30 DNA minor groove binding ligands, intercalators, saintopin, and camptothecin analogues were screened against camptothecin resistant human topoisomerase I. The previous report by our group31 suggested selective inhibition of bacterial topoisomerase over human topoisomerases. This paper deals with how the order of addition of each component (topoisomerase I enzyme, supercoiled plasmid DNA, and benzimidazole) makes a difference in the inhibitory effect of these molecules toward human and E. coli topoisomerase I. The fluorescence titration studies also confirmed differential binding modes of molecules. Besides, we have employed a molecular docking study to correlate our finding with a theoretical approach. In addition, we describe here a few clinically relevant mutants of human topoisomerase I and their activity toward DMA and Hoechst 33342. Although a number of mutations in the human topoisomerase I gene that render the enzyme inactive toward different classes of inhibitors are reported, three of those constructs (A653P, N722S, and T729P) that are resistant to camptothecin were found to be sensitive to DMA and Hoechst 33342.



METHODS Camptothecin was purchased Sigma (St. Louis, MO) and dissolved in dimethyl sulfoxide (DMSO). Synthesis of the benzimidazole derivative DMA was described previously.32 These derivatives were dissolved in methanol. pHOT1 plasmid DNA was purchased from TopoGen Inc. (Port Orange, FL). Molecular Docking Protocol. The crystal structures of E. coli topoisomerase I [Protein Data Bank (PDB) entry 1ECL] and human topoisomerase (PDB entry 1EJ9) have been obtained from RCSB PDB. The active compounds were viewed in Viewer Lite software and then brought to their energetically minimized structures by Gaussian (version 09W), utilizing a conjugate gradient method with the 6-31G MP2 force field. B

DOI: 10.1021/acs.biochem.8b01102 Biochemistry XXXX, XXX, XXX−XXX

Article

Biochemistry

Figure 1. (A) Chemical structures of Hoechst 33342 and DMA. (B) Docked pose of the binary complex of Hoechst 33342 with E. coli topoisomerase I, showing interaction of Hoechst with Glu9, Glu115, Asp113, and Tyr319. (C) Docked pose of the binary complex of DMA with E. coli topoisomerase I, showing interaction of DMA with Asp113, Tyr319, and Arg321. (D) Docked pose of the binary complex of Hoechst 33342 with human topoisomerase I, showing interaction of Hoechst with Lys532, Arg488, and Arg590. (E) Docked pose of the binary complex of DMA with human topoisomerase I showing interaction with Asp538, Ile535, Arg488, etc. All of the structures are colored by element.

studied (Figure 2A,B). Details of cloning, expression, and purification of topoisomerase proteins are provided in Figures S5−S7. It was observed that simultaneous addition of enzyme, plasmid DNA, and ligand inhibits (IC50 values) relaxation of plasmid DNA at 2.83 ± 2.8 μM camptothecin, 56.41 ± 3.5 μM DMA, and 6.72 ± 1.9 μM Hoechst 33342 (Figure 2A). Preincubation of enzyme with ligand for 5 min followed by addition of DNA inhibits plasmid relaxation using 2.83 ± 3.0 μM camptothecin, 15.7 ± 2.5 μM DMA, and 4.80 ± 4.5 μM Hoechst 33342. With E. coli topoisomerase I, simultaneous addition of enzyme, DNA, and ligand showed inhibition of plasmid DNA relaxation at 8.8 ± 2.2 μM DMA and 11.13 ± 3.7 μM Hoechst 33342 (IC50 values). Camptothecin did not show any effect on relaxation activity of E. coli topoisomerase I (Figure 2B). After preincubation of E. coli topoisomerase I with DMA and Hoechst 33342 for 5 min followed by addition of plasmid DNA, the levels of plasmid relaxation inhibition, i.e., IC50 values, were reduced to 0.74 ± 1.6 and 4.19 ± 2.5 μM for DMA and Hoechst 33342, respectively. Quantitative analysis of simultaneous and preincubation inhibition is represented in the graphs of Figures S8 and S9. To observe the effect of DMA and Hoechst 33342, the mutant protein relaxation assay was performed with topoisomerase I mutants A653P, N722S, and T729P (Figure 3A−C). When there is simultaneous addition of enzyme, plasmid DNA, and ligand, relaxation inhibition of mutated topoisomerase I (A653P) was observed at 61.3 ± 3.2 and 4.01 ± 1.3 μM with DMA and Hoechst 33342, respectively (Figure 3A), but after a 5 min preincubation of enzyme with ligand, followed by addition of plasmid DNA, the IC50 values decreased to 7.8 ± 1.0 and 1.31 ± 2.7 μM for DMA and Hoechst 33342, respectively. The N722S mutant protein inhibited plasmid relaxation at 58.41 ± 2.5 μM DMA and 30.6 ± 2.1 μM Hoechst 33342 when enzyme, plasmid DNA, and ligand were added simultaneously (Figure 3B), but after a 5 min preincubation of enzyme with ligand followed by addition of plasmid DNA, relaxation inhibition of plasmid DNA was observed at 15.9 ± 3.2 μM DMA and 18.2 ±

catalytic domain of human topoisomerase I, catalytic domain of E. coli topoisomerase I, and C-terminal domain of E. coli topoisomerase I).



RESULTS AND DISCUSSION Molecular Docking Study of Bisbenzimidazoles with E. coli and Human Topoisomerase I. Hoechst 33342 and DMA were both found to be docked in the active site of the catalytic domain of E. coli topoisomerase I. Table S1 depicts the docking scores found between topoisomerase and two benzimidazole analogues. Hoechst 33342 was found to be packed more efficiently with E. coli topoisomerase I with a docking score of −10.42, whereas DMA showed a moderate value of −8.65. In a similar way, Hoechst 33342 binds better (−8.71) with human topoisomerase than DMA does (−7.27) (Table S1). Tyr319, Glu9, Asp113, and Glu115 directly interact with the aromatic nucleus of benzimidazoles via π−π interactions and H-bonding between N−H of imidazoles and the free lone pair of Nmethylated piperazine. Earlier, it was reported32 that an acidic triad, Asp111, Asp113, and Glu115, is conserved throughout type IA DNA topoisomerases, located near the active site Tyr319, and the acidic triad participates in metal coordination, which is important for religation of DNA. The ethoxy group of Hoechst 33342 showed interaction with Ala499 (Figure 1 and Figures S1 and S2), which was not found in the case of DMA. Besides, these benzimidazoles bound in the DNA binding pocket, which is also the catalytic domain of human topoisomerase I. Lys532, His632, Phe723, and Arg488 interact with both DMA and Hoechst 33342. The other residues found in the vicinity are aromatic amino acids. Hoechst 33342 binds with Phe259, Phe274, and Phe361, whereas DMA interacts with Phe309, Phe274, Arg362, and Lys262 (Figure 1 and Figures S3 and S4). DNA Relaxation Assays of Human Topoisomerase I, E. coli Topoisomerase I, and Mutated Human Topoisomerase I. The effect of camptothecin, DMA, and Hoechst 33342 on the relaxation activity of human and E. coli topoisomerase I was C

DOI: 10.1021/acs.biochem.8b01102 Biochemistry XXXX, XXX, XXX−XXX

Article

Biochemistry

Figure 2. pBSKII(+) plasmid DNA relaxation with (A) human topoisomerase I and (B) E. coli topoisomerase I under simultaneous and preincubation conditions in the presence of (a) camptothecin, (b) Hoechst 33342, and (c) DMA: lane 1, control; lane 2, topoisomerase I and DNA; lane 3, ligand (1 μM), DNA, and topoisomerase I; lane 4, ligand (5 μM), DNA, and topoisomerase I; lane 5, ligand (10 μM), DNA, and topoisomerase I; lane 6, ligand (25 μM), DNA, and topoisomerase I; lane 7, ligand (50 μM), DNA, and topoisomerase I; lane 8, ligand (75 μM), DNA, and topoisomerase I; lane 9, ligand (100 μM), DNA, and topoisomerase I. Five hundred nanograms of plasmid DNA was incubated in the presence of 2 units of DNA topoisomerase enzyme. Abbreviations: OC, open circular; RM, relaxed monomer; SC, supercoiled.

3.7 μM Hoechst 33342. The T729P protein inhibited (IC50) plasmid relaxation at values of 71.1 ± 1.3 and 58.3 ± 3.6 μM for DMA and Hoechst 33342, respectively, when enzyme, DNA, and ligand were added simultaneously (Figure 3C), and the IC50 values decreased to 55.14 ± 1.3 and 44.9 ± 3.9 μM for DMA and Hoechst 33342, respectively, upon preincubation with enzyme. Quantitative analysis of simultaneous and preincubation inhibition is represented in the graphs in Figure S10A−C. No inhibition was observed for camptothecin under any condition at concentrations of ≤100 μM. For time-dependent relaxation, topoisomerase I was added to plasmid DNA in a ratio of 3:1 and reactions were terminated at different time intervals as mentioned above. The amount of enzyme used in the time-course assay should be adjusted to ensure that all of the plasmid substrate will be fully relaxed in

Figure 3. pBSKII(+) plasmid DNA relaxation with (A) mutated human topoisomerase I A653P, (B) mutated human topoisomerase I N722S, and (C) mutated human topoisomerase I T729P under simultaneous and preincubation conditions in the presence of (a) camptothecin, (b) Hoechst 33342, and (c) DMA: lane 1, control; lane 2, topoisomerase I and DNA; lane 3, ligand (1 μM), DNA, and topoisomerase I; lane 4, ligand (5 μM), DNA, and topoisomerase I; lane 5, ligand (10 μM), D

DOI: 10.1021/acs.biochem.8b01102 Biochemistry XXXX, XXX, XXX−XXX

Article

Biochemistry

Hoechst 33342 bind irreversibly to E. coli topoisomerase I, and the reaction kinetics was not linear (Figure 4B). This is because of the strong binding of the enzyme with the bisbenzimidazoles. DMA and Hoechst 33342 interact with free enzyme as well as free DNA, suggesting that they belong to the noncompetitive class of inhibitors. To further understand the mechanism of action of benzimidazoles, three camptothecin resistant mutant human topoisomerase I proteins (A653P, N722S, and T729P) were expressed and purified for this study. Time-dependent relaxation with mutated human topoisomerase I has shown linear enzyme kinetics (Figure 4C−E). DNA was not completely relaxed after 60 min in the presence of ligands DMA and Hoechst 33342 by all of the three mutated human DNA topoisomerases I (A653P, N722S, and T729P). This suggests that the reaction is not completely reversible in the presence of bisbenzimidazoles up to 60 min for mutated topoisomerase I (Figures S13−S15). Of the three mutations considered in our study, A653P is located in the linker domain, which is far from the ligand binding site, shown by computational modeling and X-ray crystallography.30 N722S and T729P are located in the catalytic domain of human toposiomerase I. Asn722Ser is a camptothecin resistant mutation that shows little impact on the catalytic activities of human topoisomerase I. Thr729 is buried within the C-terminal domain of the CAT region of human topoisomerase I.29 It may be hypothesized that Asn722 and Thr729 are part of a minor groove intercalating wedge and this interaction occurs at the DNA cleavage site. Although mutation of Ala653 makes topoisomerase resistant to camptothecin, this mutation is far

Figure 3. continued DNA, and topoisomerase I; lane 6, ligand (25 μM), DNA, and topoisomerase I; lane 7, ligand (50 μM), DNA, and topoisomerase I; lane 8, ligand (75 μM), DNA, and topoisomerase I; lane 9, ligand (100 μM), DNA, and topoisomerase I. Five hundred nanograms of plasmid DNA was incubated in the presence of 2 units of the DNA topoisomerase enzyme. Abbreviations: OC, open circular; RM, relaxed monomer; SC, supercoiled.

approximately 8−12 min. Time-course assays are linear for enzyme concentrations that lead to complete relaxation of the substrate within a range of 5−60 min. With the results from time-dependent relaxation, human and E. coli topoisomerase I show linear enzyme kinetics as the time required to complete the relaxation process is