Trapping of the transport-segment DNA by the ATPase domains of a type II topoisomerase

Abstract

Type II topoisomerases alter DNA topology to control DNA supercoiling and chromosome segregation and are targets of clinically important anti-infective and anticancer therapeutics. They act as ATP-operated clamps to trap a DNA helix and transport it through a transient break in a second DNA. Here, we present the first X-ray crystal structure solved at 2.83 Å of a closed clamp complete with trapped T-segment DNA obtained by co-crystallizing the ATPase domain of S. pneumoniae topoisomerase IV with a nonhydrolyzable ATP analogue and 14-mer duplex DNA. The ATPase dimer forms a 22 Å protein hole occupied by the kinked DNA bound asymmetrically through positively charged residues lining the hole, and whose mutagenesis impacts the DNA decatenation, DNA relaxation and DNA-dependent ATPase activities of topo IV. These results and a side-bound DNA-ParE structure help explain how the T-segment DNA is captured and transported by a type II topoisomerase, and reveal a new enzyme-DNA interface for drug discovery.

Fig. 1

T-segment DNA capture and transport by type IIA topoisomerases. a The proposed type II topoisomerase reaction cycle exemplified by topoisomerase IV. ParC subunits are in grey, ParE N-terminal domain is in cyan, ParE C-terminal TOPRIM domain (containing the metal binding TOPRIM fold) is in yellow. G-gate DNA is in green and the transported T-segment DNA is in mauve. ATP bound to the ATPase domain is denoted by a red dot. Drug-targetable domains within the type II topoisomerase complex are highlighted in subsections A, B and C with examples on the right-hand side of the figure. b Domain organization of S. pneumoniae topoisomerase IV. The active enzyme is a ParC₂ParE₂ tetramer. The ATPase domain (residues 1–402) is shown in light blue. c Orthogonal views of the ParE44 protein in complex with a captured 14-mer T-segment DNA. N-terminal ParE domains are shown in cartoon mode in light blue and cyan, bound AMP-PNP molecules are in red and T-segment DNA is in mauve/yellow for backbone/bases, respectively

Fig. 2

Molecular features of the ParE44 complex with hole-bound 14-mer DNA. a Amino acid side chains that interact with the DNA duplex are shown as ellipses with colouring corresponding to the two subunits of the ParE dimer. Shortest distances between the DNA and the amino acid side chains are given in Ångstroms. The 6 bp DNA section that interacts with ParE is coloured, the solvent-exposed region is not. The top and bottom DNA strands are in purple and pink and are marked by single and double asterisks. X denotes a putative DNA intercalating AMP-PNP molecule shown throughout in red. The potential weakly interacting amino acid K313 is greyed. b, c Front and top views of the DNA–protein interaction, with the uninvolved N-proximal GHKL ParE region greyed out. Arrows from/to the zoom-in d, e showing the top (*) and bottom (**) DNA strand interaction sites indicate the direction of views displayed in d, e. Residues targeted for mutagenesis are underscored

Fig. 3

X-ray crystal structure of ParE44 with side-bound 5′-GCGCGC-3′ duplex DNA and AMP-PNP. a, b, c Front, side and bottom views of the complex and d residues that interact with the DNA and were targeted for mutagenesis (underscored). The ParE44 protein forms a closed dimer with two molecules of DNA bound as a regular B-form helix but on the outside of the protein rather than in the hole. N-terminal ParE domains are shown in cartoon mode in light blue and cyan, bound AMP-PNP molecules are in red and DNA is in mauve/yellow for backbone/bases, respectively

Fig. 4

Comparison of hole-bound and side-bound ParE44-DNA structures. Composite omit electron density maps (2F_obs−F_calc) were contoured around a the entire molecular structure, b the bound DNA and c the bound AMP-PNP in the ATP binding site. Structures on the left-hand side represent the complex with the DNA bound in the hole formed by the ParE44 dimer. Structures on the right-hand side represent the side-bound DNA complex. In a the electron density map is contoured at 1.5 and 1.0 σ levels for protein and DNA in blue and yellow colours, respectively. In b the electron density map is contoured at 1.0 σ level and is in light blue. In c the electron density map is contoured at the 1.5 σ level only and is in blue. In a the protein and DNA are displayed in cartoon mode in light blue/cyan and red, respectively. In b the DNA backbone is in pink with phosphates indicated by spheres, the DNA bases are in yellow and the intercalated molecule X is in red. In c the bound AMP-PNP molecule and the coordinated magnesium are shown in yellow, while the protein residues are in light blue/cyan. Magnesium is shown as a sphere, side chains and AMP-PNP are shown using stick representation and the protein backbone is displayed in cartoon mode

Fig. 5

Mutations of ParE cavity residues affect DNA strand passage by topo IV. Complexes of topo IV reconstituted with ParE bearing cavity mutations at residues K291, R321, K346 or R353 exhibit greatly reduced DNA decatenation (a) and DNA relaxation (b) activities. a A representative decatenation experiment. For each panel, kinetoplast DNA was incubated at 37 °C for 1 h in decatenation buffer (40 mM Tris-HCl (pH 7.5), 6 mM MgCl₂, 10 mM DTT, 200 mM potassium glutamate, 50 µg ml⁻¹ BSA) with 1 mM ATP and S. pneumoniae topoisomerase IV reconstituted from a fixed amount of ParC (25 ng) (and in experiments not shown up to 2000 ng) and titrated with either wild-type or mutant ParE (as indicated) at 100, 20, 10, 5 and 2.5 ng (lanes 1–5, respectively in each panel). Reactions were terminated and the DNA products were separated and visualized by electrophoresis in 1% agarose gels. Lane C, no enzyme addition. kDNA, kinetoplast DNA; Mini denotes minicircle DNA released by topo IV. Intermediate bands are partially unlinked DNA species. b Relaxation assays for topo IV were carried as described for kDNA decatenation except supercoiled plasmid pBR322 DNA (400 ng) was used as substrate with the same level of ParC (25 ng) and wild-type or mutant ParE at 100, 40, 20, 10 and 5 ng (each panel lanes 1–5, respectively). DNA products were analysed on 1% agarose gels. Lane C, no enzyme addition. R and S, relaxed and supercoiled DNA, respectively

Fig. 6

Measuring the ATPase activity of ParE proteins. a Comparison of basal ATPase rates for wild-type (WT) and representative ParE cavity mutants and stimulation by ParC and DNA. A coupled enzyme assay was used wherein ATPase activity was measured through the change in NADH absorbance at 340 nm^,. The reaction mix contained topo IV decatenation buffer, 2 mM ATP and 100 nM ParE and was incubated at 37 °C in the presence or absence of 130 nM ParC and 10 µg pBR322 (final volume of 500 µl). b Bar chart showing the ATPase activity of WT ParE and various single mutants of ParE observed in the absence and presence of DNA and ParC (to form topo IV). Reactions were carried out in triplicate on separate days. Error bars show the standard error. Inclusion of DNA or ParC alone did not stimulate ParE ATPase activity. The ParE D269V mutant showed strong basal ATPase activity that was greatly enhanced by inclusion of ParC and DNA

Fig. 7

Short DNA duplexes stimulate ATP hydrolysis by topoisomerase IV. Using the conditions described in the legend to Fig. 6, topo IV was incubated with 2 mM ATP in the absence and presence of a variety of short annealed oligonucleotide duplexes (10 µg) and ATP hydrolysis was followed by a coupled assay. The results shown are the average of three independent measurements. In the absence of DNA, the basal ATPase activity of topo IV was 12.4 nM s⁻¹. Asymmetric E20 and V20 are 20-mers of the E and V sites that bind as G (gate DNA) segments and are strongly cleaved by topo IV^, and were produced in duplex form by annealing complementary strands as carried out for 5′-GC(TA)₄TCG-3′

Fig. 8

Overlap between the dimerized yeast topoisomerase II ATPase domain (Protein Data Bank 4GFH]) (shown in yellow) and S. pneumoniae topoisomerase IV ATPase domain (through-hole DNA bound complex; shown in blue). AMP-PNP molecules are shown in red, and DNA is shown in pink/yellow for backbone/bases, respectively. The additional eukaryotic topo II-specific loop is shown in red. The structure indicates a potential clash between the bound DNA and the 22-amino acid long insertion loop in the eukaryotic topoisomerase II structure (Supplementary Fig. 8)