Probing the differential interactions of quinazolinedione PD 0305970 and quinolones with gyrase and topoisomerase IV

Abstract

Quinazoline-2,4-diones, such as PD 0305970, are new DNA gyrase and topoisomerase IV (topo IV) inhibitors with potent activity against gram-positive pathogens, including quinolone-resistant isolates. The mechanistic basis of dione activity vis-à-vis quinolones is not understood. We present evidence for Streptococcus pneumoniae gyrase and topo IV that PD 0305970 and quinolones interact differently with the enzyme breakage-reunion and Toprim domains, DNA, and Mg2+-four components that are juxtaposed in the topoisomerase cleavage complex to effect DNA scission. First, PD 0305970 targets primarily gyrase in Streptococcus pneumoniae. However, unlike quinolones, which select predominantly for gyrA (or topo IV parC) mutations in the breakage-reunion domain, unusually the dione selected for novel mutants with alterations that map to a region of the Toprim domain of GyrB (R456H and E474A or E474D) or ParE (D435H and E475A). This "dione resistance-determining region" overlaps the GyrB quinolone resistance-determining region and the region that binds essential Mg2+ ions, each function involving conserved EGDSA and PLRGK motifs. Second, dione-resistant gyrase and topo IV were inhibited by ciprofloxacin, whereas quinolone-resistant enzymes (GyrA S81F and ParC S79F) remained susceptible to PD 0305970. Third, dione-promoted DNA cleavage by gyrase occurred at a distinct repertoire of sites, implying that structural differences with quinolones are sensed at the DNA level. Fourth, unlike the situation with quinolones, the Mg2+ chelator EDTA did not reverse dione-induced gyrase cleavage nor did the dione promote Mg2+-dependent DNA unwinding. It appears that PD 0305970 interacts uniquely to stabilize the cleavage complex of gyrase/topo IV perhaps via an altered orientation directed by the bidentate 3-amino-2,4-dione moiety.

FIG. 1.

Structures of PD 0305970 and ciprofloxacin.

FIG. 2.

Inhibitory activity of PD 0305970 against wild-type and mutant S. pneumoniae type II topoisomerases. Inhibition of DNA supercoiling by DNA gyrase. Relaxed pBR322 plasmid DNA (0.4 μg) was incubated with wild-type (wt) gyrase activity (1 U) or mutant enzyme (1 U) reconstituted with GyrB 474D or GyrB 474A (A) or GyrA S81F (B), in the presence of 1.4 mM ATP and the indicated concentrations (μM) of PD 0305970. Reactions were stopped, and DNA species were separated by electrophoresis in 1% agarose. After staining with ethidium bromide, DNA was photographed under UV light. Lanes a and b are supercoiled and relaxed pBR322 DNA, respectively. Sc, R, and N denote supercoiled, relaxed, and nicked DNA, respectively. Inhibition of DNA decatenation by S. pneumoniae topo IV. Kinetoplast DNA (0.4 μg) was incubated with wt topo IV (1U) or mutant enzyme (1 U) reconstituted with ParE 475A (C) or ParC 79F (D) and incubated with 1.4 mM ATP and PD 0305970 at the levels (μM) indicated. DNA was separated and analyzed as described above. Lane a, kDNA. M indicates released minicircles.

FIG. 3.

Capture of gyrase and topo IV cleavage complexes by PD 0305970. GyrB 474D and 474A (A) and ParE 475A (C) changes (but not GyrA 81F [B] and ParC 79F [D] mutations) inhibit DNA cleavage by S. pneumoniae gyrase and topo IV. (A and B) Supercoiled plasmid pBR322 DNA (0.4 μg) was incubated with either wt or mutant S. pneumoniae gyrase reconstituted from appropriate GyrB (1.7 μg) and GyrA subunits (0.45 μg) in the presence of PD 0305970 at the concentrations (μM) indicated. Cleavage was induced by addition of SDS and after proteinase K treatment; DNA was examined by electrophoresis in 1% agarose. Lanes a and b, supercoiled pBR322 and EcoRI-linearized pBR322. N, L, and Sc denote nicked, linear, and supercoiled DNA, respectively. (C and D) Cleavage by S. pneumoniae topo IV was conducted as for gyrase (A), but by using the respective ParE (1.7 μg) and ParC (0.45 μg) proteins. Samples were analyzed as described for DNA gyrase.

FIG. 4.

PD 0305970 induces site-specific DNA cleavage by gyrase. (A) PD 0305970 promotes gyrase-mediated double-stranded DNA cleavage. pBR322 DNA linearized with EcoRI was employed as a substrate in a cleavage assay with wild-type S. pneumoniae gyrase (as described in the legend for Fig. 3) by using PD 0305970 (QD) or ciprofloxacin (CIP) at the concentrations indicated. Lane M, DNA markers; molecular sizes in kilobase pairs (Kb) are indicated on the left. (B) Dione-induced DNA cleavage by gyrase analyzed at the sequence level. The 290-bp S fragment from the S. pneumoniae parE gene (27) was labeled with ³³P at the 5′ end of the top strand and incubated with S. pneumoniae gyrase and 1 mM ATP in the absence or presence of 100 μM gemifloxacin (Gemi) or 50 μM PD 0305970 (QD). After cleavage induction with SDS and proteinase K, DNA was precipitated with ethanol and run on a 6% denaturing polyacrylamide gel alongside ACGT chain termination sequencing products. Filled and open arrows indicate DNA fragments induced specifically by QD and gemifloxacin, respectively.

FIG. 5.

Reversal of drug-stabilized DNA cleavage in the absence (top) or presence (bottom) of ATP. (Top) Supercoiled plasmid pBR322 (0.45 μg) was incubated with S. pneumoniae GyrA (0.45 μg) and GyrB (1.7 μg) in a DNA cleavage assay in the presence of 2.5 μM PD 0305970 (QD), 20 μM gemifloxacin (Gemi), or 80 μM ciprofloxacin (CIP). After addition of SDS and proteinase K, DNA samples were purified and analyzed by electrophoresis in 1% agarose. Prior to cleavage induction, samples were incubated with 0.6 M NaCl for 15 min (lanes S), with 35 mM EDTA for 15 min (lanes E), or at 65°C for 15 min (lanes H). Lanes C, no pretreatment prior to cleavage; lane 0, cleavage with gyrase in the absence of drug; lanes a and b, supercoiled and linear pBR322, respectively. N, L, and Sc denote nicked, linear, and supercoiled DNA, respectively. (Bottom) Experiments were performed as described in the legend for the top panel, except 1 mM ATP was included and the drug concentrations were reduced to 0.83 μM PD 0305970, 6.7 μM gemifloxacin, and 26.7 μM ciprofloxacin.

FIG. 6.

Assay for DNA unwinding of closed circular DNA by PD 0305970 and ciprofloxacin. Relaxed pBR322 DNA (0.8 μg) was preincubated with calf thymus DNA topo I (10 U) at room temperature for 10 min in the presence of MgCl₂ at the concentrations (mM) indicated. After the addition or omission of PD 0305970 (QD) (A) or ciprofloxacin (CIP) (B), each at 160 μM, reaction mixtures were further incubated at 37°C for 1 h, and then the topo I was inactivated. Samples were extracted with phenol to remove bound ligand, and DNA products were separated and analyzed by electrophoresis in Tris-phosphate-EDTA buffer. Half of each DNA sample was analyzed in the absence of chloroquine (CQ−); the other half was subjected to electrophoresis in the presence of chloroquine (CQ+). In the absence of drugs, DNA samples are slightly positively supercoiled and are made more positively coiled by inclusion of chloroquine in the gel. Lanes a, untreated relaxed pBR322 DNA employed as substrate. QD and CIP were added from stock solutions in 0.1 M NaOH and in DMSO, and their concentrations were adjusted in each reaction to 10 mM and 2.5% (vol/vol), respectively.

FIG. 7.

Resistance to PD 0305970 maps to residues in the Toprim domain of the S. pneumoniae GyrB and ParE proteins. Filled arrows denote quinazolinedione resistance mutations described here; unfilled arrows denote those described by Huband et al. (23). Boldface type highlights conserved EGDSA, P(I/L)RGK, and (I/L)MTDXDXD motifs containing residues (underlined) implicated in Mg²⁺ ion binding. Residues 431 to 510 of the GyrB/ParE protein (40) are shown.