Quinolone resistance mutations in Streptococcus pneumoniae GyrA and ParC proteins: mechanistic insights into quinolone action from enzymatic analysis, intracellular levels, and phenotypes of wild-type and mutant proteins

Abstract

Mutations in DNA gyrase and/or topoisomerase IV genes are frequently encountered in quinolone-resistant mutants of Streptococcus pneumoniae. To investigate the mechanism of their effects at the molecular and cellular levels, we have used an Escherichia coli system to overexpress S. pneumoniae gyrase gyrA and topoisomerase IV parC genes encoding respective Ser81Phe and Ser79Phe mutations, two changes widely associated with quinolone resistance. Nickel chelate chromatography yielded highly purified mutant His-tagged proteins that, in the presence of the corresponding GyrB and ParE subunits, reconstituted gyrase and topoisomerase IV complexes with wild-type specific activities. In enzyme inhibition or DNA cleavage assays, these mutant enzyme complexes were at least 8- to 16-fold less responsive to both sparfloxacin and ciprofloxacin. The ciprofloxacin-resistant (Cip(r)) phenotype was silent in a sparfloxacin-resistant (Spx(r)) S. pneumoniae gyrA (Ser81Phe) strain expressing a demonstrably wild-type topoisomerase IV, whereas Spx(r) was silent in a Cip(r) parC (Ser79Phe) strain. These epistatic effects provide strong support for a model in which quinolones kill S. pneumoniae by acting not as enzyme inhibitors but as cellular poisons, with sparfloxacin killing preferentially through gyrase and ciprofloxacin through topoisomerase IV. By immunoblotting using subunit-specific antisera, intracellular GyrA/GyrB levels were a modest threefold higher than those of ParC/ParE, most likely insufficient to allow selective drug action by counterbalancing the 20- to 40-fold preference for cleavable-complex formation through topoisomerase IV observed in vitro. To reconcile these results, we suggest that drug-dependent differences in the efficiency by which ternary complexes are formed, processed, or repaired in S. pneumoniae may be key factors determining the killing pathway.

FIG. 1

SDS-PAGE analysis of highly purified S. pneumoniae wild-type (wt) GyrA, GyrA(Phe81), wild-type ParC, and ParC(Phe79) proteins. The His-tagged proteins were overexpressed in E. coli, purified by nickel resin chromatography, and examined on an SDS–7.5% polyacrylamide gel. Lane M, marker proteins (sizes in kilodaltons).

FIG. 2

DNA supercoiling activity of mutant S. pneumoniae DNA gyrase is highly refractory to inhibition by sparfloxacin (SPAR) (A) and by ciprofloxacin (CIP) (B). Relaxed pBR322 (0.4 μg) was incubated with gyrase activity (1 U) reconstituted from GyrB and either wild-type (wt) GyrA (left lanes) or GyrA(Phe81) (right lanes) in the presence of 1.4 mM ATP and the indicated amounts (in micromolar) of quinolones. Reactions were stopped, and the DNA products were separated by electrophoresis in 1% agarose. DNA was stained with ethidium bromide and photographed under UV illumination. Lanes a and b, supercoiled and relaxed pBR322 DNA, respectively. N, R, and S denote nicked, relaxed, and supercoiled DNA, respectively.

FIG. 3

S. pneumoniae topoisomerase IV containing the ParC(Phe79) subunit is resistant to inhibition by sparfloxacin (SPAR). Topoisomerase IV activity (1 U), reconstituted from recombinant ParE subunit and either wild-type (wt) ParC or ParC(Phe79) protein, was incubated with kDNA (0.4 μg) in the presence of 1.4 mM ATP and quinolones at the concentrations indicated. DNA was analyzed by agarose gel electrophoresis as described in Fig. 2. Lane a, kDNA. N and M indicate kinetoplast network DNA and released relaxed minicircles, respectively.

FIG. 4

Ser81Phe mutation in GyrA impairs DNA cleavage by gyrase promoted by sparfloxacin (SPAR) (A and C) or ciprofloxacin (B). Supercoiled pBR322 DNA (0.4 μg) was incubated with S. pneumoniae GyrB (1.7 μg) and either wild-type (wt) GyrA or GyrA(Phe81) (0.45 μg) in the absence (A and B) or presence of 1.4 mM ATP (C) and quinolones at the concentrations indicated. After addition of SDS and proteinase K, DNA samples were analyzed by electrophoresis in 1% agarose. Lanes a and b, supercoiled pBR322 and EcoRI-cut pBR322. N, L, and S indicate nicked, linear, and supercoiled DNA.

FIG. 5

DNA breakage by topoisomerase IV mediated by sparfloxacin (SPAR) (A) and by ciprofloxacin (B) is inhibited by the Phe79 mutation in ParC. DNA cleavage reactions were carried out as described in the legend to Fig. 4 using ParE (1.7 μg) and wild-type (wt) ParC or ParC(Phe79) (0.45 μg) but with a lower range of quinolone concentrations.

FIG. 6

Polyclonal antisera specific for the S. pneumoniae GyrA, ParC, GyrB, and ParE proteins. Equimolar amounts of highly purified recombinant GyrA (A), ParC (C), GyrB (B), and ParE (E) proteins were run on SDS–6% polyacrylamide gels which were either stained with Coomassie blue (panels 1 and 4) or electrotransferred to nitrocellulose filters (panels 2, 3, 5, and 6). The filters were probed with rabbit antisera made to the recombinant His-tagged GyrA (panel 2) and ParC (panel 3) or recombinant His-tagged GyrB (panel 5) or ParE subunit (panel 6). In each case, GyrA- and ParC- and GyrB- and ParE- antisera were prestripped of cross-reacting antibodies using the other homologous recombinant protein as an affinity ligand. The binding of the first antibody was visualized by using an alkaline-phosphatase-conjugated anti-rabbit IgG second antibody with colorimetric development.

FIG. 7

Quantitative immunoblotting of gyrase and topoisomerase IV subunits in protein lysates of S. pneumoniae and its quinolone-resistant mutants. (A) GyrA and GyrB are more abundant in cell extracts than ParC and ParE. Protein extracts (4 μg) from wild-type S. pneumoniae strains 7785 and R6 prepared by SDS lysis were run on four SDS–6% polyacrylamide gels alongside known amounts of recombinant GyrA, GyrB, ParC, or ParE protein used as the standard. Proteins were transferred to nitrocellulose filters and probed with specific antisera and a second antibody as described in the Fig. 6 legend. Due to the presence of histidine tags, the protein standards have a slightly lower mobility than the native proteins. (B) ParC and ParE proteins are detected using four- to eightfold higher amounts of extract. Protein extracts, prepared from strains 7785 and R6 and gyrA mutants 1S1 and 1S4, were loaded at 4 μg per lane for detection of GyrA and GyrB and at 16 μg and 32 μg per lane for quantitation of ParC and ParE, respectively.