Nonoptimal DNA topoisomerases allow maintenance of supercoiling levels and improve fitness of Streptococcus pneumoniae

Abstract

Fluoroquinolones, which target gyrase and topoisomerase IV, are used for treating Streptococcus pneumoniae infections. Fluoroquinolone resistance in this bacterium can arise via point mutation or interspecific recombination with genetically related streptococci. Our previous study on the fitness cost of resistance mutations and recombinant topoisomerases identified GyrAE85K as a high-cost change. However, this cost was compensated for by the presence of a recombinant topoisomerase IV (parC and parE recombinant genes) in strain T14. In this study, we purified wild-type and mutant topoisomerases and compared their enzymatic activities. In strain T14, both gyrase carrying GyrAE85K and recombinant topoisomerase IV showed lower activities (from 2.0- to 3.7-fold) than the wild-type enzymes. These variations of in vitro activity corresponded to changes of in vivo supercoiling levels that were analyzed by two-dimensional electrophoresis of an internal plasmid. Strains carrying GyrAE85K and nonrecombinant topoisomerases had lower (11.1% to 14.3%) supercoiling density (σ) values than the wild type. Those carrying GyrAE85K and recombinant topoisomerases showed either partial or total supercoiling level restoration, with σ values being 7.9% (recombinant ParC) and 1.6% (recombinant ParC and recombinant ParE) lower than those for the wild type. These data suggested that changes acquired by interspecific recombination might be selected because they reduce the fitness cost associated with fluoroquinolone resistance mutations. An increase in the incidence of fluoroquinolone resistance, even in the absence of further antibiotic exposure, is envisaged.

FIG. 1.

Characteristics of the strains used and of their gyrase and topo IV enzymes. (A) Amino acids that change in strains Tr7, T1, T4, T9, and T14 with respect to the strain R6 sequence are indicated, and those involved in fluoroquinolone resistance are showed in boldface and underlined. For each strain are shown their CIP MIC, mean relative fitness, and 95% CI calculated in competition experiments with R6 (4). Fitness cost categorized as high is shown in boldface and underlined. (B) Purified topoisomerase subunits revealed by staining with Coomassie blue. His-tagged proteins were overexpressed in E. coli, purified by nickel resin chromatography, and examined on an SDS-12% polyacrylamide gel. Lane Mm, molecular mass marker (kDa). Approximately 0.5 μg of each protein was loaded per well. (C) Supercoiling activity of gyrase subunits. Reaction mixtures contained no enzyme (lane 0), 142.5 fmol of GyrA (lane A) or GyrAE85K (lane AE85K), 541.5 fmol of GyrB (lane B), or reconstituted wild-type (lane A+B) and mutant (lane E85K+B) gyrase. N, nicked pBR322; R, relaxed pBR322; S, negatively supercoiled pBR322. (D) Decatenation activity of topo IV. Standard reaction mixtures contained either no enzyme (lane 0), 16 fmol of ParC (lane C), 61 fmol of ParE (lane E), or reconstituted topo IV (lane C+E). m, monomers. (E) Relaxation activity of topo IV. Standard reaction mixtures contained either no enzyme (lane 0), 2.5 pmol of ParC (lane C), 9.3 pmol of ParE (lane E), or reconstituted topo IV (lane C+E). (F) Stimulation of gyrase-mediated cleavage activity by CIP. Reaction mixtures contained either no enzyme (lanes 0), 1.1 pmol of GyrA (lane A), 5.9 pmol of GyrB (lane B), or reconstituted gyrase (lanes A+B). (G) Stimulation of topo IV-mediated cleavage activity by CIP. Reaction mixtures contained either no enzyme (lanes 0), 2.3 pmol of ParC (lane C), 8.7 pmol of ParE (lane E), or reconstituted topo IV (lanes C+E). (F and G) The different forms of plasmid pBR322 are indicated: OC, open circle; L, linear; S, negatively supercoiled.

FIG. 2.

The supercoiling decatenation and relaxation activities of mutant topoisomerases are lower than those of the wild-type enzyme. (A) Gyrase supercoiling. Subunit wt GyrA or GyrAE85K (lane AE85K) was reconstituted with wt GyrB in a 1:3.8 proportion and used to supercoil relaxed plasmid pBR322 (0.4 μg) with various amounts of reconstituted gyrase (31 to 500 fmol GyrA). Lane 0, relaxed pBR322 substrate. (B) Decatenation activity of topo IV (lanes E+C) and its ParCS79F mutant (lanes E+CS79F) reconstituted with wt ParE, recParC (lanes E+C*S79F), or recParE plus recParC (lanes E*+C*S79F) in a 3.8:1 proportion (12.5 to 200 fmol ParC). Enzymes were incubated with kDNA (0.4 μg). m, monomers. (C) Relaxation activity of topo IV and its ParCS79F mutant reconstituted with wt ParE, recParC, or recParE plus recParC in a 3.8:1 proportion (0.3 to 2.5 pmol ParC). Enzymes were incubated with 50 ng of supercoiled plasmid pBR322. N, nicked pBR322; R, relaxed pBR322; S, negatively supercoiled pBR322.

FIG. 3.

Mutant topoisomerases showed lower ciprofloxacin-promoted DNA cleavage activities than wild-type enzyme. (A) CIP-promoted DNA cleavage of gyrase. Plasmid pBR322 DNA (0.4 μg) was incubated with different amounts of reconstituted gyrase (lanes A) and its AE85K mutant (lanes AE85K), using 71 to 1,140 fmol GyrA. (B) CIP-promoted DNA cleavage with different amounts of reconstituted topo IV (lanes E+C) and its ParCS79F mutant (lanes E+CS79F), using 125 to 1,000 fmol ParC. (C) CIP-promoted DNA cleavage by topo IV with recParC (lanes E+C*S79F) or recParE plus recParC (lanes E*+C*S79F).

FIG. 4.

Supercoiling levels correlate with bacterial fitness. (A) Distribution of pLS1 topoisomers of the indicated isogenic strains after two-dimensional agarose gel electrophoresis run in the presence of 1 and 2 μg/ml chloroquine in the first and second dimensions, respectively. An empty arrowhead indicates the most abundant topoisomer. σ values are indicated below the image of one characteristic gel of each strain as the average ± standard deviation, n (number of determinations). (B) Correlation between supercoiling density and fitness at 12 h of growth in competition with R6. A black square indicates those strains carrying the GyrAE85K change.