Role of the luxS quorum-sensing system in biofilm formation and virulence of Staphylococcus epidermidis

Abstract

Nosocomial infections caused by Staphylococcus epidermidis are characterized by biofilm formation on implanted medical devices. Quorum-sensing regulation plays a major role in the biofilm development of many bacterial pathogens. Here, we describe luxS, a quorum-sensing system in staphylococci that has a significant impact on biofilm development and virulence. We constructed an isogenic DeltaluxS mutant strain of a biofilm-forming clinical isolate of S. epidermidis and demonstrated that luxS signaling is functional in S. epidermidis. The mutant strain showed increased biofilm formation in vitro and enhanced virulence in a rat model of biofilm-associated infection. Genetic complementation and addition of autoinducer 2-containing culture filtrate restored the wild-type phenotype, demonstrating that luxS repressed biofilm formation through a cell-cell signaling mechanism based on autoinducer 2 secretion. Enhanced production of the biofilm exopolysaccharide polysaccharide intercellular adhesin in the mutant strain is presumably the major cause of the observed phenotype. The agr quorum-sensing system has previously been shown to impact biofilm development and biofilm-associated infection in a way similar to that of luxS, although by regulation of different factors. Our study indicates a general scheme of quorum-sensing regulation of biofilm development in staphylococci, which contrasts with that observed in many other bacterial pathogens.

FIG. 1.

Identification of the luxS gene of S. epidermidis. Shown are the luxS gene and its surrounding region according to genetic information (58).

FIG. 2.

Complementation of the AI-2 synthesis deficiency of E. coli DH5α by luxS from S. epidermidis. The capacity to produce AI-2 was examined by an AI-2 reporter assay as described previously (47, 48). V. harveyi BB170 (AI-1 sensor⁻, AI-2 sensor⁺) served as a positive control. Cell-free medium was prepared from the strains at logarithmic growth phase, when their light production reached maximum. Luminescence was expressed as fold induction relative to the background values and normalized to cell OD of 1 at 600 nm. Data were obtained from four independent experiments.

FIG. 3.

Detection of AI-2 activity throughout the growth of S. epidermidis 1457. Cells were grown overnight and diluted 1:100 into fresh TSB medium. The absorbance (OD) at 600 nm was then examined hourly by spectrophotometry. Every 2 h, conditioned cell-free supernatants were examined for the capacity to induce light production in V. harveyi BB170. Data were obtained from four independent experiments.

FIG. 4.

AI-2 activity of S. epidermidis 1457, its isogenic ΔluxS mutant strain and/or the mutant strain with empty vector complemented with luxS. Cell-free medium was prepared from those strains during the log phase. All experiments were performed in triplicate. WT, S. epidermidis 1457 wild-type strain; M, S. epidermidis 1457 isogenic ΔluxS mutant strain; M(pBT1), S. epidermidis ΔluxS mutant strain containing vector pBT1; M(pBT1-luxS), luxS-complemented S. epidermidis mutant strain.

FIG. 5.

Semiquantitative biofilm assay. Each experiment was repeated eight times. A, Biofilm formation of S. epidermidis 1457, its isogenic ΔluxS mutant, the mutant with the empty vector, and the luxS-complemented mutant at 24 h. B, Biofilm formation of S. epidermidis 1457, its isogenic ΔluxS mutant, and the mutants with exogenous AI-2 prepared from wild-type and complemented mutant strains at 24 h using mutant with CM from mutant strain transformed with empty vector as control.

FIG. 6.

Scanning electron micrographs of S. epidermidis 1457 and its ΔluxS mutant. Twenty-four-hour biofilms were grown on hydroxyapatite disks that were deposited in 24-well cell culture clusters in TSB with 0.5% (wt/vol) glucose. Images shown are at a magnification of ×3,000.

FIG. 7.

A, TaqMan analysis of ica gene expression. TaqMan analysis was performed using an icaC probe, and cells were grown to mid-exponential phase. The levels of ica mRNA had been normalized to the level of 16S rRNA mRNA. Assays were performed in triplicate. Values given are the means plus or minus standard errors of the means. B, PIA production. PIA samples were isolated from the surface of cells grown to stationary growth phase by boiling with 0.5 mol/liter EDTA. PIA production was determined by immuno-dot blot analysis using anti-PIA antisera and quantified by photodigital analysis. Values are the means plus or minus standard errors of the means from three independent experiments with three measurements (nine dots per strain). C, Representative immuno-dot blot. First row, wild-type strain; second row, isogenic ΔluxS mutant strain; third row, empty vector control strain; fourth row, luxS-complemented strain.