Nosocomial infections that result in the formation of biofilms on the surfaces of biomedical implants are a leading cause of sepsis and are often associated with colonization of the implants byStaphylococcus epidermidis. Biofilm formation is thought to require two sequential steps: adhesion of cells to a solid substrate followed by cell-cell adhesion, creating multiple layers of cells. Intercellular adhesion requires the polysaccharide intercellular adhesin (PIA), which is composed of linear β-1,6-linked glucosaminylglycans and can be synthesized in vitro from UDP-N-acetylglucosamine by products of the intercellular adhesion (ica) locus. We have investigated a variety ofStaphylococcus aureus strains and find that all strains tested contain the ica locus and that several can form biofilms in vitro. Sequence comparison with the S. epidermidis ica genes revealed 59 to 78% amino acid identity. Deletion of the ica locus results in a loss of the ability to form biofilms, produce PIA, or mediateN-acetylglucosaminyltransferase activity in vitro. Cross-species hybridization experiments revealed the presence oficaA in several other Staphylococcus species, suggesting that cell-cell adhesion and the potential to form biofilms is conserved within this genus.

Triclosan is used widely as an antibacterial agent in dermatological products, mouthwashes, and toothpastes. Recent studies imply that antibacterial activity results from binding to enoyl (acyl carrier protein) reductase (EACPR, EC 1.3.1.9). We first recognized the ability of triclosan to inhibit EACPR from Escherichia coli in a high throughput screen where the enzyme and test compound were preincubated with NAD(+), which is a product of the reaction. The concentration of triclosan required for 50% inhibition approximates to 50% of the enzyme concentration, indicating that the free compound is depleted by binding to EACPR. With no preincubation or added NAD(+), the degree of inhibition by 150 nM triclosan increases gradually over several minutes. The onset of inhibition is more rapid when NAD(+) is added. Gel filtration and mass spectrometry show that inhibition by triclosan is reversible. Steady-state assays were designed to avoid depletion of free inhibitor and changes in the degree of inhibition. The results suggest that triclosan binds to E-NAD(+) complex, with a dissociation constant around 20-40 pM. Triclosan follows competitive kinetics with respect to NADH, giving an inhibition constant of 38 pM at zero NADH and saturating NAD(+). Uncompetitive kinetics are observed when NAD(+) is varied, giving an inhibition constant of 22 pM at saturating NAD(+). By following regain of catalytic activity after dilution of EACPR that had been preincubated with triclosan and NAD(+), the rate constant for dissociation of the inhibitor (k(off)) is measured as 1.9 x 10(-4) s(-1). The association rate constant (k(on)) is estimated as 2.6 x 10(7) s(-1) M(-1) by monitoring the onset of inhibition during assays started by addition of EACPR. As expected, the ratio k(off)/k(on) = 7.1 pM is similar to the inhibition constants from the steady-state studies. The crystal structure of E. coli EACPR in a complex with coenzyme and triclosan has been determined at 1.9 A resolution, showing that this compound binds in a similar site to the diazaborine inhibitors. The high affinity of triclosan appears to be due to structural similarity to a tightly bound intermediate in catalysis.

[3H]tobramycin bound to sodium alginate and to exopolysaccharide prepared from two mucoid strains of Pseudomonas aeruginosa. Binding to sodium alginate was similar to binding to exopolysaccharide, both in the dependence on tobramycin concentration and in the maximum binding observed at saturation. Incorporation of sodium alginate into agar plates reduced the zone sizes of growth inhibition caused by tobramycin. The reductions in zone sizes were quantitatively accounted for by the binding of tobramycin to sodium alginate during diffusion of the antibiotic away from the well in which it had been placed at the start of the experiment. However, the binding of tobramycin to the exopolysaccharide of P. aeruginosa, and the resulting inhibition of diffusion of the antibiotic, did not significantly increase the penetration time of a spherical microcolony with a radius of 125 ,um, such as might be found in the respiratory tract of a patient with cystic fibrosis (from a 90% penetration time of 12 s in the absence of exopolysaccharide to one of 35 s with an exopolysaccharide concentration of 1.0% [wt/voll).The question of whether bacterial exopolysaccharides reduce the penetration of antibiotics to their target sites (5, 22) is an important one in antibacterial chemotherapy. This is because exopolysaccharide-producing bacteria existing as biofilms are less susceptible to antibiotics than are freely suspended bacteria (6, 17), and mucoid Pseudomonas aeruginosa apparently forms microcolonies (12) when causing respiratory tract infections that are refractory to chemotherapy in patients with cystic fibrosis. Moreover, in a recent review (4), the exopolysaccharide material of the biofilm was specifically postulated to exclude antibacterial substances.Inhibition of the diffusion of aminoglycoside antibiotics occurs in the presence of alginate (21), a polyanionic polysaccharide similar in structure to the exopolysaccharide of mucoid P. aeruginosa (13), or in the presence of the exopolysaccharide from P. aeruginosa (21). A likely reason for the reduced rate of diffusion of aminoglycosides within the anionic polysaccharide matrix is that any antibiotic-binding sites act as sinks, thereby reducing the free concentration of antibiotic, which is effectively the driving force of diffusion. In apparent conflict with this suggestion, the binding of tobramycin and streptomycin to alginate has been reported (23) as not being detectable in a physiological buffer containing 0.10 M NaCl.We quantitatively investigated the inhibition of diffusion and assessed the binding of tobramycin to alginate and Pseudomonas exopolysaccharide using radiolabeled tobramycin. Here we report that in a physiological buffer solution, tobramycin binds to alginate and to exopolysaccharides isolated from two mucoid strains of P. aeruginosa and that the binding to alginate quantitatively accounts for the inhibition of diffusion reported previously (21). However, binding and consequent inhibition of diffusion cannot account for antibiotic resistance within microcolonie...

No abstract

~~ ~ ~Cells of mucoid and non-mucoid Pseudomonas aeruginosa in colonies were at least onethousandfold less sensitive to the antibiotics tobramycin or cefsulodin than were cells of the same bacteria in dispersed suspension. We did not detect any difference between the mucoid form and the non-mucoid form in the antibiotic sensitivity of colonies, from which we infer that the exopolysaccharide of the mucoid form does not contribute to colony-resistance by forming a barrier to antibiotic diffusion. Mathematical models were constructed in order to estimate timecourses of penetration of tobramycin and cefsulodin into biofilms and microcolonies of mucoid and non-mucoid P. aeruginosa. For tobramycin penetration, adsorption of antibiotic to the exopolysaccharide of the glycocalyx and antibiotic uptake by cells were taken into account in the calculations. The longest time-period for the concentration of tobramycin at the base of a biofilm 100 prn deep to rise to 90% of the concentration outside the biofilm was predicted to be 2.4 h. For cefsulodin penetration, irreversible hydrolysis catalysed by p-lactamase was taken into account, using P-lactamase levels taken from the literature. The calculations predicted that the cefsulodin concentration at the base of a biofilm 100pm deep would rise to 90% of the external concentration in 29 s when the /I-lactamase was synthesized at the basal level. For a similar biofilm of bacteria synthesizing enhanced levels of fl-lactamase ('derepressed'), the concentration of cefsulodin at the base was calculated to rise to 41% of the external concentration in about 50s and then remain at that level. This was despite the fact that cefsulodin is a poor substrate for this P-lactamase.