Search

The adherence of Staphylococcus aureus to human endothelial cells (EC) is probably an important step in the pathogenesis of systemic staphylococcal infections. We examined the influence of type 5 capsular polysaccharide (CP5) production, the global regulator agr, and the bacterial growth phase on S. aureus adherence to EC. Whereas S. aureus Newman showed maximal adherence to EC in the logarithmic phase of growth, an isogenic agr mutant showed maximal adherence in the stationary growth phase. S. aureus adherence to EC and CP5 expression were negatively correlated: a mutation in the agr locus diminished CP5 production and led to increased adherence. Likewise, induction of CP5 expression by addition of NaCl to the growth medium resulted in reduced staphylococcal adherence to EC. S. aureus Newman cells that adhered to EC did not express CP5. A Newman cap5O mutant was acapsular and showed significantly greater adherence to EC than the parental strain did (P<0.005). Complementation of the cap5O mutation in trans restored CP5 expression and reduced EC adherence to a level similar to that of the parental strain. The enhanced adherence shown by the cap5O mutant was similar in magnitude to that of the agr mutant or the cap5O agr double mutant. Cells of the cap5O mutant and cap5O agr double mutant harvested from stationary-phase cultures adhered significantly better than did cells harvested in the exponential growth phase. These data are consistent with the postexponential and agr-independent expression by S. aureus of at least one putative EC adhesin, whose binding domain may be masked by CP5.

Bacteria possess a repertoire of distinct regulatory systems promoting survival in disparate environments. Under in vitro conditions it was demonstrated for the human pathogen Staphylococcus aureus that the expression of most virulence factors is coordinated by the global regulator agr. To monitor bacterial gene regulation in the host, we developed a method for direct transcript analysis from clinical specimens. Quantification of specific transcripts was performed by competitive reverse transcription-PCR, and results were normalized against the constitutively expressed gene for gyrase (gyr). Using sputum from cystic fibrosis (CF) patients infected with S. aureus we examined the transcription of the effector molecule RNAIII of agr, of spa (protein A), generally repressed by agr, and of hla (alpha-toxin), generally activated by agr. In the CF lung RNAIII was expressed poorly, indicating an inactive agr in vivo. Despite the low level of RNAIII expression, spa was detectable only in minute amounts and an irregular transcription of hla was observed in all sputum samples. After subculturing of patient strains agr-deficient isolates and isolates with unusual expression profiles, i.e., not consistent with those obtained from prototypic strains, were observed. In conclusion, the agr activity seems to be nonessential in CF, and from the described expression pattern of spa and hla, other regulatory circuits aside from agr are postulated in vivo.

We investigated the role of Pseudomonas aeruginosa exoenzymes in cystic fibrosis lung infection in the presence and absence of specific serum antibodies. In sputa of 21 cystic fibrosis patients, concentrations of P. aeruginosa proteases and exotoxin A were determined by sensitive radioimmunoassays. In all sputa, detection of exoenzymes was negative (less than or equal to 10 ng). Positive serum antibody titers to bacterial exoenzymes were found in the majority of patients. Purified immunoglobulin G (IgG) preparations from the sera of two patients revealing specific antibody titers to the bacterial proteases neutralized these enzymes at ratios of 1,000:1 to 5,600:1 (wt/wt). Above the neutralizing capacity of IgG, proteases caused cleavage of IgG; below that level, no enzymatic activity was observed. In vitro incubation of P. aeruginosa elastase, alkaline protease, or exotoxin A with elastase derived from polymorphonuclear leukocytes showed that polymorphonuclear leukocyte elastase: (i) was cleaved by bacterial elastase, (ii) was not inactivated by alkaline protease, and (iii) inactivated exotoxin A. The results suggest that soon after the onset of P. aeruginosa lung infection in cystic fibrosis patients, bacterial proteases, but not exotoxin A, become important virulence factors. The results also suggest that exoenzymes do not directly contribute to lung damage after immune response to bacterial antigens has begun.