Assessment of the role of gamma-toxin in experimental endophthalmitis using a hlg-deficient mutant of Staphylococcus aureus.

Purified gamma-toxin is known to have a proinflammatory effect in the rabbit vitreous humor. To assess the biological role of the gamma-toxin, when expressed in vivo by Staphylococcus aureus strain Newman, the vitreous humor of rabbit eye was used as an infection model. A gamma-toxin-deficient mutant of strain Newman was constructed by allelic replacement. S. aureus Newman wild-type, its hlg-deficient derivative strain (N65) and the strain N65 complemented with the wild-type hlg+ gene were injected into the vitreous humor of rabbit eye. All three strains produced a strong proinflammatory effect in the eye conjunctiva, posterior and anterior chambers, suggesting a role for another unidentified proinflammatory component of strain Newman distinct from the gamma-toxin. These components are not the leucocidin of Panton-Valentine, beta-toxin or alpha-toxin which are not produced by this strain. Only the hlg-deficient mutant lacked the ability to cause inflammation in the eyelid, whereas the two Hlg-producing strains gave strong inflammation. These data suggest that in vivo, strain Newman produces as yet unidentified proinflammatory molecules and that the in vivo-produced HlgA, HlgB and HlgC molecules expressed by the gamma-toxin locus, contribute in part to the inflammatory process observed in vivo in the rabbit eye.

A predicted beta-sheet from class S components of staphylococcal gamma-hemolysin is essential for the secondary interaction of the class F component.

Site-directed mutagenesis was performed on genes encoding HlgA and HlgC, two of the three proteins expressed from the staphylococcal y-hemolysin locus, which originate two pore-forming toxins (HlgA + HlgB, HlgC + HlgB). As related proteins, HlgA and HlgC were found to bind first to cell membranes. Amino acid substitutions concerned residues that would predictably disrupt a 13 amino acid conserved beta-sheet of the Chou and Fasman secondary structure prediction. The mutation of a threonin into an aspartic acid residue from HlgA (T28D) and from HlgC (T30D) that would break this predicted N-terminal structure lowered dramatically the biological activities on purely lipidic vesicles, erythrocytes and polymorphonuclear cells. The change in secondary structure was confirmed by Fourier Transformed Infrared spectroscopy. The binding of mutated and native proteins at the same kind of sites onto polymorphonuclear cells was evidenced with flow cytometry and fluorescein-labelled anti-class S antibodies or wild type HlgA or HlgC. However, the subsequent binding of fluorescein-labelled HlgB to membrane-bound mutated HlgA or HlgC complexes was inhibited. In conclusion, the first binding of class S components is essential for the subsequent binding of class F components, and a predicted beta-sheet seems to be at least one of the functional domains involved.

Panton-Valentine leucocidin and gamma-hemolysin from Staphylococcus aureus ATCC 49775 are encoded by distinct genetic loci and have different biological activities.

Staphylococcus aureus ATCC 49775 produces three proteins recognized by affinity-purified antibodies against the S component of Panton-Valentine leucocidin (LukS-PV) and two proteins recognized by affinity-purified antibodies against the F component of this toxin (LukF-PV). Purification of these proteins and cloning of the corresponding genes provided evidence for the presence of two loci. The first one, encoding Panton-Valentine leucocidin, consisted of two cotranscribed open reading frames, lukS-PV and lukF-PV, coding the class S and the class F components, respectively. The second one coded for a gamma-hemolysin and consisted of two transcription units, the first one encoding an HlgA-like protein, a class S component, and the second one encoding two cotranscribed open reading frames identical to HlgC and HlgB, class S and class F components, respectively, from gamma-hemolysin from the reference strain Smith 5R. It appears that the Panton-Valentine leucocidin from S. aureus ATCC 49775 (V8 strain) should not be confused with leucocidin from ATCC 27733 (another isolate of V8 strain), which had 95% identity with HlgC and HlgB from gamma-hemolysin. The cosecretion of these five proteins led to six possible synergistic combinations between F and S components. Two of these combinations (LukS-PV-LukF-PV and HlgA-LukF-PV) had dermonecrotic activity on rabbit skin, and all six were leukocytolytic on glass-adsorbed leukocytes. Only three were hemolytic on rabbit erythrocytes, the two gamma-hemolysin combinations and the combination LukF-PV-HlgA.

Sequencing of leucocidin R from Staphylococcus aureus P83 suggests that staphylococcal leucocidins and gamma-hemolysin are members of a single, two-component family of toxins.

A 2,813-bp HincII-ClaI DNA fragment encodes the two S and F components (LukS-R and LukF-R) of leucocidin R (Luk-R) which are secreted by Staphylococcus aureus P83. The two genes (lukS-R and lukF-R) belong to a single operon. Two peptidic sequences were deduced: LukS-R is a 35,721-Da polypeptide of 315 amino acids, including a signal sequence of 29 residues, and LukF-R is a 36,838-Da polypeptide of 325 amino acids, including a signal sequence of 25 residues. LukS-R and LukF-R were expressed in Escherichia coli and purified from the periplasmic space. Luk-R exerts biological activities on polymorphonuclear cells and on erythrocytes from various animals. Comparison of the amino acid sequence of LukF-R with that of the B component of gamma-hemolysin (HlgB), those of the F and S components of another recently sequenced staphylococcal leucocidin, and those of a few peptides of the F component from Panton-Valentine leucocidin suggests that all four toxins belong to a single, two-component family of toxins.