Within 36 to 48 h, eyes injected with E. faecalis OG1RF showed marked vitreal polymorphonuclear infiltrate, cystoid changes in the ganglionic cell layer, decreased nuclear density of the inner and outer nuclear layers, mild subretinal polymorphonuclear infiltrate, and overall loss of structural integrity. In contrast, even after 48 h, eyes injected with the ΔfsrB strain showed only mild vitreal polymorphonuclear infiltrate, preserved structure of all retinal layers, and no subretinal inflammatory infiltrate (representative slides are shown in Fig. ). The histopathologic data from the group of rabbits infected with the complemented mutant were identical to those noted in the group injected with the wild type. When loss of B-wave response was 100% (for the fsrB deletion mutant, this time point was after the time points noted on Fig. ), histologic examination revealed the same degree of destruction in all three groups.
FIG. 2. Thin-section histopathology (representative slides; H-E stain). (A) Infection with E. faecalis OG1RF after 48 h, showing marked vitreal polymorphonuclear infiltrate, cystoid changes in the ganglionic cell layer, decreased nuclear density of inner and (more ...)
Gram-positive bacteria often use small peptide “pheromones” as cell-to-cell communication signals to mediate quorum sensing (4
). These communication signals are secreted from growing cells and accumulate outside the growing cells. When the concentration of the signal compound reaches a threshold level, a bacterial sensor is switched on and the expression of certain genes is activated. As a consequence of this mechanism, species- or strain-specific group behavior is controlled in response to a certain cell density (7
). The E. faecalis fsr
regulatory system appears to be functionally as well as structurally related to the staphylococcal agr
locus in that it regulates two known proteases and possibly other factors associated with virulence. The S. aureus agr
regulates expression of at least 19 exoproteins that are potentially important in the pathogenesis of endophthalmitis (1
). In the rabbit endophthalmitis model, a strain of S. aureus
defective in expression of the global regulatory locus agr
consistently resulted in a slower loss of B-wave response than in the wild-type strain (1
). The one known enterococcal toxin, the cytolysin, also mediates the precipitous loss of organ function in the endophthalmitis model (13
) and renders this infection refractory to existing antibiotic or antiinflammatory treatment (12
). The cytolysin was recently shown to be regulated by a novel autoregulated quorum system (7
) distinct from the fsr
system. The present study demonstrates that fsr
regulates additional virulence traits that contribute to the pathogenesis of enterococcal infection in the eye.
FsrB is predicted to be a membrane protein with multiple transmembrane segments (similar to AgrB in staphylococci) and also shows partial similarity to transporter proteins, such as the putative glutamate transporter of Borrelia burgdorferi
(23% identity over 101 amino acid residues) and the integral membrane component of an ABC transporter protein in Escherichia coli
(32% identity over 75 amino acid residues) (17
). FsrB may control the production of gelatinase in response to gelatinase biosynthesis-activation pheromone accumulation outside the cell.
Enterococci are an important cause of postoperative endophthalmitis, associated with significant loss of vision, and enterococcal infections have the second worst visual sequelae among all causes of endophthalmitis (6
). Study of enterococcal pathogenesis in a rabbit model of endophthalmitis may be particularly relevant for the evaluation of a quorum-sensing system such as Fsr. A very low number of E. faecalis
CFU can be used to establish infection (for example, in this study each eye was injected with ca. 102
CFU). Thus, in the rabbit endophthalmitis model, the quorum develops under in vivo conditions, as opposed to the high inocula used in most other models where quorum concentrations are present immediately upon injection. In addition, the endophthalmitis model provides an exquisitely sensitive infection system in which organ function can be directly assessed (by ERG) and the effects of the infection can be monitored as the enterococcal quorum develops. Previous reports have outlined the importance of the fsrB
gene in enterococcal virulence in a mouse peritonitis model as well as in a nonmammalian model of infection (5
). Herein, we show that deletion of the fsrB
gene significantly decreases virulence in an endophthalmitis model as well. As in every other animal model, results from this study cannot necessarily be extrapolated to the variety of human infections caused by this important pathogen. The unique physiology of the intraocular space, the tissue specificity of bacterial adherence and possible virulence factors, and the relative absence of humoral factors and mechanical clearance mechanisms should be taken into consideration. However, previous data from the mouse peritonitis (19
) and C. elegans
) models of enterococcal infection, in conjunction with the present data in the rabbit endophthalmitis model, suggest that fsrB
is important in E. faecalis
pathogenicity across a broad range of both local and systemic infections. This study, therefore, provides additional support for the conclusion that many features of bacterial virulence are conserved across a variety of different model systems.
In conclusion, the experimental endophthalmitis model is a sensitive model for assessment of E. faecalis quorum sensing. The Fsr system plays a significant role in the virulence of E. faecalis in this and other disease models, and it may provide an attractive target for development of new antimicrobial agents.