Skip to Content

Xiang Qin

Advisor: Barbara Murray, MD, George Weinstock, Ph.D.

Enterococci are one of the leading causes of nosocomial infections, and Enterococcus faecalis causes the majority of enterococcal infections. However, the mechanisms of enterococcal pathogenesis are still not yet understood. In our initial screening of E. faecalis strain OG1RF genomic libraries, autolysin and a homolog of a protein of Enterococcus faecium previously designated P54 were found to be two major antigens that reacted with human patient sera, and an antigen designated MH-1 antigen that reacted with serum from a endocarditis patient was also identified. To explore a possible role for these antigens in enterococcal infections, the genes encoding these three antigens were disrupted in Enterococcus faecalis OG1RF. Testing of these mutants in a mouse peritonitis model showed that the LD50s and time courses of killing of mice were similar to that of OG1RF, suggesting that these antigens are not important for the enterococcal infections, at least in this animal model. Other testing showed that the autolysin gene disruption mutant formed longer chains of cells compared to wild type OG1RF when grown in brain heart infusion (BHI) broth, showed a slower spontaneous autolysis rate than OG1RF, slightly less penicillin-induced lysis and less killing by penicillin than OG1RF, while the minimum inhibitory concentration (MIC) of the mutant was the same as OG1RF, suggesting that inactivation or alteration of this autolysin gene is not responsible for the tolerance of E. faecalis to penicillin observed in some clinical isolates or in vitro derived penicillin-tolerant strains.

To explore a possible role of an E. faecalis gelatinase (encoded by gelE), which belongs to a family of Zn-metalloproteases that have been shown to be virulence factors in other organisms, in enterococcal infections, an insertion mutant was constructed in OG1RF and tested in the mouse peritonitis model. The mice infected with the gelE mutant showed a significantly prolonged survival compared to the wild type strain. To study the expression of gelE, the regions flanking gelE were sequenced. Sequence analysis of the gelE flanking regions revealed three genes (fsrA, fsrB and fsrC) upstream of gelE that show homology to the genes in a locus (agr) that globally regulates the expression of virulence factors in Staphylococcus aureus and one open reading frame (sprE) with homology to bacterial serine protease downstream of gelE. To study the functions of these genes, gene disruption mutants of fsr and sprE were constructed. Serine protease activity was abolished in the sprE mutant, while both gelatinase and serine protease activities were undetectable in the gelE insertion mutant, suggesting a polar effect on sprE by the insertion in gelE. Disruption of any of fsr genes eliminated the expression of other fsr genes, gelE and sprE and the production of gelatinase and serine protease activities, suggesting auto-regulation of fsr genes and positive regulation of gelE and sprE by fsr genes. Northern blot analysis indicated that fsrB and fsrC are co-transcribed, as are gelE and sprE, which was further confirmed by RT-PCR. Testing the virulence of fsr and sprE mutants in the mouse peritonitis model showed that these mutants had significant delays in lethality compared to OG1RF, suggesting that sprE and fsr genes contribute to the virulence of E. faecalis OG1RF in this model. A gelE deletion mutant had similar LD50 and the time course of killing to those of the wild type strain. However, interpretation of the lack of an effect on virulence is complicated by the overproduction of the serine protease in the gelE deletion mutant. Northern blot analysis also showed that the fsr genes were highly expressed in post-exponential phase during growth and that the expression of fsr genes were cell-density dependent and could be induced by a secreted inducer(s) in the supernatant of OG1RF cultures. Primer extension and transcription fusion analysis suggested that a promoter is located immediately upstream of fsrA, of fsrB and of gelE. Examination of promoter activity in OG1RF and fsr mutants indicated that the fsrB and gelE promoters are fsr dependent and the fsrA promoter is an fsr-independent weak constitutive promoter. Sequence analysis of the fsrB and gelE promoters revealed two conservative 7-bp direct repeats immediately upstream of these promoters; deletions and mutations in the repeated sequences completely abolished the fsrB and gelE promoter activities, suggesting that the repeats are important for the regulatory function in the fsrB and gelE promoter regions. Analysis of surface protein profiles of fsr mutants in comparison with that of OG1RF in two-dimensional gels suggested that fsr may negatively regulate the expression of several surface proteins in OG1RF.

In conclusion, in this study of identification of possible virulence factors in E. faecalis surface and secreted proteins, of three genes encoding antigens detected by human patient sera, none could be shown to effect virulence in the mouse peritonitis model. Inactivation of one of these antigens (autolysin) was shown to slightly increase the tolerance of E. faecalis to penicillin. A serine protease and a locus (fsr) that regulates the expression of gelE and sprE were shown to be important for enterococcal infection in the mouse peritonitis model.

Search pubmed for papers by X Qin and BE Murray or GM Weinstock

Research Info

Identification and Characterization of Virulence Factors in Enterococcus faecalis