26.05 GelE/sprE are critical for Enterococcus faecalis-induced anastomotic leak in a rat model

J. N. Luo1, B. A. Shakhsheer1, R. Klabbers2, A. Zaborin1, N. Belogortseva1, O. Zaborina1, J. C. Alverdy1  1The University Of Chicago Pritzker School Of Medicine,Department Of Surgery,Chicago, IL, USA 2Radboud University Nijimegen Medical Centre,Department Of Surgery,Nijimegen, GELDERLAND, Netherlands

Introduction:  Anastomotic leak following colorectal surgery is a dreaded complication of which the cause remains unknown. Our laboratory has recently established that high collagenase producing strains of Enterococcus faecalis (E. faecalis) “bloom” in the gut following anastomotic injury and are both necessary and sufficient to cause anastomotic leak in rats.  Here we performed a mutational analysis to define the roles of gelE and sprE – co-regulated genes that control: adherence, penetration and collagenase production – on anastomotic leak.  We used a double knockout mutant lacking gelE and sprE derived from E. faecalis V583 (a vancomycin-resistant clinical isolate). We hypothesize that mutants deficient in both genes (ΔgelEΔsprE) would be attenuated in their capacity to cause leak in rats. 

Methods:  Adult rats (n=24) underwent colon resection followed by recto-sigmoid anastomosis. Following anastomotic construction, 5cc of 107 CFU/ml of either the double mutant deficient in both gelE and sprE (ΔgelEΔsprE), or the mutant complemented with gelE and sprE (ΔgelEΔsprE/gelE+sprE- termed VT07) were administered via rectal enema to inoculate the anastomosis. On postoperative day six, all rats were sacrificed and evaluated for evidence of anastomotic leak via previously established criteria. Anastomotic tissues were harvested and separated into mucosal, serosal, and perianastomotic samples for bacterial species identification via culture and phenotype analysis (collagenase activity).
 

Results: Anastomotic leakage was observed to be significantly greater in rats rectally inoculated with the complemented strain VT07 that produces high collagenase compared to its matched double mutant ΔgelEΔsprE that produces no collagenase (p<0.01). Leaks were characterized by dense perianastomotic adhesions, perianastomotic abscesses, and occasional gross anastomotic dehiscence.  The most severe leaks were identified to be associated with high adherence of E. faecalis to the mucosa, penetration into the serosa, high collagenase activity, and culture positivity of perianastomotic tissues. Dense mucosal colonization alone was not associated with leak. 
 

Conclusion: High collagenase producing E. faecalis present at and within anastomotic tissues appears to play a critical role in the pathogenesis of anastomotic leak. Although various genes are likely to confer its pathogenic potential at healing anastomotic tissues, we show in this study that gene products and pathways regulated by the gelE/sprE system that control adherence, penetration, and collagenase/protease production are likely to play a key role.  Because E. faecalis is a ubiquitous commensal organism in humans, and only certain strains of it are associated with anastomotic leak, a more complete understanding of the microbial pathogenesis and phenotypic expression could enable us to better predict leak and thus improve patient management.