Introduction: Despite improvements in operative technique and antibiotics, patients remain at risk for anastomotic leaks, especially those exposed to adjunctive chemoradiation. The aim of this study was to define potential mechanisms by which pathogenic microbes, often present in the intestinal tract of complex patients, affect healing and anastomotic leakage. Specifically, we sought to determine whether radiation affects the virulence properties of intestinal Pseudomonas aeruginosa to a degree sufficient to influence its ability to cause anastomotic leakage and attenuate wound healing. Methods: A model involving preoperative radiation and colon resection was created in which male rats (250-300 g) received fractionated pelvic radiation with a total dose of 25 Gy followed one week later by a low colon resection and anastomosis. At laparotomy, the cecum was directly injected with P. aeruginosa. Additional groups underwent radiation, resection, and anastomosis formation without P. aeruginosa injection as well as resection, anastomosis, and P. aeruginosa injection without radiation. Six days after surgery, all rats were euthanized and assessed for gross anastomotic leaks. To determine if in vivo activation of P. aeruginosa influenced the development of anastomotic leak, P. aeruginosa was isolated from anastomotic sites and non-anastomotic sites of radiated and non-radiated rats on selective media and then analyzed for virulence expression (pyocyanin production), expression of a lethal phenotype (C. elegans killing assays), and the ability to impair healing of wounded rat intestinal epithelial cell monolayers (IECs). Results: Rats exposed to preoperative pelvic radiation followed by a colorectal anastomosis and P. aeruginosa bowel contamination had a significantly higher anastomotic leak rate compared to non-radiated animals (75% vs 8%, respectively; n=12, p<0.05). However, radiated and non-radiated rats subjected to colon anastomosis without P. aeruginosa injection had no leaks, implicating a causal role for P. aeruginosa. Although P. aeruginosa preferentially segregated to the anastomotic sites of both radiated and non-radiated rats, its virulence (pyocyanin production; p<0.05) and lethal potential (C. elegans killing assays; p<0.05) were significantly enhanced in strains recovered from radiated rats. Assays performed on wounded IEC cells showed impaired cellular healing in the presence of P. aeruginosa recovered from the anastomosis of radiated animals, suggesting that radiation-induced in vivo virulence activation of P. aeruginosa plays a direct role in epithelial healing. Conclusion: The presence of pathogenic flora as a result of patient complexity, use of antibiotics, etc may play an underappreciated role in anastomotic leaks. Pathogens such as P. aeruginosa appear to be specifically up-regulated in virulence during radiation, which may enhance their ability to disrupt the healing and tissue integrity of a high risk anastomosis.

Introduction: Despite extensive investigation, disruption of intestinal anastomoses leading to leak and sepsis remains as a significant cause of surgical morbidity. While both host and microbial factors independently play roles in anastomotic failure, their dynamic bidirectional crosstalk may generate key response elements within the interacting systems (the interactome) that influence anastomotic healing. Spatial and temporal dynamics need to be accounted for, as pathogen-host interactions at the precise site of anastomotic failure produce local environmental cues leading to microbial phenotype transformation linked to epithelial injury and healing. Agent-based modeling, an object-oriented computational modeling tool, is well suited to capture and synthesize these dynamic interactions. Here we use an agent-based model (ABM) to integrate the mechanisms of by which P. aeruginosa, a microorganism often present during anastomotic failure, responds to the microenvironment of epithelial injury and healing incorporating data derived from in vitro experiments. Methods: An ABM of epithelial wound healing was constructed using NetLogo. Computational agents representing individual epithelial cells were used to populate an in silico analog of an in vitro cell culture model, which included the culture media and soluble factors in the background. Rules governing agent behavior were adapted from known mechanisms of epithelial wound healing in the literature. The in silico epithelial monolayer was then damaged as would be done using a traditional scratch assay. A second agent layer representing P. aeruginosa bacteria was introduced, and bacterial activation at the wound site was modeled based on a putative mechanism involving nitric oxide, hypoxia-inducible factor-1α and adenosine. Simulations of hypoxia and radiation effects were also performed. Results: The ABM accurately reproduced epithelial wound healing dynamics. Activation of epithelial cells via ERK-1/2 signaling and EGF-R expression led to cell migration and spreading in order to heal the damage. Wound closure speed was calibrated to the known data based on the presence or absence of various factors such as EGF in the media. Addition of P. aeruginosa to the system demonstrated patterns of bacterial adherence similar to in vitro results, with associated inhibition of epithelial migration and healing. Hypoxia augmented migration, while radiation inhibited migration and increased bacterial adherence. Conclusions: Awareness of the importance of the host-microbe interactome adds a new level of complexity in understanding differences in anastomotic healing. Addressing this complexity will likely require representation of multi-factorial interactions at a level of resolution not currently possible. Dynamic knowledge representation using agent-based modeling may provide a means of increasing our resolution in visualizing mechanistic hypotheses and integrating them into a clinically relevant context.

Introduction: During organ harvest for transplantation (TX) tissues undergo cold ischemia (C) during preservation and reperfusion during implantation. Cooling to 4 °C with specialized solutions is routinely employed to decrease tissue edema, oxidative stress and preserve membrane integrity. We have previously reported that high molecular weight polyethylene glycol (15-20,000 MW, PEG15-20) protects the intestinal epithelium against a variety of cellular stresses including radiation injury and microbial invasion by mechanisms that appear to involve lipid rafts. PEGs have also been shown to decrease lipid peroxidation and the production of reactive oxygen species in various tissues. The aim of this study was to determine if a high molecular weight PEG could more efficiently preserve the intestinal integrity of intestine grafts harvested for subsequent transplantation when compared to a standard tissue preservation solution (HTK). Methods: Twenty (20) 8 week old C57BL6 mice underwent total enterectomy in preparation for small intestinal transplantation using a well described technique of isolation and preservation of vascular inflow and outflow. Intestinal segments were preserved in either in 4C Histidine-Tryptophan-Ketoglutarate (HTK) or HTK+ 5% high molecular weight PEG co-polymer solution. Intestinal segments were analyzed histologically for tissue preservation and integrity by calculating an epithelial injury score at 9 h and 24 h following preservation. Results: Gross macroscopic exam of tissues revealed that segments from the HTK+5% PEG appeared less edematous, the vascular architecture was highly defined, and no gross necrosis was detected compared to HTK alone which displayed visible areas of necrosis and pale and ill-define vasculature (data not shown). Histology demonstrated mild lymphoplasmocytic infiltration in HTK+5% PEG compared to moderate to severe in HTK alone. Tissue architecture was intact in HTK+5% PEG and disrupted in HTK alone with visible focal ulcerations, erosions, crypt distortion, denuded epithelium and fibrosis HTK (B). Goblet cells were better preserved in HTK+5% PEG solution (data not shown).