Michael T. Reddell

A novel model of common Toll-like receptor 4- and injury-induced transcriptional themes in human leukocytes

IntroductionAn endotoxin challenge, sepsis, and injury/trauma, trigger significant changes in human peripheral blood leukocytes (PBL) gene expression. In this study, we have sought to test the hypothesis that the Toll-like receptor 4 (TLR4) induced transcription patterns elicited in humans exposed to in vivo endotoxin would parallel gene expression patterns observed in trauma patients with initial non-infectious injury. In addition, we sought to identify functional modules that are commonly affected by these two insults of differing magnitude and duration.MethodsPBL were obtained from seven adult human subject experimental groups. The groups included a group of healthy, hospitalized volunteers (n = 15), that comprised four study groups of subjects challenged with intravenous endotoxin, without or with cortisol, and two serial samplings of trauma patients (n = 5). The PBL were analyzed for gene expression using a 8,793 probe microarray platform (Gene Chip® Focus, Affymetrix). The expression of a subset of genes was determined using qPCR.ResultsWe describe sequential selection criteria of gene expression data that identifies 445 genes that are significantly differentially expressed (both P ≤ 0.05 and >1.2 fold-change) in PBL derived from human subjects during the peak of systemic inflammatory responses induced by in vivo endotoxin, as well as in PBL obtained from trauma patients at 1 to 12 days after admission. We identified two functional modules that are commonly represented by this analysis. The first module includes more than 50 suppressed genes that encode ribosomal proteins or translation regulators. The second module includes up-regulated genes encoding key enzymes associated with glycolysis. Finally, we show that several circadian clock genes are also suppressed in PBL of surgical ICU patients.ConclusionsWe identified a group of >400 genes that exhibit similar expression trends in PBL derived from either endotoxin-challenged subjects or trauma patients. The suppressed translational and circadian clock modules, and the upregulated glycolytic module, constitute a robust and long lasting PBL gene expression signature that may provide a tool for monitoring systemic inflammation and injury.

Quantitative myoelectric determination of bowel viability

A new device designed to quantitatively measure viability in ischemic bowel was studied in 20 dogs. This strain gauge device is clipped on the serosal surface and can rapidly deliver an electrical stimulus. The threshold stimulus level (TSL) was the stimulus necessary to produce a clearly defined smooth muscle contraction. TSL was compared with blood flow measured by Doppler ultrasound at 2-cm intervals in 30-cm ischemic segments as a determinant of bowel viability. Doppler readings were taken from the bowel wall (BW), peripheral arteriolar branches of the marginal artery (PA), and marginal artery (MA), itself, at each 2-cm interval. Bowel color and peristalsis were correlated with Doppler and TSL data. TSL scale ranged from 0 to 100 mA and varied from 21 +/- 2 mA in normal bowel to 98 +/- 2 mA in gangrenous bowel. At the locus of the last audible Doppler signal in the BW, PA, and MA, mean TSL readings were 27 +/- 5, 34 +/- 6, and 48 +/- 16 mA, respectively. These readings were each significantly greater than the TSL of normal bowel at P less than or equal to 0.0001 level. Portions of each ischemic segment were resected and reanastomosed. Absence of Doppler signals in the BW and PA did not preclude healing of ischemic bowel. These data suggest that TSL measurement is more sensitive than either Doppler ultrasound or gross visual evaluation in assessment of bowel viability.

Intraoperative Assessment of Bowel Viability

Viability of ischemic bowel was assessed in 30 dogs after mesenteric arterial ligation in a 40-cm length of ileum. Viability was evaluated using two gross features, color and peristalsis, and four objective methods including bowel wall surface oximetry (pO2), Doppler ultrasound, quantitative fluorescein fluorimetry, and myoelectric activity measured by a strain gauge probe. Each parameter was measured at 2-cm intervals within the 40-cm ischemic segment before resection and anastomosis was performed. There were seven fatal anastomotic leaks, all due to further bowel necrosis. Survival did not correlate with bowel color, presence of peristalsis, bowel wall pO2 Doppler ultrasound, or the myoelectric parameters. However, fluorescein fluorimetry was predictive of long-term viability. These results suggest that quantitative fluorescein fluorimetry may be a useful adjunct in intraoperative bowel viability assessment.

Bacterial Translocation After Mesenteric Ligation in Dogs

These experiments were designed to determine the relationship between translocation of Escherichia coli and viability of ischemic small bowel. Twenty beagles were gavaged with 14C-labeled E. coli at two time intervals (3 and 24 h) prior to ligation of the blood supply to a 40-cm segment of ileum. Mesenteric lymph node (MLN) biopsies and bacterial cultures of the peritoneal fluid, peripheral arterial blood, and splanchnic venous blood were taken immediately prior to ligation and 24 h later both before and after the ischemic bowel was resected and anastomosed. Biopsies of each resection margin were taken to measure translocation of E. coli into the bowel wall. Several hemodynamic hemodynamic parameters were also measured before and 24 h after ligation. Seven of the 20 dogs died of further bowel necrosis. In survivors A-alpha DO2 was significantly decreased 24 h after mesenteric ligation vs. preligation, whereas in dogs that died DO2 was significantly increased after ligation vs. preligation. The incidence of mesenteric venous cultures positive for E. coli was significantly higher 24 h after ligation vs. preligation. However, there was no correlation between survival and the incidence of positive E. coli cultures in the blood or peritoneal fluid. Mean MLN counts were significantly higher in dogs gavaged at 3 h vs. those gavaged 24 h prior to laparotomy. However, there was no correlation between survival and translocation into either the bowel wall or MLN at either time interval. Viability of ischemic small bowel in this canine model was not affected by translocation of E. coli. Hemodynamic parameters that are altered during the course of sepsis also did not correlate with survival.

Clinical administration of LPS endotoxin show changes in heart rate variability are detected prior to changes in temperature

Sepsis is a significant cause of hospital deaths, but early detection could save lives. Heart rate variability (HRV) is a measurement of the change in heart rate and may be useful as a marker for early detection of sepsis. In this study, E. coli endotoxin was administered to healthy human subjects and changes in temperature and HRV were monitored at 6 hours post administration. HRV features measured were pNN50 and multi-scale entropy (MSE). We modeled changes in baseline using a sigmoidal model with 3 variables to help quantify dynamics of HRV and temperature time series. Curves of HRV and temperature versus time demonstrated that changes in HRV features preceded changes in temperature. Our model fits indicate HRV changes both begin earlier and occur faster than temperature or heart rate. Our findings are consistent with other group findings that demonstrate changes in interbeat intervals can be detected earlier than other signs of systemic inflammation.