The gut may play an important role in orchestrating the clinical picture of sepsis and multiple organ failure. Early work looking at the gut as a motor for sepsis examined the role that various insults to barrier function produced on translocation of bacteria (1). Refinement of this model produced evidence pointing to the additional release of cytokines as being equally, if not more, important in the development of sepsis and multiple organ failure (2). Although experimental models of burns, trauma, and hemorrhagic shock have examined the mechanisms by which sepsis may be induced at the level of the gut, not much is known regarding the role of acute alcohol intoxication on these same mechanisms.
Traumatic injury is preceded by ethanol (EtOH) intake in up to 40% of all cases (3). Epidemiology aside, this relationship may cause a number of physiological disturbances, including immune dysfunction (4). The correlation between the scant mechanistic evidence of EtOH-induced immune dysfunction and the few clinical articles citing the effect of acute intoxication on septic complications has yet to be bridged. This lapse in our understanding of the repercussions of alcohol and trauma is only exacerbated by the fact that most studies have focused on the effect of chronic alcohol usage on systemic immune function.
In previous studies, injury induced by I/R to the gut followed by bacterial challenge has been shown to affect a breakdown of the mucosal barrier (5). At high concentrations of EtOH, a similar disruption of functional gut integrity has been noted (6); however, the resultant immune dysfunction that EtOH produces in a shock-induced gut milieu has not been described. We chose to examine the proinflammatory cytokine production and barrier dysfunction that low levels of EtOH might produce in a reductionist model of shock.
MATERIALS AND METHODS
Caco2 cells obtained from American Type Culture Collection (Rockville, Md) were maintained in Dulbecco's modified Eagle's medium (DMEM; Gibco BRL Products, Grand Island, NY) supplemented with 4.5g/L glucose, 10% fetal calf serum (Hyclose Laboratory, Logan, Utah), 1% nonessential amino acids (Gibco), and a 1% antibiotic-antimycotic solution (penicillin G, streptomycin, and amphotericin B; Sigma Chemical, Co., St. Louis, Mo). The cells were grown in 75-cm2 T-flasks (Fisher, St. Louis, Mo) in an incubator whose environment was maintained at 37°C along with 5% CO2. Medium was changed twice weekly, and cells were passaged every 7 to 10 days.
Confluent monolayers were harvested by washing the cells with Hanks' balanced salt solution (Gibco) followed by trypsin-EDTA solution. Caco2 cells (1 × 105) were seeded into the apical chamber of a two-chamber cell culture system (Costar Corp., Cambridge, Mass) containing a 0.1-μm-pore-size polycarbonate membrane. The formation of a complete cell monolayer was monitored by serial measurements of transepithelial electrical resistance using a Millicell electrical resistance system (Millipore, Bedford, Mass).
Supplemented DMEM was first replaced with RPMI in the Caco2 monolayer wells and then exposed to 0.1% EtOH, Escherichia coli, or both for 30 min. The dosage of alcohol was meant to simulate physiologically relevant values (4). The duration of exposure was chosen as a minimum time required for acute effects from previous work in our laboratory. These cell culture systems, which have been previously shown to be a reliable model for drug metabolism and absorption in the gut (7), were then divided into two treatment groups. The first treatment group underwent hypoxia (5% O2, 37°C) for 90 min, followed by a 90-min reoxygenation period, and the second treatment group was maintained in a normoxic environment (21% O2, 37°C).
The cell culture systems in the hypoxia/reoxygenation (H/R) group were placed in a hypoxic glove box (Coy Laboratory Products, Grass Lake, Mich) whose environment had been regulated to 5% O2 at 37°C. After 90 min of the hypoxic or normoxic culture conditions, both groups were placed in a normoxic incubator environment for 90 min. Supernatants were then collected from the monolayers and centrifuged for 10 min at 1,300 rpm at 4°C. The Caco2 monolayers were analyzed for apoptosis (as described below) immediately.
Supernatants, however, were frozen at −20°C until enzyme-linked immunosorbent assay (ELISA) analysis. The supernatants in this model are reflective of the gut milieu under the respective conditions mentioned above. In previous studies, 90 min in a 5% O2 environment has been shown to correlate to a Po2 of approximately 30 mmHg and would therefore reflect a state of severe shock (8).
At the completion of the experiment, media was cultured to confirm that the 0.1-μm pore size of the polycarbonate membranes excluded bacteria inoculated into the apical side from entering the basal chamber. These data were used to assure that neither EtOH nor a hypoxic environment caused anatomical barrier disruption in the cell model.
Escherichia coli preparation
Escherichia coli C-25, a strain representative of indigenous gut flora that is nonpathogenic, was grown overnight in trypticase soy broth, centrifuged, and resuspended in DMEM sans phenol red. Spectrophotometry was used to determine bacterial concentrations; this was verified by pour-plate assay. Final concentrations of bacteria were adjusted to 106 colony-forming units/mL in fresh DMEM for use in each experiment. This concentration was based on human studies showing a similar number of E. coli bound to between 1 × 104 and 1 × 105 enterocytes obtained from the distal ileum (9).
Caco2 cells posttreatment were trypsinized and resuspended in phosphate-buffered saline. Fluorescein isothiocyanate-conjugated annexin V and propidium iodide were added to 1 × 105 cells according to apoptosis detection kit instructions (R & D Systems, Minneapolis, Minn). Flow cytometry was then used to index apoptosis.
Cytokine determination by ELISA
Previously frozen supernatants were thawed and analyzed for TNF-α and IL-6 concentrations using a commercially available (Biosource International, Camarillo, Calif) ELISA. The minimal detectable levels of IL-6 and TNF-α were 2 and less than 1 pg/mL, respectively.
An ANOVA with a post hoc Tukey test was used to analyze the data. Statistical significance was inferred at values of P < 0.05 All data are expressed as mean ± SD.
Figure 1 shows the synergistic effect of EtOH with H/R on TNF-α production by the Caco2 cell. The production of TNF-α, as measured apically, is consistently higher than basal chamber media values. These findings are congruent with previous work done in our laboratory. Interestingly, there is nearly a 14-fold increase in apical TNF-α production by Caco2 cells with only the addition of 0.1% EtOH. With the concomitant addition EtOH and EC, there is a 40-fold increase in apical Caco2 TNF-α production. Finally, this increases to 50-fold increase when EtOH is added in the setting of 90 min of hypoxia followed by 90 min of reoxygenation.
In the basal chamber, TNF-α production by Caco2 cells is not significantly increased when exposed to EtOH alone. However, the addition of EC to the apical media result in a greater-than 10-fold increase. The basal increase is only 4- to 5-fold when EtOH is added in the setting of H/R. Moreover, when Caco2 cells are exposed to H/R in the presence of both EtOH and EC, the increase is similar to that seen with exposure to EtOH and EC alone. The findings are statistically significant at a P value of less than 0.001 for all groups when compared with Caco2 control and versus all of the other groups.
Figure 2 shows the synergistic effects of EtOH and H/R on IL-6 production. Again, apical cytokine production predominates. We can see that apical levels of IL-6 are statistically significant when compared with Caco2 control not with the addition of EtOH alone, but with the EtOH + EC and EtOH + H/R groups. Apical levels of IL-6 become 30 to 40 times control when Caco2 cells are incubated with 0.1% EtOH and EC and then subjected to H/R. Findings in the basal chamber only have statistical significance in the two groups where Caco2 cells are subjected to H/R. Compared with controls, when Caco2 cells are subjected to H/R in the presence of EtOH, we see a 3- to 4-fold increase in basal IL-6 production. When EC is added, this value becomes five to six times the control levels.
Figure 3 shows the effect of EtOH and H/R on Caco2 cell apoptosis. Caco2 cells treated with EtOH and EC show a significant increase in apoptosis versus Caco2 controls. Apoptosis further increases when Caco2 cells are treated with EtOH alone and then subjected to H/R. The greatest increase is noted when Caco2 cells are treated with EtOH and EC and then subjected to H/R. Statistical significance is set at P value less than 0.001 compared with controls and versus all other treatment groups.
A growing body of evidence points to gut injury as the initial insult in the chain of events leading to full-blown sepsis. In one such model, rodents that had undergone enterectomy had an improved mortality when subjected to hemorrhagic shock when compared with controls (10). Early work looking at the effect of the gut barrier dysfunction focused on the potential for bacterial translocation to cause seeding of distant organs as a nidus for the evolution of a systemic infection. However, this paradigm only partially explained experimental findings. More recent work points to the role of the gut in sepsis as being that of a potent cytokine producer. This cytokine release may occur both in the absence or presence of bacterial translocation and in some studies can occur without bacterial challenge.
In our reductionist model of the gut mucosal barrier, we chose to look at alcohol-mediated Caco2 production of TNF-α and IL-6 production. These cytokines have been shown to cause disruption of intestinal barrier capacity when secreted by various immune cells in the lamina propria (11). Additionally, rodent models have shown both these cytokines to be up-regulated in the intestine before such an effect in the liver and lung (12). These resultant systemic elevations of both TNF-α and IL-6 have been noted to be key mediators in the inflammation associated with sepsis (13). Additionally, there is some evidence to support that these cytokines may be linked with increased morbidity and mortality of sepsis. This idea has been corroborated in human studies where higher levels of IL-6 and TNF-α were seen at 48 h in those septic patients who did not survive their intensive care stay (14).
Although our reductionist model does demonstrate an increase in these two proinflammatory cytokines, it is difficult to extrapolate systemic repercussions. On the contrary, there is sufficient evidence corroborating our findings that gut epithelial secretion of TNF-α and IL-6 produces functional barrier disruption in an autocrine fashion (15). Previous work using this model has shown TNF-α to be a potent inducer of apoptosis in intestinal epithelial cells (IECs) (16). Furthermore, bacterial challenge and I/R injury to the IEC may be paramount in producing this apoptotic signal, providing a means of defense against potential pathogens. However, we can see that in the presence of EtOH, the IEC may respond by promulgating an excessive inflammatory response. Similarly, IL-6 has been shown in previous studies to augment the inflammatory potential of the IEC via increased expression of nuclear factor-κB (16). This may lead to altered permeability to toxins and potential pathogens from the gut lumen. Linking the autocrine effects of gut-mediated cytokine production with the temporal compartmentalization of the gut-liver-lung axis may be a bridge to understanding the relevance of these cytokines to systemic elevations.
Although difficult to extrapolate from our findings, work by Tamion et al. (17) has shown that after hemorrhagic shock and resuscitation, both IL-6 and TNF-α are found to be increased in mesenteric lymph. One can surmise this rise in the liver and the lung may be triggered by gut increases. The exact timing of TNF-α and IL-6 elevation in the gut after shock has some variability depending on the model of shock used, but the trend in all these models points to the gut ischemia as the predecessor to the septic response (18).
Not only does enterocyte release of cytokines lead to alterations in tight junction structure and localized inflammation, but it also promotes increased rates of apoptosis (as seen in Fig. 3). With the addition of bacterial products such as endotoxin or coincubation with a nonpathogenic strain of EC as in the experiments described above, there is a further exaggeration of the TNF-α-mediated permeability changes. Permeability changes may also be indirectly related to these proinflammatory cytokines. The interplay between lymphocytes in Peyer patches and gut mucosal cells in regulating permeability involves complex feedback mechanisms induced by both IL-6 and NO; this is one such example that exposes the limitations of our in vitro model (19).
Although the relationship between shock and bacterial challenge in producing proinflammatory cytokines has a great deal of evidence supporting it, the role of EtOH in producing barrier dysfunction has been more closely linked with direct cytotoxic effects. Much of this work shows disruption of tight junctions, as measured by increases in transepithelial electrical resistance to occur at relatively high concentrations of EtOH greater than 40% (20). Disruption of the actin and myosin filaments leading to breakdown of cellular backbone seemed to be the mechanism by which EtOH caused breakdown of mucosal integrity (20).
However, we can see from Figures 1 and 2 that EtOH, even at low concentrations, produces an exaggerated inflammatory cytokine release when combined with a shock insult. Furthermore, this cytokine release may be sufficient to create a microenvironment wherein apoptosis is increased in an unregulated manner (as shown in Fig. 3). Experimental models point to these changes as playing a role in the development of pulmonary inflammation. This theory has been validated not only by the fact that the mesenteric lymph may be a potent activator of hyperinflammatory neutrophils (21) but also may serve as a direct activator of pulmonary endothelium (21).
Still, the cellular mechanism by which EtOH and H/R produce breakdown of monolayer integrity has yet to be answered. Early work looking at I/R injury in the in vivo models of intestinal barrier function pointed to free radicals as playing a part in producing mucosal injury (22). Work in an animal model of necrotizing enterocolitis expanded upon the aforementioned idea; in this particular model, not only were reactive nitrogen species implicated in gut barrier failure, but these same toxic metabolites did so by depleting glutathione stores before barrier failure could ensue (23). This was validated by the same group in an immortal cell line where blockage of glutathione-dependent detoxifying enzymes produced a decrease in cellular monolayer integrity against free radical challenge compared with those monolayers where glutathione was exogenously added (23). Furthermore, it has been shown that decreases in glutathione and increases in the oxidized form, glutathione disulfide, induce nuclear factor-κb and up-regulate proinflammatory cytokines IL-6 and TNF-α (24), although there may be other pathways associated with this up-regulation (25).
A greater body of evidence for this pathway is found in work examining the effects of I/R injury on oxidant-induced glutathione depletion and subsequent injury to the pulmonary endothelium (26). Additionally, work has been done at examining chronic alcohol usage as a mechanism by which systemic glutathione depletion may lead to increased severity of acute respiratory distress syndrome when compared with matched controls (27). Whether these mechanistic findings are completely applicable to our work on Caco2 monolayers has yet to be proven; however, we believe that looking at glutathione levels in our model is the appropriate next step in proposing a mechanism for our findings.
Thus, just as I/R injury may produce altered immune states, we believe that our model begins to offer a potential mechanism by which septic complications may be increased in patients who are acutely intoxicated, although no definite human relevance can be extrapolated. Clinical work by Gentilello et al. (28) shows the subset of acutely intoxicated trauma patients with penetrating injuries harbor an increased risk of septic complications when compared with those without concomitant EtOH intake. It seems unlikely that the myriad of immune dysfunctions pertaining to patients who chronically consume alcohol are entirely applicable to the acutely intoxicated patient. It is very likely that the long-term immune suppression chronic alcohol users have differed mechanistically in some regards than the proinflammatory sequelae we feel acute alcohol intoxication brings forth.
The intestinal mucosal barrier is essential in distancing the host from potential pathogens and toxins in the gut lumen. Furthermore, disruption of gut integrity may cause unregulated release of proinflammatory cytokines that may cause more distant injury in the host.
Although the destruction of this innate aspect of mucosal defense may be mediated by acute EtOH intoxication, understanding the synergy EtOH demonstrates with H/R injury and bacterial challenge may be important in understanding how the acutely intoxicated trauma patient may more easily succumb to late septic complications.
1. Swank GM, Deitch EA: Role of the gut in multiple organ failure: bacterial translocation and permeability changes. World J Surg
2. Cruz N, Alvarez X, Berg RD, Deitch EA: Bacterial translocation across enterocytes: results of a study of bacterial-enterocyte interactions utilizing Caco-2 cells. Shock
3. Gentilello LM: Alcohol interventions in trauma centers: the opportunity and the challenge. J Trauma
4. Moore EE: Alcohol and trauma: the perfect storm. J Trauma
5. Diebel LN, Liberati DM, Dulchavsky SA: Enterocyte apoptosis and barrier function are modulated by SIgA after exposure to bacteria and hypoxia/reoxygenation. Surgery
6. Miller TA, Smith GS, Banan A, Kokoska ER: Cytoskeleton as a target for injury in damaged intestinal epithelium. Microsc Res Tech
7. Artursson P, Borchardt RT: Intestinal drug absorption and metabolism in cell cultures: Caco-2 and beyond. Pharm Res
8. Xu D, Lu Q, Kubicka R, Deitch EA: The Effect of hypoxia/reoxygenation on the cellular function of intestinal epithelial cells. J Trauma
9. Deneke C, McGowan K, Thorne G, Gorback SL: Attachment of enterotoxigenic Escherichia coli
to human intestinal cells. Infect Immun
10. Chang TW: Improvement of survival from hemorrhagic shock by enterectomy in rats: finding to implicate the role of the gut for irreversibility of hemorrhagic shock. J Trauma
11. Vazquez E, Gil A, Garcia-Olivares E, Rueda R: Weaning induces an increase in the number of specific cytokine-secreting intestinal lymphocytes in mice. Cytokine
12. Ackermann M, Reuter M, Flohe S, Bahrami S, Redl H, Schade FU: Cytokine synthesis in the liver of endotoxin-tolerant and normal rats during hemorrhagic shock. J Endotoxin Res
13. Ding J, Magnotti LJ, Huang Q, Xu DZ, Condon MR, Deitch EA: Hypoxia combined with Escherichia coli
produces irreversible gut mucosal injury characterized by increased intestinal cytokine production and DNA degradation. Shock
14. Tamion F, Richard V, Sauger F, Menard JF, Girault C, Richard JC, Thuilez C, Leroy J, Bonmarchand G: Gastric mucosal acidosis and cytokine release in patients with septic shock. Crit Care Med
15. Guo W, Magnotti LJ, Ding J, Huang Q, Xu D, Deitch EA: Influence of gut microflora on mesenteric lymph cytokine production in rats with hemorrhagic shock. J Trauma
16. Diebel LN, Liberati DM, Baylor AE 3rd, Brown WJ, Diglio CA: The pivotal role of tumor necrosis factor-alpha in signaling apoptosis in intestinal epithelial cells under shock conditions. J Trauma
17. Tamion F, Richard V, Lyoumi S, Daveau M, Bonmarchand G, Leroy J, Thuillez C, Lebreton JP: Gut ischemia and mesenteric synthesis of inflammatory cytokines after hemorrhagic or endotoxic shock. Am J Physiol
18. Grotz MR, Ding J, Guo W, Huang Q, Deitch EA: Comparison of plasma cytokine levels in rats subjected to superior mesenteric artery occlusion or hemorrhagic shock. Shock
19. Chen J, Ng CP, Rowlands DK, Xu PH, Gao JY, Chung YW, Chan HC: Interaction between enteric epithelial cells and Peyer's patch lymphocytes in response to Shigella
lipopolysaccharide: effect on nitric oxide and IL-6 release. World J Gastroenterol
20. Ma TY, Nguyen D, Bui V, Nguyen H, Hoa N: Ethanol modulation of intestinal epithelial tight junction barrier. Am J Physiol
21. Lu Q, Xu DZ, Davidson MT, Hasko G, Deitch EA: Hemorrhagic shock induces endothelial cell apoptosis, which is mediated by factors contained in mesenteric lymph. Crit Care Med
22. Granger DN, Rutili G, McCord JM: Role of superoxide radicals in intestinal ischemia. Gastroenterology
23. Kelly N, Friend K, Boyle BS, Zhang XR, Wong C, Hackam DJ, Zamora R, Ford H, Upperman JS: The role of the glutathione antioxidant system in gut barrier failure in a rodent model of experimental necrotizing enterocolitis. Surgery
24. Kirsch JD, Yi AK, Spitz DR, Krieg AM: Accumulation of glutathione disulfide mediates NF-kappa B activation during immune stimulation with CpG DNA. Antisense Nucleic Acid Drug Dev
25. Haddad JJ: Glutathione depletion is associated with augmenting a proinflammatory signal: evidence for an antioxidant/pro-oxidant mechanism in regulating cytokines in the alveolar epithelium. Cytokines Cell Mol Ther
26. Zhao X, Alexander JS, Zhang S, Zhu Y, Sieber NJ, Aw TY, Carden DL: Redox regulation of endothelial barrier integrity. Am J Physiol Lung Cell Mol Physiol
27. Yeh MY, Burnham EL, Moss M, Brown LS: Chronic alcoholism alters systemic and pulmonary glutathione redox status. Am J Respir Crit Care Med
28. Gentilello LM, Cobean RA, Walker AP, Moore EE, Wertz MJ, Dellinger EP: Acute ethanol intoxication increases the risk of infection following penetrating abdominal trauma. J Trauma