Surgical stress often induces distant organ impairment (1–3). In that regard, the gastrointestinal tract is known to play a key role in the etiology of distant organ injuries under such conditions as surgical stress, hemorrhagic shock, burn, trauma, and reperfusion injury (4–10). Indeed, compelling evidence indicates that the small intestine is the primary source of factors inducing lung injury after major digestive surgery and that the lymphatic system is the major route by which these gut-derived factors reach the pulmonary circulation (11–13). Based on observations in animal models, neutrophils, cytokines, and reactive oxygen species (ROS) have all been proposed as the major cause of lung impairment (4–10,14,15). Neutrophil accumulation and their interaction with endothelial cells are secondary events, however (1); the first response to a surgical insult is more likely increased production of cytokines or ROS. In that regard, ROS are associated with the development of acute respiratory distress syndrome resulting from direct tissue injury during the early phases of critical illness (5,11,15). We have been interested in the relationship between ROS production induced by surgical stress and lung edema. In the present study, we hypothesized that surgical stress induces production of ROS in the small intestine and that these factors then enter the mesenteric lymph and induce lung edema.
MATERIALS AND METHODS
2′,7′-Dichlorofluorescein diacetate (DCFH-DA) was purchased from Molecular Probes Inc. (Eugene, OR). 2′,7′-Dichlorofluorescein (DCF) was from Sigma Chemical Co. (St. Louis, MO). Sodium (−)-8-(3-methoxy-4-phenylsulfinylphenyl) pyrazolo [1,5-a]-1,3,5-triazine-4-olate monohydrate [(−)BOF 4272], a novel xanthine oxidase (XOD) inhibitor, was a kind gift from Otsuka Pharmaceutical Factory Inc. (Naruto, Tokushima, Japan). Evans blue was purchased from Wako Chemicals (Tokyo, Japan). Hydrogen peroxide detection kits were from OXIS International, Inc. (Portland, OR).
Animals and experimental model
All experiments were carried out in accordance with the animal guidelines of Akita University School of Medicine. Male pathogen-free Sprague-Dawley rats (280–350 g) were housed under a daily 12-h light-dark cycle and fed water and rat chow ad libitum for at least 2 weeks. They were then fasted for 24 h before experimentation.
For experimentation, rats were anesthetized by intraperitoneal injection of pentobarbital sodium (50 mg/kg body weight), after which a polyethylene cannula (diameter 2.5 mm) was inserted into the trachea, and the rats were ventilated with a mechanical ventilator (EVM-50A, Aika, Tokyo, Japan) delivering room air at a respiration volume of 0.5 mL/g/min at 80 breaths/min (i.e., tidal volume 1.9 mL/300 g rat). The femoral vein was cannulated with a polyethylene tube for administration of agents, and the femoral artery was cannulated to continuously monitor systemic arterial pressure. Surgical manipulation was carried out as described by Anup et al. (4–7). Briefly, the abdominal wall was opened with a midline incision approximately 5 cm in length, and the intestine was gently handled for 5 min over its entire length, from the ileocecal junction proximally, simulating the “inspection” that occurs in a clinical setting. The intestine was then replaced in the abdominal cavity, the abdominal wall was sutured, and the rats continued to be mechanically ventilated for 5 h. Following examination of the small intestine and lung (see below), animals were killed by administration of additional pentobarbital sodium. To avoid the influence of bile juice, in experiments in which H2O2 production in the mucosa of small intestine was measured, the common bile duct was cannulated with a polyethylene tube that led the bile to an extraabdominal collecting site. To measure H2O2 in the mesenteric lymph, the mesenteric lymphatic duct, which is situated between the superior mesenteric artery and vein, was exposed, cannulated with 24-gauge catheter connected to silicon tube (0.5 mm diameter), and exteriorized via the wound. Samples were then collected at the indicated times.
Rats were assigned to three groups: a control group in which ventilation was started without laparotomy or intestinal handling, but at the last 40 min of the 5-h ventilation, the abdomen was opened and the common bile duct cannulated (n = 21); a stress group in which rats were subjected to laparotomy and intestinal handling for 5 min, and then at the last 40 min of the 5-h ventilation, the abdomen was reopened, and the common bile duct cannulated (n = 25); and a stress + BOF group in which rats were intraperitoneally injected with 2 mg of the XOD inhibitor BOF 4272 1 h before being subjected to the surgical stress protocol (n = 16). In addition, as a positive control, we produced a rat sepsis model by giving a continuous femoral intravenous infusion of endotoxin at 4.5 mg/kg/h for 2 h, as previously reported (n = 4) (16,17).
H2O2 production in small intestine and lung
H2O2 production was detected as a function of DCF fluorescence. Within cells, DCFH-DA is hydrolyzed to nonfluorescent DCFH, which is rapidly oxidized to highly fluorescent DCF in the presence of H2O2 (18). We used this dye to visualize and measure H2O2 production in the mucosa of the small intestine and lung using a well-established digital fluoroimaging technique (16,17,19–22). The DCFH-DA (1 mg/rat) was dissolved in pure ethanol and infused over the last 40 min of the ventilation period. The rats then received a second laparotomy, and the small intestine was incised longitudinally at the center point of the overall length and rolled out flat, enabling a cover slide glass to be placed on the mucosa. Alternatively, we opened the chest with a median incision that was extended to a left lateral thoracotomy along the eighth rib, and an intralobar site within the left upper lobe of the lung was gently bonded to a glass chamber using the bonding agent Alonalpha (16–18). We then observed the selected regions using intravital fluorescence microscopy (BH-2FRC, Olympus, Tokyo, Japan). The excitation and emission wavelengths were 490 and 530 nm, respectively (excitation filter BP490 and Dichroic mirror and barrier filter DM500, Olympus, Tokyo, Japan). The fluorescence images of the intestine and lung were captured with a silicon-intensified target camera (C2741-08, Hamamatsu Photonics, Shizuoka, Japan), displayed on a video monitor, and recorded on a video cassette recorder.
Analysis of H2O2 production in the small intestine and lung
The recorded images were digitized to a resolution of 512 × 512 pixels and 256 gray levels, and the averaged fluorescence intensities in the small intestine and lung were measured using NIH Image ver. 1.56 image analysis software. To quantify the DCF fluorescence in the intestinal mucosa, we first measured the fluorescence intensities of known concentrations of DCF in a hemocytometer. These fluorescence intensities (256 gray levels) were then plotted against the DCF concentrations to construct a calibration line (Fig. 1A). Using this calibration line, we calculated the DCF fluorescence intensity in the mucosa of small intestine, expressing H2O2 production in terms of DCF concentration. To analyze the DCF in the lung, fluorescent areas in five images were selected, and the numbers of pixels counted. DCF fluorescence was then expressed in terms of “pixels/5 images.” The specificity of the DCF reaction with H2O2 was confirmed by blocking it with catalase.
Detection of H2O2 in the mesenteric lymph
In some experiments, H2O2 concentrations were measured in samples of mesenteric lymph collected 1 h after inducing surgical stress. In other experiments, changes in the H2O2 concentration in mesenteric lymph were determined by comparing the levels in lymph samples collected 1 h before induction and samples collected 2 h after induction of surgical stress. The collected samples were centrifuged at 800 ×g for 10 min at 4°C, after which total hydroperoxides were measured using a BIOXYTECH H2O2–560 Quantitative Hydrogen Peroxide Assay in the calorimetric quantitative determination mode according to the manufacturer's instructions.
Analysis of lung permeability
Pulmonary microvascular permeability was measured using a modification of the Evans blue dye extravasation technique described previously (24,25). Briefly, animals received 20 mg/kg of Evans blue by injection 2 h before being killed. At the time of sacrifice, a heparinized sample of blood was taken from the cannulated femoral vein, and the plasma was removed by centrifugation. The lungs were then perfused free of blood with 20 mL of normal 0.9% saline, after which they were removed from the thoracic cavity and surrounding mediastinal structures and weighed. The pulmonary tissue was then homogenized in 3 mL of 0.9% saline added to 2 vol of deionized formamide and incubated at 60°C for 12 h. The supernatant was then separated from the lung tissue by centrifugation at 2000 ×g for 30 min. Evans blue in plasma and lung tissue was quantified by dual-wavelength spectrophotometric analysis as described by Linderkamp et al. (26). This method corrects the specimen's absorbance at 620 nm for the absorbance of contaminating heme pigments. The correction is calculated with the following formula: corrected absorbance at 620 nm = actual absorbance at 620 nm − [1.426(absorbance at 740 nm) + 0.03]. The amount of Evans blue measured in the pulmonary tissues was then normalized to the tissue weight, after which a permeability index that reflects the degree of extravasation of Evans blue into the extravascular pulmonary tissue compartment was calculated by dividing the corrected pulmonary tissue Evans blue absorbance (620 nm/g of lung tissue) by the corrected plasma Evans blue absorbance (620 nm) (27).
Myeloperoxidase in the lung
To assess neutrophil accumulation in the lung, we measured the myeloperoxidase (MPO) content of lungs that had been stored at −80°C. The lungs were weighed, homogenized in buffer, and then centrifuged at 2500 ×g for 10 min. MPO activity was then assayed in the supernatant using a standard spectrophotometric technique (28). Test samples (0.1 mL) were mixed with 2.9 mL of 50 mM phosphate buffer (pH = 6.0) containing 0.167 mg/mL o-dianisidine dihydrochloride (Kanto Chemical, Tokyo, Japan) and 0.0005% H2O2 (Wako Chemical, Tokyo, Japan) in a final volume of 3 mL. The absorbance change at 460 nm was measured over 2 min in the cell (25°C) of a spectrophotometer (U-1100, Hitachi Instruments Service Co, Tokyo, Japan), after which the MPO concentration in the samples was calculated using a standard curve in which absorbance was plotted against known concentrations of human MPO. At the same time, protein in the lungs was estimated using a BCA Protein Assay Reagent Kit (Pierce, Rockford, US), and MPO levels were expressed as μg/mg protein.
TNF-α in the lung
Lungs that had been stored at −80°C were weighed, homogenized in buffer, and then centrifuged at 2500 ×g for 10 min. The TNF-α content of the supernatants was determined using a Cytoscreen rat TNF-α ELISA kit (Biosource International, Camarillo, CA) according to the manufacturer's instructions with a Wellreader SK601 (Seikagaku Corp., Tokyo, Japan). Again, protein in the lung was estimated, and TNF-α levels were expressed as pg/mg protein.
Data from the three groups were expressed as means ± SD. The significance of differences between groups was assessed by one-way analysis of variance (ANOVA) with Scheffe's and Fisher PLSD multiple comparison tests. Values of P < 0.05 were considered significant.
The intensity of the DCF fluorescence recorded from the mucosa of the small intestine in the stress group was significantly greater than control (Fig. 1B), indicating a significant rise in the H2O2 concentration in the mucosa of the small intestine. This effect was completely blocked by pretreatment with XOD inhibitor BOF 4272 (Fig. 1B).
Consistent with the increase in the small intestine, the concentration of H2O2 in the mesenteric lymph 2 h after surgical manipulation of small intestine was significantly higher than the levels 1 h before manipulation; no change was observed in control animals not subjected to surgical stress (Fig. 2A). In addition, the postsurgical H2O2 concentration in the mesenteric lymph was found to be significantly higher in the stressed rats than in control animals, and this effect was completely blocked by BOF 4272 (Fig. 2B). By contrast, serum H2O2 was not detected in this assay following surgical stress. Analysis of Evans blue extravasation showed that surgical manipulation of the small intestine significantly increased lung microvascular permeability, as compared with control. As with the increases in H2O2 in the small intestine and mesenteric lymph, the surgical stress-induced increase in lung microvascular permeability was blocked by BOF 4272 (Fig. 3), suggesting that a rise in mesenteric H2O2 leads to an increase in pulmonary microvascular permeability.
We measured pulmonary MPO as an indicator of neutrophil accumulation in the lung. The results were confirmed by the finding that surgical stress had no effect on neutrophil accumulation in the lung (Fig. 4). By contrast, MPO levels were markedly increased by endotoxin administered as a positive control. Moreover, few neutrophils from stressed animals exhibited DCF fluorescence, as compared with those from animals administered endotoxin, and there was no difference in the level of DCF fluorescence in neutrophils from stressed and control animals (Fig. 5). These data suggest that surgical manipulation of small intestine does not increase the neutrophil accumulation or H2O2 production in the pulmonary microcirculation. Furthermore, surgical manipulation of intestine also had no significant effect on TNF-α production in the lung and blood (Fig. 6).
In the present study, we used the well-characterized technique of intravital fluorescence microscopy (16,17,19–23) to accomplish the first real-time visualization of H2O2 production in the small intestine. In addition, we detected for the first time a surgical stress-induced rise in H2O2 in the mesenteric lymph. The fact that pretreatment with BOF 4272, a specific XOD inhibitor with no free radical scavenging ability at the dose used (19,29), completely blocked surgical stress-induced increases in H2O2 in the intestinal mucosa and mesenteric lymph confirms that the production of H2O2 was XOD-dependent. Although, it has been reported that XOD in pulmonary endothelial cells can play a major role in lung injury, that does not appear to be the main cause of lung injury in the present model (30). Indeed, Magnotti et al. demonstrated that ligation of the mesenteric lymphatic system prevented the lung injury induced by hemorrhagic shock (11). We therefore believe that activation of intestinal XOD by intestinal handling caused the increase in the pulmonary microvascular permeability. The idea that the lymphatics are the major route by which gut-derived factors reach the pulmonary circulation makes sense because the lung is the first vascular bed to be exposed to the mesenteric lymph. Once in the pulmonary circulation, these mediators (ROS) interact with vascular endothelial cells, thereby diminishing capillary vascular patency. It is likely that in blood ROS are efficiently scavenged; that capacity is apparently not present in lymph. Consequently, gut-derived ROS are able to enter the pulmonary circulation via mesenteric lymph and to play a dominant role in inducing lung edema. On the other hand, recent studies have implicated serine proteases in the pathogenesis of shock-induced gut injury and the generation of biologically active mesenteric lymph and have shown that glycine reduces systemic and local inflammatory responses following intestinal ischemia/reperfusion or endotoxin challenge. In a future study, we will determine whether gut-derived ROS contribute directly to lung injury or whether they act indirectly by stimulating production of another gut factor that in turn acts in the lung (31,32).
Thomas et al. also reported that surgical manipulation of the intestine increases the sequestration of activated neutrophils and levels of ROS in lung (5). By contrast, we detected virtually no H2O2-producing neutrophils using DCF and intravital fluorescence microscopy 30 min, 2 h, and 5h after surgery. Moreover, neither measurement of MPO levels nor histologic examination of tissue samples indicated an accumulation of neutrophils in the lung following surgical manipulation of intestine. Likewise, TNF-α activity in the lung was unaffected by surgical stress. We are not certain of the precise reason our results differed from those of Thomas et al. (5), but there were some important differences in the experimental models. For instance, in contrast to the earlier study, we provided continuous mechanical ventilation before and after surgery to protect against possible hypoventilation and hypoperfusion of the lung after surgery, and our surgical manipulation of intestine was relatively prolonged (5 min). In addition, we were careful to be sure that the arterial blood pressure was maintained throughout the experiment and that the arterial blood gases were within their normal ranges. As a result of these measures, there was little histologically detectable damage to the lungs of surgically stressed rats. Only lung edema was present, which was detected using Evans blue kinetics. We therefore believe we were able to detect an early phase of pure surgical stress-induced lung edema caused by ROS.
In summary, surgical stress induces XOD-dependent H2O2 production in the small intestine. The H2O2 then enters the mesenteric lymph and makes its way to lung, where it increases capillary permeability and thus induces lung edema. Given these findings, we suggest that a free radical inhibitor or scavenger might be an effective therapeutic agent with which to treat distant organ failure induced by surgery.
1. Ogawa M: Mechanisms of the development of organ failure following surgical insult: the second attack theory. Clin Intensive Care 7:34–38, 1996.
2. Marshall JC, Christou NV, Meakins JL: The gastrointestinal tract. The `undrained abscess` of multiple organ failure. Ann Surg 218:111–119, 1993.
3. Deitch EA: Multiple organ failure. Pathophysiology and potential future therapy. Ann Surg 216:117–134, 1992.
4. Anup R, Balasubramanian KA: Surgical stress and the gastrointestinal tract. J Surg Res 92:291–300, 2000.
5. Thomas S, Karnik S, Balasubramanian KA: Surgical manipulation of the small intestine and its effect on the lung. J Surg Res 106:145–156, 2002.
6. Anup R, Susama P, Balasubramanian KA: Role of xanthine oxidase in small bowel mucosal dysfunction after surgical stress. Br J Surg 87:1094–1101, 2000.
7. Anup R, Aparna V, Pulimood A, Balasubramanian KA: Surgical stress and the small intestine: role of oxygen free radicals. Surgery 125:560–569, 1999.
8. Pastores SM, Katz DP, Kvetan V: Splanchnic ischemia and gut mucosal injury in sepsis and the multiple organ dysfunction syndrome. Am J Gastroenterol 91: 1697–1710, 1996.
9. Wang JF, Gao YQ, Lippton H, Hyman A, Spitzer JJ: The roles of nitric oxide and hydrogen peroxide production in lipopolysaccharide-induced intestinal damage. Shock 2:185–191, 1994.
10. Parikh AA, Moon MR, Pritts TA, Fischer JE, Szabo C, Hasselgren PO, Salzman AL: IL-1beta induction of NF-kappaB activation in human intestinal epithelial cells is independent of oxyradical signaling. Shock 13:8–13, 2000.
11. Magnotti LJ, Upperman JS, Xu DZ, Lu Q, Deitch EA: Gut-derived mesenteric lymph but not portal blood increases endothelial cell permeability and promotes lung injury after hemorrhagic shock. Ann Surg 228:518–527, 1998.
12. Adams Jr, CA Hauser CJ, Adams JM, Fekete Z, Xu DZ, Sambol JT, Deitch EA: Trauma-hemorrhage-induced neutrophil priming is prevented by mesenteric lymph duct ligation. Shock 18:513–517, 2002.
13. Dayal SD, Hauser CJ, Feketeova E, Fekete Z, Adams JM, Lu Q, Xu DZ, Zaets S, Deitch EA: Shock mesenteric lymph-induced rat polymorphonuclear neutrophil activation and endothelial cell injury is mediated by aqueous factors. J Trauma 52:1048–1055, 2002.
14. Wang YR, Xiao XZ, Huang SN, Luo FJ, You JL, Luo H, Luo ZY: Heat shock pretreatment prevents hydrogen peroxide injury of pulmonary endothelial cells and macrophages in culture. Shock 6:134–141, 1996.
15. Nathens AB, Neff MJ, Jurkovich GJ, Klotz P, Farver K, Ruzinski JT, Radella F, Garcia I, Maier RV: Randomized, prospective trial of antioxidant supplementation in critically ill surgical patients. Ann Surg 236:814–822, 2002.
16. Minamiya Y, Abo S, Kitamura M, Izumi K, Kimura Y, Tozawa K, Saito S. Endotoxin-induced hydrogen peroxide production in intact pulmonary circulation of rat. Am J Respir Crit Care Med 152s:348–354, 1995.
17. Minamiya Y, Saito S, Nakamura M, Tozawa K, Saito H, Matsuzaki I, Ogawa J: Nafamostat mesilate attenuates radical formation in the rat lung infused with endotoxin. Shock 18:255–260, 2002.
18. Szejda P, Parce JW, Seeds MS, Bass DA: Flow cytometric quantitation of oxidative product formation by polymorphonuclear leukocytes during phagocytosis. J Immunol 133:3303–3307, 1984.
19. Motoyama S, Minamiya Y, Saito S, Saito R, Matsuzaki I, Abo S, Inaba H, Enomoto K, Kitamura M: Hydrogen peroxide derived from hepatocytes induces sinusoidal endothelial cell apoptosis in perfused hypoxic rat liver. Gastroenterology 114:153–163, 1998.
20. Motoyama S, Saito S, Alojado ME, Ito H, Kitamura M, Suzuki H, Saito R, Momiyama H, Nakae H, Ogawa J, Inaba H: Hydrogen peroxide induces midzonal heat shock protein 72 and apoptosis in sinusoidal endothelial cells of hypoxic rat liver. Crit Care Med 28:1509–1514, 2000.
21. Motoyama S, Saito S, Inaba H, Kitamura M, Minamiya S, Suzuki H, Saito R, Kamata R, Nakae H, Ogawa J: Red blood cells attenuate sinusoidal endothelial cell injury by scavenging xanthine oxidase-dependent hydrogen peroxide in hyperoxic perfused rat liver. Liver 20:200–208, 2000.
22. Motoyama S, Saito S, Minamiya Y, Saito R, Nakamura M, Okuyama M, Imano H, Ogawa J: Methylprednisolone inhibits low flow hypoxia-induced mitochondrial dysfunction in isolated perfused rat liver. Crit Care Med 31:1468–1474, 2003.
23. Motoyama S, Saito S, Saito R, Minamiya Y, Nakamura M, Okuyama M, Imano H, Ogawa J: Hydrogen peroxide-dependent declines in Bcl-2 induces apoptosis in hypoxic liver. J Surg Res 110:211–216, 2003.
24. Green TP, Johnson DE, Marchessault RP, Gatto CW: Transvascular flux and tissue accrual of Evans blue: effects of endotoxin and histamine. J Lab Clin Med 111:173–183, 1988.
25. Saria A, Lundberg JM: Evans blue fluorescence: quantitative and morphological evaluation of vascular permeability in animal tissues. J Neurosci Methods 8:41–49, 1983.
26. Linderkamp O, Mader T, Butenandt O, Riegel KP: Plasma volume estimation in severely ill infants and children using a simplified Evans blue method. Eur J Pediatr 125:135–141, 1977.
27. Colletti LM, Kunkel SL, Walz A, Burdick MD, Kunkel RG, Wilke CA, Strieter RM: Chemokine expression during hepatic ischemia/reperfusion-induced lung injury in the rat. The role of epithelial neutrophil activating protein. J Clin Invest 95:134–141, 1995.
28. Goldblum SE, Wu KM, Jay M: Lung myeloperoxidase as a measure of pulmonary leukostasis in rabbits. J Appl Physiol 59:1978–1985, 1985.
29. Okamoto K, Nishino T: Mechanism of inhibition of xanthine oxidase with a new tight binding inhibitor. J Biol Chem 270:7816–7821, 1995.
30. Merker MP, Pitt BR, Choi AM, Hassoun PM, Dawson CA, Fisher AB: Lung redox homeostasis: emerging concepts. Am J Physiol Lung Cell Mol Physiol 279:L413–L417, 2000.
31. Deitch EA, Shi HP, Lu Q, Feketeova E, Xu da Z: Serine proteases are involved in the pathogenesis of trauma-hemorrhagic shock-induced gut and lung injury. Shock 19:452–456, 2003.
32. Grotz MR, Pape HC, van Griensven M, Stalp M, Rohde F, Bock D, Krettek C: Glycine reduces the inflammatory response and organ damage in a two-hit sepsis model in rats. Shock 16:116–121, 2001.