Endotoxin induces bacterial translocation and increases xanthine oxidase activity

Previously, we documented that endotoxin induces bacterial translocation from the gut and that inhibition or inactivation of xanthine oxidase activity reduces endotoxin-induced bacterial translocation. Consequently, experiments were performed to correlate endotoxin-induced bacterial translocation with changes in intestinal mucosal structure and xanthine dehydrogenase and oxidase activity. Segments of the jejunum, ileum, cecum, proximal colon, distal colon, and liver were harvested from ICR mice 24 hr after IP administration of E. coli 0111:B4 endotoxin (0.1 mg). Xanthine dehydrogenase and oxidase activities were measured in these samples and correlated with intestinal morphology. Bacteria translocated from the intestines to extraintestinal organs in 70% of the mice receiving endotoxin, while the organs of control mice were sterile (p less than 0.01). Endotoxin injured primarily the ileal and cecal mucosa and increased ileal and hepatic xanthine dehydrogenase and cecal oxidase activities (p less than 0.05). These results suggest that xanthine oxidase-induced mucosal damage plays a role in endotoxin-induced bacterial translocation.     

Endotoxin-induced bacterial translocation: a study of mechanisms

Previously, we documented that nonlethal doses of endotoxin cause the translocation (escape) of bacteria from the gut to systemic organs. The purpose of this study was to determine which portion(s) of the endotoxin molecule induces bacterial translocation and to examine the role of xanthine oxidase activity in the pathogenesis of endotoxin-induced bacterial translocation. Nonlethal doses of Salmonella endotoxin preparations (wild type, Ra, or Rb), containing the terminal portion of the core polysaccharide, induced bacterial translocation, whereas those preparations lacking the terminal-3 sugars (Rc, Rd, Re, or lipid A) did not induce bacterial translocation. Additionally, only those endotoxin preparations that induced bacterial translocation injured the gut mucosa, increased ileal xanthine dehydrogenase and oxidase activity, and disrupted the normal ecology of the gut flora, resulting in overgrowth with enteric bacilli. Inhibition of xanthine oxidase activity by allopurinol prevented endotoxin (Ra)-induced mucosal injury and reduced the incidence of bacterial translocation from 83% to 30% (p less than 0.01). These results suggest that endotoxin-induced bacterial translocation requires the presence of the terminal core lipopolysaccharide moiety and that xanthine oxidase-generated oxidants are important in the pathogenesis of endotoxin-induced mucosal injury and bacterial translocation.     

Role of bacterial adherence and the mucus barrier on bacterial translocation: effects of protein malnutrition and endotoxin in rats.

OBJECTIVE: The purpose of the study was to investigate the potential relations between mucosal bacterial adherence, intestinal mucus and mucin content, and bacterial translocation. SUMMARY BACKGROUND DATA: The attachment of bacteria to mucosal surfaces is the initial event in the pathogenesis of most bacterial infections that originate at mucosal surfaces, such as the gut. The intestinal mucus layer appears to function as a defensive barrier limiting micro-organisms present in the intestinal lumen from colonizing enterocytes. Consequently, studies focusing on the biology of bacterial adherence to the intestinal mucosa likely are to be important in clarifying the pathogenesis of gut origin sepsis. METHODS: To explore the relations between intestinal bacterial adherence, mucus bacterial binding, and bacterial translocation, two models were used. One (protein malnutrition) in which profound alterations in intestinal morphology occurs in the absence of significant translocation and one (endotoxin challenge) in which bacterial translocation occurs and intestinal morphology is relatively normal. RESULTS: Protein malnutrition was not associated with bacterial translocation and measurement of enteroadherent, mucosally associated bacterial population levels documented that the total number of gram-negative enteric bacilli adherent to the ileum and cecum was less in the protein-malnourished rats than in the normally nourished animals (p < 0.01). Furthermore, there was an inverse relation between the duration of protein malnutrition and bacterial adherence to the intestinal mucosa (r = 0.62, p < 0.002). In contrast, after endotoxin challenge, the level of enteroadherent bacteria was increased and bacterial translocation was observed. The binding of Escherichia coli to immobilized ileal mucus in vitro was decreased significantly in protein-malnourished rats, whereas E. coli binding to insoluble ileal mucus was increased in the rats receiving endotoxin. CONCLUSIONS: This study indicates that the adherence of bacteria to the intestinal mucosal surface is an important factor in bacterial translocation, that intestinal mucus modulates bacterial adherence, and that increased levels of mucosally associated bacteria are associated with a loss intestinal barrier function to bacteria.     

Genetic susceptibility to mucosal damage leads to bacterial translocation in a murine burn model

Since genetic factors may be important in host resistance to infections after thermal injury, we screened the susceptibility of three mouse strains (CD-1, Balb/c, and C57/bl) to thermally induced bacterial translocation from the GI tract. Bacteria translocated to the MLNs of Balb/c but not the CD-1 or C57/bl mice receiving 25% body burns. The increased incidence of bacterial translocation in the burned Balb/c mice appeared to be due to a burn-induced gut mucosal injury, since the intestinal mucosa of the Balb/c but not the CD-1 or C57/bl mice was damaged 24 hr after the thermal injury. The mucosal injury appears to be mediated, at least in part, by xanthine oxidase-generated oxygen-free radicals, since inhibition of xanthine oxidase activity with allopurinol, or inactivation of xanthine oxidase activity by a molybdenum-free tungsten diet, prevented the mucosal injury and reduced the extent of bacterial translocation.     

Role of neutrophil and hydroxyl radical in shock-induced gut origin infection]

The relative roles of hydroxyl radical and neutrophils in the pathogenesis of shock-induced mucosal injury and gut origin infection (GOI) were determined. The incidence of GOI was higher in the shocked rats (30 mmHg for 30 min) than the sham-shock controls (87% vs 12.5%; P less than 0.01). Administration of the hydroxyl radical scavenger, dimethyl sulfoxide (DMSO) or iron chelator and deferoxamine reduced the incidence of GOI from 87% to 20% and 40% respectively (P less than 0.05). DMSO and deferoxamine appeared to prevent shock-induced GOI by blunting the magnitude of shock-induced mucosal injury. In contrast, neutrophil depletion did not prevent GOI or protect the intestinal mucosal in the shocked rats. Instead, the incidence of systemic spread of bacteria past the mesenteric lymph nodes to the livers and spleens of the shocked rats was higher in the neutrophil depleted rats (56%) than any other group (7%) (P less than 0.01). Thus, shock-induced GOI and intestinal injury appears to be mediated by xanthine oxidase generated oxidants such as hydroxyl radical rather than neutrophil-generated factors. In addition, neutrophil depletion may be clinically deleterious, since it promotes systemic sepsis rather than preventing shock-induced GOI.     

Protective Effect of a Lipoxygenase Inhibitor, Nordihydroguaretic Acid, on the Ileal Motor Disturbances Induced by Serratia marcescens Endotoxin in Rats

This study was conducted to examine the effects of peptidoleukotrienes on the ileal contractility disturbances induced by Serratia marcescens endotoxin in rats. Thirty-two male Wistar rats were divided into four groups (n = 8 each). The first group was given only an anesthetic agent (control group); the second group was given the endotoxin (endotoxin group); the third group was given a lipoxygenase inhibitor, nordihydroguaretic acid (NDGA); and the fourth group was given NDGA 10 min before administration of the endotoxin (NDGA+endotoxin group). The isolated ileum response was recorded in each group. Normal contractile activity was seen in the control group. After the endotoxin was given, the isolated ileum did not respond to acetylcholine (ACh) in the endotoxin group, but the contractile results of isolated ileum to ACh were similar to the control group results in both the NDGA and endotoxin+NDGA groups. The results of this study demostrate that leukotrienes may play a role in endotoxin-induced ileal contractility disturbances, and that the lipoxygenase inhibitor, NDGA, could be useful for the treatment of ileal motility disturbances induced by endotoxin. Received: April 24, 2000 / Accepted: November 20, 2000     

A study of the relationship among survival, gut-origin sepsis, and bacterial translocation in a model of systemic inflammation.

Several factors, including uncontrolled inflammation, gut barrier failure, and sepsis, have been implicated in the development of multiple organ failure. To investigate the relative importance and interrelationships among some of these factors, increasing doses of the inflammatory agent zymosan were used to induce a systemic inflammatory state in mice. At nonlethal doses (0.1 and 0.5 mg/g body weight), zymosan caused injury to the intestinal mucosa, increased intestinal xanthine oxidase activity, and promoted bacterial translocation in a dose-dependent fashion. Inhibition or inactivation of xanthine oxidase activity was effective in reducing mucosal injury and bacterial translocation when zymosan was injected at 0.1 mg/g but not at 0.5 mg/g body weight. At a dose of 1 mg/g, the lethal effects of zymosan appeared to be related to gut-origin sepsis, since cefoxitin (1 mg/g) reduced the seven-day mortality rate from 100% to 20% (p less than 0.01). However, at a zymosan dose of 2 mg/g, antibiotics did not improve survival. Zymosan thus induced gut barrier failure and systemic infection in a dose-dependent fashion. Additionally, the mechanism of zymosan-induced bacterial translocation and the relationship of gut-origin sepsis to survival appeared to be related to the magnitude of the inflammatory insult (the dose of zymosan).     

Endotoxin promotes the translocation of bacteria from the gut

Experiments were performed in mice to determine whether endotoxin could cause bacteria normally colonizing the gut to spread systemically, a process termed bacterial translocation. Endotoxin given intraperitoneally promoted bacterial translocation in a dose-dependent fashion from the gut to the mesenteric lymph node (MLN). The incidence of bacterial translocation to the MLN was similar whether the endotoxin was administered intramuscularly or intraperitoneally, although the number of bacteria colonizing the MLN was greater with intraperitoneal endotoxin. The incidence and magnitude of endotoxin-induced bacterial translocation were similar between CD-1 and C3H/HeJ (endotoxin-resistant) mice, indicating that bacterial translocation is not prevented by genetic resistance to endotoxin. Thus, it appears that the gut may serve as a reservoir for bacteria causing systemic infections during endotoxemia.     

Modification of oxidative stress in response to intestinal preconditioning

Previous studies have demonstrated that intestinal preconditioning protects the organ from ischemia reperfusion damage. Xanthine oxidase mediating free radical generation contributes to the development of injury associated to ischemia reperfusion. Thus, any process able to modulate the oxygen free radical generation system could attenuate the injury. Also, it is known that nitric oxide is implicated in the preconditioning response. The aim of this work is to determine: (1) the effect of intestinal preconditioning on the xanthine oxidase system, (2) the relevance of this system in the development of injury, and (3) its relationship with nitric oxide. For this purpose, we have determined the activity of the xanthine dehydrogenase/xanthine oxidase system, the levels of its substrate (xanthine), and end-product (uric acid) and oxidant stress status in rat small intestine subjected to ischemic pre-conditioning. The effects of nitric oxide inhibition have also been evaluated. Results show that the percentage of xanthine dehydrogenase to xanthine oxidase conversion, xanthine, uric acid concentration, lipoperoxides, and reduced glutathione were significantly reduced in preconditioned rats irrespectively of nitric oxide inhibition. In summary, this work shows that oxidative stress in intestinal preconditioning is reduced as consequence of the diminished conversion of xanthine dehydrogenase to xanthine oxidase, and also as a consequence of the reduced availability of xanthine.     

Minimal role of xanthine oxidase and oxygen free radicals in rat renal tubular reoxygenation injury

The role of xanthine oxidase and oxygen free radicals in postischemic reperfusion injury in the rat kidney remains controversial. Proximal tubules, the focal segment affected by ischemic renal injury, were isolated in bulk, assayed for xanthine oxidase activity, and subjected to 60 min of anoxia or hypoxia and 60 min of reoxygenation to evaluate the participation of xanthine oxidase and oxygen radicals in proximal tubule reoxygenation injury. The total xanthine oxidase in isolated rat proximal tubules was 1.1 mU/mg of protein, approximately 30% to 40% of the activity found in rat intestine and liver. Lactate dehydrogenase release, an indicator of irreversible cell damage, increased substantially during anoxia (39.8 +/- 2.3 versus 9.8 +/- 1.8% in controls) with an additional 8 to 12% release during reoxygenation. Addition of 0.2 mM allopurinol, a potent xanthine oxidase inhibitor, and dimethylthiourea, a hydroxyl radical scavenger, failed to protect against the reoxygenation lactate dehydrogenase release. Analysis of xanthine oxidase substrate levels after anoxia and flux rates during reoxygenation indicates that hypoxanthine and xanthine concentrations are in a 15-fold excess over the enzyme Km and 0.3 mU/mg of protein of xanthine oxidase activity exists during reoxygenation. Hypoxic tubule suspensions had a minimal lactate dehydrogenase release during hypoxia and failed to demonstrate accelerated injury upon reoxygenation. In conclusion, although xanthine oxidase is present and active during reoxygenation in isolated rat proximal tubules, oxygen radicals did not mediate reoxygenation injury.     

Xanthine:acceptor oxidoreductase activities in ischemic rat skin flaps

Xanthine:acceptor oxidoreductase activities were assayed in free skin flaps following prolonged preservation. In normal rat skin, xanthine dehydrogenase transfers electrons to NAD+ and accounts for 73% of total oxidoreductase activity, and xanthine oxidase transfers electrons to molecular oxygen and accounts for the remaining 27%. Xanthine oxidase activity increased significantly in skin flaps during ischemia: approximately 30 and 100% increases after 6 and 24 hr of ischemia, respectively. Allopurinol inhibited xanthine oxidoreductase activity: free skin flaps obtained from allopurinol-treated animals exhibited a low level of xanthine oxidoreductase activity throughout the period of preservation. Systemic allopurinol significantly improved the survival rate from 32 to 75% of free flaps transferred after 24 hr of preservation at room temperature. These observations suggest that the xanthine oxidase system is a major source of oxygen free radicals following ischemia/reperfusion in skin. The increase in xanthine oxidase is attributable to the conversion of xanthine dehydrogenase to oxidase, a conversion which involves sulfhydryl oxidation in skin flaps.     

Milrinone improves intestinal villus blood flow during endotoxemia

PURPOSE: To determine whether the compromised intestinal villus blood flow in a rat model of endotoxemia could be improved by continuous infusion of the phosphodiesterase (PDE) inhibitor milrinone. METHODS: Twenty-four anesthetized and ventilated rats were laparotomized and an ileal portion was exteriorized and opened by an antimesenteric incision. The ileal segment was fixed with the mucosal surface upward. Microcirculatory parameters were assessed by intravital videomicroscopy. The animals were randomly assigned to receive one of three treatments: infusion of Escherichia coli lipopolysaccharides without phosphodiesterase inhibitor pretreatment (=LPS group); or infusion of LPS with milrinone pretreatment (= milrinone group), or without infusion of LPS or milrinone (=control group). Macrohemodynamic parameters (MAP, HR) and microhemodynamic parameters of ileal mucosa (mean diameter of central arterioles = D(A) and mean erythrocyte velocity within the arterioles= V(E)) were measured 30 min before and at 0, 60, and 120 min after induction of endotoxemia. Mucosal villus blood flow was calculated from D(A) and V(E). RESULTS: In the milrinone group MAP decreased 60 min after induction of endotoxemia whereas it remained stable in the control and the LPS group. In both groups given endotoxin V(E) decreased after start of LPS infusion. In contrast, D(A) decreased in the LPS group, but increased in the milrinone group after 120 min of endotoxemia. Thus, the endotoxin-induced decrease of intestinal villus blood flow was diminished but not fully restored by milrinone infusion. CONCLUSION: Our results indicate that milrinone has some beneficial microcirculatory effects during endotoxemia. Although it contributed to systemic hypotension, it attenuated intestinal mucosal hypoperfusion.     

Splenectomy influences endotoxin-induced bacterial translocation

To determine whether splenectomy affects the antibacterial defenses of the gut, experiments were performed using bacterial translocation (BT) as a marker of intestinal barrier failure. The incidence of BT was measured 8 days after splenectomy or sham-splenectomy in mice receiving or not receiving endotoxin (0.1 mg IP). Splenectomy does not appear to promote BT from the gut, since the incidence of bacterial translocation after splenectomy or sham-splenectomy (5%) were not different. A second experiment was performed to determine whether the resistance to endotoxin-induced BT was modified after splenectomy. The incidence of endotoxin-induced BT was 73% in the unoperated control group, 59% in the sham-splenectomy group, but 23% in the splenectomy group (p less than 0.002). Thus, splenectomy but not sham-splenectomy increased the resistance of otherwise healthy mice to endotoxin-induced BT.     

急性坏死性胰腺炎时肠屏障损害及肠源性细菌和内毒素移位的实验研究

目的 观察急性坏死性胰腺炎（ANP）时肠粘膜屏障的改变和肠源性细菌和内毒素移位。方法 将Wistar大鼠随机分为正常对照组（n=6)、假手术组（n=30)和ANP组（n=39)。采用人工胆汁胰管逆行灌注法制作ANP模型。观察胰腺、小肠病理改变和小肠粘膜上皮细胞间紧密连接（冷冻蚀刻电镜）变化。动态测定血浆D-乳酸、内毒素水平,以及腹腔脏器细菌移位率。结果 ANP后小肠粘膜损伤,皮皮细胞间紧密连接破坏甚至消失,血浆D-乳酸水平上升,发病早期即出现内毒素血症;ANP发生后72h脏器细菌移位率达到59.5%。结论 ANP早期肠粘膜屏障功能受损。导致肠道细菌和内毒素移位,成为全身炎症反应和胰腺继发感染的根源。     

Differential pathophysiology of bacterial translocation after thermal injury and sepsis.

Bacterial translocation (BT) occurs transiently after thermal injury and may result from an ischemic intestinal insult. To evaluate continued intestinal ischemia in the ongoing BT associated with sepsis after injury, rats were randomized to (1) 30% burn injury with Pseudomonas wound infection (BI), (2) BI + fluid resuscitation (BI + Fluid), (3) BI after allopurinol pretreatment to inhibit xanthine oxidase (BI + Allo), or (4) BI after azapropazone pretreatment to inhibit neutrophil degranulation (BI + Aza). On postburn days (PBD) 1, 4, and 7, animals were studied for evidence of BT and intestinal lipid peroxidation. BI + Fluid, BI + Allo, and BI + Aza significantly (p less than 0.05) reduced rates of BT and ileal lipid peroxidation acutely after thermal injury (PBD 1) compared to BI. All four groups had equally high rates of BT associated with the onset of sepsis (PBDs 4 and 7), without evidence of further intestinal lipid peroxidation. These data indicate that the chronic gut barrier failure associated with sepsis after injury occurs independently of continued intestinal ischemia.     

急性坏死性胰腺炎并发感染的机理研究

目的观察急性坏死性胰腺炎时肠屏障损伤与细菌移位情况,探讨急性胰腺炎继发感染的机理。方法１５只犬于肠道内定植ＰＵＣ１８质粒菌ＪＭ１０９后,分对照组（ｎ＝７）和急性坏死性胰腺炎组（ＡＮＰ,ｎ＝８）。ＡＮＰ组经主胰管注入牛磺胆酸钠和胰蛋白酶制作ＡＮＰ模型。结果ＡＮＰ组较对照组血胰淀粉酶显著升高（Ｐ＜０．０１）;尿中乳果糖／甘露醇比值高出对照组２～１２倍;空、回、盲肠粘膜及肠内容物中大肠杆菌明显增加（Ｐ＜０．０１）,双歧杆菌、乳酸杆菌明显减少（Ｐ＜０．０１、Ｐ＜０．０５）。对照组犬血培养阴性,除２只犬肠系膜淋巴结培养出细菌外,其余脏器培养均阴性。ＡＮＰ组犬血和脏器细菌培养阳性率均为１００％,且每只犬都能检出术前定植于肠道的质粒菌ＪＭ１０９;胰腺腺泡出血、坏死;肠粘膜绒毛破坏;血浆、回肠组织二胺氧化酶活性下降。结论ＡＮＰ时肠粘膜屏障功能严重受损,发生肠道细菌移位,成为继发性胰腺感染的潜在根源。     

Changes in xanthine oxidase in ischemic rat brain

Xanthine oxidase activity in the rat brain was measured by means of high-performance liquid chromatography with electrochemical detection of uric acid. Cerebral ischemia was produced by a four-vessel occlusion method. In the control rat, the enzyme activity was 0.87 +/- 0.13 nmol/gm wet weight/min at 25 degrees C (mean +/- standard deviation), of which 92.4% was associated with the nicotinamide adenine dinucleotide (NAD)-dependent dehydrogenase form and only 7.6% with the oxygen-dependent superoxide-producing oxidase form. However, the ratio of the latter form increased to 43.7% after 30 minutes of global ischemia, despite the total xanthine oxidase activity remaining the same. Thus, it was revealed that uric acid can be synthesized in the rat brain and that cerebral ischemia induced the conversion of xanthine oxidase from an NAD-dependent dehydrogenase to an oxygen-dependent superoxide-producing oxidase. Although the xanthine oxidase pathway has been proposed as a source of oxygen-derived free radicals in various ischemic organs other than brain, the results of the present study suggest the involvement of the oxygen free radicals generated from this pathway in the pathogenesis of the ischemic injury of the rat brain.     

Role of gut-derived endotoxaemia and bacterial translocation in rats after thermal injury: effects of selective decontamination of the digestive tract

This study was performed to investigate: (1) the role of gut-derived endotoxin/bacterial translocation in the pathogenesis of sepsis, and (2) the possible effects of selective decontamination of the digestive tract (SDD) on mortality in rats following 40 per cent full-thickness scald injury. In the SDD-treated group, Enterobacteriaceae and yeasts were eradicated from the caecal mucosa, while the mucosal flora consisting of mainly anaerobes was well preserved, within 3 days. The incidence of bacterial translocation to the mesenteric lymph nodes (MLN) and viscerae was significantly lowered on postburn days 1, 3 and 5 (P < 0.05−0.01). Meanwhile, pretreatment with SDD resulted in reductions of the faecal endotoxin levels in different segments of intestinal tract to less than 0.5 per cent (0.04 – 0.45 per cent) of the untreated control; there was also a significant attenuation of the elevation of endotoxin concentrations in both portal and systemic blood. Intestinal diamine oxidase (DAO) activity returned to baseline on day 5 in rats receiving SDD but not in controls. The 5-day survival rate in the SDD-treated group was elevated by 26.7 per cent as compared with controls (P < 0.05). These data suggested that endotoxin/bacterial translocation took place early and commonly, which in turn contributed to postburn sepsis and mortality. SDD was effective in preventing gut origin endotoxaemia and bacterial translocation, and improving the survival rate in rats following severe thermal injuries.     

Differential effects of gut hormones on pancreatic and intestinal growth during administration of an elemental diet.

Liquid elemental diets (ED) will, in time, cause atrophy of the gut. Pentagastrin (PG), neurotensin (NT), and bombesin (BBS) are peptides that have trophic effects on the gut of normal rats. This study examined the effect of these three agents on gut atrophy produced by ED. Four groups of rats were given an ED and injected with either saline (control), PG (250 micrograms/kg), NT (300 micrograms/kg), or BBS (10 micrograms/kg) subcutaneously every 8 hours for 5 or 10 days. A fifth group was fed rat chow ad libidum. The rats were killed on day 6 or 11; the pancreas and segments of small intestine were removed. Atrophy of ileal mucosa was apparent on days 6 and 11, and atrophy of jejunal mucosa was manifest by day 11. Bombesin prevented jejunal mucosal atrophy and significantly increased ileal mucosal growth (compared with control). Neurotensin prevented the jejunal, but not the ileal, mucosal atrophy produced by ED. Pentagastrin had no effect on gut mucosa. Bombesin and PG, but not NT, stimulated pancreatic growth. Neurotensin stimulates pancreaticobiliary secretions (PBS), which are known to stimulate gut growth. Jejunoileal bypass was performed to determine whether trophic effects of NT on gut mucosa were mediated through stimulation of PBS. After 1 week treatment, animals were killed and segments of intestine removed. As expected NT was trophic for gut mucosa in continuity with the luminal stream; furthermore NT produced significant stimulation of growth of gut mucosa in the bypassed segment. We conclude that both BBS and NT are trophic for intestinal mucosa of rats given ED; both agents have a more pronounced effect on jejunum. The trophic effect of NT is mediated, in part, by a mechanism unrelated to stimulation of PBS. Bombesin and NT may have important regulatory functions in the adaptive growth of small bowel mucosa and in the maintenance of gut mucosal integrity.     

Is Ischemic Preconditioning a Useful Strategy in Steatotic Liver Transplantation?

This study examined the effect of preconditioning on steatotic livers for transplantation and attempted to identify the underlying protective mechanisms. Blood flow alterations, neutrophil accumulation, tumor necrosis factor alpha release and lipid peroxidation were observed in nonsteatotic livers after transplantation. Steatotic and nonsteatotic liver grafts were similar in their blood flow, neutrophil accumulation, and TNF release after transplantation. However, in the presence of steatosis, lipid peroxidation and hepatic injury increased. In addition, recipients of steatotic liver grafts were more vulnerable to lung damage associated with transplantation. The conversion of xanthine dehydrogenase to xanthine oxidase and the accumulation of xanthine during cold ischemia was greater in steatotic than in nonsteatotic liver grafts. The results obtained with xanthine oxidase inhibitors indicated that xanthine/xanthine oxidase could be responsible for the increased lipid peroxidation as well as the exacerbated liver and lung damage associated with transplantation of steatotic livers. Preconditioning reduced the xanthine accumulation and percentage of xanthine oxidase seen in steatotic liver grafts during cold ischemia, and conferred protection against liver and lung damage following transplantation. The benefits of preconditioning could be mediated by nitric oxide. These findings suggest that preconditioning could be a relevant new strategy to protect against the inherent risk of steatotic liver failure following transplantation.