Steatohepatitis/Metabolic Liver Disease
Fibrin accumulation plays a critical role in the sensitization to lipopolysaccharide-induced liver injury caused by ethanol in mice†
Article first published online: 23 JAN 2009
Copyright © 2009 American Association for the Study of Liver Diseases
Volume 49, Issue 5, pages 1545–1553, May 2009
How to Cite
Beier, J. I., Luyendyk, J. P., Guo, L., von Montfort, C., Staunton, D. E. and Arteel, G. E. (2009), Fibrin accumulation plays a critical role in the sensitization to lipopolysaccharide-induced liver injury caused by ethanol in mice. Hepatology, 49: 1545–1553. doi: 10.1002/hep.22847
Potential conflict of interest: Nothing to report.
- Issue published online: 27 APR 2009
- Article first published online: 23 JAN 2009
- Accepted manuscript online: 23 JAN 2009 12:00AM EST
- Manuscript Accepted: 11 JAN 2009
- Manuscript Received: 9 MAY 2008
- NIAAA. Grant Number: R01 AA003624
- COBRE. Grant Number: P20 RR021940
The early stages of alcohol-induced liver injury involve chronic inflammation. Whereas mechanisms by which this effect is mediated are not completely understood, it is hypothesized that enhanced sensitivity to circulating lipopolysaccharide (LPS) contributes to this process. It has recently been shown that ethanol induces activation of plasminogen activator inhibitor-1 (PAI-1). PAI-1 causes fibrin accumulation in liver by inhibiting degradation of fibrin (fibrinolysis). LPS also enhances fibrin accumulation by activating the coagulation cascade. It was therefore hypothesized that ethanol will synergistically increase fibrin accumulation caused by LPS, enhancing liver damage. Accordingly, the effect of ethanol pretreatment on LPS-induced liver injury and fibrin deposition was determined in mice. Ethanol enhanced liver damage caused by LPS, as determined by plasma parameters and histological indices of inflammation and damage. This effect was concomitant with a significant increase in PAI-1 expression. Extracellular fibrin accumulation caused by LPS was also robustly increased by ethanol preexposure. Coadministration of the thrombin inhibitor hirudin or the MEK (mitogen-activated protein kinase) inhibitor U0126 significantly attenuated the enhanced liver damage caused by ethanol preexposure; this protection correlated with a significant blunting of the induction of PAI-1 caused by ethanol/LPS. Furthermore, thrombin/MEK inhibition prevented the synergistic effect of ethanol on the extracellular accumulation of fibrin caused by LPS. Similar protective effects on fibrin accumulation were observed in tumor necrosis factor receptor 1 (TNFR-1)−/− mice or in wild-type injected with PAI-1-inactivating antibody. Conclusion: These results suggest that enhanced LPS-induced liver injury caused by ethanol is mediated, at least in part, by fibrin accumulation in livers, mediated by an inhibition of fibrinolysis by PAI-1. These results also support the hypothesis that fibrin accumulation may play a critical role in the development of early alcohol-induced liver injury. (HEPATOLOGY 2009.)
Alcohol-induced liver injury ranks among the major causes of morbidity and mortality in the world1 and affects millions of patients worldwide each year. Progression of the disease is well characterized and is actually a spectrum of liver diseases, which ranges initially from simple steatosis, to inflammation and necrosis (steatohepatitis), to fibrosis and cirrhosis. Although the progression of alcohol-induced liver injury is well characterized, there is no universally accepted therapy available to halt or reverse this process in humans. With better understanding of the mechanism(s) and risk factors that mediate the initiation and progression of this disease, rational targeted therapy can be developed to treat or prevent it in clinics.
Chronic inflammation is a hallmark of the progression of alcohol-induced liver injury. A key concept in the hepatic inflammatory response during this disease is priming and sensitization. Specifically, inflammatory cells are affected by alcohol so that they respond more robustly to an inflammatory insult (i.e., “primed”). Furthermore, target cells respond more robustly to products produced by inflammatory cells (i.e., “sensitized”). Inflammatory cytokines (e.g., tumor necrosis factor alpha [TNFα]) play key roles in both priming and sensitization in experimental alcohol-induced liver injury. For example, macrophages from alcohol-exposed animals produce more TNFα after stimulus and TNFα is more cytotoxic to hepatocytes from alcohol-exposed animals.2, 3 However, the potential roles of other aspects of the inflammatory response in alcohol-induced liver injury have been less studied.
Previous work by this group has shown that plasminogen activator inhibitor-1 (PAI-1) may mediate inflammatory effects in vivo. Experiments with PAI-1−/− mice exposed to chronic enteral ethanol showed that, in addition to blunting steatosis, genetic inhibition of PAI-1 expression also conferred profound antiinflammatory effects.4 Indeed, whereas knocking-out PAI-1 partially blunted the steatotic changes caused by ethanol, there was almost complete protection against the inflammatory changes caused by alcohol in this strain. A similar antiinflammatory effect of knocking-out PAI-1 was observed by others in a mouse model of glomerulonephritis.5 Whereas PAI-1, the main inhibitor of fibrinolysis, is well known to be induced during inflammation, how PAI-1 may actually contribute to inflammatory processes is less understood. One mechanism by which PAI-1 may contribute to inflammation is through its “classic” role of impairing fibrinolysis.6 The purpose of the current study was to test the hypothesis that ethanol enhances inflammatory liver damage caused by lipopolysaccharide (LPS) through altered fibrin degradation.
Materials and Methods
Animals and Treatments.
Six-week-old male C57BL/6J, B6.129-Tnfrsf1atm1Mak/J (TNFR-1−/−) mice were purchased from the Jackson Laboratory (Bar Harbor, ME). Mice were housed in a pathogen-free barrier facility accredited by the Association for Assessment and Accreditation of Laboratory Animal Care, and procedures were approved by the University of Louisville's Institutional Animal Care and Use Committee. Food and tap water were allowed ad libitum. Animals received ethanol (6 g/kg i.g.) or isocaloric/isovolumetric maltose-dextrin solution for 3 days.7 LPS (E. coli, serotype 055:B5; Sigma, St Louis, MO; 10 mg/kg i.p.) was injected 24 hours after the last ethanol administration as described.7 Hirudin (Refludan, Berlex, Montville, NJ, 1 mg/kg s.c.) or vehicle (saline) was given 30, 150, and 270 minutes after LPS.8 U0126 (Calbiochem, La Jolla, CA, 10 mg/kg i.p.) or vehicle (dimethyl sulfoxide) was administered 60 minutes after LPS. Some mice were injected with PAI-1-inactivating antibody (Cisthera Inc., Seattle WA) i.p. (200 μL/mouse; ≈10 mg/kg) 30 minutes prior to injection with LPS. Mice were anesthetized with ketamine/xylazine (100/15 mg/kg, i.m.) at select times up to 48 hours after injection with LPS (see timeline in Fig. 1). Blood was collected from the vena cava just prior to sacrifice by exsanguination and citrated plasma was stored at −80°C for further analysis. Portions of liver tissue were snap-frozen in liquid nitrogen, frozen-fixed in OCT-Compound (Sakura Finetek, Torrance, CA), or were fixed in 10% neutral buffered formalin for subsequent sectioning and mounting on microscope slides.
Clinical Analyses and Histology.
Plasma levels of aminotransferases (alanine aminotransferase [ALT] and aspartate aminotransferase [AST]) and hepatic levels of triglycerides were determined using standard kits (Thermotrace, Melbourne, Australia).4 Paraffin-embedded sections were stained for hematoxylin and eosin (H&E); pathology was scored in a blinded manner by a trained pathologist.8 Oil Red O staining,4 chloroacetate esterase (CAE) staining, and neutrophil accumulation in the livers were assessed as described.8, 9 The intensity and extent of CAE staining in liver tissues were quantified by counting CAE-positive neutrophils per 1000 hepatocytes. Plasma thrombin-antithrombin (TAT) concentration was determined by enzyme-linked immunosorbent assay using a kit (Dade Behring Inc., Deerfield, IL).10 The accumulation of fibrin matrices was determined immunofluorometrically as described.8
RNA Isolation and Real-Time Reverse-Transcriptase Polymerase Chain Reaction (RT-PCR).
RNA extraction and real-time RT-PCR was performed as described.4 PCR primers and probes were designed using Primer 3 (Whitehead Institute for Biomedical Research, Cambridge, MA). Primers were designed to cross introns to ensure that only complementary (c)DNA and not genomic DNA was amplified (see Table 1 for sequences).
|Forward (3′-5′)||Reverse (3′-5′)||Probe (3′-5′)|
Liver samples were homogenized in RIPA buffer (20 mM Tris/Cl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% [wt/vol] Triton X-100), containing protease and phosphatase inhibitor cocktails (Sigma). Samples were loaded onto sodium dodecyl sulfate (SDS)-polyacrylamide gels of 10% (wt/vol) acrylamide followed by electrophoresis and Western blotting onto PVDF membranes (Hybond P, GE Healthcare, Piscataway, NJ). Primary antibodies against phosphorylated and total ERK1/2 (Cell Signaling Technology, Beverly, MA) were used. Bands were visualized using an ECL kit (Pierce, Rockford, IL) and Hyperfilm (GE Healthcare). Densitometric analysis was performed using ImageQuant (GE Healthcare) software.
Analysis of variance (ANOVA) with Bonferroni's post-hoc test or the Mann-Whitney Rank Sum test was used for the determination of statistical significance among treatment groups, as appropriate. Results are reported as means ± SEM (n = 4-6).
Effect of Ethanol on Hepatic Lipid Accumulation and Triglycerides.
Figure 1 shows representative photomicrographs depicting lipid accumulation in the liver (Oil Red O stain, middle) and a time course (bottom) of hepatic triglycerides in the current model. As has been observed previously,4 lipid staining and triglyceride levels in livers from maltose-dextrin-treated animals were minimal and did not differ from staining in livers from naïve control mice. In contrast, ethanol treatment caused significant accumulation of lipid droplets and triglycerides in livers of wild-type animals increased up to 12 hours after administration (Fig. 1, time-point −12 hours). By the time of LPS injection (t = 0; 24 hours after ethanol exposure), lipid levels had almost returned to basal levels (Fig. 1).
Effect of Ethanol on Liver Damage Caused by LPS.
Plasma levels of indices of liver damage (AST, ALT) were within normal ranges in mice exposed to maltose-dextrin in the absence of LPS (Fig. 2); ethanol alone did not significantly alter AST and ALT levels compared to these control animals. LPS injection alone significantly increased the level of ALT and AST released into the plasma 24 hours after injection (Fig. 2). As has been observed previously,7 pretreatment with ethanol significantly enhanced the increase in plasma ALT and AST caused by LPS at 24 hours by a factor of ≈3 (Fig. 2).
Figure 3 shows representative photomicrographs depicting liver pathology (H&E stain, left column) and neutrophil accumulation (CAE stain, right column) 24 hours after injection with LPS. No pathological changes were observed in liver tissue after maltose-dextrin or ethanol exposure alone (Fig. 4, top). Photomicrographs of livers from maltose-dextrin-injected mice are shown to represent both of the groups (“Control”; Fig. 3, top). LPS at this dose caused no gross morphological changes to liver (Fig. 3, left column, middle), but increased the inflammatory and necrosis scores slightly (Fig. 4, top) and the number of infiltrating neutrophils (Fig. 3, right column, middle; Fig. 4, bottom). The combination of ethanol and LPS increased hepatic damage, with necroinflammatory foci now detectable macroscopically (Fig. 3, left column, bottom), which lead to significantly increased pathology scores (Fig. 4, top). Ethanol also significantly enhanced the effect of LPS (≈2.5-fold) on the recruitment of neutrophils to the liver (Fig. 3, bottom right; Fig. 4, bottom).
Effect of LPS and Ethanol on Hepatic Fibrin Deposition and Circulating TAT Levels.
LPS is known to activate the coagulation system, which can lead to fibrin deposition and subsequent hemostasis.11 The effect of LPS and ethanol on hepatic fibrin deposition was therefore determined. Figure 5 comprises representative confocal photomicrographs depicting immunofluorescent detection of fibrin deposition. LPS caused fibrin deposition in sinusoidal spaces of the liver lobule (Fig. 5, top right). Ethanol alone did not cause a visible increase in fibrin deposition. However, ethanol treatment dramatically exacerbated accumulation of fibrin caused by LPS (Fig. 5, middle left).
To determine whether enhanced fibrin deposition was a consequence of exaggerated thrombin generation, plasma TAT levels were determined (Fig. 5, middle right) as an index of thrombin activation. Plasma TAT levels were significantly increased by LPS alone (Fig. 5, middle right). However, ethanol pretreatment did not alter the increase in plasma TAT concentrations caused by LPS. Despite the lack of difference between indices of coagulation, selective inhibition of thrombin with the specific thrombin inhibitor hirudin completely prevented the increase in fibrin accumulation (Fig. 5, bottom left) and liver damage (Figs. 2 and 4, bottom) caused by ethanol after LPS injection, with values for the latter variables similar to that of LPS alone.
Effect of Ethanol on the Induction of Gene Expression Caused by LPS.
Fibrin accumulation may be enhanced not only by increasing deposition through the coagulation cascade, but also by impairing degradation through fibrinolysis (see Fig. 9). Because thrombin activation (TAT levels, Fig. 5) was not elevated, the effect of ethanol on indices of the latter pathway (i.e., fibrinolysis) was determined. Specifically, the effect of LPS and ethanol on hepatic expression of PAI-1 and TNFα (a known upstream inducer of PAI-1)12 was quantitated (Fig. 6). Ethanol preexposure alone had no effect on hepatic mRNA expression of PAI-1 or TNFα under these conditions. LPS alone significantly induced hepatic expression of PAI-1 and TNFα as early as 1 hour after LPS and was still induced after 48 hours. Exposure to ethanol prior to LPS enhanced the increase in PAI-1 mRNA expression caused by LPS at the 4- and 24-hour timepoints. In contrast, TNFα mRNA expression was not significantly enhanced by ethanol at the 4-hour timepoint, and was significantly decreased at the 24-hour timepoint.
Effect of LPS and Ethanol on the Activation of ERK1/2.
Both inflammation and cell death were enhanced by ethanol in response to LPS. It is known that mitogen-activated protein kinase (MAPK) signaling cascades are involved in both processes. Furthermore, MAPK (i.e., ERK1/2) is a known upstream mediator of PAI-1 induction.12 Previous studies by others have also shown that ERK1/2 activation in response to LPS is enhanced by alcohol preexposure in rodent liver.13 The effect of LPS and ethanol exposure on the phosphorylation (activation) status of ERK1/2 was therefore studied; representative blots (Fig. 7, top) and densitometric analysis (Fig. 7, bottom) are shown. LPS alone caused an increase in ERK1/2 activation at the 4-hour timepoint. Whereas ethanol administration alone did not significantly alter ERK1/2 phosphorylation at any timepoint, it significantly enhanced LPS-induced ERK1/2 protein phosphorylation (≈2-fold). Treatment with the upstream MAPK inhibitor U0126 significantly blocked the induction of ERK1/2 protein phosphorylation (Fig. 7). Hirudin coadministration did not significantly attenuate the increase in ERK phosphorylation caused by the combination of LPS and ethanol, with an average value 3.7-fold higher than control. Analogous to the effect of hirudin on fibrin accumulation (Fig. 5, bottom left), U0126 treatment not only blunted the accumulation of fibrin (Fig. 5, bottom right), but also blunted the increase in LPS-induced liver damage caused by ethanol (Figs. 2 and 4, bottom). Whereas it had no significant effect on TNFα mRNA expression, U0126 prevented the increase in PAI-1 expression caused by ethanol in the presence of LPS at the peak timepoint (4 hours); indeed, the values were similar to those of LPS alone (Fig. 6).
Effect of LPS and Ethanol in TNFR-1−/− Mice and Mice Treated with PAI-1 Antibody.
The data presented above support a mechanism by which TNFα (by way of ERK1/2) induces PAI-1; this induction inhibits fibrinolysis, causing fibrin to accumulate (see Fig. 9). The effect of ethanol and LPS on liver damage and fibrin accumulation in TNFR-1−/− mice or in mice injected with PAI-1 antibodies was determined (Fig. 8) to directly test this hypothesis. Analogous to findings with hirudin and U0126 (Fig. 4), the enhancement of LPS-induced liver damage caused by ethanol was almost completely abrogated in TNFR-1 knockout mice or mice treated with PAI-1 antibody (Fig. 8). Specifically, the increase in the pathology scores caused by ethanol/LPS (Fig. 4, top) and aminotransferases were significantly blunted (Fig. 8, bottom). Furthermore, the accumulation of fibrin caused by ethanol/LPS under these conditions (Fig. 8, top; see also Fig. 5) was also completely abrogated in these two groups (Fig. 8, top).
Ethanol Enhances Inflammatory Damage in Mouse Liver.
Exposure of the liver to low levels of LPS is common and occurs through multiple means. For example, both acute and chronic alcohol consumption increase circulating LPS levels in human subjects.14 Whereas inflammatory responses triggered by small doses of LPS are typically noninjurious, physiological/biochemical stresses can synergistically enhance the hepatotoxic response to LPS. Indeed, in addition to increasing circulating LPS, ethanol also enhances inflammation and liver damage caused by acute LPS injected >20 hours after ethanol exposure.7, 15 Whereas this acute exposure model is not synonymous with chronic ethanol consumption, the response to LPS in this model appears to be mechanistically similar to certain aspects of alcohol-induced liver injury,16 and it is the rationale for its employment.15 It was shown here that the enhancement of LPS-induced inflammation and liver damage caused by ethanol correlated with a robust increase in fibrin accumulation (Figs. 2–5). Furthermore, this effect of ethanol was nearly completely abrogated by blocking the initiation of the coagulation cascade with hirudin (Figs. 2–5). Taken together, the major findings of the current study indicate that ethanol enhances inflammation and liver injury owing to a bolus injection of LPS, at least in part, through thrombin-dependent fibrin deposition.
Does Ethanol Enhance Hemostasis Caused by LPS?
Previous studies have shown that pretreatment of rats with thrombin inhibitors, such as heparin or hirudin, prevented the activation of the coagulation cascade and subsequent liver injury induced by large doses of LPS.17, 18 The coagulation cascade can contribute to liver injury, in part through the generation of occlusive fibrin clots in the hepatic sinusoids (Fig. 9).19, 20 The fibrin clots could cause hemostasis, microregional hypoxia, and subsequent hepatocellular death, which may be independent of inflammatory injury.17, 18, 21 Ethanol pretreatment enhanced LPS-induced fibrin accumulation and liver injury in LPS-treated mice, both of which were prevented by hirudin (Figs. 2–5). These data indicate that de novo fibrin deposition by the coagulation cascade contributes to liver injury under these conditions. However, the exaggerated fibrin accumulation induced by ethanol was not associated with an enhancement of LPS-induced coagulation, per se, as thrombin-antithrombin levels were similar in the LPS and ethanol/LPS groups (Fig. 5). This result indicates that although fibrin accumulation in ethanol/LPS-treated mice is initiated at the level of thrombin activation (Fig. 9), the enhanced fibrin accumulation is not derived from increased deposition, but rather by another factor in fibrin homeostasis (see below).
In addition to deposition by coagulation, the level of fibrin extracellular matrix is also regulated by degradation of the existing matrix (i.e., fibrinolysis, see Fig. 9). Specifically, inhibition of fibrinolysis can cause this extracellular matrix to accumulate, even in the absence of enhanced deposition by the thrombin cascade. A major inhibitor of fibrinolysis is PAI-1, through blocking the activation of plasmin by plasminogen activators (uPA and tPA). In the current study, ethanol enhanced the increase in PAI-1 expression caused by LPS as early as 4 hours after injection of LPS (Fig. 6). Elevated PAI-1 and hepatic fibrin have correlated with enhanced LPS-induced liver damage in other models, such as idiosyncratic drug toxicity,6 surgical resection,22 or adipocytokine administration.8 As mentioned in the introduction, the “classic” role of PAI-1 in impairing fibrinolysis may also contribute to inflammation. For example, fibrin matrices have been shown to be permissive to chemotaxis and activation of monocytes and leukocytes.23, 24 In addition to altering fibrin metabolism, PAI-1 may alter the profile of other inflammatory mediators through inhibition of plasminogen activators. For example, the inhibition of plasmin activation by PAI-1 prevents the conversion of secreted latent TGFβ to its active form,25 which may mediate antiinflammatory effects, especially on monocytes/macrophages.26 These mechanisms are not mutually exclusive and may occur in tandem. Indeed, blocking active PAI-1 with an antibody prevented the recruitment of neutrophils to the liver caused by the combination of LPS and ethanol (Fig. 8).
Regulation of the expression of PAI-1 is multifaceted in the cell.27, 28 In the current study the increase in PAI-1 expression caused by LPS after ethanol (Fig. 6) was coupled with enhanced activation of ERK1/2 (Fig. 7). The ERK1/2 MAPK signaling pathway has been shown to play a role in the expression of PAI-1 in response to TNFα.29 In agreement with these studies, the increase in PAI-1 expression caused by LPS after ethanol was significantly attenuated (≈4-fold) by U0126 administration (Fig. 6). The inhibition of PAI-1 induction by U0126 was also mirrored by prevention of fibrin accumulation (Fig. 5) and liver damage (Fig. 2) under these conditions, which supports the hypothesis that exacerbated LPS-induced liver damage caused by ethanol is due to increased fibrin accumulation. Lastly, the accumulation of fibrin, neutrophils, and liver injury was significantly attenuated in TNFR-1−/− mice. Taken together, these results suggest that the increased activation of ERK1/2 by TNFα enhances LPS-induced PAI-1 expression and fibrin deposition that contribute to liver injury. However, the exact mechanism by which PAI-1 is induced after LPS is unclear. Indeed, Montes et al.30 demonstrated in rabbits that hepatic PAI-1 induction in response to LPS appears to be independent of TNFα production.
Here, although the increase in expression of TNFα caused by LPS was not detectably enhanced in liver by ethanol preexposure (Fig. 6), there was an increase in the activation of ERK1/2 activation by this preexposure (Fig. 7). Possible mechanisms to explain these results include that ethanol increases the expression of other agonists that lead to ERK activation, and/or ethanol enhances the response of ERK activation to an agonist. Data for both mechanisms are supported, at least under in vitro conditions. For example, Chen et al.31 showed that ethanol preexposure to cultured hepatocytes enhanced and prolonged the activation of ERK in response to a number of agonists. Regardless of mechanism, the results with U0126 support the hypothesis that the enhancement of LPS-induced PAI-1 expression and liver damage caused by ethanol preexposure is mediated through ERK signaling.
Taken together, the results of this study identify a new potential mechanism by which ethanol may contribute to inflammatory liver damage. Specifically, through enhancing the induction of PAI-1 and thereby impairing fibrinolysis, ethanol exacerbates inflammatory liver damage caused by LPS. Whereas fibrin deposition is stimulated by LPS, at least in part by activating the coagulation system, ethanol seems to increase LPS-induced fibrin deposition exclusively through an inhibition of fibrinolysis (Fig. 9). The results of the current study support the hypothesis that ethanol may mediate hepatic injury, at least in part, by decreasing fibrinolysis.
- 12Tumor necrosis factor-alpha and troglitazone regulate plasminogen activator inhibitor type 1 production through extracellular signal-regulated kinase- and nuclear factor-kappaB-dependent pathways in cultured human umbilical vein endothelial cells. J Pharmacol Exp Ther 2003; 307: 987–994., , , .