Tumor necrosis factor α (TNFα) has been shown to be both proapoptotic and mitogenic for hepatocytes and necessary for alcohol-induced liver injury. Ras, a known proto-oncogene, is very important in the regulation of cellular responses to TNFα. Therefore, the purpose of this study was to investigate the role of Ras in alcohol-induced pathogenesis. Male C57Bl/6 mice were fed ethanol or high-fat control diet via intragastric cannulation for 4 weeks. Ras activity was increased significantly after 4 weeks of ethanol and correlated with an increase in pathologic features. However, in mice deficient in the receptor-type 1 for TNFα (TNFR1-/-), ethanol-induced liver injury and the increase in Ras activity were significantly blunted compared with wild-type mice. Furthermore, it was demonstrated that H-, K-, and R-Ras isoforms were increased after ethanol exposure in wild-type mice. In TNFR1-/- mice, R-Ras activity remained elevated by ethanol, whereas H-Ras and K-Ras activity was blunted significantly under these conditions. Interestingly, hepatocellular proliferation, which was elevated approximately fivefold after 4 weeks of chronic ethanol in wild-type mice, was also blunted in TNFR1-/- mice given ethanol. Inhibition of Ras with adenovirus containing a dominant-negative Ras had no effect on ethanol-induced liver injury, but significantly blunted ethanol-induced hepatocyte proliferation by more than 50%. Overexpression of mitochondrial superoxide dismutase using recombinant adenovirus blunted lipid peroxidation and attenuated hepatic injury resulting from ethanol, but had no effect on Ras activation and hepatocyte proliferation caused by ethanol. In conclusion, these data support the hypotheses that hepatocellular oxidative stress leads to cell death and that TNFα-induced Ras activation is important in hepatic proliferation in response to ethanol-induced liver injury. (HEPATOLOGY 2004;39:721–731.)
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Recently, a critical role for TNFα in the early pathogenesis of alcohol-induced liver disease has been established.1 TNFα is hypothesized to signal through TNFR1 to initiate a cascade of hepatocellular events, including mitochondrial oxidative stress, activation of stress kinases such as p42/44 and p38, and a number of other responses ultimately leading to cell death. However, TNFα has been described as both proapoptotic and mitogenic for hepatocytes. Thus, it is likely that investigation of downstream signal events from the TNF receptor may explain these apparent differences.
Ras, a known proto-oncogene, is important in the regulation of cellular responses to TNFα. Recently, it was shown that TNFα-mediated signaling in hepatocytes involves activation of the Ras/mitogen activated protein kinase (MAPK) pathway.2 Inhibition of Ras using a dominant-negative mutant nearly completely inhibited TNFα-induced hepatocyte proliferation in culture. It has also been shown that TNFα-mediated hepatocyte proliferation after partial hepatectomy is dependent on Ras activation.3 Yet, it is unclear whether TNFα release caused by chronic ethanol stimulates Ras activation in vivo, and whether this phenomenon is related to ethanol-induced hepatocyte proliferation.
TNFα also causes oxidative stress in hepatocytes, most likely originating in mitochondria, leading to hepatocyte death. Overexpression of mitochondrial isoform of superoxide dismutase (MnSOD) using recombinant adenovirus blocked alcohol-induced liver injury.4 However, this effect was downstream of TNFα production, suggesting that oxidant production is an integral part of TNFα-induced hepatocyte death caused by chronic ethanol. Whether Ras activation contributes to this mechanism is not known.
Ras exists in several isoforms, which are largely redundant; however, some evidence now exists that each isoform has distinct signaling properties.5 Whether Ras activation in response to ethanol is involved in pathogenesis or proliferation is not clear. Thus, the purpose of this study was to investigate the activation of Ras isoforms under conditions of chronic ethanol exposure. It is likely that activation of multiple Ras isoforms plays different roles in pathogenesis or proliferation.
In this study, it is shown that TNFα-dependent activation of Ras caused by chronic ethanol regulates the proliferation of hepatocytes. By overexpressing dominant-negative Ras with recombinant adenovirus, it was demonstrated that Ras activation did not contribute to injury. Moreover, it is concluded based on mitochondrial MnSOD overexpression that Ras-mediated cell proliferation is independent of oxidative stress and subsequent liver injury.
TNFR1, tumor necrosis factor receptor type 1; MnSOD, mitochondrial isoform of superoxide dismutase; ALT, alanine transaminase; GST-RBD, glutathione S-transferases-fusion protein with the Ras-binding domain of Raf1; PBS, phosphate buffered saline; PCNA, proliferating cell nuclear antigen; HA, hemagglutinin.
Materials and Methods
Wild-type and TNFR1-/- mice (22–25 g; Jackson Laboratories, Bar Harbor, ME) were fed a high-fat control diet or diet containing ethanol (approximately 24 g/kg daily) via intragastric cannulation as described by Tsukamoto et al.6 and modified for mice as described by Yin et al.1
Male C57BL/6 mice (25–30 g) were obtained from Jackson Laboratories. All animals were housed in pathogen-free barrier facilities accredited by the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) and procedures used were approved by the local Institutional Animal Care and Use Committee (IACUC). Gastric cannulation in mice was performed using aseptic surgical techniques detailed elsewhere.7, 8 After surgery, animals were placed in individual sterile metabolic cages, and the gastric cannulae were connected to an infusion pump. All animals had free access to water and cellulose pellets (Harlan, Madison, WI) as a source of fiber during enteral feeding.
Diets were initiated 1 week after surgery to allow for full recovery. A liquid diet described by Thompson and Reitz,9 supplemented with lipotropes as detailed by Morimoto et al.,10 was prepared daily. The total amount of the fat in diet is much the same considered by the body weight. It contains corn oil as fat (37% of total calories). Animals were randomly divided into two experimental groups and fed either a high-fat control or ethanol-containing diet continuously for 4 weeks. Diet was infused at a rate of 0.4 mL/g daily with an infusion pump (Harvard Apparatus, Natic, MA). The degree of alcohol intoxication was assessed regularly to evaluate development of tolerance so that ethanol delivery could be increased. The amount of ethanol in the diet was increased over the course of the study from 5% to 8% to obtain optimal delivery of calories without compromising growth or survival. Previous work has demonstrated that the above-described dosing regimen leads to similar urine alcohol profiles between species.7, 11 Ethanol concentration in urine, which is representative of blood alcohol levels, was measured daily. Animals were housed in metabolic cages that separated urine from feces, and urine was collected over 24 hours in bottles containing mineral oil to prevent evaporation. Each day at 9:00 AM, urine collection bottles were changed, and a sample was stored at −20°C for later analysis. Ethanol concentration was determined by measuring absorbance at 366 nm resulting from the reduction of NAD+ to NADH by alcohol dehydrogenase. Averages from individual animals over the course of the study were pooled to determine group means.
Adenoviral Synthesis and Preparation.
Recombinant adenoviral vectors containing the transgene for either β-galactosidase (Ad.lacZ), dominant-negative Ras activation (Ad.ΔRas), or MnSOD (Ad.MnSOD) were prepared as described elsewhere.12, 13 Virus was injected at a titer of 1 × 1010 (Ad.ΔRas) or 3 × 109 (Ad.MnSOD) plaque-forming units diluted into 250 μL of lactated Ringer's solution via the femoral vein within 1 week from the initial feeding of ethanol.
Blood Collection and Transaminase Determinations.
After 4 weeks of enteral feeding, mice were anesthetized with sodium pentobarbital 75 mg/kg intraperitoneally and blood was collected from the vena cava and centrifuged. Serum was stored at −80°C until it was assayed for alanine transaminase (ALT) by standard enzymatic procedures.14
After 4 weeks of ethanol treatment, livers were formalin fixed, embedded in paraffin, and stained with hematoxylin and eosin to assess steatosis, inflammation, and necrosis. Liver pathologic characteristics were scored in a blind manner as described by Nanji et al.15 as follows: steatosis (the percentage of liver cells containing fat), <25% = 1+, <50% = 2+, <75% = 3+, 75% = 4+; inflammation and necrosis, 1 focus per low-power field = 1+; 2 or more foci = 2+.
Expression and Purification of Glutathione S-Transferases Fusion Proteins.
The expression of glutathione S-transferases-fusion protein with the Ras-binding domain of Raf1 (GST-RBD) fusion proteins was performed in an overnight culture of SOC medium containing 50 μg ampicillin/mL. Production of recombinant protein was induced with isopropyl β-D-thiogalactopyranoside, and the culture was then incubated for 3 to 4 hours at room temperature. Cells were harvested by centrifugation, resuspended, and lysed by sonication. The supernatant was mixed with glutathione-Sepharose beads (Bio-Rad, Hercules, CA) and incubated for 1 hour at 4°C. The beads were washed and resuspended in 50% (vol/vol) phosphate buffered saline (PBS). The purified protein was used for GST pull-down experiments.
Ras Pull-Down Assay.
Samples (300 μg protein) in a final volume of 550 μL were incubated with 50 μL GST-RBD fusion protein overnight at 4°C. Twenty microliters of a 50% slurry of glutathione-bound agarose beads were added, and after 30 minutes of incubation, the solution was centrifuged as above. Supernatants were washed and were subjected to SDS-PAGE and transferred to Immobilon-P membranes for Western blotting.16
Electrophoresis and Immunoblotting of Membranes.
Sample buffer (2% SDS, 100 mM Tris HCl, pH 6.8, 5% β-mercaptoethanol, 12% (vol/vol) glycerol, and 0.02% (wt/vol) bromophenol blue) was added to the protein sample, and the mixture was heated to 100°C for 5 minutes. Protein was separated by electrophoresis through a 12% SDS-polyacrylamide gel and was transferred onto a nitrocellulose membrane. Membranes were incubated for 60 minutes with the monoclonal pan Ras primary antibody (1:2,500; Chemicon) at room temperature. The membranes were washed several times with blotting buffer and were incubated for 60 minutes with secondary antibody (goat anti-rabbit immunoglobulin G, 1:1,000; Bio-Rad). The membranes were washed, and the protein bands were detected by chemiluminescence.
Immunohistochemistry of Proliferating Cell Nuclear Antigen (PCNA).
Hepatocytes in S phase were analyzed by PCNA staining. Paraffin-embedded sections of liver tissue were deparaffinized, rehydrated, and stained immunohistochemically by sequential incubation with a monoclonal anti-PCNA antibody (Alpha Diagnostic International, San Antonio, TX) in PBS (pH 7.4) containing 1% Tween 20 and 1% bovine serum albumin. Peroxidase-linked secondary antibody and diaminobenzidine (Peroxidase Envision Kit; DAKO Corp., Carpinteria, CA) were used to detect specific binding. No positive staining was detected in tissue from the same animals that were processed without primary antibody. An Axioskop 50 microscope (Carl Zeiss, Inc., Thornwood, NY) was used for evaluation in a blinded manner. Positive nuclei were counted in immunostained tissue sections at a magnification ×200.17 Only hepatocytes with dark brown-stained nuclei were counted as S phase cells. Positively stained nuclei in 10 randomly selected fields were counted (approximately 200 cells per field) and data from each tissue section were pooled to determine means.
Immunohistochemistry of 4-Hydroxynonenal-Modified Protein.
Paraffin-embedded sections of liver tissue were deparaffinized, rehydrated, and stained immunohistochemically for presence of an in vivo marker of lipid peroxidation, 4-hydroxynonenal-protein adducts, by sequential incubation with a polyclonal antibody (Alpha Diagnostic International, San Antonio, TX) in PBS (pH 7.4) containing 1% Tween 20 and 1% bovine serum albumin, as described previously.18 A peroxidase-linked secondary antibody and diaminobenzidine (Peroxidase Envision Kit; DAKO Corp., Carpinteria, CA) were used to detect specific binding. No staining was detected in tissue from the same animals that were processed without primary antibody.
The Metamorph image acquisition and analysis system (Universal Imaging Corp., Chester, PA) incorporating an Axioskop 50 microscope (Carl Zeiss, Inc.) was used for evaluation in a blinded manner. Positive area was captured and analyzed in the immunostained tissue sections at a magnification ×100.17 Color detection ranges were set for the red-brown color of the diaminobenzidine chromogen based on an intensely labeled point. The extent of 4-hydroxynonenal labeling in the liver lobule was defined as the percent of the field area within the default color range determined by the software. Data from each tissue section (10 fields per section) were pooled to determine means.
Two-way ANOVA using Bonferroni's post-hoc test was used for the determination of statistical significance as appropriate. A P value of less than 0.05 was selected before the study as the level of significance.
Steady body weight gain was observed during the 4 weeks of continuous enteral feeding of liquid diets in all groups, indicating that animals received adequate nutrition. The mean body weight gains were not significantly different among the groups (data not shown). As reported previously,19–21 alcohol levels in urine fluctuated in a cyclic pattern from zero to more than 500 mg/dL. Average urine alcohol concentrations were approximately 210 mg/dL daily, and there were no significant differences in mean urine alcohol concentrations between groups given ethanol (data not shown). In the experiments using recombinant adenovirus, no changes in this cyclic pattern or differences in urine alcohol concentration were observed.
Pathologic Evaluation and Hepatocyte Proliferation in TNFR1-/- Mice After Ethanol Exposure.
Wild-type and TNFR1-/- mice were fed a high-fat control diet or diet containing ethanol via intragastric cannulae for 4 weeks. In control groups, liver histologic scores were normal as determined by hematoxylin and eosin staining (Table 1). After 4 weeks of ethanol feeding, significant steatosis and mild necrosis developed in wild-type mice (Fig. 1A), however, these effects were significantly blunted in TNFR1-/- mice (Fig. 1B). Hepatocyte proliferation was determined by PCNA staining. After 4 weeks of enteral feeding, there were no differences in PCNA staining between wild-type mice and TNFR1-/- mice fed a high-fat control diet (Figs. 1C,D and 2). However, there was a significant increase in nuclear PCNA in wild-type mice fed an ethanol-containing diet (Figs. 1E and 2). This increase was blunted in TNFR1-/- mice fed an ethanol-containing diet (Figs. 1F and 2).
Table 1. Serum Alanine Transaminase Levels and Pathologic Score
Serum Alanine Transaminase (U/L)
NOTE. Wild-type mice were infected with recombinant adenovirus (1 × 109 pfu) containing either β-galactosidase (Ad.lacZ), dominant-negative Ras (Ad.ΔRas), or human mitochondrial superoxide dismutase (Ad.MnSOD). Mice then were fed a high-fat control diet or a diet containing ethanol for 4 weeks. Serum alanine transaminase levels were determined by a standard enzymatic biochemical assay. Liver sections were stained with hematoxylin and eosin, and pathologic features were evaluated and scored using a scale established by Nanji et al.15 on the basis of steatosis (0–4), inflammatory foci (0–2), and necrotic foci (0–2). Data are expressed as means ± SEM (n = 4–6).
P < .05 compared with mice fed a high-fat control diet.
Ras Activation between Wild-Type and TNFR1-/- Mice After Ethanol Exposure.
Because TNFα was reported to induce hepatocellular proliferation through a Ras-dependent mechanism,2 Ras activation was evaluated under these conditions. As described in Materials and Methods, activated GTP-bound Ras was measured by pull-down assay using a GST-RBD as a substrate. Precipitates were then evaluated by Western blot using antibodies for Ras.16 Chronic ethanol increased total Ras activity in wild-type mice compared with animals fed a control diet (Fig. 3A,B). However, in TNFR1-/- mice, alcohol-induced Ras activation was significantly blunted, suggesting that Ras activation was dependent on TNFα signaling through the TNFR1. Using the pull-down assay described above, activation of each Ras isoform was determined using antibodies specific for each isoform. It was demonstrated that H-Ras, K-Ras, and R-Ras levels increased after ethanol exposure in wild-type mice. Interestingly, the increase in R-Ras activity was similarly elevated in wild-type mice under these conditions, whereas ethanol-induced H-Ras and K-Ras activity was significantly blunted in TNFR1-/- mice (Fig. 3C). These data support the hypothesis that TNFα-induced Ras activation may play an important role in the progression of liver injury resulting from ethanol. Moreover, these data suggest that the isoforms of Ras are activated or regulated by different mechanisms and may have distinct roles in pathogenesis.
Inhibition of Ras Activation Using Recombinant Adenoviral Vectors.
Because Ras activation correlated with TNFα-dependent hepatocellular damage after 4 weeks of ethanol exposure, it was hypothesized that the inhibition of Ras activation would blunt ethanol-induced pathogenesis or proliferation. Also, the role of hepatocellular oxidative stress was evaluated. This was accomplished by recombinant adenoviral gene delivery of either dominant-negative Ras, which inhibits the activation of all Ras isoforms,3 or mitochondrial isoform of superoxide dismutase (MnSOD).4 Dominant-negative Ras was tagged with hemagglutinin (HA), and Western analysis and immunoblot for HA-tag expression were performed to evaluate the expressions of the transgene (Fig. 4A). Dominant-negative Ras expression was observed in Ad.ΔRas infected mice fed both high-fat and ethanol diet. Importantly, Ras activation was blocked after 4 weeks of enteral feeding in these animals (Fig. 4C). The expression of human MnSOD determined by Western blot was evident in all mice as expected but increased nearly fourfold in mice transduced with Ad.MnSOD (Fig. 4B). The level of alcohol-induced Ras activation was also determined in these animals (Fig. 4D); however, no effect of MnSOD overexpression on Ras activity was observed.
4-Hydroxynonenal Staining After Chronic Ethanol Exposure.
To determine whether inhibition of Ras effects ethanol-induced oxidative stress, lipid peroxidation was evaluated in livers from mice fed either a high-fat control diet or an ethanol containing diet by 4-hydroxynonenal (4-HNE) immunohistochemistry. Wild-type mice and TNFR1-/- mice were given either a high-fat control diet or a diet containing ethanol as previously described. Accumulation of 4-hydroxynonenal in wild-type mouse fed high-fat control diet was minimal (Fig. 5A-a) but was significantly enhanced in mice given ethanol (Fig. 5A-b,B). Interestingly, the increase in 4-HNE immunohistochemistry resulting from ethanol was significantly blunted in TNFR1-/- mice (Fig. 5A-c), compared with wild-type mice (Fig. 5B). Livers from Ad.lacZ-infected mice given ethanol also had significant 4-HNE staining (Fig. 5A-d) compared with Ad.lacZ-infected animals fed a high-fat control diet (Fig. 5C). Ethanol causes a significant increase in 4-HNE staining in livers of Ad.ΔRas-infected mice (Fig. 5A-e), similar to the lipid peroxidation observed in control mice given ethanol. However, livers from mice infected with Ad.MnSOD fed ethanol showed little immunohistochemical staining for 4-HNE (Fig. 5A-f). Image densitometry was performed to quantify the changes in ethanol-induced lipid peroxidation (Fig. 5B,C). These data suggest that ethanol induces oxidative stress independently of Ras, and more importantly, oxidative stress is not required for Ras activation.
Liver Injury After Chronic Ethanol Exposure.
Mice infected with recombinant adenovirus containing either β-galactosidase (Ad.lacZ), dominant-negative Ras (Ad.ΔRas), or mitochondrial SOD (Ad.MnSOD) were given either a control diet or a diet containing ethanol for 4 weeks. Serum ALT levels were approximately 30 IU/L after 4 weeks of a high-fat control diet (Table 1). In Ad.lacZ-infected animals, serum ALT were significantly elevated to 170 ± 37 U/L. There was no significant difference in the increase in ALT caused by ethanol between Ad.lacZ-infected mice and Ad.ΔRas-infected mice. However, the increase in serum ALT levels was significantly blunted in mice infected with Ad.MnSOD.
Liver histologic features was assessed by hematoxylin and eosin staining (Fig. 6), and pathologic features was scored using criteria established by Nanji et al.,15 which assesses steatosis, inflammation, and necrosis (Table 1). After 4 weeks of ethanol control diet, no pathologic changes were observed in mice of any groups. After 4 weeks of ethanol feeding, steatosis and severe inflammation and necrosis developed in both Ad.lacZ-infected mice and Ad.ΔRas-infected mice (Fig. 6B,D). However, in Ad.MnSOD-infected mice fed ethanol, pathologic features were blunted significantly (Fig. 6F).
Cell Proliferation After Chronic Ethanol Exposure.
Because TNFα stimulates hepatocyte proliferation in a TNFα-dependent manner,2 cell proliferation after 4 weeks of ethanol was determined by PCNA staining. There were no observable differences in PCNA staining between Ad.lacZ-infected mice, Ad.ΔRas-infected mice, and Ad.MnSOD mice fed a high-fat control diet (Figs. 7 and 8). In Ad.lacZ-infected mice, ethanol caused a significant increase in cell proliferation demonstrated by an increase in nuclear PCNA staining. Interestingly, in mice overexpressing dominant-negative Ras, cell proliferation was markedly reduced.
Because MnSOD overexpression reduced early alcohol-induced liver injury, it was hypothesized that the overexpression of MnSOD would blunt alcohol-induced cell proliferation. In mice fed a high-fat control diet, there was no significant difference in nuclear PCNA staining between those infected with Ad.lacZ or with Ad.MnSOD (Figs. 7 and 8). However, there were more PCNA-positive nuclei in Ad.lacZ-infected mice and Ad.MnSOD-infected mice fed ethanol. Interestingly, the increases in cell proliferation caused by ethanol in Ad.MnSOD-infected mice were not different from that observed in Ad.lacZ-infected mice fed ethanol. These data suggest that oxidative stress is indeed important for hepatocellular injury, as earlier reported.4 Moreover, these data support the hypothesis that liver injury is not required for TNFα-mediated cell proliferation caused by ethanol. In addition, these data suggest that there are distinct signaling mechanisms controlling TNFα-induced cell injury and cell proliferation.
The role of TNFα in alcohol-induced liver injury has been established through a variety of experimental approaches. Notably, Yin et al. showed that TNFα signaling through TNFR1 is essential for alcohol-induced pathologic features by intragastrically feeding ethanol to mice deficient in the receptor.1 Other groups have shown that in addition to the toxic effects of its signaling in liver, TNFα was responsible for hepatocyte proliferation.22, 23 Using the same strain of mice, it was also demonstrated that TNFα signaling was required for liver regeneration after partial hepatectomy.24, 25 It is unclear why TNFα acts as a mitogen to induce cell proliferation under some conditions and cell death in other situations.
Based on a number of in vitro studies, TNFα stimulates cell proliferation through a variety of signal transduction pathways including MAPK26 and nuclear factor kappa B (NFκB).5 Additionally, groups have demonstrated that Ras activation is important for hepatocyte proliferation caused by TNFα.2 Thus, the purpose of this study was to investigate the role of Ras activation in alcohol-induced TNFα production and subsequent liver disease. It was shown here that Ras is indeed activated during 4 weeks of ethanol exposure (Fig. 3A). Moreover, H-Ras and K-Ras isoforms are activated in response to ethanol in a TNFα-dependent mechanism, whereas R-Ras activation occurs independent of TNFα signaling in response to ethanol (Fig. 3C), suggesting that different Ras isoforms may have different roles in response to TNFα and other stimuli or mitogens. This indicates that Ras isoforms are not redundant and, for the most part, are activated by distinctly different mechanisms. These findings also may reflect the fact that Ras isoforms could be expressed differentially within multiple cell types in the liver.
It was also shown here that 4 weeks of ethanol exposure caused a significant increase in cell proliferation. However, in TNFR1-/- mice, this increase was reduced (Fig. 2). This suggests that hepatocyte proliferation after ethanol may be dependent on TNFα signaling, which is consistent with the mitogenic properties of TNFα. However, these findings may be because TNFR1-/- mice do not exhibit pathologic changes, and therefore a regenerative response is not induced. Because of the correlation between cell proliferation and Ras activation after chronic ethanol exposure, it is reasonable to speculate that TNFα-induced Ras activation mediates an increase in cell proliferation. Other reports have shown that chronic ethanol inhibits the capacity of the liver to regenerate.27 Moreover, it was shown that proliferation after partial hepatectomy was blunted even in a model similar to the intragastric infusion model used here. In these reports, however, cell proliferation was assessed in response to partial hepatectomy and was shown to be blunted or delayed in animals exposed to ethanol. In contrast, the basal levels of cell proliferation in response to injury is evaluated here, not the robust proliferative response generated after partial hepatectomy. Unfortunately, in these studies it is difficult to dissect the TNFα-dependent cell proliferation caused by chronic ethanol exposure and the potential inhibitory effects of ethanol itself.
Recently, a dominant-negative inhibitor of Ras was developed and was shown to inhibit hepatocyte proliferation after partial hepatectomy in mice infected with recombinant adenovirus containing the transgene.3 This approach was used to test the hypothesis that Ras activation plays a role in alcohol-induced liver injury and cell proliferation. Inhibition of Ras protected against alcohol-induced liver injury (Fig. 6), suggesting that activation of Ras is not involved in the mechanism of pathogenesis. However, alcohol-induced cell proliferation is blunted when Ras activation is inhibited (Fig. 7), indicating that Ras may be involved in tissue repair or maintenance in response to alcohol-induced liver injury. However, these data do not fully exclude the possibility that isoforms of Ras other than those evaluated here (e.g., N-Ras) contribute to this phenomenon.
It is hypothesized the increase in hepatocyte proliferation after chronic ethanol involves TNFα-induced Ras activation, which subsequently induces cyclin-dependent kinases, cyclins, and so forth. This TNFα-induced signaling mechanism apparently is distinct from the mechanism of TNFα-induced cell death, because inhibition of Ras did not effect alcohol-induced injury but did inhibit hepatocyte proliferation. These data suggest that Ras activation does not contribute to injury (Table 1 and Fig. 6) and that TNFα-dependent activation of Ras caused by chronic ethanol regulates the proliferation of hepatocytes (Figs. 7 and 8). However, it is possible that Ras activation and hepatocyte proliferation are merely responses to the tissue injury and are only coincidentally dependent on TNFα.
If Ras activation and hepatocyte proliferation are only a consequence of TNFα mediated tissue injury, then blocking injury should also block cell proliferation. Likewise, if cell proliferation is not related or dependent upon injury, then cell proliferation should occur even if alcohol-induced liver injury is inhibited, as long as TNFα signaling is not disrupted. To test this hypothesis, mitochondrial MnSOD was overexpressed in liver using recombinant adenovirus. It is important to note that in previous studies MnSOD blunted early alcohol-induced liver injury without affecting TNFα production.4 Under similar conditions, MnSOD overexpression blunted ethanol induced hepatic oxidative stress and liver injury; however, hepatocellular proliferation was not inhibited (Figs. 7 and 8). Moreover, it is concluded, based on these results, that Ras-mediated cell proliferation occurs independent of oxidative stress and subsequent liver injury.
In conclusion, it seems that injury and proliferation are both dependent on TNFα production but involve two different mechanisms downstream of TNF receptor activation. The data here support the conclusion that TNFα stimulates both injury and proliferation via TNFR1. For proliferation, TNFα stimulates Ras activation. For injury, TNFα stimulates oxidative stress, which ultimately leads to cell death.4 The fundamental question is: What conditions determine whether a cell responds with proliferation or death? The answer is still unclear, and thus further investigation into the signal transduction of this mechanism is necessary. It seems that early pathologic changes associated with chronic ethanol are in the balance of cell injury and cell proliferation. A shift in the balance may underlie susceptibility or the rate of progression of alcoholic liver disease.