Elevated interleukin-6 during ethanol consumption acts as a ... - Nature

ã

Oncogene (2002) 21, 32 ± 43 2002 Nature Publishing Group All rights reserved 0950 ± 9232/02 $25.00 www.nature.com/onc

Elevated interleukin-6 during ethanol consumption acts as a potential endogenous protective cytokine against ethanol-induced apoptosis in the liver: involvement of induction of Bcl-2 and Bcl-xL proteins Feng Hong1, Won-Ho Kim1, Zhigang Tian2,3, Barbara Jaruga1, Edward Ishac2, Xuening Shen2 and Bin Gao*,1 1

Section on Liver Biology, Laboratory of Physiologic Studies, National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health, Bethesda, Maryland, MD 20892, USA; 2Department of Pharmacology and Toxicology, Medical College of Virginia Commonwealth University, Richmond, Virginia, VA 23298, USA; 3School of Life Science, University of Science and Technology of China, Hefei, China

Elevation of serum interleukin-6 (IL-6) levels is always associated with alcoholic liver disease (ALD), but the signi®cance of such elevation is not clear. Here we show that chronic ethanol consumption induces signi®cant apoptosis in the liver of IL-6 (7/7) mice but not IL-6 (+/+) mice. IL-6 (7/7) hepatocytes are more susceptible to ethanol- and tumor necrosis factor a- (TNFa-) induced apoptotic killing, which can be corrected by IL-6. Expression of both anti-apoptotic (such as Bcl-2 and BclxL) and proapoptotic (such as Bax) proteins is markedly elevated in the liver of human ALD and chronically ethanol-fed IL-6 (+/+) mice. On the contrary, induction of Bcl-2 and Bcl-xL is not observed in the liver of chronically ethanol-fed IL-6 (7/7) mice, whereas expression of Bax protein remains elevated. Injection of IL-6 markedly induces expression of Bcl-2 and Bcl-xL but not Bax in the liver. Finally, high concentrations of ethanol inhibit IL-6-activated anti-apoptotic signal, but increasing the concentrations of IL-6 is able to overcome such inhibitory eect. These ®ndings suggest that elevated serum IL-6 levels in ALD may overcome the inhibitory eect of ethanol on IL-6-mediated anti-apoptotic signals and prevent alcohol-induced hepatic apoptosis by induction of Bcl-2 and Bcl-xL. Oncogene (2002) 21, 32 ± 43. DOI: 10.1038/sj/onc/ 1205016 Keywords: IL-6; Bcl-2; Bax; apoptosis; liver; alcoholic liver disease Introduction Alcohol is a major etiologic factor of liver disease in Western countries. Chronic and excessive consumption of alcohol can lead to fatty liver, perivenular ®brosis,

*Correspondence: B Gao, Section on Liver Biology, NIAAA/NIH, Park Building, Room 120, 12420 Parklawn Drive, MSC 8115, Bethesda, MD 20892, USA; E-mail: [email protected] Received 22 June 2001; revised 12 September 2001; accepted 1 October 2001

alcoholic foamy degeneration, alcoholic hepatitis, occlusive venous lesions, cirrhosis and hepatocellular carcinoma. Alcoholic liver disease (ALD) is a direct result of alcohol-induced hepatotoxicity coupled with impaired hepatic regenerative activity (Abittan and Lieber, 1999; Diehl, 1989; Hall, 1994; Maddrey, 2000; Walsh and Alexander, 2000). Direct liver injury by ethanol is due to several processes with deleterious eects on the liver, including intracellular accumulation of proteins, and of acetaldehyde, microsomal activation of hepatotoxins, alterations in hepatic redox state, and enhancement of lymphocyte cytotoxicity (Abittan and Lieber, 1999; Diehl, 1989; Hall, 1994; Maddrey, 2000; Walsh and Alexander, 2000). Anti-regenerative activity of ethanol in the liver is, at least in part, due to suppression of several key signals involved in liver regeneration. These signals include TNFa-activated NF-kB and JNK (Diehl, 1989, 1999a,b), IL-6-activated STAT3 (Chen et al., 1997), growth factors-activated mitogen-activated protein kinases (Chen et al., 1998), and PI3/Akt kinase (Hong et al., 2000; Sasaki and Wands, 1994). It is well-known that serum levels of many cytokines (such as TNFa, IL-1, IL-8, IL-6) are elevated in alcoholic patients and chronically ethanol-fed animals (Diehl, 1999a; Laso et al., 1999; Martinez et al., 1992; McClain et al., 1999). The essential role of TNFa in alcoholic liver injury has been clearly demonstrated in TNFa receptor 1-de®cient mice by Thurman's group (Yin et al., 1999). They demonstrated that long-term ethanol feeding caused liver injury in wild-type and TNF-R2 knockout mice but not in TNF-R1 knockout mice. High serum levels of IL-8 are correlated with increased mortality rate in alcoholic patients (Bird, 1994), suggesting that IL-8 plays a harmful role in alcoholic liver disease. Since serum IL-6 levels are also highly correlated with the severity of ALD, it is believed that IL-6 may also be involved in alcoholic liver injury (Hill et al., 1992; Sheron et al., 1991). Thus, anti-IL-6 therapy has been proposed for treatment of ALD (Fontanilla et al., 2000; McClain et al., 1999). However IL-6 has been shown to play an important role in liver regeneration (Cressman et al., 1996), liver

IL-6 protects ethanol-induced apoptosis in the liver F Hong et al

repair from injury (Kovalovich et al., 2000), antiapoptosis in the liver (Kovalovich et al., 2001) and cultured hepatocytes (Chen et al., 1999b; Kuo et al., 2001). To further determine the precise role of IL-6 in ALD, we examined the eects of chronic ethanol consumption in both IL-6 (+/+) and IL-6 (7/7) mice. We demonstrate that chronic consumption with the DeCarli-Lieber liquid diet containing ethanol causes greater apoptosis in the liver of IL-6 (7/7) mice than IL-6 (+/+) mice. Furthermore, expression of anti-apoptotic Bcl-2 and Bcl-xL proteins is signi®cantly elevated in the liver of human ALD and chronically ethanol-fed IL-6 (+/+) mice but not in the chronically ethanol-fed IL-6 (7/7) mice. These ®ndings suggest that elevated serum IL-6 levels in alcoholic patients may protect alcohol-induced hepatocyte apoptosis in the liver by induction of antiapoptotic proteins.

33

Results Serum IL-6 levels are elevated in chronically ethanol-fed IL-6 (+/+) mice To study the role of IL-6 in alcoholic liver injury, seven male IL-6 (+/+) and seven male IL-6 (7/7) mice were fed with ethanol containing liquid diet for 8 weeks as described under Materials and methods. Weight-matched littermates were pair-fed on the same liquid diet, except that ethanol was replaced by carbohydrate content. To minimize the dierences due to food intake, the same volume of isocaloric liquid diets was fed in the control and ethanol-treated mice. At the end of the 8-week feeding period, a similar body weight gain was observed in the four treatment groups (Figure 1a). The absolute liver weights and normalized liver weight (liver weight/body weight) of

Figure 1 Body weights, liver weights, liver histology, and serum IL-6 levels in ethanol (ETOH)-fed mice and pair-fed controls. Seven male IL-6 (+/+) and seven male IL-6 (7/7) mice were fed with a nutritionally adequate DeCarli-Lieber liquid diet containing 4% ethanol for 4 weeks, then switched to the same diet containing 5% ethanol for additional 4 weeks. Seven weightmatched IL-6 (+/+) littermates and seven IL-6 (7/7) littermates were pair-fed on the same liquid diet, except that ethanol was replaced by carbohydrate. At the end of 8 weeks, mice were sacri®ced, body weights (a), liver weights (b), liver histology (c), and serum IL-6 levels (d) were measured. *P50.01, in comparison with pair-fed control group (c) Oncogene


34

ethanol-fed mice are greater, but not signi®cantly, than those pair-fed mice (Figure 1b). Liver histology showed that the livers of ethanol-fed IL-6 (+/+) and IL-6 (7/ 7) mice exhibit signi®cant steatosis, whereas no hepatic necrosis, in¯ammation, or ®brosis was observed (Figure 1c). No hepatic steatosis, necrosis, in¯ammation, or ®brosis were observed in pair-fed IL-6 (+/+) or IL-6 (7/7) mice. Serum IL-6 levels between pair-fed and ethanol-fed groups were also measured. As shown in Figure 1d, serum IL-6 levels were very low (52 pg/ml) in pair-fed IL-6 (+/+) mice, pair-fed IL-6 (7/7) mice, and chronically ethanol-fed IL-6 (7/7) mice, whereas they reached to 38+11 pg/ml in chronically ethanol-fed IL-6 (+/+) mice. These ®ndings indicate that serum IL-6 levels are markedly elevated in chronically ethanol-fed mice. Chronic ethanol consumption induces greater apoptosis in the liver of IL-6 (7/7) mice than IL-6 (+/+) mice To understand the precise role of elevated serum IL-6 levels in ALD, IL-6 (+/+) and IL-6 (7/7) mice were fed with an ethanol-containing liquid diet or a control diet as described under Material and methods. TUNEL assay showed no signi®cant apoptosis in chronically ethanol-fed IL-6 (+/+) mice (Figure 2a, top panel), nor paired-fed (IL-6 (+/+) and IL-6 (7/7) mice (data not shown), which is consistent with a previous report (Rashid et al., 1999). On the contrary, chronic ethanol consumption caused signi®cant apoptosis in the liver of IL-6 (7/7) mice (Figure 2a, top panel). Signi®cant apoptosis in the liver of chronically ethanolfed IL-6 (7/7) mice was also con®rmed by TUNEL assay with ¯uorescence staining (Figure 2a, low panel). The apoptotic cells were counted from ®ve low microscopic ®elds for each animal and the data are depicted in Figure 2b. The apoptotic cells in the liver of chronically ethanol-fed IL-6 (+/+) and pair-fed IL6 (+/+) and IL-6 (7/7) mice were less than 1%, but were signi®cantly increased (10.5+1.2%) in chronically ethanol-fed IL-6 (7/7) mice. IL-6 (7/7) hepatocytes are more susceptible to ethanol killing, which can be protected by adding IL-6 The data above clearly demonstrated that IL-6 (7/7) mice were more susceptible to ethanol apoptotic killing in the liver in vivo. This suggests that IL-6 protects ethanol-induced apoptosis in hepatocytes in vivo. To further con®rm this notion, we examined whether IL-6 (7/7) hepatocytes were more susceptible to ethanolinduced apoptosis in vitro. First, we tested whether cultured hepatocytes secreted IL-6. Hepatocytes from IL-6 (+/+) and IL-6 (7/7) mice were cultured in serum-free media for 24 h, the supernatant was collected and IL-6 levels were measured. As shown in Figure 3a, as expected, IL-6 levels were very low in the supernatant of IL-6 (7/7) hepatocytes, whereas in the supernatant of IL-6 (+/+) hepatocytes, they were 58.5+4.2 pg/ml. Interestingly, treatment of IL-6 (+/ +) hepatocytes with 100 mM ethanol for 24 h

Oncogene

Figure 2 Chronic ethanol consumption induces greater apoptosis of liver in IL-6 (7/7) mice than in IL-6 (+/+) mice. (a) Liver tissues from pair-fed or ethanol-fed IL-6 (+/+) or IL-6 (7/7) mice in Figure 1 were examined for apoptosis as described under Materials and methods. The top and low rows show TUNEL results of liver apoptosis from a representative of ethanol-fed IL-6 (+/+) and IL-6 (7/7) mice. (b) The apoptotic cells and hepatocytes were counted from 5 lower microscopic ®elds for each animal. *P50.01, in comparison with ethanol-fed IL-6 (+/+) group

signi®cantly reduced the IL-6 levels. These ®ndings suggest that cultured hepatocytes secret IL-6, and that ethanol inhibits such secretion. Next we examined the eects of ethanol on apoptosis of IL-6 (+/+) and IL-6 (7/7) hepatocytes by both DNA fragmentation (Figure 3b) and FASC analysis (Figure 3c). There was no signi®cant apoptosis in both untreated IL-6 (+/+) and IL-6 (7/7) primary cultures of hepatocytes as shown in Figure 3b and c, which is consistent with a previous report (Kovalovich et al., 2001). Ethanol treatment for 48 h did not cause signi®cant apoptosis in IL-6 (+/+) hepatocytes, whereas the same treatment for 24 and 48 h induced marked apoptosis in IL-6 (7/7) hepatocytes as demonstrated by DNA fragmentation (Figure 3b). Furthermore, ethanol induction of apoptosis in mouse IL-6 (+/+) and IL-6 (7/7) hepatocytes was also analysed by FASC analysis. As shown in Figure 3c, treatment of IL-6 (7/7) hepatocytes with 100 mM


35

Figure 3 IL-6 (7/7) hepatocytes are more susceptible to ethanol killing, which can be protected by adding IL-6. (a) Mouse hepatocytes from IL-6 (+/+) and IL-6 (7/7) mice were cultured in serum-free media in the absence or presence of 100 mM ethanol for 24 h, the supernatant was collected for measurement of IL-6 levels. (b) Mouse hepatocytes from IL-6 (+/+) and IL-6 (7/7) mice were cultured in serum-free media in the absence or presence of 100 mM ethanol for various time points as indicated. Apoptosis was measured by DNA fragmentation analysis as described under Materials and methods. (c) Mouse hepatocytes from IL-6 (+/+) and IL-6 (7/7) mice were cultured in serum-free media in the absence or presence of 100 mM ethanol and/or IL-6 for 24 h. Apoptosis (M1, sub-G1 peak) was measured by propodium iodide staining as described under Materials and methods. (b) and (c) are representatives of four independent experiments

ethanol induced signi®cant apoptosis (45.22% sub-G1 peak M1 [IL-67/7] vs 13.04% sub-G1 peak M1 [IL6+/+]). Pretreatment with 100 ng/ml IL-6 markedly protected ethanol-induced apoptosis in IL-6 (7/7) hepatocytes (2.69% sub-G1 peak M1 in IL-6+ETOH group vs 45.22% sub-G1 peak M1 in ETOH group). These ®ndings suggest that IL-6 (7/7) hepatocytes are more susceptible to ethanol-induced apoptotic killing, which can be protected by IL-6. IL-6 protects ethanol- and TNFa-induced apoptosis in primary mouse hepatocytes It has been shown that pretreatment of cells with ethanol increases sensitivity to TNFa-mediated apoptosis in rat hepatocytes and HepG2 cells (Pastorino and Hoek, 2000). We wondered whether IL-6

protected ethanol and TNF-mediated apoptosis in hepatocytes. To test this hypothesis, primary mouse hepatocytes were treated with 100 mM ethanol for 24 h, followed by incubation with ethanol and TNFa for an additional 24 h. Apoptosis was then analysed by measuring DNA fragmentation (Figure 4a) or using FASC analysis (Figure 4b). As shown in Figure 4, treatment of cells with either ethanol or TNFa alone did not induce signi®cant DNA fragmentation (Figure 4a), and only caused a small increase in number of apoptotic cells (Figure 4b: 0.82, 9.04 and 6.2% sub-G1 peak M1 in control, ETOH-, and TNFa-treated groups, respectively). Treatment of cells with ethanol and TNFa together induced signi®cant DNA fragmentation (Figure 4a) and increased apoptotic cells (Figure 4b, 33.55% sub-G1 peak M1 in ETOH+TNFa group). These eects of ethanol and Oncogene


36

Figure 4 IL-6 protects ethanol- and TNFa-induced apoptosis in primary mouse hepatocyte. (a) Mouse hepatocytes from IL-6 (+/ +) mice were cultured in serum-free media in the absence or presence of 100 mM ethanol and/or IL-6 for 48 h, followed by a 24-h treatment with various concentrations of TNFa. Apoptosis was measured by DNA fragmentation analysis as described under Materials and methods. (b and c) Mouse hepatocytes from IL-6 (+/+) mice were cultured in serum-free media in the absence or presence of 100 mM ethanol and/or IL-6 for 48 h, followed by a 24-h treatment with 20 ng/ml TNFa. Apoptosis (M1, sub-G1 peak) was measured by propoidium iodide staining (b) and ROS induction was measured as described under Materials and methods. Histograms for ROS (solid line) are shown overlaid on the control (dotted line) in (c). (a),(b) and (c) are representatives of four independent experiments

TNFa were protected by pretreatment of cells with IL-6 (Figure 4a and b: 9.19% sub-G1 peak M1 in IL6+ETOH+TNFa group vs 33.55% sub-G1 peak M1 in ETOH+TNFa group). These ®ndings suggest that ethanol increases the sensitivity of primary mouse hepatocytes to TNFa-mediated apoptosis, which can be protected by IL-6. Oxidative stress has been implicated in ethanol- and TNFa-mediated hepatocyte death (Fuchs et al., 1994; Oncogene

Kurose et al., 1997), we wondered whether IL-6 protected hepatocyte death by inhibiting oxidative stress. As shown in Figure 3c, treatment of primary hepatocytes with ethanol plus TNFa caused signi®cant ROS induction, pretreatment of cells with IL-6 completely prevented such ROS induction. These ®ndings strongly suggest that IL-6 prevents ethanoland TNFa-mediated hepatocyte death by inhibiting ROS induction.


37

Figure 5 Bcl-2 and Bcl-xL are elevated in the liver of ethanol-fed IL-6 (+/+) but not ethanol-fed IL-6 (7/7) mice. (a) Five IL-6 (+/+) mice were pair-fed with control diet for 8 weeks, ®ve IL-6 (+/+) mice and ®ve IL-6 (7/7) mice were fed with 5% ethanol containing diet for 4 weeks and 6% ethanol containing diet for additional 4 weeks. Liver tissues were prepared and subjected to immunohistochemical analysis by using anti-Bcl-2, anti-Bcl-xL, and anti-Bax antibodies as indicated. (b) Mice were injected intravenously with IL-6 (1 mg/g body weight) for various time points. Liver extracts were prepared and subjected to Western blotting by using anti-Bcl-2, anti-Bcl-xL, and anti-Bax antibodies as indicated. Each slide in (a) is a representative from ®ve animals. (b) is a representative of two independent experiments

Expression of Bcl-2 and Bcl-xL is elevated in the liver of chronically ethanol-fed IL-6 (+/+) but not IL-6 (7/7) mice To study the molecular mechanism by which chronic ethanol consumption induces hepatocyte apoptosis in IL-6 (7/7) but not in IL-6 (+/+) mice, expression of several anti-apoptotic and pro-apoptotic proteins was examined in pair-fed IL-6 (+/+) mice, ethanol-fed IL6 (+/+) mice, and ethanol-fed IL-6 (7/7) mice. As shown in Figure 5a, immunohistochemistry showed that expression of Bcl-2, Bcl-xL, and Bax proteins was signi®cantly elevated in the liver of ethanol-fed IL-6 (+/+) mice as compared with pair-fed IL-6 (+/+)

mice. On the contrary, expression of Bcl-2 and Bcl-xL proteins was not signi®cantly elevated in the liver of ethanol-fed IL-6 (7/7) mice whereas Bax protein expression remained elevated. Taken together, these ®ndings suggest that chronic ethanol consumption increases the expression of anti-apoptotic Bcl-2 and Bcl-xL proteins by an IL-6-dependent mechanism. To further con®rm this notion, mice were injected with 0.5 mg/g body weight for various time points and expression of several anti-apoptotic and proapoptotic proteins in the liver were examined. As shown in Figure 5b, injection of IL-6 markedly induced Bcl-2 and Bcl-xL expression whereas the expression of Bax was not induced. Oncogene


38

consistent with the chronically ethanol-fed mice model used in this study (Figure 5). Mutual antagonism between ethanol- and IL-6-mediated signals

Figure 6 Upregulation of anti-apoptotic (Bcl-2, Bcl-xL) and proapoptotic (Bax) proteins in human ALD. (a) Whole tissue extracts from 10 normal control human liver tissues (C1 ± 10) and 10 ALD tissues (A1 ± 10) were prepared and subjected to Western blotting analysis by using various antibodies as indicated. (b) The densities of bands on a were quanti®ed with a PhosphoImager. *P50.01, P**50.05, in comparison with control group

Upregulation of anti-apoptotic (Bcl-2, Bcl-xL) and pro-apoptotic (Bax) proteins in human ALD The above data clearly demonstrated that the expression of Bcl-2, Bcl-xL and Bax was markedly elevated in the liver of chronically ethanol-fed mice. Next we examined whether the expression of these antiapoptotic and pro-apoptotic proteins was also increased in the liver of human ALD. As shown in Figure 6, Western blotting analysis demonstrated that anti-apoptotic (Bcl-2 and Bcl-xL) and proapoptotic (Bax and caspase-3) proteins were signi®cantly increased in the liver tissue from human ALD, whereas other apoptosis-related proteins (such as caspase-7, PCNA, Fas ligand, Nip1, and Nm 23) remained unchanged. Quantitation analysis showed that expression of Bcl-2, Bcl-xL, Bax, and caspase-3 in the liver of human ALD was increased by 4.23-, 2.85-, 3.65-, 2.2fold as compared with control healthy livers. These ®ndings suggest that both anti-apoptotic (Bcl-2 and Bcl-xL) and pro-apoptotic (Bax) proteins are signi®cantly elevated in the liver of human ALD, which is Oncogene

The data above data indicated that IL-6 protected ethanol-induced apoptosis in the liver, whereas we have previously demonstrated that ethanol inhibited IL-6 activation of STAT3 (Chen et al., 1997, 1999a), a critical anti-apoptotic signal. These ®ndings suggest that ethanol and IL-6 antagonize each other. To further con®rm this notion, we examined the eect of 100 mM ethanol on various concentrations of IL-6-induced STAT3 activation in hepatocytes. As shown in Figure 7a, acute exposure of freshly isolated hepatocytes with 100 mM ethanol markedly attenuated STAT3 activation induced by 10 or 20 ng/ml of IL-6, but did not signi®cantly inhibit STAT3 activation induced by 100 ± 500 ng/ml of IL-6. Figure 7b showed that chronic ethanol consumption signi®cantly attenuated STAT3 activation induced by 10 ng/ml of IL-6, but did not inhibit STAT3 activation induced by 100 ng/ml of IL-6 in hepatocytes. Furthermore, Figure 7c showed that 75 ± 100 mM ethanol treatment completely abolished STAT3 activation induced by 10 ng/ml of IL-6, whereas 50 mM ethanol slightly inhibited such activation, and 20 mM ethanol had no eect on IL-6 activation of STAT3. Taken together, these ®ndings suggest that ethanol inhibits IL-6-mediated STAT3 activation, but increasing the concentrations of IL-6 antagonizes such inhibitory eect. Discussion Although it has been shown that chronic ethanol consumption increases apoptosis of hepatocytes in mice (Goldin et al., 1993; Deaciuc et al., 2001); rats (Benedetti et al., 1988; Kurose et al., 1997; Mi et al., 2000; Slomiany et al., 1999; Yacoub et al., 1995); minipigs (Halsted et al., 1996), and humans (Zhao et al., 1997), the numbers of apoptotic hepatocytes are small. Others (Rashid et al., 1999) have shown, and we have con®rmed here (Figure 2), that there is no signi®cant apoptosis in the liver of mice fed chronically with the DeCarli-Lieber ethanol containing liquid diet. These ®ndings suggest that hepatocytes are resistant to ethanol-induced apoptosis in vivo, or some protective pathways may exist in vivo. Our ®ndings here demonstrate that IL-6 is one such protective pathway that inhibits ethanol-mediated apoptosis in vivo, and is summarized in a model in Figure 8. In this model, chronic ethanol consumption increases the expression of pro-apoptotic Bax protein that threatens hepatocyte viability, and also inhibits IL-6-activated anti-apoptotic signals (STAT3). However, chronic ethanol consumption also signi®cantly increases the serum IL-6 levels that can overcome the inhibitory eect of ethanol on IL-6 signaling, followed by stimulation of expression of anti-apoptotic Bcl-2 and Bcl-xL proteins that protect ethanol-induced apoptosis in the liver and lead to


39

Figure 7 Ethanol inhibits IL-6-activated STAT3 but increasing the concentrations of IL-6 antagonizes such inhibitory eect. (a) Freshly isolated rat hepatoctyes were incubated with or without 100 mM ethanol for 30 min, followed by a 30-min stimulation with various concentrations of IL-6 as indicated. (b) Three rats were fed with DeCarli-Lieber liquid diet containing 7% ethanol for 6 weeks; three rats were pair-fed with control diet for 6 weeks. Hepatocytes were then isolated and stimulated with 100 ng/ml or 10 ng/ml of IL-6 for 30 min. (c) Freshly isolated rat hepatocytes were incubated with various concentrations of ethanol for 30 min, followed by a 30-min stimulation with 10 ng/ml of IL-6. Whole cell extracts from a, b, and c were prepared and subjected to gel shift assay by using m67 oligo as a probe

mediated signals, and protect ethanol-mediated injury in the liver. The rationale for this model is presented as follows. IL-6 plays an important role in protection of ethanol-induced apoptosis in the liver in vivo

Figure 8 Mutual antagonism between ethanol- and IL-6mediated signals. Chronic ethanol consumption increases the expression of proapoptotic Bax protein that threatens hepatocyte viability, and also inhibits IL-6-activated STAT3 signaling by PKC- and SOCS-dependent mechanisms. However, chronic ethanol consumption also signi®cantly increases the serum IL-6 levels that overcome the inhibitory eect of ethanol on IL-6 signaling, followed by stimulation of expression of anti-apoptotic Bcl-2, Bcl-xL, and Mcl-1 proteins that protect ethanol-induced apoptosis in the liver and lead hepatocyte survival

hepatocyte survival. Therefore, elevated serum IL-6 levels in ALD patients are compensatory eects that overcome the suppressive eect of ethanol on IL-6-

Serum IL-6 levels were signi®cantly elevated in chronic ethanol feeding mice as demonstrated in Figure 1, however, the underlying mechanism is not clear. Impaired hepatocyte endocytosis of IL-6 (Tuma et al., 1996) during chronic ethanol consumption and elevated endotoxin that stimulates IL-6 expression (Panesar et al., 1999) could be mechanisms responsible for elevation of serum IL-6 levels. In contrast, acute exposure of primary hepatocytes to ethanol decreased the IL-6 secretion, which is probable due to an increase of IL-6 uptake after acute ethanol exposure (Deaciuc et al., 1996). Furthermore, we provided several lines of evidence suggesting that elevated IL6 protects ethanol-induced apoptosis in the liver. First, we have demonstrated that chronic ethanol consumption induced greater apoptosis in the liver of IL-6 (7/ 7) mice than IL-6 (+/+) mice. It has also been shown that IL-6 (7/7) mice show greater hepatic apoptosis than IL-6 (+/+) mice after bile duct ligation- or Fas-induced liver injury (Ezure et al., 2000; Kovalovich et al., 2001). Taken together, these ®ndings suggest that IL-6 generally protects liver apoptosis induced by a wide variety of toxicants including alcohol. Second, we demonstrated that IL-6 (7/7) hepatocytes were more susceptible to ethanol Oncogene


40

apoptotic killing, which can be corrected by IL-6. In addition to ethanol, Fas also induced much higher apoptosis in IL-6 (7/7) hepatoctyes as demonstrated recently by Taub's group (Kovalovich et al., 2001). They further claimed that this is at least in part due to lack of expression of anti-apoptotic (such as Bcl-2, Bcl-xL, and FLIP) proteins in IL-6 (7/7) hepatocytes, which may also contribute to the susceptibility of IL-6 (7/7) hepatocytes to ethanol-induced apoptotic killing. In addition to this pre-existing anti-apoptotic defect, lack of IL-6 secretion in IL-6 (7/7) hepatocytes may also contribute to ethanol apoptotic killing in IL-6 (7/7) hepatocytes as shown in Figure 3. Third, IL-6 is able to attenuate hepatocyte apoptosis induced by other toxicants such as TGFb (Chen et al., 1999b; Kuo et al., 2001). In Figure 3, we showed that 100 ng/ml of IL-6 prevented ethanol-induced apoptosis in primary hepatocytes. Although this concentration appear to be high as compared to serum levels of IL-6 in Figure 1, expression and secretion of IL-6 by hepatocytes and Kuper cells in alcoholic liver were markedly enhanced (Gonzalez-Quintela et al., 2000; Kamimura and Tsukamoto, 1995; Pennington et al., 1997), which could form paracrine and autocrine loops in the liver, causing high IL-6 levels within the alcoholic liver. Upregulation of the expression of proapoptotic Bax protein in the liver of chronically ethanol-fed mice and human ALD We have demonstrated that expression of Bax was markedly elevated in the liver of chronically ethanolfed IL-6 (+/+) and IL-6 (7/7) mice, as well as human ALD (Figures 5 and 6). Furthermore, injection of IL-6 did not signi®cantly induce Bax expression in the liver (Figure 5). Taken together, these ®ndings suggest that upregulation of Bax during chronic ethanol consumption is not mediated by IL-6. Since Bax is one of the key factors involved in induction of apoptosis (Adams and Cory, 2001), upregulation of Bax during chronic ethanol consumption may contribute to ethanol-induced apoptosis in the liver. Ethanol inhibits IL-6-activated STAT3 signaling and increasing the concentrations of IL-6 overwhelms such inhibitory effect It has been shown that IL-6-activated STAT3 plays an important role in liver regeneration, protection of liver injury, and anti-apoptosis in the liver (Cressman et al., 1996; Kovalovich et al., 2000, 2001). We have demonstrated that both acute and chronic ethanol signi®cantly inhibited IL-6-activated STAT3 (Chen et al., 1997, 1999a). Inhibition of PKC partially antagonized the inhibitory eect of ethanol on IL-6activated STAT3 (unpublished data), suggesting that ethanol attenuates IL-6-activated STAT3 by a PKCdependent mechanism. We have also demonstrated that induction of suppressor of cytokine signaling (SOCS) proteins may be involved in chronic ethanol

Oncogene

consumption-mediated suppression of IL-6 activation of STAT3 (unpublished data). Although ethanol inhibits IL-6-mediated activation of STAT3, increasing the concentrations of IL-6 is able to overcome such inhibitory eect as demonstrated in Figure 7. In this ®gure, we showed that as low as 20 ng/ml of IL-6 partially overcame whereas 100 ng/ml of IL-6 completely abolished the inhibitory eect of ethanol on IL-6-activated STAT3. Thus, it is plausible that elevated serum IL-6 levels in alcoholic patients and chronically ethanol-fed mice are compensatory eects that overcome the inhibitory eect of ethanol on IL-6activated STAT3, and protect from ethanol-induced liver injury. Elevated serum IL-6 levels induce the expression of anti-apoptotic Bcl-2, Bcl-xL proteins in the liver of chronically ethanol-fed mice and human ALD We have demonstrated that expression of Bcl-2 and Bcl-xL was markedly elevated in the liver of chronically ethanol-fed mice and human ALD (Figures 5 and 6). However, Bcl-2 and Bcl-xL were not signi®cantly increased in the liver of chronically ethanol-fed IL-6 (7/7) mice (Figure 5). Furthermore, injection of IL-6 markedly induced expression of Bcl-2 and Bcl-xL in the liver (Figure 5). These ®ndings strongly suggest that elevated serum IL-6 levels in ethanol-fed mice are responsible for the increased Bcl-2 and Bcl-xL. It has been reported that expression of Bcl-2 and Bcl-xL was controlled by STAT (Grad, 2000), and that IL-6 stimulated the expression of Bcl-2 and Bcl-xL by a transcription-dependent mechanism. Recently, Kovalovich et al. (2001) demonstrated that FLIP, Bcl-2, and Bcl-xL protein levels were markedly decreased in the liver of IL-6 (7/7) mice, but mRNA levels remained unchanged when compared with IL-6 (+/+) mice, suggesting that IL-6 may also regulate these antiapoptotic proteins by a post-transcriptional mechanism. Induction of Bcl-2 by IL-6 was also observed in acute lung injury (Ward et al., 2000) and in enterocytes during hemorrhage shock (Rollwagen et al., 1998). Although we did not provide direct evidence for a role of Bcl-2 or Bcl-xL in the eects of ethanol in vivo, or with cultured hepatocytes, there are several references showing that overexpression of Bcl-2 and Bcl-xL in the liver or hepatic cells prevented apoptosis induced by anti-Fas antibody, Fas ligand, LIGHT, interferon-g, and ethanol (Chang and Xu, 2000; Lacronique et al., 1996; Zhang et al., 2000; Chen et al., 2000). Taken together, these ®ndings suggest that upregulation of Bcl-2 and Bcl-xL is involved in IL-6 protection of organ injury induced by a variety of toxicants, including ethanol. In summary, our ®ndings here suggest that elevated serum IL-6 levels in alcoholic patients may overcome the inhibitory eect of ethanol on IL-6-mediated antiapoptotic signals and prevent alcohol-induced hepatic apoptosis by inducing anti-apoptotic proteins. Recently, it has been shown that IL-6 gene therapy signi®cantly protects D-galactosamine-induced hepato-


cellular injury (Galun et al., 2000; Hecht et al., 2001), hypoxia-induced acute lung injury (Ward et al., 2000), and hemorrhage-induced enterocyte apoptosis (Rollwagen et al., 1998). Thus, IL-6 could be used in the treatment of fulminant alcoholic hepatic failure, which is often fatal. Furthermore, anti-IL-6 therapy should be carefully designed.

Materials and methods Materials IL-6 (+/+) and IL-6 (7/7) mice were purchased from Jackson Laboratory (Bar Harbor, Maine, USA). Anti-Bcl-2, anti-Bcl-xL and anti-Bax antibodies were purchased from Pharmingen (San Diego, CA, USA). Anti-caspase-3, anticaspase-7, anti-PCNA, anti-Fas ligand, anti-Nip1, anti-Nm23 antibodies were obtained from Transduction Laboratories (Lexington, KY, USA). IL-6, TNFa, ethanol were purchased from Sigma (St. Louis, MO, USA). Primary mouse hepatocyte isolation and culture Primary mouse hepatocytes were isolated and cultured as described previously (Kim et al., 2001). Brie¯y, IR mice weighing 30 ± 35 g were anesthetized with sodium pentobarbital (30 mg/kg intraperitoneally), and the portal vein was cannulated under aseptic conditions. The liver was perfused with EGTA solution (5.4 mM KCl, 0.44 mM KH2PO4, 140 mM NaCl, 0.34 mM Na2HPO4, 0.5 mM EGTA, 25 mM Tricine, pH 7.2) and Dulbecco's modi®ed Eagle medium (Gibco BRL, NY, USA), and digested with 0.075% collagenase solution for mouse liver. The isolated mouse hepatocytes were then cultured in Hepato ± ZYME ± SFM media in rat-tail collagen coated plates and treated with IL-6 and/or ethanol and/or TNFa for various time points as indicated in the text. Human ALD specimens Human ALD specimens were provided by LTPDS program. These patients had history of more than 15 years heavy alcohol drinking and had no viral hepatitis B and C infection. Normal health liver specimens were also provided by LTPDS program. The protocol was approved by LTPDS program. ELISA IL-6 was measured by using standard ELISA sandwich kit as speci®ed by the manufacturer (Biosource International, Camarillo, CA, USA). Western blotting Western blot analysis was performed as described previously (Hong et al., 2001). Brie¯y, liver tissues were lysed in lysis buer (30 mM Tris, pH 7.5, 150 mM sodium chloride, 1 mM phenylmethylsulfonyl ¯uoride, 1 mM sodium orthovanadate, 1% Nonidet P-40, 10% glycerol) for 15 min at 48C, vortexed and centrifuged at 16 000 r.p.m. at 48C for 10 min. The supernatants were mixed in Laemmli running buer, boiled for 4 min, and then subjected to SDS ± PAGE. After electrophoresis, proteins were transferred onto nitrocellulose membranes and blotted against primary antibodies for 16 h. Membranes were washed with TPBS (0.05% [vol/vol] Tween

20 in phosphate-buered saline [pH 7.4]) and incubated with a 1 : 4000 dilution of horseradish peroxidase-conjugated secondary antibodies for 45 min. Protein bands were visualized by an enhanced chemiluminescence reaction (Amersham Pharmacia Biotech, Piscataway, NJ, USA).

41

Flow cytometry After trypsination, approximately 106 cells were collected by centrifugation at 1000 g for 5 min. Cells were then washed in PBS followed by resuspension and ®xation in 70% ethanol for approximately 2 h. Next, cells were washed with phosphatebuered saline and resuspended in 500 ml phosphate-buered saline containing 100 mg RNase, followed by a 30-min incubation at 378C. Cellular DNA was then stained by the addition of 10 mg propidium iodide, and cells were analysed on a FACScan by utilizing Cellquest software (Becton Dickinson). DNA fragmentation in cultured hepatoctyes Cultured hepatocytes were washed twice with phosphatebuered saline and lysed in buer (100 mM NaCl, 10 mM TrisHCl, pH 8.0, 25 mM EDTA, 0.5% SDS, and 0.1 mg/ml proteinase K) at 378C for 18 h. DNA was extracted with an equal volume of phenol/chloroform (1 : 1) and precipitated at 7708C. DNA pellets were resuspended in 10 ml of 10 mM Tris (pH 7.8), 1 mM EDTA buer and incubated for 1 h at 378C with 1 mg/ml RNase (Roche Molecular Biochemicals) to remove RNA. DNA pellets were electrophoresed for 2 ± 3 h at 90 v on 1.8% agarose gels. The gel was stained with ethidium bromide and the DNA fragments were visualized under ultraviolet light. Terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) The liver tissues were removed and ®xed for 24 h in 4% paraformaldehyde, dehydrated in graded ethanol and embedded in paran. Paran sections of 4 mm were adhered to slides pretreated with a 0.01% aqueous solution of poly-Llysine (Mr. 300 000, Sigma) and then processed for staining with hematoxylin and eosin. Apoptotic cells on sections of mouse livers were detected by using Apop Tag (Oncor, Gaitherburg, MD, USA), an in situ apoptosis detection kit. Brie¯y, sections prepared as described above were incubated with terminal transferase and digoxigenin-labeled-dUTP, followed by incubation with anti-digoxigenin mAb conjugated to horseradish peroxidase. The slides were developed with peroxidase substrate and counterstained with methylene green. The percentage of apoptotic hepatocytes was quantitated by determining the mean number of apoptotic hepatoctyes as a percentage of total hepatocytes in 5 low power ®elds (magni®cation 2006). Apoptosis was also determined by TUNEL staining with terminal transferase and avidine-labeled-dUTP, followed by incubated with FITClabeled anti-avidine antibody. The slides were then examined under ¯uorescence microscope. Immunohistochemistry Formalin-®xed, paran-embedded tissues were deparanized and rehydrated with phosphate-buered saline, followed by hyaluronidase treatment (1 mg/ml in 100 mmoll sodium acetate, 150 mmoll NaCl [pH 5.5]; 30 min at room temperature). Tissue sections were permeabilized using 0.1% Triton X-100 (diluted in 0.1% sodium citrate, 2 min on ice), and endogenous biotin was blocked using the DAKO biotin blocking system (DAKO A/S, Glostrup, Denmark). UnOncogene


42

speci®c binding sites were blocked by 30-min incubation in 50% fetal calf serum/0.1% bovine serum albumin (BSA), followed by another 30 min in 1% BSA/0.1% ®sh gelatin. Sections were incubated with 1 : 50 diluted primary antibodies for 1.5 h at 378C. Secondary antibodies (biotinylated; DAKO A/S) were diluted 1 : 50, and sections were incubated for 45 min at 378C, followed by incubation with streptavidinalkaline phosphatase (1 : 50 in 0.1% BSA, 30 min at 378C; DAKO A/S). Color development was induced using fast-red substrate (DAKO A/S) during a 5 to 30 min incubation period. Using this substrate, speci®c staining can be visualized by using light microscopy.

Assessment of reactive oxidative species (ROS) production ROS was determined with the cell-permeable ¯uorogenic probe 2',7'-dichloro¯uorescin diacetate (DCF-DA). The assay is based on the ¯uorescence detection of dichloro¯uorescein (DCF), which is formed by hydrogen peroxide oxidation of the non-¯uorescent precursor dichloro¯uorescin. Primary mouse hepatocytes were incubated with 10 mg/ml DCF-DA for 1 h at 378C. After incubation, cells were harvested by trypsination, washed with RPMI 1640 and resuspended in PBS. Intracellular DCF was analysed with a FACSscan by utilizing Cellquest software (Becton Dickinson).

Chronic ethanol consumption mice model

Statistics analysis

Seven male IL-6 (+/+) (mean initial weight 30 g) and seven male IL-6 (7/7) mice (mean initial weight 29.3 g), were fed with a nutritionally adequate DeCarli-Lieber liquid diet (Bioserv, Frenchtown, NJ, USA) containing 4% ethanol for 4 weeks, then switched to the same diet containing 5% ethanol for additional 4 weeks. Seven weight-matched IL-6 (+/+) littermates (mean initial weight 31 g) and seven IL-6 (7/7) littermates (mean initial weight 30.2 g) were pair-fed on the same liquid diet, except that ethanol was replaced by carbohydrate.

For comparing values obtained in three or more groups, onefactor analysis of variants was used, followed by Tukey's post hoc test, and P50.05 was taken to imply statistical signi®cance.

Gel mobility shift assay

Abbreviations ALD, alcoholic liver disease; Bax, Bcl-2-associated X protein; Bcl-2, B cell lymphoma-2 associated oncogene; ETOH, ethanol; TNFa; tumor necrosis factor a; IL-6, interleukin-6; IL-6 (7/7), IL-6 null; IL-6 (+/+), IL-6 wild type; STAT3, signal transducer and activator transcription factor 3.

Activation of STAT3 DNA binding was determined by gel mobility shift assay, as described previously (Chen et al., 1997), using m67 (a high anity serum induce element m67) (5'-GTG CAT TTC CCG TAA ATC TTG TCT ACA-3').

Acknowledgments This work is supported by National Institutes of Health intramural program and grant R01AA12637.

References Abittan CS and Lieber CS. (1999). Curr. Treat Options Gastroenterol., 2, 72 ± 80. Adams JM and Cory S. (2001). Trends Biochem. Sci., 26, 61 ± 66. Benedetti A, Brunelli E, Risicato R, Cilluo T, Jezequel AM and Orlandi F. (1988). J. Hepatol., 6, 137 ± 143. Bird G. (1994). Acta. Gastroenterol. Belg., 57, 255 ± 259. Chang YC and Xu YH. (2000). Cell Res., 10, 233 ± 242. Chen J, Bao H, Sawyer S, Kunos G and Gao B. (1997). Biochem. Biophys. Res. Commun., 239, 666 ± 669. Chen J, Ishac EJ, Dent P, Kunos G and Gao B. (1998). Biochem. J., 334, 669 ± 676. Chen J, Kunos G and Gao B. (1999a). FEBS Lett., 457, 162 ± 168. Chen MC, Hsu TL, Luh TY and Hsieh SL. (2000). J. Biol. Chem., 275, 38794 ± 38801. Chen RH, Chang MC, Su YH, Tsai YT and Kuo ML. (1999b). J. Biol. Chem., 274, 23013 ± 23019. Cressman DE, Greenbaum LE, DeAngelis RA, Ciliberto G, Furth EE, Poli V and Taub R. (1996). Science, 274, 1379 ± 1383. Deaciuc IV, Alappat JM, McDonough KH and D'Souza NB. (1996). Alcohol. Clin. Exp. Res., 20, 293 ± 301. Deaciuc IV, Fortunato F, D'Souza N, Hill D and McClain C. (2001). Hepatol. Res., 19, 306 ± 324. Diehl AM. (1989). Med. Clin. North Am., 73, 815 ± 830. Diehl AM. (1999a). Alcohol. Clin. Exp. Res., 23, 1419 ± 1424. Diehl AM. (1999b). Clin. Biochem., 32, 571 ± 578. Ezure T, Sakamoto T, Tsuji H, Lunz JG, III, Murase N, Fung JJ and Demetris AJ. (2000). Am. J. Pathol., 156, 1627 ± 1639. Oncogene

Fontanilla CV, Faunce DE, Gregory MS, Messingham KA, Durbin EA, Duner LA and Kovacs EJ. (2000). Alcohol. Clin. Exp. Res., 24, 1392 ± 1399. Fuchs D, Gruber A, Uberall F and Wachter H. (1994). Immunol. Today, 15, 496. Galun E, Zeira E, Pappo O, Peters M and Rose-John S. (2000). FASEB J., 14, 1979 ± 1987. Goldin RD, Hunt NC, Clark J and Wickramasinghe SN. (1993). J. Pathol., 171, 73 ± 76. Gonzalez-Quintela A, Dominguez-Santalla MJ, Perez LF, Vidal C, Lojo S and Barrio E. (2000). Cytokine, 12, 1437 ± 1440. Grad J, Zeng X and Boise L. (2000). Curr. Opin. Oncol., 12, 543 ± 549. Hall PD. (1994). Alcohol. Suppl., 2, 303 ± 313. Halsted CH, Villanueva J, Chandler CJ, Stabler SP, Allen RH, Muskhelishvili L, James SJ and Poirier L. (1996). Hepatology, 23, 497 ± 505. Hecht N, Pappo O, Shouval D, Rose-John S, Galun E and Axelrod JH. (2001). Mol. Ther., 3, 683 ± 687. Hill DB, Marsano L, Cohen D, Allen J, Shedlofsky S and McClain CJ. (1992). J. Lab. Clin. Med., 119, 547 ± 552. Hong F, Nguyrn VA and Gao B (2001). FASEB J., 15, 1595 ± 1597. Hong F, Nguyen VA, Shen X, Kunos G and Gao B. (2000). Biochem. Biophys. Res. Commun., 279, 974 ± 979. Kamimura S and Tsukamoto H. (1995). Hepatology, 22, 1304 ± 1309. Kim W, Hong F, Jaruga B, Hu Z, Fan S, Liang J and Gao B. (2001). FASEB J., 15, in press


Kovalovich K, DeAngelis RA, Li W, Furth EE, Ciliberto G and Taub R. (2000). Hepatology, 31, 149 ± 159. Kovalovich K, Li W, DeAngelis R, Greenbaum LE, Ciliberto G and Taub R. (2001). J. Biol. Chem., 276, 26605 ± 26613. Kuo ML, Chuang SE, Lin MT and Yang SY. (2001). Oncogene, 20, 677 ± 685. Kurose I, Higuchi H, Miura S, Saito H, Watanabe N, Hokari R, Hirokawa M, Takaishi M, Zeki S, Nakamura T, Ebinuma H, Kato S and Ishii H. (1997). Hepatology, 25, 368 ± 378. Lacronique V, Mignon A, Fabre M, Viollet B, Rouquet N, Molina T, Porteu A, Henrion A, Bouscary D, Varlet P, Joulin V and Kahn A. (1996). Nat. Med., 2, 80 ± 86. Laso FJ, Iglesias-Osma C, Ciudad J, Lopez A, Pastor I and Orfao A. (1999). Alcohol. Clin. Exp. Res., 23, 1306 ± 1311. Maddrey WC. (2000). Clin. Liver Dis., 4, 115 ± 131, vii. Martinez F, Abril ER, Earnest DL and Watson RR. (1992). Alcohol, 9, 455 ± 458. McClain CJ, Barve S, Deaciuc I, Kugelmas M and Hill D. (1999). Semin. Liver Dis., 19, 205 ± 219. Mi LJ, Mak KM and Lieber CS. (2000). Alcohol. Clin. Exp. Res., 24, 207 ± 212. Panesar N, Tolman K and Mazuski JE. (1999). J. Surg. Res., 85, 251 ± 258. Pastorino JG and Hoek JB. (2000). Hepatology, 31, 1141 ± 1152. Pennington HL, Hall PM, Wilce PA and Worrall S. (1997). J. Gastroenterol. Hepatol., 12, 305 ± 313.

Rashid A, Wu TC, Huang CC, Chen CH, Lin HZ, Yang SQ, Lee FY and Diehl AM. (1999). Hepatology, 29, 1131 ± 1138. Rollwagen FM, Yu ZY, Li YY and Pacheco ND. (1998). Clin. Immunol. Immunopathol, 89, 205 ± 213. Sasaki Y and Wands JR. (1994). Biochem. Biophys. Res. Commun., 199, 403 ± 409. Sheron N, Bird G, Goka J, Alexander G and Williams R. (1991). Clin. Exp. Immunol., 84, 449 ± 453. Slomiany A, Piotrowski E, Grabska M, Piotrowski J and Slomiany BL. (1999). Alcohol. Clin. Exp. Res., 23, 334 ± 343. Tuma DJ, Todero SL, Barak-Bernhagen M and Sorrell MF. (1996). Alcohol. Clin. Exp. Res., 20, 579 ± 583. Walsh K and Alexander G. (2000). Postgrad. Med. J., 76, 280 ± 286. Ward NS, Waxman AB, Homer RJ, Mantell LL, Einarsson O, Du Y and Elias JA. (2000). Am. J. Respir. Cell. Mol. Biol., 22, 535 ± 542. Yacoub LK, Fogt F, Griniuviene B and Nanji AA. (1995). Alcohol. Clin. Exp. Res., 19, 854 ± 859. Yin M, Wheeler MD, Kono H, Bradford BU, Gallucci RM, Luster MI and Thurman RG. (1999). Gastroenterology, 117, 942 ± 952. Zhang B, Zhang D and Ren H. (2000). Zhonghua Gan Zang Bing Za Zhi, 8, 215 ± 217. Zhao M, Laissue JA and Zimmermann A. (1997). Virchows Arch, 431, 337 ± 344.

43

Oncogene