- Email: [email protected]
Nitric oxide production by monocytes in alcoholic liver disease
Journal of Hepatology, 1992; 14: 146- 150 6 1992Elsevier Science Publishers B.V. All rights reserved. 0168-8278/92/$05.00
146 HEPAT 00995
Nitric oxide production by monocytes in alcoholic liver disease
N.C.A. Hunt and R.D. Goldin Department of Histopathology, St. Mary’s Hospital Medical School, London, United Kingdom
(Received 7 January 1990)
Nitric oxide, initially described as an endothelial-derived relaxing factor, has recently been recognised as a mediator of macrophage function. We have studied the production of nitric oxide by peripheral blood monocytes from both normal volunteers and alcoholics. This was measured indirectly by assessing nitrite formation. Normal monocytes were found to produce a basal level of nitrite, which could be stimulated more than 6-fold using endotoxin. This effect was abrogated by the addition of nitric oxide synthesis inhibitor, L-n-monomethyl-arginine. A striking difference was observed in the monocytes obtained from alcoholics with and without evidence of alcoholic hepatitis. Whereas the latter behaved in a similar manner to the controls, the former had markedly increased basal levels. In the hepatitis group there was also substantial inhibition of production by t_-n-monomethyl-arginine. We believe that these results indicate that nitric oxide derived from monocytes may play a role in the pathogenesis of alcoholic liver disease, especially alcoholic hepatitis.
Much work has been carried out on the role of various cells, including hepatocytes, neutrophils and lymphocytes, in the pathogenesis of alcoholic liver disease (ALD) (1). There is increasing evidence to suggest that cells of the mononuclear phagocyte family are involved, as effector cells, in this process (2). Of particular interest are the hepatic sinusoidal macrophages (Kupffer cells) and, more recently, peripheral blood monocytes; both, when stimulated, are known to produce a variety of mediators including interleukin-1, and tumour necrosis factor (cachectin) (3). Early reports suggested a decrease in the number of hepatic sinusoidal macrophages, as assessed by measuring the number of lysozyme-positive cells, in patients with increasingly severe ALD (4). Thus, the decrease was greatest in patients with cirrhosis and least in those with steatosis alone. Later work supported this decrease in lysozyme-positive cells. However, using a number of other antibodies it was also shown that certain classes of hepatic sinusoidal macrophages were in fact more abundant and overall there is an increase in num-
bers of hepatic sinusoidal macrophages, even in the earliest stages of ALD (5). It has been suggested that functional impairment of hepatic sinusoidal macrophages may be of greater importance in ALD than the decrease in cell number. In support of this it has been shown that alcohol-exposed macrophages produce decreased amounts of lysozymes (6). Decreased macrophage function in conjunction with portosystemic shunting, in certain groups of ALD patients, is believed to contribute to the endotoxaemia seen in these people (6,7). Endotoxin is itself a known activator of monocytes and macrophages (8). The fact that endotoxin stimulates the production of nitric oxide by macrophages, including hepatic sinusoidal macrophages has been well documented (9). It has been shown that endotoxin can stimulate monocytes to produce increased amounts of tumour necrosis factor and, moreover, the monocytes of alcoholics show an increased basal production of this mediator (10). The object of the present study was to investigate the production of nitric oxide
Correspondence: R.D. Goldin, Department of Histopathology, St. Mary’s Hospital Medical School, London W2, U.K.
MONOCYTJZS, NITRIC OXIDE
AND ALCOHOLICS
147
by monocytes in patients with ALD and to see to what extent they, and normal monocytes, responded to endotoxin.
NMMA: Sigma). All samples were incubated for 24 h at 37 “C, 5% CO,/95% air, before the supematants were removed for subsequent nitrite assay
Materiais and Methods
Nitrite assay Nitric oxide production was measured indirectly using a quantitative, calorimetric assay based on the Griess reaction, a sensitive assay for nitrite ions (11). Nitrite and nitrate ions are the stable, anionic, products formed in a molar ratio of 3:2 from nitric oxide as a result of its decompos8ttion (12). In the present study triplicate 100~,ulaliquots of supernatants were added to an equal volume of freshly prepared Griess reagent. In this system nitrite ions react with 1% sulphanilamide in 5% ortho-phosphoric acid/O. 1% N-l-naphthylethylenediamine dihydrochloride to yield an azo-chromophore, the absorbance of which was measured at 550 nm using a Titretek 96well plate reader. All samples were assayed against a blank comprising RPMI-1640 medium incubated for 24 h in the same plates as the samples but in the absence of cells.
Subjects .4 total of 54 subjects were recruited for the study. The control group consisted of 15 normal volunteers (mean age = 37.7 years, age range = 19-61, eight males and seven females). The alcoholic hepatitic group consisted of 22 patients, all with histories of excessive alcohol intake (mean age = 42.8 years, age range = 18-65,14 males and eight females). All were out-patients with histological diagnosis of alcoholic hepatitis with varying degrees of fibrosis. In addition, a third group of 17 alcohol abusers with non-hepatic liver disease were studied (mean age = 39.0 years, age range = 23-57, ten males and seven females). These patients ail had liver biopsies showing fatty change with varying degrees tif fibrosis but no evidence of alcoholic hepatitis. In each case serological markers of viral hepatitis, including hepatitis C virus, were negative. All biopsies were reviewed prior to inclusion in the study in order to confirm the histological diagnosis (RDG.) Monocyte extraction and culture Monocytes were extracted from peripheral blood on a density gradient materiai (IEstopaque, Sigma). This and all subsequent procedures required specific precautions in order to minimise contamination by LPS. Such measures included baking of all laboratory glassware, use of pyrogen-free disposables and endotoxin-free culture media. Briefly, peripheral blood mononuclear cells were removed from the density gradient, washed twice in endotoxin-free Hanks’ Balanced Salts Solution, and finally resuspended in RPMI-1640 (RPM]-1640, endotoxin-free. Phenol red-free; Sigma) supplemented with 10% decomplemented fetal calf serum (low endotoxin; Flow, U.K.) and 2 mmol t_-glutamine (Sigma). Cells were then plated out at a concentration of 2-4-l@ per ml in individual 16mm wells of tissue cuiture plates (24 well; Nunclon). A Trypan blue exclusion analysis was performed at this stage and showed a mean viability of 95% with a range of 92-97%. Monocyte enrichment was achieved by allowing the cells to adhere to the plates for 2 h at 37 “C, after which time the medium was removed by aspiration after vigorous washing to remove any non-adherent cells. For the final incubation cells were cultured in either RPMI-1640 (supplemented as above) alone, medium with the addition of 1 &ml endotoxin (Escherichia co/i lipopolysaccharide, L4516; Sigma), or medium with 1 ,@ml endotoxin plus 300 ,uM t-n-monomethyl-arginine (L-
Protein assay Briefly, PBMCs were removed from the density gradient, washed twice in endotoxin-free Hanks’ Balanced Salts Solution, and finally resuspended in RMPI-1640 medium (RMPI-1640, low endotoxin, Phenol red-free; Sigma Chemicals) supplemented with 10% decomplemented fetal bovine serttm (low endotoxin; Flow, U.K.). Cells iyere then plated out at a concentration of plates (Nunclan, 24 well). A Trypan blue exclusion analysis was performed at this stage and showed a mean viability of 95% with a range of 92-97%. Monocyte enrichment was achieved by allowing the cells to adhere to the plates for 2 h at 37 “C, after which time the medium was removed by aspiration after vigorous washing to displace any non-adherent cells. For the final incubation cells were cultured in either RPM&1640 alone, RPMI-1640 with the addition of 1 ,ug/ml endotoxin (addition of 1 pg/ml endotoxin, E. coli, L4516; Sigma) or RPMI-1640 with the addition of endotoxin (as before), plus 3OOpM L-NMMA (Sigma). All samples were incubated for 24 h at 37 “C, 5% C02/95% air before the supematants were removed for subsequent nitrite (NO,-) assay. The concentration of endotoxin used, 1 @ml, stimulates maximal nitrite production. Concentrations of endotoxin in the range O.l-lOO&ml were also studied, but produced no increase in nitrite production (results not shown). Statktical analysis To assess the significance of the differences means a modified Student’s t-test was used.
between
N.C.A. HUNT AND R.D. GOLDIN
148 Results Normal monocytes The monocytes from these 15 volunteers showed a basal nitrite production with a mean value of 7.5 pmol&g protein and a range of 0.0-11 .Opmo&g (Table 1). The addition of endotoxin led to a marked stimulation of nitrite production by the cells (Table l), with a mean’of 55.5 pmol&g protein and a range of 25.7-96.8 pmollpg. T-test analysis showed the difference between these means to be highly significant p << 0.001). When incubated with both LPS and L-NMMA there was a clear inhibition of nitrite production (Table 1) with a mean of 2.5 pmol/pg protein and a range of 0.0-11.5 pmol&g protein. This represents a highly significant decrease in nitrite production (p << 0.001) compared to samples incubated with endotoxin alone.
Monocytesfrom alcoholics
(a) The monocytes obtained from 17 patients without an hepatitic component to their liver disease showed a mean basal nitrite production of 11.6 pmo@g protein (Table 1) and a range of 0.0-19.3 pmol/pg protein. This was not statistically different from the control results from the normal monocyte samples @ > 0.5). In the presence of endotoxin the mean nitrite production rose to 49.6 pmol/pg protein (Table 1) with a range of 29.0-105.0 pmol&g protein. Again this result was not significantly different from that for the unstimulated normal samples (p > 0.5). (b) The monocyte samples isolated from 22 patients with alcoholic hepatitis showed a mean basal nitrite production of 35.3 pmol/pg protein with a range of 18.1-121.3 pmol&g (Table 1). The difference between the mean in this case and that of the normal basal production was compared using t-test analysis and found to be highly significant @ << 0.001). The addition of LPS gave
TABLE 1 Nitrite production by unstimuiated monocytes, monocytes plus LPS, and monocytes plus LPS and L-NMMA isolated from control and alcoholic subjects Controls
Unstimulated Plus LPS Plus LPS and L-NMMA Sample size
Alcoholics Non-hepatic
Hepatic
7.5 (0.9)
11.6 (2.7)
35.3 (4.6)
55.5 (7.3)
49.6 (12.3)
41.0 (3.9)
2.5 (1.2)
2.9 (1.4)
2.0 (0.9)
15
17
Results are expressed as mean (S.E.) in pmo&g protein.
22
a mean of 41.0 pmol/pg protein with a range of 20.5-110.0 pmol&g (Table 1). However, this did not achieve statistical significance in a t-test analysis. Once again the effect of LPS was abrogated by L-NMMA giving a mean basal nitrite production of 2.0 pmol&g protein (Table 1) with a range of 0.0-7.0 pmollpg. This was not significantly different from the value for the normal samples.
Discussioar
To date, most of the work on mononuclear phagocyte involvement in alcoholic liver disease has focused on the macrophage, and in particular, the hepatic sinusoidal macrophage (2). However, many of the effects of ethanol and its metabolites are manifest outside the liver (7). This widespread tissue damage suggests the activity of a more widely acting agenz. Possible candidates include circulating acetaldhyde-albumin complexes and tumour necrosis factor (13). Work on tumour necrosis factor production by monocytes has shown that these cells are activated in the circulation of alcoholic patients and are producing higher than normal basal levels of this mediator. The same work also shows that this production can be stimulated by endotoxin (10). Our present study suggests that nitric oxide produced by circulating peripheral blood monocytes may also be important particularly in patients with alcoholic hepatitis and therefore evidence of active liver disease. Nitric oxide is a pleotropic mediator perhaps best known as the moiety responsible for endothelium-derived relaxing factor activity (14), although this has been called into question recently (15). It has many other roles, for example, as the effector molecule in cytotoxic-activated macrophage activity and also in certain types of neurotransmission (16). In most cases nitric oxide is generated from L-arginine, via a series of intermediates, by a cytosolic, NADPH-dependent activity called, putatively, nitric oxide synthetase, which probably exists as a series of isoenzymes (16). It is at the level of this enzyme that the Larginine analogue L-NMMA acts as a competitive inhibitor of nitric oxide production (17). The nitric oxide thus generated is a highly reactive species and is rapidly degraded. Two stable ion species result from this breakdown, namely nitrite and nitrate ions; normally in a molar ratio of three nitrite to every two nitrate ions (12). It is for this reason that we are able to obtain a reasonable indication of nitric oxide production by measuring nitrite and this is a widely accepted indirect assay for nitric oxide (II). The assay system used for measuring nitrite allows de-
MONOCYTES, NITRIC OXIDE AYD ALCOHOLICS
tection of this ion down to about 1 ,umol/pg protein. As mentioned earlier every effort was made to reduce endotoxin contamination in the system. Despite this it is probably impossible to eliminate all endotoxin. The use of polymyxin B to inactivate endotoxin is advocated by certain workers. Our studies showed that the basal production of nitrite by normal monocytes was very significantly increased by endotoxin showing that they were not maximally stimulated. Furthermore, not all workers are agreed OII the efficacy of this molecule (18) and its use would have added another, unnecessary variable to control for. As mentioned earlier, it has been shown that there appears to be a difference in the TNF production between normal monocytes and those from patients with alcoholic hepatitis. This work suggested that half of the 16 alcoholic’s samples exhibited basal TNF production, compared to only two of the 16 normal (10). In addition, it would seem that the monocytes of alcoholics are more sensitive to the addition of endotoxin, in contrast with our own results (Table 1). The authors of this work do not suggest that any steps were taken to specifically reduce endotoxin contamination in their system. Other factors may also be important when interpreting the results of in vitro studies of monocyte function. It has been found that monocytes isolated from heparinised blood produce more tumour necrosis factor than when EDTA is used as the anticoagulant (19). Based on this observation it has been suggested that heparin directly activates macrophages or that calcium ions are needed for at least some macrophage functions (20). Another point which is made by McClain and Cohen is that ethanol causes a suppression in the production of TNF by both normal and alcoholics’ monocytes (10). This suggests that perhaps TNF production by these cells is not so important in the development of alcoholic liver disease as such patients would be expected to have high blood and tissue ethanol levels. The monocytes from controls and alcoholics without hepatitis behaved similarly. In contrast, the monocytes from patients with alcoholic hepatitis showed a significantly higher basal production of nitrite compared to either of these groups. This shows that these cells are circulating in an activated state, at least with respect to nitric oxide production. There are a number of mechanisms by
References 1 MacSween FWM. Alcoholic liver disease. In: Thomas HC, MacSweeo RNM, eds. Recent Advances in Hepatology 1. Edinburgh; Churchill Livingstone, 1983. 2 Wickramasinghe SN. Role of macrophages in the pathogenesis of
149 which these cells may be activated in vivo. These include endotoxin and y-interferon. There is no evidence to suggest that endototin levels are higher in the systemic circulation of patients with alcoholic hepatitis rather than fatty change alone. However, it has been shown that liver biopsies of patients with alcoholic hepatitis contain increased numbers of CD4 + cells (21) which could act as rich sources of monocyte stimulating cytokmes including y-interferon . A number of effects of nitric oxide at the cellular level are of particular interest in the case of alcoholic liver disease and alcohol-related systemic pathology. First, one of the features of alcoholic liver disease is a decrease in hepatocyte protein synthesis (1). One of the known effects of nitric oxide is its ability to inhibit protein synthesis in a variety of target cells including certain tumour lines and, more importantly, in hepatocytes (22). Another effect is the depression of Krebs’ cycle activity (23), again a recog nised biochemical defect in alcoholic liver disease (1). The fact that the monocyte is a circulating blood cell opens the distinct possibility that nitric oxide from this source may be involved in the extrahepatic effects of long term ethauol abuse as well. Such features as the hyperkinetic circulation and encephalopathy seen in these patients are examples where nitric oxide, horn circulating monocytes, may be involved due to its activity as a vasodilator and its possible activity in certain types of neurotransmission. The evidence presented above is, we feel, an indication that the role of nitric oxide in alcohniic liver disease and related systemic pathology is worthy of further investigation. Of particular importance is the potential role of nitric oxide in mediating liver cell damage and the possibility of therapeutic intervention by manipulating the synthetic pathway. The existence of isoenzymes of nitric oxide synthetase (24) gives the option of highly specific therapy and all the advantages that affords.
Acknowledgement N.C.A. Hunt was supported Council grant.
by a lVIedica1 Research
alcoholinducedtissuedamage. Br Med J 1987; 294: 1137. 3 Le .I, Vilcek J. Tumour necrosis factor and interleukin I: cytokines with multiple overlapping biological activities. Lab Invest 1987; 56: 234. 4 Mills LR, Scheuer PJ. Hepatic sinusoidal macrophages in alcohotic liver disease. J Pathol1985; 147: 127.
150 5 Karakucuk I, Dilly SA, Maxwell JD. Portal tract macrophages are increased in alcoholic liver disease. Histopathology 1989; 14: 245. 6 McCarthy SP, Lewis CE, McGee, JO’D. Effects of ethanol on human monocyte/macrophage lysozyme storage and release: implications for the pathobiology of alcoholic liver disease. J Path01 1990: 160: 2: 165a (Abstr.). 7 Sherlock S. Diseases of the Liver and Biliary System. 8th Edn. London: Blackwell Scientific Publications Limited, 1989. 8 Gordon S, Biology of the macrophage. J Cell Sci (Suppl.) 1986; 4~267. 9 Stuehr DJ, Marletta MA. Mammalian nitrate biosynthesis: mouse macrophages produce nitrite and nitrate in response to F. coli lipopolysaccharide. Proc Nat1 Acad Sci USA 1985; 82: 7738. 10 McClain CJ, Cohen DA. Increased tumour necrosis factor production by monocytes in alcoholic hepatitis. Hepatology 1989; 9: 349. 11 Green LC, Wagner PJ, Glogowski J, Skipper PL, Wishnok JS, Tannenbaum SR. Analysis of nitrate, nitrite and (15N) nitrate in biological fluids. Anal Biochem 1982; 126: 131. 12 Marletta MA, Poksyn SY, Iyengar R, Leaf CD, Wishnok JS. Macrophage oxidation of L-arginine to nitrite and nitrate: nitric oxide is an intermediate. Biochemistry 1988; 27: 8706. 13 Wickramasinghe SM, Gardner S, Barden G. Circulating cytotoxic protein generated after ethanol consumption: identification and mechanism of reaction with cells. Lancet 1987; ii: 122. 14 Palmer RMJ, Ferrige AG, Moncada S. Nitric oxide release accounts for the biological activity of endothelium-derived relaxing factor. Nature 1987; 327: 524. 15 Myers PR, Minor RL, Guerra R, Bates JM, Harrison BG. Vasorelaxant properties of the endothelium derived relaxing factor
N.C.A. HUNT AND R.D. GOLDTN more closely resemble S-nitrosocysteine than nitric oxide. Nature 1990; 345: 161. 16 Moncada S, Palmer RMJ, Higgs AE. Biosynthesis of nitric oxide from L-arginine: a pathway for cell function and communication. Biochem Parmacoll989; 38; 11: 1709. 17 Hibbs JB, Vavrin 2, Taintor RR. t.-Arginine is required for the expression of the activated macrophage effector mechanism causing selective inhibition in target cells. J Immunol 1987; 138: 550. 18 Weinburg JB. Endotoxin contamination and in vitro monocytejmacrophage function: methods of detecting, detoxifying and eliminating endotoxin. In: Adams DO, Edelson PJ, Karen HS, eds. Methods for Studying Mononuclear Phagocytes. New York: Academic Press, 1981. 19 Freeman R, Wheeler J, Robertson H, Paes ML, Laidler J. In-vitro production of TNF-y in blood samples. Lancet 1990; 336: 312. 20 Goldii RD, Hunt NCA. In-vitro production of cytokines in serum. Lancet 1990; 336: 688. 21 Lusheng S, Whiteside TL, Schade PR, Van Thiel D. Lymphocyte subsets studied with monoclonal antibodies in liver tissues of patients with alcoholic liver disease. Alcoholism: Clin Exp Res 1983; 7; 4: 431. 22 Billiar TR, Curran RD, Stuehr DJ, West MA, Bentx BG, Simmons RL. An L-arginine-dependent mechanism mediates Kupffer cells inhibition of hepatocyte protein synthesis. J Exp Med 1989; 169: 1467. 23 Stuehr DJ, Nathan CF. Nitric oxide: a macrophage product responsible for cytostasis and respiratory inhibition in tumour target cells. J Exp Med 1989; 169: 1543. 24 Bredt DS, Hwang PM, Snyder SH. Localization of nitric oxide synthetase indicating a neural role for nitric oxide. Nature 1990; 347: 768.
146 HEPAT 00995
Nitric oxide production by monocytes in alcoholic liver disease
N.C.A. Hunt and R.D. Goldin Department of Histopathology, St. Mary’s Hospital Medical School, London, United Kingdom
(Received 7 January 1990)
Nitric oxide, initially described as an endothelial-derived relaxing factor, has recently been recognised as a mediator of macrophage function. We have studied the production of nitric oxide by peripheral blood monocytes from both normal volunteers and alcoholics. This was measured indirectly by assessing nitrite formation. Normal monocytes were found to produce a basal level of nitrite, which could be stimulated more than 6-fold using endotoxin. This effect was abrogated by the addition of nitric oxide synthesis inhibitor, L-n-monomethyl-arginine. A striking difference was observed in the monocytes obtained from alcoholics with and without evidence of alcoholic hepatitis. Whereas the latter behaved in a similar manner to the controls, the former had markedly increased basal levels. In the hepatitis group there was also substantial inhibition of production by t_-n-monomethyl-arginine. We believe that these results indicate that nitric oxide derived from monocytes may play a role in the pathogenesis of alcoholic liver disease, especially alcoholic hepatitis.
Much work has been carried out on the role of various cells, including hepatocytes, neutrophils and lymphocytes, in the pathogenesis of alcoholic liver disease (ALD) (1). There is increasing evidence to suggest that cells of the mononuclear phagocyte family are involved, as effector cells, in this process (2). Of particular interest are the hepatic sinusoidal macrophages (Kupffer cells) and, more recently, peripheral blood monocytes; both, when stimulated, are known to produce a variety of mediators including interleukin-1, and tumour necrosis factor (cachectin) (3). Early reports suggested a decrease in the number of hepatic sinusoidal macrophages, as assessed by measuring the number of lysozyme-positive cells, in patients with increasingly severe ALD (4). Thus, the decrease was greatest in patients with cirrhosis and least in those with steatosis alone. Later work supported this decrease in lysozyme-positive cells. However, using a number of other antibodies it was also shown that certain classes of hepatic sinusoidal macrophages were in fact more abundant and overall there is an increase in num-
bers of hepatic sinusoidal macrophages, even in the earliest stages of ALD (5). It has been suggested that functional impairment of hepatic sinusoidal macrophages may be of greater importance in ALD than the decrease in cell number. In support of this it has been shown that alcohol-exposed macrophages produce decreased amounts of lysozymes (6). Decreased macrophage function in conjunction with portosystemic shunting, in certain groups of ALD patients, is believed to contribute to the endotoxaemia seen in these people (6,7). Endotoxin is itself a known activator of monocytes and macrophages (8). The fact that endotoxin stimulates the production of nitric oxide by macrophages, including hepatic sinusoidal macrophages has been well documented (9). It has been shown that endotoxin can stimulate monocytes to produce increased amounts of tumour necrosis factor and, moreover, the monocytes of alcoholics show an increased basal production of this mediator (10). The object of the present study was to investigate the production of nitric oxide
Correspondence: R.D. Goldin, Department of Histopathology, St. Mary’s Hospital Medical School, London W2, U.K.
MONOCYTJZS, NITRIC OXIDE
AND ALCOHOLICS
147
by monocytes in patients with ALD and to see to what extent they, and normal monocytes, responded to endotoxin.
NMMA: Sigma). All samples were incubated for 24 h at 37 “C, 5% CO,/95% air, before the supematants were removed for subsequent nitrite assay
Materiais and Methods
Nitrite assay Nitric oxide production was measured indirectly using a quantitative, calorimetric assay based on the Griess reaction, a sensitive assay for nitrite ions (11). Nitrite and nitrate ions are the stable, anionic, products formed in a molar ratio of 3:2 from nitric oxide as a result of its decompos8ttion (12). In the present study triplicate 100~,ulaliquots of supernatants were added to an equal volume of freshly prepared Griess reagent. In this system nitrite ions react with 1% sulphanilamide in 5% ortho-phosphoric acid/O. 1% N-l-naphthylethylenediamine dihydrochloride to yield an azo-chromophore, the absorbance of which was measured at 550 nm using a Titretek 96well plate reader. All samples were assayed against a blank comprising RPMI-1640 medium incubated for 24 h in the same plates as the samples but in the absence of cells.
Subjects .4 total of 54 subjects were recruited for the study. The control group consisted of 15 normal volunteers (mean age = 37.7 years, age range = 19-61, eight males and seven females). The alcoholic hepatitic group consisted of 22 patients, all with histories of excessive alcohol intake (mean age = 42.8 years, age range = 18-65,14 males and eight females). All were out-patients with histological diagnosis of alcoholic hepatitis with varying degrees of fibrosis. In addition, a third group of 17 alcohol abusers with non-hepatic liver disease were studied (mean age = 39.0 years, age range = 23-57, ten males and seven females). These patients ail had liver biopsies showing fatty change with varying degrees tif fibrosis but no evidence of alcoholic hepatitis. In each case serological markers of viral hepatitis, including hepatitis C virus, were negative. All biopsies were reviewed prior to inclusion in the study in order to confirm the histological diagnosis (RDG.) Monocyte extraction and culture Monocytes were extracted from peripheral blood on a density gradient materiai (IEstopaque, Sigma). This and all subsequent procedures required specific precautions in order to minimise contamination by LPS. Such measures included baking of all laboratory glassware, use of pyrogen-free disposables and endotoxin-free culture media. Briefly, peripheral blood mononuclear cells were removed from the density gradient, washed twice in endotoxin-free Hanks’ Balanced Salts Solution, and finally resuspended in RPMI-1640 (RPM]-1640, endotoxin-free. Phenol red-free; Sigma) supplemented with 10% decomplemented fetal calf serum (low endotoxin; Flow, U.K.) and 2 mmol t_-glutamine (Sigma). Cells were then plated out at a concentration of 2-4-l@ per ml in individual 16mm wells of tissue cuiture plates (24 well; Nunclon). A Trypan blue exclusion analysis was performed at this stage and showed a mean viability of 95% with a range of 92-97%. Monocyte enrichment was achieved by allowing the cells to adhere to the plates for 2 h at 37 “C, after which time the medium was removed by aspiration after vigorous washing to remove any non-adherent cells. For the final incubation cells were cultured in either RPMI-1640 (supplemented as above) alone, medium with the addition of 1 &ml endotoxin (Escherichia co/i lipopolysaccharide, L4516; Sigma), or medium with 1 ,@ml endotoxin plus 300 ,uM t-n-monomethyl-arginine (L-
Protein assay Briefly, PBMCs were removed from the density gradient, washed twice in endotoxin-free Hanks’ Balanced Salts Solution, and finally resuspended in RMPI-1640 medium (RMPI-1640, low endotoxin, Phenol red-free; Sigma Chemicals) supplemented with 10% decomplemented fetal bovine serttm (low endotoxin; Flow, U.K.). Cells iyere then plated out at a concentration of plates (Nunclan, 24 well). A Trypan blue exclusion analysis was performed at this stage and showed a mean viability of 95% with a range of 92-97%. Monocyte enrichment was achieved by allowing the cells to adhere to the plates for 2 h at 37 “C, after which time the medium was removed by aspiration after vigorous washing to displace any non-adherent cells. For the final incubation cells were cultured in either RPM&1640 alone, RPMI-1640 with the addition of 1 ,ug/ml endotoxin (addition of 1 pg/ml endotoxin, E. coli, L4516; Sigma) or RPMI-1640 with the addition of endotoxin (as before), plus 3OOpM L-NMMA (Sigma). All samples were incubated for 24 h at 37 “C, 5% C02/95% air before the supematants were removed for subsequent nitrite (NO,-) assay. The concentration of endotoxin used, 1 @ml, stimulates maximal nitrite production. Concentrations of endotoxin in the range O.l-lOO&ml were also studied, but produced no increase in nitrite production (results not shown). Statktical analysis To assess the significance of the differences means a modified Student’s t-test was used.
between
N.C.A. HUNT AND R.D. GOLDIN
148 Results Normal monocytes The monocytes from these 15 volunteers showed a basal nitrite production with a mean value of 7.5 pmol&g protein and a range of 0.0-11 .Opmo&g (Table 1). The addition of endotoxin led to a marked stimulation of nitrite production by the cells (Table l), with a mean’of 55.5 pmol&g protein and a range of 25.7-96.8 pmollpg. T-test analysis showed the difference between these means to be highly significant p << 0.001). When incubated with both LPS and L-NMMA there was a clear inhibition of nitrite production (Table 1) with a mean of 2.5 pmol/pg protein and a range of 0.0-11.5 pmol&g protein. This represents a highly significant decrease in nitrite production (p << 0.001) compared to samples incubated with endotoxin alone.
Monocytesfrom alcoholics
(a) The monocytes obtained from 17 patients without an hepatitic component to their liver disease showed a mean basal nitrite production of 11.6 pmo@g protein (Table 1) and a range of 0.0-19.3 pmol/pg protein. This was not statistically different from the control results from the normal monocyte samples @ > 0.5). In the presence of endotoxin the mean nitrite production rose to 49.6 pmol/pg protein (Table 1) with a range of 29.0-105.0 pmol&g protein. Again this result was not significantly different from that for the unstimulated normal samples (p > 0.5). (b) The monocyte samples isolated from 22 patients with alcoholic hepatitis showed a mean basal nitrite production of 35.3 pmol/pg protein with a range of 18.1-121.3 pmol&g (Table 1). The difference between the mean in this case and that of the normal basal production was compared using t-test analysis and found to be highly significant @ << 0.001). The addition of LPS gave
TABLE 1 Nitrite production by unstimuiated monocytes, monocytes plus LPS, and monocytes plus LPS and L-NMMA isolated from control and alcoholic subjects Controls
Unstimulated Plus LPS Plus LPS and L-NMMA Sample size
Alcoholics Non-hepatic
Hepatic
7.5 (0.9)
11.6 (2.7)
35.3 (4.6)
55.5 (7.3)
49.6 (12.3)
41.0 (3.9)
2.5 (1.2)
2.9 (1.4)
2.0 (0.9)
15
17
Results are expressed as mean (S.E.) in pmo&g protein.
22
a mean of 41.0 pmol/pg protein with a range of 20.5-110.0 pmol&g (Table 1). However, this did not achieve statistical significance in a t-test analysis. Once again the effect of LPS was abrogated by L-NMMA giving a mean basal nitrite production of 2.0 pmol&g protein (Table 1) with a range of 0.0-7.0 pmollpg. This was not significantly different from the value for the normal samples.
Discussioar
To date, most of the work on mononuclear phagocyte involvement in alcoholic liver disease has focused on the macrophage, and in particular, the hepatic sinusoidal macrophage (2). However, many of the effects of ethanol and its metabolites are manifest outside the liver (7). This widespread tissue damage suggests the activity of a more widely acting agenz. Possible candidates include circulating acetaldhyde-albumin complexes and tumour necrosis factor (13). Work on tumour necrosis factor production by monocytes has shown that these cells are activated in the circulation of alcoholic patients and are producing higher than normal basal levels of this mediator. The same work also shows that this production can be stimulated by endotoxin (10). Our present study suggests that nitric oxide produced by circulating peripheral blood monocytes may also be important particularly in patients with alcoholic hepatitis and therefore evidence of active liver disease. Nitric oxide is a pleotropic mediator perhaps best known as the moiety responsible for endothelium-derived relaxing factor activity (14), although this has been called into question recently (15). It has many other roles, for example, as the effector molecule in cytotoxic-activated macrophage activity and also in certain types of neurotransmission (16). In most cases nitric oxide is generated from L-arginine, via a series of intermediates, by a cytosolic, NADPH-dependent activity called, putatively, nitric oxide synthetase, which probably exists as a series of isoenzymes (16). It is at the level of this enzyme that the Larginine analogue L-NMMA acts as a competitive inhibitor of nitric oxide production (17). The nitric oxide thus generated is a highly reactive species and is rapidly degraded. Two stable ion species result from this breakdown, namely nitrite and nitrate ions; normally in a molar ratio of three nitrite to every two nitrate ions (12). It is for this reason that we are able to obtain a reasonable indication of nitric oxide production by measuring nitrite and this is a widely accepted indirect assay for nitric oxide (II). The assay system used for measuring nitrite allows de-
MONOCYTES, NITRIC OXIDE AYD ALCOHOLICS
tection of this ion down to about 1 ,umol/pg protein. As mentioned earlier every effort was made to reduce endotoxin contamination in the system. Despite this it is probably impossible to eliminate all endotoxin. The use of polymyxin B to inactivate endotoxin is advocated by certain workers. Our studies showed that the basal production of nitrite by normal monocytes was very significantly increased by endotoxin showing that they were not maximally stimulated. Furthermore, not all workers are agreed OII the efficacy of this molecule (18) and its use would have added another, unnecessary variable to control for. As mentioned earlier, it has been shown that there appears to be a difference in the TNF production between normal monocytes and those from patients with alcoholic hepatitis. This work suggested that half of the 16 alcoholic’s samples exhibited basal TNF production, compared to only two of the 16 normal (10). In addition, it would seem that the monocytes of alcoholics are more sensitive to the addition of endotoxin, in contrast with our own results (Table 1). The authors of this work do not suggest that any steps were taken to specifically reduce endotoxin contamination in their system. Other factors may also be important when interpreting the results of in vitro studies of monocyte function. It has been found that monocytes isolated from heparinised blood produce more tumour necrosis factor than when EDTA is used as the anticoagulant (19). Based on this observation it has been suggested that heparin directly activates macrophages or that calcium ions are needed for at least some macrophage functions (20). Another point which is made by McClain and Cohen is that ethanol causes a suppression in the production of TNF by both normal and alcoholics’ monocytes (10). This suggests that perhaps TNF production by these cells is not so important in the development of alcoholic liver disease as such patients would be expected to have high blood and tissue ethanol levels. The monocytes from controls and alcoholics without hepatitis behaved similarly. In contrast, the monocytes from patients with alcoholic hepatitis showed a significantly higher basal production of nitrite compared to either of these groups. This shows that these cells are circulating in an activated state, at least with respect to nitric oxide production. There are a number of mechanisms by
References 1 MacSween FWM. Alcoholic liver disease. In: Thomas HC, MacSweeo RNM, eds. Recent Advances in Hepatology 1. Edinburgh; Churchill Livingstone, 1983. 2 Wickramasinghe SN. Role of macrophages in the pathogenesis of
149 which these cells may be activated in vivo. These include endotoxin and y-interferon. There is no evidence to suggest that endototin levels are higher in the systemic circulation of patients with alcoholic hepatitis rather than fatty change alone. However, it has been shown that liver biopsies of patients with alcoholic hepatitis contain increased numbers of CD4 + cells (21) which could act as rich sources of monocyte stimulating cytokmes including y-interferon . A number of effects of nitric oxide at the cellular level are of particular interest in the case of alcoholic liver disease and alcohol-related systemic pathology. First, one of the features of alcoholic liver disease is a decrease in hepatocyte protein synthesis (1). One of the known effects of nitric oxide is its ability to inhibit protein synthesis in a variety of target cells including certain tumour lines and, more importantly, in hepatocytes (22). Another effect is the depression of Krebs’ cycle activity (23), again a recog nised biochemical defect in alcoholic liver disease (1). The fact that the monocyte is a circulating blood cell opens the distinct possibility that nitric oxide from this source may be involved in the extrahepatic effects of long term ethauol abuse as well. Such features as the hyperkinetic circulation and encephalopathy seen in these patients are examples where nitric oxide, horn circulating monocytes, may be involved due to its activity as a vasodilator and its possible activity in certain types of neurotransmission. The evidence presented above is, we feel, an indication that the role of nitric oxide in alcohniic liver disease and related systemic pathology is worthy of further investigation. Of particular importance is the potential role of nitric oxide in mediating liver cell damage and the possibility of therapeutic intervention by manipulating the synthetic pathway. The existence of isoenzymes of nitric oxide synthetase (24) gives the option of highly specific therapy and all the advantages that affords.
Acknowledgement N.C.A. Hunt was supported Council grant.
by a lVIedica1 Research
alcoholinducedtissuedamage. Br Med J 1987; 294: 1137. 3 Le .I, Vilcek J. Tumour necrosis factor and interleukin I: cytokines with multiple overlapping biological activities. Lab Invest 1987; 56: 234. 4 Mills LR, Scheuer PJ. Hepatic sinusoidal macrophages in alcohotic liver disease. J Pathol1985; 147: 127.
150 5 Karakucuk I, Dilly SA, Maxwell JD. Portal tract macrophages are increased in alcoholic liver disease. Histopathology 1989; 14: 245. 6 McCarthy SP, Lewis CE, McGee, JO’D. Effects of ethanol on human monocyte/macrophage lysozyme storage and release: implications for the pathobiology of alcoholic liver disease. J Path01 1990: 160: 2: 165a (Abstr.). 7 Sherlock S. Diseases of the Liver and Biliary System. 8th Edn. London: Blackwell Scientific Publications Limited, 1989. 8 Gordon S, Biology of the macrophage. J Cell Sci (Suppl.) 1986; 4~267. 9 Stuehr DJ, Marletta MA. Mammalian nitrate biosynthesis: mouse macrophages produce nitrite and nitrate in response to F. coli lipopolysaccharide. Proc Nat1 Acad Sci USA 1985; 82: 7738. 10 McClain CJ, Cohen DA. Increased tumour necrosis factor production by monocytes in alcoholic hepatitis. Hepatology 1989; 9: 349. 11 Green LC, Wagner PJ, Glogowski J, Skipper PL, Wishnok JS, Tannenbaum SR. Analysis of nitrate, nitrite and (15N) nitrate in biological fluids. Anal Biochem 1982; 126: 131. 12 Marletta MA, Poksyn SY, Iyengar R, Leaf CD, Wishnok JS. Macrophage oxidation of L-arginine to nitrite and nitrate: nitric oxide is an intermediate. Biochemistry 1988; 27: 8706. 13 Wickramasinghe SM, Gardner S, Barden G. Circulating cytotoxic protein generated after ethanol consumption: identification and mechanism of reaction with cells. Lancet 1987; ii: 122. 14 Palmer RMJ, Ferrige AG, Moncada S. Nitric oxide release accounts for the biological activity of endothelium-derived relaxing factor. Nature 1987; 327: 524. 15 Myers PR, Minor RL, Guerra R, Bates JM, Harrison BG. Vasorelaxant properties of the endothelium derived relaxing factor
N.C.A. HUNT AND R.D. GOLDTN more closely resemble S-nitrosocysteine than nitric oxide. Nature 1990; 345: 161. 16 Moncada S, Palmer RMJ, Higgs AE. Biosynthesis of nitric oxide from L-arginine: a pathway for cell function and communication. Biochem Parmacoll989; 38; 11: 1709. 17 Hibbs JB, Vavrin 2, Taintor RR. t.-Arginine is required for the expression of the activated macrophage effector mechanism causing selective inhibition in target cells. J Immunol 1987; 138: 550. 18 Weinburg JB. Endotoxin contamination and in vitro monocytejmacrophage function: methods of detecting, detoxifying and eliminating endotoxin. In: Adams DO, Edelson PJ, Karen HS, eds. Methods for Studying Mononuclear Phagocytes. New York: Academic Press, 1981. 19 Freeman R, Wheeler J, Robertson H, Paes ML, Laidler J. In-vitro production of TNF-y in blood samples. Lancet 1990; 336: 312. 20 Goldii RD, Hunt NCA. In-vitro production of cytokines in serum. Lancet 1990; 336: 688. 21 Lusheng S, Whiteside TL, Schade PR, Van Thiel D. Lymphocyte subsets studied with monoclonal antibodies in liver tissues of patients with alcoholic liver disease. Alcoholism: Clin Exp Res 1983; 7; 4: 431. 22 Billiar TR, Curran RD, Stuehr DJ, West MA, Bentx BG, Simmons RL. An L-arginine-dependent mechanism mediates Kupffer cells inhibition of hepatocyte protein synthesis. J Exp Med 1989; 169: 1467. 23 Stuehr DJ, Nathan CF. Nitric oxide: a macrophage product responsible for cytostasis and respiratory inhibition in tumour target cells. J Exp Med 1989; 169: 1543. 24 Bredt DS, Hwang PM, Snyder SH. Localization of nitric oxide synthetase indicating a neural role for nitric oxide. Nature 1990; 347: 768.