CHRONIC ALCOHOL CONSUMPTION INCREASES SERUM LEVELS OF CIRCULATING ENDOTHELIAL CELL/LEUCOCYTE ADHESION MOLECULES E-SELECTIN AND ICAM-1

E. Sacanella, R. Estruch*, E. Badía, J. Fernández-Sola, J. M. Nicolás and A. Urbano-Márquez

Department of Internal Medicine, Hospital Clínic, Institut d'Investigació Biomèdica Agustí Pi Sunyer (IDIBAPS), University of Barcelona, Barcelona, Spain

Received 29 January 1999; in revised form 12 April 1999; accepted 24 April 1999


    ABSTRACT
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
A group of 30 chronic alcoholics without alcohol-related diseases and 30 controls (teetotallers) were selected to measure serum levels of endothelial adhesion molecules (AMs) (ICAM-1, VCAM-1, and E-selectin). ICAM-1 and E-selectin serum levels were higher in alcoholics, whereas VCAM-1 serum levels were similar in both groups. There was a significant correlation between daily alcohol intake and serum level of ICAM-1 (r = 0.49, P = 0.003) and E-selectin (r = 0.41, P = 0.02). A significant positive correlation between E-selectin and total lifetime dose of ethanol was also observed (r = 0.52, P = 0.003). These changes in serum levels of endothelial AMs of chronic alcoholics may reflect endothelial and/or immune activation, and could interfere with the reactions between immune cells and the endothelium.


    INTRODUCTION
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
The endothelium is involved in several homeostatic mechanisms, such as the maintenance of a non-thrombotic surface, the metabolism of lipoproteins and in immune response (Ross, 1993Go). Several conditions may induce activation of endothelial cells, which leads to the appearance of adhesion molecules (AMs) on the cell surface. AMs, such as E-selectin, intercellular adhesion molecule-1 (ICAM-1), and vascular cell adhesion molecule-1 (VCAM-1), mediate the attachment of certain leucocytes to the endothelial surface and may be important in controlling the extravasation of leucocytes from the circulation to sites of inflammation (Bevilacqua, 1993Go). Endothelial AMs may also be detected in a soluble form in the serum, usually reflecting higher expression on the cell membrane, and are increased in inflammatory processes (septic shock, vasculitis, atherosclerosis) (Gearing and Newman, 1993Go).

Chronic alcohol intake has been shown to be associated with modulation of immune defence mechanisms (Watson et al., 1994Go). In this sense, peripheral blood mononuclear cells (PBMCs) of chronic alcoholics without liver disease are basally activated. This state is manifested by a higher expression of activation antigens, CD25, CD69, and HLA-DR (Sacanella et al., 1998Go) and the AMs CD29, VLA proteins, CD11b, L-selectin, CD31, and CD57 (Cook et al., 1995Go; Sacanella et al., 1999Go) on PBMCs of alcoholic patients, compared to PBMCs of controls. However, whether the activation of immune cells in alcoholic patients without ethanol-related diseases is accompanied by endothelial cell activation manifested by raised serum levels of vascular AMs is unknown.

Thus, a study was carried out to analyse whether chronic alcohol consumption is related to the serum levels of E-selectin, ICAM-1 and VCAM-1. A group of 30 well-nourished chronic alcoholics without liver disease were compared with 30 subjects who did not drink any amount of ethanol (teetotallers).


    PATIENTS AND METHODS
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Patient and control selection
Over a 12-month period, a total of 420 patients with chronic alcoholism (DSM-IIIR) (American Psychiatric Association, 1987Go) were seen in the Alcoholism Unit of our institution. This unit deals with ambulatory patients who seek assistance in terminating their dependence upon alcohol, but who do not suffer from other overt diseases. On the Monday of each week, the first male patient to register with a daily alcohol intake above 100 g in the previous 2 years up to the day prior to admission was selected for study. Initially, 40 patients were included. Among these patients, two had cirrhosis, four alcoholic hepatitis and fatty liver, and four had antibodies against hepatitis B or C virus; all were excluded from the study, leaving a total of 30 male alcoholics.

Control subjects were obtained from the medical and auxiliary personnel of our institution who reported not drinking any amount of ethanol (teetotallers). These subjects were matched for gender (male) and age (±3 years) with the alcoholic patients and were studied in the same way as the alcoholics. All patients and controls were Caucasian males of Spanish descent. No patients or controls objected to being in the study, and all gave informed consent for the various procedures. The study protocol was approved by our Institutional Review Board.

Exclusion criteria
Alcoholic and control subjects were excluded from the study if they met one of the following criteria: human immunodeficiency virus infection or other causes of immunodeficiency, hepatitis B or C virus infection, cirrhosis, acute alcoholic hepatitis (AAH), malnutrition, neoplastic diseases, acute infectious diseases, consumption of drugs with effect on immunity (steroids, immunosuppressants), or chronic diseases such as diabetes mellitus.

Clinical, laboratory and nutritional studies
A detailed clinical history of ethanol intake, smoking, and dietary habits was obtained by two physicians (E.S. and J.F.-S.) using a structured questionnaire, and the data were confirmed by consultation with family members (Sobell et al., 1979Go).

Within 24 h of admission to the hospital, blood samples were obtained to measure soluble endothelial AM and for laboratory studies. These included red-cell and leucocyte count, liver-function tests, total protein, albumin, prealbumin, retinol-binding protein, and serum folate, vitamin B12 and transketolase activity. The levels of ethanol in blood and urine, as well as serum antibodies against human immunodeficiency virus (HIV) and hepatitis B and C viruses were determined in all patients.

Hepatic ultrasonography was performed in all patients and an echo-guided percutaneous needle liver biopsy was also performed in those patients with hepatomegaly, echographic evidence of chronic liver disease and/or aminotransferases greater than 120 U/l after 2 months of complete ethanol abstinence.

Overall nutrition was assessed in terms of the proportion of actual to ideal weight. The lean body mass was calculated from the circumference of the upper non-dominant arm and the thickness of the tricipital skin fold. Nutritional protein status was estimated according to the values obtained for lymphocyte count, haemoglobin, total protein, albumin, prealbumin, and retinol-binding protein. Caloric malnutrition was considered if the body weight was less than 90% of ideal weight and/or if lean body mass was more than 10% below the normal value. Protein malnutrition was considered if four of the six parameters mentioned above were below normal values (Nicolás et al., 1993Go).

Quantification of soluble adhesion molecules
Serum levels of ICAM-1, VCAM-1, and E-selectin were determined by a sandwich ELISA technique. Commercially available kits from British Bio-technology Products, Abingdon, Oxon, UK for E-selectin and kits from Bender MedSystems, Vienna, Austria for ICAM-1 and VCAM-1 were used. The procedures were performed according to the manufacturers' instructions. Briefly, a horseradish peroxidase-conjugated monoclonal antibody against ICAM-1, VCAM-1, and E-selectin was added to microtitre plates coated with a murine monoclonal IgG antibody recognizing a different epitope of the corresponding molecule. After incubation with samples or standards in appropriate dilution, the colour reaction was developed with tetramethylenbenzidine, and the plates were read on an automated multiscanner at 450 nm. All measurements were performed in duplicate.

For ICAM-1 and VCAM-1, the intra-assay coefficients of variation were 4.1 and 3.1% respectively (Bender MedSystems) and for E-selectin 4.8% (R&D Systems). The inter-assay coefficient of variation was 7.4% for E-selectin, 7.66% for ICAM-1, and 5.2% for VCAM-1.

Statistical analysis
The data were analysed using the SPSS-PC 4.0 statistical software (SPSS, Chicago, IL, USA). Differences between groups were analysed using the two-tailed t-test, the {chi}2-test and Fisher's exact test. Non-parametric tests, such as the Mann– Whitney test, were performed when necessary. Correlation studies were obtained by Pearson's correlation coefficient and regression analysis. All variables are expressed as means ± SD. Statistical significance was considered when the P value was below 0.05.


    RESULTS
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Clinical data, nutritional status, and lipid analysis
The alcoholic patients and the controls had a mean age (± SD) of 40.1 ± 7.3 years and 37.3 ± 6.1 years respectively. The daily ethanol intake of alcoholics was 196 ± 54 g/day (range 110–383 g), for a period of 18.5 ± 11.4 years. The mean total lifetime dose of ethanol was 19.4 ± 9.6 kg of ethanol/kg body wt. The pattern of drinking was continuous excessive ethanol intake as a part of everyday life. No differences were found in smoking habits and nutritional status between both groups (Table 1Go).


View this table:
[in this window]
[in a new window]
 
Table 1. Clinical and nutritional data and liver-function tests of 30 chronic alcoholics without liver disease and their matched controls
 
None of the alcoholics or the controls had clinical or laboratory evidence of caloric or protein malnutrition, although the mean circumference of the upper non-dominant arm and the thickness of the tricipital skin fold were lower in patients than in controls (P < 0.02, for both). Blood and urine levels of ethanol determined at the same time as the other laboratory analyses were negative in all alcoholics. In addition, serum levels of triglycerides (132 ± 64 vs 122 ± 47 mg/dl, P = 0.4) and total cholesterol (225 ± 12 vs 209 ± 38 mg/dl, P = 0.1) were similar in alcoholics and controls.

Alcoholic liver disease (ALD)
Serum {gamma}-glutamyl transpeptidase and alanine and aspartate aminotransferases were significantly higher in the alcoholics, compared to controls (P < 0.02) (Table 1Go). Hepatic ultrasonography was performed in all alcoholic cases and disclosed mild hepatomegaly in five patients, and 11 other patients showed liver steatosis. In the remaining 20 patients, liver echography was entirely normal. Liver biopsy was performed in patients with hepatomegaly (n = 5) and in those with serum aminotransferases remaining greater than 120 U/l after 2 months of complete ethanol abstinence (n = 2). The liver biopsy performed in these seven patients showed fatty liver and alcoholic hepatitis in four, cirrhosis in two, but in one the biopsy was normal. The former six patients were excluded from the analysis. All the alcoholic patients were followed up over an average period of 18 months in the out-patient clinic and none exhibited clinical or laboratory evidence of liver disease.

Circulating adhesion molecules in alcoholic patients and controls
The mean serum levels of ICAM-1 (408.4 ± 134.2 vs 292.5 ± 92.4 ng/ml) and E-selectin (118.6 ± 74.7 vs 54.3 ± 22.5 ng/ml) were significantly higher in alcoholics than in controls (P < 0.001), whereas the serum concentration of VCAM-1 was similar in both groups (Fig. 1Go and Table 2Go). No differences were detected in serum levels of AMs among alcoholic patients with normal or abnormal liver function tests. In addition, alcoholic patients with normal liver function tests also had significantly higher serum levels of E-selectin and ICAM-1 than controls (P < 0.01, for both). Finally, serum levels of ICAM-1, E-selectin, and VCAM-1 among smoker and non-smoker alcoholics and among alcoholics with normal and abnormal lipid analysis were similar (Table 2Go). Similar results were obtained in the control group.



View larger version (11K):
[in this window]
[in a new window]
 
Fig. 1. Scatter plots indicating individual and mean (horizontal line) values of ICAM-1 and E-selectin serum levels in 30 chronic alcoholics and 30 controls.

 

View this table:
[in this window]
[in a new window]
 
Table 2. Serum levels of ICAM-1, E-selectin, and VCAM-1 in controls and alcoholics
 
Correlation between alcohol intake and serum concentrations of ICAM-1 and E-selectin
Alcoholics who reported a daily ethanol intake greater than 200 g/day showed a significantly higher serum level of ICAM-1 than their counterparts (489.2 ± 108.7 vs 350.7 ± 121.9 ng/ml, P = 0.001). In addition, those patients with a total life-time dose of ethanol greater than 19 kg of ethanol/kg of body wt exhibited a significantly higher serum level of ICAM-1 (468.6 ± 127.3 vs 332 ± 122.3 ng/ ml, P < 0.01) and E-selectin (137.5 ± 84.3 vs 84.9 ± 34.9 ng/ml, P = 0.033) than their counterparts. A significant positive correlation between ICAM-1 with daily ethanol intake was also observed (r = 0.49, P = 0.003), whereas the E-selectin serum level correlated with daily ethanol intake (r = 0.41, P = 0.02) and with the total life-time dose of ethanol (r = 0.52, P < 0.01) (Fig. 2Go). No correlation was found between VCAM-1 and ethanol intake.



View larger version (14K):
[in this window]
[in a new window]
 
Fig. 2. Correlation between total lifetime dose of ethanol and serum level of E-selectin in a group of 30 chronic alcoholics without liver disease.

 

    DISCUSSION
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Raised serum levels of vascular cell AMs (E-selectin, ICAM-1, and VCAM-1) have been observed mainly in patients with infectious, inflammatory, and neoplastic diseases (Gearing and Newman, 1993Go). In addition, these molecules are considered as markers of atherosclerosis (Hwang et al., 1997Go) and specifically of ischaemic heart disease (Ghaisas et al., 1997Go). In these clinical settings, measuring serum levels of AMs may be useful to detect the subclinical activity and prognosis of these diseases.

In alcoholic patients, analysis of the serum levels of endothelial AMs has been focused on subjects with severe ALD. Patients with AAH or cirrhosis exhibited significantly higher serum levels of these three AMs, compared to non-drinking controls (Adams et al., 1994Go). Serum levels were higher in patients with AAH and in those actively drinking than in those with stable cirrhosis or abstainers (Shimada et al., 1993Go; Adams et al., 1994Go). Moreover, correlations between serum levels of vascular AMs and expression of these molecules on the hepatic endothelium (Adams et al., 1994Go), degree of histologic injury (Ishii et al., 1994Go; Douds et al., 1997Go) or even clinical evolution of AAH (Shimada et al., 1993Go) have been reported. Thus, some authors have suggested that the type of the inflammatory infiltrate (mononuclear or polynuclear cells) seen in the liver of ALD may depend, at least in part, on the nature of AM expressed on the hepatic endothelium (Adams, 1994Go; Adams et al., 1994Go).

Studies in alcoholic patients without liver disease are scarce with the samples being too small (n < 9) and the results controversial. One of these studies did not include a control group (Adams et al., 1994Go) and the other two only analysed ICAM-1 (Shimada et al., 1993Go; Douds et al., 1997Go). Moreover, the results of these studies were negative, since no significant differences were observed in endothelial AM serum levels. In the current study, significantly higher serum levels of E-selectin and ICAM-1 were observed in asymptomatic chronic alcoholics, suggesting endothelial and/or immune activation. Other confusing factors, such as smoking habit and hyperlipaemia, were also controlled. Moreover, the dose-dependent relationship between alcohol intake and serum levels of these AMs suggests a direct or indirect effect of ethanol on the endothelium. Raised levels of E-selectin always reflect endothelial activation due to its unique origin, whereas increased levels of ICAM-1 could reflect endothelial or immune activation (T-lymphocyte activation) (Gearing and Newman, 1993Go). In fact, the existence of activated T-cells in chronic alcoholics has already been described (Sacanella et al., 1998Go). Increased E-selectin and ICAM-1 without changes in VCAM-1 serum levels has already been observed in patients with asthma (Montefort et al., 1994Go) or ischaemic heart disease (Hwang et al., 1997Go).

How alcohol exerts its effect on immune cells is unknown (Watson et al., 1994Go), as is the mechanism by which ethanol may enhance the expression of AM on the endothelium. Proinflammatory cytokines (IL-1, TNF-{alpha}) and endotoxin are potent stimulators of AM expression on endothelium (Bevilacqua, 1993Go). In ALD, the mechanism responsible for the different pattern of AM expression on the hepatic endothelium could be a change in local cytokine release (Adams, 1994Go; Adams et al., 1994Go). A similar mechanism could act in alcoholic patients without ALD. In this sense, it is well established that alcoholic patients without ALD have raised serum levels of TNF-{alpha} and endotoxin compared to abstainers (Khoruts et al., 1991Go).

Finally, the clinical significance of higher endothelial AM serum levels in chronic alcoholics remains to be elucidated. It may be useful in detecting patients developing ALD (quiescent phase) and also as an early marker of atherosclerosis. In the latter case, several studies have suggested that alcohol abusers have a significantly higher risk of undergoing ischaemic stroke or of death from cardiovascular disease than abstainers, whereas moderate consumption of ethanol may have a beneficial effect on the cardiovascular system (Constant, 1997Go; Sacco et al., 1999Go). It would be interesting to determine the serum levels of endothelial AMs in subjects who consume one to three drinks a day in order to establish whether they show significantly lower levels of these markers.


    ACKNOWLEDGEMENTS
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
This work was supported in part by grant FIS 94/114.


    FOOTNOTES
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
* Author to whom correspondence should be addressed at: Department of Internal Medicine, Hospital Clinic, Villarroel 170, 08036 Barcelona, Spain. Back


    REFERENCES
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Adams, D. H. (1994) Leucocyte adhesion molecules and alcoholic liver disease. Alcohol and Alcoholism 29, 249–260.[Abstract]

Adams, D. H., Burra, P., Hubscher, S. G., Elias, E. and Newman, W. (1994) Endothelial activation and circulating vascular adhesion molecules in alcoholic liver disease. Hepatology 19, 588–594.[ISI][Medline]

American Psychiatric Association (1987) Diagnostic and Statistical Manual of Mental Disorders, 3rd edn, revised. American Psychiatric Association Press, Washington, DC.

Bevilacqua, M. P. (1993) Endothelial-leukocyte adhesion molecules. Annual Reviews in Immunology 11, 767– 804.

Constant, J. (1997) Alcohol ischemic heart disease and the French paradox. Coronary Artery Disease 10, 645–649.

Cook, R., Ballas, Z. K., Waldschmidt, T. J., Vandersteen, D., LaBrecque, D. R. and Cook, B. L. (1995) Modulation of T-cell adhesion markers, and the CD45R and CD57 antigens in human alcoholics. Alcoholism: Clinical and Experimental Research 19, 555–563.[ISI][Medline]

Douds, A. C., Lim, A. G., Jazrawi, R. P., Finlayson, C. and Maxwell, J. D. (1997) Serum intercellular adhesion molecule-1 in alcoholic liver disease and its relationship with histological disease severity. Journal of Hepatology 26, 280–286.[ISI][Medline]

Gearing, A. J. H. and Newman, W. (1993) Circulating adhesion molecules in disease. Immunology Today 14, 506–512.[ISI][Medline]

Ghaisas, N. K., Shahi, C. N., Foley, B., Goggins, M., Crean, P., Kelly, A., Kelleher, D. and Walsh, M. (1997) Elevated levels of circulating soluble adhesion molecules in peripheral blood of patients with unstable angina. American Journal of Cardiology 80, 617–619.[ISI][Medline]

Hwang, S. J., Ballantyne, C. M., Sharrett, A. R., Smith, L. C., Davis, C. E., Gotto, A. M. and Boerwinkle, E. (1997) Circulating adhesion molecules VCAM-1, ICAM-1 and E-selectin in carotid atherosclerosis and incident coronary heart disease cases: the Atherosclerosis Risk in Communities (ARIC). Circulation 96, 4219–4225.[Abstract/Free Full Text]

Ishii, K., Furedera, S., Kumashiro, R., Seo, J., Koga, Y., Sata, M. and Tanikawa, K. (1994) Role of serum interleukin-8 and intercellular adhesion molecule-1 in the severity of alcoholic hepatitis. Alcohol and Alcoholism 29 (Suppl. 1), 81–85.

Khoruts, A., Stahnke, L., McClain, C., Logan, G. and Allen, J. I. (1991) Circulating tumor necrosis factor, interleukin-1 and interleukin-6 concentrations in chronic alcoholic patients. Hepatology 13, 267–276.[ISI][Medline]

Montefort, S., Lai, C. K., Kapahi, P., Leung, J., Chan, H. S., Haskard, D. O., Howarth, P. H. and Holgate S. T. (1994) Circulating adhesion molecules in asthma. American Journal of Respiratory and Critical Care Medicine 149, 1149–1152.[Abstract]

Nicolás, J. M., Estruch, R., Antúnez, E., Sacanella, E. and Urbano-Márquez, A. (1993) Nutritional status in chronically alcoholic men from the middle socioeconomic class and its relation to ethanol intake. Alcohol and Alcoholism 28, 551–558.[Abstract]

Ross, R. (1993) The pathogenesis of atherosclerosis: a perspective for the 1990s. Nature 362, 801–809.[ISI][Medline]

Sacanella, E., Estruch, R., Gayà, A., Fernández-Solà, J., Antúnez, E. and Urbano-Márquez, A. (1998) Activated lymphocytes (CD25+ CD69+ cells) and decreased CD19+ cells in well-nourished chronic alcoholics without ethanol-related diseases. Alcoholism: Clinical and Experimental Research 22, 897–901.[ISI][Medline]

Sacanella, E., Estruch, R., Gayà, A., Ferrer, K., Fernández-Solà, J., Alonso, J. R., Nicolás, J. M. and Urbano-Márquez A. (1999) Upregulated expression of VLA proteins and CD29 in peripheral blood lymphocytes of chronic alcoholics without ethanol-related diseases. Alcoholism: Clinical and Experimental Research 22, 371–375.

Sacco, R. L., Elkind, M., Boden-Albala, B., Lin, I-F., Kargman, D. E., Hauser, W. A., Shea, S. and Paik, M. C. (1999) The protective effect of moderate alcohol consumption on ischemic stroke. Journal of the American Medical Association 281, 53–60.[Abstract/Free Full Text]

Shimada, S., Yamauchi, M. and Toda, G. (1993) Serum levels of intercellular adhesion molecule-1 in patients with alcoholic liver disease. Alcohol and Alcoholism 28 (Suppl. 1B), 47–51.

Sobell, L. C., Maisto, S. A., Sobell, M. B. and Cooper, A. M. (1979) Reliability of alcohol abusers' self-reports of drinking behaviors. Behaviour Research Therapy 17, 157–160.[ISI][Medline]

Watson, R. R., Borgs, P., Witte, M., McCuskey, R. S., Lantz, C., Johnson, M. I., Mufti, S. I. and Earnest, D. L. (1994) Alcohol, immunomodulation, and disease. Alcohol and Alcoholism 29, 131–139.[Abstract]