Acetaldehyde-Related Pathology: Bridging the Trans-Disciplinary Divide (Novartis Foundation Symposium 285)

Novartis Foundation Symposium 285

ACETALDEHYDERELATED PATHOLOGY: BRIDGING THE TRANSDISCIPLINARY DIVIDE

ACETALDEHYDE-RELATED PATHOLOGY: BRIDGING THE TRANS-DISCIPLINARY DIVIDE

The Novartis Foundation is an international scientific and educational charity (UK Registered Charity No. 313574). Known until September 1997 as the Ciba Foundation, it was established in 1947 by the CIBA company of Basle, which merged with Sandoz in 1996, to form Novartis. The Foundation operates independently in London under English trust law. It was formally opened on 22 June 1949. The Foundation promotes the study and general knowledge of science and in particular encourages international co-operation in scientific research. To this end, it organizes internationally acclaimed meetings (typically eight symposia and allied open meetings and 15–20 discussion meetings each year) and publishes eight books per year featuring the presented papers and discussions from the symposia. Although primarily an operational rather than a grantmaking foundation, it awards bursaries to young scientists to attend the symposia and afterwards work with one of the other participants. The Foundation’s headquarters at 41 Portland Place, London W1B 1BN, provide library facilities, open to graduates in science and allied disciplines. Media relations are fostered by regular press conferences and by articles prepared by the Foundation’s Science Writer in Residence. The Foundation offers accommodation and meeting facilities to visiting scientists and their societies. Information on all Foundation activities can be found at http://www.novartisfound.org.uk

Novartis Foundation Symposium 285

ACETALDEHYDERELATED PATHOLOGY: BRIDGING THE TRANSDISCIPLINARY DIVIDE

Copyright © Novartis Foundation 2007 Published in 2007 by John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester PO19 8SQ, UK National 01243 779777 International (+44) 1243 779777 e-mail (for orders and customer service enquiries): [email protected] Visit our Home Page on http://www.wileyeurope.com or http://www.wiley.com All Rights Reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except under the terms of the Copyright, Designs and Patents Act 1988 or under the terms of a licence issued by the Copyright Licensing Agency Ltd, 90 Tottenham Court Road, London W1T 4LP, UK, without the permission in writing of the Publisher. Requests to the Publisher should be addressed to the Permissions Department, John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ, England, or emailed to [email protected], or faxed to (+44) 1243 770620. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the Publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. Other Wiley Editorial Offices John Wiley & Sons Inc., 111 River Street, Hoboken, NJ 07030, USA Jossey-Bass, 989 Market Street, San Francisco, CA 94103-1741, USA Wiley-VCH Verlag GmbH, Boschstr. 12, D-69469 Weinheim, Germany John Wiley & Sons Australia Ltd, 33 Park Road, Milton, Queensland 4064, Australia John Wiley & Sons (Asia) Pte Ltd, 2 Clementi Loop #02-01, Jin Xing Distripark, Singapore 129809 John Wiley & Sons Canada Ltd, 6045 Freemont Blvd, Mississauga, Ontario, Canada L5R 4J3 Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Novartis Foundation Symposium 285 x + 276 pages, 41 figures, 24 tables

Anniversary Logo Design: Richard J Pacifico British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN: 978-0-470-05766-7 Typeset in 10½ on 12½ pt Garamond by SNP Best-set Typesetter Ltd., Hong Kong Printed and bound in Great Britain by T. J. International Ltd, Padstow, Cornwall. This book is printed on acid-free paper responsibly manufactured from sustainable forestry, in which at least two trees are planted for each one used for paper production.

Contents

Symposium on Acetaldehyde-related pathology: bridging the trans-disciplinary divide, held at the Novartis Foundation, London 5–7th September 2006 Editors: Derek J. Chadwick (Organizer) and Jamie Goode This symposium is based on a proposal made by Victor Preedy, Peter Emery and Mikko Salaspuro Peter Emery Chair’s introduction 1 David W. Crabb and Suthat Liangpunsakul Acetaldehyde generating enzyme systems: roles of alcohol dehydrogenase, CYP2E1 and catalase, and speculations on the role of other enzymes and processes 4 Discussion 16 Richard A. Deitrich, Dennis Petersen and Vasilis Vasiliou Removal of acetaldehyde from the body 23 Discussion 40 Shih-Jiun Yin and Giia-Sheun Peng Acetaldehyde, polymorphisms and the cardiovascular system 52 Discussion 63 Jun Ren Acetaldehyde and alcoholic cardiomyopathy: lessons from the ADH and ALDH2 transgenic models 69 Discussion 76 Mikko Salaspuro Interrelationship between alcohol, smoking, acetaldehyde and cancer 80 Discussion 89 v

vi

CONTENTS

Hiroto Matsuse, Chizu Fukushima, Terufumi Shimoda, Sadahiro Asai and Shigeru Kohno Effects of acetaldehyde on human airway constriction and inflammation 97 Discussion 106 Helmut K. Seitz The role of acetaldehyde in alcohol-associated cancer of the gastrointestinal tract 110 Discussion 119 Robert Tardif The determination of acetaldehyde in exhaled breath 125 Discussion 133 Mostofa Jamal, Kiyoshi Ameno, Mitsuru Kumihashi, Weihuan Wang, Ikuo Uekita and Iwao Ijiri Ethanol and acetaldehyde: in vivo quantitation and effects on cholinergic function in rat brain 137 Discussion 141 Ville Salaspuro Pharmacological treatments and strategies for reducing oral and intestinal acetaldehyde 145 Discussion 153 Victor R. Preedy, David W. Crabb, Jaume Farrés and Peter W. Emery Alcoholic myopathy and acetaldehyde 158 Discussion 177 Onni Niemelä Acetaldehyde adducts in circulation 183 Discussion 193 General discussion 198 M. Apte, J. McCarroll, R. Pirola and J. Wilson pathways and acetaldehyde 200 Discussion 211

Pancreatic MAP kinase

Shivendra D. Shukla, Youn Ju Lee, Pil-hoon Park and Annayya R. Aroor Acetaldehyde alters MAP kinase signalling and epigenetic histone modifications in hepatocytes 217 Discussion 224 Paul J. Thornalley Endogenous α-oxoaldehydes and formation of protein and nucleotide advanced glycation endproducts in tissue damage 229 Discussion 243

CONTENTS

C. J. Peter Eriksson Measurement of acetaldehyde: what levels occur naturally and in response to alcohol? 247 Discussion 256 Final discussion 261 Contributors Index 265 Subject index 267

vii

Participants

Emanuele Albano Dipartimento di Scienze Mediche, Università Amedeo Avogadro del Piemonte Orientale, Via Solaroli 17, 28100 Novara, Italy Minoti V. Apte Pancreatic Research Group, South Western Sydney Clinical School, The University of New South Wales, Room 517, Level 5, Wallace Wurth Building, UNSW, Sydney, NSW 2052, Australia Agustin Aranda Departamento de Bioquimica y Biologia Molecular, Universidad de Valencia y Departamento de Biotecnologia, Instituto de Agroquimica y Tecnologia de Alimentos, Apdo 73, Burjassot, 46100, Valencia, Spain David W. Crabb Indiana University School of Medicine, 545 Barnhill Drive, Emerson Hall, Room 317, Indianapolis, IN 46202-5124, USA Richard A. Deitrich University of Colorado Health Sciences Center at Fitsimons, Department of Pharmacology, Alcohol Research Center, 12800 E. 19th Ave, P O Box 6211, Mail stop 8303, Aurora, CO 80045-0508, USA Peter Emery (Chair) Department of Nutrition and Dietetics, King’s College London, The Franklin Wilkins Building, 150 Stamford Street, London SE1 9NH, UK C. J. Peter Eriksson Department of Mental Health and Alcohol Research, National Public Health Institute, KTL/ATY, POB 33, Helsinki, FIN-00251, Finland Mostofa Jamal Department of Forensic Medicine, Faculty of Medicine, Kagawa University, 1750-1 Ikenobe, Miki, Kita, Kagawa, 761-0793, Japan Hiroto Matsuse Second Department of Internal Medicine, Nagasaki University School of Medicine, 1-7-1 Sakamoto, Nagasaki, 852-8501, Japan viii

PARTICIPANTS

ix

John B. Morris Department of Pharmaceutical Sciences, School of Pharmacy, University of Connecticut, 69 N Eagleville Road, Box U-3092, Storrs, CT 062693092, USA Onni Niemelä Department of Laboratory Medicine and Medical Research Unit, Seinäjoki Central Hospital and University of Tampere, FIN-60220 Seinäjoki, Finland Tomonori Okamura Department of Health Science, Shiga University of Medical Science, Seta Tsukinowa-cho, Otsu City, Shiga, 520-2192, Japan Victor R. Preedy Department of Nutrition and Dietetics, School of Life Sciences, King’s College London, The Franklin Wilkins Building, 150 Stamford Street, London SE1 9NH, UK Etienne Quertemont Psychologie Quantitative, Department of Cognitive Sciences, Boulevard du Rectorat 5/B32, University of Liège, Liège 4000, Belgium Radhakrishna K. Rao Department of Physiology, University of Tennessee, 894 Union Ave, Nash 426, Memphis, TN 38163, USA Jun Ren Center for Cardiovascular Research and Alternative Medicine, Division of Pharmaceutical Sciences, University of Wyoming, 1000 E. University Avenue, Department 3375, Laramie, WY 82071, USA Mikko Salaspuro Research Unit of Substance Abuse Medicine, Biomedicum Helsinki, Haartmaninkatu 8, Helsinki, 00029 HUS, Finland Ville Salaspuro Research Unit of Substance Abuse Medicine, Biomedicum Helsinki, Haartmaninkatu 8, Helsinki, 00029 HUS, Finland Helmut K. Seitz Laboratory of Alcohol Research, Liver Disease and Nutrition, Department of Medicine, Salem Medical Centre Heidelberg and University of Heidelberg, Heidelberg, Germany Shivendra D. Shukla Department of Medical Pharmacology & Physiology, School of Medicine, University of Missouri, Columbia, MO 65212, USA Sophie Tambour (Novartis Foundation Bursar) University of Liege, Département des Sciences Cognitives, Laboratoire de Neuroscience Comportementale

x

PARTICIPANTS

et de Psychopharmacologie, Boulevard du Rectorat 5/B32, B-4000 Liege, Belgium Robert Tardif Department of Occupational and Environmental Health, Faculty of Medicine, University of Montreal, 2375 Cote St-Catherine, Montreal, Québec, Canada H3T 1A8 Paul J. Thornalley Protein Damage and Systems Biology Research Group, Warwick Medical School & Systems Biology Centre, Clinical Sciences Research Institute, University of Warwick, University Hospital, Coventry CV2 2DX, UK Simon Worrall Alcohol Research Unit, Biochemistry and Molecular Biology, School of Molecular and Microbial Sciences, University of Queensland, Brisbane, QLD 4072, Australia Shih-Jiun Yin Department of Biochemistry, National Defense Medical Center, 161 Min-Chuan East Road, Section 6, Taipei 114, Taiwan

Chair’s introduction Peter Emery Department of Nutrition and Dietetics, King’s College London, The Franklin Wilkins Building, 150 Stamford Street, London SE1 9NH, UK

Although there are three proposers listed for this symposium, the idea for this meeting came largely from Victor Preedy, and his enthusiasm caught up Mikko Salaspuro and myself. I have been working with Victor for a few years on a number of projects, one of which has involved looking at aspects of acetaldehyde protein adducts. As we were working on this project, the idea came to us that although many people are interested in acetaldehyde because of its role as a metabolite of alcohol, there are in fact many other aspects of acetaldehyde that are equally important. The idea here was to bring together people who are approaching acetaldehyde from a number of different perspectives to try to understand more about what these different approaches can bring. Acetaldehyde is an appealingly simple small molecule that is pretty reactive in vitro, and has some fairly serious effects in vivo in a variety of physiological systems. I would like to outline some of the questions that we might want to explore during this meeting. As a nutritionist I tend to start from a simple point of view, asking questions such as ‘where does it come from?’ Many people will think of it mainly as the first metabolite of alcohol. Perhaps the key to understanding many of the damaging effects of acetaldehyde is the distribution of the alcohol dehydrogenase (ADH) enzymes in tissues. We know that there is a fair amount of acetaldehyde produced quite separately from the ingestion of ethanol. In particular, the bacteria in the oral cavity and throughout the gut can produce it. They may be contributing significant amounts of acetaldehyde: does this have systemic effects or is it a local phenomenon? We also know that acetaldehyde is a product of cigarette smoke. We will hear about the effects of acetaldehyde from this source and its interaction with alcohol intake. There are also increasing amounts of acetaldehyde in the air: it is a volatile molecule produced by combustion of hydrocarbons. As we live in an increasingly polluted world, more of our exposure to acetaldehyde may be coming from the air. There is also acetaldehyde in some foodstuffs, particularly fermented foods. We don’t know whether this is a significant source of acetaldehyde intake as well. 1

2

EMERY

Once it is in the body, how is it metabolized? It is metabolized by various dehydrogenase and oxidase enzymes. There are background levels of these activities, which may increase considerably in response to exposure. We will hear a lot about polymorphisms of these different enzymes, which give us useful biological models for studying exposure to acetaldehyde. People produce very different amounts of acetaldehyde in response to the same amount of alcohol intake. There are other aldehydes present, and some of the systems we have for metabolizing acetaldehyde will also metabolize other aldehydes. The interaction with other aldehydes could be key to understanding some of the physiological actions of acetaldehyde. This leads us to consider the whole question of the variety of antioxidant defences that may be induced and up-regulated when we are exposed to acetaldehyde. It comes in, it is metabolized, but what we really need to measure is how much acetaldehyde is present in various tissues and for how long. This is difficult with a molecule like this that is short-lived and moves around. Instead of tissue concentrations we may have to look at proxies such as blood levels. Saliva may be useful: of course, this will reflect what is produced in the mouth, but it could also reflect systemic production and exposure to an extent. Acetaldehyde is a volatile molecule so we may be able to measure it in the breath. We may be able to measure products of acetaldehyde metabolism in the urine; this could be a long-term integrative measure of exposure, rather than reflecting acute changes. A key question from a biochemical viewpoint is the mechanism by which damage is caused. I’m sure we’ll hear a lot about this fairly vague term ‘oxidative stress’. Hopefully, we can be more precise in our discussion to clarify what is meant by this term. As an electrophilic molecule, acetaldehyde will attack many nucleophilic centres in a variety of important molecules, particularly forming adducts with DNA and protein, and indeed the lipid components. The question then becomes, what is the subsequent damage caused by production of these adducts? Which sorts of proteins may be affected? Proteins in signalling pathways may be amplifying the signal and the damage that is caused. There could be effects within the nucleus through transcription factors or DNA repair enzymes, or epigenetic effects on histone decoration. We will consider the tissues that are affected. Victor Preedy always teaches his students about the effects of alcohol on different tissues. When we look at where acetaldehyde might be having its damaging effects, it concerns a great variety of tissues, and not just the liver. There is the gut, and many cancers through the gastrointestinal tract may relate to acetaldehyde damage. There could be effects on the brain, which could lead not only to behavioural effects but also degenerative diseases. There are effects on the heart and cardiovascular system, and effects on skeletal muscle. The lung will be exposed to acetaldehyde from the atmosphere, and asthma, bronchitis and emphysema have all been linked with acetaldehyde.

CHAIR’S INTRODUCTION

3

Breast cancer development seems to be sensitive to differences in acetaldehyde production. Finally, the question we would like to move towards: what can we do about it? We are developing the idea that acetaldehyde is a damaging molecule, so are there ways of avoiding or minimizing this damage? Are there agents that could be used to bind acetaldehyde and reduce the amount that is present? Can we affect the amounts that are produced by reducing bacterial populations or manipulating them to change their characteristics? What can we do with the enzymes that are involved in producing or removing acetaldehyde to minimize the amount of acetaldehyde we are exposed to? The following list highlights some of the questions about acetaldehyde that we might want to explore over the next few days: • • • • • •

How does it get into the body? How is it metabolized? How much is there? How does it cause damage? Which tissues are affected? What can we do about it?

So let’s start the story where it needs to begin, by looking at alcohol dehydrogenase and other enzymes involved in acetaldehyde production.

Acetaldehyde generating enzyme systems: roles of alcohol dehydrogenase, CYP2E1 and catalase, and speculations on the role of other enzymes and processes David W. Crabb and Suthat Liangpunsakul Indiana University School of Medicine and Roudebush VA Medical Center, Emerson Hall Room 317, 545 Barnhill Drive, Indianapolis, IN 46202, USA

Abstract. Most acetaldehyde is generated in the liver by alcohol dehydrogenase (ADH) during ethanol metabolism. Polymorphic variants of these genes encode enzymes with altered kinetic properties, and pathophysiological effects of these variants may be mediated by accumulation of acetaldehyde. Two additional pathways of acetaldehyde generation are by the cytochrome P450 2E1 (CYP2E1) and catalase. While the amount of ethanol oxidized by these enzymes comprises a small fraction of total body ethanol clearance, the local formation of acetaldehyde by these enzymes may have important effects. Additional sources of acetaldehyde include other minor enzymes (nitric oxide synthase, other cytochrome P450s, P450 reductase, xanthine oxidoreductase) as well as non-enzymatic pathways (formation of hydroxyethyl radicals from the reaction of ethanol with hydroxyl radical, and its subsequent decomposition to acetaldehyde). Acetaldehyde may have effects locally (in the cells generating it), or when delivered to other cells by the blood stream or saliva, or by diffusion from the lumen of the gastrointestinal tract. The ultimate determinants of acetaldehyde toxicity include rates of its formation, rates of oxidation, and the capacity of cellular systems to prevent or repair chemical effects of acetaldehyde (e.g. formation of protein adducts or modification of nucleic acid bases). 2007 Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Foundation Symposium 285) p 4 –22

Chronic consumption of large amounts of ethanol has well known effects on the heart, liver, brain, muscles, fetus and pancreas, and is involved in the pathogenesis of a number of neoplasms. The susceptibility of individuals to the ill effects of alcohol consumption is due to complex interactions of genes and the environment. Many of the effects of ethanol are mediated by acetaldehyde, which is mainly 4

ACETALDEHYDE GENERATING ENZYME SYSTEMS NAD+

5

NADH

Ethanol

Acetaldehyde Alcohol dehydrogenase NADP+

NADPH, O2

Acetaldehyde

Ethanol

CYP2E1 H2O2

H2O

Acetaldehyde

Ethanol

Catalase

FIG. 1. Major enzymatic pathways for acetaldehyde formation. The major pathways of acetaldehyde formation, alcohol dehydrogenase, cytochrome P450 2E1 (CYP2E1), and catalase are shown with their cofactors, substrates and products.

generated by alcohol dehydrogenase (ADH), cytochrome P450 2E1 (CYP2E1) and catalase (Fig. 1). Some of the enzymes involved in acetaldehyde formation are genetically polymorphic; when the polymorphism alters the enzymatic properties of the enzyme or the level of its expression, different individuals may generate varying amounts of acetaldehyde in a given tissue. Furthermore, the enzyme activities are in some cases regulated by transcriptional or translational mechanisms, controls, concentrations of substrates and products, and by post-translational modification (Table 1). These enzymes, the regulation of their activity, and tissue distribution, as well as some minor enzymatic processes that form acetaldehyde, are the subject of this overview. Enzymology of acetaldehyde formation Alcohol dehydrogenases General description. The enzymes responsible for the bulk of alcohol oxidation are the ADHs. All are dimeric enzymes with subunit molecular weight of about 40 kDa; subunits are identified by Greek letters. These enzymes are grouped into classes based upon enzymatic properties and the degree of sequence similarities. Enzyme subunits belonging to the same class can heterodimerize. The general properties of these enzymes are summarized in Table 2. Class I contains α, β, and γ isozymes. These enzymes have a low Km for ethanol and are highly sensitive to inhibition by pyrazole derivatives. They are very abundant in liver, and play a major role in alcohol metabolism. Class II ADH (πADH) is also abundant in liver, has a higher Km for ethanol, and is less sensitive to pyrazole inhibition than class I enzymes (Ehrig

6

CRABB & LIANGPUNSAKUL

TABLE 1 Properties of alcohol dehydrogenases (ADHs) in humans Gene locus Class I ADH1 ADH2 ADH3 Class II ADH4 Class III ADH5 Class IV* ADH7 Class V* ADH6

New nomenclature

Subunit type

Km (ethanol)

Vmax

4 0.05–34** 0.6–1**

54 — —

Tissue distribution

ADH1A ADH1B ADH1C

α β γ

ADH4

π

34

ADH5

χ

1000

ADH7

σ, µ

20

1510

Stomach, oesophagus, other mucosae

ADH6

−

30

?

Liver, stomach

40 —

Liver Liver, lung Liver, stomach Liver, cornea Most tissues

* Tentative assignments based upon sequence homologies. Km values are given in mM and Vmax values are given in terms of turnover number (min−1). Tissue distributions indicate tissues with relatively high expression; see Fig. 3 for more detailed distribution information. ** Kinetic constants vary with isozyme, see Table 2.

TABLE 2 Properties of polymorphic forms of human alcohol dehydrogenase (ADH) Gene locus ADH2 *1 (ADH1B *1) ADH2 *2 (ADH1B *2) ADH2 *3 (ADH1B *3) ADH3 *1 (ADH1C *1) ADH3 *2 (ADH1C *2)

Subunit type

Km (ethanol)

Vmax

Population

β1 β2 β3 γ1 γ2

0.05 0.9 34 1.0 0.63

9 400 300 87 35

Caucasians, African-Americans Asians African-Americans All groups Caucasians

The kinetic constants are noted for the homodimers of the subunits listed (Ehrig et al 1990). Heterodimers behave as if the active sites were independent. The Km values are in mM and the Vmax values are given in terms of turnover numbers (min−1), as in Table 1. The column labelled population indicates which populations have high allele frequencies for these variants. The alleles are not limited to those populations.

et al 1990). Class III ADH (χADH) is present in nearly all tissues, is virtually inactive with ethanol, but can metabolize longer chain alcohols, ω-hydroxy-fatty acids, and formaldehyde. A recent paper suggested that class III ADH might be more active towards ethanol in a hydrophobic environment, and argues that liver cytosol may be such an environment (Haseba et al 2006). The class IV enzyme was purified from stomach and oesophagus (Pares et al 1994). σADH has the highest Vmax of the known ADHs and is very active towards

ACETALDEHYDE GENERATING ENZYME SYSTEMS Genes

Expression variants: promoter, mRNA stability, translation efficiency

Protein

7

Post-translational modification Substrate/product

Altered enzyme mass

Enzyme activity

Altered kinetic properties

Protein stability

Saturation effects (high Km variants) Product inhibition

Coding region variants

Protein stabilization

FIG. 2. Factors which control the rate of enzymatic generation of acetaldehyde. Genetic variation can influence the expression of the gene (transcriptional effects) or the stability or translational efficiency of the mRNA, as well as alter the coding sequence. Transcriptional or mRNA effects will result in varying amounts of active enzyme, and thus determine the maximum flux through the pathway. Coding region variants for ADH have widely varying kinetic properties. Post-translational modifications can influence the activity of an enzyme or its susceptibility to degradation, as can the degree of substrate binding to the enzyme, as in the case of CYP2E1 stabilization by substrate. The enzyme activity is ultimately determined by the concentrations of substrate and product, the kinetic constants for each isozyme, and the total activity of the enzyme, as defined by the kinetic rate equation for the enzyme (Crabb et al 1983).

retinol. This may be relevant to its expression in numerous epithelia which are dependent on retinol for their integrity. Class V ADH, encoded by the ADH6 gene, is expressed in liver and in stomach, but the enzyme itself has not been purified. In vitro expressed enzyme had a high Km for ethanol (about 30 mM), and moderate sensitivity to pyrazole inhibition (Cheng & Yoshida 1991). Class VI ADH was reported in deer, mouse and rat liver; class VII ADH was cloned from chicken, but the human homologues have not been found. Genetic variants. The nomenclature for ADH genes was recently revised. The ADH1, 2 and 3 genes are now designated ADH1A, ADH1B, and ADH1C genes, respectively. Two of the seven human ADH gene loci are polymorphic, and the prevalence of the alleles depends on continental origin. The kinetic properties and population distributions of these allelic enzymes are shown in Fig. 3. The isozymes encoded by the three ADH1B alleles, differing at single amino acids, vary markedly in Km for ethanol and Vmax. β1 is most common in Caucasians, has a low Vmax and a very low Km for ethanol. β2 is found in Asians and Ashkenazi Jews. It has a substantially higher Vmax and somewhat higher Km compared with β1. The β3 isozyme was first detected in samples from African-Americans, and has also been found in Southwest Native Americans. It has a high Km for ethanol and high Vmax. Smaller differences in enzymatic properties are observed between the products of the ADH1C alleles. The γ1 isozyme has about twice the Vmax of the γ2 isozyme,

8

CRABB & LIANGPUNSAKUL ADH1C

Tissue

ADH4

ADH6

ADH7

CYP2E1

CAT

blood

0

17

0

0

53

367

bone

13

0

0

0

13

55

bone marrow

0

0

0

0

0

634

brain

27

0

1

0

19

47

connective tissue

74

0

0

0

0

65

adipose tissue

4251

0

0

0

0

144

liver

1930

729

252

0

843

319

pancreas

36

4

4

0

0

95

adrenal gland

611

0

0

0

0

32

thyroid

0

0

0

0

18

163

placenta

16

0

0

0

0

121

eye

9

0

0

19

0

67

cervix

62

0

20

0

0

41

ovary

0

0

9

0

28

0

uterus

217

0

8

0

4

62

prostate

32

0

0

0

6

51

testis

28

0

11

0

8

48

bladder

132

0

0

33

0

99

kidney

56

0

84

0

0

79

tongue

30

0

15

90

0

30

larynx

32

0

0

32

0

98

pharynx

0

0

0

0

0

0

salivary gland

0

0

48

0

0

146

heart

602

0

55

0

0

100

lymph node

10

0

0

0

0

146

spleen

416

0

0

0

0

37

thymus

135

0

0

0

13

0

mammary gland

450

29

23

0

29

58

muscle

122

0

8

17

8

69

lung

169

0

0

40

28

69

trachea

1444

0

0

288

0

20

skin

21

0

0

0

0

85

vascular

118

0

0

0

0

157

small intestine

1558

22

90

0

0

22

colon

153

0

14

0

0

84

stomach

254

0

48

9

0

19

esophagus

472

0

52

996

0

0

nerve tissue

550

0

0

0

39

118

FIG. 3. Tissue distribution of ADH, CYP2E1, and catalase transcripts reflected by the abundance of expressed sequence tags (ESTs). Tissue distribution of ESTs for the noted genes were obtained from the NCBI Unigene Database using the EST Profile Viewer (e.g. http://www.ncbi. nlm.nih.gov/UniGene/ESTProfileViewer.cgi?uglist=Hs.78989 for ADH5). The number in each cell is the number of transcripts per million, a measure of the abundance of the transcripts. The spot intensity is based on the transcripts per million.

ACETALDEHYDE GENERATING ENZYME SYSTEMS

9

while the Kms for ethanol are similar. γ1 ADH is found at high frequency in Asians and African-Americans; Caucasians have about equal frequency of γ1 and γ2 ADH alleles (Burnell & Bosron 1989). The other ADH loci have not been found to be polymorphic to date. Individuals expressing ADH1B *2 and ADH1B *3 would be predicted to metabolize ethanol more rapidly and generate more acetaldehyde; however, effects of the polymorphism on ethanol elimination rates are small. Different ADH1B *2 genotypes are correlated with only a small fraction of the between-individual differences in alcohol elimination rates. The ADH1B *3 polymorphism confers a 10% increase in the rate of ethanol metabolism; both it and ADH1B *2 are protective against alcoholism (Edenberg et al 2006). The ADH1C polymorphism did not affect alcohol elimination rate, but recent data link the ADH1C *1 allele with head and neck, oesophageal, breast and hepatocellular carcinomas (Homann et al 2006), which could reflect increased rates of acetaldehyde formation. An additional ADH genetic variant is a Pvu II restriction fragment length polymorphism (RFLP) in an intron of the ADH1B gene. It is not known if the variant alters expression of the gene or is linked to another susceptibility locus; the B allele was found at higher frequency in alcoholics and in patients with alcoholic cirrhosis (Sherman et al 1993). Single nucleotide polymorphisms (SNPs) presumed to influence expression of the ADH4 gene have been linked to risk of alcoholism (Edenberg et al 2006); one polymorphism in the promoter affects gene expression (Edenberg et al 1999). Similarly, sequence variants in the promoter of ADH1C may affect its expression (Chen et al 2005). Control of expression of ADHs. The ADH1 promoters are all active in liver. They interact with ubiquitous transcription factors (e.g. TATAA binding factors, upstream stimulatory factor [USF], CTF/NF-I and Sp1-like factors), as well as tissue-specific factors (e.g. hepatocyte nuclear factor 1 [HNF-1], D-box binding protein [DBP] and CCAAT-enhancer binding proteins [C/EBPα and β]). An HNF-1 site was recently reported to serve as a master control for all three of the class I genes (Su et al 2006). The ADH5 and ADH7 promoters lack TATAA boxes. The ADH5 promoter is G+C rich, a characteristic of housekeeping genes and consistent with its ubiquitous expression. Binding sites for thyroid hormone, retinoic acid and glucocorticoid receptors have been identified in the upstream regions of ADH1 genes. In in vitro experiments, retinoic acid and glucocorticoids activated the promoters and thyroid hormone antagonized the effect of retinoic acid; these hormones had less dramatic effects in vivo. Growth hormone increased ADH activity in rats and cultured hepatocytes, while androgens and thyroid hormones decreased it. Chronic ethanol consumption can affect the expression of ADH. Ethanol increased hepatic ADH activity in male rats by reducing testosterone levels. The amount of ethanol consumed from conventional liquid diets did not alter liver ADH activity, whereas higher doses achieved by intragastric ethanol infusion induced liver

10

CRABB & LIANGPUNSAKUL

ADH activity. This resulted from induction of the transcription factor C/EBPβ and suppression of C/EBPγ and a truncated, inhibitory form of C/EBPβ called LIP (He et al 2002). In addition, chronic intragastric infusion of ethanol increases portal vein endotoxin, which can induce ADH mRNA via increased binding of USF (Potter et al 2003). In humans, the amount of ADH in the liver was not induced by chronic drinking; however, with fasting, protein malnutrition and liver disease, ADH activity and the ethanol elimination rate were decreased. Orchiectomy increased alcohol elimination rates in humans. Little is known about expression of extrahepatic ADH, with the exception of gastric ADH, which is reduced with age, in women, and with heavy drinking (Seitz et al 1993). Post-translational modifications. No post-translational modifications of the ADH enzyme are recognized. However, peroxynitrite can oxidize the active site, causing disulfide formation and release of zinc, inactivating the enzyme (Daiber et al 2002); whether this is physiologically relevant remains to be seen. Role of substrate and product concentrations. The ADH isozymes with high Km for ethanol, e.g. β3, π, and σ will be more active when blood ethanol concentrations are high or in tissues of the upper gastrointestinal (GI) tract that are directly exposed to beverage ethanol. Modelling of alcohol oxidation in rat liver indicated that ADH activity was controlled by the total activity of the enzyme as well as product inhibition by NADH and acetaldehyde (Crabb et al 1983); thus ADH operates below its Vmax at steady state. Our laboratory determined the rate of ethanol oxidation by cells expressing ADH1B *1, ADH1B *2 and ADH1B *3. The inhibition constants for β1, β2 and β3 ADH were 1.5 ± 0.1, 22 ± 14 and 210 ± 5 µM, respectively (Matsumoto et al, unpublished data), indicating that activity of β1 and β2 ADH could be limited by the accumulation of acetaldehyde. Tissue distribution. ADHs are expressed in a variety of tissues. High levels of class I ADH mRNA were found in kidney, stomach, duodenum, colon and uterus of rats, with lower levels in many organs including the lung, small intestine and hepatic Ito cells, and much lower levels were found in brain, thymus, muscle or heart (Estonius et al 1996). Cytosolic ADH has been found in parotid gland, and chronic alcohol use was associated with parotid steatosis (Maier et al 1986). Class I ADH is found in blood vessels, which may be relevant to alcohol-induced flushing and cardiovascular effects of ethanol consumption. Class II ADH was detected in liver and duodenum (Estonius et al 1996). Gastric mucosa contains several ADHs (γ-, χ-, and σADH). σADH is absent in the stomach biopsies of about 30% of Asians, and those lacking this enzyme had lower first pass metabolism of ethanol (Dohmen et al 1996), suggesting that σADH is important in gastric oxidation of ethanol.

ACETALDEHYDE GENERATING ENZYME SYSTEMS

11

Relative expression of various mRNAs can be estimated from the frequency of expressed sequence tags (ESTs) detected in cDNA libraries. Figure 3 shows the relative expression of ADH1C, ADH4, ADH6 and ADH7 transcripts in a number of tissues. Microorganisms express numerous forms of alcohol dehydrogenase, which can contribute to the formation of acetaldehyde in the lower GI tract, or wherever microbial overgrowth occurs. Cytochrome P450 2E1 General description. Ethanol can be metabolized by microsomal ethanol oxidizing systems, predominantly via cytochrome P450 2E1 (CYP2E1). Other cytochromes, CYP1A2 and CYP3A4, also contribute to a lesser extent (Lieber 2004). CYP2E1 is associated with NADPH-cytochrome P450 reductase in the endoplasmic reticulum, and reduces molecular oxygen to water as ethanol is oxidized to acetaldehyde. It is responsible for perhaps 10% of ethanol elimination. CYP2E1 is inducible by chronic drinking especially in the perivenular zone, and it may contribute to the increased rates of ethanol elimination in heavy drinkers. CYP2E1 is induced in fasting, diabetes and by a diet high in fat, which may relate to its ability to oxidize the ketone body acetone. Its Km for ethanol is about 10 mM; thus CYP2E1 may assume a greater role in ethanol metabolism at high blood alcohol levels. CYP2E1 is unusually ‘leaky’ and generates reactive oxygen species (ROS) including hydroxyl radical (OH• ), superoxide anion (O2 −), hydrogen peroxide (H2O2), and hydroxyethyl radical (HER• ). Thus, CYP2E1 is a major source of oxidative stress. CYP2E1 knockout animals had longer sleep times than normal counterparts, suggesting a role for CYP2E1 in brain sensitivity to ethanol (Vasiliou et al 2006). Genetic variants. An Rsa I (−1053C > T) polymorphism (the Rsa I+ allele is also named the c1 allele) is located in the 5′-flanking region of the CYP2E1 gene (Hayashi et al 1991) in a region interacting with HNF-1. The Rsa I− allele (c2) was more active in in vitro transcriptional assays, although a corresponding increase in CYP2E1 activity in vivo has not been unequivocally confirmed using the clearance of chlorzoxazone as a probe. The frequency of this polymorphism depends on continental origin: the c2 variant is found in 2–8% of Caucasians and in 25–36% of East Asians. Another polymorphism, detectable with the Dra I restriction enzyme, is located in intron 6. The distribution of the variant genotype (lacking the Dra I site) also depends on continental origin: 40–50% of East Asians carry this genotype, while only 10% of Caucasians lack the Dra I site. A more recently described polymorphism is the −71G > T polymorphism in exon 1, which has been associated with enhanced transcriptional activity of promoter constructs in HepG2 cells. Heterozygosity for this allele occurs in about 10% of Caucasians. The effects of the various genotypes on alcohol pharmacokinetics or risk of alcoholic

12

CRABB & LIANGPUNSAKUL

complications have been inconsistent, and there is no direct evidence of differences in rates of acetaldehyde formation. Control of expression of CYP2E1. The human CYP2E1 gene spans 11 kb, contains 9 exons, and contains a typical TATAA box. HNF-1 is critical for its expression. Expression is also controlled both at the level of mRNA (high concentrations of ethanol can induce transcription of the CYP2E1 gene [Takahashi et al 1993]) and by the stabilization of the protein, as observed for ethanol, acetone and pyrazole derivatives, which reduce the rate of proteasomal degradation (Lieber 2004). Recent data suggest that additional signals may affect its expression. For instance, CYP2E1 can be induced by interleukin (IL) 4 in liver (Lagadic-Gossmann et al 2000) and by phorbol ester and other cellular stresses in astrocytes (Tindberg 2003). Insulin post-transcriptionally reduced the expression of CYP2E1 by destabilizing its mRNA. Role of substrate and product concentrations. Since CYP2E1 has a high Km for ethanol, it will generate more acetaldehyde when ethanol concentrations are elevated. There is no evidence that acetaldehyde is a product inhibitor of CYP2E1; in fact, CYP2E1 can oxidize acetaldehyde to acetate, although probably not in the presence of ethanol. Post-translational modification. CYP2E1 is reported to be a substrate for cAMPdependent protein kinase A (PKA). Phosphorylation of a serine residue inactivates the enzyme (Oesch-Bartlomowicz et al 1998). Whether this plays a physiological role in controlling activity of this enzyme is not clear, although in several conditions in which CYP2E1 activity is low (fasting, diabetes), hepatic PKA activity is high. Tissue distribution. CYP2E1 is expressed at highest levels in the liver, as well as numerous other tissues, as demonstrated by western blotting, mRNA, or EST analyses (Fig. 3). These include kidney, lung, oesophagus, biliary epithelium, pancreas, uterus, leukocytes, breast, brain, colon, urinary bladder, nasal mucosa and pancreatic beta cells. Western blots and activity assays have confirmed expression of CYP2E1 in oesophagus, pancreas and lung, among others. In brain, CYP2E1 was reported to be expressed in neurons and inducible by ethanol administration (Tindberg & Ingelman-Sundberg 1996). Catalase General description. The peroxisomal catalase is a tetrameric, haem-containing enzyme. In addition to converting hydrogen peroxide (H2O2) to water and oxygen, it can oxidize ethanol to acetaldehyde in an H2O2-dependent fashion. This pathway is not

ACETALDEHYDE GENERATING ENZYME SYSTEMS

13

thought to be a major elimination pathway under most physiological conditions, but it may be important in certain tissues such as brain; in fact, acatalasaemic mice had longer sleep times than their normal counterparts (Vasiliou et al 2006). Genetic variation. The absence of active catalase (acatalasaemia) is encountered in Asian populations. A number of SNPs in the 5′ untranslated region and introns are reported, but there are no known effects of these variants on expression or activity of the enzyme, nor on responses to ethanol. Control of expression. Little is known regarding transcriptional control of catalase expression in mammalian cells. The rat catalase gene is a single-copy gene spanning 33 kb. The promoter region lacks a TATAA box and an initiator consensus sequence, contains multiple CCAAT boxes and GC boxes, and contains multiple transcription initiation sites, consistent with its housekeeping function. Chronic ethanol feeding was reported to increase catalase activity (Orellano et al 1998). The rat catalase promoter contains a peroxisome proliferator responsive element (PPRE [Girnun et al 2002]) and can be induced by peroxisome proliferators. Post-translational modification. In cells exposed to H2O2, Abl and Arg (non-receptor protein tyrosine kinases) associate with catalase and can activate it by phosphorylating two tyrosine residues. However, at higher concentrations of H2O2, phosphorylation of these residues can stimulate ubiquitination and proteasomal degradation of the enzyme (Cao et al 2003). Control by substrate and product levels. The activity of catalase depends upon the availability of H2O2. This was observed with perfused rat liver: when fatty acids were added to the perfusate, peroxisomal β oxidation generated H2O2 and stimulated ethanol oxidation. This raises the possibility that under conditions of oxidant stress (and H2O2 production) catalase-mediated ethanol oxidation may be increased. Tissue distribution. Catalase is expressed in nearly all tissues (Fig. 3). Catalase is also expressed by colonic micro-organisms and contributes to the formation of acetaldehyde from ethanol in the lower GI tract (Tillonen et al 1998). Other pathways of acetaldehyde generation A number of minor pathways of acetaldehyde generation have been suggested. Nitric oxide synthases 1 and 2 were reported to generate 1-hydroxyethyl radical from ethanol in the presence of NADPH and arginine. This is perhaps not

14

CRABB & LIANGPUNSAKUL

surprising given the presence of a CYP motif within the structure of the enzymes. 1-Hydroxyethyl radical can break down to form acetaldehyde (Porasuphatana et al 2006). Castro et al (2001) reported that cytosolic xanthine oxidoreductase is capable of oxidizing ethanol to acetaldehyde. CYP reductase (in the absence of CYP proteins such as CYP2E1) was reported to oxidize ethanol to 1-hydroxyethyl radical and acetaldehyde, possibly via the semiquinone form of FAD (Diaz-Gomez et al 2000). Other investigators report formation of acetaldehyde from ethanol in tissue extracts for which the responsible enzymes have not been identified in studies using different cofactors and inhibitors. It is possible that other oxidant species (hydroxyl radical) formed non-enzymatically might be able to oxidize ethanol to acetaldehyde. In addition, acetaldehyde can be formed during the degradation of threonine, putatively by threonine aldolase.

Summary Three major enzymes and several minor enzymes can generate acetaldehyde when ethanol is present. These enzymes are present in virtually all cells; thus, the ability of acetaldehyde to alter cellular function or to modify DNA or proteins, will depend on the rate of acetaldehyde formation (related to ethanol concentration, activity of the enzyme, and the presence or absence of inhibitors of the enzymes), and of its further oxidation by aldehyde dehydrogenases.

Acknowledgements This work was supported in part by P60 A07611 to DWC and a Young Investigator Award from the Richard Roudebush VA Medical Center to SL.

References Burnell JC, Bosron WF 1989 Genetic polymorphism of human liver alcohol dehydrogenase and kinetic properties of the isoenzymes. In: Crow KE, Batt RD (eds) Human metabolism of alcohol. 11th Edn. CRC Press, Boca Raton, FL, p 65–75 Cao C, Leng Y, Liu X, Yi Y, Li P, Kufe D 2003 Catalase is regulated by ubiquitination and proteosomal degradation. Role of the c-Abl and Arg tyrosine kinases. Biochemistry 42: 10348–10353 Castro GD, Delgado de Layno AM, Costantini MH, Castro JA 2001 Cytosolic xanthine oxidoreductase mediated bioactivation of ethanol to acetaldehyde and free radicals in rat breast tissue. Its potential role in alcohol-promoted mammary cancer. Toxicology 160: 11–18 Chen HJ, Tian H, Edenberg HJ 2005 Natural haplotypes in the regulatory sequences affect human alcohol dehydrogenase 1C (ADH1C) gene expression. Hum Mutat 25:150–155 Cheng C-S, Yoshida A 1991 Enzymatic properties of the protein encoded by newly cloned human alcohol dehydrogenase ADH6 gene. Biochem Biophys Res Comm 181:743–747

ACETALDEHYDE GENERATING ENZYME SYSTEMS

15

Crabb DW, Bosron WF, Li T-K 1983 Steady-state kinetic properties of purified rat liver alcohol dehydrogenase: application to predicting alcohol elimination rates in vivo. Arch Biochem Biophys 224:299–309 Daiber A, Frein D, Namgaladze D, Ullrich V 2002 Oxidation and nitrosation in the nitrogen monoxide/superoxide system. J Biol Chem 277:11882–11888 Diaz Gomez MI, Castro GD, de Layno AM, Costantini MH, Castro JA 2000 Cytochrome P450 reductase-mediated anaerobic biotransformation of ethanol to 1-hydroxyethyl-free radicals and acetaldehyde. Toxicology 154:113–122 Dohmen K, Baraona E, Ishibashi H et al 1996 Ethnic differences in gastric sigma-alcohol dehydrogenase activity and ethanol first-pass metabolism. Alcohol Clin Exp Res 20: 1569–1576 Edenberg HJ, Jerome RE, Li M et al 1999 Polymorphism of the human alcohol dehydrogenase 4 (ADH4) promoter affects gene expression. Pharmacogenetics 9:25–30 Edenberg HJ, Xuei X, Chen HJ et al 2006 Association of alcohol dehydrogenase genes with alcohol dependence: a comprehensive analysis. Hum Mol Genet 15:1539–1549 Ehrig T, Bosron WF, Li T-K 1990 Alcohol and aldehyde dehydrogenase. Alcohol Alcohol 25:105–116 Estonius M, Svensson S, Hoog JO 1996 Alcohol dehydrogenase in human tissues: localisation of transcripts coding for five classes of the enzyme. FEBS Lett 397:338–342 Girnun GD, Domann FE, Moore SA, Robbins ME 2002 Identification of a functional peroxisome proliferator-activated receptor response element in the rat catalase promoter. Mol Endocrinol 16:2793–2801 Haseba T, Duester G, Shimizu A, Yamamoto I, Kameyama K, Ohno Y 2006 In vivo contribution of Class III alcohol dehydrogenase (ADH3) to alcohol metabolism through activation by cytoplasmic solution hydrophobicity. Biochim Biophys Acta 762:276–283 Hayashi S, Watanabe J, Kawajiri K 1991 Genetic polymorphisms in the 5′-flanking region change transcriptional regulation of the human cytochrome P450IIE1 gene. J Biochem (Tokyo) 110:559–565 He L, Ronis MJ, Badger TM 2002 Ethanol induction of class I alcohol dehydrogenase expression in the rat occurs through alterations in CCAAT/enhancer binding proteins beta and gamma. J Biol Chem 277:43572–43577 Homann N, Stickel F, Konig IR et al 2006 Alcohol dehydrogenase 1C*1 allele is a genetic marker for alcohol-associated cancer in heavy drinkers. Int J Cancer 118:1998–2002 Lieber CS 2004 The discovery of the microsomal ethanol oxidizing system and its physiologic and pathologic role. Drug Metab Rev 36:511–529 Lagadic-Gossmann D, Lerche C, Rissel M et al 2000 The induction of the human hepatic CYP2E1 gene by interleukin 4 is transcriptional and regulated by protein kinase C. Cell Biol Toxicol 16:221–233 Maier H, Born IA, Veith S, Adler D, Seitz HK 1986 The effect of chronic ethanol consumption on salivary gland morphology and function in the rat. Alcohol Clin Exp Res 10: 425–427 Oesch-Bartlomowicz PR, Padma R, Becker B et al 1998 Differential modulation of CYP2E1 activity by cAMP-dependent protein kinase upon Ser129 replacement. Exp Cell Res 242:294–302 Orellana M, Rodrigo R, Valdes E 1998 Peroxisomal and microsomal fatty acid oxidation in liver of rats after chronic ethanol consumption. Gen Pharmacol 31:817–820 Pares X, Cederlund E, Moreno A, Hjelmqvist L, Jornvall H 1994 Mammalian class IV alcohol dehydrogenase (stomach alcohol dehydrogenase): structure, origin, and correlation with enzymology. Proc Natl Acad Sci USA 91:1893–1897 Porasuphatana S, Weaver J, Rosen GM 2006 Inducible nitric oxide synthase catalyzes ethanol oxidation to alpha-hydroxyethyl radical and acetaldehyde. Toxicology 223:167–174

16

CRABB & LIANGPUNSAKUL

Potter JJ, Rennie-Tankersley L, Mezey E 2003 Endotoxin enhances liver alcohol dehydrogenase by action through upstream stimulatory factor but not by nuclear factor-kappa B. J Biol Chem 278:4353–4357 Seitz HK, Egerer G, Simanowski UA et al 1993 Human gastric alcohol dehydrogenase activity: effect of age, sex, and alcoholism. Gut 34:1433–1437 Sherman DIN, Ward RJ, Warren-Perry M, Williams R, Peters TJ 1993 Association of restriction fragment length polymorphism in alcohol dehydrogenase 2 gene with alcohol-induced liver damage. Br Med J 307:1388–1390 Su JS, Tsai TF, Chang HM, Chao KM, Su TS, Tsai SF 2006 Distant HNF1 site as a master control for the human class I alcohol dehydrogenase gene expression. J Biol Chem 281:19809– 19821 Takahashi T, Lasker JM, Rosman AS, Lieber CS 1993 Induction of cytochrome P-4502E1 in the human liver by ethanol is caused by a corresponding increase in encoding messenger RNA. Hepatology 17:236–245 Tillonen J, Kaihovaara P, Jousimies-Somer H, Heine R, Salaspuro M 1998 Role of catalase in in vitro acetaldehyde formation by human colonic contents. Alcohol Clin Exp Res 22: 1113–1119 Tindberg N 2003 Phorbol ester induces CYP2E1 in astrocytes, through a protein kinase C- and tyrosine kinase-dependent mechanism. J Neurochem 86:888–895 Tindberg N, Ingelman-Sundberg M 1996 Expression, catalytic activity, and inducibility of cytochrome P450 2E1 (CYP2E1) in the rat central nervous system. J Neurochem 67: 2066–2073 Vasiliou V, Ziegler TL, Bludeau P, Petersen DR, Gonzalez FJ, Deitrich RA 2006 CYP2E1 and catalase influence ethanol sensitivity in the central nervous system. Pharmacogenet Genomics 16:51–58

DISCUSSION Deitrich: I was wondering about the protection against alcoholism given by the ADH variants. Yedi Israel has pointed out that we have been measuring steady state acetaldehyde an hour or two after alcohol has been given, but in UChA rats which have a defective ALDH2 enzyme, the major effect on acetaldehyde levels is seen in 30 min or less (Quintanilla et al 2005, Israel et al 2006). He proposes that it is the burst of acetaldehyde rather than the steady-state level which is preventing people from going ahead and drinking, and not the steady state levels. When were these acetaldehyde levels measured? It could be the burst rather than the steady state level that is important. Crabb: This thought occurred to us as we were looking at the data from the cells that have ALDH2 and ADH. It seems possible that during the first pass of alcohol through the liver, there would not be that restraining effect of acetaldehyde or alcohol oxidation, and a pre-steady-state burst of acetaldehyde might come out in the hepatic veins. I think we need someone to do the hepatic vein catheterizations as were done in Finland many years ago, to catch that early time point. M Salaspuro: The question as to why some ADH isoforms may protect from alcoholism is very interesting. We are used to working with hepatocytes, and know

ACETALDEHYDE GENERATING ENZYME SYSTEMS

17

very well how acetaldehyde formation and ethanol oxidation is regulated in the liver. But we don’t know much about how they are regulated in the mouth, especially in different populations. And we don’t know how either ADH and ALDH are expressed in various cell lines of the mouth mucosa. If acetaldehyde is released in the mouth some of its effects may be much more potent in the brain. Acetaldehyde for example may release histamine from the mast cells or mucosal cells and in this case it escapes the liver. Acetaldehyde may get to the CNS via the arterial tree. Crabb: I think you can say the same about the oesophagus. Acetaldehyde made there won’t pass through the liver, either. If the EST data are correct, acetaldehyde might even be formed beyond where we have been measuring it (i.e. in the hepatic veins) and closer to the brain—even in the arterial tree. Eriksson: On the other hand, there is work showing that when 4-methylpyrazole is used in normal conditions, there isn’t any effect on salivary acetaldehyde. This suggests that no measurable ADH-dependent levels are formed during normal conditions. If acetaldehyde is elevated, e.g. by deficient ALDH activity, then there is an effect of 4-methylpyrazole. I will speak more about this aspect in my paper. Apte: With regard to the local production of acetaldehyde, the pancreas is a bit of a forgotten organ in terms of its ability to produce acetaldehyde locally. It has been shown that the pancreas can metabolize alcohol. It has ADH. Interestingly, the kinetics of ADH in the acinar cells of the pancreas seems to match most closely to ADH5. It has a very high Km. In the cells I am interested in, the stellate cells which produce fibrosis, we think we have found ADH1. The problem I have with alcohol and acetaldehyde experiments is that when I read the literature I can’t work out whether people are using the concentrations of either ethanol or acetaldehyde that the cells may actually be exposed to in vivo. People use concentrations big enough to get an effect. This has always been a worry of mine: I’m concerned that in our own work we are using concentrations as high as 200 µM acetaldehyde, and we justify this by saying that local production during a burst of acetaldehyde might reach as high as that, and in the 30 min it persists for it has enough time to produce these toxic effects. Should we be looking more at steady-state levels? I also have a point regarding your table about ESTs. We have found CYP2E1 protein expression in the pancreas. Not only is it present, but it is also inducible in alcohol-fed rats. Crabb: I work with neurochemists, and they do interesting things such as in vitro microdialysis to get a sense of concentrations present at the pericellular level. I don’t know whether this has been done with the liver or other solid organs. If it has been, I don’t know whether our analytical methods are sensitive enough to detect acetaldehyde in those dialysates. We could do all sorts of things if we could get real time acetaldehyde concentrations. We need engineers and physicists to give us this kind of instrumentation.

18

CRABB & LIANGPUNSAKUL

Emery: Are there any people here who have experience of microdialysis in tissues? Apte: Lieber’s lab used to use GC-MS to measure acetaldehyde in hepatocytes in culture. Emery: The real question is what is happening in vivo. Eriksson: This is a relevant aspect. Acetaldehyde is formed at a specific site in the cell. At that site the concentration is extremely high. In the cytosol, nature has created aldehyde dehydrogenases with high Km. Thus, 200 µM is quite appropriate for studying the effects of acetaldehyde. Also, Asian flushers may have average hepatic concentrations this high during alcohol drinking. Apte: I’m pleased about this because I have always worried about the concentrations we’ve used. If we talk about alcohol in a general meeting, people might think about their own drinking and then be surprised about the concentrations that we use in vitro. Eriksson: I’d like to add a point. There’s a source of acetaldehyde that is forgotten, which comes with drinking the alcoholic beverage. Some beverages contain millimolar concentrations of acetaldehyde. We are talking here about micromolar concentrations, so it means that this is a significant source. Systemically, the body has a fantastic capacity to remove acetaldehyde. Also, in liver the efficacy of removal is striking: more than 99% is removed directly. Emery: I think we can be comfortable with these sorts of concentrations extracellularly. What is the concentration going to reach within the cell, and within particular compartments where the effects occur? This comes back to the question of the local production within the cell and organelle. Niemelä: With regard to the possible associations between the expression of alcohol-metabolizing enzymes and organ damage in alcoholics, recent studies have shown that overexpression of CYP2E1 creates damage in the liver (Caro & Cederbaum 2004). Is there any evidence linking ADH expression with end-organ damage or certain ethanol-induced disease states? Crabb: Even more broadly than that, many years ago Christopher Day at the University of Newcastle upon Tyne and I were looking at these gene variants as risk factors for alcohol pathology. The literature contains a number of papers reporting such associations. Dr Day has continued to work in this area and he concludes that so far the only genetic risk factors for alcoholic liver disease are the ALDH2 *2 allele in Asians, and a polymorphism in the TNFα promoter, even though we have numerous publications from various groups (Stewart et al 2001). There are no other findings that have been reproduced in other laboratories and with other populations. He kids that I did well to stop working in this area because he has been frustrated by the inability to identify stronger correlations. He has one of the best databases of patients who have been well-characterized for alcoholic liver disease without hepatitis C. I believe one comes close to having a

ACETALDEHYDE GENERATING ENZYME SYSTEMS

19

genetic risk factor if you combine several studies from Japan. It seems that ALDH2deficient individuals or β2 expressing individuals had somewhat higher risk of liver injury if they persisted in drinking heavily. This fits with our preconceptions of the role of acetaldehyde in liver injury. If you look at this from the other direction, the Consortium on Genetics of Alcoholism (COGA) have been working for 20 years, and only in the last couple of years a number of candidate genes came out of this study, which contribute to the risk of alcoholism. A couple of them are neurotransmitter receptors (which is kind of satisfying), one is a taste receptor for bitter substances, and the final one is ADH4, which is the gene for πADH. Because of its high Km and predominant expression in the liver, this was not thought to be a candidate gene before COGA, and it didn’t have coding region variants that we thought were important. Coming at it from this non-biased genome-wide screen, an enzyme of alcohol metabolism popped up that we hadn’t expected. Our next step in the field is to genotype people for ADH4 with these risk haplotypes, and then see what the difference is in their alcohol and acetaldehyde metabolism. Worrall: In your table of ESTs you had ‘brain’. Of course, the various regions of the brain vary in their metabolic capacity. Conventional wisdom is that brain has two to three orders less ethanol metabolic capacity than the liver. We have been measuring some adducts in alcoholic cerebellar degeneration. We are finding a lot more than one would expect. Is there another metabolizing system lurking in the brain that we are not seeing? Crabb: The compiled data in those Unigene sets will just say ‘brain’ without any further anatomical division. I don’t know whether it was libraries made from different regions all combined together, or whether it was whole brain. The first thing would be to go back and find existing data on an area of interest, or test for it in a specific area by measuring protein or RNA abundance. I think that Dr Deitrich is more able to answer the question about whether there are non-ADH generators of acetaldehyde than I am, because of what he published on catalase. Deitrich: We found that brain tissue will oxidise alcohol to acetaldehyde primarily through catalase, with some contribution from CYP2E1, but there’s still something that we can’t account for (Zimatkin et al 1999, 2001a,b, 2006, Zimatkin & Deitrich 1997, Gill et al 1992). I don’t know whether it is our deficiency in being able to measure those enzymes, or whether there is some other oxidative enzyme present. Person from the University of Washington thinks that there is another mechanism that is oxygen dependent which catalyses the formation of acetaldehyde from ethanol (Person et al 2000). It’s true that it’s a tiny amount compared with the liver, but it may be critical to the brain. Thornalley: I want to comment on the in situ reactivity of acetaldehyde in in vivo and in vitro experiments. In vitro experiment design can actually lead to studying acetaldehyde in a more activated state because a factor influential on aldehyde

20

CRABB & LIANGPUNSAKUL

reactivity is binding to thiols in albumin. In most cell culture systems there is 10% serum, not 100% plasma as in vivo. The aldehyde has a lower binding to thiols extracellularly and is thereby activated. This is very important in the case of dicarbonyls. The in situ reactivity of dicarbonyls is increased 10-fold in medium with 10% serum in vitro compared to 100% plasma in vivo. In vivo these aldehydes are mostly reversibly bound to cysteine 34 in albumin. Albano: Returning to the issue concerning the genetic factors that might influence the susceptibility to alcoholic liver diseases, Zintzaras et al (2006) have recently published a meta-analysis showing that none of the polymorphisms of alcohol or acetaldehyde dehydrogenases-coding genes increase the risk of developing alcoholic cirrhosis. Also, in the case of CYP2E1, none of the polymorphisms so far characterized has been shown to have any influence on the risk of progression of alcoholic liver disease or an appreciable influence on ethanol metabolism (Zintzaras et al 2006, Hu et al 1997). Nonetheless, we have to consider that in humans there are phenotypic variances in the expression of CYP2E1. In a study performed together with the late Francois Mènez in a group of heavy drinkers from Brest (France) the functional evaluation of CYP2E1 activity by the chloroxazone oxidation test has shown that in about 20% of alcohol consumers CYP2E1 activity was not induced, despite high alcohol intake. In these subjects the formation of the hydroxyethyl radical was also significantly lower than in the patients with induced CYP2E1, and comparable to non-drinking controls (Dupont et al 1998). The presence of inter-individual differences in the inducibility of CYP2E1 has been subsequently confirmed by a study performed in Professor Seitz’ laboratory (Oneta et al 2002). The genetic basis of such phenotypic variability has not yet been established, but it is not dependent upon any of the SNPs so far detected in the CYP3E1 gene (Hu et al 1997). Its characterization might be important to provide more insight on genetic factors influencing alcohol toxicity because the low-inducible CYP2E1 phenotype is prevalent among the subjects with less severe liver disease. Seitz: We don’t see differences in genetics for liver disease and ADHs, but we may see differences with respect to other diseases such as cancer. Quite a number of studies did not show cancer in ADH1C homozygotes, but there are other studies including our own which showed an increased risk of certain cancers in ADH1C homozygotes, but not of liver disease. Shukla: You raised an interesting point about the differences between human liver cell lines versus the rat in terms of ADH activity. Some of the human cell lines have very poor ADH activity. One of the issues that comes to my mind is whether G2 cells also have similarly low ADH activity. Is this due to the nature of the transformed cell line? If one considers the normal human liver cells, is the metabolic capacity the same as the transformed human liver cell line? There may be an important difference here.

ACETALDEHYDE GENERATING ENZYME SYSTEMS

21

Crabb: I think that is the case. The class I ADH expression is just as much a differentiated phenotype of a liver cell as making albumin or having these other functions. In cultured primary rat hepatocytes, ADH is lost fairly soon. If you want to use liver cell lines you need to test them to make sure they have the alcoholmetabolizing capacity you hope they do, or at least acknowledge the difference. Until we had measured alcohol metabolism and acetaldehyde generation in the experiments that I showed, we didn’t really know that much about the cells we had been working on for quite a long time. By learning that HeLa cells seem to be making acetaldehyde, we now have to double check what we are thinking when we say that if a certain phenomenon occurs in hepatoma cells but doesn’t occur in HeLa cells, that the explanation is that the phenomenon must be due to alcohol metabolism. We may have been misled. Eriksson: It is difficult to extrapolate to in vivo when we go to purer in vitro systems. You can see the difficulty. Based on enzyme kinetics in isolated pure systems the hepatic ratio between acetaldehyde and alcohol is surprisingly high to enable efficient alcohol oxidation. If you were to have a pure isolated system with only the ADH present there would be a shift in the oxidation/reduction equilibrium reaction towards alcohol at current in vivo conditions. Yet in vivo there is an alcohol oxidation taking place. This means that acetaldehyde is not in the same form in vivo and in vitro. The in vivo system is far more complicated.

References Caro AA, Cederbaum AI 2004 Oxidative stress, toxicology, and pharmacology of CYP2E1. Annu Rev Pharmacol Toxicol 44:27–42 Dupont I, Lucas D, Clot P, Ménez C, Albano E 1998 Cytochrome P4502E1 inducibility and hydroxyethyl radical formation among alcoholics. J Hepatol 28:564–571 Gill K, Menez JF, Lucas D, Deitrich RA 1992 Enzymatic production of acetaldehyde from ethanol in rat brain tissue. Alc Clin Exp Res 16:910–915 Hu Y, Oscarson M, Johansson I et al 1997 Genetic polymorphism of human CYP2E1: characterization of two variant alleles. Mol Pharmacol 51:370–376 Israel Y, Quintanilla ME, Sapag A, Tampier L 2006 Autosomal and maternal genes influence alcohol intake in alcohol drinker and nondrinker rat lines: role of the ‘acetaldehyde burst’. Alcohol Clin Exp Res 30:276A Oneta CM, Lieber CS, Li J et al 2002 Dynamics of cytochrome P4502E1 activity in man: induction by ethanol and disappearance during withdrawal phase. J Hepatology 36:47–52 Person RE, Chen H, Fantel AG, Juchau MR 2000 Enzymic catalysis of the accumulation of acetaldehyde from ethanol in human prenatal cephalic tissues: Evaluation of the relative contributions of CYP2E1, alcohol dehydrogenase, and catalase/peroxidases. Alcohol Clin Exp Res 24:1433–1442 Quintanilla ME, Tampier L, Sapag A, Israel Y 2005 Polymorphisms in the mitochondrial aldehyde dehydrogenase gene (Aldh2) determine peak blood acetaldehyde levels and voluntary ethanol consumption in rats. Pharmacogenet Genomics 15:427–431 Stewart S, Jones D, Day CP 2001 Alcoholic liver disease: new insights into mechanisms and preventative strategies. Trends Mol Med 7:408–413

22

CRABB & LIANGPUNSAKUL

Zimatkin SM, Deitrich RA 1997 Ethanol metabolism in the brain. Addict Biol 2:387–392 Zimatkin SM, Liopo AV, Deitrich RA 1999 Oxidation of ethanol to acetaldehyde in brain and the possible behavioral consequences. Adv Exp Med Biol 463:231–236 Zimatkin SM, Liopo AV, Satanovskaya VI, Bardina LR, Deitrich RA 2001a Relationship of brain ethanol metabolism to the hypnotic effect of ethanol. II: Studies in selectively bred rats and mice. Alcohol Clin Exp Res 25:982–988 Zimatkin SM, Liopo AV, Slychenkov VS, Deitrich RA 2001b Relationship of brain ethanol metabolism to the hypnotic effect of ethanol. I: Studies in outbred animals. Alcohol Clin Exp Res 25:976–981 Zimatkin S, Pronko SP, Vasiliou V, Gonzalez FJ, Deitrich RA 2006 Enzymatic mechanisms of ethanol oxidation in the brain. Alcohol Clin Exp Res 30:1500–1505 Zintzaras E, Stefanidis I, Santos M, Vidal F 2006 Do alcohol-metabolizing enzyme gene polymorphisms increase the risk of alcoholism and alcoholic liver disease? Hepatology 43:352–361

Removal of acetaldehyde from the body Richard A. Deitrich, Dennis Petersen* and Vasilis Vasiliou* Department of Pharmacology, University of Colorado School of Medicine, University of Colorado, HSC, PO Box 6511, Mail Stop 8303, Aurora, CO 80045 and *Department of Pharmaceutical Sciences, University of Colorado School of Pharmacy, 4200 E. 9th Ave, C238 Denver, CO 80262, USA Abstract. The reduction of acetaldehyde back to ethanol via NAD-linked alcohol dehydrogenase is an important mechanism for keeping acetaldehyde levels low following ethanol ingestion. However, this does not remove acetaldehyde from the body, but just delays its eventual removal. Acetaldehyde is removed from the body primarily by oxidation to acetate via a number of NAD-linked aldehyde dehydrogenase (ALDH) enzymes. There are nineteen known ALDHs in humans, but only a few of them appear to be involved in acetaldehyde oxidation. There are many analogous enzymes in other organisms. Genetic polymorphisms of several ALDHs have been identified in humans that are responsible for several hereditary defects in the metabolism of normal endogenous substrates. The best known ALDH genetic polymorphism is in ALDH2 gene, which encodes a mitochondrial enzyme primarily responsible for the oxidation of the ethanol-derived acetaldehyde. This common polymorphism is known to dominantly inhibit its enzymatic activity resulting in reduced ability to clear acetaldehyde in both homozygote and heterozygote individuals. These individuals are generally protected against alcohol abuse but are susceptible to oesophageal cancer. For those enzymes that are capable of reacting with acetaldehyde, they may do so at the expense of their normal substrates, resulting in abnormal accumulation of these substrates. Examples of this are the aldehydes of the biogenic amines, dopamine, noradrenaline, adrenaline, serotonin and long chain lipid aldehydes such as nonenal. Not all of these enzymes are capable of efficient oxidation of acetaldehyde; however, it is possible that acetaldehyde may function as an inhibitor of these enzymes as well. The aldehydes whose metabolism is interfered with may also serve as inhibitors of ALDHs as well. However, this aspect of aldehyde function has not been extensively studied. A number of other mechanisms for the removal of acetaldehyde exist. For example, reaction of acetaldehyde with protein or nucleic acids is responsible for the disappearance of a small amount of acetaldehyde, but may be responsible for some pathological effects of acetaldehyde. There are a few other enzymes such as aldehyde oxidase, xanthine oxidase, cytochrome P450 oxidase and glyceraldehyde-3-phosphate dehydrogenase that are capable of oxidizing acetaldehyde. However, these enzymes account for only a small fraction of the total activity. 2007 Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Foundation Symposium 285) p 23–51

Removal of acetaldehyde is essential to the metabolism of ethanol and critical to the survival of the animal. Ethanol could not be metabolized unless acetaldehyde 23

24

DEITRICH ET AL

was removed since the equilibrium constant is far in the direction of ethanol. Survival of the organism following ethanol intake depends on the removal of highly toxic acetaldehyde, either by reduction back to ethanol, oxidation to acetate, or removal via respiration, kidney excretion or protein binding. Acetaldehyde removal by reduction to ethanol: importance to toxicity of acetaldehyde While it is normally considered as the major producer of acetaldehyde from ethanol, the enzyme alcohol dehydrogenase can also serve an important function in limiting the accumulation of acetaldehyde since the equilibrium constant for ethanol to acetaldehyde is far in the direction of ethanol (Backlin 1958). In fact the ethanol concentration will be 5000 times greater than that of acetaldehyde in tissues that contain alcohol dehydrogenase, even if no acetaldehyde is removed by other means. This limits the amount of acetaldehyde present in those individuals with a deficiency of aldehyde dehydrogenase or in those who have ingested aldehyde dehydrogenase inhibitors. Acetaldehyde oxidation Aldehyde dehydrogenases (ALDHs) Mitochondrial ALDH (ALDH2). Early papers dealing with ALDH used various preparations of whole tissues such as acetone powders (Racker 1949), or whole tissue homogenates as well as a variety of substrates. These studies most likely dealt primarily with the mitochondrial enzyme since that enzyme is very active and has a low Km value. Later studies separated the enzymes on the basis of their tissue subcellular localization and their kinetic properties. It was not until the advent of molecular genetics that it became clear just how many ALDH enzymes there are (Sophos & Vasiliou 2003, Vasiliou & Nebert 2005, Vasiliou et al 1995). Early studies surveyed various tissues for NAD-dependent aldehyde oxidizing capacity utilizing indole-3-aldehyde as substrate because a sensitive assay could be developed by measuring the fluorescence of indole-3-acetic acid (Deitrich 1966). ALDH activity was observed in rat, rabbit, monkey and dog liver, lung and kidneys. ALDH activity was also found in the liver of other species such as cow, sheep and pig. The major work was in the rat where ALDH activity was observed in whole homogenates of liver, kidney, gonads, adrenal, brain, small intestine, heart, uterus and adipose tissue. Liver mitochondrial, cytosolic and microsomal fractions were found to have substantial ALDH activity; however, no enzymatic activity was observed in blood, perhaps because of a large blank value. An older review of the role of ALDHs in the removal of acetaldehyde from the body is available (Weiner 1979). The conclusion from this and other studies

ACETALDEHYDE REMOVAL

25

•LIVER ETHANOL 30,000 nmoles/gm

• ACETALDEHYDE • 200 nmoles/gm

(Eriksson, 1977)

FIG. 1. Liver metabolism of ethanol.

(Eriksson 1977, Eriksson & Sippel 1977) as well as a more recent review (Crabb et al 2004) is that the liver is primarily responsible for both generation and elimination of acetaldehyde. Only a small percentage of the acetaldehyde that is generated in the liver escapes into the blood. The mitochondrial ALDH enzyme has a very low Km value, of the order of 1 µM and, given the large amount of the enzyme, removes most of the acetaldehyde before it reaches the blood. However, this also demonstrates that the rate of production of acetaldehyde by the liver exceeds the ability of that organ to completely remove it (Fig. 1). The presence of the enzyme in liver, the site of the major production of acetaldehyde from ethanol, is protective of this and other organs of the body since relatively little acetaldehyde escapes. The presence of the enzyme in other tissues serves not only to protect these tissues from acetaldehyde borne by the blood or produced locally, but also to protect the tissues from other biogenic or exogenous aldehydes as well since the enzyme has broad substrate specificity (Allali-Hassani & Weiner 2001, Deitrich et al 1962, Petersen et al 1991, Wroczynski & Wierzchowski 2000, Yin et al 1995). On the other hand, acetaldehyde will compete with these other aldehydes (such as those from the biogenic amines, catechols amines and serotonin) for ALDH2 (Hellstrom & Tottmar 1982, MacKerell et al 1986, Nilsson 1988, Von Wartburg et al 1975). This inhibition may have behavioural consequences related to the actions of ethanol (Deitrich & Erwin 1980, 1975). Other endogenous substrates such as 4-hydroxynonenal also function as inhibitors of the enzyme (Hartley et al 1995a,b, Hartley & Petersen 1997, Honzatko et al 2005, Luckey et al 1999, Mitchell & Petersen 1991, Nguyen & Picklo 2003). Later studies showed that ALDH2 is located in the mitochondrial matrix (Tottmar et al 1973, Deitrich & Siew 1974, Siew & Deitrich 1976). Thus acetaldehyde, produced outside the mitochondria by alcohol dehydrogenase, cytochrome P450 or catalase must diffuse into the mitochondria in order to be oxidized by this enzyme. While the liver has the highest concentration of this enzyme in most animals, it exists in other tissues as well (reviewed in Agarwal et al 1989, Berkovitz et al 2001,

26

DEITRICH ET AL

Braun et al 1987, Harada et al 1980, Sladek 2002), including brain (Erwin & Deitrich 1966). A deficiency of ALDH2 in Asian populations is protective against development of alcoholism (Goedde et al 1992, Harada et al 1999, Xiao et al 1996, Yamamoto et al 2000) and perhaps carotid atherosclerosis (Narita et al 2003). However, this deficiency also has potentially severe consequences for those individuals who continue to consume ethanol and possibly in those that do not consume ethanol at all. There is a comprehensive analysis of the relationship between ADH1 and ALDH2 genotypes and head and neck cancer (Brennan et al 2004). The conclusion is that there is an increased risk factor for ALDH2-deficient individuals for head and neck cancer. There are also reports of increased incidence of cancer of the gastrointestinal tract (reviewed in Salaspuro 2003) in hepatocellular carcinoma (Kato et al 2003, Sakamoto et al 2005), oesophageal cancer (Morita et al 2005, Watanabe et al 2002) but not in colorectal adenomas (Hirose et al 2005). There is enhanced oxidative stress (Ohsawa et al 2003) perhaps related to late onset Alzheimer’s disease (Ohta et al 2004) although this is disputed (Kim et al 2004, Shin et al 2005). There has been a meta analysis of the relationship of inactive ALDH2 genotype with maternal inheritance of diabetes (Murata et al 2004). The conclusion was that there is an association between ALDH2 inactivity and maternal inheritance. The authors proposed an interesting hypothesis to explain this finding. ‘ALDH2 is the key enzyme for detoxifying the various aldehydes such as, not only acetaldehyde, but also 4-hydroxy-2-nonenal (4-HNE). 4-HNE is the by-product of lipid peroxidation. Therefore, in a chronic situation of diabetes (in the mother), increased tissue levels of 4-HNE could result from inactive ALDH2. Because 4-HNE reacts preferentially with sulfhydryl (SH)-containing molecules, the mitochondrial inner membrane is easily adducted with 4-HNE. The mitochondrial inner membrane adducted by 4-HNE alters the electron transfer chain system, which may increase free radicals production, leading to the production of mitochondrial DNA damage. The acquired mitochondrial DNA damages can be inherited exclusively from mother to offspring, and in the offspring, the genetic handicap may lead to the maternal inheritance of diabetes.’ This study points out that ALDH2 is normally responsible for the oxidation of many endogenous aldehydes. If the enzyme activity is compromised either by the presence of inhibitors or for genetic reasons, these aldehydes must be dealt with in other ways, such as reduction to the alcohol, oxidation by other ALDH enzymes or by reaction with proteins and nucleic acids. It is the latter possibility that may be responsible for damage even in the absence of exogenous acetaldehyde. The availability of Aldh2 knockout mice (Isse et al 2002) should help us to work out the mechanisms of these effects. It is known that mitochondria are damaged by chronic ethanol ingestion (Lieber 1997). Inhibition of ALDH in mitochondria could be accomplished by ‘mutually

ACETALDEHYDE REMOVAL

27 ACETATE

ACETALDEHYDE ALDH NONENAL

NONEONIC ACID

FIG. 2. Inhibition of ALDH in mitochondria can be accomplished by ‘mutually assured destruction’ by 4-HNE and acetaldehyde since they are competitive substrates and inhibit each other’s oxidation by ALDH.

assured destruction’ by 4-HNE and acetaldehyde (Fig. 2) since they are competitive substrates and inhibit each other’s oxidation by ALDH. In addition 4-HNE is a potent inhibitor as well as a substrate for ALDH2 (Doorn et al 2006). In spite of the studies on ALDH2 and the relationship between acetaldehyde and other endogenous aldehydes, studies have been largely confined to those enzymes where acetaldehyde is known to be a substrate. While a great deal is now known about the myriad of ALDH enzymes, relatively little is known about the effect of acetaldehyde on those ALDH enzymes for which acetaldehyde is not a substrate or a very poor substrate because of very high Km values. For many of these enzymes we do not know if acetaldehyde functions as an inhibitor since it is almost never studied in this light. Of course, it is necessary to measure the acid product concentration of the aldehyde substrate, not just the NAD(P)H produced when both aldehydes are added to the reaction mixture. Acetaldehyde certainly has the structure to function as a competitive inhibitor at the active site of ALDHs, even if it, for some reason, is not a substrate. While one would expect that the Km of acetaldehyde as a substrate would be the same as the Ki for acetaldehyde when it functions as a competitive substrate, this also has not been studied carefully. This consideration may be important in our understanding not only of the pathways of acetaldehyde removal, but also of the many effects that acetaldehyde has via the surrogate aldehydes allowed to escape metabolism by their preferred enzyme. This would result in a ‘multiplying’ or ‘amplification’ effect of the relatively small amount of acetaldehyde present. Other ALDHs There are several sources of information concerning the numerous ALDH enzymes in humans and other organisms (Vasiliou et al 1995, 1999, Vasiliou & Nebert 2005). In the tables that follow we have largely restricted the material to human ALDH enzymes with some reference to rat and mouse enzymes. It can be seen in Tables 1 and 2 that the myriad of aldehyde dehydrogenase enzymes outside of ALDH2 are unlikely to play a significant role in the direct removal of acetaldehyde from the body. However, they may play a role in the toxic effects

28

TABLE 1 ALDH enzymes for which acetaldehyde (or propionaldehyde) has been reported to be a substrate Approved gene symbol

Trivial name

Km/Ki (mM) acetaldehyde

Endogenous substrate(s) Km (mM)

Exogenous substrates Km (mM)

Notes. Consequences of inhibition

ALDH1A1

ALDH1

50, 118

Retinal, 4-HNE

Aldophosphamide 2; 7-methoxy-1napthaldehyde; 0.85; 6-methoxy2-napthaldehyde; <0.3

Cytosolic, and red blood cells. Deficiency of retinoic acid

ALDH1A3

RALDH3

220

All-trans-retinal Km = 200

Propionaldehyde 324

ALDH1A7

ND

Propionaldehyde 1400 Phenylacetaldehyde ND Propionaldehyde 30 Propionaldehyde 636

Referencea

(Saari et al 1995, Vasiliou et al 2004, Sladek 1999, Sugata et al 1988, Ueshima et al 1993, Wierzchowski et al 1997) Phenobarbital (Dunn et al 1989, induced. Deficiency Deitrich et al 1972, of retinoic acid 1977a, Vasiliou et al 2004) Phenobarbital induced. (Lindahl & Evces Deficiency of 1984a,b) retinoic acid

ALDH5 TFDH

ND ND

Acetaldehyde? 10formyltetrahydrofolate Km = 5.5

Mitochondrial Cytosol. Deficiency of folate

ALDH2

ALDH2

1, 0.6

Acetaldehyde, biogenic aldehydes, 4-HNE

28

ALDH2

4-HNE

0.5 Ki

Mitochondria. Flushing, protection from alcoholism. Parkinson’s? Mouse liver

ALDH2

GSH-4-oxnon-2-enal

0.18

Human

(Stewart et al 1995) (Cook et al 1991, Vasiliou et al 2004, Krupenko et al 1995, 1997) (Meyer et al 2004)

(Hartley & Petersen 1997, Mitchell & Petersen 1991) (Doorn et al 2006)

DEITRICH ET AL

ALDH1B1 ALDH1L1

ALDH3A1 ALDH3A1

ALDH3A1 ALDH3A1 ALDH3A2

ALDH3A2 ALDH4A1

Propionaldehyde Benzaldehyde Propionaldehyde Benzaldehyde 4-HNE

150 1600 150 50000 Aldophosphamide 526

Rat Liver I Rat liver I Rat liver II Rat liver II TCDD induced, tumour associated. Cornea

Rat liver cytosol

(Deitrich et al 1977)

Rat liver cytosol

(Lindahl & Evces 1984a,b) (Kelson et al 1997, Vasiliou et al 2004)

Leukotriene B4

Propionaldehyde 7034 Hexanal 4736 Octanal 18 Decanal 0.5 Dodecanal 0.8 4-Hydroxynonenal 0.5 Malondialdehyde 45 Benzaldehyde 6550 20 910 4100 500 Dihydrophytal 6 Benzaldehyde 1440 Crotonaldehyde 800 Glutaraldehyde 1700 Phytol

(Deitrich et al 1977, Sladek 1999, Vasiliou et al 2004) (Pappa et al 2003)

Glutamate γsemialdehyde Km = 170, 100; succinic semialdehyde Km = 30, 10

Propionaldehyde 24000, 9400; 3nitrobenzaldehyde 1300; Octanal 4000

ALDH3 TCDD inducible TCDD inducible

2600, 7095 200

Nonenal Km = 1 4-HNE Km = 45

TCDD inducible TCDD inducible FALDH ALDH10

2200

Phenylacetaldehyde Propionaldehyde Propionaldehyde Benzaldehyde Fatty aldehydes C = 6–24: Km = 18–50 µM

FALDH ALDH10 ALDH4

870 2500

22000, 5000

Microsomal

Microsomal Mitochondrial matrix. Hyperprolinemia, induced by p53

(Lindahl & Evces 1984a,b)

29

(van den Brink et al 2005) (Brunner & Neupert 1969, Farres et al 1998, ForteMcRobbie & Pietrusko 1989, Valle et al 1979, Vasiliou et al 2004, Yoon et al 2004)

ACETALDEHYDE REMOVAL

ALDH2

30

TABLE 1 (continued) Approved gene symbol

Trivial name

Km/Ki (mM) acetaldehyde

Endogenous substrate(s) Km (mM)

Exogenous substrates Km (mM)

ALDH5A1

SSDHVa

330

Aldophosphamide 560

ALDH7A1

ATQ1

2000 (grass carp, sea bream)

Succinicsemialdehyde, γaminobutyraldehyde, 4-HNE δ1-piperideine-6carboxylate/αaminoadipicsemialdehyde.

Propionaldehyde 1450, benzaldehyde 770, (grass carp, sea bream)

Pyridoxine dependent seizures: Meniere disease

ALDH8A1

RALDH4

9.6

9-cis-retinal Km = 2.3

Deficiency of retinoic acid

ALDH9A1

ALDH9

50

betaine aldehyde, γaminobutyraldehyde

Benzaldehyde Decanal Succinic semialdehyde Hexanal Octanal Propanal Glutaraldehyde Acetaldehyde Propionaldehyde

Referencea (Sladek 1999, Testore et al 1999, Vasiliou et al 2004)

Not all references are included. ND indicates that the compound has been reported to be a substrate, but no kinetics were reported.

(Chan et al 2003, Lynth & Cameron 2002, Mills et al 2006, Skvorak et al 1997, Tang et al 2002) (Lin et al 2003, Lin & Napoli 2000)

(Hjelmqvist et al 2003), (Vasiliou et al 2004)

DEITRICH ET AL

a

Notes. Consequences of inhibition

Approved gene symbol ALDH1A2

Trivial name

Endogenous substrate(s) Km (mM)

Exogenous substrates Km (mM)

Notes/consequences of inhibition

Referencea

Retinal, Km = 0.7

Cytosolic

(Lee et al 1991)

Dopaldehyde?

Cytosolic Schizophrenia? Microsomal, parotid gland

(Lee et al 1991, Sun et al 2005) (Hsu et al 1997, Hsu & Chang 1996, Yoshida et al 1998) (Goodwin et al 1999, Vasiliou et al 2004)

ALDH3B1

RALDH2, human ALDH11 ALDH7

ALDH3B2

ALDH8

ALDH6A1

MMSDH

Malonate- and, methylmalonatesemialdehyde

CoA dependent. Mitochondria; methylmalonic acidemia

ALDH16A1 ALDH18A1

P5CS

Glutamate (enzyme functions as reductase in synthesis of proline and arginine)

Mitochondrial matrix. Hyperammonemia, mental retardation

a

ACETALDEHYDE REMOVAL

TABLE 2 ALDH enzymes for which acetaldehyde has not been reported to be a substrate, or its Km not determined

Vasiliou & Nebert 2005 (Baumgartner et al 2005, Hu et al 1999, 1992)

Not all references are included.

31

32

DEITRICH ET AL

of acetaldehyde if they are inhibited, either directly via substrate competitive inhibition by acetaldehyde, or by some surrogate aldehyde whose oxidation by ALDH2 has been inhibited by acetaldehyde. Perhaps the prime example is 4-HNE. The cometabolism of 4-HNE and ethanol-derived acetaldehyde by ALDH2 could have a significant impact on the hepatic removal of acetaldehyde. This proposition is supported by the similar apparent Km values of ALDH2 for 4-HNE and acetaldehyde which are approximately 14 and 1.4 µM, respectively. The relative affinity of these substrates for this enzyme are apparent from detailed kinetic studies using purified rat liver ALDH2 which demonstrate that 4-HNE is a competitive or mixed-type inhibitor of acetaldehyde oxidation with a calculated Ki of 0.50 µM (Mitchell & Petersen 1991). Thus, assuming that the steady-state level of 4-HNE in the liver reaches 0.50 µM, one can predict that acetaldehyde oxidation rates by ALDH2 would be decreased by 50%. Although it is difficult to calculate the steady state levels of 4-HNE in the liver, estimates have appeared in the literature indicating the concentrations of this aldehyde produced during lipid peroxidation may approach 10 µM. Therefore, it is probable that the hepatic removal of ethanol-derived acetaldehyde is compromised by the co-metabolism of 4-HNE as well as other biogenic aldehydes. A wide variety of metabolic imbalances can result from such inhibition, should it occur. The oxidation of retinaldehyde by a variety of ALDH enzymes is of importance in the production of retinoic acid, which is vital to cell growth and differentiation. The oxidation of numerous toxic aldehydes in the environment could be inhibited. The results of deficiency of various ALDH enzymes can be seen when there is a genetic polymorphism that decreases or eliminates an enzyme. The most well known is the deficiency of ALDH2 among some Asian populations. However, there are several other, less widespread polymorphisms as well. Thus deficiency of ALDH4A1 (ALDH4) leads to hyperprolinaemia, ALDH6A1 deficiency is responsible for methylmalonic acidaemia, deficiency of ALDH7A1 is responsible for pyridoxine-dependent seizures (Meniere disease), and hyperammonaemia and resultant mental retardation is caused by a deficiency of ALDH18A1. It is unknown whether any symptoms of any of these conditions are present in individuals that consume large amounts of ethanol. Of more interest perhaps would be to determine if any of these conditions exist in ALDH2 negative individuals or in those taking ALDH inhibitors.

Other acetaldehyde metabolizing enzymes Glyceraldehyde-3-phosphate dehydrogenase Ryzlak & Pietruszko (1989) reported on the activity of a number of glyceraldehyde3-phosphate (G3P) dehydrogenases in brain. The cytosolic enzyme was inactive with acetaldehyde as a substrate but isozymes in the mitochondria were active.

ACETALDEHYDE REMOVAL

33

The Km varied from 300 to 2000 µM for acetaldehyde. Earlier, unpublished observations indicated that G3P dehydrogenase from Ehrlich ascites tumour cells was capable of oxidizing propionaldehyde (Deitrich & Hellerman 1959). Oxidases Aldehyde and xanthine oxidases. Aldehyde and xanthine oxidases are molybdenumcontaining enzymes that are capable of oxidizing acetaldehyde and, in the process, produce highly reactive oxygen species that have been implicated in hepatic lipid peroxidation (Shaw & Jayatilleke 1990b). The Km for acetaldehyde oxidation by xanthine oxidase is greater than 30 mM (Fridovich 1966), while the Km with aldehyde oxidase is about 1 mM (Rajagopalan & Handler 1964). An interesting cycle capable of producing reactive oxygen species was proposed by Mira et al (1995). It was observed that aldehyde oxidase not only oxidized acetaldehyde but also NADH and produced superoxide. Thus alcohol (and acetaldehyde) dehydrogenases would produce NADH which would be oxidized to NAD, thus providing more NAD for ethanol (and acetaldehyde) oxidation. This points out another interesting and recent observation, that a polymorphism in complex I of mitochondria from UChA rats, selectively bred for ethanol avoidance in a free choice situation, is partially responsible for decreased rates of acetaldehyde metabolism in the liver mitochondria of these rats (Quintanilla et al 2005). The mechanism is that there is less efficient reoxidation of NADH to NAD by the defective electron transport chain in the mitochondria. When the availability of NAD becomes the rate-limiting factor in acetaldehyde oxidation, more acetaldehyde escapes. The generation of reactive oxygen species via oxidation of aldehyde by aldehyde oxidase and xanthine oxidase has been implicated in a number of pathological conditions including oxidative stress in the heart (Oei et al 1986), lipid peroxidation (Kato et al 1990, Kera et al 1988) and folate cleavage and release of iron (Shaw et al 1989, Shaw & Jayatilleke 1990a). In addition the metabolism of aldehydes has been implicated in mammary (Castro et al 2001a) and prostate (Castro et al 2001b) cancer as well as alcoholic pancreatitis (Nordback et al 1991). Possible genetic tools for study of these enzymes were found when it was observed that DBA/2 mice are selectively deficient in expression of aldehyde oxidase homologues 1 and 2 (Vila et al 2004). Aldehyde oxidase has been found to oxidize retinaldehyde to retinoic acid as well as the numerous ALDH enzymes that will carry out this reaction (Ambroziak et al 1999, Lee et al 1991). Cytochrome P450 While it has been reported that the ethanol-inducible CYP2E1 can also utilize acetaldehyde as a substrate (Terelius et al 1991), it is unlikely that this would occur in the face of a high ethanol concentration. However, the oxidation of other

34

DEITRICH ET AL

aldehydes by this enzyme, in the absence of ethanol should be considered. The whole area of free radicals as mediators of alcohol toxicity was reviewed by Mantle & Preedy (1999). Protein binding As with the oxidases and cytochrome P450 oxidations, the amount of acetaldehyde removed by protein or nucleic acid binding is small compared to the total, but the effects of such binding can be devastating. Binding to proteins can lead to the formation of antibodies to the modified protein (Hoffmann et al 1993, Kervinen et al 1991, Thiele et al 1998, Willis et al 2002). Likewise, the binding of acetaldehyde or surrogate aldehydes to nucleic acids can have severe consequences (Becker et al 1996, Inagaki et al 2003, Kuykendall & Bogdanffy 1992). Concluding remarks While the ‘king of the hill’ in removal of acetaldehyde from the body remains ALDH2, the complexity of acetaldehyde interactions with other ALDH and aldehyde oxidizing enzymes is now apparent. Once relegated to a minor role in the toxicity of ethanol ingestion, acetaldehyde has achieved the dubious distinction of being a force to be dealt with in the overall assessment of ethanol damage. Acknowledgements Supported by grants (R01AA- 11464, AA09300 and AA11885) from the National Institute on Alcohol Abuse and Alcoholism of the National Institutes of Health.

References Agarwal DP, Eckey R, Rudnay A-C, Volkens T, Goedde HW 1989 ‘High Km’ aldehyde dehydrogenase isozymes in human tissues: constitutive and tumor-associated forms. Prog Clin Biol Res 290:119–131 Allali-Hassani A, Weiner H 2001 Interaction of human aldehyde dehydrogenase with aromatic substrates and ligands. Chem Biol Interact 130:125–133 Ambroziak W, Izaguirre G, Pietruszko R 1999 Metabolism of retinaldehyde and other aldehydes in soluble extracts of human liver and kidney. J Biol Chem 274:33366–33373 Backlin K-I 1958 The equilibrium constant of the system ethanol, aldehyde, DPN+, DPNH and H+. Acta Chem Scand 12:1279–1285 Baumgartner M, Rabier D, Nassogne M-C et al 2005 Delta 1-pyrroline-5-carboxylate synthase deficiency: neurodegeneration, cataracts and connective tissue manifestations combined with hyperammonaemia and reduced ornithine, citrulline, arginine and proline. Eur J Pediatr 164:31–36 Becker TW, Krieger G, Witte I 1996 DNA single and double strand breaks induced by aliphatic and aromatic aldehydes in combination with copper (II). Free Radic Res 24:325–332

ACETALDEHYDE REMOVAL

35

Berkovitz BKB, Maden M, McCaffery P, Barrett AW 2001 The distribution of retinaldehyde dehydrogenase-2 in rat and human orodental tissues. Arch Oral Biol 46:1099–1104 Braun T, Bober E, Schaper J, Agarwal DP, Singh S, Goedde HW 1987 Human mitochondrial aldehyde dehydrogenase: mRNA expression in different tissues using a specific probe isolated from a cDNA expression library. Alcohol Alcohol Suppl 1:161–165 Brennan P, Lewis S, Hashibe M et al 2004 Pooled analysis of alcohol dehydrogenase genotypes and head and neck cancer: A HuGE review. Am J Epidemiol 159:1–16 Brunner G, Neupert W 1969 Localisation of proline oxidase and delta pyrroline-5-carboxylic acid dehydrogenase in rat liver. FEBS Lett 3:283–286 Castro GD, De Layño AMAD, Costantini MH, Castro JA 2001a Cytosolic xanthine oxidoreductase mediated bioactivation of ethanol to acetaldehyde and free radicals in rat breast tissue. Its potential role in alcohol-promoted mammary cancer. Toxicology 160:11–18 Castro GD, De Layño AMAD, Costantini MH, Castro JA 2001b Rat ventral prostate xanthine oxidase bioactivation of ethanol to acetaldehyde and 1-hydroxyethyl free radicals: Analysis of its potential role in heavy alcohol drinking tumor-promoting effects. Teratog Carcinog Mutagen 21:109–119 Chan W-M, Tang WK, Cheng CHK, Fong WP 2003 Purification, N-terminal sequence determination and enzymatic characterization of antiquitin from the liver of grass carp. Comp Biochem Physiol B Biochem Mol Biol 136:443–450 Cook RJ, Lloyd S and Wagner C 1991 Isolation and characterization of cDNA clones for rat liver 10-formyltetrahydrofolate dehydrogenase. J Biol Chem 266:4965–4973 Crabb DW, Matsukmoto M, Chang D, You M 2004 Overview of the role of alcohol dehydrogenase and aldehyde dehydrogenase and their variants in the genesis of alcohol-related pathology. Proc Nutr Soc 63:49–63 Deitrich RA 1966 Tissue and subcellular distribution of mammalian aldehyde-oxidizing capacity. Biochem Pharmacol 15:1911–1922 Deitrich RA, Siew C 1974 Localization and function of aldehyde dehydrogenases. In: Thurman R, Yonetani T, Williamson JR, Chance B (eds). Alcohol and Aldehyde Metabolizing Systems. Plenum Press, New York, p 125–135 Deitrich RA, Erwin VG 1975 Involvement of biogenic amine metabolism in ethanol addiction. Fed Proc 34:1962–1968 Deitrich RA, Erwin V 1980 Biogenic amine-aldehyde condensation products: tetrahydroisoquinolines and tryptolines (beta-carbolines). Annu Rev Pharmacol Toxicol 20:55–58 Deitrich RA, Hellerman L, Wein J 1962 Diphosphopyridine nucleotide-linked aldehyde dehydrogenase. I. Specificity and sigma-rho function. J Biol Chem 237:560–564 Deitrich RA, Collins AC, Erwin VG 1972 Genetic influence upon phenobarbital induced increase in rat liver supernatant aldehyde dehydrogenase activity. J Biol Chem 247: 7232–7236 Deitrich RA, Bludeau P, Stock T, Roper M 1977 Induction of different rat liver supernatant aldehyde dehydrogenases by phenobarbital and tetrachlorodibenzo-p-dioxin. J Biol Chem 252:6169–6176 Doorn JA, Hurley TD, Petersen DR 2006 Inhibition of human mitochondrial aldehyde dehydrogenase by 4-hydroxynon-2-enal and 4-oxonon-2-enal. Chem Res Toxicol 19:102–110 Dunn TJ, Koleske AJ, Lindahl R, Pitot HC 1989 Phenobarbital-inducible aldehyde dehydrogenase in the rat. cDNA sequence and regulation of the mRNA by phenobarbital in responsive rats. J Biol Chem 264:13057–13065 Eriksson CJ 1977 The distribution and metabolism of acetaldehyde in rats during ethanol oxidation—II. Biochem Pharmacol 26:249–252 Eriksson CJP, Sippel HW 1977 The distribution and metabolism of acetaldehyde in rats during ethanol oxidation—I. The Distribution of acetaldehyde in liver, brain, blood and breath. Biochem Pharmacol 26:241–247

36

DEITRICH ET AL

Erwin VG, Deitrich RA 1966 Brain aldehyde dehydrogenase. Localization, purification, and properties. J Biol Chem 241:3533–3539 Farres J, Julia P, Pares X 1998 Aldehyde oxidation in human placenta. Biochem J 256:467 Forte-McRobbie CM, Pietrusko R 1989 Human glutamic-gamma-semialdehyde dehydrogenase: Kinetic mechanism. Biochem J 261:943 Fridovich I 1966 The mechanism of the enzymatic oxidation of aldehydes. J Biol Chem 241:3126–3128 Goedde HW, Agarwal DP, Fritze G et al 1992 Distribution of ADH2 and ALDH2 genotypes in different populations. Hum Genet 88:344–346 Goodwin GW, Rougraff PM, Davis EJ, Harris RA 1999 Purification and characterization of methylmalonate-semialdehyde dehydrogenase from rat liver: Identity to malonatesemialdehyde dehydrogenase. J Biol Chem 264:14965–14971 Harada S, Agarwal DP, Goedde HW 1980 Electrophoretic and biochemical studies of human aldehyde dehydrogenase isozymes in various tissues. Life Sci 26:1773–1780 Harada S, Okubo T, Nakamura T et al 1999 A novel polymorphism (−357 G/A) of the ALDH2 gene: Linkage disequilibrium and an association with alcoholism. Alcohol Clin Exp Res 23:958–962 Hartley DP, Petersen DR 1997 Co-metabolism of ethanol, ethanol-derived acetaldehyde, and 4-hydroxynonenal in isolated rat hepatocytes. Alcohol Clin Exp Res 21:298–304 Hartley DP, Lindahl R, Petersen DR 1995a Covalent modification of class 2 and class 3 aldehyde dehydrogenase by 4-hydroxynonenal. Adv Exp Med Biol 372:93–102 Hartley DP, Ruth JA, Petersen DR 1995b The hepatocellular metabolism of 4-hydroxynonenal by alcohol dehydrogenase, aldehyde dehydrogenase, and glutathione S-transferase. Arch Biochem Biophys 316:197–205 Hellstrom E, Tottmar O 1982 Effects of aldehyde dehydrogenase inhibitors on enzymes involved in the metabolism of biogenic aldehydes in rat liver and brain. Biochem Pharmacol 31: 3899–3905 Hirose M, Kono S, Tabata S et al 2005 Genetic polymorphisms of methylenetetrahydrofolate reductase and aldehyde dehydrogenase 2, alcohol use and risk of colorectal adenomas: Selfdefense forces health study. Cancer Sci 96:513–518 Hjelmqvist l, Norin A, El-Ahmad M, Griffiths W, Jornvall H 2003 Distinct but parallel evolutionary patterns between alcohol and aldehyde dehydrogenases: addition of fish/human betaine aldehyde dehydrogenase divergence. Cell Mol Life Sci 60:2009–2016 Hoffmann T, Meyer RJ, Sorrell MF, Tuma DJ 1993 Reaction of acetaldehyde with proteins: Formation of stable fluorescent adducts. Alcohol Clin Exp Res 17:69–74 Honzatko A, Brichac J, Murphy TC et al 2005 Enantioselective metabolism of trans-4hydroxy-2-nonenal by brain mitochondria. Free Radic Biol Med 39:913–924 Hsu LC, Chang WC 1996 Sequencing and expression of the human ALDH8 encoding a new member of the aldehyde dehydrogenase family. Gene 174:319–322 Hsu LC, Chang WC, Yoshida A 1997 Human aldehyde dehydrogenase genes, ALDH7 and ALDH8: Genomic organization and gene structure comparison. Gene 189:89–94 Hu CA, Delauney AJ, Verma DP 1992 A bifunctional enzyme (delta 1-pyrroline-5-carboxylate synthetase) catalyzes the first two steps in proline biosynthesis in plants. Proc Natl Acad Sci USA 89:8354–9358 Hu CA, Lin W-W, Obie C, Valle D 1999 Molecular enzymology of mammalian delta 1-pyrroline5-carboxylate synthase. J Biol Chem 274:6754–6762 Inagaki S, Esaka Y, Deyashiki Y, Sako M, Goto M 2003 Analysis of DNA adducts of acetaldehyde by liquid chromatography-mass spectrometry. J Chromatogr A 987:341–347 Isse T, Oyama T, Kitagawa K, Matsuno K et al 2002 Dimished alcohol preference in transgenic mice lacking aldehyde dehydrogenase activity. Pharmacogentics 12:621–626 Kato S, Kawase T, Alderman J, Inatomi N, Lieber CS 1990 Role of xanthine oxidase in ethanolinduced lipid peroxidation in rats. Gastroenterology 98:203–210

ACETALDEHYDE REMOVAL

37

Kato S, Tajiri T, Matsukura N et al 2003 Genetic polymorphisms of aldehyde dehydrogenase 2, cytochrome p450 2E1 for liver cancer risk in HCV antibody-positive Japanese patients and the variations of CYP2E1 mRNA expression levels in the liver due to its polymorphism. Scand J Gastroenterol 38:886–893 Kelson TL, McVoy JRS, Rizzo WB 1997 Human liver fatty aldehyde dehydrogenase: Microsomal localization, purification, and biochemical characterization. Biochim Biophys Acta 1335: 99–110 Kera Y, Ohbora Y, Komura S 1988 The metabolism of acetaldehyde and not acetaldehyde itself is responsible for in vivo ethanol-induced lipid peroxidation in rats. Biochem Pharmacol 37:3633–3638 Kervinen K, Savolainen MJ, Tikkanen MJ, Kesäniemi YA 1991 Low density lipoprotein derivatization by acetaldehyde affects lysine residues and the B/E receptor binding affinity. Alcohol Clin Exp Res 15:1050–1055 Kim JM, Stewart R, Shin IS, Jung JS, Yoon JS 2004 Assessment of association between mitochondrial aldehyde dehydrogenase polymorphism and Alzheimer’s disease in an older Korean population. Neurobiol Aging 25:295–301 Krupenko SA, Wagner C, Cook RJ 1995 Recombinant 10-formyltetrahydrofolate dehydrogenase catalyses both dehydrogenase and hydrolase reactions utilizing the synthetic substrate 10formyl-5,8-dideazafolate. Biochem J 306:651–655 Krupenko SA, Wagner C, Cook RJ 1997 Expression, purification, and properties of the aldehyde dehydrogenase homologous carboxyl-terminal domain of rat 10-formyltetrahydrofolate dehydrogenase. J Biol Chem 272:10266–10272 Kuykendall JR, Bogdanffy MS 1992 Efficiency of DNA-histone crosslinking induced by saturated and unsaturated aldehydes in vitro. Mutat Res 283:131–136 Lee M-O, Manthey CL, Sladek NE 1991 Identification of mouse liver aldehyde dehydrogenases that catalyze the oxidation of retinaldehyde to retinoic acid. Biochem Pharmacol 42: 1279–1285 Lieber CS 1997 Ethanol metabolism, cirrhosis and alcoholism. Clin Chim Acta 257:59–84 Lin M, Napoli JL 2000 cDNA cloning and expression of a human aldehyde dehydrogenase (ALDH) active with 9-cis-retinal and identification of a rat ortholog, ALDH12. J Biol Chem 275:40106–40112 Lin M, Zhang M, Abraham M, Smith SM, Napoli JL 2003 Mouse retinal dehydrogenase 4 (RALDH4) molecular cloning, cellular expression, and activity in 9-cis-retinoic acid biosynthesis in intact cells. J Biol Chem 278:9861 Lindahl R, Evces S 1984a Rat liver aldehyde dehydrogenase II. Isolation and characterization of four inducible isozymes. J Biol Chem 259:11991–11996 Lindahl R, Evces S 1984b Rat liver aldehyde dehydrogenase I. Isolation and characterization of four high Km normal liver isozymes. J Biol Chem 259:11986–11990 Luckey SW, Tjalkens RB, Petersen DR 1999 Mechanism of inhibition of rat liver class 2 ALDH by 4-hydroxynonenal. Adv Exp Med Biol 463:71–77 Lynch M, Cameron TL, Knight M et al 2002 Structural and mutational analysis of Antiquitin as a candidate gene for Meniere disease. Am J Med Genet 110:397–399 MacKerell AD Jr, Blatter EE, Pietruszko R 1986 Human aldehyde dehydrogenase: kinetic identification of the isozyme for which biogenic aldehydes and acetaldehyde compete. Alcohol Clin Exp Res 10:266–270 Mantle D, Preedy VR 1999 Free radicals as mediators of alcohol toxicity. Adverse Drug React Toxicol Rev 18:235–252 Meyer MJ, Mosley DE, Amarnath V, Picklo MJ Sr 2004 Metabolism of 4-hydroxy-trans-2-nonenal by central nervous system mitochondria is dependent on age and NAD+ availability. Chem Res Toxicol 9:1272–1279 Mills PB, Struys E, Jakobs C et al 2006 Mutation in antiquitin in individuals with pyridoxinedependent seizures. Nat Med 12:307–309

38

DEITRICH ET AL

Mira L, Maia L, Barreira L, Manso CF 1995 Evidence for free radical generation due to NADH oxidation by aldehyde oxidase during ethanol metabolism. Arch Biochem Biophys 218: 53–58 Mitchell DY, Petersen DR 1991 Inhibition of rat hepatic mitochondrial aldehyde dehydrogenasemediated acetaldehyde oxidation by trans-4-hydroxy-2-nonenal. Hepatology 13:728–734 Morita M, Oyama T, Kagawa N et al 2005 Expression of aldehyde dehydrogenase 2 in the normal esophageal epithelium and alcohol consumption in patients with esophageal cancer. Front Biosci 10:2319–2324 Murata C, Taniyama M, Kuriyama S et al 2004 Meta-analysis of three diabetes population studies: association of inactive ALDH2 genotype with maternal inheritance of diabetes. Diabetes Res Clin Pract 66:S145–147 Narita M, Kitagawa K, Nagai Y et al 2003 Effects of aldehyde dehydrogenase genotypes on carotid atherosclerosis. Ultrasound Med Biol 29:1415–1419 Nguyen E, Picklo MJ Sr 2003 Inhibition of succinic semialdehyde dehydrogenase activity by alkenal products of lipid peroxidation. Biochim Biophys Acta 1637:107–112 Nilsson GE 1988 Studies on the role of biogenic aldehydes in brain. In: Studies on the role of biogenic aldehydes in brain. Reprocentralen HSC, Uppsala, Sweden, pp 1–47 Nordback IH, MacGowan S, Potter JJ, Cameron JL 1991 The role of acetaldehyde in the pathogenesis of acute alcoholic pancreatitis. Ann Surg 214:671–678 Oei HH, Zoganas HC, McCord JM, Schaffer SW 1986 Role of acetaldehyde and xanthine oxidase in ethanol-induced oxidative stress. Res Commun Chem Pathol Pharmacol 51:195–203 Ohsawa I, Kamino K, Nagasaka K et al 2003 Genetic deficiency of a mitochondrial aldehyde dehydrogenase increases serum lipid peroxides in community-dwelling females. J Hum Genet 48:404–409 Ohta S, Ohsawa I, Kamino K, Ando F, Shimokata H 2004 Mitochondrial ALDH2 deficiency as an oxidative stress. Ann NY Acad Sci 1011:36–44 Pappa A, Estey T, Manzer R, Brown D, Vasiliou V 2003 Human aldehyde dehydrogenase 3A1 (ALDH3A1): biochemical characterization and immunohistochemical localization in the cornea. Biochem J 376:615–623 Petersen DR, Hjelle JJ, Mitchell DY 1991 Aldehydic products of lipid peroxidation: substrates or inhibitors of hepatic aldehyde dehydrogenase? Adv Exp Med Biol 284:67–73 Quintanilla ME, Tampier L, Valle-Prieto A, Sapag A, Israel Y 2005 Complex I regulates mutant mitochondrial aldehyde dehydrogenase activity and voluntary ethanol consumption in rats. FASEB J 19:36–42 Racker E 1949 Purification of aldehyde dehydrogenase. J Biol Chem 177:883 Rajagopalan KV, Handler P 1964 Hepatic aldehyde oxidase. 3. The substrate-binding site. J Biol Chem 239:2027–2035 Ryzlat MT, Pietruszko R 1989 Human brain glyceraldehyde-3-phosphate dehydrogenase, succinic semialdehyde dehydrogenase and aldehyde dehydrogenase isozymes: Substrate specificity and sensitivity to disulfiram. Alcohol Clin Exp Res 13:755–761 Saari JC, Champer RJ, Asson-Batres MA et al 1995 Characterization and localization of an aldehyde dehydrogenase to amacrine cells of bovine retina. Vis Neurosci 12:263–272 Sakamoto T, Hara JM, Higaki Y et al 2005 Influence of alcohol consumption and gene polymorphisms of ADH2 and ALDH2 on hepatocellular carcinoma in a Japanese population. Int J Cancer 118:1501–1507 Salaspuro MP 2003 Alcohol consumption and cancer of the gastrointestinal tract. Best Pract Res Clin Gastroenterol 17:679–694 Shaw S, Jayatilleke E, Herbert V, Colman N 1989 Cleavage of folates during ethanol metabolism. Role of aldehyde/xanthine oxidase-generated superoxide. Biochem J 257:277–280 Shaw S, Jayatilleke E 1990a Ethanol-induced iron mobilization: Role of acetaldehyde-aldehyde oxidase generated superoxide. Free Radic Biol Med 9:11–17

ACETALDEHYDE REMOVAL

39

Shaw S, Jayatilleke E 1990b The role of aldehyde oxidase in ethanol-induced hepatic lipid peroxidation in the rat. Biochem J 268:579–583 Shin IS, Stewart R, Kim JM et al 2005 Mitochondrial aldehyde dehydrogenase polymorphism is not associated with incidence of Alzheimer’s disease. Int J Geriatr Psychiatry 20: 1075–1080 Siew C, Deitrich RA, Erwin VG 1976 Localization and characteristics of rat liver mitochondrial aldehyde dehydrogenases. Arch Biochem Biophys 176:638–649 Skvorak A, Robertson NG, Yi Y et al 1997 An ancient conserved gene expressed in the human inner ear: Identification, expression analysis, and chromosomal mapping of human and mouse antiquitin (ATQ1). Genomics 46:191–199 Sladek NE 1999 Aldehyde dehydrogenase-mediated cellular relative insensitivity to the oxazaphosphorines. Curr Pharm Des 5:607–625 Sladek N 2002 Human aldehyde dehydrogenases: potential pathological, pharmacological, and toxicological impact. J Biochem Mol Toxicol 17:7–23 Sophos NA, Vasiliou V 2003 Aldehyde dehydrogenase gene superfamily: the 2002 update. Chem Biol Interact 143:5–22 Stewart MJ, Malek K, Xiao Q, Dipple KM, Crabb DW 1995 The novel aldehyde dehydrogenase gene, ALDH5, encodes an active aldehyde dehydrogenase enzyme. Biochem Biophys Res Commun 211:144–151 Sugata K, Takada A, Takase S, Tsutsumi M 1988 Determination of aldehyde dehydrogenase isozyme activity in human liver. Alcohol 5:39–43 Sun X, Jia Y, Zhang X, Xu Q, Shen Y, Li Y 2005 Multi-locus association study of schizophrenia susceptibility genes with a posterior probability method. Sci China C Life Sci 48:263–269 Tang WK, Cheng CHK, Fong WP 2002 First purification of the antiquitin protein and demonstration of its enzymatic activity. FEBS Lett 516:183–186 Terelius Y, Norsten-Höög C, Cronholm T, Ingelman-Sundberg M 1991 Acetaldehyde as a substrate for ethanol-inducible cytochrome P450 (CYP2E1). Biochem Biophys Res Commun 179:689–694 Testore G, Cravanzola C, Bedino S 1999 Aldehyde dehydrogenase from rat intestinal mucosa: purification and characterization of an isozyme with high affinity for gamma-aminobutyraldehyde. Int J Biochem Cell Biol 31:777–786 Thiele GM, Tuma DJ, Willis MS et al 1998 Soluble proteins modified with acetaldehyde and malondialdehyde are immunogenic in the absence of adjuvant. Alcohol Clin Exp Res 22:1731–1739 Tottmar SOC, Pettersson J, Kiessling K-H 1973 The subcellular distribution and properties of aldehyde dehydrogenases in rat liver. Biochem J 135:577–586 Ueshima Y, Matsuda Y, Tsutsumi M, Takada A 1993 Role of the aldehyde dehydrogenase-1 isozyme in the metabolism of acetaldehyde. Alcohol Alcohol Suppl 28:15–19 Valle D, Goodman SI, Harris SC, Phang JM 1979 Genetic evidence for a common enzyme catalyzing the second steop in the degradation of proline and hydroxyproline. J Clin Invest 64:1365–1370 van den Brink DM, van Miert JM, Wanders RJA 2005 A novel assay for prenatal diagnosis of Sjorgren-Larsson syndrome. J Inherit Metab Dis 28:965–969 Vasiliou V, Nebert DW 2005 Analysis and update of the human aldehyde dehydrogenase (ALDH) gene family. Hum Genomics 2:138–143 Vasiliou V, Weiner H, Marselos M, Nebert DW 1995 Mammalian aldehyde dehydrogenase genes: Classification based on evolution, structure and regulation. Eur J Drug Metab Pharmacokinet Special issue:53–64 Vasiliou V, Bairoch A, Tipton KE, Nebert DW 1999 Eukaryotic aldehyde dehydrogenase (ALDH) genes: human polymorphisms, and recommended nomenclature based on divergent evolution and chromosomal mapping. Pharmacogenetics 9:421–434

40

DEITRICH ET AL

Vasiliou V, Pappa A, Estey T 2004 Role of human aldehyde dehydrogenases in endobiotic and xenobiotic metabolism. Drug Metab Rev 36:279–299 Vila R, Kurosaki M, Barzago MM et al 2004 Regulation and biochemistry of mouse molybdoflavoenzymes. The DBA/2 mouse is selectively deficient in the expression of aldehyde oxidase homologues 1 and 2 and represents a unique source for the purification and characterization of aldehyde oxidase. J Biol Chem 279:8668–8683 Von Wartburg JP, Berger D, Ris MM, Tabakoff B 1975 Enzymes of biogenic aldehyde metabolism. In: Gross MM (ed) Alcohol intoxication and withdrawal. Plenum Press, New York and London, p 119–138 Watanabe S, Sasahara K, Kinekawa F et al 2002 Aldehyde dehydrogenase-2 genotypes and HLA haplotypes in Japanese patients with esophageal cancer. Oncol Rep 9:1063–1068 Weiner H 1979 Acetaldehyde metabolism. In: Majchrowicz E, Noble EP (eds) Biochemistry and pharmacology of ethanol. Plenum Press, New York, p 125–144 Wierzchowski J, Wroczynski P, Laszuk K, Interewicz E 1997 Fluorimetric detection of aldehyde dehydrogenase activity in human blood, saliva, and organ biopsies and kinetic differentiation between class I and class III isozymes. Anal Biochem 245:69–78 Willis MS, Klassen LW, Tuma DJ, Sorrell MF, Thiele GM 2002 Adduction of soluble proteins with malondialdehyde-acetaldehyde (MAA) induces antibody production and enhances T-cell proliferation. Alcohol Clin Exp Res 26:94–106 Wroczynski P, Wierzchowski J 2000 Aromatic aldehydes as fluorogenic indicators for human aldehyde dehydrogenases and oxidases: substrate and isozyme specificity. Analyst 125: 511–516 Xiao Q, Weiner H, Crabb DW 1996 The mutation in the mitochondrial aldehyde dehydrogenase (ALDH2) gene responsible for alcohol-induced flushing increases turnover of the enzyme tetramers in a dominant fashion. J Clin Invest 98:2027–2032 Yamamoto H, Tanegashima A, Hosoe H, Fukunaga T 2000 Fatal acute alcohol intoxication in an ALDH2 heterozygote: a case report. Forensic Sci Int 112:201–207 Yin S-J, Wang M-F, Han C-L, Wang S-L 1995 Substrate binding pocket structure of human aldehyde dehydrogenases: A substrate specificity approach. Adv Exp Med Biol 372:9–16 Yoon KA, Nakamuira KA, Arakawa H 2004 Identification of ALDH4 as a p53-inducible gene and its protective role in cellular stresses. J Hum Genet 49:134–140 Yoshida A, Rzhetsky A, Hsu LC, Chang C 1998 Human aldehyde dehydrogenase gene family. Eur J Biochem 251:549–557

DISCUSSION Seitz: You showed that 4-HNE is inhibiting ALDH. Are there any data indicating that induction of CYP2E1 leads to an increase in acetaldehyde? We would suppose that 4-HNE would inhibit ALDH and may cause an increase in acetaldehyde. Do we have any data showing that this is the case in humans? Deitrich: Not that I know of. There is an animal model of this: the deer mouse, which has no ADH (Burnett & Felder 1978). They have huge levels of CYP2E1 (Alderman et al 1989), but I’m not sure that their acetaldehyde levels have been measured. Eriksson: There are indirect data. Alcoholics often have high acetaldehyde levels. This is one of the few conditions in which you can measure it in the venous blood. It seems that this may be associated with induced alcohol metabolism by CYP2E1. In abstinent alcoholics who have already normalized this, this isn’t usually found.

ACETALDEHYDE REMOVAL

41

Apte: I want to follow up on the 4-HNE, which is a marker of oxidative stress. It was interesting to hear that 4-HNE actually does inhibit ALDH, so there are both toxic products increasing in the cell with oxidant stress and acetaldehyde. Have there been any data on antioxidants and fetal alcohol syndrome? Have people looked at the effects of antioxidants that reduce the effects of ethanol oxidation within the cells and allow acetaldehyde to be oxidised further to acetate? Deitrich: My assumption would be that these studies have been done. But I’m not familiar with them. Preedy: We have carried out some studies using α-tocopherol (vitamin E), trying to ameliorate changes in alcoholic muscle disease (Reilly et al 2000). We have found that it wasn’t preventive. In control animals the α-tocopherol itself seemed to be causing muscle damage. There were two independent studies: one with acute dosage of a few days, and a second study carried out for four weeks by Dr Michael Koll, Prof Helmut Seitz and Prof Timothy Peters (Koll et al 2003). There was no preventive effect, but control animals had reduced muscle protein content and there was an increase in protein turnover in those muscles. Apte: What sort of doses were used? Preedy: We used 30 mg/kg body weight, which is fairly high, but we felt that we needed to overcome the criticism that some of the dosage regimes with αtocopherol may have been ineffective because they couldn’t raise tissue and plasma levels. Seitz: Perhaps there is another aspect, with respect to antioxidative measures. We looked at hyperproliferation in the colon. We believe from our data that acetaldehyde correlates with hyperproliferation. In the study with Dr Preedy we gave vitamin E and found that it decreased the hyperproliferation. This was strange. It seems that with vitamin E we were decreasing lipid peroxidation products, which can then affect ALDH, lowering acetaldehyde, which may lead to a decrease in hyperproliferation. Morris: Oxidation of acetaldehyde produces protons as well in the respiratory tract. Intracellular acidification is thought to be important in toxicity (Morris 1997). Is this thought to contribute in other organ systems as well? Deitrich: There are ALDHs in almost every tissue. The efficient aldehyde oxidation takes place inside the mitochondrion. With the shuttle mechanisms, the reducing equivalents of NADH can come out, but the initial oxidation takes place within the mitochondria. Albano: With regard to the minor pathways of acetaldehyde metabolism involving xanthine oxidase or aldehyde oxidase, we have observed that the production of reactive oxygen species during acetaldehyde oxidation by xanthine oxidase is responsible for the conversion of acetaldehyde to a free radical species that has been identified as a methyl-carbonyl free radical (Albano et al 1994). Because of this reactivity, this acetaldehyde-derived radical significantly contributes to the binding of the aldehyde to proteins.

42

DEITRICH ET AL

Apte: Regarding the vitamin E and the lack of prevention in your model, was the vitamin E given with the induction of damage, or prior to it? Preedy: We gave it at the same time as we commenced the alcohol feeding regime. I agree that there are different models that we could have used. Apte: If any of this has to have in vivo translation for treatment, most people wouldn’t come to you before they started drinking! Preedy: Built in with this experimental model was a pre-treatment with tocopherol for four weeks, with an acute dosage of alcohol on top of this. It was rather a complex study, with chronic and acute regimens, and a superimposition of acute upon chronic treatments. It didn’t have any preventive effect on the muscle, but there was a preventive effect on the gut. In interpreting data on acetaldehydeinduced damage, we must be careful because the mechanisms differ among tissues. Worrall: I analysed the livers from these experiments. We got some rather unusual results. Adduct levels for things like lipoperoxidation products such as malondialdehyde were knocked down, as you might expect, but acetaldehyde adducts themselves were also knocked down by α-tocopherol treatment. We would have expected these to stay the same or go up. In fact, they went down. Perhaps it is a case of different tissues, different effects. Apte: Couldn’t that be explained differently? If you inhibit the lipid peroxidation you are facilitating the acetaldehyde to acetate transformation and not giving it time to form adducts. Worrall: It may depend on the relative rates of acetaldehyde oxidation. There seems to be evidence now that specific proteins have specific target sites that react readily with acetaldehyde. The acetaldehyde can bind very quickly to these sites even at low concentrations, such that at physiological concentrations, you almost build up a steady-state modification level over quite a short period. As long as you don’t knock the acetaldehyde level down too low, it doesn’t seem to affect the level of modification too much. At much higher (non-physiological) concentrations of acetaldehyde other sites on proteins become modified, leading to great increases in the level of modification. Emery: I’d like to return to the point Emanuele Albano made about the radical product from acetaldehyde coming from minor oxidase pathways. Do you have any data on the quantities of that species that are formed? Albano: Unfortunately not, because it is quite difficult to estimate the rates of free radical formation using EPR spectroscopy. However, the reaction appears to be quite efficient. Using xanthine oxidase the formation of methylcarbonyl radical was evident at concentrations of acetaldehyde as low as 0.1 mmol/litre. This concentration is compatible with those present in the liver during the exposure to ethanol. The reaction is dependent upon the formation of hydrogen peroxide and superoxide anion originating during the oxidation of acetaldehyde itself (Albano et al 1994).

ACETALDEHYDE REMOVAL

43

Aldehyde oxidase, which has a much lower Km for acetaldehyde than xanthine oxidase, has also been shown to generate reactive oxygen species. Thus, it is possible that the formation of methylcarbonyl free radical might have biological relevance in tissues rich in these enzymes. Ren: You made a statement that the surrogate aldehydes do more damage than acetaldehyde itself. Would that balance be different in individuals with defective ALDH? Is this the reason for more severe tissue damage? Deitrich: It could be. The one example I gave was the hypothesis that liver damage in diabetes is due to the increased production of 4-HNE. It turns out for various reasons that this may explain the inheritance of diabetes from the mother. It seems possible that the damage is due to the surrogate aldehydes as much as acetaldehyde, because they are much more reactive even though their concentrations may be lower. Crabb: One commonly reads that acetaldehyde induces oxidative stress. My understanding of the xanthine and aldehyde oxidases is that their Kms for acetaldehyde were pretty high. Deitrich: For xanthine oxidase it is, but for aldehyde oxidase it is fairly low, at around 30 µM. Under normal circumstances acetaldehyde won’t get that high in blood, but it is much higher than that in tissue—an explanation we always fall back on if we are stuck for an explanation. Crabb: If you are measuring these peroxidatic lipids, you might conclude that the oxidative stress was greater and you produced more, as opposed to this cross-talk with ALDH. One of my colleagues, Tom Hurley, has been working on the ability of nitrate to inhibit ALDH. Apparently this can happen with nitroglycerin and NO. I have heard of others looking at this, and they have found reversible nitrosylation of the reactive cysteine in ALDH2. The interesting thing is that this reverses when the enzyme is isolated with buffers containing DTT. Some of what we think we are measuring when we look at whether this enzyme activity might account for flux through the alcohol metabolism pathway may not be right, because we can’t measure what the activity of the enzyme was before we disrupted the cells. I don’t know whether this is the case with the other aldehydes. Deitrich: The 4-HNE is an irreversible inhibitor of the ALDH (Doorn et al 2006). Eriksson: We speak a lot about oxidative stress, but what is the primary effect of acetaldehyde and its metabolism? It is the same as with ethanol. They are both putting reductive stress into the system. Acetaldehyde is increasing intramitochondrial redox. How much is this affecting these different oxidative processes? This is often forgotten. Albano: Oxidative stress induced by alcohol is the result of the combined impairment of antioxidant defences and the stimulation of the production of reactive oxygen species by several sources. To list the main ones I would mention first

44

DEITRICH ET AL

the leakage of reactive oxygen species caused by the partial uncoupling of the mitochondrial respiratory chain (Bailey & Cunningham 2003). This mechanism has a role in acute intoxication, but it is even more important during chronic exposure to alcohol because the accumulation of mitochondrial DNA mutations induced by reactive oxygen species further impairs the functions of the respiratory enzymes (Hoek et al 2002). The second main pathway involves the induction of CYP2E1 which, in the absence of substrate, has an NADPH oxidase activity responsible for the generation of hydrogen peroxide and superoxide anion. The importance of CYP2E1 in causing oxidative stress during alcohol intoxication has been substantiated by a number of studies by Cederbaum’s group using hepatic cells over-expressing CYP2E1 (Caro & Cederbaum 2004). Furthermore, CYP2E1 is responsible for the conversion of ethanol to hydroxyethyl free radicals (Albano 2002). Last but not least, the activation of granulocytes and macrophages by inflammatory mechanisms is now a well recognized source of reactive oxygen species and NO during alcoholic liver disease (Arteel 2003). The generation of reactive oxygen species by the minor pathways of acetaldehyde metabolism involving xanthine oxidase and aldehyde oxidase might also play a role as suggested by Shaw & Jayatilleke (1990). It is possible that these latter mechanisms might be particularly relevant in some extra-hepatic tissues such as skeletal muscle and the myocardium where oxidative stress appears to be dependent on the presence of acetaldehyde. Emery: So you are saying that a dose of alcohol by itself isn’t going to cause oxidative damage, but it is the effect of repeated exposure which is a problem. Albano: Yes. There are a number of early studies in rodents reporting on the detection of oxidative stress following acute ethanol intoxication (see Nordmann et al 1992 for review). Nonetheless much of the oxidative damage occurs during chronic exposure to alcohol in relation to mitochondrial impairment, CYP2E1 induction and inflammation. Apte: I guess it affects the antioxidant status within the cell too, and I’m not sure what role is played by acetaldehyde–glutathione adducts. These prevent the glutathione from adopting its normal antioxidant role. An acute alcohol dose may cause oxidative stress, but it might be overcome by the cell’s natural defences. Prolonged exposure could be different. M Salaspuro: How much acetaldehyde must be given to experimental animals in order to produce some of these CNS effects, e.g. dementia? Deitrich: The problem with acetaldehyde is that you can’t maintain much of a blood level, even if it is given by infusion. If you do that, it is extremely toxic. I guess we could get deer mice and make an ALDH knockout; then we could study just acetaldehyde without it being reduced or oxidized. Eriksson: The rat model is a bad one to compare with humans because rats have a buffer capacity for binding acetaldehyde in fantastic amounts. It would be hard

ACETALDEHYDE REMOVAL

45

to produce any effects. If you want to study the toxic effects in the liver, for example, you shouldn’t choose rats. I have another question. You didn’t say much about ALDH1*A1, which is now quite a hot topic. There is a Finnish association between this gene and alcoholism. I believe one of the most toxic effects of acetaldehyde is that it is one of the causation factors for producing alcohol addiction. People don’t get addicted to the alcohol, but to the acetaldehyde. We should keep in mind that acetaldehyde has such harmful effects. Deitrich: I was aware of the 1A1 polymorphism, but I got the impression at least from the publication that there wasn’t any evidence that it was related to alcoholism. Eriksson: It was an association. There is something strange about 1A1. The normal association between ALDH2, for example, is that if you have inhibited enzyme activity, this is protective. However, ALDH1*A1 is much lower down in alcoholics and yet they drink: it is not protective at all. It seems it is associated with the state of drinking. The difference between these two ALDHs is noteworthy. 1A1 seems to be especially meaningful for the brain aldehyde metabolism. Quertemont: I want to comment on the effects of acetaldehyde on the CNS. We got some surprising results. We daily injected high amounts of acetaldehyde (>100 mg/kg intraperitoneally) into mice for 10 days and they seemed fine. But we noticed that these mice showed a severe anterograde amnesia. The mice were unable to learn new tasks, especially operant tasks. Perhaps there is a toxic effect of acetaldehyde on the brain that has yet to be investigated. Is there a reason to believe that damage to the brain can be caused by acetaldehyde? Or should we be looking at an effect on the liver with a consequence for brain functioning. Deitrich: Ostrowska et al (2004) have shown that the 4-HNE levels in the brain are much higher than in the liver. Emery: Did you measure the levels of any of the other aldehydes in your animals? Quertemont: No. Thornalley: In physiological situations in which there is oxidative stress, the physiological situation does not stay unresponsive. It responds by inducing genes with antioxidant response elements in their promoters. One of the activators of this transcription factor-mediated event is 4-HNE. If this hypothesis is correct, we would expect to see alcoholism associated with a burst of induction of antioxidant response element-linked gene expression. Is there any evidence for this? Crabb: There is in the cell model that Art Cederbaum has used. These cells have CYP2E1 expressed, and most of the expected antioxidant gene battery was turned on. He was following levels of glutathione and worked backwards to this. Thornalley: You mentioned that acetaldehyde can be reversed to ethanol. Which enzymes are involved? Are aldoketo reductases involved?

46

DEITRICH ET AL

Deitrich: No, just plain old ADH. I have a question for David Crabb on the product inhibition of ADH by acetaldehyde. Is this a real dead-end product inhibition, or is it reversal of the reaction? Crabb: Kinetically speaking, it is not the equilibrium term; rather it is due to competitive product inhibition. This is why you get different answers for the different enzymes. If it was just a matter of reversing the equilibrium, it should apply to all of the enzymes, since the equilibrium is independent of the kinetics of the enzymes. Apte: Is the protective role of decreased ALDH2 due to the fact that these people drink less since they avoid alcohol because of its unpleasant side-effects? Eriksson: The heterozygotes do drink. They can become alcoholics. They are much more likely to develop cancers because of increased acetaldehyde exposure: the odds ratio is very much increased. With regard to liver damage the situation is less clear. Someone mentioned earlier this new meta analysis that shows that in Asians there isn’t a relationship. I did a similar meta analysis some years ago and found an association. Thinking about the telescoping effect, acetaldehyde is euphoric as well as being toxic. This can be seen in any restaurant in Japan or China. The flushers become very happy, just as non-flushers become with much higher doses of alcohol. There is no other reason than the acetaldehyde. Crabb: The group in San Diego has changed my way of looking at acetaldehyde. The story used to be that acetaldehyde makes you feel sick so you don’t drink. Now it seems that you are at most risk of becoming alcoholic when you don’t respond much to drinking, whether this is because of something you inherited, or some other reason (Luczak et al 2002). Eriksson: This is a different aspect. It is well known that people who are tolerant of alcohol have a higher risk of becoming alcoholics, but this could be simply explained by the fact that those who are tolerant drink more. The acetaldehyde connection is a different issue. Apte: You mentioned that there is acetaldehyde in different alcoholic beverages. Do different beverages affect people differently? Eriksson: This has been studied, but people doing epidemiological investigations find that people who drink wine or spirits tend to have different lifestyles and personalities, especially in northern countries. The studies done with different beverages are quite confounded. A lot of these complicating factors haven’t been acknowledged. Deitrich: David Crabb raised the issue of disulfiram (Antabuse). At least in the USA judges hand out disulfiram like candy. Alcoholics who offend often have the choice of going to jail or taking Antabuse. The medical profession should look into this because the people taking Antabuse will be in essence ALDH negative. The assumption is that they are not going to be drinking alcohol, but the drug also inhibits the ALDH for all these other substrates.

ACETALDEHYDE REMOVAL

47

Preedy: As I understand it, the ALDH2 inhibitors can have quite severe long-term side effects. This could be due to the drug, or because this knocks out the other enzymes that the tissue needs for its normal metabolism. Deitrich: It is not an innocuous drug. Eriksson: If we want to know about the consequences of knocking out ALDH2, 7% of the Asian population have this genotype. We have recently published a study about the Chinese population (Luo et al 2005). There wasn’t really any difference in factors such as mortality. It would be interesting to look at some of the rarer syndromes. They could be enriched to a degree that they could be picked up in a population study. Preedy: A study published some years ago (Ohta et al 2004) looked at the ALDH2deficient subjects, and hypothesized that because of the deficiency in this enzyme they had a higher risk of other diseases, such as Alzheimer’s. They related this to the role of ALDH2 in the oxidation of other aldehydes. Eriksson: On the general mortality level, we don’t find a significant effect of ALDH2. On the other hand, there may be negative effects which are balanced out by positive effects. If you are born with a deficient ALDH you don’t become an alcoholic, for example. There could be a combination of plus or minus effects. Deitrich: The difference between ALDH-negative subjects and people treated with Antabuse is that Antabuse has many other effects. It hits many of the other ALDHs. Crabb: And CYP2E1. All our inhibitors are turning out to be non-specific. Rao: If I remember correctly, did you say that alcoholic beverages have millimolar concentrations of acetaldehyde? Eriksson: There is a huge variation. The maximum level is 5 mM, which is unbelievably high. Rao: In tissues we are talking about levels one thousandth of this. Is the acetaldehyde level in some of these alcoholic drinks a matter of real concern? Eriksson: It is an interesting question. Where is the cancer formed? It is formed in the mouth and upper digestive tract—places that come into contact with the beverages. Alcohol doesn’t so much increase the other forms of cancer that come from deeper inside. It has never been surveyed well epidemiologically because of the confounding factors. The myth is that the worst alcoholic drinks are vodka because of its strength, but perhaps the truth is that those alcoholic beverages with high acetaldehyde levels could be worse for you. M Salaspuro: There is some epidemiological evidence coming from France. In Normandy, it has been common for people to drink a lot of Calvados together with hot coffee. After adjustment for other alcoholic beverages this habit appeared to explain almost half of the peak incidence of oesophageal cancer in the Northwest of France compared to the regions with lower oesophageal cancer risk (Launoy et al 1997). We have recently measured the acetaldehyde concentration of Calvados

48

DEITRICH ET AL

produced by factories and also made in small farms. The acetaldehyde level of these beverages was in the millimolar range. This is the only epidemiological evidence we have so far. There are many other food products that also have considerable levels of acetaldehyde. Companies making dairy products sometimes try to increase the acetaldehyde level in order to give a special aromatic taste. One example is yoghurt, where some have acetaldehyde levels as high as 800 µmoles/l. Apte: Did the Calvados study tease out smoking levels? M Salaspuro: Yes, it was one of the possible confounding factors that was excluded. Aranda: Acetaldehyde is an off flavour in wine caused by ethanol oxidation, and producers aim to keep it as low as possible. In sherry wines, however, the amount is pretty high at around 300–400 mg per litre (7–9 mM), because these wines are made in an oxidative style. The process of making these wines results in high levels of acetaldehyde that are good for the organoleptic properties of these wines. Seitz: Dr Salaspuro pointed out that acetaldehyde is present in certain beverages. However, the concentration of alcohol may also be important. 40% alcohol may have a local toxic effect, leading to mucosal hyper-regeneration. In this situation biogenic amines such as spermidine and spermine are produced. In the presence of these amino acids there is increased formation of acetaldehyde DNA adducts with mutagenic and carcinogenic properties. Thus, with high concentrated alcohols, hyper-regeneration and the production of certain acetaldehyde DNA adducts are favoured. Emery: Yes, we can’t just focus on alcohol. Perhaps we need to think of acetaldehyde coming from sources other than alcohol, too. Okamura: I’d like to add some information from a Japanese study concerning beverage types (Okamura et al 2004). The effect on serum lipids and blood pressure level is the same across beverages for the same level of ethanol intake. In Japanese populations, wine drinkers show the highest serum level of cholesterol because their eating pattern is quite different. We have some cohort study data concerning the relationship between ALDH2 genotypes and total mortality (personal communication). We don’t see a significant difference among ALDH2 genotypes at the same alcohol drinking level. In non-drinkers the mortalities are quite similar. Shukla: How do some of the contaminants that are found in alcoholic drinks, such as methanol, affect the ALDH levels? Would the metabolism of surrogate alcohols increase the toxicity? Deitrich: They certainly would, but I have no idea what the Km values would be for methanol to formaldehyde. Formaldehyde is a poor substrate for ALDH2. The other longer chain aldehydes are reasonably good substrates until you get to very long chain; then the fatty aldehyde dehydrogenase takes over and metabolizes those.

ACETALDEHYDE REMOVAL

49

Shukla: I remember that isopentanol, for example, has been a problem as a contaminant for some alcoholic drinks. Deitrich: I have a question about the acatalasaemic individuals that David Crabb mentioned. There is an acatalasaemic mouse, but it is missing catalase only in the blood. The levels in the brain are about 50% and the kidney and liver are normal. What is the situation with humans? What are the catalase levels in other tissues in acatalasaemic humans? Okamura: I major in epidemiology, so I can’t answer some of these questions about the more basic science. At least in humans aged 40 and over, we don’t observe significant differences in the relationship of some genotypes such as ALDH2 and mortality or other risk factors (Amamoto et al 2002). Eriksson: If you look at people dying aged 40 years, the cause of death at such an early age will bear a high relationship with alcohol consumption. If you don’t see such a relationship in Japan, then there is something strange. The ALDH2 homozygous deficient people are non-drinkers, and if they still die at the same frequency aged 40, we should investigate this. Okamura: We need more information about early death due to alcohol drinking in Asia. It is difficult to perform this sort of study in some social groups, but it is a good suggestion. I will try to analyse some data sets of cohort study with information of ALDH2 genotypes. However, the life expectancy in Japan at age 20 is longer than almost all other developed or developing countries. This suggests that there is no burden on life expectancy in Japan by ALDH2 deficit. Niemelä: Earlier on we discussed acetaldehyde toxicity and its association with oxidative stress. Studies both in experimental models and in humans have indicated that high fat diets or excess iron modulate the generation of acetaldehyde and associated tissue damage. It is quite possible that in humans, conditions such as obesity or diabetes also create the status of enhanced oxidative stress thereby adding to acetaldehyde toxicity in alcohol consumers. Apte: I’d like to be a bit provocative about the catalase issue. It’s my understanding that catalase plays a minor role in alcohol oxidation. Would it matter if people are catalase deficient, and is there any point in doing a large study on this? Deitrich: Catalase is the major alcohol oxidizing enzyme in the brain. Catalasedeficient individuals might have interesting things happening in the brain. Brain doesn’t have very much ALDH, either. The issue here is mainly one of CNS effects of alcohol as related to acetaldehyde production in the brain via catalase and CYP2E1. Eriksson: So the point is that catalase may matter, because if the underlying causation is the alcohol metabolism in the brain, then those systems that are specifically working in the brain are of great interest. Worrall: Catalase is 90–95% of the metabolic capacity in certain areas of the brain. CYP2E1 is vanishingly small.

50

DEITRICH ET AL

Emery: Presumably, in catalase-deficient animals other enzymes would be up-regulated. This would be different from acute pharmacological effects. Jamal: I have done these experiments looking at the effects of catalase, ADH and ALDH inhibitors on brain acetaldehyde concentrations after ethanol dosing, which we measured with in vivo microdialysis in the striatum of free-moving rats. Catalase has an effect, but it doesn’t make a major contribution. High peripheral acetaldehyde levels may be able to cross the blood–brain barrier and may produce a greater contribution to brain acetaldehyde accumulation. When there is more than 70 µM acetaldehyde in the blood it does cross the blood–brain barrier. Other enzymes such as CYP2E1 and ADH would be up-regulated in catalase-deficient animals, but it is not clear yet. It is well-established that >90% of ethanol metabolism occurs via ADH in the liver. CYP2E1, a microsomal oxidative enzyme, has a small contribution to the total metabolism (about 2 or 3%) in acute ethanol ingestion in the brain. Eriksson: Many years ago we investigated the break point, when acetaldehyde is also found in the brain. You mentioned 70 µM. This may be quite close to the cut-off point after which you will have increasing concentrations in the brain. However, under this level, the capacity for removing acetaldehyde at the blood– brain barrier is very efficient. However, there are parts of the brain where there isn’t such an efficient barrier, such as the hypothalamic area. In those individuals who are heterozygotes acetaldehyde can get up to levels which get into the brain. Under normal conditions we don’t find so much except maybe in the hypothalamic area. Worrall: Isn’t there some evidence that high blood alcohol levels make this barrier leaky? High acetaldehyde plus high alcohol might lead to more acetaldehyde getting in. Eriksson: That’s an interesting possibility. To my knowledge we don’t have any human data on this. Worrall: I think there’s some literature on this (see for example Haorah et al 2005). References Albano E 2002 Free radicals and alcohol-induced liver injury. In: Sherman CDIN, Preedy VR, Watson RR (eds) Ethanol and the liver. Taylor and Francis, London, p 153–190 Albano E, Clot P, Comoglio A, Dianzani MU, Tomasi A 1994 Free radical activation of acetaldehyde and its role in protein alkylation. FEBS Lett 348:65–69 Alderman J, Kato S, Lieber CS 1989 The microsomal ethanol oxidizing system mediates metabolic tolerance to ethanol in deermice lacking alcohol dehydrogenase. Arch Biochem Biophys 271:33–39 Amamoto K, Okamura T, Tamaki S et al 2002 Epidemiologic study of the association of low-Km mitochondrial acetaldehyde dehydrogenase genotypes with blood pressure level and the prevalence of hypertension in a general population. Hypertens Res 25:857–864

ACETALDEHYDE REMOVAL

51

Arteel GE 2003 Oxidant and antioxidant in alcohol-induced liver disease. Gastroenterology 124:778–790 Bailey SM, Cunningham CC 2002 Contribution of mitochondria to oxidative stress associated with alcohol liver disease. Free Radic Biol Med 32:11–16 Burnett KG, Felder MR 1978 Genetic regulation of liver alcohol dehydrogenase in peromyscus. Biochem Genet 16:443–454 Caro AA, Cederbaum AI 2004 Oxidative stress, toxicology and pharmacology of CYP2E1 Ann Rev Pharmacol Toxicol 44:27–42 Doorn JA, Hurley TD, Petersen DR 2006 Inhibition of human mitochondrial aldehyde dehydrogenase by 4-hydroxynon-2-enal and 4-oxonon-2-enal. Chem Res Toxicol 19:102–110 Haorah J, Knipe B, Leibhart J, Ghorpade A, Persidsky Y 2005 Alcohol-induced oxidative stress in brain endothelial cells causes blood-brain barrier dysfunction. J Leukocyte Biol 78:1223–1232 Hoek JB, Cahill A, Pastorino JG 2002 Alcohol and mitochondria: a dysfunctional relationship. Gastroenterology 122:2049–2063 Koll M, Beeso JA, Kelly FJ et al 2003 Chronic alpha-tocopherol supplementation in rats does not ameliorate either chronic or acute alcohol-induced changes in muscle protein metabolism. Clin Sci (Lond) 104:287–294 Launoy G, Milan C, Day NE, Faivre J, Pienkowski P, Gignoux M 1997 Oesophageal cancer in France: potential importance of hot alcoholic drinks. Int J Cancer 71:917–923 Luczak SE, Elvine-Kreis B, Shea SH, Carr LG, Wall TL 2002 Genetic risk for alcoholism relates to level of response to alcohol in Asian-American men and women. J Stud Alcohol 63:74–82 Luo HR, Israel Y, Tu GC, Eriksson CJ, Zhang YP 2005 Genetic polymorphism of aldehyde dehydrogenase 2 (ALDH2) in a Chinese population: gender, age, culture, and genotypes of ALDH2. Biochem Genet 43:223–227 Morris JB 1997 Dosimetry, toxicity and carcinogenicity of inspired acetaldehyde in the rat. Mutat Res 380:113–124 Nordmann R, Ribière C, Rouach H 1992 Implication of free radical mechanisms in ethanol induced cellular injury. Free Radic Biol Med 12:219–240 Ohta S, Ohsawa I, Kamino K, Ando F, Shimokata H 2004 Mitochondrial ALDH2 deficiency as an oxidative stress. Ann NY Acad Sci 1011:36–44 Okamura T, Tanaka T, Yoshita K et al 2004 Specific alcoholic beverage and blood pressure in a middle-aged Japanese population: the High-risk and Population Strategy for Occupational Health Promotion (HIPOP-OHP) Study. J Hum Hypertens 18:9–16 Ostrowska J, Luczaj W, Kasacka I, Rózanski A, Skrzydlewska E 2004 Green tea protects against ethanol-induced lipid peroxidation in rat organs. Alcohol 32:25–32 Reilly ME, Patel VB, Peters TJ, Preedy VR 2000 In vivo rates of skeletal muscle protein synthesis in rats are decreased by acute ethanol treatment but are not ameliorated by supplemental alphatocopherol. J Nutr 130:3045–3049 Shaw S, Jayatilleke E 1990 The role of aldehyde oxidase in ethanol-induced hepatic lipid peroxidation in the rat. Biochem J 268:579–583

Acetaldehyde, polymorphisms and the cardiovascular system Shih-Jiun Yin and Giia-Sheun Peng* Departments of Biochemistry and *Neurology, National Defense Medical Center, 161 Min-Chuan East Road Section 6, Taipei 114, Taiwan

Abstract. To date, the only genes that have been consistently replicated across racial and ethnic groups to influence alcoholism vulnerability are polymorphisms in the alcoholmetabolizing enzymes, i.e. cytosolic alcohol dehydrogenase 1B (ADH1B) and mitochondrial aldehyde dehydrogenase 2 (ALDH2). Both the variant ADH1B *2 and ALDH2 *2 alleles significantly protect against developing alcoholism. The protection has been thought to result from accumulation of acetaldehyde after drinking. Unlike ALDH2 *2, direct correlation between ADH1B *2 and blood acetaldehyde has not been verified. ALDH2 *2/ *2 homozygosity appeared to almost completely protect against alcoholism, whereas ALDH2 *1/ *2 heterozygosity appeared to reduce risk of the disease only about threefold. Direct correlations of blood ethanol and acetaldehyde concentrations, cardiovascular haemodynamic responses, and the subjective perceptions after challenge with low (0.2 g/kg) to moderate (0.5 g/kg) alcohol in individuals with different ALDH2 genotypes support the notion that full protection against alcoholism by ALDH2 *2/ *2 may derive from either abstinence or deliberate moderation in alcohol consumption due to strong discomfort from physiological and psychological responses caused by persistently elevated blood acetaldehyde after ingestion of a small amount of alcohol, and that the partial protection by ALDH2 *1/ *2 can be ascribed to significantly lower acetaldehyde build-up in blood and the according adverse reactions. 2007 Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Foundation Symposium 285) p 52–68

Alcoholism is a multifactorial, polygenic behavioural disorder involving complex gene–gene and gene–environment interactions (Li 2000). The pharmacological and toxicological effects of ethanol depend on the duration of exposure and the concentrations of ethanol and its metabolite acetaldehyde in body fluids and tissue. Alcohol dehydrogenase (ADH; EC 1.1.1.1, alcohol:NAD+ oxidoreductase) and aldehyde dehydrogenase (ALDH; EC 1.2.1.3, aldehyde:NAD+ oxidoreductase) are the principal enzymes responsible for ethanol metabolism in humans, catalysing the conversion of ethanol to acetaldehyde, then acetaldehyde to acetate, respectively (see this volume: Crabb & Liangpunsakul 2007, Deitrich et al 2007). Interestingly, both ADH and ALDH exhibit genetic polymorphisms, which may 52

POLYMORPHISMS AND CARDIOVASCULAR SYSTEM

53

influence drinking behaviour as well as vulnerability to development of alcohol dependence (Yin & Peng 2005, Yin et al 2006). In this chapter, we present the distribution of functional allelic variations of the alcohol-metabolism genes among racial populations, the protection against developing alcoholism by variant alleles ADH1B *2 and ALDH2 *2 in East Asians, and the pharmacokinetic and pharmacodynamic consequences in individuals with different ALDH2 genotypes following a low to moderate dose of ethanol, with emphasis on cardiovascular haemodynamic effects and the subjective perceptions. This work provides a physiological perspective for the roles of acetaldehyde in full protection by ALDH2 *2/ *2 homozygosity versus partial protection by ALDH2 *1/ *2 heterozygosity against development of alcoholism. Experimental procedures Recruitment of subjects and diagnosis of alcohol dependence were as described previously (Chen et al 1999a,b, Peng et al 1999, 2002). Determination of functional single-nucleotide polymorphic sites at the ADH1B, ADH1C and ALDH2 genes was carried out as described previously (Chen et al 1999a). Blood ethanol was determined by gas chromatography and blood acetaldehyde was determined as a fluorescent adduct by high-performance liquid chromatography (Peng et al 1999). Parameters for cardiac function before and after ingestion of alcohol were measured in the supine position by M-mode, two-dimensional Doppler echocardiography (Peng et al 1999). Systolic and diastolic pressure of the left brachial artery was measured using a sphygmomanometer. Extracranial and intracranial arterial blood flow were measured by Doppler ultrasonography (Peng et al 1999). Assessment of subjective responses to alcohol was performed by slight modification of the Subjective High Assessment Scale (SHAS) by Schuckit (1984) as described previously (Peng et al 1999). The chi-square test and logistic-regression analysis for genotype data were performed using the SPSS statistics program (SPSS, Chicago, USA). The pharmacokinetic and pharmacodynamic data were analysed by SPSS one-way ANOVA with the Scheffe’s test or with the Tamhane’s T2 test and the subjective self-ratings were evaluated by SPSS nonparametric Mann– Whitney test (Peng et al 1999). Values for the data are expressed as mean ± SEM. Results and discussion Functional polymorphism and alcoholism Human ADH constitutes a complex enzyme family (Duester et al 1999, Yin et al 2006). Class I ADH1A, ADH1B, and ADH1C, and class II ADH contribute to

54

YIN & PENG

hepatic metabolism of ethanol, whereas Class I ADH1C and Class IV ADH participate in the gastric metabolism (Lee et al 2006). Class III ADH may be involved in gastrointestinal first-pass metabolism at high ethanol concentrations (Lee et al 2003, 2006). Allozymes ADH1B2 and ADH1B3, encoded by variant alleles ADH1B *2 and ADH1B *3 respectively, exhibit much higher activity than that of ADH1B1. ADH1C1, encoded by ADH1C *1, exhibits a higher activity than ADH1C2. Table 1 shows the distribution of ADH1B and ADH1C gene alleles in racial populations. ADH1B *1 is a predominant allele among the Caucasians, African Americans and American Indians, whereas ADH1B *2 is predominant among the East Asians including Han Chinese, Japanese, Koreans, and Taiwanese aborigines. ADH1B *3 occurs uniquely in the African populations (Osier el al 2002). It is interesting to note that among Caucasians, the Jews and Russians appear to have considerably higher frequencies of ADH1B *2, ∼20% (Hasin et al 2002) and 41% (Ogurtsov et al 2001), respectively. ADH1C *1 is predominant among the East Asians, African Americans and Taiwanese aborigines, but ADH1C *1 and ADH1C *2 are about equally distributed among the Caucasians and American Indians. Cytosolic ALDH1 and mitochondrial ALDH2 are the principal members of ALDH superfamily responsible for oxidation of acetaldehyde in mammals (this volume: Deitrich et al 2007). There is a common point mutation of the ALDH2 gene, denoted ALDH2 *2, which results in a glutamic acid to lysine substitution (Agarwal & Goedde 1992). Allele frequencies of ALDH2 *2 are relatively high in the Vietnamese (35%), Han Chinese (24%), Japanese (24%), and Koreans (16%) (Chen et al 1994, Higuchi et al 1995, Lee et al 1997, Chen et al 1999a), but rare in other ethnic groups including Caucasians, black Americans, American Indians, and TABLE 1 Allele frequencies of ADH1B and ADH1C in racial populations Allele frequency Population East Asians Caucasians African Americans American Indians Taiwanese Aborigines

ADH1B *1

ADH1B *2

∼0.30 ∼0.90

∼0.70 ∼0.10

0.84 0.93

0 0.01

0.09

0.91

ADH1B *3

ADH1C *1

ADH1C *2

∼0.90 ∼0.60

∼0.10 ∼0.40

0.16 0.06

0.85 0.56

0.15 0.44

0

0.99

0.01

0 0

Data are taken from Agarwal & Goedde (1992), Borràs et al (2000), Burnell & Bosron (1989), Chen et al (1999a), Higuchi et al (1995), Thomasson et al (1994) and Wall et al (1997). The ADH1B and ADH1C genes were previously denoted ADH2 and ADH3, respectively. For new nomenclature of human ADH gene family, see Duester et al (1999).

POLYMORPHISMS AND CARDIOVASCULAR SYSTEM

55

Polynesians (Yin & Peng, 2005). Interestingly, autopsy livers from individuals homozygous for ALDH2 *2 or from those heterozygous for ALDH2 *2 and ALDH2 *1 lacked ALDH2 activity on starch gel electrophoresis, suggesting a dominance on loss of the enzyme activity by the variant allele (Crabb et al 1989). Kinetic and x-ray crytallographic studies support a partial dominance model of a dimer-ofdimers molecule (Larson et al 2005). The model predicts that residual activities of ALDH2 in mutant homozygotes and the heterozygotes would account for 0% and 25%, respectively, of the total activity in normal homozygotes. A recent study with surgical liver specimens reported that a considerably lower ALDH2 activity was found in the heterozygotes as compared to that in normal homozygotes, and that no measurable activity of the enzyme was found in mutant homozygotes (Wang et al 2002). We have genotyped allelic variations of ADH1B, ADH1C and ALDH2 from leukocyte DNA samples of 340 Han Chinese alcohol-dependent subjects and 545 non-alcoholic controls in Taiwan (Chen et al 1999a). The frequency of ADH1B *2 allele was significantly lower in alcoholics than that in controls, suggesting a protection against alcoholism by the variant allele. The ADH1C *1 allele frequency was also found significantly decreased in the alcoholic group but further haplotype analysis indicated that this is virtually due to linkage disequilibrium between the ADH1C *1 and ADH1B *2 alleles in ADH gene cluster on chromosome 4. The frequency of ALDH2 *2 variant allele decreased threefold in alcoholics compared to that in controls (Chen et al 1999a). Strikingly, no single alcoholic patient with homozygous ALDH2 *2/ *2 was found in this study. The reduction of risk for alcoholism by ADH1B *2 has also been found in the Japanese (Higuchi et al 1995), Atayal Natives of Taiwan (Thomasson et al 1994) and Caucasians (Borràs et al 2000, Ogurtsov et al 2001, Hasin et al 2002). Recent haplotype/diplotype studies examining a battery of single-nucleotide polymorphisms across the seven ADH genes in Caucasian alcoholics and controls indicate that class II, III and IV ADH genes may also independently influence susceptibility to alcohol dependence (Edenberg et al 2006, Luo et at 2006). The underlying functional causality for these candidate genes, however, remains unclear. It is noteworthy that Higuchi et al (1994) surveyed 1300 Japanese alcoholdependent subjects and found no single patient identified with ALDH2 *2/ *2. On the basis of genotype frequencies in the Japanese general population, it is expected 118 of the above alcoholic individuals would be ALDH2 *2 homozygotes. Thus, genetic epidemiological studies strongly suggest that homozygosity of variant ALDH2 *2 may fully protect against alcoholism both in Japanese and Han Chinese (Table 2). Intriguingly, the frequency of heterozygous ALDH2 *1/ *2 genotype varied from 2.5 to 13% for the Japanese alcoholics and from 10 to 18% for the Han Chinese alcoholics surveyed from different registry periods with a trend of higher frequency in more recent years (Table 2). This finding suggests that the heterozygos-

56

YIN & PENG

TABLE 2 Protection against alcoholism by heterozygosity of ALDH2*1/*2 in East Asians Genotype frequency Alcoholics Japanese Han Chinese

Patients (n)

Registry period

ALDH2*1/*1

ALDH2*1/*2

ALDH2*2/*2

400 400 500 49 291

1979 1986 1992 1989–90 1994–7

0.975 0.92 0.87 0.90 0.82

0.025 0.08a 0.13b 0.10 0.18

0 0 0 0 0

Data for the Japanese and Han Chinese alcohol-dependent patients are taken from Higuchi et al (1994) and Chen et al (1999a), respectively. a P < 0.001 (1986 versus 1979); b P < 0.005 (1992 versus 1986).

ity can only afford partial protection against alcoholism, permitting other biological and environmental factors to have an influence. Unexpectedly, one case of a homozygous ALDH2 *2 alcoholic patient was discovered in a survey study in Taiwan (Chen et al 1999b). This patient displayed a unique drinking pattern to accommodate his inborn error of acetaldehyde metabolism, i.e. (a) drinking beer, instead of wine or spirit liquors, as his favourite alcoholic beverage; (b) sipping alcoholic beverages almost continuously throughout the day rather than fast, binge drinking; (c) consuming relatively low amounts of alcohol with three to five bottles (i.e. 350 ml of 4.5% by volume of ethanol or 12.4 g of ethanol per bottle) of beer per day. To address the question of interaction between functional polymorphisms of ADH1B and ALDH2 in protecting against alcoholism, we performed logistic regression analysis of the combinational genotypes for Han Chinese alcohol-dependent and control subjects (Chen et al 1999b). The risks for alcoholism are in the following order:ADH1B*2/*2–ALDH2*2/*2
POLYMORPHISMS AND CARDIOVASCULAR SYSTEM

57

TABLE 3 Blood ethanol and acetaldehyde concentrations 30 min after complete oral administration of ethanol to men with different ALDH2 genotypes Blood ethanol (mM) Ethanol dose (g/kg) 0.2 0.3 0.5

Blood acetaldehyde (mM)

ALDH2

ALDH2

Subjects (n)

*1/*1

*1/*2

*2/*2

*1/*1

*1/*2

*2/*2

6 6 8

1.7 ± 0.2 7.1 ± 0.6 8.1 ± 1.3

2.6 ± 0.3 8.2 ± 0.5 12.2 ± 1.3

3.5 ± 0.5a —d —d

0.2 ± 0.2 0.2 ± 0.2 2.4 ± 0.6

16.1 ± 1.9b 59.7 ± 14.3a 67.4 ± 5.1b

63.7 ± 10.1a,c —d —d

All subjects had the genotype of ADH1B *2/ *2 and ADH1C *1/ *1. Values are mean ± SEM. a P < 0.01 versus ALDH2 *1/ *1; b P < 0.001 versus ALDH2 *1/ *1; c P < 0.05 versus ALDH2 *1/ *2; d not determined due to these doses too being high to be acceptable to the ALDH2 *2/ *2 subjects. Data are adapted from Peng et al (1999, 2002), and taken from G.-S. Peng & S.-J. Yin (unpublished work 2006).

lated pharmacokinetics of ethanol and its metabolite acetaldehyde with the pharmacodynamic effects and psychological responses in age and body-mass index matched male healthy Han Chinese with different ALDH2 genotypes controlling for the ADH genotypes of ADH1B *2 and ADH1C *1, after ingestion of a low to moderate dose of ethanol (i.e. 0.2, 0.3 or 0.5 g/kg) (Peng et al 1999, 2002, G.-S. Peng & S.-J. Yin, unpublished work 2006). Following challenge with a small amount of ethanol (0.2 g/kg, roughly equivalent to a bottle of beer), the variant homozygous ALDH2 *2 individuals exhibited persistently and significantly higher blood acetaldehyde levels than those of the heterozygous and the normal homozygous ALDH2 *1 individuals (Peng et al 1999). Interestingly, two hours after drinking such a small amount of alcohol, blood acetaldehyde levels (17.0 ± 2.7 µM) in variant homozygotes were still similar to that of the peak concentration (23.6 ± 1.3 µM) in heterozygotes. In contrast, normal homozygotes exhibited near zero blood acetaldehyde. The area under the blood acetaldehyde concentration–time curve of variant homozygotes appeared to be five- and 220-fold greater than that of the heterozygotes and normal homozygotes, respectively. Blood ethanol levels in the ALDH2*2 homozygotes were significantly higher than those in the ALDH2*1 homozygotes (Peng et al 1999), suggesting a slower ethanol elimination in variant homozygotes than can be attributed to product inhibition of ADH activity by acetaldehyde. The blood ethanol and acetaldehyde peaked between 20–30 min after finishing ingestion of ethanol. (It took 10 min with 2 min intervals for a total of 5 aliquots of 10% ethanol solution freshly prepared in natural orange juice.) Table 3 shows the blood ethanol and acetaldehyde concentrations 30 min after complete oral administration of various doses of alcohol.

58

YIN & PENG

The peak blood acetaldehyde levels for individuals with heterozygous ALDH2 *1/ *2 and normal homozygous ALDH2 *1/ *1 were determined to be 59.7 ± 14.3 µM and 0.17 ± 0.17 µM, respectively, after 0.3 g/kg ethanol (Peng et al 2002), and 74.4 ± 2.4 and 4.0 ± 0.6 µM, respectively, after 0.5 g/kg ethanol (G.-S. Peng & S.-J. Yin, unpublished work 2006). The peak ethanol concentrations were not significantly different between these two ALDH2 genotypes following a dose of either 0.2, 0.3 or 0.5 g/kg ethanol. It is notable that oral administration of 0.5 g/kg to the ALDH2 *1/ *2 heterozygotes yielded a peak blood acetaldehyde very similar to that observed in the ALDH2*2 homozygotes (75.4 ± 10.6 µM) receiving 0.2 g/kg ethanol (Peng et al 1999). This would enable us to reasonably compare the pharmacodynamic effects of acetaldehyde between these two genotypes. The pharmacodynamic effects of ethanol per se after different doses can be evaluated from those observed for ALDH2 *1 homozygotes who always had very low blood acetaldehyde. The peak blood acetaldehyde level were not significantly different between ADH1B *1 homozygotes and ADH1B *2 homozygotes both either with ALDH2 *1/ *1 or with ALDH2 *1/ *2 after 0.3 g/kg ethanol (Peng et al 2002), and both with ALDH2*1/*1 after 0.5 g/kg ethanol (G.-S. Peng & S.-J. Yin, unpublished work 2006). This indicates that functional polymorphism of the ADH1B gene did not appreciably contribute to the accumulation of blood acetaldehyde, a similar finding to that reported for Japanese after an oral dose of 0.4 g/kg ethanol (Mizoi et al 1994). Pharmacodynamics of acetaldehyde Alcohol sensitivity commonly seen in East Asians has been attributed to the acetaldehyde-induced symptoms, including facial flushing, tachycardia, orthostatic hypotension, dizziness, headache, nausea, vomiting (Yin & Peng 2005). An elevation of plasma catecholamines following ingestion of alcoholic beverage was found in Japanese individuals deficient with ALDH2 activity (Mizoi et al 1989) and in Caucasian individuals treated with the alcohol aversive drug nitrefazole, an ALDH inhibitor (Suokas et al 1985). It has been proposed that the elevation of plasma catecholamines is a compensatory reaction to the acetaldehyde-induced decreases in diastolic blood pressure and total peripheral resistance (Mizoi et al 1989). The mechanisms underlying acetaldehyde-induced vasodilation are not precisely known. Both central (neurally mediated vasodilation) and peripheral (local vasodilation due to circulating substances) mechanisms may be involved (Yin & Peng 2005). As expected, East Asians possessing variant ALDH2 *2 alleles, a built-in Antabuse, consumed significantly smaller amounts of alcohol and showed less heavy episodic drinking as compared with the normal homozygotes (Muramatsu et al 1995, Luczak et al 2001) due to the general discomforts of their alcohol sensitivity reaction. Alterations in cardiovascular haemodynamic responses and the subjective perceptions are shown in Figs. 1 and 2, respectively, for men with different ALDH2

POLYMORPHISMS AND CARDIOVASCULAR SYSTEM 60

A

20

b

59

C

Diastolic blood pressure (% change)

c

Heart rate (% change)

40 b 20

0

b c

60 a 40

20

0

0.5 g/kg 0.2 g/kg Ethanol dose

Mean velocity of facial artery (% change)

B 80

Cardiac output (% change)

0

-10 a -20

-20

-20

10

d b

D 160 c,d 120

80

40

0

-40

0.5 g/kg 0.2 g/kg Ethanol dose

FIG. 1. Alterations in heart rate (A), cardiac output (B), diastolic blood pressure (C), and mean velocity of facial artery (D) in men with different ALDH2 genotypes 30 min after oral administration of ethanol. ALDH2 *1/ *1 (open bar); ALDH2 *1/ *2 (hatched bar); ALDH2 *2/ *2 (closed bar). Bars represent mean ± SEM. For subject number in each genotypic group and the corresponding blood ethanol and acetaldehyde levels, see Table 3. Percentage changes are differences between the values obtained after alcohol ingestion, minus that of preingestion divided by the latter value and multiplied by 100. aP < 0.05 versus ALDH2 *1/ *1 treated with the same dose; b P < 0.01 versus ALDH2 *1/ *1 treated with the same dose; cP < 0.001 versus ALDH2 *1/ *1 treated with the same dose; dP < 0.01 versus ALDH2 *1/ *2 treated with the same dose. No significant difference between ALDH2 *2/ *2 (0.2 g/kg) and ALDH2 *1/ *2 (0.5 g/kg) was found. Data are adapted from Peng et al (1999), and taken from G.-S. Peng & S.-J. Yin (unpublished work 2006).

genotypes at 30 min after administration of a low (0.2 g/kg) or a moderate (0.5 g/kg) dose of ethanol, which correspond to the blood acetaldehyde concentrations in Table 3. Following challenge with a low dose (0.2 g/kg ethanol), compared to normal ALDH2 *1 homozygotes the ALDH2 *2 homozygotes exhibited significant increases in heart rate and cardiac output during the postingestion period of 120 min, and in mean velocity of facial artery between 10 and 60 min postingestion, and significant decreases in diastolic blood pressure between 10 and 30 min postingestion (Peng et al 1999). The heterozygotes exhibited considerably less intense cardiovascular responses than did mutant homozygotes. This can be ascribed to the much

60

YIN & PENG 3

A

Palpitations (relative rating unit)

b 3

b

2

1

Terrible feelings overall (relative rating unit)

4

C a b

2

1

0

0 -1

B Facial warming (relative rating unit)

b

2

a

1

Great feelings overall (relative rating unit)

3 a

3

D

2

1

0

-1

b

0 -2

0.5 g/kg 0.2 g/kg Ethanol dose

0.5 g/kg 0.2 g/kg Ethanol dose

FIG. 2. Responses in subjective perceptions of palpitation (A), facial warming (B), terrible feelings overall (C), and great feelings overall (D) in men with different ALDH2 genotypes 30 min after oral administration of ethanol. ALDH2 *1/ *1 (open bar); ALDH2 *1/ *2 (hatched bar); ALDH2 *2/ *2 (closed bar). Bars represent mean ± SEM. For subject number in each genotypic group and the corresponding blood ethanol and acetaldehyde levels, see Table 3. aP < 0.05 versus ALDH2 *1/ *1 treated with the same dose; bP < 0.01 versus ALDH2*1/*1 treated with the same dose. No significant difference between ALDH2 *2/ *2 (0.2 g/kg) and ALDH2 *1/ *2 (0.5 g/kg) was found. Data are adapted from Peng et al (1999), and taken from G.-S. Peng & S.-J. Yin (unpublished work 2006).

lower blood acetaldehyde levels in the former genotype. Consistent with the observed haemodynamic effects, ALDH2 *2 homozygotes exhibited significant subjective perceptions of palpitation, facial warming and terrible feeling overall during 120 min postingestion compared to that perceived by ALDH1 *1 homozgotes (Peng et al 1999). The heterozygotes only significantly perceived facial warming 10–30 min postingestion that corresponded to the period of high blood acetaldehyde levels. Thus, the striking responses to a small amount of ethanol, as evidenced by the pronounced cardiovascular hemodynamic effects as well as subjective feelings of general discomfort for as long as 2 hours following ingestion, constitute the physiological basis for strong protection against alcoholism by ALDH2 *2 homozygosity.

POLYMORPHISMS AND CARDIOVASCULAR SYSTEM

61

Following challenge with a moderate dose of ethanol (0.5 g/kg), the ALDH2 *1/ *2 heterozygotes exhibited significant increases in heart rate, cardiac output and mean velocity of facial artery, and significant decreases in diastolic blood pressure compared to the normal homozygotes (G.-S. Peng & S.-J. Yin, unpublished work 2006). The heterozygotes also significantly perceived palpitations, facial warming, and terrible feelings overall but not great feelings overall. Thus, heterozygous individuals, receiving 0.5 g/kg ethanol, exhibited similar pharmacokinetics of blood acetaldehyde, cardiovascular responses and subjective feelings compared to the mutant homozygotes receiving 0.2 g/kg ethanol. The 2.5-fold higher ethanol dosage to attain a similar pharmacokinetic and pharmacodynamic consequence of acetaldehyde may explain the partial protection against alcoholism by ALDH2 *1/ *2 heterozygosity. Conclusion The functional polymorphisms of alcohol-metabolism genes, ADH1B and ALDH2, exhibit a complex pattern of influences on vulnerability to alcoholism in East Asians. The full protection against the disease by ALDH2 *2 homozygosity may derive from either abstinence or deliberate moderation in alcohol consumption due to strong alcohol sensitivity reaction that is caused by a prolonged and large accumulation of acetaldehyde in blood resulting from an almost total loss of the enzyme activity in liver. The partial protection by heterozygosity can be ascribed to significantly lower accumulation of acetaldehyde in blood due to hepatic residual activity of ALDH2, which reduces the aversion reaction to alcohol and hence permits other biological as well as environmental factors to contribute to development of the disease. The mechanism for protection against alcoholism by ADH1B polymorphism remains largely unclear and it warrants further studies. Acknowledgements The work of the authors’ laboratory was supported by grants from the National Science Council and the National Health Research Institutes, Republic of China.

References Agarwal DP, Goedde HW 1992 Pharmacogenetics of alcohol metabolism and alcoholism. Pharmacogenetics 2:48–62 Borràs E, Coutelle C, Rosell A et al 2000 Genetic polymorphism of alcohol dehydrogenase in Europeans: the ADH2*2 allele decreases the risk for alcoholism and is associated with ADH3 *1. Hepatology 31:984–989 Burnell JC, Bosron WF 1989 Genetic polymorphism of human liver alcohol dehydrogenase and kinetic properties of the isoenzymes. In: Crow KE, Batt RD (eds) Human metabolism of alcohol, vol 2. CRC Press, Boca Raton, p 65–75

62

YIN & PENG

Chen CC, Lu RB, Chen YC et al 1999a Interaction between the functional polymorphisms of the alcohol-metabolism genes in protection against alcoholism. Am J Hum Genet 65: 795–807 Chen SH, Zhang M, Wang NS, Scott CR 1994 Gene frequencies of alcohol dehydrogenase2 (ADH2) and aldehyde dehydrogenase2 (ALDH2) in five Chinese minorities. Hum Genet 94:571–572 Chen YC, Lu RB, Peng GS et al 1999b Alcohol metabolism and cardiovascular response in an alcoholic patient homozygous for the ALDH2 *2 variant gene allele. Alcohol Clin Exp Res 23:1853–1860 Crabb DW, Liangpunsakul S 2007 Acetaldehyde generating enzyme systems: roles of alcohol dehydrogenase, CYP2E1 and catalase, and speculations on the role of other enzymes and processes. In: Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Found Symp 285), p 4 –16 Crabb DW, Edenberg HJ, Bosron WF, Li TK 1989 Genotypes for aldehyde dehydrogenase deficiency and alcohol sensitivity: the inactive ALDH2(2) allele is dominant. J Clin Invest 83:314 –316 Deitrich RA, Petersen D, Vasiliou V 2007 Overview of aldehyde degrading enzymes. In: Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Found Symp 285), p 23 – 40 Duester G, Farrés J, Felder MR et al 1999 Recommended nomenclature for the vertebrate alcohol dehydrogenase gene family. Biochem Pharmacol 58:389–395 Edenberg HJ, Xuei X, Chen HJ et al 2006 Association of alcohol dehydrogenase genes with alcohol dependence: a comprehensive analysis. Hum Mol Genet 15:1539–1549 Hasin D, Aharonovich E, Liu X et al 2002 Alcohol dependence symptoms and alcohol dehydrogenase 2 polymorphism: Israeli Ashkenazis, Sephardics, and recent Russian immigrants. Alcohol Clin Exp Res 26:1315–1321 Higuchi S, Matsushita S, Imazeki H, Kinoshita T, Takagi S, Kono H 1994 Aldehyde dehydrogenase genotypes in Japanese alcoholics. Lancet 343:741–742 Higuchi S, Matsushita S, Murayama M, Takagi S, Hayashida M 1995 Alcohol and aldehyde dehydrogenase polymorphisms and the risk for alcoholism. Am J Psychiatry 152:1219–1221 Larson HN, Weiner H, Hurley TD 2005 Disruption of the coenzyme binding site and dimer interface revealed in the crystal structure of mitochondrial aldehyde dehydrogenase ‘Asian’ variant. J Biol Chem 280:30550–30556 Lee KH, Kwak BY, Kim JH, Yoo SK, Yum SK, Jeong HS 1997 Genetic polymorphism of cytochrome P-4502E1 and mitochondrial aldehyde dehydrogenase in a Korean population. Alcohol Clin Exp Res 21:953–956 Lee SL, Wang MF, Lee AI, Yin SJ 2003 The metabolic role of human ADH3 functioning as ethanol dehydrogenase. FEBS Lett 544:143–147 Lee SL, Chau GY, Yao CT, Wu CW, Yin SJ 2006 Functional assessment of human alcohol dehydrogenase family in ethanol metabolism: significance of first-pass metabolism. Alcohol Clin Exp Res 30:1132–1142 Li TK 2000 Pharmacogenetics of responses to alcohol and genes that influence alcohol drinking. J Stud Alcohol 61:5–12 Luczak SE, Wall TL, Shea SH, Byun SM, Carr LG 2001 Binge drinking in Chinese, Korean, and White college students: genetic and ethnic group differences. Psychol Addict Behav 15:306–309 Luo X, Kranzler HR, Zuo L, Wang S, Schork NJ, Gelernter J 2006 Diplotype trend regression analysis of the ADH gene cluster and the ALDH2 gene: multiple significant associations with alcohol dependence. Am J Hum Genet 78:973–987 Mizoi Y, Fukunaga T, Adachi J 1989 The flushing syndrome in Orientals. In: Crow KE, Batt RD (eds) Human metabolism of alcohol, vol. 2. CRC Press, Boca Raton, p 219–229

POLYMORPHISMS AND CARDIOVASCULAR SYSTEM

63

Mizoi Y, Yamamoto K, Ueno Y, Fukunaga T, Harada S 1994 Involvement of genetic polymorphism of alcohol and aldehyde dehydrogenases in individual variation of alcohol metabolism. Alcohol Alcohol 29:707–710 Muramatsu T, Wang ZC, Fang YR et al 1995 Alcohol and aldehyde dehydrogenase genotypes and drinking behavior of Chinese living in Shanghai. Hum Genet 96:151–154 Ogurtsov PP, Garmash IV, Miandina GI, Guschin AE, Itkes AV, Moiseev VS 2001 Alcohol dehydrogenase ADH2-1 and ADH2-2 allelic isoforms in the Russian population correlate with type of alcoholic disease. Addict Biol 6:377–383 Osier MV, Pakstis AJ, Soodyall H et al 2002 A global perspective on genetic variation at the ADH genes reveals unusual patterns of linkage disequilibrium and diversity. Am J Hum Genet 71:84–99 Peng GS, Wang MF, Chen CY et al 1999 Involvement of acetaldehyde for full protection against alcoholism by homozygosity of the variant allele of mitochondrial aldehyde dehydrogenase gene in Asians. Pharmacogenetics 9:463–476 Peng GS, Yin JH, Wang MF, Lee JT, Hsu YD, Yin SJ 2002 Alcohol sensitivity in Taiwanese men with different alcohol and aldehyde dehydrogenase genotypes. J Formos Med Assoc 101:769–774 Schuckit MA 1984 Subjective responses to alcohol in sons of alcoholics and control subjects. Arch Gen Psychiatry 41:879–884 Suokas A, Kupari M, Pettersson J, Lindros K 1985 The nitrefazole-ethanol interaction in man: cardiovascular responses and the accumulation of acetaldehyde and catecholamines. Alcohol Clin Exp Res 9:221–227 Thomasson HR, Crabb DW, Edenberg HJ et al 1994 Low frequency of the ADH2 *2 allele among Atayal natives of Taiwan with alcohol use disorders. Alcohol Clin Exp Res 18:640–643 Wall TL, Garcia-Andrade C, Thomasson HR, Carr LG, Ehlers CL 1997 Alcohol dehydrogenase polymorphisms in Native Americans: identification of the ADH2 *3 allele. Alcohol Alcohol 32:129–132 Wang RS, Nakajima T, Kawamoto T, Honma T 2002 Effects of aldehyde dehydrogenase-2 genetic polymorphisms on metabolism of structurally different aldehydes in human liver. Drug Metab Dispos 30:69–73 Yin SJ, Peng GS 2005 Overview of ALDH polymorphism: relation to cardiovascular effects of alcohol. In: Preedy VR, Watson RR (eds) Comprehensive handbook of alcohol related pathology, vol. 1. Elsevier Science Ltd, London, p 411–426 Yin SJ, Lee SL, Han CL, Chou CF, Wang MF 2006 Pharmacogenetic determinants of alcohol metabolism and alcoholism in the human alcohol dehydrogenase family. In: Weiner H, Lindahl R, Plapp BV, Maser E (eds) Enzymology and molecular biology of carbonyl metabolism 12. Purdue University Press, West Lafayette, p 161–170

DISCUSSION Crabb: I’ll throw this question open to anyone who works in the Far East. The haemodynamic responses you get suggest that there ought to be beer-induced angina pectoris in people who have greatly increased cardiac output and tachycardia and low diastolic blood pressure, for people with fixed coronary disease. Do people say that when they drink they get chest pain? Yin: No. Alcohol-induced chest pain appears very rarely in ALDH2-deficient East Asians with coronary heart disease. It seems that this can be countered by the increased cardiac output, i.e. increased cardiac perfusion. As to the reason why

64

YIN & PENG

homozygotic individuals can’t tolerate alcohol, I think it is due to diastolic pressure drop. It makes them feel dizzy so they just can’t drink that much. This is caused by the acetaldehyde-induced peripheral vasodilatation. Crabb: In the patient homozygous for ALDH2 *2, how was the diagnosis of alcoholism made? You said he drinks three or four bottles of beer a day, but he probably doesn’t ever have high blood alcohol levels. Why is he alcoholic? Yin: This is based on the diagnosis from DSM-III-R. The patient showed physical and psychological dependence with acute-stage withdrawal symptoms, including insomnia, tremor, anxiety and agitation (Chen et al 1999). Apte: People like us who are looking at alcohol-related organ damage think of alcoholics as at least drinking 80 g alcohol per day for a long time, but your definition of alcohol dependence seems to be more psychological. Do we need to make a differentiation here in terms of definitions of alcoholism? Yin: That’s a good question. This diagnosis was by a psychiatrist, and the patient still has normal liver function. Emery: Were there any other indicators of alcohol-induced damage? Yin: Not for this patient. Actually, we select this as a criterion: if the liver functions have already been affected, then it is not longer possible to compare the pharmacokinetics. Preedy: Such individuals are quite rare: you have one individual. How rare is rare in this case? Yin: For the ALDH2 *2 homozygotes the frequency is only about 5% in the general population of Taiwan. Usually in East Asia when friends get together around a large table, about one will not drink alcohol at all. These are likely to be homozygous for ALDH2 *2. Preedy: Of those, how many would have a diagnosis of alcoholism by addiction criteria? Presumably they would be quite rare. Yin: Yes. About 10% alcoholics in Japan and Taiwan are heterozygous. This is interesting: apparently, they overcome the protection this genotype would normally give against alcoholism by other positive candidate genes for alcoholism, plus social and environmental factors. These patients also showed very high blood acetaldehyde levels. It will be interesting to see the acetaldehyde-related disease in those patients. Eriksson: We all know that psychiatrists might be a little indoctrinated! This is not the right way of applying the DSM-III-R criteria. This guy is definitely not addicted to alcohol, but to acetaldehyde. Professor Jean-Pierre von Wartburg has another term: ‘acetaldehydic’. We shouldn’t talk about someone as an alcoholic if they never experience the intoxicating effect of alcohol. Yin: I agree with you. This homozygotic patient should be called an acetaldehydic. But other heterozygous alcoholics drink to a level comparable to that of a normal alcoholic.

POLYMORPHISMS AND CARDIOVASCULAR SYSTEM

65

Eriksson: Your purpose was to investigate why the heterozygotes are protected. There are a number of symptoms, but of these, flushing itself is not aversive. But palpitation and tachycardia are definitely aversive. I have investigated Finnish families, and out of many we found three with heterozygotes. One was a woman who came to the study because she’d been taken to hospital three times for severe tachycardia and heart arrest. The factor that is commonly aversive is nausea. In China or Japan where there are many heterozygotes, the hangover starts for these people during intoxication. It’s common to see people vomiting outside because of this nausea. In Japan, the western influence is going to increase the risk of cancer. They shouldn’t adopt our European lifestyles. Yin: The fact that is most relevant to the discomfort is the drop of diastolic pressure. This makes people feel very uncomfortable at the initial phase of drinking. M Salaspuro: It may be dangerous to change the diagnostic criteria for alcoholism on the basis of one case report. Especially when we realize how common underreporting of alcohol consumption is. I hesitate to use the amount of reported alcohol consumption as one of the critera of alcoholism At least in Finland, we then would have very few alcoholics. Underreporting is very common especially among heavy consumers. Have you tried any medication to treat the symptoms caused by acetaldehyde? They can be prevented by decreasing the rate of acetaldehyde production by inhibiting ADH with 4-methylpyrazole. Another method is to use antihistamines or beta blockers. Yin: One of my colleagues, Dr Yi-Chyan Chen, is a psychiatrist. Currently he works on disulfiram treatment. Alcoholics taking this end up with even higher blood acetaldehyde levels. This suggests that ALDH1 is also inhibited, and contributes to acetaldehyde metabolism. This is reasonable: the Km for ALDH1 is about 30 µM, which can be reached in liver. ALDH2 has a Km of 0.2 µM. Both are responsible for acetaldehyde metabolism in vivo, in human. M Salaspuro: Disulfiram is an ALDH-inhibitor and increases blood acetaldehyde. What about the other treatments, such as antihistamines or 4-methapyrazole? Yin: We haven’t tried this. Antabuse isn’t good, because it only inhibits ethanol metabolism and has nothing to do with craving. Patients still want to drink alcohol, and this can be dangerous. Dr Yi-Chyan Chen now uses a very small test amount deliberately to trick the patient, letting the patient experience the outcome of the treatment. Then they experience discomfort when they drink. M Salaspuro: We have used 4-methylpyrazole and it alleviates the acetaldehydeinduced symptoms such as flushing, tachycardia and decrease in blood pressure. Examples of this include alcoholics on disulfiram treatment who have taken some alcohol and come to the emergency clinic because of a disulfiram–alcohol reaction. Quertemont: Presumably, those patients came to you to seek help. If you were to take a sample from the general population, would you find the same degree of protection against alcoholism?

66

YIN & PENG

Yin: This is a case control study. Patients come to hospital and we then draw blood for genotyping. Control subjects in this study are the students without alcohol misuse from our medical school. Quertemont: In the general population, do patients deficient for ALDH drink lower levels of alcohol, but still drink regularly? Yin: This is true only for normal individuals in both Japanese and Chinese populations (Muramatsu et al 1995, Higuchi et al 1996). The drinking pattern for heterozygous and normal homozygous alcoholics looks similar. There was a recent paper from Dr Ru-Band Lu and co-workers (Hahn et al 2006) reporting that heterozygous alcoholics showed more positive psychological expectancies for alcohol. Eriksson: In the heterozygotes, you still see a level of about 60 µM acetaldehyde. One can calculate that based on the level of alcohol oxidation, there is still more than 95% efficiency in the oxidation. It seems that the ALDH2 is working in these people. It is not like a full inhibition, especially in the heterozygotes. Yin: Yes. Based on kinetic and molecular modelling studies, ALDH2 activity in the heterozygotes would be about 25% that of the normal homozygotes. The other thing is that acetaldehyde is a metabolite, it must derive from ethanol. If you drink more ethanol the production of acetaldehyde is still quite limited by the hepatic ADH activity. This is why ALDH2 is low Km and located in the mitochondria that are quite efficient in removing the acetaldehyde produced in the liver. Cytosolic ALDH1 can participate in acetaldehyde removal, especially in individuals with the heterozygous and mutant homozygous genotypes. Eriksson: I have three brief points. The comment by Dr Salaspuro might have been misinterpreted: I don’t think he suggested methylpyrazole for everyone. It is more like an emergency drug, because it has some toxicity. A heterozygote might like to drink all the time by using methylpyrazole chronically, but I wouldn’t encourage this! Antihistamine has been tested: it takes away the histamine component of the flushing, but there is more than just histamine involved in the mechanism behind the acetaldehyde effects. Finally, there was a fantastic paper you referred to which showed the acetaldehyde expectancy (Hahn et al 2006). If you have a genotypic difference in regard to the alcohol expectancy, you can see that the acetaldehyde causes a positive, reinforcing effect of alcohol. It was an impressive study. Yin: My personal experience also shows this. I’m heterozygous and I can drink some alcohol, but just a little. With one glass of wine I feel high, but if I drink any more than this I feel dizzy. With acetaldehyde, the open question is whether it is also a reinforcer in the CNS and not just aversive. M Salaspuro: I agree with Dr Eriksson. 4-Methylpyrazole is a generally accepted treatment form for alcohol–disulfiram reaction in many countries. In addition, 4-methylpyrazole can be used for the treatment of methanol or ethylene glycol poisoning.

POLYMORPHISMS AND CARDIOVASCULAR SYSTEM

67

Quertemont: I have a comment about the reinforcing effect of acetaldehyde. In animal studies, when the animals are given short bursts of acetaldehyde, the reinforcing effect predominates. In contrast, long-lasting brain acetaldehyde at high concentrations causes aversive effects. I’m convinced that acetaldehyde has reinforcing effects at some concentrations and some durations. Eriksson: It is a balance between the good and evil! Deitrich: In the ALDH2-negative individuals it seemed that their blood alcohol levels were much higher. Why? Yin: Yes. A reasonable explanation is acetaldehyde product inhibition. Deitrich: I can understand the explanation, but I’d like to see the mathematics of that increase. It looked like it was a greater increase than it should have been if all you are depending on is the back reaction or product inhibition. Yin: That is a good question. Human liver has so many ADH forms it is hard to do a model study with complete kinetic equations to show this kind of reverse reaction and product inhibition. Deitrich: It might be interesting to do an alcohol clamp study with ALDHnegative people. Eriksson: We have to understand that in these cases we are dealing with very low alcohol concentrations. This means that ADH activity is now rate-limiting. In normal conditions there are other factors that are rate limiting. In a condition like this, then even small effects affect the alcohol concentration. It is totally explainable by the equilibrium reaction. Okamura: Do you think that the effect of acetaldehyde on blood pressure has disappeared a few hours after alcohol ingestion in this study? Usually after 30 min or so the blood pressure level is almost the same across all genotypes, so the effect is limited to an acute one. Eriksson: There are many Japanese studies that have shown the development of tolerance to these effects. Yin: The late Professor Yasuhiko Mizoi had proposed that when ALDH2-negative individuals drink alcohol they show flushing, which means that the peripheral resistance decreases. Then catecholamines are increased which causes that cardiac output increases and heart rate increases to compensate for this. The molecular mechanisms involved in acetaldehyde-related vasodilation are still not clear. Okamura: It is strange that in many clinical studies alcohol drinking decreases blood pressure level in the short term, but many studies show that drinkers have higher blood pressure levels than non-drinkers. Yin: It’s acute versus long-term effects on the cardiovascular system. References Chen YC, Lu RB, Peng GS et al 1999 Alcohol metabolism and cardiovascular response in an alcoholic patient homozygous for the ALDH2 *2 variant gene allele. Alcohol Clin Exp Res 23:1853–1860

68

YIN & PENG

Hahn CY, Huang SY, Ko HC et al 2006 Acetaldehyde involvement in positive and negative alcohol expectancies in Han Chinese persons with alcoholism. Arch Gen Psychiatry 63: 817–823 Higuchi S, Matsushita S, Muramatsu T, Murayama M, Hayashida M 1996 Alcohol and aldehyde dehydrogenase genotypes and drinking behavior in Japanese. Alcohol Clin Exp Res 20: 493–497 Muramatsu T, Wang ZC, Fang YR et al 1995 Alcohol and aldehyde dehydrogenase genotypes and drinking behavior of Chinese living in Shanghai. Hum Genet 96:151–154

Acetaldehyde and alcoholic cardiomyopathy: lessons from the ADH and ALDH2 transgenic models Jun Ren Center for Cardiovascular Research and Alternative Medicine, University of Wyoming, 1000 E. University Avenue, Department 3375, Laramie, WY 82071, USA

Abstract. Alcoholic cardiomyopathy is manifested as ventricular dysfunction although its pathogenesis remains obscure. The major ethanol metabolite acetaldehyde is suspected to play a culprit role in the onset of this myopathic state. To explore the role of acetaldehyde in alcoholic cardiomyopathy, we generated transgenic mice with overexpression of the alcohol-metabolizing enzyme alcohol dehydrogenase (ADH) and the acetaldehydemetabolizing enzyme mitochondrial aldehyde dehydrogenase (ALDH2), driven by myosin heavy chain and chicken β-actin promoters, respectively. While neither transgene overtly affected the phenotype and intrinsic cardiomyocyte contractile properties of the background FVB mice, they altered the course of chronic alcohol ingestion-elicited alcoholic cardiomyopathy. Following an 8–12 week feeding with 4% alcoholic diet, cardiomyocyte mechanical function was depressed in FVB cardiomyocytes characterized by reduced peak shortening, impaired myocyte relengthening, and dampened intracellular Ca2+ release and sarcoplasmic reticulum Ca2+ re-uptake. This was associated with enhanced oxidative stress, lipid peroxidation and protein carbonyl formation in alcohol consuming FVB mice. Strikingly, ADH exaggerated whereas ALDH2 attenuated alcohol-induced mechanical and intracellular Ca2+ defects, oxidative stress, lipid peroxidation and protein damage. These data revealed that enhanced acetaldehyde production may be detrimental whereas facilitated acetaldehyde breakdown may be beneficial to alcoholic cardiomyopathy, indicating a possible therapeutic target against acetaldehyde in alcoholic tissue damage. 2007 Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Foundation Symposium 285) p 69–79

Alcohol, acetaldehyde and heart disease Although light to moderate alcohol intake appears to be beneficial to cardiovascular health, chronic alcohol use or binge drinking often result in cardiac dysfunction and arrhythmias (Preedy et al 2001, Spies et al 2001). Almost one out of every three alcoholics display some degree of heart problems manifested as alcoholic cardiomyopathy, a dilated heart muscle disease discernable by hypertrophy, 69

70

REN

myofibrillary disruption, reduced contractility, decreased ejection fraction and stroke volume (Spies et al 2001). To date, several theories have been postulated for the pathogenesis of alcoholic cardiomyopathy including toxicity of ethanol or its metabolites, build-up of reactive oxygen species and oxidative stress, protein– aldehyde adducts, accumulation of fatty acid ethyl esters, modification of lipoprotein and apolipoprotein particles, and change in neurohumoral or hormonal factors (Hannuksela et al 2002, Niemela et al 2003, Zhang et al 2004). Although these theories have provided explanations for the mechanism of action behind alcohol-induced tissue damage, they have not been fully validated by clinical and experimental evidence. The validity of one theory often relies heavily on the concept of another theory. For example, oxidative stress may serve as a permissive factor for ethanol toxicity whereas ethanol-induced oxidative stress mandates ethanol metabolism into more reactive molecules such as acetaldehyde in the first place. Acetaldehyde (CH3CHO), a chemically reactive small molecule with a low boiling point, is formed by oxidation of ethanol primarily through cytosolic alcohol dehydrogenase (ADH). Although liver is considered the major site for ethanol oxidation, other organs including heart also expresses ADH enzyme to participate in ethanol metabolism. In addition to ethanol metabolism, acetaldehyde can also be generated through lipid peroxidation, glycation and amino acid oxidation (Uchida 2000). Acetaldehyde is further oxidized to acetic acid mainly through aldehyde dehydrogenase (ALDH). It may react with amino, hydroxyl, and sulfhydryl groups to interfere with or modify structure and function of macromolecules such as proteins and enzymes. In addition to these two major ethanol metabolizing enzymes, other ethanol oxidation pathways including catalase, microsomal ethanol-oxidizing system (MEOS/ CYP2E1), and the non-oxidative pathway, which generates fatty acid ethyl esters (FAEEs), appear to play a minor role for alcohol metabolism. ADH and ALDH exhibit genetic polymorphism and ethnic variation, which may contribute to alcoholinduced tissue damage and alcoholic complications. Circulating and tissue levels of acetaldehyde were found disparately elevated depending on ADH and ALDH polymorphism following alcohol intake (Harcombe et al 1995, Kajander et al 2001). Blood acetaldehyde plateau level was much higher in alcohol-dependent (42.7 µM) than non-alcohol-dependent (26.5 µM) individuals possibly due to lessened ability of ALDH to metabolize acetaldehyde. This is consistent with the observation that blood acetaldehyde levels were 10-times higher in individuals with defective mitochondrial ALDH (ALDH2, 30–125 µM) compared with normal individuals (5 µM) (Nishimura et al 2002). However, the jury is still out as to whether elevated acetaldehyde levels participate in the aetiology of alcoholic cardiomyopathy or simply result from alcohol metabolism. Evidence from our lab as well as others has provided evidence that acute acetaldehyde exposure directly compromises cardiac contractile function (Brown et al 1999, Ren et al 1997), Moreover, the acetaldehyde-induced

ALCOHOLIC CARDIOMYOPATHY

71

cardiac depressant effect cannot be abolished by reserpine pretreatment or adrenoceptor antagonists, thus not favouring the potential contribution of autonomic regulation on the acetaldehyde-induced heart contractile effect. Our recent findings suggest that acetaldehyde-induced negative inotropic effect in the heart may be underscored by depressed cardiomyocyte contractile capacity and dampened sarcoplasmic reticulum (SR) Ca2+ release (Aberle & Ren 2003, Ren et al 1997, 2001). These effects elicited by acetaldehyde are somewhat reminiscent of the ones from ethanol in cardiomyocytes and myocardium, favouring the notion that acetaldehyde may be a candidate toxin in the pathogenesis of alcoholic cardiomyopathy. Transgenic modification of acetaldehyde metabolism Although the ‘acetaldehyde toxicity theory’ for alcoholic complications and tissue injury has been around for decades, the advance of this theory in the aetiology of alcohol cardiomyopathy has been subtle over the past years due to the lack of a suitable way to manoeuvre acetaldehyde levels in vivo. Although blood acetaldehyde levels may reach up 150 µM (in rare cases to 500 µM) following alcohol administration in Asians and African Americans due to ALDH polymorphism (Tsukamoto et al 1989, Yoshida 1992), intolerance to alcohol ingestion makes these individuals practically ineligible for clinical study. On the other hand, using metabolic inhibitors to alter acetaldehyde levels has been proven to be non-specific, ineffective, toxic and difficult to manage (Preedy et al 2002). For assessing the role of acetaldehyde in alcohol-induced tissue damage, it is possible to genetically modify the ethanol metabolizing enzymes ADH and ALDH to artificially change the levels of acetaldehyde exposed to the hearts in experimental animals. If acetaldehyde toxicity is permissive to development of alcoholic cardiomyopathy, then mice overexpressing ADH should have exacerbated alcoholic cardiomyopathy development following alcohol consumption and mice with ALDH2 transgene overexpression should be spared the progression of alcoholic cardiomyopathy following alcohol consumption. This assumption was largely based on human data on alcohol metabolizing enzyme polymorphisms. Allelic variation of ADH and ALDH genes, especially deficiency in ALDH2 due to point mutation in the active ALDH2*1 gene, significantly alters blood acetaldehyde levels and vulnerability for alcoholism (Peng et al 1999, 2002). Up to 50% of Asians carry mutant alleles of ALDH (ALDH2*2/1 and ALDH2*2/2) resulting from a single point mutation of the active ALDH2*1 gene (Nishimura et al 2002, Yoshida 1992). Blood acetaldehyde levels were ∼10-fold higher in the ALDH2-deficient than ALDH2intact populations following alcohol consumption (Nishimura et al 2002). On the other hand, using blood pressure, serum lipids and uric acid as indices for risk of coronary heart diseases, individuals with ADH2*1/2*1 genotype were found to suffer from less negative effects of drinking (Hashimoto et al 2002).

72

REN

The ALDH2*2/2 gene may protect against development of alcohol dependence and alcohol-related disease. Nonetheless, this epidemiological study fails to provide direct information regarding the role of acetaldehyde on cardiac function, largely due to intolerance to alcohol among these individuals with genetic polymorphisms. In addition to the direct cardiac depressant response elicited by acetaldehyde as mentioned earlier, contribution of acetaldehyde to alcoholic cardiomyopathy was also substantiated by the fact that the ALDH inhibitor cyanamide potentiates alcohol intake-induced rise of plasma cardiac troponin T levels, a key index for myocardial cell death. It should be indicated that homozygosity for the allele ALDH2*2 may help to inhibit the development of alcoholism in Asians. After a small dose of alcohol, cardiac and extracranial/intracranial arterial haemodynamic parameters as well as self-rated subjective sensations were strikingly responsive in homozygous ALDH2*2 individuals as evidenced by pronounced cardiovascular haemodynamic effects as well as subjective perception of general discomfort for as long as 2 h after alcohol ingestion. The accumulated blood acetaldehyde as a result of low-dose alcohol hypersensitivity may provide discomfort feelings that discourage further heavy drinking in these individuals. ADH In the cardiac-specific ADH overexpression transgenic model, ADH activity was increased by ∼40-fold in the heart of ADH transgenic mice, which results in a four to six fold increase in cardiac acetaldehyde production after alcohol ingestion (Hintz et al 2003, Liang et al 1999). Both acute (5 min) and chronic (8 weeks) alcohol administration promoted ethanol-induced cardiac contractile depression and development of alcoholic cardiomyopathy (Duan et al 2002, Hintz et al 2003, Liang et al 1999). Ethanol depressed cardiac contractile and intracellular Ca2+ response with maximal inhibitions of 23.3% and 23.4%, respectively, in cardiomyocytes from wildtype FVB mice. Interestingly, ethanol-induced cardiac depressant effects were significantly augmented in myocytes from ADH mice, with maximal inhibitions of 43.7% and 40.6% in cardiac contractile and intracellular Ca2+ responses, respectively (Duan et al 2002). Expression of mRNAs for ANP and α-skeletal actin were much higher in alcohol-consuming ADH mice compared with FVB mice consuming alcohol (Liang et al 1999). Not surprisingly, morphological and functional damage including cardiac enlargement, cardiac ultrastructure disruption, reduced whole heart and cardiomyocyte contractility were more severe in ADH mice compared to FVB mice consuming alcohol (Hintz et al 2003, Liang et al 1999). The more pronounced problem in ADH mice following alcohol consumption was associated with dampened intracellular Ca2+ release and SR Ca2+ load as well as enhanced lipid peroxidation and protein carbonyl formation in ADH mice (Hintz et al 2003). These results support the notion that acetaldehyde plays a significant

ALCOHOLIC CARDIOMYOPATHY

73

role in alcoholic cardiac injury. The ADH transgene itself does not affect morphological, mechanical and intracellular Ca2+ properties, indicating that the transgene is not innately harmful. The NAD+/NADH ratio was similar in ADH and FVB mice consuming alcohol (Liang et al 1999), indicating that depletion of NAD+ was unlikely to be an adequate factor for enhanced cardiac damage in ADH transgenic mice after alcohol intake. More recent data revealed that crossing ADH with an antioxidant such as catalase transgene cancelled ADH-induced exaggeration of alcohol-elicited oxidative stress and cardiomyocyte contractile depression (Dong et al 2006), indicating that antioxidants may effectively antagonize ADH-induced enhanced cardiac depression in response to ethanol. ALDH ALDH facilitates acetaldehyde removal by converting it into acetate. This is supported by the evidence that blood acetaldehyde levels are ∼10-fold higher in human beings with defective ALDH2 than normal individuals (Nishimura et al 2002). Using a transgenic mouse model (Aldh2−/−) with inactive ALDH2, Isse and colleagues compared ethanol and acetaldehyde concentrations in blood, brain, and liver between Aldh2−/− and Aldh2+/+ wild-type mice following a similar dose of ethanol gavage. Significantly higher blood acetaldehyde but not ethanol concentrations were found in Aldh2−/− mice 1 h after ethanol administration. ALDH2 metabolized 94% of acetaldehyde produced from ethanol as calculated from the area under the curve of acetaldehyde (Isse et al 2005). These data indicate that ALDH2 is a major enzyme for acetaldehyde metabolism, which explained high acetaldehyde levels in Aldh2−/− mice following ethanol gavage. To evaluate the role of facilitated acetaldehyde metabolism on alcohol or acetaldehyde-induced tissue and cell injury, we overexpressed ALDH2 driven by chicken β-actin promoter in human umbilical vein endothelial cells (HUVEC) and fetal human cardiac myocytes. Our results demonstrated that ALDH2 overexpression significantly attenuates ethanol and acetaldehyde-induced oxidative stress and apoptosis (Li et al 2004, 2006), suggesting that facilitation of acetaldehyde breakdown lessens or detoxifies its cellular toxicity. These results support the notion that acetaldehyde may directly elicit cell injury since facilitation of its metabolism by ALDH2 alleviates cellular toxicity. We went on to make transgenic mice overexpressing low Km ALDH2 using the same chicken β-actin promoter. The cardiac-specific α-MHC promoter was not chosen since diffusion of acetaldehyde from peri-cardiac regions would easily offset the facilitated acetaldehyde removal from the cardiac tissue. In the absence of alcohol intake, the ALDH2 transgene did not exhibit an effect on cardiomyocyte function or tissue oxidative damage, consistent with the notion that the ADH transgene is not innately harmful to heart function. Following feeding ALDH2 and FVB mice a 4% alcoholic or control diet for 12 weeks, myocytes from alcohol-fed mice

74

REN

showed a significantly reduced peak shortening, velocity and duration of relengthening (diastole), intracellular Ca2+ release and SR Ca2+ uptake. Interestingly, myocytes from ALDH2 transgenic mice displayed attenuated alcohol-induced cardiomyocyte depression. Oxidative stress, lipid peroxidation and protein carbonyl formation were significantly elevated in hearts and other tissues in FVB mice following chronic ethanol consumption, the effects of which were attenuated by the ALDH2 transgene. These preliminary data from the author’s laboratory suggest that facilitated acetaldehyde breakdown with overexpression of ALDH2 may protect the heart against alcohol-induced detrimental effects, indicating the therapeutic potential of ALDH2 enzyme in alcoholic tissue damage. Recent advances in both ADH and ALDH2 transgenes and transgenic mice have made it possible for us to artificially alter ethanol metabolism to evaluate the role of acetaldehyde in the progression of alcoholic cardiomyopathy. It should be cautioned that the acetaldehyde toxicity theory has its own share of deficiency. Often acetaldehyde initiates cell and tissue injury at a level of 50–100 µM or higher. However, the concentrations of acetaldehyde usually achieved in the body are in the low micromolar range following moderate ethanol intoxication. Certain tissues such as brain are exposed to an even lower level of acetaldehyde. Therefore, the jury is still out as to whether acetaldehyde is the main mediator of ethanol-induced cytotoxic effects. Other hypotheses postulated for alcoholic injury including oxidative damage, lipid peroxidation and altered membrane integrity may work in concert with acetaldehyde to facilitate the detrimental effect of ethanol on the function of protein and membrane phospholipids (Cederbaum et al 2001, Mantle & Preedy 1999). Based on our preliminary findings using ADH and ALDH2 transgenes, it may be concluded that elevated acetaldehyde levels during acute and chronic alcohol ingestion participate in the development of alcoholic cardiomyopathy via alterations in excitation–contractility coupling, myocardial function, oxidative stress and protein damage. Since a convincing human case study on heart function following chronic alcohol intake is still lacking, it is still premature to conclude that acetaldehyde is the ultimate cause of alcoholic cardiomyopathy. Further studies are warranted to depict the association of blood or tissue acetaldehyde levels and ventricular contractile function following alcohol ingestion. Acknowledgement The research in the author’s laboratory was supported in part by NIH R15 AA13575-01, NIH R01 AA13412, NIH/NCRR BRIN RR-16474 and NIH/NCRR COBRE P20 RR15640.

References Aberle NS, Ren J 2003 Short-term acetaldehyde exposure depresses ventricular myocyte contraction: role of cytochrome P450 oxidase, xanthine oxidase, and lipid peroxidation. Alcohol Clin Exp Res 27:577–583

ALCOHOLIC CARDIOMYOPATHY

75

Brown RA, Jefferson L, Sudan N, Lloyd TC, Ren J 1999 Acetaldehyde depresses myocardial contraction and cardiac myocyte shortening in spontaneously hypertensive rats: role of intracellular Ca2+. Cell Mol Biol (Noisy-le-grand) 45:453–465 Cederbaum AI, Wu D, Mari M, Bai J 2001 CYP2E1-dependent toxicity and oxidative stress in HepG2 cells. Free Radic Biol Med 31:1539–1543 Dong F, Fang CX, Yang X, Zhang X, Lopez FL, Ren J 2006 Cardiac overexpression of catalase rescues cardiac contractile dysfunction induced by insulin resistance: role of oxidative stress, protein carbonyl formation and insulin sensitivity. Diabetologia 49:1421–1433 Duan J, McFadden GE, Borgerding AJ et al 2002 Overexpression of alcohol dehydrogenase exacerbates ethanol-induced contractile defect in cardiac myocytes. Am J Physiol Heart Circ Physiol 282:H1216–1222 Hannuksela ML, Liisanantti MK, Savolainen MJ 2002 Effect of alcohol on lipids and lipoproteins in relation to atherosclerosis. Crit Rev Clin Lab Sci 39:225–283 Harcombe AA, Ramsay L, Kenna JG et al 1995 Circulating antibodies to cardiac proteinacetaldehyde adducts in alcoholic heart muscle disease. Clin Sci (Lond) 88:263–268 Hashimoto Y, Nakayama T, Futamura A, Omura M, Nakarai H, Nakahara K 2002 Relationship between genetic polymorphisms of alcohol-metabolizing enzymes and changes in risk factors for coronary heart disease associated with alcohol consumption. Clin Chem 48: 1043–1048 Hintz KK, Relling DP, Saari JT et al 2003 Cardiac overexpression of alcohol dehydrogenase exacerbates cardiac contractile dysfunction, lipid peroxidation, and protein damage after chronic ethanol ingestion. Alcohol Clin Exp Res 27:1090–1098 Isse T, Matsuno K, Oyama T, Kitagawa K, Kawamoto T 2005 Aldehyde dehydrogenase 2 gene targeting mouse lacking enzyme activity shows high acetaldehyde level in blood, brain, and liver after ethanol gavages. Alcohol Clin Exp Res 29:1959–1964 Kajander OA, Kupari M, Perola M et al 2001 Testing genetic susceptibility loci for alcoholic heart muscle disease. Alcohol Clin Exp Res 25:1409–1413 Li SY, Gomelsky M, Duan J et al 2004 Overexpression of aldehyde dehydrogenase-2 (ALDH2) transgene prevents acetaldehyde-induced cell injury in human umbilical vein endothelial cells: role of ERK and p38 mitogen-activated protein kinase. J Biol Chem 279:11244–11252 Li SY, Li Q, Shen JJ, Dong F, Sigmon VK, Liu Y, Ren J 2006 Attenuation of acetaldehyde-induced cell injury by overexpression of aldehyde dehydrogenase-2 (ALDH2) transgene in human cardiac myocytes: role of MAP kinase signaling. J Mol Cell Cardiol 40:283–294 Liang Q, Carlson EC, Borgerding AJ, Epstein PN 1999 A transgenic model of acetaldehyde overproduction accelerates alcohol cardiomyopathy. J Pharmacol Exp Ther 291:766–772 Mantle D, Preedy VR 1999 Free radicals as mediators of alcohol toxicity. Adverse Drug React Toxicol Rev 18:235–252 Niemela O, Parkkila S, Worrall S, Emery PW, Preedy VR 2003 Generation of aldehyde-derived protein modifications in ethanol-exposed heart. Alcohol Clin Exp Res 27:1987–1992 Nishimura FT, Fukunaga T, Kajiura H et al 2002 Effects of aldehyde dehydrogenase-2 genotype on cardiovascular and endocrine responses to alcohol in young Japanese subjects. Auton Neurosci 102:60–70 Peng GS, Wang MF, Chen CY et al 1999 Involvement of acetaldehyde for full protection against alcoholism by homozygosity of the variant allele of mitochondrial aldehyde dehydrogenase gene in Asians. Pharmacogenetics 9:463–476 Peng GS, Yin JH, Wang MF, Lee JT, Hsu YD, Yin SJ 2002 Alcohol sensitivity in Taiwanese men with different alcohol and aldehyde dehydrogenase genotypes. J Formos Med Assoc 101: 769–774 Preedy VR, Adachi J, Peters TJ et al 2001 Recent advances in the pathology of alcoholic myopathy. Alcohol Clin Exp Res 25:54S–59 Preedy VR, Adachi J, Asano M et al 2002 Free radicals in alcoholic myopathy: indices of damage and preventive studies. Free Radic Biol Med 32:683–687

76

REN

Ren J, Davidoff AJ, Brown RA 1997 Acetaldehyde depresses shortening and intracellular Ca2+ transients in adult rat ventricular myocytes. Cell Mol Biol (Noisy-le-grand) 43:825–834 Ren J, Natavio M, Jefferson L, Pavlik ML, Brown RA 2001 Loss of cardiac contractile response to tetrahydropapaveroline with advanced age and hypertension. Cell Mol Biol (Noisy-le-grand) 47 Online Pub:OL15–22 Spies CD, Sander M, Stangl K et al 2001 Effects of alcohol on the heart. Curr Opin Crit Care 7:337–343 Tsukamoto S, Muto T, Nagoya T, Shimamura M, Saito M, Tainaka H 1989 Determinations of ethanol, acetaldehyde and acetate in blood and urine during alcohol oxidation in man. Alcohol Alcohol 24:101–108 Uchida K 2000 Role of reactive aldehyde in cardiovascular diseases. Free Radic Biol Med 28:1685–1696 Yoshida A 1992 Molecular genetics of human aldehyde dehydrogenase. Pharmacogenetics 2:139–147 Zhang X, Li SY, Brown RA, Ren J 2004 Ethanol and acetaldehyde in alcoholic cardiomyopathy: from bad to ugly en route to oxidative stress. Alcohol 32:175–186

DISCUSSION Rao: You showed that acetaldehyde can cause apoptosis of cardiac myocytes, while it can also induce cardiac myocyte relaxation. At what concentration does it induce apoptosis? Ren: The threshold we found is about 10 µM. Below this we can’t see any detrimental effect on cell survival. For apoptosis we need at least 12 h incubation time and have to use a sealed container to avoid evaporation of the acetaldehyde. Albano: I have a question regarding the mechanisms of damage that you have investigated in your experiments with myocytes. It is interesting to see that apoptosis is associated with the activation of MAP kinases and, particularly, JNK and p38 MAPK. Have you considered performing experiments blocking any of those kinases with selective inhibitors to prevent apoptosis? The sustained activation of p38 MAPK and JNK is known to lead to apoptosis in many cell types. Ren: This is another set of data that I find difficult to explain. Inhibition of either of these pathways using specific inhibitors can modify the acetaldehyde-induced apoptotic effects. This is hard to explain because each signalling pathway seems to be 100% responsive for the apoptosis. Apte: It is not surprising at all that you get 100% inhibition when you inhibit either of those. Cell signalling is so redundant. Albano: You also showed depressed Ca2+ release in your experimental systems. Have you worked out a mechanism responsible? Is this due to alteration in Ca2+ fluxes through the plasma membranes, or is this just a failure to release Ca2+ from the entoplasmic reticulum stores? Ren: We don’t have any solid data to support this. We are doing some ryanodine receptor studies. From what we know so far, it seems that there is a reduction of the SR Ca2+ storage.

ALCOHOLIC CARDIOMYOPATHY

77

Apte: I have a question to do with the increase in cardiac size. If acetaldehyde increases apoptosis, then is the increase in size due to increased proliferation of myocytes, or lengthening, or both? And is it reversible? If you stopped giving the mice alcohol would the hearts revert back to normal size? Ren: Cardiomyocytes don’t proliferate, so it will likely be through an increase in size. We are trying to revert to normal diet for these animals. We have only done it over the short term (4 weeks) and we didn’t see a complete reversal of the problem. Short-term withdrawal of alcohol doesn’t seem to cause a full reversal of cardiac dysfunction. Seitz: I remember some older literature on mitochondrial damage in the myofibrils after alcohol consumption (Bing & Tillmann 1977). Did you look at mitochondrial-associated apoptosis, membrane potential and cytochrome c? Ren: We looked by transmission electron microscopy, and saw enlargement of mitochondria. We haven’t done any measurement of mitochondrial membrane potential. Preedy: I want to comment on the ADH model, because this offers lots of advantages. What was especially noticeable was the enlarged hearts. One of the criticisms of the models of alcoholic heart damage is the fact that you can’t induce this enlarged heart unless you feed alcohol for many months. The fact that you have managed to get this increased size in 12 weeks is quite dramatic. Have you tried any shorter periods? Ren: We began by trying to treat for just a month, but found that there was no change in terms of cardiomyocyte function. We have been using 8 to 12 weeks as a minimum duration for chronic feeding to make the effect more severe. Shukla: One of the questions which has yet to be decided is whether acetaldehyde is good or bad or both. As with alcohol, there are suggestions that under some conditions acetaldehyde can be protective, as your data suggest. In the catalase ADH1 mouse have you tried to see the effects of other agonists on these hearts, such as adenosine? Ren: We haven’t tried this. Emery: Is your model only using isolated cardiomyocytes, or have you done perfusions of isolated hearts? Ren: We recently started to use echocardiography, but we haven’t perfused whole hearts. Niemelä: One of the features in clinical cardiomyopathy is fibrosis, which is considered irreversible. Have you seen evidence of enhanced collagen production in your transgenic model? Ren: We haven’t done this ourselves, but I think it has been shown by other labs. This is one of the reasons for the reduced contractile capacity in the hearts (Wange et al 2005). Rao: What about effects on vascular smooth muscle? Would you expect to see a similar effect of acetaldehyde on this?

78

REN

Ren: This study was started with smooth muscle cells for ALDH2 gene transfection, but for some reason there was a poor percentage of transfection so we don’t have any results. Apte: Is there any change in the contractile protein mass in the cells? Is this the reason for their poor performance? Ren: We haven’t done this, but Victor Preedy can comment on this. Preedy: We get a reduction in the contractile protein content, but it seems that the models you use are different. In our models, we use 6 weeks alcohol feeding as a time point and don’t get some of the changes you do. You are trying to mimic in a few weeks what happens for many years in a chronic alcohol misuser. You see an increase in protein mass in patients with cardiomegaly induced by chronic alcoholism, but we don’t see this in our rat model. For example, in six weeks alcohol feeding we see a reduction in contractile protein content of the rat heart. Helmut Seitz, you did a study to look at contractile function, didn’t you? Seitz: We did echocardiography in rat hearts and there was not a clear answer. It was difficult to measure. Preedy: I do recall that in order to see contractile dysfunction in the rat many months of alcohol feeding are needed, which is impractical. One colleague was saying that they did some functional studies which they couldn’t believe because the heart function of the alcohol exposed heart was greater than the controls. Rao: I have a general question. Is it likely that the acetaldehyde-mediated relaxation of smooth muscle or cardiac muscle is one of the mechanisms involved in the beneficial effect of moderate alcohol consumption? Ren: There were some articles in the late 1990s showing that low level acetaldehyde forms some kind of a conjugate with cardiac relaxation proteins, or a protein responsible for pumping Ca2+ out of cells or back into SR. This can enhance the functions of the Ca2+ removal. But this only occurs at low levels. Eriksson: The whole concept of the U-shaped alcohol consumption/mortality curve got a bit of a blow from a large meta-analysis published early this year which identified many confounding factors (Fillmore et al 2006). The criticism is that the non-drinkers aren’t a good baseline group. Seitz: The U-shaped curve depends on the zero point—the control population of non-drinkers. If someone reports that they don’t drink, there are three possibilities: they could be lying to you, or be former alcoholics, or there is another reason for not drinking such as disease. There was a recent article from the Copenhagen group in which they studied 3.5 million till receipts from supermarkets ( Johansen et al 2006). They found that those who bought wine had a different pattern of food purchasers compared to other groups. This is a complex issue. Preedy: I’d like to connect that with the contractility of the heart. Prof UrbanoMarquez’s group has shown that the function of the heart decreases in proportion

ALCOHOLIC CARDIOMYOPATHY

79

to the lifetime cumulative alcohol intake. In terms of contractile function there doesn’t appear to be a U-shaped curve. Emery: The point about the baseline of the curve is well made, but there are points in between zero and the bottom of the curve. It doesn’t completely negate your argument, but this should be borne in mind. Eriksson: You are right, but very moderate drinking is probably selective for some underlying lifestyle factors. The person who can drink moderately can also do other things moderately. They may be more relaxed. The closer you come to the biological mechanism we realize that there is no single mechanism that can go in different directions. So we need to propose different mechanisms for each phase of the curve. Emery: Most people in the room can think of factors that do cause opposite effects at different directions, but I do take your point that the epidemiology tells you very little because there are so many confounding factors. It is very difficult to disentangle factors such as lifestyle and other aspects of diet. References Bing RJ, Tillmann H 1977 The effect of alcohol on the heart. In: Lieber CS (ed) Metabolic aspects of alcoholism. MTP Press, Lancaster Fillmore KM, Kerr WC, Stockwell T, Chikritzhs T, Bostrom A 2006 Moderate alcohol use and reduced mortality risk: Systematic error in prospective studies. Addiction Res Theor 14:101–132 Johansen D, Friis K, Skovenborg E, Gronbaek M 2006 Food buying habits of people who buy wine or beer: cross sectional study. BMJ 332:519–522 Wang L, Zhou Z, Saari JT, Kang YJ 2005 Alcohol-induced myocardial fibrosis in metallothioneinnull mice: prevention by zinc supplementation. Am J Pathol 167:337–344

Interrelationship between alcohol, smoking, acetaldehyde and cancer Mikko Salaspuro Research Unit of Substance Abuse Medicine, Biomedicum Helsinki, University of Helsinki, Helsinki, Finland

Abstract. In industrialized countries alcohol and tobacco are the main risk factors of upper digestive tract cancer. With regard to the pathogenesis of these cancers, there is strong epidemiological, biochemical and genetic evidence supporting the role of the first metabolite of alcohol oxidation—acetaldehyde—as a common denominator. Alcohol is metabolized to acetaldehyde locally in the oral cavity by microbes representing normal oral flora. Poor oral hygiene, heavy drinking and chronic smoking modify oral flora to produce more acetaldehyde from ingested alcohol. Also, tobacco smoke contains acetaldehyde, which during smoking becomes dissolved in saliva. Via swallowing, salivary acetaldehyde of either origin is distributed from oral cavity to pharynx, oesophagus and stomach. Strongest evidence for the local carcinogenic action of acetaldehyde provides studies with ALDH2deficient Asian drinkers, who form an exceptional human model for long-term acetaldehyde exposure. After drinking alcohol they have an increased concentration of acetaldehyde in their saliva and this is associated with over 10-fold risk of upper digestive tract cancers. In conclusion, acetaldehyde derived either from ethanol or tobacco appears to act in the upper digestive tract as a local carcinogen in a dose-dependent and synergistic way. 2007 Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Foundation Symposium 285) p 80–96

Acetaldehyde is the first metabolite of alcohol oxidation. But acetaldehyde is also found in tobacco smoke and is an important component of food flavourings (National Toxicology Program 2003). Acetaldehyde is reasonably anticipated to be a human carcinogen based on sufficient evidence of carcinogenicity in experimental animals (IARC 1999). On the other hand, acetaldehyde is considered to be a GRAS (generally recognized as safe) compound for the intended use as a flavouring agent and adjuvant (National Toxicology Program 2003). The GRAS classification of acetaldehyde should, however, be reconsidered in the light of increasing evidence indicating that acetaldehyde is carcinogenic also in humans. Strongest evidence for the local carcinogenic action of acetaldehyde comes from biochemical studies with individuals who have either a decreased ability to detoxify or an enhanced ability to produce acetaldehyde. ALDH2-deficient Asian alcohol 80

ALCOHOL, SMOKING, ACETALDEHYDE AND CANCER

81

consumers and Caucasian heavy drinkers homozygous for the fast ADH1C*1 allele have a markedly increased risk for upper digestive tract cancers (Yokoyama et al 1988, Homann et al 2006). These individuals also express considerably elevated salivary acetaldehyde levels after alcohol drinking as compared to those with the normal enzymes (Väkeväinen et al 2000a, Visapää et al 2004). Accordingly, these genetic variants form an exceptional human model for long-term acetaldehyde exposure, which is associated with a particularly high risk of upper digestive tract cancers. Tobacco and alcohol are two other independent and interactive causes of upper digestive tract cancers. Strong epidemiological and biochemical evidence indicates that acetaldehyde acts as a common denominator in the pathogenesis of these cancers. In the oral cavity alcohol is metabolized to acetaldehyde by many microbes representing normal oral flora (Salaspuro 2003). On the other hand, tobacco smoke contains high levels of acetaldehyde that during smoking become in part dissolved in saliva (Salaspuro & Salaspuro 2004). According to indisputable evidence acetaldehyde derived either from alcohol or tobacco appears to act in the digestive tract as a local carcinogen in a dose-dependent and synergistic way. Via swallowing, salivary acetaldehyde of either origin is distributed further to the pharynx, oesophagus and stomach and may thus explain the tobacco- and alcohol-related cancer risk of these organs. Epidemiological, biochemical and genetic interrelationships between alcohol, smoking, salivary acetaldehyde and cancer appear to be convincing. With regard to cancer prevention it is important to characterize all those factors that may have an effect on the local concentration of acetaldehyde in saliva. In addition to the abovementioned hereditary factors, these include smoking and drinking habits, possible dietary exposure to acetaldehyde-containing foodstuffs or beverages, individual differences in the characteristics of oral microflora, salivary alcohol levels and the presence of acetaldehyde binding agents.

Epidemiological interrelationships Independent and synergistic effects of alcohol and tobacco on cancer risk It has been estimated that in industrialized countries up to 80% of upper digestive tract cancers can be prevented by abstaining from alcohol drinking and smoking. In a meta-analysis concerning alcohol related cancers and including 235 studies (over 117 000 cases) strong trends in risk were observed for cancers of oral cavity, pharynx, larynx and oesophagus (Table 1). A weaker direct relationship was observed for cancer of the stomach (RR 1.32/100 g alcohol daily). For all these cancers a significantly increased risk was found also for ethanol intake of 25 g (about two drinks) per day.

82

SALASPURO

TABLE 1 Pooled relative risks of upper digestive tract cancers as related to the amount of alcohol consumed Daily alcohol consumption (g) Type of cancer Oral cavity and pharynx Larynx Oesophagus

25

50

100

1.76 1.38 1.51

2.87 1.94 2.21

6.10 3.95 4.23

Source: Bagnardi et al (2001). A bottle of wine corresponds approximately to 80 g of alcohol.

The overall risk of oral cancer among smokers has been estimated to be 7–10 times higher than for never-smokers (Warnakulasuriya et al 2005). The risk increases with increasing tobacco consumption. The relative risk of oral, pharyngeal and laryngeal cancers was 3.9 for those smoking 10–19 g and 15.4 for those smoking over 30 g daily (Brugere et al 1986). In another study the relative risk of oesophageal cancer for those smoking 0–10 g daily was 1.0 and for those smoking over 30 g daily it was 7.8 (n = 200) ( Tuyns et al 1977). 28% of stomach cancer deaths in men and 14% among women have been estimated to be attributable to tobacco use in the USA (Chao et al 2002). Similar results were obtained in a prospective European study including 521 468 individuals from 10 European countries (Conzales et al 2003). In this study the hazard ratio for neversmokers was 1.45 and for current smokers it was 1.73 in males and 1.87 in females. The risk of stomach cancer increased with intensity and duration of smoking. The synergistic effect of alcohol and tobacco on digestive tract cancer risk has been demonstrated in many studies and confirmed in a recent meta-regression analysis including 14 studies (4585 cases) (Zeka et al 2003). On the relative risk scale the carcinogenic effects of alcohol and tobacco were found to be multiplicative. The relative risks for individuals consuming over 30 cigarettes and four or more drinks daily were as follows: 21.2 oropharynx, 35.6 pharynx, 34.6 larynx and 12.7 oesophagus. Biochemical interrelationships Accumulation of acetaldehyde in the saliva after alcohol drinking After its ingestion alcohol is absorbed from the stomach and upper duodenum, and is thereafter transported by blood circulation to the liver and other organs including mucous membranes and salivary glands. Ethanol is evenly distributed to the whole aqueous phase of the human body and therefore alcohol

ALCOHOL, SMOKING, ACETALDEHYDE AND CANCER

83

concentrations are equal in the blood, saliva and other parts of the digestive tract ( Jones 1979). In the saliva ethanol is effectively metabolized to acetaldehyde by many microbes representing normal oral flora (Homann et al 1997). However, the capacity of oral microbes and mucous membranes to detoxify acetaldehyde is limited. This results in markedly elevated acetaldehyde concentrations in the saliva and intestinal contents during and after drinking of alcoholic beverages (Salaspuro 2003). Via swallowing, salivary acetaldehyde is transported from the mouth to the mucous membranes of the pharynx, oesophagus and stomach. Microbial acetaldehyde production in the saliva Many microbes representing normal oral flora possess alcohol dehydrogenase (ADH) activity. Under anaerobic conditions these microbes produce energy from glucose through alcoholic fermentation. The reaction runs as follows and is associated with a significant acetaldehyde production:

Alcoholic fermentation ADH Glucose

Acetaldehyde

Ethanol

Under aerobic or microaerobic conditions prevailing in many parts of the mouth and close to the mucous membranes, the microbial ADH reaction is reversed and starts to produce acetaldehyde (Salaspuro et al 1999). Consequently, mutagenic amounts of acetaldehyde (50–150 µM) can be detected in the saliva of healthy volunteers after ingestion of a moderate dose of ethanol (Homann et al 1997). In vitro studies with human saliva show that salivary acetaldehyde production mediated by microbial ADHs is strongly associated with the ethanol concentration, is pH dependent, and inhibitable by 4-methylpyrazole, an ADH-inhibitor (Homann et al 1977). Accordingly, at higher blood and salivary ethanol concentrations, salivary acetaldehyde levels are also higher. This may explain the well established epidemiological finding of increased cancer risk associated with heavier and more intoxicating drinking. There are significant differences between different strains of oral bacteria regarding their acetaldehyde production capability and their detected ADH activity (Kurkivuori et al 2006). The marked acetaldehyde-producing capacity of the clinical strain of Streptococcus salivarius may be particularly important, since this bacterium colonizes the mucosal surfaces of oral cavity, which rarely is colonized by other normal flora bacteria (Kurkivuori et al 2006).

84

SALASPURO

FIG. 1. The synergistic effect of alcohol drinking (0.8 g/kg body weight) and smoking on acetaldehyde levels in the saliva. Note the sevenfold increase in the salivary acetaldehyde concentration during each 5 min period of active smoking. Adapted from Salaspuro & Salaspuro (2004).

Acute effects of smoking on salivary acetaldehyde Acetaldehyde is formed during the burning process and may be one of the most toxic compounds in cigarette smoke condensates (Smith & Hansch 2000, Seeman et al 2002). During the period of active cigarette smoking, the mean in vivo acetaldehyde concentration of saliva is close to 400 µM (Fig. 1). This finding confirms that smoking is one of the most important factors in the regulation of salivary acetaldehyde levels and proves that acetaldehyde of tobacco smoke becomes at least in part dissolved in the saliva. Effect of chronic smoking on microbial acetaldehyde production in the saliva Chronic smoking (about 20 cigarettes daily) increases in vitro salivary acetaldehyde production by about 50% (Homann et al 2000). After drinking alcohol, smokers have about double the salivary acetaldehyde concentration in vivo than non-smokers (Salaspuro & Salaspuro 2004). This implies changes in the capacity of oral microflora to produce acetaldehyde from ethanol. An increased incidence of oral yeast infections and Gram-positive bacteria has been found among chronic smokers (Colman et al 1976). On the other hand, in smokers these same groups of microbes

ALCOHOL, SMOKING, ACETALDEHYDE AND CANCER

85

show an enhanced in vitro capacity to produce acetaldehyde from ethanol in the saliva (Homann et al 2001). Effect of heavy alcohol consumption on microbial acetaldehyde production in the saliva Heavy drinking, here meaning more than four alcohol doses per day, increases the in vitro salivary acetaldehyde production dose-dependently by about 50% (Homann et al 2001). Moreover, the effect of chronic smoking and heavy drinking on salivary acetaldehyde production appears to be additive. Together, they increase salivary acetaldehyde production both in vitro and in vivo by about 100% as compared to non-smokers and moderate drinkers (Homann et al 2001, Salaspuro & Salaspuro 2004). Due to the high peak level of acetaldehyde in the saliva during active smoking (Fig. 1) the main risk factors for upper digestive tract cancer—smoking and drinking—appear to increase the salivary acetaldehyde concentration independently, jointly and synergistically. The local cancer promoting effect of salivary acetaldehyde is also supported by our earlier findings showing increased in vitro acetaldehyde production in the mouth washings of patients with oral cavity, laryngeal and pharyngeal cancer (Jokelainen et al 1996).

Effect of poor oral hygiene on microbial acetaldehyde production in the saliva Poor oral hygiene, tooth loss and insufficient oral hygiene habits are additional but weak risk factors for oral cancer (Zheng et al 1990). After adjustment for alcohol consumption, smoking, gender and age, poor oral hygiene appeared to be associated with about a twofold increase in salivary acetaldehyde production from ethanol in an in vitro study including 132 volunteers (Homann et al 2001). In vitro studies on human saliva producing either very high or low acetaldehyde levels from ethanol indicate that mainly aerobic bacteria and yeasts are responsible for high acetaldehyde production (Tillonen et al 1999, Homann et al 2001).

Biochemical interactions mediated by genes So far two gene polymorphisms have been shown to be able to modify salivary acetaldehyde concentration after drinking of alcohol. After ingestion of a moderate dose of alcohol (0.5 g/kg body weight) salivary acetaldehyde levels are two to three times higher in flushing ALDH2-deficient Asians (heterozygotes) than in those with the normal enzyme (Fig. 2) (Väkeväinen et al 2000a). As stated in the introduction, ALDH2-deficient Asian heavy drinkers have up to 10-fold risk of upper digestive tract cancer (Yokoyama et al 1988) and thus form an exceptional human

86

SALASPURO

Salivary acetaldehyde, uM

70 Flushers

60

Nonflushers

50 40 30 20 10 0 40

80

120

160

200

240

Time (min) FIG. 2. The effect of ALDH2 deficiency on the concentration of acetaldehyde in the saliva after a dose of alcohol (0.5 g/kg body weight) in flushing (flushers) ALDH2-deficient Chinese volunteers as compared to those with the normal ALDH2 enzyme (non-flushers). Adapted from Väkeväinen et al (2000a).

model for long term acetaldehyde exposure. It is not yet known whether ALDH2 deficiency could also increase salivary acetaldehyde during smoking. In addition to ALDH2-deficient Asians, salivary acetaldehyde levels have also been shown to be significantly elevated after drinking different doses of alcohol among Caucasians who are homozygous for the fast ADH1C*1 allele (Visapää et al 2004). There is increasing evidence that this genotype is an independent risk factor for the development of alcohol-associated tumours among heavy drinkers (Homann et al 2006). Acetaldehyde in the stomach Achlorhydric atrophic gastritis is considered to be the major premalignant condition of gastric cancer (Morson et al 1980). The pathogenetic mechanism behind increased gastric cancer risk in these patients, however, is still without final explanation. Correa’s hypothesis proposing that hypochlorhydria permits gastric bacterial colonization, the reduction of nitrates to nitrites and the formation of potentially carcinogenic N-nitroso-compounds remains controversial. An enhanced local microbial production of endogenous acetaldehyde could be another explanatory factor for the increased gastric cancer risk among patients with achlorhydria. Normal human stomach is free of microbes, because of its low pH. However, microbes may survive and even proliferate at intragastric pH levels

ALCOHOL, SMOKING, ACETALDEHYDE AND CANCER

87

(Stockbruegger et al 1984). In patients with achlorhydric atrophic gastritis the stomach is frequently colonized with microbes representing the normal oral flora. In these individuals bacterial overgrowth results, in the presence of glucose, in the formation of minor concentrations of endogenous ethanol and acetaldehyde in the gastric juice (Väkeväinen et al 2002). Furthermore, after administration of a small amount of alcohol intragastric acetaldehyde production increases to 6.5-fold in achlorhydric patients as compared to the healthy controls (Väkeväinen et al 2002). Hypochlorhydria induced by cimetidine leads to intragastric formation of endogenous ethanol, up to 27 mM (Bode et al 1984). On the other hand, one week treatment with proton pump inhibitors results in intragastric production of acetaldehyde from ethanol, which associates in a marked overgrowth of viridans group streptococci (Väkeväinen et al 2000b). The possible effect of tobacco smoke on the acetaldehyde concentration of the gastric juice is so far not known. However, during active smoking considerable amounts of salivary acetaldehyde can be expected to reach the stomach via swallowing. Helicobacter pylori is an established risk factor for gastric cancer. This association has been confirmed in several meta-analyses (Xue et al 2003). Many H. pylori strains possess significant ADH-activity and are able to produce acetaldehyde from ethanol under microaerophilic conditions (Salmela et al 1994). So far it is not known whether local acetaldehyde production by H. pylori could contribute to the pathogenesis of alcohol-related gastric cancer.

References Bagnardi V, Blangiardo M, La Vecchia C et al 2001 A meta-analysis of alcohol drinking and cancer risk. Br J Cancer 2001:1700–1705 Bode JC, Rust S, Bode C 1984 The effect of cimetidine on ethanol formation in human stomach. Scand J Gastroenterol 19:853–856 Brugere J, Guenel P, Leclerc A et al 1986 Differential effects of tobacco and alcohol in cancer of the larynx, pharynx, and mouth. Cancer 1986:391–395 Chao A, Thun MJ, Henley J et al 2002 Cigarette smoking, use of other tobacco products and stomach cancer mortality in US adults: The cancer prevention study II. Int J Cancer 101:380–389 Colman G, Beighton D, Chalk AJ et al 1976 Cigarette smoking and the microbial flora of the mouth. Aust Dent J 21:111–118 Conzales CA, Pera G, Agudo A et al 2003 Smoking and the risk of gastric cancer in the European prospective investigation into cancer and nutrition (EPIC). Int J Cancer 107:629–634 Homann N, Jousimies-Somer H, Jokelainen K et al 1997 High acetaldehyde levels in saliva after ethanol consumption: Methodological aspects and pathogenetic implications. Carcinogenesis 18:1739–1743 Homann N, Tillonen J, Meurman JH et al 2000 Increased salivary acetaldehyde levels in heavy drinkers and smokers: a microbiological approach to oral cavity cancer. Carcinogenesis 22:663–668

88

SALASPURO

Homann N, Tillonen J, Rintamäki H et al 2001 Poor dental status increases the acetaldehyde production from ethanol in saliva: a possible link to the higher risk of oral cancer among alcohol-consumers. Oral Oncol 37:153–158 Homann N, Stickel F, König IR et al 2006 Alcohol dehydrogenase 1C*1 allele is a genetic marker for alcohol-associated cancer in heavy drinkers. Int J Cancer 118:1998–2002 IARC 1999 Re-evaluation of some organic chemicals, hydrazine and hydrogen peroxide. In: Monographs on the evaluation of the carcinogenic risk of chemicals to humans. Acetaldehyde. vol. 77, Lyon: International Agency for Research on Cancer, p 319–335 Jokelainen K, Heikkonen E, Roine R et al 1996 Increased acetaldehyde production by mouthwashings from patients with oral cavity, laryngeal, or pharyngeal cancer. Alcohol Clin Exp Res 20:1206–1210 Jones AW 1979 Distribution of ethanol between saliva and blood in man. Clin Exp Pharmacol Physiol 6:53–59 Kurkivuori J, Salaspuro V, Kaihovaara P et al 2007 Acetaldehyde production from ethanol by oral streptococci. Oral Oncol 43:181–186 Morson BC, Sobin LH, Grundmann E et al 1980 Precancerous conditions and epithelial dysplasia in the stomach. J Clin Path 33:711–721 National Toxicology Program 2003 Acetaldehyde CAS No. 75-0-0 Report on carcinogens, Eleventh Edition; U.S. Department of Health and Human Services, Public Health Service Salaspuro M 2003 Acetaldehyde, microbes, and cancer of the digestive tract. Crit Rev Clin Lab Med 40:183–208 Salaspuro V, Salaspuro M 2004 Synergistic effect of alcohol drinking and smoking on in vivo acetaldehyde concentration in saliva. Int J Cancer 111:480–483 Salaspuro V, Nyfors S, Heine R et al 1999 Ethanol oxidation and acetaldehyde production in vitro by human intestinal strains of Escherichia coli under aerobic, microaerobic, and anaerobic conditions. Scand J Gastroenterol 34:967–973 Salmela KS, Roine RP, Höök-Nikanne J et al 1994 Acetaldehyde and ethanol production by Helicobacter pylori. Scand J Gastroenterol 29:309–312 Seeman JI, Dixon M, Haussman H-J 2002 Acetaldehyde in mainstream tobacco smoke: formation and occurrence in smoke and bioavailability in the smoker. Chem Res Toxicol 15: 1332–1350 Smith CJ, Hansch C 2000 The relative toxicity of compounds in mainstream cigarette smoke condensate. Food Chem Toxicol 38:637–646 Stockbruegger RW, Cotton PB, Menon GG et al 1984 Pernicious anaemia, intragastric bacterial overgrowth and possible consequences. Scand J Gastroenterol 19:355–364 Tillonen J, Homann N, Rautio M et al 1999 Role of yeasts in the salivary acetaldehyde production from ethanol among risk groups for ethanol-associated oral cavity cancer. Alcohol Clin Exp Res 23:1409–1415 Tuyns AJ, Pequignot G, Jensen OM 1977 Les cancers del’oesophage an Ille-et-Villaine en fonction de niveaux de consommation d’alcool et de tabac. Des risques qui se multiplient. Bull Cancer 64:45–60 Visapää J-P, Götte K, Benesova M et al 2004 Increased cancer risk in heavy drinkers with the alcohol dehydrogenase 1C*1 allele, possibly due to salivary acetaldehyde. Gut 53:871– 876 Väkeväinen S, Tillonen J, Agarwal DP et al 2000a High salivary acetaldehyde after a moderate dose of alcohol in ALDH2-deficient subjects: strong evidence for the local carcinogenic action of acetaldehyde. Alcohol Clin Exp Res 24:873–877 Väkeväinen S, Tillonen J, Salaspuro M et al 2000b Hypochlorhydria induced by a proton pump inhibitor leads to intragastric microbial production of acetaldehyde from ethanol. Aliment Pharmacol Ther 14:1511–1518

ALCOHOL, SMOKING, ACETALDEHYDE AND CANCER

89

Väkeväinen S, Mentula S, Nuutinen H et al 2002 Ethanol-derived microbial production of carcinogenic acetaldehyde in achlorhydric atrophic gastritis. Scand J Gastroenterol 37:648–655 Warnakulasuriya S, Sutherland G, Scully C 2005 Tobacco, oral cancer, and treatment of dependence. Oral Oncol 41:244–260 Xue F-B, Xu Y-Y, Pan B-R et al 2003 Association of H.pylori infection with gastric carcinoma: a meta analysis. World J Gastroenterol 7:801–804 Zeka A, Gore R, Kriebel D 2003 Effects of alcohol and tobacco on aerodigestive tract cancer risk: a meta-regression analysis. Cancer Causes Control 14:897–906 Zheng TZ, Boyle P, Hu HF et al 1990 Dentition, oral hygiene, and risk of oral cancer: a case-control study in Beijing, People’s Rebublic of China. Cancer Causes Control 1:235–241 Yokoyama A, Muramatsu T, Ohmori T et al 1998 Alcohol-related cancers and aldehyde dehydrogenase-2 in Japanese alcoholics. Carcinogenesis 19:1383–1387

DISCUSSION Preedy: I have a general question about the level of mutagenicity which you cited. Where does this figure come from? M Salaspuro: Regarding earlier carcinogenicity studies, they were by and large in vitro studies and rather high acetaldehyde concentrations were mostly often used. Recently a good NIH study demonstrated that 100 µM acetaldehyde is able to produce, together with polyamines, mutagenic 1,N 2-propanodeoxyguanosine adducts (Theruvathu et al 2005). On the other hand, we also have indirect evidence from our human studies. Salivary acetaldehyde concentrations among drinking ALDH2-deficient individuals range from 50–200 µM. That is up to three times higher than among those with the normal ALDH2 enzyme (Väkeväinen et al 2000). On the other hand, their risk for oesophageal cancer is 12-fold compared with those with the normal enzyme (Yokoyama et al 1998). Morris: Just a point about terminology: formation of an adduct is not a mutation. Seitz: There are classic studies from Obe & Ristow (1977) in which the authors applied acetaldehyde to CHO cells. They were looking for sister chromatid exchanges. They found this occurring with 88 µM acetaldehyde. This is somewhat higher than 50 µM, but is still in the same range. Dr Salaspuro, there are some reports from Japan about blood alcohol levels for people who have Candida infections. Microbes may produce alcohol from carbohydrates via fermentation. This alcohol production in the gastrointestinal tract can lead to blood alcohol levels. What are expected alcohol blood levels just from bacterial ethanol production? M Salaspuro: This is an interesting comment and takes us back to the alcoholic fermentation reaction. There are several case reports from Japan of humans who had had some special gastrointestinal condition or operation favouring gastrointestinal bacterial overgrowth (Kaji et al 1984). Sometimes when they took a rice meal they reported that they felt drunk. Their blood alcohol concentrations were

90

SALASPURO

determined under controlled hospital conditions after a rice meal and up to 20 mM blood alcohol levels could be detected in those individuals. Their blood ethanol started to rise in about 30 minutes after the meal and the peak level was achieved within two hours. There is other evidence from Dr Bode in Germany who reported in 1985 that people who are using H2 blockers to decrease gastric acid secretion often have bacterial overgrowth in their stomach. When he infused them nasogastrically with nutrients he was able to demonstrate marked ethanol production in the stomach (Bode et al 1984). Unfortunately, neither blood nor intragastric acetaldehyde concentrations were measured in those studies. Apte: Has anyone looked at whether H. pylori can metabolize alcohol? M Salaspuro: Actually, we started our microbiological studies with H. pylori. The question was raised about how much alcohol is metabolised in the stomach, and the idea was that some of the first-pass oxidation of ethanol could be due to H. pylori. We went through several strains of H. pylori. Some had rather high ADH activity, others none. Under microaerobic conditions ADH-positive H. pylori strains were able to produce marked amounts of acetaldehyde from ethanol in vitro (Salmela et al 1993). This could be one possible factor in the pathogenesis of gastric cancer, since H. pylori is an established risk factor for gastric cancer. On the other hand, H. pylori infection often leads atrophic gastritis which is associated with endogenous formation of both ethanol and acetaldehyde mediated by bacteria or yeasts (Väkeväinen et al 2002). Seitz: When we saw the data from Dr Salaspuro we were interested in first-pass metabolism of alcohol. We thought that H. pylori might contribute to first-pass metabolism. However, H. pylori damages the mucosa of the stomach, leading to a decrease in mucosal ADH. Consequently, the first-pass metabolism in the presence of H. pylori was lower instead of higher because the Helicobacter by itself could not completely compensate for the loss of ADH in the mucosa. This means that the Helicobacter is probably responsible for local effects but not for a general effect with respect to ethanol metabolism. Rao: I want to raise the issue of the difficulty in measuring the acetaldehyde concentration in biological samples, largely because of its volatile nature. How confident are we that we are measuring it accurately? Could it be greater than what we measure? M Salaspuro: The method we are using is very reliable, and we can exclude the possibility of artefactual acetaldehyde formation from ethanol. Dr Seitz used another method for acetaldehyde determination in his studies and found more or less the same levels in saliva. But you are right: some of salivary acetaldehyde may be lost via evaporation. Seitz: Part of our experiment was performed in Germany and the other part was performed in Finland (Helsinki). We measured acetaldehyde using gas chromatography and our HPLC method with fluorescence, arriving at similar results.

ALCOHOL, SMOKING, ACETALDEHYDE AND CANCER

91

Eriksson: This whole set of studies is really impressive, but there is one aspect to your findings that is a little puzzling. It seems that 4-methylpyrazole had no effect on controls. On the other hand, you showed that the genetic polymorphism of ADH1C affected the acetaldehyde levels. How do you explain this? The share of this is very little, but it seemed big in your data. How much of the saliva acetaldehyde is systemic ADH-mediated, and how much is from other sources? M Salaspuro: That’s an important point which we have to consider. There is no transport of acetaldehyde from blood to the saliva. I think that in most of the Caucasians with normal ADH salivary acetaldehyde is produced by microbes. Eriksson: In one of your slides with active ADH, the acetaldehyde levels were almost 50% bigger. This is a common genotype in Finland. The results may depend on the condition under which you take the saliva. I’m guessing that you didn’t rinse your samples with antibacterial solutions in these genetic studies. In spite of this the acetaldehyde levels were much higher in individuals with the active ADH genotype. This implies that the ADH coming from sources other than the mucosa can be important. M Salaspuro: Perhaps we should do some additional inhibitor studies to find out the effect of various ADH isoforms on salivary acetaldehyde after alcohol is drunk. So far we don’t know how the rate of ethanol oxidation and acetaldehyde production is regulated at the mucosal level. There are too many open questions. But attempts to sterilize the mouth are futile because there are 500 or so different strains of bacteria in the mouth. We could reduce these populations by 50–70% or so at best. In the large intestine we can almost completely inhibit the intracolonic microbial acetaldehyde production by giving experimental animals or humans ciprofloxacin, an antibiotic that drastically decreases the number of faecal aerobic bacteria (Tillonen et al 1999, Homann et al 2000). On the other hand, antibiotics have only minor effects on the acetaldehyde production in saliva. Emery: You talk about ADH in the oral mucosa, but presumably there is also ADH in the salivary glands themselves. M Salaspuro: We couldn’t distinguish between the contribution from the mucosa and the salivary glands. Both can be contributing. Our parotid gland cannulation studies, however, indicate that at least in ALDH2-deficient subjects parotid glands contribute acetaldehyde to the saliva. Morris: I’m trying to sort out the dosimetry with respect to the tobacco smoke. Acetaldehyde is a component of tobacco smoke but it is among the least potent of the carcinogens in this smoke. What is the mass of acetaldehyde released by cigarette smoke and do these numbers add up with the amount seen in the saliva? Perhaps there is an indirect mechanism by which salivary acetaldehyde is being increased. M Salaspuro: I don’t know. The burning process in general produces acetaldehyde. There is no acetaldehyde production by tobacco that is chewed rather than smoked.

92

SALASPURO

Emery: The point is, those increases in salivary acetaldehyde were pretty massive compared with what you might expect to be produced by the cigarette. Morris: If there is only a milligram total acetaldehyde in the cigarette (Seeman et al 2002), the total dose passing via the saliva to the gut might be quite small. There are probably 60 other carcinogens present in cigarette smoke that are much more potent than acetaldehyde if we are trying to search for an explanation for some potential synergistic interaction. M Salaspuro: I don’t know the reliability of the analytical methods that have been used for the determination of the total acetaldehyde concentration of a cigarette. Anyway, the major point is that we do have a ‘genetic human knockout’ model only for acetaldehyde and not for the other potential carcinogens of tobacco smoke. V Salaspuro: If we can eliminate the acetaldehyde both from alcohol and tobacco, these studies could help determine the role of acetaldehyde. I will talk about this in my paper. Crabb: Is there any evidence that after you smoke there is an increase in blood acetaldehyde? Eriksson: There is no evidence for this. What we are missing are the calculations. The partition ratio is about 190. If you know the total amount from the smoke, you can calculate the increase in saliva. Regarding the carcinogenic effect of acetaldehyde, the studies by the Japanese (Yokoyama and collaborators) show that the heterozygotes have a fantastically increased risk of some cancers. There is a theoretical possibility that some of this cancer is coming from the beverages themselves: if you are drinking a lot, there are lots of moments when the acetaldehyde levels are quite high. M Salaspuro: That is a good point. If we go through the epidemiological data on the worldwide incidence of upper digestive tract cancers, there are certain areas where the incidence of these cancers is very high. One is north Iran, where people are frequently opium smokers. We don’t know whether or not this smoke contains acetaldehyde. In China, in Linxian about 20% of the population dies from oesophageal cancer. The incidence of oesophageal cancer there is 150 times higher than among white Americans. Researchers from the NIH have been working there for 20 years and have tried to find reasons for this high incidence. The two factors they have pinpointed are the presence of smoke in the house (this is an area where there are no chimneys) and the widespread use of pickled vegetables, which are produced by fermentation. These vegetables may have a very high acetaldehyde concentration, and perhaps some ethanol too. An additional risk factor is poor oral hygiene. Seitz: I have a comment about the number of carcinogens in tobacco smoke. We have been interested in this for a long time. We looked at the induction of CYP2E1 and activation of procarcinogens, and showed that nitrosamines are activated by CYP2E1. If we look at lung cancer, there are some reports of CYP2E1 induction

ALCOHOL, SMOKING, ACETALDEHYDE AND CANCER

93

in bronchial epithelia, which may lead to increased activation of nitrosamines in cigarette smoke. However, lung cancer is not associated very much with alcohol drinking, and therefore this is probably not a mechanism related to alcohol and lung cancer. Apte: In contrast to that, you mentioned oral tobacco use, which is common in India. There is an increased risk of oral cancer associated with this, but you said there is no acetaldehyde formation. Could you do a study comparing the incidence of cancer in oral versus smoked tobacco? M Salaspuro: That is an excellent idea. Indeed, in India the incidence of oral cancer is particularly high. The high incidence has been related to smoking, reverse smoking and chewing of betel nuts. Betel nuts and other ingredients including nicotine, lime and sweeteners are wrapped in a leaf that is kept in the cheek pocket even for several hours. This could lead to local production of acetaldehyde via alcoholic fermentation. Morris: To add another layer of complexity, vinyl acetate is an oral carcinogen in rodent species (USEPA 1997), and is metabolized to acetaldehyde. I don’t think there’s any denying that acetaldehyde is an intriguing player in oral cancer, but I don’t think it’s the one ingredient in cigarette smoke that is responsible. Albano: Just to complete the story, tobacco chewing in India often involves the use of mixtures with betel leaves that have been recognized by IARC (International Agency of Research on Cancer) as a human carcinogen. Apte: The mixture chewed also has a high concentration of lime. Albano: I’d like to get a general view from the people in the field. Could we devise a hypothesis on the pathogenesis of upper digestive tract and oral cancer related to alcohol that sees three main players: (a) acetaldehyde produced in the saliva by bacteria; (b) alcohol itself; and (c) oxidative stress? How might the genetic polymorphism in ALDH interfere with the catabolism of other aldehydes, particularly those derived from lipid peroxidation as malondialdehyde and 4-HNE? Is there any information about the possibility that these aldehydes might contribute to mutagenesis in the upper digestive mucosa by making adducts with DNA? Seitz: At the moment we are studying exocyclic etheno-DNA adducts, which come from 4-HNE and other lipid peroxidation products. We have studied this in the liver but we have not looked in other tissues. This would be interesting, because if we find increased levels it could come from oxidative stress, which may then have an additional role with respect to what you mentioned. It would be interesting to study both etheno-DNA adducts resulting from oxidative stress and acetaldehydeDNA adducts to compare them quantitatively and to find out which of these adducts are more relevant with respect to cancer development. Eriksson: Next spring is the IARC new update, and we are going to focus on the carcinogenic effects of acetaldehyde. So this meeting is at a good time.

94

SALASPURO

M Salaspuro: In the 1980s the IARC concluded that ethanol itself is not carcinogenic. In 1999 they concluded that acetaldehyde is carcinogenic in animals and reasonably anticipated to be a human carcinogen, but at that time the evidence on humans was not yet so strong. Since 1999 a lot of new information has arisen. It is likely that acetaldehyde will soon be considered to be carcinogenic in human, too. It is important that this question will be re-evaluated. Apte: You mentioned the amount of acetaldehyde in yoghurt, which surprised me. How much is there, and is it in all yoghurts? M Salaspuro: Lactobacillus bulgariensis and Streptococcus thermophilus are microbes capable of alcoholic fermentation and are used worldwide for the production of yoghurts. The former has been used in Bulgaria for hundreds of years. Manufacturers are searching for new bacteria strains that are even better acetaldehyde producers. We have detected up to 800 µM acetaldehyde concentrations in some yoghurts. But on the other hand there are some other fermented dairy products with no acetaldehyde. In these cases other bacteria not capable for alcoholic fermentation are used. The time of exposure must play an essential role in cancer pathogenesis. I don’t think it is dangerous to have one yoghurt per day. Preedy: Are there any data on acetaldehyde in cheese? M Salaspuro: We have analysed some cheeses but so far we have not been able to find any acetaldehyde in them. We do find some in smoky fish. Emery: How much is in pickled vegetables and sauerkraut? M Salaspuro: In commercially available pickled vegetables it’s in the range of 0–3500 µM. In some areas these are used daily. Eriksson: One point that hasn’t been raised is that some alcoholic beverages with high acetaldehyde are taken as hot drinks, which will affect the membranes. Sometimes calvados is drunk with hot coffee. Okamura: Some sake is served warm and is drunk with salty food. This would also have a harmful effect on the membranes. M Salaspuro: With regard to calvados, the old habit has been for centuries that people make their own calvados, and this is often drunk as half calvados and half hot coffee. Seitz: The hot alcoholic drink might create some damage, followed by hyperregeneration. Earlier on it was mentioned that acetaldehyde is an inhibitor of retinoic acid production. Some studies from Japan show that acetaldehyde in the oesophagus blocks retinoic acid production (Shiraishi-Yokoyama et al 2006). This creates an increase in AP1 gene expression in the hyper-regeneration, at least in the liver. Hyper-regeneration is a good platform for acetaldehyde acting as a toxin or carcinogen, since under hyper-regenerative conditions in the presence of biogenic amines an increase in acetaldehyde DNA adducts occurs. Rao: One of the issues concerning the carcinogenicity of different factors is the combinatorial effect where two different factors work synergistically. Acetalde-

ALCOHOL, SMOKING, ACETALDEHYDE AND CANCER

95

hyde could have effects at a lower concentration in synergy with other potential carcinogens. Perhaps the IARC should consider this. Emery: It is difficult to quantify this. It would be impossible to specify in terms of limits or guidelines. M Salaspuro: Epidemiological evidence suggests that the risk of oesophageal cancer is increased 150-fold among those who are consuming one and a half bottles of wine a day and smoking over 30 cigarettes a day, as compared with someone who doesn’t smoke or drink (Tuyns et al 1977). Heavy, intoxicating drinking results in much higher salivary acetaldehyde levels because microbial ADHs are not saturated with ethanol. The stronger and longer exposure to acetaldehyde is associated with higher cancer risk. Accordingly, there is a good agreement between the biochemical and epidemiological evidence. Seitz: They had problems finding the control group in Brittany because very few people didn’t drink or smoke! The control group is therefore small. Ren: You used 4-methylpyrazole to nullify the difference between the flushers and non-flushers. What dose did you use? And after you remove the ADH are you suggesting that the remaining amount of acetaldehyde is being produced by CYP2E1 and catalase? Or is the inhibition of ADH incomplete? M Salaspuro: The dose of 4-methylpyrazole was 10–15 mg/kg orally 2 hours before the experiments. The inhibition of ADH by 4-methylpyrazole is competitive, but the dose we used was enough to prevent the spill-over of acetaldehyde from the salivary glands to the saliva. The remaining acetaldehyde was produced from salivary ethanol by the oral microbes. Preedy: I have a question about poor oral hygiene. I went to a meeting at which someone suggested that the use of mouthwashes is a risk factor for oral cancer. If you have poor oral hygiene and you want to use a mouthwash you are replacing one risk with another. M Salaspuro: It depends what kind of mouthwash you are using. If it contains ethanol there will be immediate local acetaldehyde production. There are other types that don’t contain ethanol so these don’t have the same risk. References Bode JC, Rust S, Bode C 1984 The effect of cimetidine treatment on ethanol formation in the human stomach. Scand J Gastroenterol 19:853–856 Homann N, Tillonen J, Salaspuro M 2000 Microbially produced acetaldehyde from ethanol may increase the risk of colon cancer via folate deficiency. Int J Cancer 86:169–173 Kaji H, Asanuma Y, Yahara O et al 1984 Intragastrointestinal alcohol fermentation syndrome: report of two cases and review of the literature. J Forensic Sci Soc 24:461–471 Obe G, Ristow H 1977 Acetaldehyde, but not ethanol, induces sister chromatid exchanges in Chinese hamster cells in vitro. Mutat Res 56:211–213 Salmela KS, Roine RP, Koivisto T, Hook-Nikanne J, Kosunen TU, Salaspuro M 1993 Characteristics of Helicobacter pylori alcohol dehydrogenase. Gastroenterology 105:325–330

96

SALASPURO

Seeman JI, Dixon M, Haussmann H-J 2002 Acetaldehyde in mainstream tobacco smoke: formation and occurrence in smoke and bioavailability in the smoker. Chem Res Toxicol 15:1331–1350 Shiraishi-Yokoyama H, Yokoyama H, Matsumoto M, Imaeda H, Hibi T 2006 Acetaldehyde inhibits the formation of retinoic acid from retinal in the rat esophagus. Scand J Gastroenterol 41:80–86 Theruvathu JA, Jaruga P, Nath RG, Dizdaroglu M, Brooks PJ 2005 Polyamines stimulate the formation of mutagenic 1,N2-propanodeoxyguanosine adducts from acetaldehyde. Nucleic Acids Res 33:3513–3520 Tillonen J, Homann N, Rautio M, Jousimies-Somer H, Salaspuro M 1999 Ciprofloxacin decreases the rate of ethanol elimination in humans. Gut 44:347–352 Tuyns AJ, Pequignot G, Jensen OM 1977 Esophageal cancer in Ille-et-Vilaine in relation to levels of alcohol and tobacco consumption. Risks are multiplying. Bull Cancer 64:45–60 United States Environmental Protection Agency (USEPA) 1997 Carcinogenesis study of vinyl acetate (drinking water study) in rats and mice with cover letter dated 01/31/1997, EPA/OTS, FYI-OTS-0297-1286 Vakevainen S, Tillonen J, Agarwal DP, Srivastava N, Salaspuro M 2000 High salivary acetaldehyde after a moderate dose of alcohol in ALDH2-deficient subjects: strong evidence for the local carcinogenic action of acetaldehyde. Alcohol Clin Exp Res 24:873–877 Vakevainen S, Mentula S, Nuutinen H et al 2002 Ethanol-derived microbial production of carcinogenic acetaldehyde in achlorhydric atrophic gastritis. Scand J Gastroenterol 37:648–655

Effects of acetaldehyde on human airway constriction and inflammation Hiroto Matsuse, Chizu Fukushima, Terufumi Shimoda, Sadahiro Asai and Shigeru Kohno Second Department of Internal Medicine, Nagasaki University School of Medicine, 1-7-1 Sakamoto, Nagasaki 852-8501, Japan

Abstract. The purpose of the present study was to determine the effects of acetaldehyde on airway smooth muscle constriction and inflammation. An oral ethanol provocation test was performed in Japanese asthmatics to measure pulmonary function, blood ethanol, acetaldehyde and histamine. Acetaldehyde dehydrogenase 2 (ALDH2) genotype was determined by polymerase chain reaction (PCR) and ethanol patch test. Human bronchi and mast cells were stimulated with acetaldehyde in vitro. Mite allergen-sensitized mice were inoculated with intranasal acetaldehyde. Approximately half the asthmatic subjects developed bronchoconstriction with concomitant increases in blood acetaldehyde and histamine, which was associated with genetically reduced ALDH2 activities. In vitro acetaldehyde stimulation induces bronchoconstriction and degranulation of human mast cells. It also induced granulocyte macrophage colony stimulating factor (GM-CSF) production and nuclear factor (NF)-κB activation in human bronchi and increased mite allergen-sensitized allergic inflammation in a murine model of asthma. We conclude that acetaldehyde has potential effects on human airway by two distinct mechanisms. As a metabolite of alcohol, its elevation following alcohol consumption induces airway mast cells to release histamine, which results in exacerbation of asthma in susceptible populations. And as an air pollutant contained in cigarette smoke, for example, its inhalation potentially increases airway inflammation. 2007 Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Foundation Symposium 285) p 97–109

A small group of European white subjects and approximately half of the Japanese population are estimated to experience exacerbation of asthma after consumption of alcoholic drinks; however, the underlying mechanisms differ in these two groups. Certain antigens, preservatives, or both, present in alcoholic beverages cause asthma exacerbation in the former (Dahl et al 1986). In contrast, acetaldehyde, a metabolite of alcohol, seems to play a critical role in alcohol-induced asthma in the latter group, which is associated with genetically controlled enzymatic activity of acetaldehyde dehydrogenase 2 (ALDH2), a primary catabolic enzyme of acetaldehyde. Additionally, acetaldehyde is not only a metabolite of alcohol but is also present in many products such as plastic and rubber goods, and in cigarette smoke. Acetaldehyde is 97

98

MATSUSE ET AL

thus considered as an air pollutant that causes airway injury (Wyatt et al 1999). Thus acetaldehyde potentially causes two distinct pathological conditions in the airway, as a primary mediator involved in alcohol-induced asthma in a subset of asthmatics and as an air pollutant that increases airway inflammation. Here we show the mechanisms of alcohol-induced asthma utilizing human asthmatic subjects and isolated human bronchi. We also show the effects of acetaldehyde on the airway inflammation in isolated human bronchi and in a murine model of asthma. Methods Oral ethanol provocation test in asthmatic subjects Pure ethanol was dissolved in 5% glucose solution to yield a 10% ethanol solution. Asthmatic and healthy subjects drank 300 ml of this solution in 5 to 10 minutes. Determinations of pulmonary function, blood ethanol, acetaldehyde and histamine were performed before and 15, 30, 60 and 120 minutes after ethanol challenge. Patients showing a 20% or greater fall in forced expiratory volume in 1 second (FEV1) following oral ethanol provocation were considered responders, as described elsewhere (Shimoda et al 1996). Pretreatment with a histamine H1 receptor antagonist A 1 week wash-out period was allowed for positive responders to the oral ethanol provocation test. During the following week, the responder was given 2 mg of azelastine hydrochloride, a histamine H1 receptor antagonist, and the second oral ethanol provocation test was performed the day after the last dose of azelastine, as described elsewhere (Takao et al 1999). Determination of ALDH2 genotype by PCR and ethanol patch test To determine ALDH2 genotype, we subjected blood samples from asthmatics and healthy subjects to polymerase chain reaction (PCR) analysis as described elsewhere (Takao et al 1998). PCR results were stratified into type NN (normal homozygote) or NM (mutant heterozygote) or MM (mutant homozygote). Ethanol patch test was also performed on the inner surface of the upper arm of the same subjects. A patch area that showed erythema of 15 mm represented a positive result for reduced ALDH2 activity as described elsewhere (Matsuse et al 2001). In vitro acetaldehyde stimulation to human bronchi and airway mast cells Human bronchi were prepared from the resected lung tissues from lung cancer patients and were suspended in a Magnus bath containing Krebs-Henseleit solu-

ACETALDEHYDE IN HUMAN AIRWAY DISEASE

99

tion. The upper surface of the specimen was suspended by an isometric transducer to record the contractile tension by a pen recorder. The suspended bronchi were stimulated with acetaldehyde and the contractile responses were recorded. 30 minutes after stimulation, buffer samples were collected and were subjected for determination of histamine by radioimmunoassay (RIA). Additionally, airway mast cells were isolated from lung tissue by means of immunomagnetic methods. These mast cells were directly stimulated with acetaldehyde and the concentrations of histamine in the cultured medium were determined by RIA as previously described (Kawano et al 2004). Effects of acetaldehyde on airway inflammation Human bronchi were cultured in the presence of acetaldehyde for 24 hours and the concentrations of proinflammatory cytokines were determined. These tissues were also immunohistochemically stained for nuclear factor (NF)-κB p65 as described elsewhere (Machida et al 2003). Intranasal inoculation of acetaldehyde to murine model of allergic asthma Female BALB/c mice were intraperitoneally sensitized with mite allergen followed by intranasal (i.n.) mite allergen challenge to develop allergic airway inflammation. Thereafter, these mice were inoculated i.n. with low concentration (3%, 50 µl/ mouse) of acetaldehyde. Haematoxylin and eosin (H&E)-stained pulmonary pathology was compared among four groups of mice; control, acetaldehyde inoculated only, mite allergen sensitized only, and mite allergen sensitized plus acetaldehyde inoculated. Results and discussion Mechanism of alcohol-induced asthma in vivo 55% of asthmatic subjects were regarded as responders by the oral ethanol provocation test. The remaining asthmatics and healthy subjects did not show a significant fall in their pulmonary function and were regarded as non responders (Fig. 1A). Blood acetaldehyde and histamine, but not ethanol, were significantly higher in responders than in non-responders (Fig. 1B, C, D). These results indicated that an increase in blood acetaldehyde, but not ethanol, induces increased blood histamine levels, which cause bronchoconstriction in a subset (approximately half ) of Japanese asthmatics. To further confirm this mechanism, responders were pretreated with azelastine hydrochloride, a histamine H1 receptor antagonist, and were subjected to a second oral ethanol provocation test. Pretreatment of azelastine

100

MATSUSE ET AL

FIG. 1. Serial changes in (A) FEV1, (B) serum ethanol, (C) acetaldehyde and (D) histamine following oral ethanol provocation test. Bars represent mean (n = 21 for responders, n = 25 for non-responders, n = 40 for healthy subjects) ±SEM. *P < 0.05 and **P < 0.01 vs. non-responders.

completely inhibited ethanol-induced bronchoconstriction, suggesting a critical role of histamine in this phenomenon. Genetic mechanism of alcohol-induced asthma Similarly to asthma, other responses to alcohol differ among races. Mongoloid populations often show facial flushing, palpitation and nausea after a small amount of alcohol consumption (Wolff 1972). In contrast European white populations generally drink large amounts of alcohol. This difference in alcohol metabolism is based on differences in ALDH2 enzymatic activity among races. The gene encoding ALDH2 is located in the long arm of chromosome 12. ALDH2 becomes inactive when residue 487 (glutamic acid) is replaced with lysine as a result of point mutation of the 12th exon. ALDH2 is a tetramer, and all four subunits must be normal for the enzyme to retain its activity. In the type MM (mutant homozygote) ALDH2 gene, all four tetramers are absent, and thus no ALDH2 activity is present. In the type NM (mutant heterozygote) ALDH2 gene, there are only a few normal tetramers, resulting in a low ALDH2 activity (Enomoto & Takada 1990). Types NN, NM and MM make up 56.4%, 39.4%, and 4.2% of the Japanese populations, respectively. In contrast, type NN is present in almost all of European white

ACETALDEHYDE IN HUMAN AIRWAY DISEASE

101

TABLE 1 The distribution of ALDH2 genotype determined by PCR in asthmatic and healthy subjects ALDH2 genotype: Asthmatic Healthy

NN

NM

MM

23 (50.0) 22 (55.0)

20 (43.5) 16 (40.0)

3 (6.5) 2 (5.0)

Results are shown as n (%) of NN (normal homozygote), NM (mutant heterozygote), MM (mutant homozygote).

TABLE 2 Relationship between ALDH2 genotype determined by PCR and results of oral ethanol provocation test in asthmatic subjects ALDH2 genotype: Positive Negative

NN

NM

MM

4 (17.4) 19 (82.6)

14 (70.0) 6 (30.0)

3 (100) 0 (0)

Results are shown as n (%) of NN (normal homozygote), NM (mutant heterozygote), MM (mutant homozygote).

TABLE 3 Relationship between ALDH2 genotype determined by PCR and results of ethanol patch test in asthmatic subjects ALDH2 genotype Positive Negative

NN

NM

MM

4 (5.9) 80 (100)

56 (82.4) 0 (0)

8 (11.8) 0 (0)

Results are shown as n (%) of NN (normal homozygote), NM (mutant heterozygote), MM (mutant homozygote).

populations (Harada 1990). PCR analysis for ALDH2 genotypes was performed to determine why only a subset of subjects developed increased blood acetaldehyde levels following oral ethanol provocation. Results demonstrated that significant differences did not occur in the frequency of ALDH2 genotypes between asthmatics and healthy subjects (Table 1), while the percentage of responders to oral ethanol provocation test was higher among those with inactive ALDH2 genotypes (NM and MM) than those with the normal active types (NN) (Table 2). Results of ethanol patch testing correlated well with ALDH2 genotype determined by PCR (Table 3) and further confirmed that reduced ALDH2 activity based on ALDH2 genotype differences was found to be significantly higher in responders.

102

MATSUSE ET AL

Taken together, we propose the following mechanism of alcohol-induced asthma. Orally taken ethanol is decomposed by alcohol dehydrogenase to yield acetaldehyde, which is then degraded by ALDH2 to yield acetic acid. In approximately half of Japanese subjects, the enzymatic activities of ALDH2 are genetically reduced and thus blood acetaldehyde levels are markedly elevated following alcohol consumption. Acetaldehyde induces histamine release, resulting in facial hot flushes and bronchoconstriction, i.e. alcohol-induced asthma. In contrast, ALDH2 activity is normal in almost all of European Caucasian whites, thus they develop neither facial hot flushes nor alcohol-induced asthma since blood acetaldehyde does not markedly increase after alcohol consumption.

Mechanism of alcohol-induced asthma in vitro Although it is likely that mast cells play a critical role in acetaldehyde-induced bronchoconstriction via production of histamine in human asthma, as mentioned above, there had been few studies to evaluate the direct effects of acetaldehyde on mast cells (Koivisto et al 1999, Ruiz & Gomes 2000). Thus, we stimulated human resected bronchi and isolated mast cells with acetaldehyde in vitro. Acetaldehyde increased airway muscle tone (Fig. 2), which was associated with a significant increase in the release of histamine (medium stimulation 26.3 ± 4.8 M/g

FIG. 2. Human bronchi were stimulated with either medium (upper) or acetaldehyde (lower). Smooth muscle constriction was recorded by polygraph. Arrows indicate points of stimulation.

ACETALDEHYDE IN HUMAN AIRWAY DISEASE

103

vs. acetaldehyde stimulation 98.7± 8.9 M/g; P < 0.05). Acetaldehyde also directly induced a significant histamine release from isolated human airway mast cells (Fig. 3). Taken together, these in vitro experiments have repeatedly shown and confirmed the in vivo experiments that acetaldehyde directly stimulates mast cells to release histamine, which causes airway smooth muscle constriction.

Effects of acetaldehyde on airway inflammation Since acetaldehyde could be involved in airway inflammation as an air pollutant, we measured production of an inflammatory cytokine, granulocyte macrophage colony stimulating factor (GM-CSF) and activation of NF-κB, a transcription factor, from acetaldehyde-stimulated human bronchi. Acetaldehyde significantly increased GM-CSF production from human bronchi and nuclear translocation of NF-κB in airway epithelium (Fig. 4A,B). Both GM-CSF and NF-κB are critically involved in the development and maintenance of airway inflammation, and environmental acetaldehyde might enhance allergic airway inflammation observed in human asthma. In fact, a low concentration of acetaldehyde, which per se could not cause airway inflammation, significantly enhanced pre-existing mite allergen-induced allergic

FIG. 3. Acetaldehyde-induced histamine secretion from isolated human airway mast cells. Bars represent means ±SEM of eight independent experiments. *P > 0.05 vs. before stimulation.

104

MATSUSE ET AL

FIG. 4. (A) Acetaldehyde-induced GM-CSF production from human bronchi. Bars represent mean ± SEM of eight independent experiments. *P > 0.05 vs. medium stimulation. (B) Acetaldehyde-induced NF-κB nuclear translocation. Representative micrographs from eight independent experiments are shown.

airway inflammation (Fig. 5). Although the underlying mechanism is yet to be determined, it is possible that acetaldehyde-induced GM-CSF could activate inflammatory cells including eosinophils, macrophages and dendritic cells. Concluding remarks Collectively, our experimental results indicate that acetaldehyde has potential effects on human airway by two distinct mechanisms. As a metabolite of alcohol, its elevation following alcohol consumption induces airway mast cells to release histamine, which results in exacerbation of asthma in susceptible populations. As an air pollutant contained in cigarette smoke, for example, its inhalation potentially increases airway inflammation. Acknowledgements This study is supported by a grant-in-aid for scientific research (No. 12670563) by the Japanese Society for the Promotion of Science, and a grant from the Kanae Foundation for Life & Socio-Medical Science.

ACETALDEHYDE IN HUMAN AIRWAY DISEASE

105

FIG. 5. Pulmonary pathology of (A) control, (B) acetaldehyde inoculated, (C) mite allergen sensitized, (D) mite allergen sensitized and acetaldehyde inoculated BALB/c mice. Representative photomicrographs of each group (n = 4) are shown.

References Dahl R, Henriksen JM, Henning H 1986 Red wine asthma, a controlled challenge study. J Allergy Clin Immunol 78:1126–1129 Enomoto N, Takada A 1990 Acetaldehyde metabolism and aldehyde dehydrogenase 2 gene. J Exp Med 154:823–828 Harada S 1990 Racial and genetic factors in ethanol and aldehyde metabolism. J Exp Med 154:817–822 Kawano T, Matsuse H, Kondo Y et al 2004 Acetaldehyde induces histamine release from human airway mast cells to cause bronchoconstriction. Int Arch Allergy Immunol 134:233–239 Koivisto T, Kaihovaara P, Salaspuro M 1999 Acetaldehyde induces histamine release from purified rat peritoneal mast cells. Life Sci 64:183–190 Machida I, Matsuse H, Kondo Y et al 2003 Acetaldehyde induces granulocyte macrophage colony-stimulating factor production in human bronchi through activation of nuclear factor-κB Allergy Asthma Proc 24:367–371 Matsuse H, Shimoda T, Fukushima C et al 2001 Screening for acetaldehyde dehydrogenase 2 genotype in alcohol-induced asthma by using the ethanol patch test. J Allergy Clin Immunol 108:715–719

106

MATSUSE ET AL

Ruiz CM, Gomes JC 2000 Effects of ethanol, acetaldehyde, and acetic acid on histamine secretion in guinea pig lung mast cells. Alcohol 20:133–138 Shimoda T, Kohno S, Takao A et al 1996 Investigation of the mechanism of alcohol-induced bronchial asthma. J Allergy Clin Immunol 97:74–84 Takao A, Shimoda T, Kohno S, Asai S, Harada S 1998 Correlation between alcohol-induced asthma and acetaldehyde dehydrogenase-2 genotype. J Allergy Clin Immunol 101:576–580 Takao A, Shimoda T, Matsuse H et al 1999 Inhibitory effects of azelastine hydrochloride in alcohol-induced asthma. Ann Allergy Asthma Immunol 82:390–394 Wyatt TA, Heires AJ, Sanderson SD, Floreanni AA 1999 Protein kinase C activation is required for cigarette smoke-enhanced C5a-mediated release of interleukin-8 in human bronchial epithelial cells. Am J Respir Cell Mol Biol 21:283–288 Wolff PH 1972 Ethnic differences in alcohol sensitivity. Science 175:449–450

DISCUSSION M Salaspuro: The blood ethanol levels in your experiments were rather low, at around 5–10 mM, even though you gave subjects a rather high dose of ethanol. The blood acetaldehyde levels ranged from 10–30 µM, which is similar to the range we see in our experiments. How did you measure your blood acetaldehyde? Matsuse: The concentration of blood acetaldehyde was measured by the outside lab. They used GC. M Salaspuro: In your study the flushing and other effects occurred immediately after the end of the experiment, and was associated with a rapid histamine release. Could you repeat your study but ask subjects to keep alcohol only in their mouth and not swallow it. In this case microbial acetaldehyde production starts immediately and might result in the local release of histamine from the mucosal mast cells without any systemic effects. Matsuse: We had never done such a study. Thank you for your good suggestion. Morris: Your results are intriguing because we don’t think of acetaldehyde as being that potent an irritant by itself. You might want to be careful in your animal models. I have shown that if you inhibit ALDH, this inhibits sensory nerve stimulation by acetaldehyde in the rat (Stanek et al 2001). Rodents may not be a good model for the human. In the rodent it seems that acetic acid is the key player for the respiratory tract effects. Matsuse: We agree with you. Actually, I wanted to develop an animal model of alcohol-induced asthma in mice, but mice aren’t good for histamine research. Morris: Was your effect specific for GM-CSF? Matsuse: That is all I have measured. Apte: As John mentioned it was surprising to see that there is an induction of NF-κB, but then no induction of TNF or interleukins. In terms of your results in the human studies you saw blood acetaldehyde levels going up and coming down around the 120 min mark. But the histamine levels keep rising, so could there be something other than acetaldehyde which keeps the mast cells in a state of degranulation.

ACETALDEHYDE IN HUMAN AIRWAY DISEASE

107

Matsuse: We have measured other mediators such as leukotrienes, but only histamine was elevated. I have no idea why this happens. Apte: Do mast cells themselves have ADH, or is acetaldehyde positively chemotactic for mast cells? As far as I could tell your patients didn’t have active asthma at the time. The asthma was under control, so there was presumably less inflammation and fewer mast cells, yet there is a huge histamine release. M Salaspuro: We have used isolated human mast cells and were able to show histamine release caused by acetaldehyde of about 100 µM. Matsuse: I have no idea whether mast cells themselves have ADH, or whether acetaldehyde is positively chemotactic for mast cells. Morris: The time course is interesting. Was bronchoconstriction immediate? Was there enough time to get mast cells coming in? Apte: It happened within 30 min. Is this enough time for mast cell migration? Morris: It is for neutrophils; I don’t know about mast cells. Matsuse: We don’t think histamine is released from migrating mast cells but it is released from mucosal mast cells. Apte: There are tissue mast cells that are already present. But even if it is all tissue mast cells, there needs to be some chemotaxis. Eriksson: The first point made by Mikko Salaspuro is important. This finding has been documented before in Japanese populations. It is well established that acetaldehyde is responsible. But which mast cells are involved, and where are they located? The first flushing is in the face and neck. It would be interesting to see where it starts and to uncover the mechanism. Rao: It is interesting that a small dose of acetaldehyde by itself doesn’t do anything, but in allergen-mediated inflammation it can have an effect. This touches on the discussion we had yesterday about the carcinogenic concentration of acetaldehyde. A low dose of acetaldehyde can, in concert with other factors, be injurious. Deitrich: You mentioned that there is acetaldehyde in plastic and rubber. There are a lot of other things that leach out of plastic bottles into alcohol, such as phthalates and triphenol phosphate (Goldstein et al 1987). Niemelä: I am curious about the possible immunological basis of this phenomenon. If you repeat the experiments several times on the same subjects, would the symptoms remain similar or would they get worse? In mast cell reactions, there may be other immunological players involved which we don’t yet know about. Matsuse: We repeated the experiment at least twice in the same subjects and found similar symptoms. Morris: Mast cells put out a variety of pro-inflammatory mediators, not just histamine. Continuous stimulation of mast cells would exacerbate the inflammatory aspects of asthma. Apte: Could you desensitize the mast cells by repeated challenge?

108

MATSUSE ET AL

Morris: In the process of desensitizing inflammation might get worse, but it might get better in the long term. Matsuse: Another group have used inhaled acetaldehyde to cause bronchoconstriction. They showed desensitisation with repeated challenge. Thornalley: In cigarette smoke there are many other aldehydes, some of which are far more potent than acetaldehyde and which have been shown to induce histamine secretion. These are also substrates for ALDH. Morris: In that context, asthmatic subjects are hyper-responsive to a wide variety of irritants. In the mouse models this is also the case. It may not be specific to acetaldehyde. Emery: This also brings us back to the comment made about pollution. Acetaldehyde may be one of the irritant molecules, but do we have any feel for how this relates to other air pollutants? Morris: In terms of respiratory tract irritation acetaldehyde is not considered to be a key player. There are other more potent irritants present in potentially much more hazardous concentrations. Matsuse: The number one irritant in human environments is formaldehyde. The next one is acetaldehyde. Does anyone know whether ALDH2 can metabolise formaldehyde? Morris: No, it requires formaldehyde dehydrogenase. Deitrich: Formaldehyde is a poor substrate for ALDH2. Morris: As an irritant formaldehyde is a 1000 times more potent than acetaldehyde. Deitrich: Formaldehyde is almost completely hydrated, and acetaldehyde is only about 50% hydrated. This may be why formaldehyde is not a good substrate for ALDHs. Apte: Just a point of clarification. Were your human subjects non-smokers? Or had they just stopped smoking? This could be a confounding factor. Matsuse: Some are current smokers and some are ex-smokers. Apte: So you had a mixture of people, some of whom were still smoking when they entered the study. That’s something to be careful of. Crabb: A prolonged increase in histamine would turn on gastric acid secretion if the concentration is high enough. Is there a connection with alcohol flushing and duodenal ulcers? Matsuse: I have no idea about duodenal ulcers. Emery: Are there any genetic data on ulcers? Could this be related to the polymorphisms? Yin: We published a paper in 1992 about human lung ADH and ALDH (Yin et al 2002). ADH1B is the only form present in human lung for ethanol oxidation. ADH1B2 is high activity. It is possible that high activity of ADH1B2 produces acetaldehyde that can be accumulated in the lung cells in individuals with deficient

ACETALDEHYDE IN HUMAN AIRWAY DISEASE

109

ALDH2 activity. The locally produced acetaldehyde may also contribute to bronchoconstriction and alcohol-related cytotoxicity. References Goldstein DB, Feistner GJ, Faull KF, Tomer KB 1987 Plasticizers as contaminants in commercial ethanol. Alcohol Clin Exp Res 11:521–524 Stanek J, Symanowicz PT, Olsen JE, Gianutsos G, Morris JB 2001 Sensory-nerve-mediated nasal vasodilatory response to inspired acetaldehyde and acetic acid vapors. Inhal Toxicol 13:807–822 Yin SJ, Liao CS, Chen CM, Fan FT, Lee SC 1992 Genetic polymorphism and activities of human lung alcohol and aldehyde dehydrogenases: implications for ethanol metabolism and cytotoxicity. Biochem Genet 30:203–215

The role of acetaldehyde in alcohol-associated cancer of the gastrointestinal tract Helmut K. Seitz and Nils Homann* Laboratory of Alcohol Research, Liver Disease and Nutrition, Department of Medicine, Salem Medical Center Heidelberg and University of Heidelberg, Heidelberg, Germany and *Medical University of SchleswigHolstein, Department of Gastroenterology, Luebeck, Germany

Abstract. Acetaldehyde has been classified as a carcinogen in experimental animal research. Acetaldehyde is highly toxic, mutagenic and carcinogenic. Acetaldehyde causes point mutations, sister chromatid exchanges and gross chromosomal aberrations. In the liver, acetaldehyde binds to DNA and the formation of stable adducts represents one mechanism by which acetaldehyde could trigger the occurrence of replication errors and/or mutations in oncogenes or tumour suppressor genes. In experimental colorectal carcinogenesis the inhibition of acetaldehyde dehydrogenase with elevated acetaldehyde levels results in an acceleration of cancer development. The production of acetaldehyde is reduced when germ-free animals are studied, emphasizing the role of bacteria in the generation of colorectal acetaldehyde. Acetaldehyde levels in the colorectum correlate with crypt cell production rate and result in hyper-regeneration, a precancerous condition. Genetic linkage studies give further evidence for acetaldehyde as a carcinogen. Individuals who accumulate acetaldehyde due to polymorphism and/or mutations in the genes coding for enzymes responsible for acetaldehyde generation and detoxification have an increased cancer risk. This is true for Asians with low acetaldehyde dehydrogenase 2 and for Caucasians with alcohol dehydrogenase 1C*1/1. In conclusion, there is an enormous body of evidence from in vitro studies, animal experiments and genetic linkage studies, that acetaldehyde is the major factor responsible for tumour development in alcohol-associated carcinogenesis of the gastrointestinal tract. 2007 Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Foundation Symposium 285) p 110–124

Alcohol per se is not a carcinogen, while acetaldehyde has been identified as a mutagen and carcinogen. In this overview the mechanisms by which acetaldehyde acts in alcohol-associated carcinogenesis and the evidence for acetaldehyde as being the major carcinogenic principle in alcohol-associated cancer development will be discussed. Carcinogenesis of the upper alimentary tract and the liver will be considered only. It is beyond the scope of this review to discuss acetaldehyde effects 110

GI TRACT CANCER

111

on other tissues such as the breast as well as effects of acetaldehyde on metabolic and signal transduction pathways involved in carcinogenesis in detail. Acetaldehyde, a carcinogen The first and major metabolite of ethanol oxidation is acetaldehyde (AA). There is increasing evidence that AA rather than alcohol itself is responsible for the cocarcinogenic effect of alcohol (Pöschl & Seitz 2004). Numerous in vitro and in vivo experiments in prokaryotic and eukaryotic cell cultures as well as in animal models have identified AA as highly toxic, mutagenic and carcinogenic. Inhalation of AA in rats and hamsters resulted in the occurrence of carcinomas in the nasal mucosa and in the larynx (Woutersen et al 1986, Feron et al 1982). AA interferes at many sites with DNA synthesis and repair and may, consequently, result in tumour development (IARC 1999). AA causes point mutations in the hypoxanthine-guanine-phosphorybosyl transferase locus in human lymphocytes, induces sister chromatid exchanges and gross chromosomal aberration (for review see Pöschl & Seitz 2004). It induces inflammation and metaplasia of tracheal epithelium, delays cell cycle progression and enhances cell injury associated with hyperregeneration (Simanowski et al 1994). AA binds to proteins resulting in structural and functional alterations. AA weakens the antioxidative defence systems by binding to glutathione and thus increases oxidative stress indirectly. AA also injures microtubules and mitochondria. Decreased mitochondrial function results in an inhibition of fatty acid oxidation and ATP formation. Both factors favour the occurrence of fatty liver. In addition, mitochondrial damage induces apoptosis but also survival factors such as NF-κB. It has also been shown that AA interferes with the DNA repair machinery. AA directly inhibits O6-methylguanosyl transferase, an enzyme important for the repair of DNA adducts (Espina et al 1988). In the liver, AA forms adducts with intracellular proteins and DNA resulting in morphological and functional impairment of the cell and a humeral immune reaction towards de novo generated antigens. The binding to DNA and the formation of stable adducts represents one of the mechanisms by which AA could trigger the occurrence of replication errors and/or mutations in oncogenes or tumour suppressor genes. The occurrence of stable DNA adducts has been shown in different organs of alcohol-fed rodents and in leukocytes of alcoholics (Fang & Vaca 1997). It has been shown that the major stable DNA adduct N2-ethyl desoxyguanosine (N2-Et-dG) serves as a substrate of eukaryotic DNA polymerase (Matsuda 1999). Although N2-Et-dG has been shown to be present in DNA samples from white blood cells of human alcoholics and in the liver of rats (Wang et al 2006) that were administered ethanol in the drinking water, there is relatively little evidence that this lesion is mutagenic and its biological significance is unclear. However, this lesion can be detected in human urine samples,

112

SEITZ ADH1C*1/1

ETHANOL

ALDH2*1/2

ACETALDEHYDE

ACETATE Antioxidant defence

Methyl transfer

DNA adducts

DNA repair

IL6

NF-κB

Apoptosis FIG. 1. Effect of acetaldehyde on important factors in carcinogenesis. Acetaldehyde accumulates due to either increased production or decreased detoxification. Acetaldehyde inhibits methyl transfer at various stages (e.g. folate deficiency and reduced activation of methionine to S-adenosyl-methionine) and injures the antioxidative defence system. Acetaldehyde forms DNAadducts and inhibits DNA-repair at the same time. Subsequently acetaldehyde results in an increased NF-κB, which is associated with inhibition of apoptosis and hyper-regeneration (Seitz & Stickel 2006).

suggesting that it may be useful as a biomarker of AA-related DNA damage. More recent data have shown that in the presence of basic amino acids or histones, AA reacts with deoxyguanosine in DNA to form a different DNA adduct, 1,N 2-propano-dG (PdG) (Brooks & Theruvathu 2005). In contrast to N2-Et-dG, PdG has been shown to be a mutagenic DNA lesion in vivo in mammalian cells. These AA-associated effects occurred at AA concentrations from 40–1000 µM ( Theruvathu et al 2005), which are similar to concentrations observed in human saliva following alcohol ingestion. It has been recently shown that the polyamines spermine and spermidine can also facilitate the formation of PdG from AA and dG at AA concentrations as low as 80 µM. The action of AA on DNA metabolism is summarized in Fig. 1. According to the International Agency for Research on Cancer there is sufficient evidence to classify AA as a carcinogen in experimental animals (IARC 1999). Acetaldehyde production in the gastrointestinal tract AA in the gastrointestinal tract can be produced either by mucosal enzymes or by gastrointestinal bacteria. Mucosal alcohol dehydrogenase (ADH) isozymes vary within the gastrointestinal tract (GIT) (Seitz & Oneta 1998). While ADH4 is the major ADH present in the mucosa of the upper GIT (oropharynx, oesophagus and the stomach), AHD1C is predominantly present in the stomach and in the small and large intestine. ADH3 is present in the entire GIT and ADH1B can be found in the muscular layer of the mucosa. The contribution of the various ADH

GI TRACT CANCER

113

isozymes to ethanol metabolism depends on their kinetic properties and on their distribution within the GIT. Gastrointestinal ethanol metabolism is dependent on the ethanol concentration in the alcoholic beverage consumed. Metabolism increases with increasing concentrations of ethanol. ADH1C with a Km of approximately 1 mM is usually saturated and contributes fully to ethanol metabolism. ADH4 with its intermediate Km value of 41 mM at ph 7.4 contributes only partly. ADH3 is insaturable for alcohol and contributes only at rather high alcohol concentrations. Moreover, genetic polymorphism in ADH1C and ethnic differences in the expression of ADH4 have been described and may further modify gastrointestinal ethanol metabolism. It is interesting that a high percentage of Asians lack class IV ADH in the stomach. All these ADHs are influenced by various factors in their activity including genetics, ethnicity, gender, age, concentration of alcohol, drugs and mucosal injury (Seitz & Oneta 1998). The net amount of AA accumulating in mucosal cells depends on the production of AA from ethanol and on its detoxification by acetaldehyde dehydrogenase (ALDH). Thus, the ratio of ADH to ALDH activity is of pathogenetic importance and an imbalance of this ratio has been reported for the distal colon and the rectum. Beside the fact that alcohol is metabolized to AA in the mucosal cells AA is also produced by bacteria (Homann 2001, Salaspuro et al 2006). This bacterial metabolism takes place in the upper GIT (Homann et al 1997), and in the large intestine (Jokelainen et al 1996a). Many microbes representing normal oral flora possess ADH activity with individual kinetic characteristics. Under aerobic or microaerobic conditions in the mouth the ADH reaction is reversed with AA being the end product (Salaspuro et al 1999). After drinking of alcohol, salivary ethanol is metabolised to AA via this reversed reaction and AA concentrations up to 100–150 µM can be detected in the saliva of healthy volunteers (Homann et al 1997). This AA can be reduced significantly by an antiseptic mouthwash underlining the essential role of bacteria in the production of salivary AA. The kinetic characteristics of microbial ADHs may vary to a great extent. Some microbial ADHs are not saturated with ethanol; therefore an increasing salivary ethanol concentration or production of AA is enhanced. Accordingly, at higher blood and salivary ethanol concentrations salivary AA levels are also higher. This may explain the well established epidemiological finding of increased cancer risk associated with heavier and more intoxicating drinking. In addition to that, smoke contains high amounts of AA and smoking also changes the bacterial oral flora towards a flora with a higher capacity to produce AA. Thus, smoking, poor oral hygiene and drinking alcoholic beverages with high concentrations of AA, such as Calvados, are associated with an increased risk for cancer (Salaspuro et al 2006). It has been shown that patients with an oropharyngeal cancer had indeed a higher concentration of AA in their saliva as compared to healthy controls (Jokelainen et al 1996b).

114

SEITZ

AA can also be produced by colonic bacteria (Jokelainen et al 1996a). The amount of AA per gram of mucosal tissue in the colon is significantly higher than in other tissues following ethanol ingestion (up to 200 µM). This is primarily due to the production of AA from ethanol by faecal bacteria, as shown in experiments using germ-free rats (Seitz et al 1990). The toxic mucosal effects of AA result in decreasing cell numbers in the functional compartment of the colonic crypt, being answered secondarily by compensatory hyper-regeneration with increased crypt cell production rates and an extension of the proliferative compartment towards the lumen of the crypt (Simanowski et al 1986, 1994, 2001). This observation was paralleled by a significant increase in rectal mucosal ornithine decarboxylase activity (Seitz et al 1990). The alterations of crypt cell dynamics caused by AA favour the development of colorectal cancer (Seitz et al 1990). As the alcohol-associated hyper-regeneration of the colonic mucosa is especially pronounced with increasing age, chronic alcohol consumption during the lifetime may additionally result in an elevated risk of developing colorectal cancer. In contrast, cell differentiation with regard to the cytokeratin expression pattern was not influenced by chronic alcohol consumption as well as regulatory factors involved in carcinogenesis and/or apoptosis (Simanowski et al 2001). Changes in colonic cell regeneration caused by ethanol and AA have been observed in animal studies and in humans (Simanowski et al 1986, 1994, 2001). Evidence for the carcinogenic role of AA in alcohol-associated cancer Animal experiments It has been shown in rats that the administration of cyanamide, a potent ALDH inhibitor, increases AA concentrations in the blood but also in the colon accelerates chemically induced colorectal carcinogenesis. Under these conditions an earlier occurrence of tumours is observed (Seitz et al 1990). This is associated with an increased proliferation status of the colorectal mucosa which correlates with AA levels (Simanowski et al 1994). Such a hyper-regeneration of the gastrointestinal mucosa is also observed following chronic AA feeding resembling that following chronic ethanol feeding (Homann et al 1997, Simanowski et al 1993). Furthermore, colonic acetaldehyde concentrations correlate with mucosal folate levels in the colon. This is of major importance since folate deficiency is associated with an increased risk for distal colorectal cancer when it coincides with alcohol drinking (Giovannucci et al 1995). Genetic linkage studies Recent and striking evidence for the causal role of AA in ethanol-associated GIT carcinogenesis derives from genetic linkage studies in alcoholics. Individuals who

GI TRACT CANCER

115

accumulate AA because of polymorphism and/or mutation in the genes coding for enzymes responsible for AA generation or detoxification have an increased cancer risk. In Japan as well as in other Asian countries, a high percentage of individuals carry a mutation of ALDH2. Mitochondrial ALDH2 is primarily responsible for AA oxidation. Human ALDH2 is polymorphic, with two distinct alleles: ALDH2 *1 and ALDH2 *2. ALDH2 *2 results from a single-point mutation in chromosome 6. Blood AA levels of ALDH2 *2 homozygous individuals are 6–20 times higher compared to ALDH2 *1 individuals, in whom AA is hardly detectable after alcohol consumption. The elevated AA concentrations cause unpleasant side effects (flush syndrome) that protect these individuals from alcoholism. However, heterozygous individuals may become heavy drinkers or even alcoholics. Yokoyama et al (1998) were the first to report that the heterozygous mutation of the ALDH2 gene (ALDH2 *1,2) is a strong risk factor for oesophageal cancer in everyday drinkers and alcoholics. A comprehensive study of the ALDH2 genotype and cancer prevalence in Japanese alcoholics showed that the frequency of inactive ALDH2 increased remarkably among alcoholics with cancer of the oral cavity, oropharynx, hypopharynx, larynx, oesophagus and colorectum (Yokoyama et al 1998). Many epidemiological studies have uniformly shown that the risk of alcohol related GIT cancers is markedly increased in Asians with the low-activity ALDH2 enzyme. After adjustment for confounders the relative risks to those with the normal enzyme were 11.1 for oropharyngolaryngeal, 12.5 oesophageal, 3.5 stomach, 3.4 colon and 8.2 for lung cancer (Yokoyama et al 1998). In a recent meta-analysis including seven studies and 905 cases carried out in Japan, Taiwan and Thailand, these findings with regard to the risk of oesophageal cancer were confirmed (Lewis & Smith 2005). The review provided additional evidence for the important role of alcohol intake in the risk of oesophageal cancer. Individuals whose genotype results in markedly lower alcohol intake (homozygotic flushers) appear to be protected. However, the most important message of this meta-analysis was that AA may play an important carcinogenic role in the pathogenesis of oesophageal cancer. It is important to note that these individuals also have high AA levels in their saliva, and thus AA is delivered directly to the surface mucosa of the upper GIT in such individuals (Väkeväinen et al 2000). In addition to the mutation of the ALDH2 gene, polymorphisms of ADH1B and ADH1C may also modulate AA levels. Whereas the ADH1B *2 allele encodes for an enzyme that is approximately 40 times more active than the enzyme encoded by the ADH1B *1 allele, ADH1C *1 transcription leads to an ADH isoenzyme 2.5 times more active than that from ADH1C *2. However, the ADH1B *2 allele frequency is high in Asians but low in Caucasians. It protects from alcoholism because of the high amount of AA produced and its toxic side effects. Because of the low ADH1B *2 allele frequency and the lack of ALDH2 mutations in Caucasians, ADH1C polymorphism and its role in alcohol-associated carcinogenesis is ideally investigated in Caucasian populations.

116

SEITZ

Studies on ADH1C polymorphism in Caucasians and GIT cancer have shown contradictory results. Whereas an increased risk of oropharyngeal and laryngeal cancer in individuals with the ADH1C*1 allele has been reported (Harty et al 1997, Coutelle et al 1997), others could not confirm such an association in case-control studies (Olshan et al 2001, Sturgis et al 2001, Schwarz et al 2001, Zavras et al 2002). One reason for this discrepancy is the fact that in all these studies, the percentage of cancer patients with high alcohol intake was rather low, sometimes extremely low. In the study by Sturgis et al (2001) the amount of alcohol ingested was not even reported. Thus, it is not surprising that a pooled analysis of all the studies published so far led to the conclusion that the ALDH1C allele is not a risk factor for alcohol-associated carcinogenesis (Brennan et al 2004). Visapää et al (2004) studied 107 alcoholic patients with high alcohol ingestion and oropharyngeal, laryngeal, hypopharyngeal and oesophageal cancer to compare their ADH1C genotype with 103 age-matched alcoholics with a similar alcohol consumption but without cancer, and he observed a significantly increased cancer risk in individuals with the ADH1C*1 allele. This was found to be associated with significantly elevated AA levels in the saliva of individuals homozygous for ADH1C*1 Increased salivary AA levels in these individuals as in individuals with ineffective ALDH activity may explain their increased cancer risk, because AA comes into direct contact with the mucosa. In this context, it is interesting to note that AA-fed rats showed a severe hyper-regeneration of the upper GIT mucosa (Homann et al 1997); this is very similar to the morphological changes observed after chronic consumption. These changes were only observed when the animals had functionally intact salivary glands. After sialoadenectomy, this proliferation disappeared, which supports the hypothesis that salivary AA is involved in carcinogenesis (Simanowski et al 1993). In this context, it has to be pointed out that chronic alcohol consumption alters salivary morphology and function. In addition, a more recent study in 818 patients with alcohol-associated cancers gives further strong evidence for AA being an important factor in the development of upper GIT and liver cancer (Homann et al 2006). Summary and Conclusion There is overwhelming evidence for the involvement of AA in alcohol-associated carcinogenesis which may be summarized as follows: • High AA levels occur in the saliva and in the colon following alcohol consumption. • Oral bacteria are capable of oxidizing AA from ethanol, and salivary AA levels are elevated in patients with upper GIT cancer as well as in individuals with a poor dental state and cigarette smokers, both conditions also favouring cancer risk.

GI TRACT CANCER

117

• AA leads to mucosal hyper-regeneration in the upper and lower GIT, and the colonic crypt cell production rate significantly correlates with AA levels in the colonic mucosa. • Animal experiments have shown an acceleration of colorectal carcinogenesis induced by the specific locally acting carcinogen acetoxymethyl-methylnitrosamine (AMMN), when cyanamide, an ALDH inhibitor, was applied and AA levels were increased. • Colonic AA levels show a significant inverse correlation with mucosal folate concentrations supporting in vitro data showing a destruction of folate by AA. • Individuals with the inactive form of ALDH2 resulting in elevated AA concentrations exhibit an increased risk for upper and lower GIT cancer due to chronic alcohol consumption. • In individuals homozygous for the ADH1C *1 allele, salivary AA levels are elevated thus increasing the risk of developing upper GIT cancer.

References Brennan P, Lewis S, Hashibe M et al 2004 Pooled analysis of alcohol dehydrogenase genotypes and head and neck cancer. A HuGE review. Am J Epidemiol 15:1–16 Brooks PJ, Theruvathu JA 2005 DNA adducts from acetaldehyde: implications for alcohol-related carcinogenesis. Alcohol 35:187–193 Coutelle C, Ward PJ, Fleury B et al 1997 Laryngeal and oropharyngeal cancer, and alcohol dehydrogenase 3 and glutathione S-transferase M1 polymorphism. Hum Genet 99:319–325 Espina N, Lima V, Lieber CS, Garro AJ 1988 In vitro and in vivo inhibitory effect of ethanol and acetaldehyde on O6-methylguanine transferase. Carcinogenesis 9:761–766 Fang JL, Vaca CE 1997 Detection of DNA adducts of acetaldehyde in peripheral white blood cells of alcohol abusers. Carcinogenesis 18:627–632 Feron VJ, Kruysse A, Woutersen RA 1982 Respiratory tract tumours in hamsters exposed to acetaldehyde vapour alone or simultaneously to benzo(a)pyrene or diethylnitrosamine. Eur J Cancer Clin Oncol 18:13–31 Giovannucci E, Rimm EB, Ascherio A, Stampfer MJ, Colditz GA, Willett WC 1995 Alcohol, low-methionine-low-folate diets and risk of colon cancer in men. J Natl Cancer Inst 87: 265–273 Harty LC, Caporaso NE, Hayes RB et al 1997 Alcohol dehydrogenase 3 genotype and risk of oral cavity and pharyngeal cancers. J Natl Cancer Inst 89:1698–1705 Homann N 2001 Alcohol and upper gastrointestinal tract cancer: the role of local acetaldehyde production. Addict Biol 6:309–323 Homann N, Jousimies-Somer H, Jokelainen K, Heine R, Salaspuro M 1997 High acetaldehyde levels in saliva after ethanol consumption: methodological aspects and pathogenetic implications. Carcinogenesis 18:1739–1743 Homann N, Kärkkäinen P, Koivisto T, Nosova T, Jokelainen K, Salaspuro M 1997 Effects of acetaldehyde on cell regeneration and differentiation of the upper gastrointestinal tract mucosa. J Natl Cancer Inst 89:1692–1697 Homann N, Stickel F, Konig IR et al 2006 Alcohol dehydrogenase 1C*1 allele is a genetic marker for alcohol-associated cancer in heavy drinkers. Int J Cancer 118:1998–2002

118

SEITZ

IARC 1999 Re-evaluation of some organic chemicals, hydrazine and hydrogen peroxide. In: Monographs on the evaluation of the carcinogenic risk of chemicals to humans. Acetaldehyde. vol. 77, p 319–335, International Agency for Research on Cancer, Lyon Jokelainen K, Matysiak-Budnik T, Mäkisalo H, Höckerstedt K, Salaspuro M 1996a High intracolonic acetaldehyde values produced by a bacteriocolonic pathway for ethanol oxidation in piglets. Gut 39:100–104 Jokelainen K, Heikkonen E, Roine R, Lehtonen H, Salaspuro M 1996b Increased acetaldehyde production by mouthwashings from patients with oral cavity, laryngeal or pharyngeal cancer. Alcohol Clin Exp Res 20:1206–1210 Lewis SJ, Smith GD 2005 Alcohol, ALDH2 and esophageal cancer: a meta-analysis which illustrates the potentials and limitations of a Mendelian randomization approach. Cancer Epidemiol Biomarkers Prev 14:1967–1971 Matsuda T, Terashima I, Matsumoto Y, Yabushita H, Matsui S, Shibutani S 1999 Effective utilization of N2-ethyl-2′-deoxyguanosine triphosphate during DNA synthesis catalyzed by mammalian replicative DNA polymerases. Biochemistry 38:929–935 Olshan AF, Weissler MC, Watson MA, Bel DA 2001 Risk of head and neck cancer and the alcohol dehydrogenase 3 genotype. Carcinogenesis 22:57–61 Pöschl G, Seitz HK 2004 Alcohol and cancer. Alcohol Alcohol 39:155–165 Salaspuro M, Salaspuro V, Seitz HK 2006 Interaction of alcohol and tobacco in upper aerodigestive tract and stomach cancer. In: Cho CH, Purohit V (eds) Alcohol, tobacco and cancer. Karger Basel, p 48–62 Salaspuro V, Nyfors S, Heine R, Siitonen A, Salaspuro M, Jousimies-Somer H 1999 Ethanol oxidation and acetaldehyde production in vitro by human intestinal strains of Escherichia coli under aerobic, microaerobic, and anaerobic conditions. Scand J Gastroenterol 34:967–973 Schwartz SM, Doody DR, Fitzgibbons ED, Rick S, Porter PL, Chen C 2001 Oral squamous cell cancer risk in relation to alcohol consumption and alcohol dehydrogenase-3 genotypes. Cancer Epidemiol Biomarkers Prev 10:1137–1144 Seitz HK, Oneta CM 1998 Gastrointestinal alcohol dehydrogenase. Nutr Rev 56:52–60 Seitz HK, Simanowski UA, Garzon FT et al 1990 Possible role of acetaldehyde in ethanol related rectal carcinogenesis in the rat. Gastroenterology 98:1–8 Simanowski UA, Seitz HK, Baier B, Kommerell B, Schmidt-Gayk H, Wright NA 1986 Chronic ethanol consumption selectively stimulates rectal cell proliferation in the rat. Gut 127:278–282 Simanowski UA, Suter P, Stickel F et al 1993 Oesophageal epithelial hyperregeneration following chronic ethanol ingestion: effect of age and salivary gland function. J Natl Cancer Inst 85: 2030–2033 Simanowski UA, Suter P, Russell RM et al 1994 Enhancement of ethanol induced rectal mucosal hyperregeneration with age in F244 rats. Gut 35:1102–1106 Simanowski UA, Homann N, Knuhl M et al 2001 Increased rectal cell proliferation following alcohol abuse. Gut 49:418–422 Sturgis EM, Dahlstrom KR, Guan Y et al 2001 Alcohol dehydrogenase genotype is not associated with risk of squamous cell carcinoma of the oral cavity and pharynx. Cancer Epidemiol Biomarkers Prev 10:273–275 Theravathu JA, Jaruga P, Nath RG, Dizdaroglu M, Brooks PJ 2005 Polyamines stimulate the formation of mutagenic 1,N2-propanodeoxyguanosine adducts from acetaldehyde. Nucleic Acids Res 33:3513–3520 Väkeväinen S, Tillonen J, Agarwal D, Srivastava N, Salaspuro M 2000 High salivary acetaldehyde after a moderate dose of alcohol in ALDH2-deficient subjects: strong evidence for the local carcinogenetic action of acetaldehyde. Alcohol Clin Exp Res 24:873–877 Visapää JP Gotte K, Benesova M et al 2004 Increased cancer risk in heavy drinkers with the alcohol dehydrogenase 3*1-allele possibly due to salivary acetaldehyde. Gut 53:871–876

GI TRACT CANCER

119

Wang M, Yu N, Chen L, Villalta PW, Hochalter JB, Hecht SS 2006 Identification of an acetaldehyde adduct in human liver DNA and quantitation as N2-ethyldeoxyguanosine. Chem Res Toxicol 19:319–324 Woutersen RA, Appelman LM, van Garderen-Hoetmer A, Feron JV 1986 Inhalation toxicity of acetaldehyde in rats. III. Carcinogenicity study. Toxicology 41:213–231 Yokoyama A, Muramatsu T, Ohmori T et al 1998 Alcohol-related cancers and aldehyde dehydrogenase-2 in Japanese alcoholics. Carcinogenesis 19:1383–1387 Zavras AI, Wu T, Laskaris G et al 2002 Interaction between a single nucleotide polymorphism in the alcohol dehydrogenase 3 gene, alcohol consumption and oral cancer risk. Int J Cancer 97:526–530

DISCUSSION Crabb: The mutagen you showed (AMN) looked like an adduct of acetaldehyde with nitrosamine. Does this actually form from acetaldehyde and other nitrosamines? Seitz: This is a nitrosamine which does not need activation, as the normal nitrosamines do. It is a compound that is very stable and acts directly. We don’t need any metabolic activation. Crabb: Could it be created in, say, the upper GIT with acetaldehyde and other nitrosamines? Seitz: Possibly. Emery: What did you mean by saying that acetaldehyde destroys folate? Seitz: The study was done in vitro by Spencer Shaw (Shaw et al 1989). The authors applied different amounts of acetaldehyde to folate, and couldn’t recover the folate as a folate molecule any more. I’m unsure of the chemical mechanisms. They found no effect with 50–100 µM acetaldehyde, but did when they went to 200 µM. Worrall: There are a couple of mechanisms that have been proposed. One is that acetaldehyde opens one of the rings. Another is that there are large modifications on the rings that alter the properties of the folate. Thornalley: If this is of physiological importance you might expect to see increased expression of the reduced folate carrier in these colonic epithelial cells. The reduced folate carrier is also the carrier by which thiamine monophosphate is exported from cells. This could be one of the mechanisms by which thiamine levels become depleted in alcoholism. Seitz: That could well be, but there could be other mechanisms of thiamine depletion. Thornalley: There are many adducts of dicarbonyls, such as glyoxal and methylglyoxal. These increase in oxidative stress. Conceivably in these sorts of situations there are several types of mutagenic adducts present. Seitz: I agree. I didn’t mention adducts created by oxidative stress. We are working on exocyclic etheno –DNA adducts coming from 4-HNE, for example. We measure

120

SEITZ

them. I am not sure which of these adducts are more important. We should measure both types in a certain tissue. Thornalley: There are imidazopurinone adducts which are tricyclic adducts produced from glyoxal and methylglyoxal, which normally exist at levels similar to those of 8-hydroxydeoxyguanosine. These are possibly increased as well as adducts of acetaldehyde. If you do DNA digests with nuclease at high pH you destroy these. You need to do an acidic digest to see them. Rao: The evidence you have provided for the role of acetaldehyde in carcinogenesis is compelling. Is there a way to examine whether the acetaldehyde effect is more a promotion of carcinogenesis, or whether by itself it can induce carcinogenesis? Seitz: That is an important question. Although we see that there are adducts, I am not sure whether these adducts are more important than those coming from oxidative stress. The data we have indicate that it is more of a promoting effect of acetaldehyde than an initiating effect. However, we can’t exclude the latter possibility completely. Rao: Can you also analyse other types of mutations in these patients? Are there oncogenic mutations? Seitz: We haven’t specified certain genes at that time point. Shukla: The information you provided on hypermethylation is very relevant to carcinogenesis. Has anyone looked at the CpG island methylations and p53? Seitz: We have looked at p53 in the colon and haven’t found an effect. Apte: But there are hypermethylations of CpG islands in colon cancer. Seitz: Yes, but that isn’t related to p53. In p53 there is nothing. Shukla: There is some evidence that in relation to alcohol, hypomethylation of DNA can occur. Seitz: I agree. Emery: Are specific genes being hypomethylated? Shukla: It is more global. Apte: In addition to the proliferative effect you see, is there an effect on apoptosis? Is it a double effect with increased proliferation and decreased apoptosis? Seitz: We see an acetaldehyde effect on apoptosis in the liver. We also see an effect on NF-κB in the liver. We haven’t shown this clearly in the GI tract. We have looked at some antiapoptotic proteins in the colon of individuals, such as Bcl-2. There was no effect. We have also looked at cytokeratins and haven’t seen an effect. Apte: Have there been any in vitro studies with colonocytes? Seiz: No. Rao: One of the mechanisms involved in carcinogenesis and tumour invasion is the loss of cadherin based cell–cell contact. Studies in our laboratory have shown that acetaldehyde even at low concentrations (100 µM) can disrupt the cadherin–

GI TRACT CANCER

121

catenin complex by inducing phosphorylation of E-cadherin and β-catenin on tyrosine residues. Deitrich: In relation to the question about the nitrosamines, I’d like to point out that ethanol will react with NO to give ethyl nitrite (Deng et al 2004). I don’t know whether or not this is carcinogenic. It was marketed at the turn of the century as sweet spirits of nitre. It fell out of favour. The concentration in vivo is very low, but it might not need to be high to be carcinogenic. M Salaspuro: Yesterday we were discussing that we should be able to measure acetaldehyde in situ in different tissues. What kind of method did you use for this? Do you think your findings reflect something that is happening in the colonic mucosa, or do you think your results reflect the acetaldehyde concentration inside the colon, and then part of that is transported to the mucosal cells? We tried to infuse high acetaldehyde concentrations intracolonically. We could not detect anything in the portal blood. This means that the colon mucosa must be able to metabolize acetaldehyde efficiently. Seitz: I don’t think the enzymic machinery in the colon mucosa is sufficient to produce that high amount of acetaldehyde. We have done bacterial studies and believe that colonic bacteria are largely responsible for the acetaldehyde production. If we look at ADHs in individuals, these are usually the class I. We studied polyps in patients, and found sigma ADH (ADH7) in the polyps. We never found this in the normal mucosa of the colon, but we found it in the polyps. The idea was that polyps as a precancerous lesion probably need retinoic acid more for differentiation and to reduce proliferation than anywhere else. The idea is that there may have been a re-expression of ADH7 to produce retinoid acid which is lacking. This is another observation that needs following up. So to come back to your question, I think it is mostly bacteria producing acetaldehyde, which is then bound to the mucosa when we measured it in the mucosa. The method we are using is the one developed by Timothy Peters’ group, which is an HPLC method measuring adduct formation of acetaldehyde by fluorescence detection. We have compared this method with gas chromatography and they both produce equivalent results. Emery: So there is evidence that acetaldehyde can be transported into the colonic mucosa, but probably only if it is produced by the bacteria that are intimately associated with the mucosa. Yin: In your genetic study, you show that ADH1C polymorphism is associated with colon, liver and oesophageal cancer. Have you looked at the phenotype? We need to consider the enzyme kinetic parameters, the enzyme amount and the overall ADH family expression pattern in target tissue. In colon mucosa ADH1C is the major form but both the Vmax difference between ADH1C1 and ADH1C2 and the total enzyme amount don’t seem that high. In the upper digestive tract ADH1C seems very minor. In the liver it is only one of the ADH isozymes involved and its total contribution is not that high.

122

SEITZ

Seitz: For the colon it isn’t clear. We did some phenotyping studies and found ADH1C present, probably not as high as in other tissues, but it is still there. These results are preliminary, and I’m still not sure of the mechanism. For the liver the only explanation I have is that in the Japanese studies there is a mixture of other risk factors such as hepatitis C in the control patients. Ren: Is there any evidence linking ALDH2 polymorphisms to homocysteine? Seitz: There is evidence that alcohol leads to homocysteinaemia. This is well known. But I am not aware that ALDH2 heterozygotes have a change in their homocysteine levels. Okamura: Two years ago, to my knowledge, there was no study looking at the relationship between ALDH2 genotypes and homocysteine. The relationship between alcohol drinking per se and homocysteine is controversial. Related to this, I have a question. As you mentioned, folate is a very important protective factor for acetaldehyde-related carcinogenesis. Vitamin B12 is also an important determinant for plasma homocysteine levels. How do we know about the relationship between vitamin B12 or folate and carcinogenesis related to alcohol drinking or acetaldehyde? In a study published a few years ago on a Japanese population (Moriyama et al 2002), we have examined the determinants of serum plasma homocysteine. Vitamin B12 is a strong determinant of plasma homocysteine in Japanese populations, especially when there is a genetic deficiency of MTHFR. Seitz: To my knowledge there are data on B12 effects on the relationship between alcohol and carcinogenesis. There are data from Dr Mason in Boston showing that B12 may modulate colon carcinogenesis. But I am not aware of any such data with alcohol. There are data on an increased risk of colon cancer in vitamin B6 deficiency (Larsson et al 2005). We know that acetaldehyde can lead to B6 destruction (Lumeng & Li 1974). Apte: I’d like to return to the colonic bacteria story. Earlier we heard how lactobacilli produce acetaldehyde and they are part of the normal colonic flora. There is a background level of acetaldehyde production all the time in the colon. Do you think the increase in acetaldehyde from this baseline is enough to cause the toxicity that we are talking about? Seitz: I am not sure. Yesterday we heard about the level of toxicity or carcinogenicity, which is between 50 and 100 µM. The level we find is far above this, while the background level is below this. If you drink regularly for a long time this could play an important role. Emery: Are there other factors in food that would stimulate bacterial production of acetaldehyde in the colon? Apte: People talk about not eating enough vegetables and fresh fruit as opposed to eating a lot of red meat. Is this connected to fermentation? Seitz: This is an important question. This is what we are thinking: the diet may modulate bacterial flora, and in turn this may modulate acetaldehyde production. If

GI TRACT CANCER

123

bacteria are modulated by an antibiotic such as metronidazole, this can kill anaerobics and increase aerobics in the colon. This would result in increased acetaldehyde production. This is probably one of the reasons that metronidazole plus alcohol increases acetaldehyde levels. Probiotics such as lactulose change the pH of the colon. If pH is decreased then acetaldehyde production is reduced. Eriksson: It seems that since the odds ratio of colon cancer is not so dramatic with alcohol, there is an effect but it doesn’t compare with the oesophageal cancer rates in Japanese populations with ALDH mutations. There may be a risk factor to start with so the increase is not that big regarding colon cancer in Europeans. If you differentiate different control populations you might be able to find this out. It means that acetaldehyde is still one of the main players, but its role isn’t so obvious because of the existing endogenous baseline production. This would also be interesting to study in the Japanese population. Controls there should have endogenously quite high acetaldehyde levels. On the other hand, their diet is a little different from European diets. Rao: The GI tract has evolved mucosal defence mechanisms to protect the mucosal tissue from a variety of injurious insults. There needs to be a balance between the protective factors and the injurious factors. Weak mucosal protection may increase the risk for acetaldehyde-mediated tissue injury. Worrall: Is there an increase of gut motility in these people? Seitz: We have not measured it. But alcohol per se is changing motility significantly in the whole GI tract. There is hypermotility in the small intestine and also quite a disturbed motility in the colon. Motility is affected by alcohol. Emery: Is that independent of other aspects of their diet? Seitz: We know that a lot of dietary factors influence motility, so probably not. Worrall: Presumably that low motility in the large intestine is giving the bacteria more chance to metabolize things? Seitz: Alcohol is actually having the opposite effect. One cause of alcoholassociated diarrhoea is hypermotility. Preedy: Although we have not measured motility disturbances in the rat, rats fed on an alcohol-containing diet using the Lieber-DeCarli regimen show reduced smooth muscle loss from the small intestine, but not the rectum or colon. This may contribute to your overall motility disturbances. The smooth muscle part of the gut is very sensitive to alcohol in terms of protein synthesis. Albano: You showed nice data on the effects of ADH polymorphisms on the risk of upper digestive tract cancer. However, I am a little confused. If a considerable amount of acetaldehyde is coming from the bacterial fermentation of alcohol, how can we see this strong influence of genetic background? This suggests that the ADH polymorphism is probably influencing the metabolism of other aldehydes. Seitz: That goes in the same direction as Dr Yin proposed. The genetic modification causes acetaldehyde to be produced within the cell (the colonocyte). The rest

124

SEITZ

happens in the colonic lumen outside the colonocyte by bacteria. It may take several steps for the acetaldehyde to enter the cell and then the nucleus. Even if a low amount of acetaldehyde is produced in the colonocyte by genetic modification, this may be relevant because it is closer to the target. Apte: The retinol story is interesting. In the liver we know there is a small window for vitamin A reinforcement in alcoholics because it can cause toxicity. Since retinol is a precursor for retinoic acid, is there any place for retinol replacement? Seitz: The story has now been completed. If retinol or retinoic acid is given to alcohol-fed animals, both compounds are metabolized by CYP2E1, which is induced by alcohol. As a result of this increased metabolism apoptotic metabolites of retinol and retinoic acid appear in the liver. This is the major mechanism by which the combined application of alcohol, β-carotene, retinol or retinoic acid create toxicity. The therapeutic window for β-carotene is very small and replacement therapy of β-carotene, retinol or retinoic acid would be the same. We don’t recommend it. References Deng XS, Bludeau P, Deitrich RA 2004 Formation of ethyl nitrite in vivo after ethanol administration. Alcohol 34:217–223 Larsson SC, Giovannucci E, Wolk A 2005 Vitamin B6 intake, alcohol consumption, and colorectal cancer: a longitudinal population-based cohort of women. Gastroenterology 128:1830– 1837 Lumeng L, Li TK 1974 Vitamin B6 metabolism in chronic alcohol abuse. Pyridoxal phosphate levels in plasma and the effects of acetaldehyde on pyridoxal phosphate synthesis and degradation in human erythrocytes. J Clin Invest 53:693–704 Moriyama Y, Okamura T, Kajinami K et al 2002 Effects of serum B vitamins on elevated plasma homocysteine levels associated with the mutation of methylenetetrahydrofolate reductase gene in Japanese. Atherosclerosis 164:321–328 Shaw S, Jayatilleke E, Herbert V, Colman N 1989 Cleavage of folates during ethanol metabolism. Role of acetaldehyde/xanthine oxidase-generated superoxide. Biochem J 257:277–280

The determination of acetaldehyde in exhaled breath Robert Tardif Department of Occupational & Environmental Health, Faculty of Medicine, University of Montreal, 2375 Cote St-Catherine, Montreal, Québec, Canada H3T 1A8

Abstract. Breath acetaldehyde has been used to investigate the production of acetaldehyde after ethanol ingestion in ALDH2-deficient individuals, to compare ethanol and acetaldehyde metabolism, to study the toxicological outcome of metabolic inhibitors of ALDH2, and as a biomarker of exposure to ethanol vapours. A number of approaches have been developed to collect representative breath samples (mixed air or alveolar air) and to measure breath acetaldehyde. For instance, the highest breath concentration of acetaldehyde (∼50 nmoles/l) measured during pulmonary ethanol exposure (1000 ppm, 6 hours) is of the same magnitude as those measured after ingestion of 0.4–0.8 g/kg (∼60–80 nmoles/ l), whereas endogenous levels rarely exceed 1 nmole/l. The interpretation of breath acetaldehyde is compounded by several factors; smoking, ALDH2 polymorphism and alcohol drinking habits are associated with higher breath/blood levels. Some authors have considered that breath acetaldehyde, particularly low levels, cannot be used to estimate blood acetaldehyde. Despite the problems associated with its determination, breath acetaldehyde could be an interesting tool for estimating ethanol or acetaldehyde exposure. However, some additional research efforts will be necessary in order to standardize the technique used for breath sampling and to control the influence of the factors that are known to affect breath acetaldehyde determination. 2007 Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Foundation Symposium 285) p 125–136

The determination of volatile organic chemicals (VOCs), such as acetaldehyde, in exhaled breath is useful for several applications in toxicology, epidemiology and pathology. Among other things, breath testing, which involves a safe and noninvasive approach, is being used to assess environmental or occupational exposure to chemicals, in several metabolic studies, and in the diagnosis of several pathologies or diseases. Indeed, the concentration of a given chemical in a breath sample, such as alveolar air (end-expired air), reflects its blood concentration and, as such, the internal dose. Hence, the determination of VOCs in breath samples represents a useful approach for estimating the internal exposure to toxic chemicals and the health risk as well. 125

126

TARDIF

Acetaldehyde, the first metabolite of ethanol, is a very reactive chemical and has been shown to be mutagenic and carcinogenic in animals. It has been reported that acetaldehyde is produced by the lung cancer cell lines SK-MES and CALU-1 under in vitro conditions (Smith et al 2003). Acute toxic effects (e.g. irritation, bronchoconstriction, flushing syndrome) have been associated with the ingestion of ethanol or with the inhalation of ethanol or acetaldehyde vapours. Hald and Jacobsen were the first, in 1948, to isolate acetaldehyde chemically from the expired air of individuals previously treated with disulfiram (Antabuse)—a known metabolic inhibitor of acetaldehyde—and ethanol (Hald & Jacobsen 1948). A number of epidemiological and experimental studies have provided convincing evidence of the role of acetaldehyde as a significant risk factor for various diseases. These observations resulted in an increasing interest, particularly during the 1980s, in the determination of acetaldehyde in biological fluids (e.g. blood and plasma) for the assessment of acetaldehyde exposure resulting from local or systemic ethanol metabolism. However, several reports pointed out the difficulties affecting acetaldehyde measurement in blood/plasma, particularly due to the low levels usually encountered. Acetaldehyde, like other volatiles, crosses the alveolar–capillary membrane of the lungs and is in equilibrium with pulmonary blood. Therefore, in order to overcome the problems associated with blood/plasma analysis, some researchers considered the possibility of measuring acetaldehyde in breath samples as a surrogate to blood/ plasma acetaldehyde (Freund & O’Halloren 1965, Stowell et al 1980). Several aspects of acetaldehyde determination in breath have been previously reviewed by Jones (1995). The objective of this paper is to provide a historical perspective of the developments regarding acetaldehyde analysis in exhaled breath, with special emphasis on: (1) the various approaches that have been used to collect breath samples, (2) the analytical techniques applied for the detection/quantification of acetaldehyde in breath samples, and (3) the main factors that are known to influence the production/elimination kinetics of acetaldehyde and the levels of acetaldehyde in exhaled breath. Utility of breath acetaldehyde determination Breath acetaldehyde has been primarily used in studies aimed at characterizing the production of acetaldehyde after ethanol ingestion in normal individuals or in individuals with atypical forms of alcohol dehydrogenase (ADH2) and/or aldehyde dehydrogenase (ALDH2) (Freund & O’Hollaren 1965, Stowell et al 1980, Couchman & Crow 1980, Jones et al 1984, Stowell et al 1984, Sarkola et al 2002). Similarly, other researchers have used breath acetaldehyde to compare ethanol and acetaldehyde metabolism or to investigate the toxicological/toxicokinetic

ACETALDEHYDE LEVELS IN BREATH

127

outcome resulting from the administration of metabolic inhibitors of ALDH2 (e.g. disulfiram, calcium carbamide) (Miozi et al 1980, Hald & Jacobsen 1948, Stowell et al 1984, Sarkola et al 2002). In some of these studies, breath acetaldehyde or ethanol was used to indirectly estimate their equivalent blood concentration (Stowell et al 1980, Jones et al 1984). Recently, acetaldehyde has been measured, and compared to ethanol levels, in breath samples obtained from volunteers exposed by inhalation to low levels (25–1000 ppm) of ethanol vapours (Tardif et al 2004). Analytical approaches Breath air sample The type of breath sample, either mixed (dead space) or end-exhaled air (alveolar), used for acetaldehyde determination differs from one study to another (see Tables 1–3). Moreover, from the authors’ description, it was sometimes impossible to clearly identify which type was used. Although it still remains difficult to assess the impact of breath sample type, it is reasonable to consider that it could influence the levels of acetaldehyde measured, depending on the dose of ethanol and the time of breath sampling. For instance, Pikkarainen et al (1981) measured acetaldehyde levels in dead space and alveolar air in volunteers who were administered ethanol (0.8 g/kg); compared to dead space levels, the levels measured in alveolar air (60–120 min post administration) showed a downward trend (−15% to −42%), although the differences were not statistically significant. Analysis of acetaldehyde Since the first study that reported acetaldehyde in breath, the analytical methodology used to collect breath samples and to measure breath acetaldehyde has considerably improved and has contributed to an increase in the sensitivity/accuracy of this analysis. In 1948, Hald and Jacobson collected breath samples through a valve connected to a U-shaped glass tube filled with an absorber, and measured acetaldehyde by spectrophotometry after a colorimetric reaction with 2,4dinitrophenylhydrazine (Hald & Jacobsen 1948). Later, acetaldehyde was analysed by injecting breath air directly into a gas chromatograph (Freund & O’Hollaren 1965). Other authors included a prior concentrating step, using either a cold trap (Dannecker et al 1981, Shaskan & Dolinsky 1985) or drawing a known volume of breath air through a glass tube containing a derivatizing agent such as 2hydroxymethylpiperidine (Tardif et al 2004). Similarly, Lin et al (1995) measured acetaldehyde by high performance liquid chromatography after reaction with 2,4-dinitrophenylhydrazine. More recently, some researchers took advantage of the high specificity and sensitivity provided by selected-ion flow-tube mass

128

TARDIF

TABLE 1 Highest breath concentrations of acetaldehyde following ethanol administration, from studies directly reporting acetaldehyde concentrations in breath Ethanol dose

Subjects

n

Breath samples

Acetaldehyde (nmoles/l)

0.87 g/kg

Healthy men

6

Alveolar

147–227

38–70 g

Healthy men

6

Alveolar

6–27

0.4 g/kg 0.4 g/kg

Healthy men Healthy men (smokers) Healthy men (non-smokers) Healthy men Healthy men Healthy women (−OCc) Healthy women (+OC) Healthy men (n = 3) and women (n = 2)

29 11

Mixed air Alveolar

0.3 g/kg 0.5 g/kg

1000 ppm × 6h (∼0.2 g/ kg)d

Alveolar Alveolar

14 –36a 28 ± 18b 22 ± 15

Wong et al (1992) Sarkola et al (2002)

23 ± 12

12 5

Freund & O’Hallaren (1965) Couchman & Crow (1980) Mizoi et al (1980) Jauhonen et al (1982)

23 ± 3

12 5 13 10

58 ± 35 (SD) 52 ± 8 (SEM)

Reference

Alveolar

48 ± 10.8 (SD)e Tardif et al (2004)

a Values reported in Wong et al’s papers are in µmoles/l, which is impossible considering the dose administered.

b

Measured 75 minutes after ethanol administration. Women using (+OC) or not using (−OC) oral contraceptives. d Estimated considering 100% pulmonary absorption. e Concentration measured at 4 hours after the start of exposure. c

spectrometry (SIFT-MS) (Diskin et al 2003, Turner et al 2006). Lately, Mitsubayashi et al (2005) developed bioelectronic sniffers (gas sensors) incorporating enzymes such as ALDH or alcohol oxidase for the measurement of acetaldehyde or ethanol, respectively, in expired air. Levels of acetaldehyde in breath Almost all data on breath acetaldehyde are from studies where volunteers ingested various doses of ethanol. The concentrations obtained in some studies are reported in Tables 1 and 2. Some of them are directly reporting values of acetaldehyde (Table 1). Other values, however, have been indirectly estimated from the blood levels provided in the papers, since these blood levels were calculated from breath level values measured in volunteers; these blood levels have been divided by the

ACETALDEHYDE LEVELS IN BREATH

129

TABLE 2 Highest breath concentrations of acetaldehyde following ethanol administration, from studies reporting blood acetaldehyde estimated from breath concentrationsa Ethanol dose 1.2 g/kg 0.4 g/kg 0.8 g/kg

0.25 g/kg 0.25 g/kg

Subjects

n

Healthy men Healthy men Healthy men (smokers) Healthy men (non-smokers) Healthy men Healthy men

5 5 3

Breath samples

Acetaldehyde (nmoles/l)

Mixed Mixed Not specified

32–84 32–159 70 ± 11 (SEM)

Alveolar Alveolar

18 ± 3 (SEM) 9–34

3 10 10

31 ± 5

Reference Stowell et al (1980) Stowell et al (1984) Pikkarainen et al (1981)b Jones et al (1987) Jones et al (1988)

a

Breath concentrations were estimated indirectly from reported blood levels as follows: BreathACETALDEHYDE = BloodACETALDEHYDE ∏ Acetaldehyde blood:air partition coefficient (Pb : air = 189) (Stowell et al 1980; Jones 1995). b Values reported in Jones 1995.

TABLE 3 Endogenous breath concentrations of acetaldehyde reported in humans Subjects

n

Breath samples

Acetaldehyde (nmoles/l)

Reference

Healthy men (n = 9) and women (n = 5) NA-NS NA-S A-NS A-S Healthy men (n = 3) and women (n = 2) Healthy men (n = 3) and women (n = 2) Healthy men (n = 19) and women (n = 11)

14

Mixed air

0.016–0.25

Dannecker et al (1981)

14 12 16 15 5

Not specified

0.10 ± 0.02 (SEM) 0.37 ± 0.06 0.33 ± 0.06 0.57 ± 0.05 0.045–0.11

Shaskan & Dolinsky (1985)

Not specified

5

Alveolar

30

Mixed air

0.57 ± 0.16 (SD) (0.48–0.81) 0.55 ± 0.39 (SD)

Diskin et al (2003) Tardif et al (2004) Turner et al (2006)

A, alcoholic; NA, non-alcoholic; S, smoker; NS, non-smoker.

value of acetaldehyde blood:air partition coefficient (Table 2) ( Jones 1995). Several of these studies reported breath values measured in normal individuals who received acetaldehyde metabolic inhibitors of ALDH2 prior to ethanol administration or in individuals carrying a deficient form of this enzyme (flushing individuals); both situations result in particularly high breath acetaldehyde levels as a result of reduced acetaldehyde metabolism. These values, however, are not presented in this paper.

130

TARDIF

There is only one study that reported breath acetaldehyde concentrations after inhalation of ethanol vapours (Tardif et al 2004). Interestingly, this study showed that the highest concentration of acetaldehyde (∼50 nmoles/l) measured during an exposure to 1000 ppm of ethanol for 6 hours is of the same magnitude as those measured after ingestion of 0.4–0.8 g/kg of ethanol (∼60–80 nmoles/l) (Table 1). Nonetheless, few papers report endogenous levels of breath acetaldehyde measured in humans. In general, the levels reported are below 1 nmole/l and are strongly influenced by the individuals’ smoking habits and alcohol use (Shaskan & Dolinsky 1985) (Table 3). Factors modifying acetaldehyde in breath air Despite numerous studies on ethanol metabolism, we still do not have a clear idea of the extent to which breath acetaldehyde is representative of blood acetaldehyde and about how to reduce/avoid the influence of some factors that are known to affect acetaldehyde analysis in breath. Indeed, different sources besides acetaldehyde in blood contribute to acetaldehyde in breath. Pïkkarainen et al (1980, 1981) showed that acetaldehyde originates, in part, from lung microsomal enzymes (likely CYP2E1 known to be enhanced by chronic alcohol consumption) and from microbial ethanol oxidation in the oropharynx. This was confirmed by Homann et al (1997) who showed that the local production of acetaldehyde by oral microflora can be reduced by the use of antiseptic mouthwash. Lately, Turner et al (2006) reported that they have occasionally observed large concentrations of ethanol in breath that they attributed to the consumption of sweet drinks prior to breath sampling. They showed that whereas breath ethanol was enhanced by washing the mouth with sugary drinks, there was no apparent enhancement in the production of acetaldehyde within the mouth. However, this raises the question as to what extent the consumption of sugary food or drinks by individuals with poor oral hygiene might enhance the production of acetaldehyde in the mouth. Jauhonen et al (1982), in a study aimed at evaluating the suitability of breath acetaldehyde for estimating blood acetaldehyde, found that most of the acetaldehyde in the end-expired air of non-Oriental subjects (e.g. Caucasian), who produced low concentrations of acetaldehyde in blood after alcohol ingestion, was produced in the respiratory tract and did not correlate with blood levels. In such situations, calculations of blood acetaldehyde based on breath analysis result in overestimation of the actual concentration. They concluded that breath acetaldehyde is not a good indicator of blood levels when these levels are below 30–50 µM, which are very high concentrations compared to those generally reached after drinking alcohol (Tables 1–2). This was confirmed in a subsequent report by Stowell et al (1984),

ACETALDEHYDE LEVELS IN BREATH

131

who concluded that blood acetaldehyde ‘. . . cannot under all circumstances be predicted from breath analysis with the same accuracy that blood ethanol concentration can.’ However, Jauhonen et al (1982) reported that the production of acetaldehyde in the respiratory tract, following alcohol ingestion, was clearly exaggerated in Japanese flushers (ALDH2 deficient), and by long-term cigarette smoking, which is also known to induce CYP2E1. Interestingly, Shaskan & Dolinsky (1985) reported that the endogenous levels of acetaldehyde in breath is higher in smokers, and that being an alcoholic smoker results in even higher levels (Table 3). McLaughlin et al (1990) also reported this positive contribution of cigarette smoking. Unfortunately, however, there are no data available for evaluating the impact of ALDH2 deficiency on endogenous levels of acetaldehyde in breath.

Potential impact of ethanol-blended gasoline on breath acetaldehyde The anticipated escalating use of ethanol-blend gasoline in certain countries will result in increasing the ethanol and acetaldehyde exposure of the general population. As a result, there is a demand by regulatory agencies for health-based evidence to assess the potential health risk associated with inhalation of ethanol vapours. Exposure to 25 ppm of ethanol vapours (for 6 hours), which is lower than the maximum ambient air level (46 ppm) reported to occur upon refuelling a vehicle, produced measurable levels of acetaldehyde in breath air (Tardif et al 2004). Accordingly, there is a need to understand better the potential impact of such long-term exposure on endogenous breath acetaldehyde levels and to assess the health risks associated with such exposure, particularly in individuals with ALDH2 deficiency or pulmonary diseases (e.g. asthma).

Conclusion The assessment of the health risk associated with exposure to ethanol or acetaldehyde in the general population or in more susceptible individuals requires a proper estimate of the exposure. Despite the problems associated with its determination, breath acetaldehyde remains an interesting tool for estimating ethanol or acetaldehyde exposure following ingestion or inhalation, with good toxicological relevance, since acetaldehyde is more directly related to toxic effects. In this case, however, some additional research efforts will be necessary in order, among other things, to standardize the technique used for breath sampling and to control/assess the influence of the several factors that are known to affect breath acetaldehyde determination and interpretation.

132

TARDIF

References Couchman KG, Crow KE 1980 Breath acetaldehyde levels after ethanol consumption. Adv Exp Med Biol 132:451–457 Dannecker JR Jr, Shaskan EG, Phillips M 1981 A new highly sensitive assay for breath acetaldehyde: detection of endogenous levels in humans. Anal Biochem 114:1–7 Diskin AM, Spanel P, Smith D 2003 Time variation of ammonia, acetone, isoprene and ethanol in breath: a quantitative SIFT-MS study over 30 days. Physiol Meas 24:107–119 Freund G, O’Hollaren P 1965 Acetaldehyde concentrations in alveolar air following a standard dose of ethanol in man. J Lipid Res 6:471–477 Hald J, Jacobsen E 1948 The formation of acetaldehyde in the organism after ingestion of Antabuse (Tetraethylthiuramdisulphide) and alcohol. Acta Pharmacol Toxicol 4:305–310 Homann N, Jousimies-Somer H, Jokelainen K, Heine R, Salaspuro M 1997 High acetaldehyde levels in saliva after ethanol consumption: methodological aspects and pathogenetic implications. Carcinogenesis 18:1739–1743 Jauhonen P, Baraona E, Miyakawa H, Lieber CS 1982 Origin of breath acetaldehyde during ethanol oxidation. Effect of long-term cigarette smoking. J Lab Clin Med 100:908–916 Jones AW, Skagerberg S, Borg S, Anggard E 1984 Time course of breath acetaldehyde concentrations during intravenous infusions of ethanol in healthy men. Drug Alcohol Depend 14:113–119 Jones AW, Neiman J, Hillbom M 1987 Elimination kinetics of ethanol and acetaldehyde in healthy men during the calcium carbimide-alcohol flush reaction. Alcohol Alcohol Suppl 1:213–217 Jones AW, Neiman J, Hillbom M 1988 Concentration-time profiles of ethanol and acetaldehyde in human volunteers treated with the alcohol-sensitizing drug, calcium carbimide. Br J Clin Pharmacol 25:213–221 Jones AW 1995 Measuring and reporting the concentration of acetaldehyde in human breath. Alcohol Alcohol 30:271–285 Lin Y, Dueker SR, Jones AD, Ebeler SE, Clifford AJ 1995 Protocol for collection and HPLC analysis of volatile carbonyl compounds in breath. Clin Chem 41:1028–1032 McLaughlin SD, Scott BK, Peterson CM 1990 The effect of cigarette smoking on breath and whole blood-associated acetaldehyde. Alcohol 7:285–287 Mitsubayashi K, Matsunaga H, Nishio G, Toda S, Nakanishi Y 2005 Bioelectronic sniffers for ethanol and acetaldehyde in breath air after drinking. Biosens Bioelectron. 20:1573–1579 Mizoi Y, Hishida S, Ijiri I et al 1980 Individual differences in blood and breath acetaldehyde levels and urinary excretion of catecholamines after alcohol intake. Alcohol Clin Exp Res 4:354–360 Pikkarainen P, Baraona E, Seitz H, Lieber CS 1980 Breath acetaldehyde: evidence of acetaldehyde production by oropharynx microflora and by lung microsomes. Adv Exp Med Biol 132:469–474 Pikkarainen PH, Baraona E, Jauhonen P, Seitz HK, Lieber CS 1981 Contribution of oropharynx microflora and of lung microsomes to acetaldehyde in expired air after alcohol ingestion. J Lab Clin Med 97:631–636 Sarkola T, Iles MR, Kohlenberg-Mueller K, Eriksson CJ 2002 Ethanol, acetaldehyde, acetate, and lactate levels after alcohol intake in white men and women: effect of 4-methylpyrazole. Alcohol Clin Exp Res 26:239–245 Shaskan EG, Dolinsky ZS 1985 Elevated endogenous breath acetaldehyde levels among abusers of alcohol and cigarettes. Prog Neuropsychopharmacol Biol Psychiatry 9:267–272 Smith D, Wang T, Sule-Suso J, Spanel P, Haj AE 2003 Quantification of acetaldehyde released by lung cancer cells in vitro using selected ion flow tube mass spectrometry. Rapid Commun Mass Spectrom 17:845–850

ACETALDEHYDE LEVELS IN BREATH

133

Stowell AR, Lindros KO, Salaspuro MP 1980 Breath and blood acetaldehyde concentrations and their correlation during normal and calcium carbamide-modified ethanol oxidation in man. Biochem Pharmacol 29:783–787 Stowell A, Johnsen J, Aune H, Vatne K, Ripel A, Morland J 1984 A reinvestigation of the usefulness of breath analysis in the determination of blood acetaldehyde concentrations. Alcohol Clin Exp Res 8:442–447 Tardif R, Liu L, Raizenne M 2004 Exhaled ethanol and acetaldehyde in human subjects exposed to low levels of ethanol. Inhal Toxicol 16:203–207 Turner C, Spanel P, Smith D 2006 A longitudinal study of ethanol and acetaldehyde in the exhaled breath of healthy volunteers using selected-ion flow-tube mass spectrometry. Rapid Commun Mass Spectrom 20:61–68. Wong MK, Scott BK, Peterson CM 1992 Breath acetaldehyde following ethanol consumption. Alcohol 9:189–192

DISCUSSION1 Worrall: When you talk about blood acetaldehyde levels, what are you referring to? There must be a complex set of equilibria between the plasma and red cells, and protein bound and non-bound acetaldehyde. Eriksson: Yes. I’ll say more about this in my paper. Preedy: This difference between bound and non-bound acetaldehyde is quite important. This ‘storage’ that you alluded to is interesting. I am aware that some studies have examined acetaldehyde dosing and noticed that long after the dosing there is acetaldehyde within the tissue (Heap et al 1995). This tissue was the brain. After the dosing period, the acetaldehyde was measurable in the brain but not in the liver or blood. This could be due to the fact that it was being released from another tissue. Eriksson: This is an important issue. I am speaking about the ‘free’ or ‘loosely bound’ acetaldehyde. Because acetaldehyde is so reactive, there is an enormous amount of different kinds of binding. Some are in equilibrium with the partition and some binding is released with a little bit of heating. It is certainly bound in such a way that it can accumulate. It could be interesting to take different types of tissue to see with different methods how the acetaldehyde could be released. You have gone another way, trying to measure the adducts themselves directly. One could also try to release the acetaldehyde itself. Preedy: In the measurement of breath acetaldehyde, do you have a similar pre-oral hygiene regime? Do subjects have to wash out their mouths? Eriksson: No, but if you look at Mikko Salaspuro’s data the washout affects the saliva acetaldehyde. We have done studies where we take breath measurements and saliva measurements at the same time. The early part of the breath sample could 1

Dr Tardif was unable to attend the meeting, and in his absence his paper was presented by Dr Eriksson, who also fielded the questions in this discussion session.

134

TARDIF

have more resemblance to the salivary acetaldehyde and the last part of the alveolar air could match more closely the systemic acetaldehyde. No one has done this very systematically. Seitz: Has anyone taken the saliva and incubated it with headspace chromatography at 37 °C, then heated it up and measured again? In other words, how much acetaldehyde is released from the saliva at body temperature? Eriksson: Not exactly, but there is a lot of acetaldehyde formation at 37 °C which complicates the exact measurement of in vivo saliva acetaldehyde levels. Since the saliva sampling takes some time the detected acetaldehyde levels may overestimate the in vivo levels. Seitz: What percentage of saliva acetaldehyde could be released into the air? Eriksson: It seems that all free or quite loosely bound acetaldehyde can be released from saliva. V Salaspuro: There are salivary proteins that acetaldehyde might bind to. It isn’t clear. M Salaspuro: The biggest problem is the continuous microbial production of acetaldehyde. Evaporation occurs at about 18 degrees, which is a severe confounding factor. Has anyone measured breath acetaldehyde? We have many patients who are intubated, and in these we could easily measure breath acetaldehyde without bacterial contamination. Eriksson: I don’t think this has been reported. The breath acetaldehyde itself is partly of microbiological origin. After it is taken, nothing further happens. The evaporation problem isn’t as big as you propose. The boiling point is very low and initially the saliva is warmer than this. But the acetaldehyde doesn’t evaporate so dramatically. In five minutes you can have about 10% loss. This implies that in the saliva part of the acetaldehyde is in the form of Schiff bases, and in such a form that real partition is not so freely available. Worrall: You suggested trying to drive off acetaldehyde from reversible adducts. We are working on some mass spectrometry, and it doesn’t seem to work: what seems to happen is that some of the reversible adducts are converted into irreversible adducts—we actually change the adduct chemistry. Eriksson: How have you done this? Worrall: So far we have just been trying a modest temperature rise, going from say 30 to 50 °C. Eriksson: Try going a little bit higher. You get as much acetaldehyde as you want! The problem is you get it from other sources, through other chemical reactions. Worrall: That’s why we weren’t trying to go so high. Eriksson: What acetaldehyde you get depends on the binding and the chemistry. A few years ago we published this heating process and what it could achieve (Fukunaga et al 1993). The result depends very much on the temperature level.

ACETALDEHYDE LEVELS IN BREATH

135

Apte: Once you have the sample what is the stability of acetaldehyde? Eriksson: That’s a good question. We chill everything, and we always try to measure acetaldehyde the same day we do the study. Acetaldehyde can disappear very easily. But it can be preserved at cold temperatures and in the right sort of container. Apte: Do you see any clinical role for this, given all the confounding factors and difficulties with measurement? Eriksson: Yes. First of all it is non-invasive. You can blow it with a tube directly into the machine. It simplifies measurements. It also measures something occurring inside the body. Rao: There is increasing concern over passive smoking. Does this increase acetaldehyde levels? Eriksson: The acetaldehyde content in the air will not be all that high. This needs to be investigated. Morris: In a smoky car, the airborne acetaldehyde concentration gets to be about 40 nM/litre (1 part per million). This is as high as it can get. In a smoky bar it is usually 100-fold lower. Eriksson: That’s about 8 µM in the blood if it is all absorbed. The lung is quite good at absorbing things. People are now using alcohol vapour to study chronic effects of alcohol. Morris: The partition coefficient of alcohol is about 2000 versus 200 for acetaldehyde. This favours the absorption of alcohol. Thornalley: About 10 years ago we addressed the issue of reversible aldehyde binding to plasma proteins with methylglyoxal. We distinguished between free, reversibly bound and irreversibly bound methylglyoxal by using ultrafiltration. If albumin is incubated with methylglyoxal and this is followed by limited ultrafiltration where there is less than 10% change of volume, the methylglyoxal in the first ultrafiltrate is the free form. If you then wash exhaustively three or four times by ultrafiltration then the reversibly bound methylglyoxal comes off the protein into the combined ultrafiltrate and can be quantified. Then there are the irreversible adducts. The main one is an arginine adduct which is quasi-reversible but with a half-life of about two weeks under physiological conditions. So for all intents and purposes this is irreversibly bound. Eriksson: If you have a biological sample you need to do things fast. So while this distinction between reversible and irreversible binding is a useful one, it has its practical limits. Thornalley: In measuring volatiles in breath, can you reference the acetaldehyde in breath to any other volatile that would be useful? Can you measure acetone, for example? Eriksson: Yes. We have an internal standard, but it is the alcohol. If you take saliva or blood at the same time you can overcome these difficulties quite elegantly.

136

TARDIF

Ren: I have a point for clarification. Dr Tardif used the term ‘background’ levels of acetaldehyde. What was he referring to? Eriksson: Acetaldehyde is sometimes found as a pollutant in the air. Controls are absolutely essential, therefore. We used to work in the Finnish alcohol monopoly, and in that whole building there was an endogenous ethanol and acetaldehyde level in the air! Ren: How high are background levels? Morris: Background acetaldehyde concentrations in the USA are about 10 parts per billion, which is 0.4 nM/litre. Eriksson: It must depend very much on the environmental conditions. This summer we were faced with a problem because of the forest fires in Russia. Apte: You said earlier that venous blood acetaldehyde levels are almost non-detectable. Eriksson: They are almost undetectable in most conditions. The problem is not the detection limit but more the fact that any treatment of blood, even control blood, involves some artefactual acetaldehyde formation from internal or external alcohol which has to be controlled for. References Heap L, Ward RJ, Abiaka C et al 1995 The influence of brain acetaldehyde on oxidative status, dopamine metabolism and visual discrimination task. Biochem Pharmacol 50:263–270 Fukunaga T, Sillanaukee P, Eriksson CJP 1993 Problems involved in the determination of endogenous acetaldehyde in human blood. Alcohol Alcohol 28:535–541

Ethanol and acetaldehyde: in vivo quantitation and effects on cholinergic function in rat brain Mostofa Jamal, Kiyoshi Ameno, Uekita Ikuo, Mitsuru Kumihashi, Weihuan Wang and Iwao Ijiri Department of Forensic Medicine, Faculty of Medicine, Kagawa University, Ikenobe, Miki, Kita, Kagawa 761-0793, Japan

Abstract. First, ethanol (EtOH) and acetaldehyde levels were determined simultaneously in the striatum of free-moving rats after administration of their major oxidative enzyme inhibitors followed by EtOH. The results showed that acetaldehyde was present in the cyanamide (CY) + EtOH, CY + 4-methylpyrazole (4-MP) + EtOH and CY + sodium azide + EtOH groups. The CY + EtOH-induced peak acetaldehyde level was 195.2 ± 19.4 µM, and this value was significantly higher than those in the other groups. The peak EtOH level was 25.9 ± 2.3 mM in the CY + 4-MP + EtOH group, and this level was considerably higher than the value in EtOH. No significant difference in brain EtOH levels was found in any of the other groups studied. Second, the effects of EtOH and acetaldehyde on choline acetyltransferase (ChAT) and acetylcholinesterase (AChE) were investigated in the frontal cortex and hippocampus of high acetaldehyde-producing rats using RT-PCR and Western blot. The results showed that EtOH and acetaldehyde decreased ChAT expression at 40 and 240 min after EtOH dosing in the brain. The acetaldehyde-induced reduction in ChAT expression was significantly higher than that induced by EtOH. No remarkable alteration of AChE expression was observed. The study suggested that catalase made a significant contribution to acetaldehyde formation in the rat brain, and that EtOH and acetaldehyde decreased ChAT expression at 40 and 240 min after EtOH dosing. 2007 Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Foundation Symposium 285) p 137–144

Brain acetaldehyde accumulation after ethanol (EtOH) intake has been a topic of debate. Two observations have been made with respect to acetaldehyde formation as well as acetaldehyde transport to the brain: (1) peripherally formed high acetaldehyde may be able to cross the blood–brain barrier (BBB) and enter the brain (Hoover et al 1981), and (2) the enzyme catalase is the chief mediator of acetaldehyde formation in the brain (Aragon et al 1992). It is, therefore, essential to measure the brain acetaldehyde level after inhibition of the major oxidative pathways of EtOH and acetaldehyde metabolism. The oxidation of EtOH occurs 137

138

JAMAL ET AL

via several enzymatic pathways that include alcohol dehydrogenase (ADH), catalase and CYP2E1 (Ramchandani et al 2001). Another enzyme, aldehyde dehydrogenase (ALDH), breaks the acetaldehyde down into acetate. In this study, a very simple brain microdialysis method was employed for simultaneous quantitation of in vivo EtOH and acetaldehyde in the brain of free-moving rats after administration of their oxidative enzyme inhibitors. The neurons of the brain are adversely affected by the ingestion of alcohol. It is, therefore, of considerable interest to know whether these central effects are due, in part, to acetaldehyde, a metabolite of EtOH, in the brain. EtOH has a substantial effect on the cholinergic system, and this effect may share some features with Alzheimer’s disease (Arendt et al 1988). Therefore, the cholinergic deficits after EtOH consumption have been well documented, but acetaldehyde’s effects on the CNS have remained a matter of debate. Thus, we focused on choline acetyltransferase (ChAT) and acetylcholinesterase (AChE) expression at 40 and 240 min after EtOH dosing in the brain of high acetaldehyde-producing rats in order to clarify the effects of EtOH and acetaldehyde on these enzymes. Materials and methods Animal Male Wistar rats (9–10 weeks-old, 250–300 g) were used throughout the study. For the quantitation of EtOH and acetaldehyde, the rats were divided into the following 7 experimental groups: (a) EtOH (1 g/kg), (b) cyanamide (CY, 50 mg/kg, an ALDH inhibitor) + EtOH, (c) CY + 4-methylpyrazole (4MP, 82 mg/kg, an ADH inhibitor) + EtOH, (d) CY + sodium azide (10 mg/kg, a catalase inhibitor) + EtOH, (e) 4MP + EtOH, (f) sodium azide + EtOH, and (g) CY alone. For the ChAT and AChE analyses, the rats were divided into saline (0.9%), EtOH, and CY + EtOH groups. In both trials, rats received an i.p. injection of EtOH (20% v/v) 30 and 60 min after a dose of sodium azide and CY or 4MP, respectively. Microdialysis The microdialysis procedure and chromatographic conditions were as previously described ( Jamal et al 2003). Using this method, no artefactual acetaldehyde was detectable in the dialysate. In vitro probe recovery of EtOH and acetaldehyde were 72.2 ± 3.6 and 52.6 ± 5.9%, respectively, and the values were corrected. RT-PCR and Western blot The procedures used for the ChAT and AChE analyses have been described in detail previously ( Jamal et al 2006).

BRAIN ACETALDEHYDE

139

Results and discussion Table 1 shows the EtOH and acetaldehyde levels 30 min after EtOH dosing in the striatum. Acetaldehyde was detected in the CY + EtOH, CY + 4MP + EtOH, and CY + sodium azide + EtOH groups. The CY + EtOH-induced peak acetaldehyde level (195.2 ± 19.4 µM) was significantly greater than the values in the other groups. Treatment with inhibitors did not modify the levels of EtOH that reached the brain from the bloodstream, with the exception that treatment with CY + 4MP + EtOH showing the peak EtOH level (25.9 ± 2.3 mM) was significantly higher than the value for EtOH. Therefore, the results of this study support the hypothesis that catalase and the high level of peripherally-formed acetaldehyde contributed to brain acetaldehyde accumulation in the CY + EtOH group. Peripherally-formed high acetaldehyde played a role in brain acetaldehyde accumulation in the CY + sodium azide + EtOH and catalase played a role in brain acetaldehyde accumulation in the CY + 4MP + EtOH group. Thus, the data obtained from this study demonstrated that catalase made a remarkable contribution to brain acetaldehyde formation. Table 2 shows the percentage of ChAT and AChE expression at 40 and 240 min after EtOH injection in the brain. RT-PCR analysis revealed a significant decrease in ChAT levels at 40 min in the EtOH and CY + EtOH groups, respectively, in the frontal cortex and hippocampus. The hippocampal reduction in ChAT was significantly higher in the CY + EtOH group than in the EtOH group. These findings were consistent with the results of the Western blot analysis, but one discrepancy was found. ChAT mRNA expression was reduced significantly at 240 min in the frontal cortex and hippocampus, respectively, in the CY + EtOH group as revealed by RT-PCR. The reduction in ChAT levels was markedly higher in the CY + EtOH

TABLE 1 The Peak EtOH and acetaldehyde (ACe) levels in the brain Groups CY + EtOH CY + 4MP + EtOH CY + sodium azide + EtOH Sodium azide + EtOH 4MP + EtOH EtOH CY

EtOH level (mM)

ACe level (mM)

17.8 ± 1.1 25.9 ± 2.3 17.9 ± 1.9 16.3 ± 1.7 19.7 ± 1.9 20.1 ± 1.9§ ND

195.2 ± 19.4 129.3 ± 12.1* 76.9 ± 7.3†‡ ND ND ND ND

Data represent mean ± SD (n = 5), analysed by student’s t-test. § P < 0.05, for the difference between EtOH and CY + 4MP + EtOH. * P < 0.05, for the difference between CY + EtOH and CY + 4MP + EtOH, † P < 0.05, for the difference between CY + EtOH and CY + sodium azide + EtOH, and ‡ P < 0.05, for the difference between CY + 4MP + EtOH and CY + sodium azide + EtOH. ND, not determined.

140

TABLE 2 ChAT and AChE mRNA, and ChAT protein expression in the rat frontal cortex (A) and hippocampus (B) mRNAs (% of control) 40 min Groups (A) (B)

Saline EtOH CY + EtOH Saline EtOH CY + EtOH

ChAT 100 ± 10.1 72.8 ± 8.2* 71.6 ± 4.1† 100 ± 7.1 76.5 ± 6.3* 53.0 ± 4.2†‡

Protein (% of control) 240 min

AChE 100 ± 15.0 87.5 ± 19.0 82.8 ± 17.2 100 ± 19.0 115.0 ± 18.3 95.0 ± 15.0

ChAT 100 ± 15.3 92.2 ± 17.1 62.0 ± 9.0†‡ 100 ± 18.2 97.6 ± 17.5 65.5 ± 12.2†‡

AChE 100 ± 18.0 85.6 ± 10.0 99.8 ± 15.0 100 ± 9.1 102.9 ± 18.0 88.2 ± 15.3

40 min

240 min

ChAT

ChAT

100 ± 9.1 49.0 ± 8.2* 34.5 ± 5.0†‡ 100 ± 9.0 54.9 ± 6.1* 32.0 ± 8.3†‡

100 ± 9.1 86.5 ± 15.4 34.1 ± 9.2†‡ 100 ± 11.0 87.8 ± 13.0 49.6 ± 8.1†‡

Data represent mean ± SD (n = 5) as percentage of control, analysed by two-way ANOVA followed by post hoc Tukey-Kramer test. * P < 0.005, for the difference between saline and EtOH, † P < 0.005, for the difference between saline and CY + EtOH, and ‡ P < 0.005, for the difference between EtOH and CY + EtOH.

JAMAL ET AL

BRAIN ACETALDEHYDE

141

group than in the EtOH group. The EtOH group did not show any significant change in the ChAT mRNA levels at 240 min. These findings were also consistent with the results of the Western blot analysis. No significant difference of AChE mRNA was found in either group. The brain EtOH levels were 20.1 ± 1.9 and 0.65 ± 0.1 mM at 30 and 240 min, respectively, and the acetaldehyde levels were 195.2 ± 19.4 and 51.7 ± 11.6 µM at 30 and 240 min, respectively. The data obtained from this study demonstrated that EtOH and acetaldehyde reduced ChAT expression at 40 and 240 min after EtOH injection in the rat brain, and the acetaldehydeinduced decrease in ChAT expression was significantly higher than the decrease by EtOH. Summary We conclude that the enzyme catalase contributed to the accumulation of acetaldehyde in the rat brain. EtOH and acetaldehyde both decreased ChAT expression in the rat brain, and the acetaldehyde-induced decrease in ChAT expression was considerably higher than EtOH-induced. Acknowledgement This study was funded by a Grant-in-Aid for Scientific Research (c) (No-18590637) from the Ministry of Education, Science and Culture, Japan.

References Aragon CM, Rogan F, Amit Z 1992 Ethanol metabolism in rat brain homogenates by a catalase H2O2 system. Biochem Pharmacol 44:93–98 Arendt T, Henning D, Gray JA, Marchbanks R 1988 Loss of neurons in the rat basal forebrain cholinergic projection system after prolonged intake of ethanol. Brain Res Bull 21:563–569 Hoover DJ, Brien JF 1981 Acetaldehyde concentration in rat blood and brain during the calcium carbimide–ethanol interaction. Can J Physiol Pharmacol 59:65–70 Jamal M, Ameno K, Kumihashi M et al 2003 Microdialysis for the determination of acetaldehyde and ethanol concentration in the striatum of freely moving rats. J Chromatogr B 798: 155–158 Jamal M, Ameno K, Ameno S et al 2007 Changes in cholinergic function in the frontal cortex and hippocampus of rat exposed to ethanol and acetaldehyde. Neuroscience 144: 232–238 Ramchandani VA, Bosron WF, Li TK 2001 Research advances in ethanol metabolism. Pathol Biol 49:676–682

DISCUSSION Quertemont: In your experiments, when you tried to measure acetaldehyde in the brain after an injection of ethanol alone, i.e. without any other drug that modulates

142

JAMAL ET AL

ethanol metabolism, you found non-detectable levels of acetaldehyde. Many papers have been published showing that catalase inhibitors reduced many different alcohol-induced behaviours. In the brain, the major part of acetaldehyde is produced from ethanol metabolism by the enzyme catalase. Therefore, catalase inhibitors are believed to exert their behavioural effects through a decrease in brain acetaldehyde concentrations after alcohol administration, and these studies often interpret their results in terms of the role of acetaldehyde in alcohol’s effects (Quertemont et al 2005). One of the problems with this explanation is that so far it hasn’t been possible to show in vivo that catalase inhibitors reduce the concentrations of acetaldehyde within the brain. As you have an undetectable level of acetaldehyde after ethanol administration alone, would it be possible with your methodology to show a significant reduction of brain acetaldehyde concentrations with catalase inhibitors without the concomitant administration of cyanamide? Such a result would greatly improve our understanding of the role of acetaldehyde in the behavioural effects of alcohol. Jamal: I used a different group, cyanamide plus ethanol, which results in a remarkable increase in acetaldehyde in the blood and brain. In the ethanol only group, there is no detection of acetaldehyde in the brain. The ethanol in the brain is highly metabolized to acetaldehyde and then metabolized to acetic acid. This is the reason I couldn’t detect any acetaldehyde in the ethanol group. Catalase does important things. It produces acetaldehyde in the brain, but we didn’t know how much. So I used the catalase inhibitor powered by ethanol. Catalase inhibitor in combination with cyanamide produced acetaldehyde in the brain, but the cyanamide plus ethanolinduced acetaldehyde production was significantly greater. Eriksson: This was an old argument about the acetaldehyde levels in brain. It has been clearly shown that there is a cut-off point in brain (Eriksson 1977) after which it starts to begin to be detected. This is close to 200 µM. This doesn’t mean that acetaldehyde is not formed in the brain. Indeed, Dr Jamal’s data show that there is active catalase present. It works also during the normal metabolism, but the acetaldehyde is rapidly metabolised to acetate and is also bound or condensated in brain, which may be important. It may be the immediate reactions which are important in the brain and these are hard to study. Preedy: I think that’s an important point. It would be interesting to look at the level of adduct formation. You could correlate this with the actual levels of acetaldehyde. Eriksson: That would be perfect if it was possible, but so far it has not been possible to replicate adduct formation that can be correlated with normal concentrations of acetaldehyde. Worrall: I have measured adduct levels in brain from ethanol-fed rats and from human cerebellar degeneration. I’ll mention this in my paper.

BRAIN ACETALDEHYDE

143

Deitrich: There is another complication with cyanamide. If you inhibit catalase, then cyanamide will not be effective since it is required to convert cyanamide to an active inhibitor of ALDH. There is a cross-talk with cyanamide which is an inhibitor of ALDH after reaction with catalase, and catalase inhibition, which renders cyanamide ineffective. There is therefore a complication in using the catalase inhibitor with cyanamide as an ALDH inhibitor (Deitrich 2004). Yin: Also, azide can inhibit cytochrome oxidase. There are a lot of complications with azide use. Quertemont: One of the problems with the studies that have been published on catalase inhibitors is that many of them have a lot of non-specific effects. Some of them are also toxic. Eriksson: Did I understand correctly that alcohol had dual effects, first increasing the acetylcholine, and then decreasing it after some time? Jamal: Yes. Our studies showed this. I would like to find out the reason for the short-term increase in acetylcholine. The results of the present findings have confirmed that the brief elevation in acetylcholine release is presumably due to the arousal from the injection of EtOH (Jamal et al 2005). Eriksson: This came to my mind: that there could be an initial reaction, which in itself would be an interesting finding. Jamal: The other thing I need to investigate is if the choline acetyltransferase and acetylcholinesterase transport proteins are responsible for the mechanism of acetylcholine reduction in the rat brain after ethanol and acetaldehyde exposure. Choline acetyltransferase catalyses the biosynthesis of acetylcholine in the cytoplasm of presynaptic terminals, whereas acetylcholinesterase is responsible for degradation of acetylcholine to acetate and choline in the synaptic cleft. Apte: I was interested in the ethanol effect on choline acetyltransferase at 40 min, which seemed to be lost at 240 min. What is the explanation? The acetaldehyde persisted. Jamal: Ethanol produces a short-term elevation in acetylcholine at 40 min followed by a decrease ( Jamal et al 2005). The present results indicated an opposite decrease in choline acetyltransferase at 40 min while ethanol reached a peak and acetaldehyde was not detected. This ethanol-induced elevation did not correspond with the result of choline acetyltransferase. There needs to be further studies to confirm this. Eriksson: There is acute tolerance to alcohol effects. 240 min is enough to get some tolerance. References Deitrich R 2004 Acetaldehyde: Deja Vu du Jour. J Stud Alcohol 65:557–572 Eriksson CJP 1977 Acetaldehyde metabolism in vivo during ethanol oxidation. Adv Exp Med Biol 85a:319–341

144

JAMAL ET AL

Quertemont E, Tambour S, Tirelli E 2005 The role of acetaldehyde in the neurobehavioral effects of ethanol: a comprehensive review of animal studies. Prog Neurobiol 75:247–274 Jamal M, Ameno K, Wang W et al 2005 Inhibition of acetaldehyde metabolism decreases acetylcholine release in medial frontal cortex of freely moving rats. Brain Res 1039: 90–96

Pharmacological treatments and strategies for reducing oral and intestinal acetaldehyde Ville Salaspuro Research Unit of Substance Abuse Medicine, University of Helsinki, Biomedicum Helsinki, Finland

Abstract. Strong epidemiological, genetic and biochemical evidence indicates that local acetaldehyde exposure is a major factor behind gastrointestinal cancers especially associated with alcohol drinking and smoking. Thus, reducing the exposure to carcinogenic acetaldehyde either by decreasing the production or by eliminating acetaldehyde locally might offer a preventive strategy against acetaldehyde-induced gastrointestinal cancers. Thiol products, such as the amino acid cysteine, are known to be able to protect against acetaldehyde toxicity. Cysteine is able to bind acetaldehyde efficiently by forming a stable thiazolidine–carboxylic acid compound. Special cysteine preparations (such as lozenge and chewing gum) have already been developed to bind smoking and alcohol drinking derived acetaldehyde from the oral cavity. Most importantly, these type of drug formulations offer a novel method for intervention studies aimed to resolve the eventual role of acetaldehyde in the pathogenesis of upper digestive tract cancers. Acetaldehyde exposure could also be influenced by modifying the acetaldehyde producing microbiota. With regard to the upper digestive tract, acetaldehyde production from ingested ethanol could be significantly reduced by using an antiseptic mouthwash, chlorhexidine. In the large intestine acetaldehyde production could be markedly decreased either by reducing the Gram-negative microbes by ciprofloxacin antibiotic or by lowering the intraluminal pH by lactulose. 2007 Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Foundation Symposium 285) p 145–157

Alcohol and tobacco are the main risk factors for upper digestive tract cancers ( Bagnardi et al 2001, Castellsagué et al 1999, Zeka et al 2003). A plausible explanation for the high cancer risk in alcoholics and smokers is the increased exposure to high local acetaldehyde concentrations in the upper digestive tract (Salaspuro & Salaspuro 2004, Salaspuro 2003). Carcinogenic acetaldehyde is produced from ingested alcohol by oral microbes during ethanol metabolism. On the other hand, during tobacco smoking acetaldehyde from the smoke becomes easily dissolved into the saliva. Both of these result in significant acetaldehyde concentration in the saliva which, via swallowing, is distributed to all parts of the upper digestive 145

146

SALASPURO

tract. Thus, combined exposure to acetaldehyde derived from tobacco smoke and microbial alcohol metabolism might explain the multiplicative and synergistic risk effect. There is also a causal link between alcohol consumption and cancers of the colon and rectum (Kune & Vitetta 1992, Scheppach et al 1999). In the large intestine there are many aerobic and/or facultative anaerobic bacteria capable of ethanol oxidation with subsequent production of acetaldehyde (Salaspuro 1996). Therefore ethanol consumption leads to marked intracolonic acetaldehyde concentration ( Jokelainen et al 1996, Seitz et al 1990). During ethanol metabolism acetaldehyde is produced by normal digestive tract microbiota. This microbially mediated pathway for acetaldehyde production can be modulated by treatment with antiseptics, antibiotics or prebiotics, which affect the microbial counts or the metabolic activity of the microbes. On the other hand, acetaldehyde can be inactivated by locally administered cysteine. Cysteine is a sulfur-containing amino acid present in dietary proteins. Decades ago it was found in experimental studies that free cysteine is able to protect against the lethal effects of acetaldehyde (Sprince et al 1974). Cysteine is able to ameliorate the toxicity of acetaldehyde by forming a stable 2-methyl-thiazolidine-4-carboxylic acid (MTCA) adduct (Cederbaum & Rubin 1976). This review presents known strategies and a novel experimental approach to decrease the local acetaldehyde exposure of the upper and lower digestive tracts, with an eventual goal of reducing the prevalence of digestive tract cancers.

Reducing the acetaldehyde exposure in the upper digestive tract Chlorhexidine After ingestion, ethanol is present in saliva in concentrations comparable to blood ethanol levels. Human oral flora is known to contain over 350 cultivatable species of bacteria and many of them are capable of producing mutagenic (>50 µM) amounts of acetaldehyde from ethanol. For instance, certain Streptococcus and Neisseria species present in the oral flora possess high alcohol dehydrogenase (ADH) activity and are able to produce high levels of acetaldehyde from ethanol in vitro (Kurkivuori et al 2006, Muto et al 2000). Poor oral hygiene is an accepted risk factor for oral cancer, which is associated with bacterial overgrowth and an increased in vitro salivary acetaldehyde production from ethanol (Homann et al 2000). Acetaldehyde produced in the saliva by oral microbes can be significantly reduced by using an antiseptic mouthwash, chlorhexidine. In a study of 10 volunteers salivary acetaldehyde levels after ethanol drinking decreased by about 40% after three-day treatment with chlorhexidine (Homann et al 1997). The acetaldehyde peak levels decreased from 143.3 µM to 49 µM. There is recent evidence suggesting that

ELIMINATION OF ACETALDEHYDE

147

acetaldehyde levels over 50 µM are mutagenic ( Theruvathu et al 2005). In addition to chlorhexidine mouthwash, an invention involving chlorhexidine, formulated as a controlled-release chip fixed with a dental device has been hypothesized to be a rational strategy for reducing acetaldehyde production by oral microbiota (Rota & Poggi 2003). Cysteine Cysteine is a sulfur-containing amino acid. With respect to human physiology, it is considered as semiessential, since it can be synthesized from methionine and serine. Cysteine is normally consumed as a component of dietary proteins. The estimated average intake of cysteine is about 1 g/day. Cysteine, cystine (two cysteine molecules attached by disulfide linkage) and cysteine-containing peptides are absorbed after digestion from the small intestine. The importance of cysteine is related to the presence of a sulfur-containing functional thiol group in its side chain. Cysteine plays a key role in the regulation of cellular redox state and is also a rate-limiting amino acid in the synthesis of glutathione (Shoveller et al 2005). Earlier data strongly suggested that cysteine should be administered locally in order to be effective in the elimination of acetaldehyde. Due to its reactivity, it is unlikely that free, effective cysteine will reach the desired site of action through systemic delivery. Thus, the local liberation of cysteine to bind acetaldehyde should be preferred. This type of in situ method is associated with several benefits including (i) cysteine would not have to be absorbed from the gut, (ii) formed acetaldehyde is directly and immediately bound at the site of its formation and (iii) the total amount of delivered cysteine can be kept low. The condensation product, MTCA, is formed from the reaction of acetaldehyde with cysteine according to the following equation:

COOH CH3CHO +

H2N

CH

COOH

CH2 CH 3

SH Acetaldehyde

+

Cysteine

↔

HN

CH

CH

CH2

+

H 2O

S

2-methyl-thiazolidine-

+

water

4-carboxylic acid (MTCA)

There is some evidence that this thiazolidine derivative may undergo nitrosation in vivo, forming nitroso-thiazolidine-4-carboxylic acid. This substance is detected

148

SALASPURO

in vivo from human urine especially from smokers without cysteine supplementation, indicating endogenous deactivation of acetaldehyde by cysteine (Ohshima & Bartsch 1984). There exits a strong and linear correlation between salivary acetaldehyde and ethanol levels (Homann et al 1997). Therefore, acetaldehyde is continuously formed in the saliva during ethanol challenge. In order to bind the acetaldehyde during ethanol oxidation in the oral cavity, a special buccal drug formulation with Lcysteine is required. A tablet which slowly releases intact L-cysteine enables continuous and direct local binding of reactive acetaldehyde. In a human in vivo study, up to two-thirds of acetaldehyde could be removed from saliva after ethanol intake with this type of buccal tablet (Fig. 1). Tablets containing 100 mg of L-cysteine reduced the acetaldehyde exposure (expressed as area under the curve, AUC) of the volunteers from 162.3 ± 34.2 µM × h to 54.3 ± 11 µM × h (P = 0.003) as compared to placebo tablet, after ingestion of ethanol (0.8 g/kg of body weight) (Salaspuro et al 2002). This finding implies that this drug formulation could potentially be used for the prevention of the local toxic effects of acetaldehyde during alcohol consumption in the oral cavity and upper digestive tract. In addition to ethanol, another major source of acetaldehyde is tobacco, since acetaldehyde is one of the major toxic components of tobacco smoke (Fowles &

50 Placebo Cysteine

Acetaldehyde (µM)

40

30

20

10

0 0

50

100

150

200

250

300

min

FIG. 1. In vivo acetaldehyde levels (mean ± SEM) in saliva of volunteers with placebo or Lcysteine-containing buccal drug formulation after a dose of alcohol. Differences between concentrations are significant at all time points from 20 min to 320 min (P ≤ 0.001). Adapted from Salaspuro & Salaspuro (2004).

ELIMINATION OF ACETALDEHYDE

149

Dybing 2003, Hoffmann & Hoffmann 1997). Acetaldehyde is easily dissolved from the mainstream tobacco smoke into saliva and is further distributed to the entire upper digestive tract of smokers via swallowing. During a smoking period, in vivo salivary acetaldehyde level increases markedly ranging from 100 µM to 400 µM. In a placebo controlled in vivo study with smokers it was shown that carcinogenic acetaldehyde can be totally inactivated by a lozenge containing 5 mg of L-cysteine (Salaspuro et al 2006). In addition, there was an inverse dose–response relationship between the cysteine content of the lozenge and salivary acetaldehyde concentration (Fig. 2). As acetaldehyde is totally eliminated from saliva the carcinogenic potential of acetaldehyde is probably eliminated also in the oesophagus and stomach. Most importantly, these type of drug formulations offer a novel method for intervention studies aimed to resolve the real role of acetaldehyde in the pathogenesis of upper digestive tract cancers. The idea of the chemopreventive mechanism of this tablet is that it is sucked during every smoking period. Because of the resolving time of about 6 minutes, the tablet designed for tobacco is not ideal for longer smoking periods as in the case of cigar smoking. After the tablet has dissolved, the salivary acetaldehyde increases after three minutes to the placebo level (Fig. 3). The use of the L-cysteine tablet is safe because L-cysteine, as a non-essential amino acid, has no known adverse effects in the concentrations described. Furthermore, the daily dose of L-cysteine would be very small even if one is a heavy smoker because an efficient tablet contains only 5 mg of L-cysteine.

Salivary acetaldehyde (µM)

400

300

200

* 100

**

**

**

5mg

10mg

0 0mg

1,25mg

2.5mg

Cysteine concentration in the tablet

FIG. 2. Salivary acetaldehyde levels immediately after tobacco smoking with placebo- or Lcysteine-containing tablet (*P = 0.007, **P < 0.001). Adapted from Salaspuro et al (2006).

150

SALASPURO 350

Acetaldehyde (µM)

300

250

200

150

100

50

0 0

3

6

9

12

15

18

21

min Smoking started with L-cysteine (5mg) tablet

Tablet totally dissolved

FIG. 3. Salivary acetaldehyde concentration in vivo during cigar smoking with a lozenge containing 5 mg of L-cysteine. Adapted from Salaspuro et al (2006).

Reducing the acetaldehyde exposure in the large intestine Intracolonic acetaldehyde emerges from microbial metabolism due to anaerobic fermentation or due to the oxidation of ingested ethanol. There is substantial evidence for microbial production of acetaldehyde from ingested ethanol under aerobic or microaerobic conditions prevailing near the mucosal surface of the large intestine ( Jokelainen et al 1996, Seitz et al 1990). After alcohol consumption ethanol is absorbed from the intestine and is thereafter distributed by blood circulation into the whole water compartment of the human body. Therefore, ethanol concentrations in the colon are comparable with that found in the blood. In the large intestine ethanol can be metabolized by intracolonic microbes to acetaldehyde. Consequently, high acetaldehyde levels have been found after alcohol administration in the colon of experimental animals. These acetaldehyde concentrations are significantly lower in germ-free animals than in conventional rats, highlighting the essential role of microbes in intracolonic acetaldehyde production (Seitz et al 1990). On the other hand, there are measurable levels of acetaldehyde in the large intestine even without ingested ethanol. This acetaldehyde is probably produced

ELIMINATION OF ACETALDEHYDE

151

endogenously during alcoholic fermentation or from fermented ethanol (Visapää et al 1998). Thus, by modifying the microbiota or the environment with antibiotics or prebiotics it is possible to enhance or reduce the local acetaldehyde production in the large intestine. Ciprofloxacin Ciprofloxacin, an antibiotic, possesses excellent antibacterial activities against most aerobic and facultative anaerobic bacteria. It has been shown that ciprofloxacin treatment reduces the total ethanol elimination rate by 9–10% in both rats and humans, which is associated with a reduction in faecal aerobic flora and faecal ADH activity ( Jokelainen et al 1996, Tillonen et al 1999). In this way the formation of endogenous acetaldehyde in the large intestine could be abolished almost totally. Furthermore, the intracolonic production of acetaldehyde from ingested ethanol could also be almost totally inhibited by ciprofloxacin treatment. The mean intracolonic acetaldehyde concentration in the rats after a dose of ethanol (1.5 g/kg of body weight i.p.) was reduced by ciprofloxacin treatment from 483 ± 169 µM to 23 ± 15 µM (Visapää et al 1998). Lactulose Lactulose is a non-absorbable disaccharide, which is rapidly metabolized by bacteria representing normal large bowel microbiota to lactic acid. This results in a significant decrease in the pH of the contents of the large intestine. By acidifying the intracolonic environment with lactulose, the pH-dependent ADH-mediated microbial ethanol oxidation and acetaldehyde production is proportionally inhibited. It has been shown in rats that lactulose feeding effectively inhibits ethanol oxidation by colonic microbes and subsequently reduces intracolonic acetaldehyde levels (Zidi et al 2003). Interestingly, lactulose has also been shown to significantly decrease the recurrence rate of colorectal adenomas (Roncucci et al 1993), which thus could be related to the ability of lactulose to decrease the concentration of carcinogenic acetaldehyde in the large intestine. In conclusion, numerous studies emphasize the potential role of acetaldehyde in the carcinogenesis of the digestive tract in humans. L-cysteine effectively binds reactive acetaldehyde when used as different drug formulations designed to be used during alcohol drinking and/or tobacco smoking. As a safe amino acid cysteine could be used to reduce the local acetaldehyde exposure, especially among certain high risk groups, e.g. ALDH2-deficient Asians, smokers and heavy drinkers. Thus, these cysteine preparations warrant further clinical intervention studies to find out the eventual role of acetaldehyde in the pathogenesis of upper digestive tract cancers.

152

SALASPURO

In order to reduce the acetaldehyde exposure derived from microbial metabolism, good oral health is recommended. Also the use of antiseptic mouthwash (without alcohol) efficiently reduces the number of oral microbes which results in a decreased acetaldehyde exposure after alcohol consumption. With regard to the large intestine, experimental results show that ciprofloxacin and lactulose are efficient in reducing intracolonic acetaldehyde levels. In particular, lactulose, as a harmless prebiotic, should be considered as a tentative agent for further clinical trials.

References Bagnardi V, Blangiardo M, La Vecchia C, Corrao G 2001 A meta-analysis of alcohol drinking and cancer risk. Br J Cancer 85:1700–1705 Castellsagué X, Muñoz N, De Stefani E et al 1999 Independent and joint effects of tobacco smoking and alcohol drinking on the risk of esophageal cancer in men and women. Int J Cancer 82:657–664 Cederbaum AI, Rubin E 1976 Protective effect of cysteine on the inhibition of mitochondrial functions by acetaldehyde. Biochem Pharmacol 25:963–973 Fowles J, Dybing E 2003 Application of toxicological risk assessment principles to the chemical constituents of cigarette smoke. Tob Control 12:424–430 Hoffmann D, Hoffmann I 1997 The changing cigarette, 1950–1995. J Toxicol Environ Health 50:307–364 Homann N, Jousimies-Somer H, Jokelainen K, Heine R, Salaspuro M 1997 High acetaldehyde levels in saliva after ethanol consumption: methodological aspects and pathogenetic implications. Carcinogenesis 18:1739–1743 Homann N, Tillonen J, Meurman JH et al 2000 Increased salivary acetaldehyde levels in heavy drinkers and smokers: a microbiological approach to oral cavity cancer. Carcinogenesis 21:663–668 Jokelainen K, Matysiak-Budnik T, Mäkisalo H, Höckerstedt K, Salaspuro M 1996 High intracolonic acetaldehyde values produced by a bacteriocolonic pathway for ethanol oxidation in piglets. Gut 39:100–104 Kune GA, Vitetta L 1992 Alcohol consumption and the etiology of colorectal cancer: a review of the scientific evidence from 1957 to 1991. Nutr Cancer 18:97–111 Kurkivuori J, Salaspuro V, Kaihovaara P et al 2007 Acetaldehyde production from ethanol by oral streptococci. Oral Oncol 43:181–186 Muto M, Hitomi Y, Ohtsu A et al 2000 Acetaldehyde production by non-pathogenic Neisseria in human oral microflora: implications for carcinogenesis in upper aerodigestive tract. Int J Cancer 88:342–350 Ohshima H, Bartsch H 1984 Monitoring endogenous nitrosamine formation in man. In: Berlin A, Draper M, Hemminki K, Vainio H (eds) Monitoring human exposure to carcinogenic and mutagenic agents. International Agency for Research on Cancer, Oxford University Press, p 233–246 Roncucci L, Di Donato P, Carati L et al 1993 Antioxidant vitamins or lactulose for prevention of the recurrence of colorectal adenomas. Dis Colon Rectum 36:227–234 Rota MT, Poggi P 2003 Reduction of oral acetaldehyde levels using a controlled-release chlorhexidine chip as a prevention strategy against upper digestive tract cancer. Med Hypotheses 60:856–858 Salaspuro M 1996 Bacteriocolonic pathway for ethanol oxidation: characteristics and implications. Ann Med 3928:195–200

ELIMINATION OF ACETALDEHYDE

153

Salaspuro M 2003 Alcohol consumption and cancer of the gastrointestinal tract. Best Pract Res Clin Gastroenterol 17:679–694 Salaspuro V, Salaspuro M 2004 Synergistic effect of alcohol drinking and smoking on in vivo acetaldehyde concentration in saliva. Int J Cancer 111:480–483 Salaspuro V, Hietala J, Kaihovaara P, Pihlajarinne L, Marvola M, Salaspuro M 2002 Removal of acetaldehyde from saliva by a slow-release buccal tablet of L-cysteine. Int J Cancer 97: 361–364 Salaspuro VJ, Hietala JM, Marvola ML, Salaspuro MP 2006 Eliminating carcinogenic acetaldehyde by cysteine from saliva during smoking. Cancer Epidemiol Biomarkers Prev 15:146–149 Scheppach W, Bingham S, Boutron-Ruault MC et al 1999 WHO consensus statement on the role of nutrition in colorectal cancer. Eur J Cancer Prev 8:57–62 Seitz HK, Simanowski UA, Garzon FT et al 1990 Possible role of acetaldehyde in ethanol-related rectal cocarcinogenesis in the rat. Gastroenterology 98:406–413 Shoveller AK, Stoll B, Ball RO, Burrin DG 2005 Nutritional and functional importance of intestinal sulfur amino acid metabolism. J Nutr 135:1609–1612 Sprince H, Parker C, Smith G, Gonzales L 1974 Protection against acetaldehyde toxicity in the rat by L-cysteine, thiamin and L-2-methylthiazolidine-4-carboxylic acid. Agents Actions 4:125–130 Theruvathu JA, Jaruga P, Nath RG, Dizdaroglu M, Brooks PJ 2005 Polyamines stimulate the formation of mutagenic 1,N2-propanodeoxyguanosine adducts from acetaldehyde. Nucleic Acids Res 33:3513–3520 Tillonen J, Homann N, Rautio M, Jousimies-Somer H, Salaspuro M 1999 Ciprofloxacin decreases the rate of ethanol elimination in humans. Gut 44:347–352 Visapää JP, Jokelainen K, Nosova T, Salaspuro M 1998 Inhibition of intracolonic acetaldehyde production and alcoholic fermentation in rats by ciprofloxacin. Alcohol Clin Exp Res 22:1161–1164 Zeka A, Gore R, Kriebel D 2003 Effects of alcohol and tobacco on aerodigestive cancer risks: a meta-regression analysis. Cancer Causes Control 14:897–906 Zidi SH, Linderborg K, Väkeväinen S, Salaspuro M, Jokelainen K 2003 Lactulose reduces intracolonic acetaldehyde concentration and ethanol elimination rate in rats. Alcohol Clin Exp Res 27:1459–1462

DISCUSSION Preedy: I have read that acetaldehyde or some protein adducts are actually toxic to some cells. Is it possible that the adduct that you are forming in very high concentrations could be toxic? V Salaspuro: Not that I am aware of. I have examined the rather sparse literature on this topic and there is no evidence of toxicity. Preedy: It would be interesting to test it in vitro in cell lines. Emery: Is the effect specific to cysteine or can you do a similar job with other amino acids? V Salaspuro: We have tried this with several amino acids and substances, such as N-acetylcysteine. This also binds to acetaldehyde but not so effectively.

154

SALASPURO

Worrall: We have done some kinetic studies on thiazolidine adduct formation and breakdown. The amino group is needed near the thiol to get the ring structure, otherwise it just makes a thio-hemi-acetal, which is not stable. I have a question. We use acid to break down the thiazolidine, so are you shifting the acetaldehyde from the mouth further down the gut into the stomach? V Salaspuro: This has occurred to me. If in vivo it is nitrosated, this is a stable compound. But I don’t know the in vivo role of this. In our in vitro studies we have not seen any release of acetaldehyde even in acidic conditions. That might be due to the surplus of cysteine as compared to acetaldehyde. Eriksson: Penicillamine is doing the same job as for making the ring structure. How much would then be hydrolysed in the stomach is an interesting question. Albano: What about the effect of peptides containing cysteine such as glutathione? V Salaspuro: We have tested glutathione and it is not as effective. Reactions would be quite rapid in the mouth and gluthatione seems to act quite slowly in the binding process. Worrall: Again, the cysteine is needed right on the end of the molecule for it to work. Internal cysteines don’t work. As you say, glutathione doesn’t seem to form adducts very well. Eriksson: The rat haemoglobin has these reactive SH groups with NH groups in proximity. They bind acetaldehyde very nicely. Rao: Is this effect stereospecific? Worrall: Both L and D work in vitro. Rao: You could use D-cysteine in addition to the alcoholic beverages. It won’t be absorbed. Worrall: You still have the potential problem of shifting the acetaldehyde further down. Crabb: Are there any bacteria that could live on acetaldehyde? Eriksson: Yes, they make alcohol out of it. V Salaspuro: You can modify ethanol and acetaldehyde production and metabolism by bacteria with different ADHs. Crabb: You could take acetaldehyde to acetate by a bacterium with high expression of an aldehyde oxidase, but it would be a genetically modified food. Deitrich: Making alcohol from acetaldehyde is an energy requiring system since a molecule of NADH is oxidized to NAD. Seitz: What was the background of the study with the lactulose and the polyps? How long was the lactulose given to these individuals? This is an interesting aspect. Given long-term, lactulose results in a softer stool. V Salaspuro: In the Roncucci study (Roncucci et al 1993) they gave lactulose to polypectomized patients 20 mg/day for 18 months and the treatment lowered the recurrence rate of the adenomas. Emery: Do other prebiotics have a similar effect on acetaldehyde?

ELIMINATION OF ACETALDEHYDE

155

V Salaspuro: We haven’t tested this. Crabb: Do you know why pH has that effect on acetaldehyde formation? I take it the inside of a bacterium is somehow pH regulated. What is it about the pH that is changing what the bacteria do? Apte: As a corollary, are there any microbes that have inducible ADH or CYP2E1 activity? V Salaspuro: Yes. These changes have been discovered in smokers, for example. They have increased acetaldehyde production in smoking and ethanol consumption. There must be selection for bacteria with enhanced acetaldehyde production. We don’t know whether it is a selection of bacteria with higher ADH or induced ADH activity. Apte: The idea of having a cysteine lozenge during smoking seems attractive, but tobacco smoke has many other carcinogens and it would be dangerous to lull people into a false sense of security about their smoking. V Salaspuro: I agree. This helps us to find out the role of acetaldehyde, though. Eriksson: This statement is relevant, because many arguments are more derived by emotion rather than thinking. These are legal drugs, although they are causing harm. The idea is that if you reduce the harmful effects it is worthwhile. Rao: We showed that epidermal growth factor (EGF) can prevent acetaldehydemediated toxicity in intestinal epithelia (Seth et al 2004, Sheth et al 2004). Saliva secretes EGF at a very high level, nearly 1000-fold its concentration required for biological activity. Unfortunately, cigarette smoking also reduces this level of EGF secretion. Crabb: I was interested to come across some work from the 1990s which showed that alcohol feeding of animals caused some important changes in the salivary glands (Maier et al 1990). They get fatty parotid glands, and we know that human drinkers get big mumps-like glands. We have heard that it does not lead to acetaldehyde being formed in the saliva, but does it affect the EGF or other factors in the saliva? Seitz: We did a study in which we removed the salivary glands. We thought that this had some effect on growth factors. We couldn’t come up with a clear answer at that time. This was the idea we had. M Salaspuro: I have a comment on this. We cannulated the parotid gland duct on those with either normal or deficient ALDH2. Only those individuals with the deficient enzyme secreted acetaldehyde in the parotid gland saliva. This proves that salivary glands are metabolizing alcohol to acetaldehyde. When they aren’t capable of metabolizing acetaldehyde further, there is a spill-over to saliva. Crabb: Is the level in saliva higher than in the blood? M Salaspuro: Yes, about six times higher. Eriksson: I am not sure this has been proven. Acetaldehyde is penetrating the membranes very easily, so it can come from elsewhere. Why shouldn’t they find any acetaldehyde in the parotid gland without a local alcohol metabolizing system?

156

SALASPURO

And if there is a local system making acetaldehyde, one would expect to find it there. M Salaspuro: Penetrating from where? Eriksson: The saliva, the blood and the vicinity. M Salaspuro: If we put the cannula into the duct, we get just the pure saliva. Eriksson: This pure saliva has still been in the gland for some time and the acetaldehyde could have penetrated into this region. M Salaspuro: Saliva coming from the parotid gland is sterile and reflects the acetaldehyde concentration of the parotid gland cells. There is spill-over of acetaldehyde to saliva only in ALDH2-deficient individuals. In those who have the normal ALDH2 enzyme all acetaldehyde formed in the gland is metabolized further to acetate. Because blood acetaldehyde levels are much lower this is the only way to explain our findings. Eriksson: Then why should it only be formed when there is deficient ALDH? M Salaspuro: Because salivary gland cells can’t remove acetaldehyde efficiently in this case. Aranda: Do the cysteine tablets taste nice? V Salaspuro: No, because of the sulfur. Aranda: I am worried that if you put cysteine in the mouth that bacteria present can use cysteine as a nitrogen source. This will cause the production of all sorts of smelly sulfur products. V Salaspuro: Because of the sulfur cysteine doesn’t taste nice, so we have to add blackcurrant flavour. Smokers, with their reduced taste ability, can’t tell the difference between placebo and one with 5 mg of L-cysteine. Aranda: If you were to add a nitrogen source that is preferred by bacteria you could avoid the bacteria using cysteine as a nitrogen source. Preedy: In terms of taste, have you looked at the effects of eating food on oral acetaldehyde production? Does the change in flora that results change the acetaldehyde production? V Salaspuro: Our experiments are done after fasting. Eating causes a wash-out of bacteria, but this is hard to measure. Crabb: Is acetaldehyde only formed when tobacco is burned, or is there acetaldehyde present with snuff or smokeless tobacco? V Salaspuro: In Sweden a moist snuff is used, and there isn’t any acetaldehyde in this. Eriksson: Yes, acetaldehyde is produced by the burning process. Emery: The antibiotic experiments you reported were interesting. What effect does ciprofloxacin have on other bacteria within the gut? V Salaspuro: In humans a 7 day ciprofloxacin treatment reduced the number of faecal aerobic bacteria from 1.1 × 108 cfu/g to 6.5 × 106 cfu/g. Before the drug administration Enterobacteriaceae was the predominant aerobic flora, and

ELIMINATION OF ACETALDEHYDE

157

Enterococcus sp. were also found. After the ciprofloxacin however, these species totally disappeared from the stool samples. In addition to this, species-specific changes were also detected but those were the major changes. References Maier H, Seitz HK, Mayer B, Adler D, Mall G, Born IA 1990 Lipomatous atrophy of the parotid gland in chronic alcohol consumption. Laryngorhinootologie 69:600–604 Roncucci L, Di Donato P, Carati L et al 1993 Antioxidant vitamins or lactulose for prevention of the recurrence of colorectal adenomas. Dis Colon Rectum 36:227–234 Seth A, Sheth P, Basuroy S, Rao RK 2004 L-glutamine ameliorates acetaldehyde-induced paracellular permeability in Caco-2 cell monolayer. Am J Physiol 287:G510–G517 Sheth P, Seth A, Thangavel M, Basuroy S, Rao RK 2004 Epidermal growth factor prevents acetaldehyde-induced disruption of tight junctions in Caco-2 cell monolayer. Alcohol Clin Exp Res 28:797–804

Alcoholic myopathy and acetaldehyde Victor R. Preedy*, David W. Crabb†, Jaume Farrés‡ and Peter W. Emery* * Nutritional Sciences Research Division, Department of Nutrition and Dietetics, The Franklin Wilkins Building, 150 Stamford Street, London SE1 9NH, UK, † Indiana University School of Medicine, Roudebush VA Medical Center, Emerson Hall, 545 Barnhill Drive, Indianapolis, IN 46202, USA and ‡ Department of Biochemistry and Molecular Biology, Faculty of Sciences, Universitat Autonoma de Barcelona, E-08193 Barcelona, Spain

Abstract. Alcoholic myopathy is characterized by biochemical and morphological lesions within muscle, ranging from impairment of muscle strength and loss of lean tissue to cellular disturbances and altered gene expression. The chronic form of the disease is five times more common than cirrhosis and is characterized by selective atrophy of type II (anaerobic) fibres: type I (aerobic) fibres are relatively protected. Although the causative agent is known (i.e. ethanol), the intervening steps between alcohol ingestion and the development of symptoms and lesions are poorly understood. However, acetaldehyde appears to have an important role in the aetiology of the disease. For example, alcohol is a potent perturbant of muscle protein synthesis in vivo, and this effect is exacerbated by cyanamide pre-dosage, which raises acetaldehyde concentrations. Acetaldehyde alone also reduces muscle protein synthesis in vivo and proteolytic activity in vitro. The formation of acetaldehyde protein adducts is another mechanism of putative importance in alcoholic myopathy. These adducts are formed within muscle in response to either acute or chronic alcohol exposure and the adducts are located preferentially within the sarcolemmal and sub-sarcolemmal regions. However, the significance of protein adduct formation is unclear since we do not currently know the identity of the adducted muscle proteins nor whether adduction alters the biochemical or functional properties of skeletal muscle proteins. 2007 Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Foundation Symposium 285) p 158–182

In this paper we first define then review the nature of the disease entity ‘alcoholic myopathy’ and speculate on the role of acetaldehyde. We focus our attention on (i) the enzymes responsible for acetaldehyde formation and oxidation; (ii) protein adducts, (iii) protein degradation and (iv) protein synthesis. This does not mean to say that there are only four processes affected in alcoholic myopathy. Indeed, we have always argued that in alcohol toxicity, all pathways have the potential to be affected at the cellular and molecular levels and there may be hitherto unexplored inter-tissue mechanisms (Preedy & Watson 2005). For example, we have recently shown molecular involvement in that acute alcohol moderates the expression of at least 400 genes (Arno et al 2006). Furthermore, when endogenous 158

ALCOHOLIC MYOPATHY AND ACETALDEHYDE

159

acetaldehyde levels are raised with cyanamide + ethanol, there are significant increases in muscle mRNA encoding c-myc, which may represent a pre-apoptotic effect, or even a non-specific cellular stress response (Nakahara et al 2003). Studies by other groups in vitro have shown that acetaldehyde impairs biomechanical factors of muscle, such as irreversibly reducing the twitch and the tetanic tension in isolated muscle (Khan 1981). Acetaldehyde also interferes in the interaction of actin and myosin in muscle contraction and the dissociation of the actomyosin complex in a reversible manner in vitro (Puszkin & Rubin 1975). Change in muscle biomechanical features may be due to perturbations in calcium channel proteins and/or calcium regulation (Ohlendieck et al 2003, Oba & Maeno 2004). However, at present, the role of acetaldehyde in modulating muscle contraction and force generation in vivo has not been proven. Limitations in space preclude us from describing all the facets of alcoholic myopathy, and in this article we focus on the role of acetaldehyde in our studies. There are, however, some generalized reviews available elsewhere (Lang et al 2001, 2005, Preedy et al 2001b, 2001a, 2002, 2003, Adachi et al 2003, Urbano-Marquez & Fernandez-Sola 2004, Fernandez-Sola et al 2005).

Effect of alcohol on skeletal muscle Alcohol-induced muscle disease (AIMD; Table 1) is arguably the most prevalent skeletal muscle disorder in the Western Hemisphere but, paradoxically, one of the

TABLE 1 Definitions and features of alcoholic-induced muscle disease Alcohol induced muscle disease (AIMD) Acute alcoholic myopathy

Chronic alcoholic myopathy

A composite term to describe any pathology (molecular, biochemical, structural or physiological) affecting muscle as a consequence of either acute or chronic alcohol ingestion. A rare condition affecting approximately 1% of alcoholics, characterized by swollen painful muscles, usually with myoglobinuria, strikingly raised serum creatine kinase activities and often renal impairment. It occurs after severe alcoholic binges in malnourished alcoholics and affects both red and white muscle fibres (rhabdomyolysis). A common complication of alcoholism affecting approximately 50% of patients. It primarily affects the white muscle fibres of the pectoral and pelvic girdles i.e. a proximal metabolic myopathy though other muscle groups may be involved. Approximately 50% of affected patients show proximal muscle weakness. It is not associated with nutritional, vitamin or mineral deficiencies. The atrophy of the muscle fibres does not lead to an increase in serum creatine kinase and is reversible with 6–12 month abstinence.

160

PREEDY ET AL

least studied. It occurs in up to two-thirds of all chronic alcohol misusers and is thus about four or fives times more common than alcoholic cirrhosis and ranks with the most common alcohol-induced pathologies (Fernandez-Sola et al 2005, Patussi et al 2005). In simple terms, AIMD can be divided into chronic and acute alcoholic myopathy. Working definitions of these are provided in Table 1. In chronic alcoholic myopathy, the diameters of fast-twitch white fibres are markedly reduced: in contrast the slow twitch red fibres are more resilient (Martin et al 1985) (Fig. 1). Patients may lose up to 30% of their muscle mass (Duane & Peters 1988). There may be enhancement of lipid deposition within the muscle, but no overt fibrosis, inflammation or necrosis (Martin et al 1985). In most patients with alcoholic myopathy mitochondrial changes are absent (Martin et al 1985), though there is reduced skeletal muscle strength (Martin et al 1985, Urbano-Marquez et al 1989, Aagaard et al 2003). Muscle strength assessment can form the basis of diagnostic tests, in the absence of facilities for biopsy and histology. Patients with overt liver disease such as cirrhosis also have reduced muscle strength. However, it is important to emphasise that the myopathy can arise independently of liver disease (Martin et al 1985, Aagaard et al 2003), neuropathy (Mills et al 1986) or malnutrition

Control subject Type I Type II

Alcoholic patient Type I Type II

FIG. 1. Type I and II fibres in a control and alcoholic patient. Micrographs show a specimen from a normal control and an alcoholic patient. Muscles have been stained for myosin-ATPase. Thus the lighter fibres are type I whereas the darker fibres are type II.

ALCOHOLIC MYOPATHY AND ACETALDEHYDE

161

(Duane & Peters 1988). The metabolic superimposition of concomitant disease or malnutrition will exacerbate the myopathy (Nicolas et al 2003). Acute myopathy is relatively rare, occurring in less that 1% of alcohol misusers (Martin et al 1985) (Table 1). Despite this, many of the older textbooks on skeletal muscle concentrate on the acute form of the disease. The full spectrum of rhabdomyolysis leading to renal failure can be fatal (Louboutin et al 1995, Hewitt & Winter 1995, Riggs 1998, De Francesco et al 2005). Animal models of alcohol-induced muscle disease In carrying out patient-based studies there are numerous constraints of a practical (such as limitations in obtaining sufficient material for biochemical analysis) or ethical (such as dosing of alcohol misusers with ethanol or drugs which will potentiate the effects of alcohol) nature. Moreover, the coexistence of many pathologies or malnutrition complicates the interpretation of clinical data. The use of animal models with pair-feeding regimens resolves many of these problems, as alcohol can be administered in defined amounts without concomitant nutrient deficiencies: controls are given identical amounts of the same diet in which ethanol is replaced with isoenergetic carbohydrate (for example Ohlendieck et al 2003) or fat (for example Hunter et al 2003). In such models, anatomically distinct skeletal muscles are taken to represent the different muscle fibre types. To investigate the type I fibres, the soleus is examined, whereas the plantaris or gastrocnemius are used to represent the type II fibres (Preedy & Peters 1988a, Preedy et al 1990) (see also Table 2). The suitability of this has been affirmed by studies showing that feeding rats a nutritionally complete liquid diet, containing ethanol as 35% of total dietary energy, causes a marked reduction in the weight of the plantaris compared to those muscles taken from control rats fed identical amounts of the same diet in which ethanol is replaced by isocaloric glucose, whereas the weight of the soleus is relatively unaffected by alcohol feeding (Preedy & Peters 1988b) (Fig. 2). It could be argued that these changes in muscle weight could be due to some hitherto unrecognized phenomena related to the anatomical location of the plantaris muscles. However, we examined the type II fibres within the soleus muscle and showed that the relative abundance of fibres with a small diameter increased, with a corresponding decrease in the proportion of fibres with a larger diameter (Fig. 3). This strongly suggests that the myopathy is related to fibre characteristics rather than the location of the muscle (Preedy et al 1989). The biochemical and physiological differences in the two fibre types are displayed in Table 2. Of potential interest in understanding the aetiology of AIMD disease is the higher activities of antioxidant enzyme systems such as catalase, superoxide dismutase and glutathione peroxidase in the type I fibres when compared to the more susceptible type II fibres.

162

PREEDY ET AL

TABLE 2 Features of type I and II fibres or muscle types

Myoglobin Twitch/Myosin heavy chains Metabolism Glycogen Mitochondria Capillary density Muscle fibre diameter Antioxidant enzymes Superoxide dismutase Glutathione peroxidase Catalase Alpha-tocopherol concentration Imidazole dipeptides Myosin ATPase ALDH2

Red (type I)

White (type II)

High Slow Oxidative-aerobic Low High High Smaller

Low Fast Glycolytic-anaerobic High Low or absent Low Larger

High High High High Low Low High

Low Low Low Low High High Low

Muscle weight (mg)

In this table, the differences between the fibre and muscle types are compared. We have defined ‘muscle types’ as muscle that contains a particular prominence of fibre types. It is very rare for muscles to contain exclusively a single fibre type, and most anatomically distinct muscles contain a mixture of fibre types. Much of the information contained in this table is derived from animals where a muscle with a predominance of a particular fibre type has been dissected and analysed. For example the plantaris, EDL (extensor digitorum longus) or white portions of the gastrocnemius muscle can be taken to represent the white muscle. For red muscle, the soleus or red portions of the gastrocnemius are analysed. Compiled from various sources.

500 400 300 NS 200 100 0 Soleus (Type I)

P<0.001 Control Alcoholic

Plantaris (Type II)

FIG. 2. The effect of chronic ethanol feeding on skeletal muscle weights. Male Wistar rats were fed nutritionally complete liquid diets in which ethanol comprised 35% of total dietary energy. Controls were pair-fed the same diet in which ethanol was replaced by isocaloric glucose. At the end of 6 weeks the rats were killed and soleus (type I fibre predominant) and plantaris (type II fibre predominant) were dissected and weighed. Data are mean ± SEM, n = 6–7. Data from Preedy & Peters (1988b).

ALCOHOLIC MYOPATHY AND ACETALDEHYDE

163

Percent of type II fibres in soleus

Soleus 100

Type I

P<0.05

Type II

80 60

P<0.05

40

Control

20

Alcoholic

0

Small Large (0-39 mm) (40-79 mm)

FIG. 3. The effect of chronic ethanol feeding on the diameter of type II fibres. Control and alcohol-fed rats were fed nutritionally complete liquid diets containing 35% of total energy as glucose (controls) or ethanol (treated). At the end of 6 weeks rats were killed and soleus (a type I fibre predominant muscle which also contains a smaller number of type II fibres) were dissected and stained for myosin ATPase. Data are mean ± SEM (n = 6). The insert shows a section of soleus in which type I and type II fibres have been identified. Data from Preedy et al (1989).

Enzymes responsible for catalysing acetaldehyde formation and oxidation The enzymes responsible for catalysing acetaldehyde formation and removal have been well described (Crabb 1995, Crabb et al 2004). It is possible that, in the genesis of alcoholic myopathy, the muscles that are most susceptible (such as the plantaris) will have greater amounts of the enzymes that generate acetaldehyde (i.e. alcohol dehydrogenase [ADH], microsomal ethanol oxidising system [MEOS], catalase) and less of the enzymes that remove acetaldehyde (mainly aldehyde dehydrogenase [ALDH]). Present work on ADHs in skeletal muscle types is preliminary. We assessed ADH activity in muscle using glycine-NaOH buffer at pH 10.0 by monitoring the increase of absorbance at 340 nm with NAD+ as a coenzyme and ethanol as a substrate. However, there was no measurable ADH activity in either the soleus or the plantaris ( J. Farrés et al unpublished observations). In the same studies, ADH1 and ADH3 were assessed by isoelectrofocusing on polyacrylamide gels followed by activity staining. Using 0.1 M 2-buten-1-ol as a substrate, the data showed that ADH3, a glutathione-dependent formaldehyde dehydrogenase, was present in both soleus and plantaris, whereas by using 0.2 M ethanol, ADH1 was found in soleus but not in plantaris muscle ( J. Farrés et al unpublished observations). In these studies, the activity of ADH in the livers from the same animals was comparable to previously published values from other groups, and there were significant ADH1 and ADH3 activities ( J. Farrés et al unpublished observations), thus

164

PREEDY ET AL

giving overall credibility to the negative data in muscle. These results substantiate much earlier studies, which reported negligible levels of ADH in skeletal muscle (Zahlten et al 1981). In general, studies appear to show that the cytochrome P450s in skeletal muscle are either absent or have low activities, especially compared to liver (Crosbie et al 1997). Nevertheless, their location on the skeletal muscle sarcoplasmic reticulum may in some way be connected to the genesis of the myopathy (Riggs 1998) and may also be related to the generation of protein adducts, which are predominantly located within the sarcolemmal region (Worrall et al 2001). Studies have been carried out on CYP1A1/2 and CYP2E1 in skeletal muscle of rats subjected to alcohol treatment (Smith et al 2000). These studies showed that moderate amounts of CYP1A1/2 were located in skeletal muscle which were markedly increased upon 2 weeks feeding of alcohol using the Lieber-DeCarli pairfeeding protocol (Smith et al 2000). However, the authors ascribed these changes to solubilisation of plastic or metal material from the feeding system, as induction of CYP1A1/2 could not be replicated using all-glass feeding systems. Furthermore the authors failed to detect constitutive CYP2E1 in either soleus or plantaris and there was no significant induction of muscle CYP2E1 after 2 weeks alcohol feeding (Smith et al 2000). Studies by other groups have shown that catalase, which is also capable of oxidising ethanol to acetaldehyde, is higher in the soleus than the plantaris (Riley et al 1988). Although catalase is reported to play a role in ethanol oxidation in the liver, its overall contribution to the conversion of ethanol to acetaldehyde has not been determined. Using Western blotting, we have shown that the concentration of ALDH2 in the soleus was about three-fold higher than in the plantaris, though levels in these two muscles were considerably less than those occurring in liver (Fig. 4) (Fischer et al 2002). Lower concentrations of ALDH2 in the plantaris may reflect the fact that plantaris muscles have fewer mitochondria. Overall, these results show that there is a higher potential for both formation and removal of acetaldehyde in the type I muscles (i.e. soleus). The lower concentration of ALDH2 in the type II muscles (i.e. plantaris) appears to offer some support to the idea that acetaldehyde contributes to the myopathy, in that potentially reduced activity of systems for removing acetaldehyde may lead to increased levels of damaging acetaldehyde following ethanol ingestion. On the other hand, the data on ADH are contrary to our expectations. However, consideration needs to be given to the fact that (I) there are no studies that have systematically examined the acetaldehyde levels per se in these muscle types under different conditions; (II) there are some limitations in the present analysis, it is quite possible that in the absence of mitochondria, the cytosolic form of ALDH, i.e. ALDHA1, may play a prominent role in the plantaris.

ALCOHOLIC MYOPATHY AND ACETALDEHYDE

165

A Soleus

Plantaris

Liver

B ALDH 2 protein (AU x10-6 per 0.1 mg protein)

P<0.001 P<0.001

20 15 10

P<0.001

5 0

Liver

Soleus (Type I)

Plantaris (Type II)

FIG. 4. Western blot of ALDH2 abundance in liver, soleus and plantaris. Skeletal muscle and liver proteins were analysed using Western blot-FluorImager analysis. Data are mean ± SEM (n = 8). Differences between means were assessed by ANOVA. Data from Fisher et al (2002).

Protein adducts The importance of protein adduct formation in the aetiology of alcoholic organ damage has been described and reviewed recently (Freeman et al 2005, Worall & Thiele 2005). Acetaldehyde can cause protein adduct formation in two ways. The first pertains to adduction of acetaldehyde to proteins in either reducing or nonreducing conditions (i.e. acetaldehyde-protein adducts). The second pertains to indirect mechanisms where oxidative stress, arising as a consequence of raised acetaldehyde or acetaldehyde metabolism, will initiate the formation of volatile species such as those derived from lipid peroxidation (i.e. malondialdehyde–, 4-hydroxy-nonenal– and malondialdehyde–acetaldehyde–protein adducts) (Freeman et al 2005). There are also other adduct species that arise in alcohol toxicity such as hydroxyethyl radical–protein adducts (Freeman et al 2005).

166

PREEDY ET AL

Acetaldehyde-protein adducts (arbitrary scores)

The importance of adduct formation pertains to the fact that (I) adduction may impair the functional activity of the proteins, (II) adducts may cause autoimmune responses via the formation of neoantigens and (III) adducts may be toxic to cells (Freeman et al 2005, Worall & Thiele 2005). In a model of chronic alcoholic myopathy, we analysed skeletal muscles from rats fed nutritionally complete liquid diets containing ethanol as 35% of total dietary energy; control rats were fed the same diet in which ethanol was replaced by isocaloric glucose. ELISA analyses for protein adducts showed increased amounts of unreduced-acetaldehyde adducts in soleus (P < 0.025) and plantaris (P < 0.025) muscles (Fig. 5B) (Worrall et al 2001). Immunohistochemical analysis using an antiserum reacting against both reduced and unreduced acetaldehyde adducts also showed that adducts were increased in soleus (P < 0.05) and plantaris (P < 0.025), confirming the ELISA analysis (Fig. 5A) (Worrall et al 2001). Adducts were prefer-

2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

P<0.05

A

P<0.025

Control Alcoholic

Soleus Type I

Plantaris Type II

Acetaldehyde-protein adducts (unreduced) (Absorbance at 405nm)

B 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05

P<0.025

P<0.025 Control Alcoholic

Soleus Type I

Plantaris Type II

FIG. 5. Quantification of acetaldehyde–protein adducts assessed by immunohistochemistry and ELISA. Male Wistar rats were fed nutritionally complete liquid diets in which 35% of total dietary energy was provided by glucose or ethanol. At the end of 6 weeks rats were killed and soleus (type I fibre predominant) and plantaris (type II fibre predominant) muscles were dissected and acetaldehyde–protein adducts were quantified by immunohistochemistry (A) or ELISA (B). Data are mean ± SEM (n = 6). P values relate to differences between control and alcohol-fed rats. Data from Worrall et al (2001).

ALCOHOLIC MYOPATHY AND ACETALDEHYDE

167

entially located within the sarcolemmal and subsarcolemmal (i.e. muscle membrane) regions (Worrall et al 2001) (Fig. 6). In these chronic studies both plantaris and soleus behaved similarly in their response to chronic alcohol feeding (Fig. 5) (Worrall et al 2001). This does not mean that there are no differences between the soleus and the plantaris in all metabolic and biochemical responses to alcohol toxicity. Rather, differential effects on muscle may occur in the final or rate limiting stages in the development of the myopathy. For example, it is quite possible that differences between the soleus and the plantaris will be seen in the rates of protein degradation or synthesis (see below). Thus, protein adduct formation could be considered a ‘priming reaction’. We also investigated whether acetaldehyde-derived adducts are formed in muscle of rats as a result of acute exposure to ethanol and acetaldehyde. Treatment of rats with ethanol and cyanamide + ethanol for 2.5 hours increased the amount of aldehyde-derived protein adducts in both soleus and plantaris muscle (Fig. 7) (Patel et al 2005). However, polyclonal antibodies are subject to the criticism that they lack specificity as they are composed of multiple antibody species and recognise, within the antigen, dissimilar epitopes (the domain on the protein). Monoclonal

Control

Alcoholic

Soleus

Soleus

Plantaris

Plantaris

FIG. 6. Immunohistochemical staining of acetaldehyde–protein adducts in soleus and plantaris muscles of control and ethanol-fed rats. Acetaldehyde–protein adducts in soleus and plantaris muscles from the same studies described in Fig. 5. Notice heavier staining around the sarcolemmal and sub-sarcolemmal regions. Quantification of adduct density is summarized in Fig. 5. Bar = 50 µm. Reproduced from Worrall et al (2001). Protein adducts in type I and type II fibre predominant muscles of the ethanol-fed rat: Preferential localisation in the sarcolemmal and subsarcolemmal region. Eur J Clin Invest 31:723–730. Copyright 2001, Blackwell Publishing Ltd.

PREEDY ET AL

Acetaldehyde-protein adducts (unreduced) (Absorbance at 405nm)

Acetaldehyde-protein adducts (unreduced) (Absorbance at 405nm)

168 Soleus 2 way ANOVA

0.75

Treatments Cyanamide Ethanol Interaction

0.5

P 0.020 <0.001 0.003

0.25 0

[A]

[B]

Saline +Saline

[C]

Cyanamide Saline +Saline +Ethanol

[D] Cyanamide +Ethanol

Plantaris 2 way ANOVA

0.75 0.5

P 0.005 <0.001 NS

Treatments Cyanamide Ethanol Interaction

0.25 0

[A] Saline +Saline

[B]

[C]

Cyanamide Saline +Saline +Ethanol

[D] Cyanamide +Ethanol

FIG. 7. Acetaldehyde–protein adducts in muscles of rats treated acutely with ethanol with or without cyanamide. The experimental protocol was divided into two parts, namely: a pre-treatment stage for 30 min followed by a treatment stage of 2 h 30 min. Thus, the groups were (pre-treatment + treatment): [A] Saline + saline; [B] Saline + ethanol; [C] Cyanamide + saline; [D] Cyanamide + ethanol. In the pre-treatment stage, rats were injected intraperitoneally with (0.5 ml/100 g body weight) of either saline or cyanamide. At 30 min the rats were injected in the treatment stage, with an intraperitoneal injection (1 ml/100 g body weight) of either saline or ethanol. After a further 2 h 30 min the rats were killed, soleus and plantaris muscles were dissected and acetaldehyde– protein adducts were measured by ELISA. The following doses were used: saline, 0.15 mol of NaCl/l; ethanol, 75 mmol/kg body weight; cyanamide, 0.50 mmol/kg body weight. Data are optical density units per 10 ng protein, shown as mean ±SEM of n = 7–8 observations in each group. Data from Patel et al (2005).

antibodies, on the other hand, have high specificity and homogeneity and recognise the same epitope species. Use of a polyclonal antibody and the RT1.1 mouse monoclonal antibody showed that the acetaldehyde adducts were unreduced (Patel et al 2005). As with the chronic studies we could not confirm greater sensitivity of the plantaris (Patel et al 2005), although, paradoxically, greater sensitivity of this muscle in acetaldehyde adduct formation was apparent when data were analysed using

ALCOHOLIC MYOPATHY AND ACETALDEHYDE

169

immunohistochemistry (Niemela et al 2002). In interpreting this, consideration should be given to the fact that both methods of quantifying adducts have their advantages and limitations. For example, ELISA methods utilize objective assessments of adducts, though there is no identification of regions affected by adduction. Immunohistochemical methods are more subjective, but have the advantage of identifying affected regions such as their preferential location around the sarcolemmal regions of muscle. Protein degradation The concept of protein turnover implies that the reductions in muscle mass must be due to changes in the rates of protein breakdown and/or protein synthesis. With regard to protein breakdown, there is a considerable body of evidence to show that this process decreases in alcoholic myopathy. For example, urinary 3-methylhistidine excretion (a marker of contractile protein degradation) decreases in myopathic alcoholics (Martin & Peters 1985). The rate of protein degradation in rats subjected to pair-feeding with the Lieber-DeCarli protocol has also been shown to be reduced (Preedy & Peters 1989). A role of acetaldehyde is also implicated as the addition of acetaldehyde to human muscle preparations in vitro reduced the activities of a wide spectrum of lysosomal and non-lysosomal proteases (Mantle et al 1999) (Table 3). The physiological significance of this is rather unclear as the

TABLE 3 Effect of acetaldehyde on protease activities Effect of acetaldehyde (% change)

Cytoplasmic Alanyl aminopeptidase Arginyl aminopeptidase Diaminopeptidase IV Triaminopeptidase Proline endopeptidase Lysosomal Diaminopeptidase I Diaminopeptidase II Cathepsin B Cathepsin H Cathepsin L

17 mmol/l

170 mmol/l

−63* −72* −11 −39* −1

−81* −90* −44* −71* −34*

−39* −34* 0 −64* −1

−74* −56* +1 −79* −3

Homogenates of human muscle were incubated in vitro with acetaldehyde. The minus prefix indicates a reduction in mean enzyme activity, whereas a positive prefix indicates an increase. *, P < 0.05 in comparison with control preparations without acetaldehyde. Data adapted from Mantle et al (1999).

170

PREEDY ET AL

inhibitory effects of acetaldehyde in vitro were only achieved with high supraphysiological levels of acetaldehyde. Acetaldehyde in concentrations of 1.7 mmol/l reduced arginyl aminopeptidase activity by 50% (P < 0.05) (Mantle et al 1999). When acetaldehyde concentrations were increased to 17 and 170 mmol inhibitory effects were observed for a variety of proteolytic enzymes (Mantle et al 1999) (Table 3). However, there was no effect of chronic 6 week alcohol feeding in vivo on a variety of proteases in rat muscle, including the lysosomal cathepsins B, D, H and L, dipeptidyl aminopeptidases I and II, non-lysosomal alanyl aminopeptidase, arginyl aminopeptidase, leucyl aminopeptidase, dipeptidyl aminopeptidase IV, tripeptidyl aminopeptidase and proline endopeptidase as well as enzymes of the Ca2+dependent calpastatin–calpain system (Koll et al 2002). When an attempt was made to increase acetaldehyde levels in vivo with cyanamide + ethanol, there was similarly no effect of acetaldehyde (Koll et al 2002). These observations are at variance with the in vitro studies described above but consideration needs to be given to several factors: (I) artificial substrates in the in vitro assay do not actually reflect the constituent proteins in vivo and there may be localized co-factors which regulate proteolysis; (II) in in vitro systems optimal assay conditions are achieved in terms of enzyme– substrate relationships and presence of co-factors and these conditions are not necessarily achieved in vivo; and (III) it is possible that adduction may influence proteolysis. For example, acetaldehyde-adducted proteins appear to be more stable and have reduced rates of protein degradation (Nicholls et al 1994). Finally, the use of artificially high levels of acetaldehyde in in vitro systems does not readily match in vivo levels (Oba et al 2005). Thus there is a pressing need to elucidate the effects of acetaldehyde in intact animal models or human beings if possible. Protein synthesis We have shown, using stable isotopes, that alcoholic myopathy in human is characterized by a decrease in protein synthesis (Pacy et al 1991). Reductions in protein synthesis have also been reproduced in laboratory animals. For example, we have shown that in acute ethanol dosage there is a decrease in the fractional rate of protein synthesis, particularly in type II fibre-predominant muscles such as the plantaris. In the soleus, reductions in protein synthesis were less marked (Preedy & Peters 1988a). Changes in protein synthesis have been ascribed principally to defects in translation (Lang et al 2003, 2004, 2005). For example, ethanol causes impairment of initiation factor availability (i.e. eIF4E) in skeletal muscle, including a decrease in the active eIF4EeIF4G complex and an increase in the inactive eIF4E4E-BP1 complex (reviewed in Lang et al 2001, 2005). The phosphorylation of S6 kinase 1 and 4E-BP1 are also reduced (Lang et al 2004).

ALCOHOLIC MYOPATHY AND ACETALDEHYDE

171

Impaired amino acid availability is not thought to play a role per se (Parry-Billings et al 1992). The caveat to this is the possibility that the response of skeletal muscle to amino acid stimulation of protein synthesis may be impaired, i.e. an amino acid resistance (Lang et al 2003). Certainly, the fasting–feeding response of muscle protein synthesis is impaired in the presence of ethanol (Sneddon et al 2003). In follow up studies we injected rats with ethanol, with or without inhibitors of ADH (4-methylpyrazole) or ALDH (cyanamide) (Preedy et al 1992). Fractional rates of protein synthesis (ks, the percentage of the protein pool renewed each day) in muscle were reduced by ethanol by approximately 25%. Pre-treatment of ethanol-dosed rats with 4-methylpyrazole reduced protein synthesis rates by approximately 35% (Preedy et al 1992) (Fig. 8). Much greater effects were seen with pre-treatment of ethanol-dosed rats with cyanamide, which reduced ks by approximately 65% (Preedy et al 1992) (Fig. 8). The exacerbation of ethanol’s effects by 4-methylpyrazole was somewhat surprising and may be related to the fact that the 4-methylpyrazole pre-treatment raised ethanol levels (Preedy et al 1992). Greater insights into the contribution of acetaldehyde in the role of ethanolinduced impairment of protein synthesis have been reported by others (Lang et al 2004). These studies show that ethanol impairs insulin-like growth factor 1 (IGF1) signalling via IGF1-induced phosphorylation of S6K1 and ribosomal protein S6. This effect was not modulated by 4-methylpyrazole treatment, leading to the conclusion that the aforementioned changes occur independently of alcohol oxidation.

Protein synthesis rate, ks (%/day)

25 20

-18%

-27% -45%

15

-68%

10 5

5

0 Control

Ethanol

4MP+ ethanol

Cyanamide Acetaldehyde +ethanol

FIG. 8. Summary of changes in protein synthesis in vivo due to ethanol, inhibitors of alcohol and aldehyde dehydrogenases and acetaldehyde. For experimental details of pre-treatment and treatments see legend to Fig. 7. The dose of 4-methylpyrazole was identical to that of cyanamide. For acetaldehyde, rats were injected with acetaldehyde alone at a dose of 2.8 mmol/kg body weight. Fractional rates of protein synthesis were measured with tritiated phenylalanine. Data from Preedy et al (1992).

172

PREEDY ET AL

In other words, the data suggest that acetaldehyde is not necessarily needed for these changes to occur (Lang et al 2004). However, in the aforementioned studies, no corresponding studies were made using cyanamide to block acetaldehyde oxidation and there is also limited data on acetaldehyde levels per se when using 4methylpyrazole treatments. Although we have used 4-methylpyrazole ourselves, we now believe that some caution is needed in its usage and interpretation. This is because acetaldehyde levels in skeletal muscle are not altered 2.5 hours after acute ethanol dosage (75 mmol/kg body weight) though they are significantly increased in blood and liver (Fig. 9). Indeed, mean acetaldehyde concentrations in muscle of control rats are much higher than levels seen in blood or liver, the cause of which is unknown (Adachi et al 2001). Use of 4-methylpyrazole is also problematical as it induces CYP2E1 in tissues such as liver and kidney, though there are no corresponding studies in muscle (Wu & Cederbaum 1993, 1994). However, induction of CYP2E1 by 4-methylpyrazole could also explain why some studies appear to show that 4methylpyrazole exacerbates the metabolic effect of ethanol in muscle (for example see Xu et al 1993). Acetaldehyde alone at a dose of 3 mmol/kg body weight reduces protein synthesis rates in vivo by approximately 15%, supporting the contention that it is a protein synthetic perturbant (Preedy et al 1992) (Fig. 8). The role of acetaldehyde in reducing protein synthesis in vitro has also been shown by other groups using 0.2 mM acetaldehyde (Hong-Brown et al 2001). Overall, the data suggest that both ethanol and acetaldehyde may, possibly independently, mediate the ethanol-induced inhibition of skeletal muscle protein synthesis. Conclusions Alcohol induced muscle disease is a multifaceted processes, affecting a number of key regulatory mechanisms and compounded by environmental factors such as genetics, nutrition and the coexistence of other diseases that so commonly occur in alcoholics. Although not fatal per se, it is about five times more common than cirrhosis and contributes significantly to the enhanced morbidity seen in alcoholic patients. There is a body of evidence to show the involvement of acetaldehyde, but at present the precise role or steps where acetaldehyde acts is unresolved. Acknowledgements We are indebted to Mr Sergio Porté for performing the activity assays and isoelectrofocusing experiments with ADH in the laboratory of Drs Xavier Parés and Jaume Farrés (Université Autònoma de Barcelona, E-08193 Bellaterra, Barcelona, Spain).

Acetaldehyde (nmol/g)

Acetaldehyde (nmol/g)

Acetaldehyde (nmol/ml)

ALCOHOLIC MYOPATHY AND ACETALDEHYDE

173

Blood 50

P<0.001

40 30 20 10 0

NaCl (Control)

Ethanol

Muscle 100

NS

80 60 40 20 0

NaCl (Control)

Ethanol

Liver 120 100

P<0.001

80 60 40 20 0

NaCl (Control)

Ethanol

FIG. 9. Acetaldehyde levels in different tissue compartments of the rat and in response to ethanol. Male Wistar rats were injected intraperitoneally with ethanol at a dose of 75 mmol/kg body weight and sacrificed after 2.5 hours. Controls were injected with 0.15 mol/l NaCl. Data are mean ± SEM. Acetaldehyde was measured with 2,4-dinitrophenylhydrazine-1,3indanedione-1-azine (DIH). For determination of muscle acetaldehyde, the combined gastrocnemius and plantaris was rapidly dissected and immediately homogenized in DIH trapping agent. The combined analysis of both the gastrocnemius and plantaris was necessary in order to minimize the period between death and suspension of tissue homogenate in DIH. Adapted from Adachi et al (2001).

174

PREEDY ET AL

References Aagaard NK, Andersen H, Vilstrup H, Clausen T, Jakobsen J, Dorup I 2003 Decreased muscle strength and contents of Mg and Na,K-pumps in chronic alcoholics occur independently of liver cirrhosis. J Intern Med 253:359–366 Adachi J, Asano M, Ueno Y et al 2001 Acute effect of ethanol on 7-hydroperoxycholesterol in muscle and liver. Lipids 36:267–271 Adachi J, Asano M, Ueno Y et al 2003 Alcoholic muscle disease and biomembrane perturbations. J Nutr Biochem 14:616–625 Arno MJ, Lin CI, Wiseman H, Preedy VR 2007 Use of microarrays to determine the effects of ethanol on gene expression in skeletal muscle. Alcohol Clin Exp Res Suppl [In press] Crabb DW 1995 Ethanol oxidizing enzymes: roles in alcohol metabolism and alcoholic liver disease. Prog Liver Dis 13:151–172 Crabb DW, Matsumoto M, Chang D, You M 2004 Overview of the role of alcohol dehydrogenase and aldehyde dehydrogenase and their variants in the genesis of alcohol-related pathology. Proc Nutr Soc 63:49–63 Crosbie SJ, Blain PG, Williams FM 1997 Metabolism of n-hexane by rat liver and extrahepatic tissues and the effect of cytochrome P-450 inducers. Hum Exp Toxicol 16: 131–137 De Francesco DE, Bezerra dS Jr, Menezes BD, Borges PL, Pinheiro BG 2005 Rhabdomyolysis and acute renal failure after strenuous exercise and alcohol abuse: Case report and literature review. Sao Paulo Med J 123:33–37 Duane P, Peters TJ 1988 Nutritional status in alcoholics with and without chronic skeletal muscle myopathy. Alcohol Alcohol 23:271–277 Fernandez-Sola J, Nicolas JM, Estruch R, Sacanella E, Grau JM, Urbano-Marquez A 2005 Alcoholic myopathy: clinical aspects. In: Preedy VR, Watson RR (eds) Comprehensive handbook of alcohol-related pathology. Academic Press, London p 671–692 Fischer M, Crabb DW, You M, Hunter RJ, Martin CR, Preedy VR 2002 Decreased mitochondrial aldehyde dehydrogenase (ALDH2) level in Type II fibre-predominant muscle: a possible factor in alcoholic myopathy. Alcohol Clin Exp Res 26, Suppl:186A Freeman TL, Tuma DJ, Thiele GM et al 2005 Recent advances in alcohol-induced adduct formation. Alcohol Clin Exp Res 29:1310–1316 Hewitt SM, Winter RJ 1995 Rhabdomyolysis following acute alcohol intoxication. J Accid Emerg Med 12:143–144 Hong-Brown LQ, Frost RA, Lang CH 2001 Alcohol impairs protein synthesis and degradation in cultured skeletal muscle cells. Alcohol Clin Exp Res 25:1373–1382 Hunter RJ, Neagoe C, Jarvelainen HA et al 2003 Alcohol affects the skeletal muscle proteins, titin and nebulin in male and female rats. J Nutr 133:1154–1157 Khan AR 1981 Influence of ethanol and acetaldehyde on electro-mechanical coupling of skeletal muscle fibers. Acta Physiol Scand 111:425–430 Koll M, Ahmed S, Mantle D et al 2002 Effect of acute and chronic alcohol treatment and their superimposition on lysosomal, cytoplasmic, and proteosomal protease activities in rat skeletal muscle in vivo. Metabolism 51:97–104 Lang CH, Kimball SR, Frost RA, Vary TC 2001 Alcohol myopathy: Impairment of protein synthesis and translation initiation. Int J Biochem Cell Biol 33:457–473 Lang CH, Frost RA, Deshpande N et al 2003 Alcohol impairs leucine-mediated phosphorylation of 4E-BP1, S6K1, eIF4G, and mTOR in skeletal muscle. Am J Physiol Endocrinol Metab 285: E1205–1215

ALCOHOLIC MYOPATHY AND ACETALDEHYDE

175

Lang CH, Pruznak AM, Deshpande N, Palopoli MM, Frost RA, Vary TC 2004 Alcohol intoxication impairs phosphorylation of S6K1 and S6 in skeletal muscle independently of ethanol metabolism. Alcohol Clin Exp Res 28:1758–1767 Lang CH, Frost RA, Summer AD, Vary TC 2005 Molecular mechanisms responsible for alcoholinduced myopathy in skeletal muscle and heart. Int J Biochem Cell Biol 37:2180–2195 Louboutin JP, Nataf S, Jardel-Boulssiere J et al 1995 Evidence for low production of lactate and pyruvate in alcoholic rhabdomyolysis. Muscle Nerve 18:784–785 Mantle D, Falkous G, Peters TJ, Preedy VR 1999 Effect of ethanol and acetaldehyde on intracellular protease activities in human liver, brain and muscle tissues in vitro. Clin Chim Acta 281:101–108 Martin FC, Peters TJ 1985 Assessment in vitro and in vivo of muscle degradation in chronic skeletal muscle myopathy of alcoholism. Clin Sci 68:693–700 Martin F, Ward K, Slavin G, Levi J, Peters TJ 1985 Alcoholic skeletal myopathy, a clinical and pathological study. Q J Med 55:233–251 Mills KR, Ward K, Martin F, Peters TJ 1986 Peripheral neuropathy and myopathy in chronic alcoholism. Alcohol Alcohol 21:357–362 Nakahara T, Hashimoto K, Hirano M, Koll M, Martin CR, Preedy VR 2003 Acute and chronic effects of alcohol exposure on skeletal muscle c-myc, p53, and Bcl-2 mRNA expression. Am J Physiol Endocrinol Metab 285:E1273–1281 Nicholls RM, Fowles LF, Worrall S, de Jersey J, Wilce PA 1994 Distribution and turnover of acetaldehyde-modified proteins in liver and blood of ethanol-fed rats. Alcohol Alcohol 29:149–157 Nicolas JM, Garcia G, Fatjo F et al 2003 Influence of nutritional status on alcoholic myopathy. Am J Clin Nutr 78:326–333 Niemela O, Parkkila S, Koll M, Preedy VR 2002 Generation of protein adducts with malondialdehyde and acetaldehyde in muscles with predominantly type I or type II fibers in rats exposed to ethanol and the acetaldehyde dehydrogenase inhibitor cyanamide. Am J Clin Nutr 76: 668–674 Oba T, Maeno Y 2004 Acetaldehyde alters Ca2+ release channel gating and muscle contraction in a dose-dependent manner. Am J Physiol Cell Physiol 286:C1188–1194 Oba T, Maeno Y, Ishida K 2005 Differential contribution of clinical amounts of acetaldehyde to skeletal and cardiac muscle dysfunction in alcoholic myopathy. Curr Pharm Des 11: 791–800 Ohlendieck K, Harmon S, Koll M, Paice AG, Preedy VR 2003 Ca2+-regulatory muscle proteins in the alcohol-fed rat. Metabolism 52:1102–1112 Pacy PJ, Preedy VR, Peters TJ, Read M, Halliday D 1991 The effect of chronic alcohol ingestion on whole body and muscle protein synthesis—a stable isotope study. Alcohol Alcohol 26: 505–513 Parry-Billings M, Preedy VR, Opara E, Liu CT, Newsholme EA 1992 Acute ethanol administration and the metabolism of glutamine by skeletal muscle of the rat: implications for ethanolinduced reductions in protein synthesis. Alcohol Alcohol 27:613–618 Patel VB, Worall S, Emery PW, Preedy VR 2005 Protein adduct species in muscle and liver of rats following acute ethanol administration. Alcohol Alcohol 40:485–493 Patussi V, Mezzani L, Scafato E 2005 An overview of the pathologies occurring in alcohol abusers. In: Preedy VR, Watson RR, (eds) Comprehensive handbook of alcohol-related pathology. Academic Press, London p 253–260 Preedy VR, Peters TJ 1988a Acute effects of ethanol on protein synthesis in different muscles and muscle protein fractions of the rat. Clin Sci 74:461–466 Preedy VR, Peters TJ 1988b The effect of chronic ethanol ingestion on protein metabolism in type-I- and type-II-fibre-rich skeletal muscles of the rat. Biochem J 254:631–639

176

PREEDY ET AL

Preedy VR, Peters TJ 1989 The effect of chronic ethanol ingestion on synthesis and degradation of soluble, contractile and stromal protein fractions of skeletal muscles from immature and mature rats. Biochem J 259:261–266 Preedy VR, Watson RR 2005 Comprehensive handbook of alcohol-related pathology: volumes 1–3. Academic Press, London Preedy VR, Bateman CJ, Salisbury JR, Price AB, Peters TJ 1989 Ethanol-induced skeletal muscle myopathy: biochemical and histochemical measurements on Type I and Type II fibre-rich muscles in the young rat. Alcohol Alcohol 24:533–539 Preedy VR, Gove CD, Panos MZ et al 1990 Liver histology, blood biochemistry and RNA, DNA and subcellular protein composition of various skeletal muscles of rats with experimental cirrhosis: implications for alcoholic muscle disease. Alcohol Alcohol 25:641– 649 Preedy VR, Keating JW, Peters TJ 1992 The acute effects of ethanol and acetaldehyde on rates of protein synthesis in type I and type II fibre-rich skeletal muscles of the rat. Alcohol Alcohol 27:241–251 Preedy VR, Adachi J, Ueno Y et al 2001a Alcoholic skeletal muscle myopathy: definitions, features, contribution of neuropathy, impact and diagnosis. Eur J Neurol 8:677–687 Preedy VR, Paice A, Mantle D, Dhillon AS, Palmer TN, Peters TJ 2001b Alcoholic myopathy: Biochemical mechanisms. Drug Alcohol Depend 63:199–205 Preedy VR, Adachi J, Asano M et al 2002 Free radicals in alcoholic myopathy: Indices of damage and preventive studies. Free Radic Biol Med 32:683–687 Preedy VR, Ohlendieck K, Adachi J et al 2003 The importance of alcohol-induced muscle disease. J Muscle Res Cell Motil 24:55–63 Puszkin S, Rubin E 1975 Adenosine diphosphate effect on contractility of human muscle actomyosin: inhibition by ethanol and acetaldehyde. Science 188:1319–1320 Riggs JE 1998 Alcohol-associated rhabdomyolysis: ethanol induction of cytochrome P450 may potentiate myotoxicity. Clin Neuropharmacol 21:363–364 Riley DA, Ellis S, Bain JL 1988 Catalase-positive microperoxisomes in rat soleus and extensor digitorum longus muscle fiber types. J Histochem Cytochem 36:633–637 Smith C, Stamm SC, Riggs JE et al 2000 Ethanol-mediated CYP1A1/2 induction in rat skeletal muscle tissue. Exp Mol Pathol 69:223–232 Sneddon AA, Koll M, Wallace MC et al 2003 Acute alcohol administration inhibits the refeeding response after starvation in rat skeletal muscle. Am J Physiol Endocrinol Metab 284: E874–882 Urbano-Marquez A, Fernandez-Sola J 2004 Effects of alcohol on skeletal and cardiac muscle. Muscle Nerve 30:689–707 Urbano-Marquez A, Estruch R, Navarro-Lopez F, Grau JM, Mont L, Rubin E 1989 The effects of alcoholism on skeletal and cardiac muscle. N Engl J Med 320:409–415 Worall S, Thiele GM 2005 Modification of proteins by reactive ethanol metabolites: adduct structure, functional and pathological consequences. In: Preedy VR, Watson RR (eds) Comprehensive handbook of alcohol-related pathology. Academic Press, London p 1209– 1222 Worrall S, Niemela O, Parkkila S, Peters TJ, Preedy VR 2001 Protein adducts in type I and type II fibre predominant muscles of the ethanol-fed rat: Preferential localisation in the sarcolemmal and subsarcolemmal region. Eur J Clin Invest 31:723–730 Wu D, Cederbaum AI 1993 Induction of liver cytochrome P4502E1 by pyrazole and 4methylpyrazole in neonatal rats. J Pharmacol Exp Ther 264:1468–1473 Wu D, Cederbaum AI 1994 Characterization of pyrazole and 4-methylpyrazole induction of cytochrome P4502E1 in rat kidney. J Pharmacol Exp Ther 270:407–413 Xu D, Heng JK, Palmer TN 1993 The mechanism(s) of the alcohol-induced impairment in glycogen synthesis in oxidative skeletal muscles. Biochem Mol Biol Int 30:169–176

ALCOHOLIC MYOPATHY AND ACETALDEHYDE

177

Zahlten RN, Nejtek ME, Jacobson JC 1981 Ethanol metabolism in guinea pig: In vivo ethanol elimination, alcohol dehydrogenase distribution, and subcellular localization of acetaldehyde dehydrogenase in liver. Arch Biochem Biophys 207:371–379

DISCUSSION Eriksson: It may be a naïve question, but human alcoholics or animals chronically treated with alcohol might have weakened muscles because they don’t use them very much. How do you deal with this potential confounding factor? Preedy: That’s a good question. There are two answers. The first relates to the specificity of the fibre type. If you immobilize the muscles you should get type I atrophy. Some studies show that inactivity causes a type II atrophy, but this could be caused by malnutrition. We have also taken these animals and exercised them, and this doesn’t prevent the myopathy. We have made the rats swim, and they just eat more. Eriksson: It is well known in sport physiology that you can’t create muscles while you are drinking alcohol. Preedy: I agree that it is important to look at activity. There is a study which looked at exhaustive training (Vila et al 2001) and showed that this could reverse or prevent the alcohol-induced myopathy, but they never really looked at the imposition of feeding and diet. Albano: How much do inflammatory reactions contribute in the pathogenesis of alcoholic myopathy in humans? It is known that alcoholics have high levels of circulating TNFα, and this is a major signal for muscle fibre catabolism. Have you thought to explore alcohol myopathy in the model of intragastric alcohol administration devised by Tzukamoto and French ( Tsukamoto et al 1985)? According to the data originating from the late Ron Thurman’s lab circulating endotoxins and pro-inflammatory cytokines are high in these alcohol-fed animals ( Thurman 1998). It would be interesting to see whether myopathy is worse in these animals? Preedy: It is possible. We have looked at the effects of lipopolysaccharide (LPS) on protein synthesis, and compared this with LPS administration with alcohol. When we gave LPS and alcohol, and looked at protein synthesis, the LPS ameliorated some of the effects of alcohol. We think this may be a thermic effect. We haven’t looked at TNFα in this particular model. M Salaspuro: It isn’t so easy to diagnose alcoholic myopathy. Peripheral neuropathy may induce similar symptoms. Therefore, in most cases you need a muscle biopsy or EMG to diagnose myopathy. It may be that less severe cases escape diagnosis. Do you have any data with regard to the incidence of this disease in different populations? Preedy: That’s a good question. The incidence in the UK has been studied by Timothy Peters (Martin et al 1985) who looked at about 150 alcoholics and found that

178

PREEDY ET AL

two-thirds were myopathic as determined by fibre diameter reduction. UrbanoMarquez’s group showed (Urbano-Marquez et al 1989) that the incidence was about 50%. I have also seen reports on myopathy and a reduction in type II fibres from India, Canary Islands, Denmark and Taiwan, so it seems to occur everywhere. In terms of the ALDH2 polymorphism, we put in a grant to the Royal Society last year to look at this. Our hypothesis was that raised acetaldehyde levels would perturb muscle protein synthesis. Unfortunately this grant was turned down. We believe that skeletal muscle myopathy is a feature of a disease affecting all muscle types. Seitz: The muscle has a lot of mitochondria. Is there any evidence for acetaldehyde interfering with ATP generation or causing apoptosis from the mitochondria? Preedy: Good question. In extreme cases—patients with severe alcoholism—the type I muscle fibres start to go and we see mitochondrial involvement. One study (Martin et al 1984) looked at mitochondrial enzymes in controls and alcoholics and found no change. The NMR spectroscopy study with Dr Jules Griffin (University of Cambridge) showed changes in a mitochondrial TCA cycle intermediate in alcohol exposed muscle, which suggests there is a defect in one of the key enzymes which is preferentially located in the mitochondria, so I need to look at this a bit more. In terms of apoptosis, we have looked at two models using different techniques. We used acute dosage up to 24 h with ethanol, and the Lieber-DeCarli regimen for 6 weeks. We found no apoptosis in either case. There are increases in preapoptotic genes, such as c-Myc. Recent Affymetrix data suggest that there are changes in both the preapoptotic and anti-apoptotic pathways. Emery: The point is that the mitochondria are mainly in the type I fibres. Presumably they are used to generating ROS and have antioxidant defences, which could be why they are more protected. Preedy: When we looked at the ALDH2, all we were looking at was the relative amounts of mitochondria which are higher in type I muscle. Apte: The early stages of alcohol-induced liver disease are reversible if you stop drinking. Is this also the case for myopathy? Preedy: We have always had plans to look at the reversibility of alcoholic myopathy. Lang’s group have shown that it is reversible (Vary et al 2004) in terms of the signalling processes. Having said this, Urbano-Marquez’s group in Spain have looked at the muscle strength, and they showed that five years after abstinence, while muscle strength improved in many there was a group of individuals who never attain the normal range (Estruch et al 1998). Shukla: Acetaldehyde seems to be the evil chemical in our discussions, thus far. But it also has cardioprotective effects, and there is good evidence that it is somehow involved in reconditioning. Could some of the observations you have made be related to its protective effect, especially at a level of moderate drinking? Preedy: That is a good question. In our animal model we haven’t carried out low alcohol feeding. There are rapid changes in response to alcohol. Urbano-Marquez’s

ALCOHOLIC MYOPATHY AND ACETALDEHYDE

179

group have some nice data looking at cumulative intake of alcohol. They concluded there was an inverse correlation: the more you drink over your lifetime, the greater the decrease in muscle strength. There was no evidence for a biphasic curve. Shukla: There is evidence that resveratrol or other components of wine are cardioprotective. But there is also evidence that ethanol itself also has cardioprotective effects. Preedy: Yes, but in some ways this is like the heat shock protein story. Many people have shown that it is increased in liver, but we showed that heat shock protein is decreased in chronic ethanol-exposed skeletal muscle, so it has a peculiar response. That is, muscle doesn’t behave like other tissues such as the liver or heart in terms of its response to either alcohol or acetaldehyde. Shukla: Is there any situation where acetaldehyde can be a beneficial molecule? Worrall: There are some data from the Loughborough group that it stops AGE formation (Al-Abed et al 1999). It intercepts part the way down the pathway leading to the very complex adducts (AGEs). Preedy: There were some studies looking at adduct formation and protein turnover (Nicholls et al 1994). Didn’t you show that adduct formation didn’t reduce the protein degradation? Worrall: It depends where you look. In cells it increases the half-life by about 10fold. If it is in the circulation it decreases it by about 50%. Emery: How is the low dose of alcohol supposed to protect the cardiovascular system? Shukla: Many hypotheses have been proposed. One is the involvement of PKCε. Perhaps some of the other kinases may be involved. It is an area which needs to be worked out. Thornalley: There was a controversial paper about how alcohol consumption could decrease AGEs by Tony Cerami and colleagues (Al-Abed et al 1999). It claimed to explain the French paradox of red wine decreasing cardiovascular disease risk. They suggested that acetaldehyde was binding to amino groups in proteins that would otherwise become glycated by glucose. Acetaldehyde was preventing the glucosederived adduct formation. Emery: Is the idea that acetaldehyde-conjugated proteins are less damaging than glycated proteins? Thornalley: There may be mechanisms linked to alcohol intoxication that actually increase protein damage by other routes. What is the chemical nature of the adducts that you are measuring here? Preedy: Simon Worrall has more information on this. Worrall: The lack of highly chemically-defined adducts is a major problem in the field. There are some which are well defined, such as malondialdehyde–acetaldehyde adducts (MAA; especially the dihydropyridine type), N-ethylated amino groups and thiazolidine derivatives. Generally we tend to work with those which are formed

180

PREEDY ET AL

under non-reducing or reducing conditions, as well as those generated by things like malondialdehyde or hydroxyethyl radical-type adducts. Preedy: In the muscle we only found the unreduced acetaldehyde adducts increased in response to chronic ethanol feeding. Depending on the dose regimen, in some cases we also saw an increase in the malondialdehyde protein adducts in muscle. When you ask about the nature, do you mean the specific proteins? Thornalley: No, I mean the precise chemical adducts. We need to define the adducts chemically and quantify them independently by LC-MS/MS multiple reaction monitoring and using stable isotopic standards. This will result in very reliable data. Rao: It is also important to know whether the proteins of the contractile apparatus are modified. Preedy: That’s a good point. Worrall: Tuma’s group showed this a long time ago: they are actually poor substrates for acetaldehyde reactivity. In vitro very high concentrations of acetaldehyde are needed to get these contractile proteins modified. Tubulin can be modified easily in comparison. Seitz: When I was at V.A. Medical Center in the Bronx I saw some of these myopathy patients, and when I was at Northwick Park with Tim Peters I also saw them. In southern Germany, where people drink the same amount in the form of wine, we haven’t seen many. Do we underdiagnose them or are there differences with respect to different alcoholic beverages? Preedy: That is a good point. There was an initiative by the Spanish group to carry out a European study looking at dietary differences between different countries and different patterns of alcohol intake. The idea was to look at the same diseases in different European locations. There may be some dietary differences. Another previous idea was to take a group of alcoholic patients, and look for those that had myopathy compared with those who didn’t. They found only two differences. One difference was in an enzyme called carnosinase, which is responsible for catalysing the cleaving of the imidazole dipeptide carnosine which is an important antioxidant. The other difference was in α-tocopherol. They noticed that UK alcoholics who were myopathic had lower α-tocopherol in the plasma. Then there is selenium and α-tocopherol deficiency. Separately they cause myopathy. The tocopherol story could explain the differences, but when one looks at the tocopherol in Spanish alcoholics with myopathy, it has been found that, because of the Mediterranean diet, there are no differences in tocopherol compared to controls. Niemelä: Myopathy may well be underdiagnosed, because hospital registers typically indicate a prevalence of less than 10% for alcoholic cardiomyopathy among cardiomyopathy patients. If we look more systematically and take a careful alcohol history in each case, the prevalence tends to increase several-fold. Thus, there are

ALCOHOLIC MYOPATHY AND ACETALDEHYDE

181

many patients who go around with the diagnosis of cardiac insufficiency only, but the alcohol as a causative factor behind this may have been neglected. Albano: You presented some data showing an increase in CYP2E1 gene expression in the muscles of alcohol-fed rats. How does this increase in CYP2E1 correlate with the presence of oxidative stress markers? Preedy: Actually, I am not aware of studies showing that there definitely is CYP2E1 in skeletal muscle. There are conflicting results. If there is some, it is very low compared with the liver. I can’t recall any CYP2E1 studies in oxidative stress in skeletal muscle. In more severe stress and using other agents we may be able to look for CYP2E1 in muscle. Crabb: You mentioned the loss of muscle with cumulative alcohol consumption: could this be one of the major reasons for loss of lean body mass with age? Preedy: There may be common aetiological mechanisms for reduction in muscle mass. I did a literature search some time ago looking at the imposition of alcohol in the ageing but there wasn’t much published. Many diseases are characterized by type II atrophy. Seitz: What is the prevalence of myopathy in ALDH heterozygote people in Asia? Preedy: We haven’t done this work yet, but we have been planning a study. Ren: I have a comment in defence of the idea that in some situations acetaldehyde can be good. Some labs have found that a low dose of acetaldehyde can release catecholamines, which could have cardiovascular benefits. Also, low dose acetaldehyde induces stress signalling such as MAPK which should be beneficial in ischaemia–reperfusion injury.

References Al-Abed Y, Mitsuhashi T, Li H et al 1999 Inhibition of advanced glycation endproduct formation by acetaldehyde: role in the cardioprotective effect of ethanol. Proc Natl Acad Sci USA 96:2385–2390 Estruch R, Sacanella E, Fernandez-Sola J, Nicolas JM, Rubin E, Urbano-Marquez A 1998 Natural history of alcoholic myopathy: a 5-year study. Alcohol Clin Exp Res 22:2023–2028 Martin FC, Slavin G, Levi AJ, Peters TJ 1984 Investigation of the organelle pathology of skeletal muscle in chronic alcoholism. J Clin Pathol 37:448–454 Martin F, Ward K, Slavin G, Levi J, Peters TJ 1985 Alcoholic skeletal myopathy, a clinical and pathological study. Q J Med 55:233–251 Nicholls RM, Fowles LF, Worrall S, de Jersey J, Wilce PA 1994 Distribution and turnover of acetaldehyde-modified proteins in liver and blood of ethanol-fed rats. Alcohol Alcohol 29:149–157 Thurman RG 1998 Alcoholic liver injury involves activation of Kupffer cells by endotoxins. Am J Physiol 275:G605–G611 Tsukamoto H, French SW, Benson N et al 1985 Severe and progressive steatosis and focal necrosis in rat liver induced by continuous intragastric infusion of ethanol and low fat diet. Hepatology 5:224–232

182

PREEDY ET AL

Urbano-Marquez A, Estruch R, Navarro-Lopez F, Grau JM, Mont L, Rubin E 1989 The effects of alcoholism on skeletal and cardiac muscle. N Engl J Med 320:409–415 Vary TC, Nairn AC, Lang CH 2004 Restoration of protein synthesis in heart and skeletal muscle after withdrawal of alcohol. Alcohol Clin Exp Res 28:517–525 Vila L, Ferrando A, Voces J, Cabral de Oliveira C, Prieto JG, Alvarez AI 2001 Effect of chronic ethanol ingestion and exercise training on skeletal muscle in rat. Drug Alcohol Depend 64:27–33

Acetaldehyde adducts in circulation Onni Niemelä Department of Laboratory Medicine and Medical Research Unit, Seinäjoki Central Hospital and University of Tampere, FIN-60220 Seinäjoki, Finland

Abstract. Studies over the past two decades have indicated that protein modifications by acetaldehyde, the first metabolite of ethanol, are generated in the circulation as a result of alcohol abuse and play a major role in the pathogenesis of ethanol-induced diseases. Acetaldehyde can react with nucleophilic groups forming both stable and unstable adducts and several preferred target proteins have already been identified. Protein adducts have been found from the erythrocytes of heavy drinkers and from non-alcoholic volunteers after heavy drinking bouts. Elevated adduct levels also occurred in erythrocytes of women who continued to drink during pregnancy and subsequently gave birth to children with fetal alcohol effects. Evidence of adduct formation has also been observed in plasma proteins, including albumin and lipoproteins. Consequently, there may be interference of cellular functions, breakdown of immune tolerance and induction of autoantibodies towards the resulting neoantigens. Upon ethanol-induced oxidative stress, more abundant amounts and multiple species of adducts may be generated from aldehydic products of lipid peroxidation and through the formation of hybrid adducts. Studies in both human alcoholics and experimental animals have further demonstrated adduct deposition in tissues including the liver, brain, gut, muscle and heart thereby aggravating ethanol toxicity in such organs. 2007 Acetaldehyde-related pathology: bridging the transdisciplinary divide. Wiley, Chichester (Novartis Foundation Symposium 285) p 183–197

Mounting evidence both from human alcoholics and from experimental animal models has shown that acetaldehyde, the first metabolite of ethanol, and aldehydic products of lipid peroxidation can bind to proteins forming both stable and unstable adducts (Freeman et al 2005). This may lead to interference with cellular functions and generation of immunological responses towards the aldehyde-modified epitopes, and could cause injury to a wide variety of ethanol-exposed tissues. To date, several distinct adduct types have been described, including those with acetaldehyde, malondialdehyde (MDA), 4-hydroxynonenal (4-HNE), hybrid adducts of malondialdehyde and acetaldehyde (MAA), and hydroxyethyl radicals. The precise chemical nature of the various adducts and their relative importance in creating ethanol toxicity has, however, not been elucidated. The purpose of this contribution is to summarize recent progress in studies on acetaldehyde-derived protein adducts and their pathogenic and diagnostic 183

184

NIEMELÄ

implications in alcoholism. For more comprehensive information on adduct structures and tissue distributions readers are referred to recent reviews in this field (Freeman et al 2005, Niemelä 2001, Tuma 2002, Worrall & Thiele 2001). Characteristics of protein adducts Upon excessive drinking, both ethanol metabolism and associated oxidative stress may lead to the generation of aldehydic products, which are capable of adduct formation. Acetaldehyde can bind to reactive lysine residues, some aromatic amino acids and to cysteine (Tuma 2002, Worrall & Thiele 2001). At physiological conditions, acetaldehyde reacts with nucleophilic groups, such as ε-amino groups of internal lysine residues and the α-amino group on the N-terminal amino acid of unblocked proteins, forming unstable Schiff base adducts (Freeman et al 2005). Although conflicting data have been reported on the stabilization of the Schiff bases to form the N-ethyl lysine adduct, it appears that under appropriate reducing conditions proteins with abundant amounts of reactive lysine residues are modified at acetaldehyde concentrations that can occur in vivo as a result of excessive alcohol consumption (<200 µM) (Eriksson 2001). In the absence of reducing agents stable cyclic imidazolidinone structures are formed between acetaldehyde and the free α-amino group of the N-terminal valine of haemoglobin. Such adducts may be stabilized by bond rearrangement to form the cyclic 2-methyl-imidazolidin-4-one derivative and accumulate in the blood of alcoholics (Worrall & Thiele 2001). When reacting with N-terminal cysteine, acetaldehyde will form a thiazolidine derivative, but the reaction may not result in significant formation of the thiohemiacetal adduct in vivo (Freeman et al 2005). Induction of oxidative stress and enhanced lipid peroxidation appears to play a central role in ethanol toxicity (Caro & Cederbaum 2004). MDA, the aldehydic product of lipid peroxidation, can also react with amino groups of proteins and form protein adducts (Freeman et al 2005, Niemelä 2001, Worrall & Thiele 2001). The free radical-mediated oxidation of long-chain polyunsaturated fatty acids generates 4-HNE, which can react with various amino acid residues. Of these, lysine, cysteine, and histidine appear to be the major amino acids that participate in the formation of stable 4-HNE adducts. Tuma et al (2001) have further demonstrated the formation of hybrid adducts with acetaldehyde and malondialdehyde, which have been designated as MAA adducts. These consist of 2 equivalents of malondialdehyde with 1 equivalent of acetaldehyde (MAA 2 : 1 adduct), or 1 equivalent of malondialdehyde and 1 equivalent of acetaldehyde (MAA 1 : 1 adduct). Functional implications of adduct formation There seem to be preferential target proteins for aldehyde-derived modifications in vivo (Niemelä 2001). In vitro studies have shown that erythrocyte membrane

CIRCULATING ALDEHYDE ADDUCTS

185

proteins and haemoglobin can bind acetaldehyde. Adduct formation has also been described to occur with major serum proteins that are synthesised in the liver, including albumin, transferrin and lipoproteins. Proteins binding relatively high amounts of acetaldehyde include tubulin, ethanol-metabolizing cytochrome P450IIEI (CYP2E1) enzyme, collagens and ketosteroid reductase, which is a key enzyme in bile acid biosynthesis. As a consequence of adduct formation, the physicochemical properties of proteins, nucleic acids and lipids may be altered leading to disturbed cellular functions and aggravation of ethanol toxicity (Niemelä 2001, Tuma 2002). The interference of protein function has been well characterized for several proteins with abundant amounts of lysine residues in critical locations, such as tubulin or lysinedependent enzymes. Impaired microtubule function following acetaldehyde–tubulin condensation may affect protein secretion and plasma membrane assembly. Reactions between acetaldehyde and lipoproteins may activate apolipoprotein E synthesis in macrophages and thereby promote atherogenesis in alcohol abusers (Lin et al 1995). Acetaldehyde and the aldehydic products of lipid peroxidation also increase collagen mRNA levels and the expression of connective tissue proteins. An increasing body of evidence has further shown that aldehyde–protein adducts stimulate immunological responses directed against the specific types of modifications (Israel et al 1986, Latvala et al 2005, Rolla et al 2000, Stewart et al 2004, Worrall et al 1991).

Adducts in blood cells and molecules Erythrocytes Protein–acetaldehyde adducts have been detected in red cells of chronic alcoholics using either HPLC or immunological techniques. The latter approaches have been based on antibodies, which recognize acetaldehyde–protein condensates independently of the nature of the carrier protein (Israel et al 1986). It has been postulated that such adducts would be ideal markers for alcohol consumption since they represent specific metabolites of ethanol as an integral part of the analyte. However, as yet no routine applications have been developed. Comparisons between the sensitivities of the adduct measurements have also been limited. Several studies have reported elevated levels of haemoglobin–acetaldehyde adducts in chronic alcoholics (Chen et al 1995, Gross et al 1992, Hazelett et al 1998, Takeshita & Morimoto 2000). Lin et al (1993) found a sensitivity of 71% and a specificity of 96% for such adducts in comparisons contrasting chronic alcoholics and teetotallers. In larger materials, approximately 50% of alcohol abusers showed elevated values (Niemelä & Israel 1992). Acetaldehyde adducts also appear to accumulate in the erythrocytes of non-alcoholic volunteers even after occasional heavy

186

NIEMELÄ

drinking bouts (Niemelä et al 1990). Experiments on both acute ethanol intake and on hospitalized chronic alcoholics have indicated that adducts in erythrocytes remain elevated after blood ethanol has been eliminated. In chronic alcoholics, the values returned to normal after 1–3 weeks of abstinence (Niemelä & Israel 1992). Acetaldehyde adducts have also been found in erythrocytes of women who continued to drink during pregnancy and subsequently gave birth to children with fetal alcohol effects (Niemelä et al 1991a, Stoler et al 1998). Immunocytochemical studies of peripheral blood erythrocytes and bone marrow cells of alcoholics have revealed acetaldehyde-derived epitopes both inside the red cells and on cell membranes, showing an association with morphological abnormalities (Niemelä & Parkkila 2004). The cell membrane appeared to be the primary site of adduct deposition. Separation of the erythrocyte proteins by HPLC has revealed fast-eluting haemoglobin fractions that also react with antibodies against acetaldehyde adducts. Bone marrow aspirates have shown positive staining for acetaldehyde adducts in the late (non-nucleated) erythropoietic cells. It remains to be established, however, whether the formation of adducts would be associated with ethanolinduced aberrations in haematopoiesis in general. The degree of red cell macrocytosis has not been found to correlate with the amount of erythrocyte adducts (Niemelä & Parkkila 2004). A specific effect of acetaldehyde on erythrocytes is, however, supported by several lines of observations (Niemelä & Parkkila 2004, Tyulina et al 2006). Ethanol can permeate cell membranes and interfere with cell structure and metabolism. Although erythrocytes lack alcohol-metabolizing enzymes, except for catalase activity, high acetaldehyde concentrations have been found from the erythrocytes of alcoholics, which suggests a role for these cells as bioreactors for removing acetaldehyde (Hernández-Muñoz et al 1989). Interestingly, patients with high mean cell volumes of erythrocytes also frequently present with autoantibodies against the acetaldehyde-modified epitopes, suggesting that immunological mechanisms may also contribute to abnormalities in circulating blood cells in alcoholics (Koivisto et al 2006). Adducts in plasma proteins Although in vitro studies suggest that adduction of several serum proteins may also occur in the blood of alcoholics, only a few studies have reported increased adduct levels from plasma (Lin et al 1990, Tyulina et al 2006). In vitro, acetaldehyde binds to albumin more efficiently than to erythrocyte proteins (Israel et al 1986). Acetaldehyde binding to lipoproteins has also been described in a number of in vitro studies. Lipoproteins may contain different types of modifications, which could create immunological responses, activate apolipoprotein E synthesis in macrophages and induce inflammatory responses (Lin et al 1995). Studies have also pointed to a possible atherogenic role of ‘oxidized LDL’, which represents a

CIRCULATING ALDEHYDE ADDUCTS

187

mixture of aldehyde–LDL adduction. Indeed, adducts of proteins with MAA have been previously demonstrated from arterial vessel walls of atherosclerotic lesions (Hill et al 1998). Autoimmune responses towards protein adducts Aldehyde–protein adducts also trigger immunological responses directed against the specific neoantigens. Ethanol administration to experimental animals leads to the generation of circulating immunoglobulins with anti-acetaldehyde adduct, anti-MDA adduct, or anti-MAA adduct specificity (Israel et al 1986, Rolla et al 2000, Viitala et al 2000, Worrall et al 1991). Although antibody responses representing several immunoglobulin isotypes also appear to be generated in human alcoholics, the IgA responses seem to be typical for acetaldehyde-derived antigens (Fig. 1). Studies have indicated that anti-acetaldehyde adduct IgA titres are elevated in 60–70% of patients with alcoholic liver disease (ALD). The titres in ALD are significantly higher than in patients with non-ALD, non-drinking controls, or heavy drinkers without clinical and biochemical signs of liver disease. Anti-adduct IgGs have also been found in some patients with ALD. The antibody titres have also shown an association with the severity of ALD, as measured with clinical and laboratory indices or with the presence of inflammation and necrosis in biopsy specimens (Viitala et al 2000).

p < 0.001

1000

3000

750

p < 0.01

Anti-adduct IgG (U/l)

Anti-adduct IgA (U/l)

p < 0.001

p < 0.001

500

p < 0.05

250

0

Alcoholics with Heavy liver disease drinkers

Moderate drinkers

Abstainers

2000

1000

0

Alcoholics with Heavy liver disease drinkers

Moderate drinkers

Abstainers

FIG. 1. Antibodies to acetaldehyde derived adducts in circulation in patients with alcoholic liver disease (n = 54), heavy drinkers without apparent liver disease (n = 64), healthy moderate drinkers (n = 25) and in abstainers (n = 16). The antibody titres were measured by enzyme-linked immunosorbent assays using acetaldehyde-modified red cell protein as the test antigen as previously described (Latvala et al 2005, Hietala et al 2006). The IgA isotype antibodies are especially characteristic for the alcoholic patients. The bars indicate the mean (SD) titres of the antibodies representing different immunoglobulin isotypes.

188

NIEMELÄ

The generation of IgA responses against acetaldehyde adducts coincides with the induction of interleukin (IL)6 and IL10 cytokines and occurs early in the sequence of events leading from excessive alcohol consumption to clinical signs of alcoholic liver injury (Latvala et al 2005). It may be assumed that IgA antibodies contribute to exclusion and neutralization of antigens resulting from acetaldehyde modification of proteins during ethanol oxidation. The gastrointestinal tract is rich in enzymes capable of metabolizing ethanol. IgA antibodies may result from intestinally induced B cell responses since the epithelial tissues in alcoholics are continuously exposed to ethanol and mucosal immunity is highly adaptable to the antigenic load of the environment. Consistent with the above view, the IgA antibodies have also shown a strong correlation with the actual amounts of recent ethanol ingestion and change in parallel in the follow-up of alcoholics with abstinence (Latvala et al 2005, Hietala et al 2006). In alcoholics with severe liver disease, both IgA and IgG antibodies may occur together with excessive amounts of both pro- and anti-inflammatory cytokines suggesting that a disturbed balance between ethanol-derived antigen loading, immune responses and the ratios of pro- and anti-inflammatory cytokines play a role in the evolution of liver injury. It should be noted, however, that such immunological responses in vivo may show significant gender dependence. Autoantibodies against MDA adducts (Viitala et al 2000) and against MAA adducts (Rolla et al 2000) have also been found in the blood of human alcoholics. The highest titres of such antibodies are found in patients with severe liver disease. In patients with ALD, studies have also described the occurrence of specific T cell responses to adducted proteins (Stewart et al 2004). Deposition of protein adducts in tissues Liver The liver seems to be a predominant site of adduct deposition in vivo (Table 1) (Niemelä 2001). Immunohistochemical studies have revealed the occurrence of acetaldehyde–protein adducts in zone 3 hepatocytes in the early phase of liver disease (Niemelä et al 1991b, 1994). Protein adducts were present in heavy drinkers with no obvious clinical, biochemical or histological signs of alcoholic liver disease (Holstege et al 1994, Niemelä et al 1991b, 1994). In a follow-up study in a micropig model of alcoholic liver disease progressive accumulation of adducts has been demonstrated upon continuous ethanol intake. There is also a co-occurrence of acetaldehyde adducts and collagen deposition in the early phase of liver disease. Acetaldehyde-modified epitopes have also been detected from hepatocyte surfaces by flow cytometry. The presence of sinusoidal acetaldehyde adducts in human alcohol consumers has been suggested to contribute to the prognosis of the patients (Holstege et al 1994).

CIRCULATING ALDEHYDE ADDUCTS

189

TABLE 1 Aldehyde-derived adducts and their distributions in circulation and in tissues in alcoholics Adduct type

Location

Acetaldehyde

Erythrocytes and plasma proteins Liver Hepatocytes with zone 3 predominance Microsomal proteins Hepatocyte plasma membranes Sinusoids Areas of fatty deposition and hepatitis, Ito cells Brain Gut Muscle Heart Blood proteins Liver Hepatocytes with zone 3 predominance Co-localizes with acetaldehyde adducts in ALD Heart Liver Microsomes Atherosclerotic plaques

Malondialdehyde (MDA)

Malondialdehyde– acetaldehyde hybrid (MAA)

References: Niemelä (2001), Niemelä & Parkkila (2004).

High fat diets seem to stimulate the formation of protein adducts (Niemelä 2001). When alcohol was administered to rats together with such a diet, adduct formation was found in the early phase of liver damage co-occurring with increased expression of CYP2E1. When the ethanol-containing high fat diet is further supplemented with iron, an enhancer of oxidative stress, several types of adducts in more abundant amounts are seen coinciding with elevated levels of serum liver-derived enzymes ( Tsukamoto et al 1995). Liver pathology may also be promoted if ethanol is given together with a folate-deficient diet (Halsted et al 2002). Oxidative modifications of proteins with MDA and 4-HNE have been demonstrated from hepatic tissue in alcoholic liver disease (Niemelä et al 1994). Double immunofluorescence stainings for acetaldehyde and MDA adducts have revealed a significant co-localization between these two types of adducts and histological tissue damage. MAA adducts are also generated in alcoholic livers ( Tuma 2002). Extrahepatic tissues Muscle. Acetaldehyde adducts have also been implicated in the pathogenesis of muscle dysfunction in alcoholics, as evidenced by the demonstration of adducts in

190

NIEMELÄ

experimental models of alcohol-induced myopathy. These observations are discussed in detail by Preedy et al (2007, this volume). Heart. Recent studies have demonstrated that covalent binding of reactive aldehydic species to proteins also occurs in the heart in response to ethanol exposure (Niemelä et al 2003). The amounts of MDA and unreduced acetaldehyde adducts were found to be significantly increased in heart homogenates of rats treated with ethanol or cyanamide or both. Immunohistochemical studies revealed sarcolemmal adducts in those treated with ethanol and cyanamide in addition to intracellular adducts which were also present in the group treated with ethanol alone, indicating that enhanced lipid peroxidation and the generation of protein-aldehyde condensates may also have a role in the molecular mechanisms of cardiac dysfunction in alcoholics. Brain. Evidence of acetaldehyde-protein adduct formation in brain has been obtained in studies on experimental animal models. Adduct deposition after lifelong ethanol consumption was found in the white matter, large neurons in the deep layers of the frontal cortex, and in the molecular layer of the cerebellum, particularly in a rat strain (ANA), which usually shows high concentrations of acetaldehyde during ethanol oxidation. Gastrointestinal tract. Although intestinal bacteria are able to metabolize ethanol and thereby increase local acetaldehyde concentrations, toxicity and risk for carcinogenesis, only limited information has so far been available on the occurrence of protein adducts in the alimentary tract. In pancreas, ethanol intake leads to the formation of lipid peroxidation-derived adducts, suggesting a role for free radical generation in alcoholic pancreatitis. Conclusions Recent data have indicated that acetaldehyde, the first metabolite of ethanol, may play a role in the adverse effects of ethanol through the formation of protein adducts in circulation and ethanol-exposed tissues. Such modifications of proteins may lead to breakdown of immune tolerance and autoantibody generation. The characteristic immune responses and cytokine profiles in alcoholic patients indicate that these mechanisms are key players in the sequence of events leading from excessive ethanol consumption to tissue injury. Measurements of adducts and associated autoimmune responses may provide new diagnostic applications for more specific detection and biological staging of ethanol-induced diseases. Future studies appear warranted to explore the possibility

CIRCULATING ALDEHYDE ADDUCTS

191

that modulation of the biological responses to adduct generation could also improve the treatment of the alcoholic patients. Acknowledgements The original studies in the author’s laboratory were supported by the Finnish Foundation for Alcohol Studies.

References Caro AA, Cederbaum AI 2004 Oxidative stress, toxicology, and pharmacology of CYP2E1. Annu Rev Pharmacol Toxicol 44:27–42 Chen HM, Scott BK, Braun KP, Peterson CM 1995 Validated fluorimetric HPLC analysis of acetaldehyde in hemoglobin fractions separated by cation exchange chromatography: three new peaks associated with acetaldehyde. Alcohol Clin Exp Res 19:939–944 Eriksson CJP 2001 The role of acetaldehyde in the actions of alcohol (update 2000). Alcohol Clin Exp Res 25:15–32S Freeman TL, Tuma DJ, Thiele GM et al 2005 Recent advances in alcohol-induced adduct formation. Alcohol Clin Exp Res 29:1310–1316 Gross MD, Gapstur SM, Belcher JD, Scanlan G, Potter JD 1992 The identification and partial characterization of acetaldehyde adducts of hemoglobin occurring in vivo: a possible marker of alcohol consumption. Alcohol Clin Exp Res 16:1093–1103 Halsted CH, Villanueva JA, Devlin AM et al 2002 Folate deficiency disturbs hepatic methionine metabolism and promotes liver injury in the ethanol-fed micropig. Proc Natl Acad Sci USA 99:10072–10077 Hazelett SE, Liebelt RA, Brown WJ, Androulakakis V, Jarjoura D, Truitt EB Jr 1998 Evaluation of acetaldehyde-modified hemoglobin and other markers of chronic heavy alcohol use: effects of gender and hemoglobin concentration. Alcohol Clin Exp Res 22:1813–1819 Hernández-Muñoz R, Baraona E, Blacksberg I, Lieber CS 1989 Characterization of the increased binding of acetaldehyde to red blood cells in alcoholics. Alcohol Clin Exp Res 13: 654–659 Hietala J, Koivisto H, Latvala J, Anttila P, Niemelä O 2006 IgAs against acetaldehyde-modified red cell protein as a marker of ethanol consumption in male alcoholics, moderate drinkers and abstainers. Alcohol Clin Exp Res 30:1693–1698 Hill GE, Miller JA, Baxter BT et al 1998 Association of malondialdehyde-acetaldehyde (MAA) adducted proteins with atherosclerotic-induced vascular inflammatory injury. Atherosclerosis 141:107–116 Holstege A, Bedossa P, Poynard T et al 1994 Acetaldehyde-modified epitopes in liver biopsy specimens of alcoholic and nonalcoholic patients: localization and association with progression of liver fibrosis. Hepatology 19:367–374 Israel Y, Hurwitz E, Niemelä O, Arnon R 1986 Monoclonal and polyclonal antibodies against acetaldehyde-containing epitopes in acetaldehyde-protein adducts. Proc Natl Acad Sci USA 83:7923–7927 Koivisto H, Hietala J, Anttila P, Parkkila S, Niemelä O 2006 Long-term ethanol consumption and macrocytosis: diagnostic and pathogenic implications. J Lab Clin Med 147:191–196 Latvala J, Hietala J, Koivisto H, Järvi K, Anttila P, Niemelä O 2005 Immune responses to ethanol metabolites and cytokine profiles differentiate alcoholics with or without liver disease. Am J Gastroenterol 100:1303–1310

192

NIEMELÄ

Lin RC, Lumeng L, Shahidi S, Kelly T, Pound DC 1990 Protein-acetaldehyde adducts in serum of alcoholic patients. Alcohol Clin Exp Res 14:438–443 Lin RC, Shahidi S, Kelly TJ, Lumeng C, Lumeng L 1993 Measurement of hemoglobin-acetaldehyde adduct in alcoholic patients. Alcohol Clin Exp Res 17:669–674 Lin RC, Dai J, Lumeng L, Zhang MY 1995 Serum low density lipoprotein of alcoholic patients is chemically modified in vivo and induces apolipoprotein E synthesis by macrophages. J Clin Invest 95:1979–1986 Niemelä O 2001 Distribution of ethanol-induced protein adducts in vivo: relationship to tissue injury. Free Radic Biol Med 31:1533–1538 Niemelä O, Israel Y 1992 Hemoglobin-acetaldehyde adducts in human alcohol abusers. Lab Invest 67:246–252 Niemelä O, Parkkila S 2004 Alcoholic macrocytosis–is there a role for acetaldehyde and adducts? Addict Biol 9:3–10 Niemelä O, Israel Y, Mizoi Y, Fukunaga T, Eriksson CJP 1990 Hemoglobin-acetaldehyde adducts in human volunteers following acute ethanol ingestion. Alcohol Clin Exp Res 14:838–841 Niemelä O, Halmesmäki E, Ylikorkala O 1991a Hemoglobin-acetaldehyde adducts are elevated in women carrying alcohol-damaged fetuses. Alcohol Clin Exp Res 15:1007–1010 Niemelä O, Juvonen T, Parkkila S 1991b Immunohistochemical demonstration of acetaldehydemodified epitopes in human liver after alcohol consumption. J Clin Invest 87:1367–1374 Niemelä O, Parkkila S, Ylä-Herttuala S et al 1994 Covalent protein adducts in the liver as a result of ethanol metabolism and lipid peroxidation. Lab Invest 70:537–546 Niemelä O, Parkkila S, Worrall S, Emery PW, Preedy VR 2003 Generation of aldehyde-derived protein modifications in ethanol-exposed heart. Alcohol Clin Exp Res 27:1987–1992 Preedy VR, Crabb DW, Farrés J, Emery PW 2007 Alcoholic myopathy and acetaldehyde. In: Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Found Symp 285) p 158 –182 Rolla R, Vay D, Mottaran E et al 2000 Detection of circulating antibodies against malondialdehyde-acetaldehyde adducts in patients with alcohol-induced liver disease. Hepatology 31: 878–884 Stewart SF, Vidali M, Day CP, Albano E, Jones DE 2004 Oxidative stress as a trigger for cellular immune responses in patients with alcoholic liver disease. Hepatology 39:197–203 Stoler JM, Huntington KS, Peterson CM et al 1998 The prenatal detection of significant alcohol exposure with maternal blood markers. J Pediatr 133:346–352 Takeshita T, Morimoto K 2000 Accumulation of hemoglobin-associated acetaldehyde with habitual alcohol drinking in the atypical ALDH2 genotype. Alcohol Clin Exp Res 24:1–7 Tsukamoto H, Horne W, Kamimura S et al 1995 Experimental liver cirrhosis induced by alcohol and iron. J Clin Invest 96:620–630 Tuma DJ 2002 Role of malondialdehyde-acetaldehyde adducts in liver injury. Free Radic Biol Med 32:303–308 Tuma DJ, Kearley ML, Thiele GM et al 2001 Elucidation of reaction scheme describing malondialdehyde-acetaldehyde-protein adduct formation. Chem Res Toxicol 14:822–832 Tyulina OV, Prokopieva VD, Boldyrev AA, Johnson P 2006 Erythrocyte and plasma protein modification in alcoholism: a possible role of acetaldehyde. Biochim Biophys Acta 1762: 558–563 Viitala K, Makkonen K, Israel Y et al 2000 Autoimmune responses against oxidant stress and acetaldehyde-derived epitopes in human alcohol consumers. Alcohol Clin Exp Res 24: 1103–1109 Worrall S, Thiele GM 2001 Protein modification in ethanol toxicity. Adverse Drug React Toxicol Rev 20:133–159 Worrall S, de Jersey J, Shanley BC, Wilce PA 1991 Antibodies against acetaldehyde-modified epitopes: an elevated IgA response in alcoholics. Eur J Clin Invest 21:90–95

CIRCULATING ALDEHYDE ADDUCTS

193

DISCUSSION Thornalley: My group quantifies glycation adducts in diabetes and other diseases. We measure about 12 different glycation adducts, four oxidation adducts and nitrotyrosine. The experience of the glycation field has shown that even if you produce monoclonal antibodies to defined chemical adducts in highly modified proteins, using a highly modified immunogen that is not well characterized, even for monoclonal antibodies you can misidentify the epitope that is being recognized. In this field, when people produce antibodies for ELISA they are making a highly modified immunogen to get a good response. If you make an antibody to something like N-ethyl lysine, in vivo from glycation there is N-carboxyethyl-lysine. You only need a little of this decarboxylated form in your sample and you can get a positive response. There are many of these adducts in any proteins that are used as blocking proteins in a Western or ELISA. You can get high background responses. This type of work is only put on a sure footing when you can corroborate your antibody response to liquid chromatography mass spectroscopy (LCMS)/MS quantitation. Even then it is restricted to a precise assay matrix. It is very important to use LCMS/MS to quantify these specific adducts. The other thing I’d like to comment about is looking at responses of monocytes and other inflammatory cells to modified proteins. Two errors have been made over the decades. One is not to use endotoxin-free protein. Also, highly modified proteins respond profoundly differently to the minimally modified proteins that are generated in vivo. Most of the receptor responses of highly glycated albumin are probably to a small fraction of the albumin that is highly damaged and aggregated. This is what you are looking at responses to: not to a glycation adduct on a particular protein. It is quite conceivable that the people who have done studies on activation of Kupffer cells with albumin or other proteins highly modified by acetaldehyde are looking at responses to a tiny fraction of the protein that has been highly damaged and aggregated. Then you can get activation through several different types of receptor. RAGE (receptor for advanced glycation end products) is one that we now know responds to the aggregated protein produced. There are many potential pitfalls in this kind of work. It would be tremendously valuable if you could get a well validated marker of acetaldehyde or a related metabolite of haemoglobin or albumin that could be used to check on alcohol consumption in clinic. The ‘gold-standard’ reference method for quantitation should be LCMS/MS-based or a similar mass spectrometric method. Worrall: With regard to modification of proteins, you were right about the situation 10 years ago when people were using a couple of hundred millimolar acetaldehyde in their modification protocols. But these days most people are using physiological concentrations of a half to one millimolar or so. This makes the massive over-modification you talk about less likely. I agree with what you said about LPS. This can be a major issue in the use of modified proteins in some

194

NIEMELÄ

experiments, but it probably doesn’t matter when you are making antibodies against these adducts, and may in fact help with the immunoreactivity of the immunogen. Thornalley: The number of adducts per protein molecule is an important factor. In the glycation field people were increasing the mass of albumin by 7 kDa— approximately 10% of the mass comprised 40 adducts added per molecule. In vivo there is one adduct on every 200 molecules. If you could miniaturize yourself and stand on a vascular endothelial cell, and watch the albumin molecules floating past, approximately one in every 100 albumin molecules floating past has an advanced glycation end product (AGE) attached to it. The molecules created in vitro floating past would each have 40 AGEs attached. It is challenging the physiological relevance of the experiment if you do this. Worrall: We have tried similar work in which we used sandwich ELISAs looking at two different adduct epitopes on the same protein. Only a small fraction of the proteins that were modified actually carried more than one modified epitope. With some types of modification, such as the MAA adduct, there may be only one site on the protein which can actually make that adduct. To form this adduct you appear to need several reactive amino acid residues in the correct threedimensional conformation to initially make two adducts which then combine to make one. Niemelä: I agree that it is important to know the precise chemical nature of the adducts. They are indeed largely unknown at this time. This is why I like to broadly speak about acetaldehyde-derived adducts. When you generate adducts in vitro it has been shown that you create different types of structures depending on the concentration of acetaldehyde and on the presence or absence of a reducing agent. An important feature of the antibodies generated against the acetaldehyde-derived adducts is, however, that they recognize acetaldehyde protein modifications independently of the nature of the carrier protein. In my hands, the antibodies which have been produced with in vitro modifications in the range of 100 µM to 1 mM have provided the best results in detecting acetaldehyde-derived modifications from alcoholics. The adducts created in these conditions seem to be rather close to those that are generated in vivo. On the other hand, Larry Lumeng’s group have made adducts with 240 mM modification of keyhole limpet haemocyanin and got antibodies that differentiated between alcoholics and non-alcoholics, indicating that modifications prepared at high concentrations of acetaldehyde may also create structures that share the important immunological determinants (Lin et al 1993). However, there still seems to be a long journey ahead before the adduct determinations are ready for routine laboratory use. One should also remember that in the glycosylation field, the first evidence that proteins were glycosylated came over 50 years ago, but the first antibodies that were functional in clinical laboratories only came a few years ago.

CIRCULATING ALDEHYDE ADDUCTS

195

Thornalley: The quantitation of these adducts by immunoassays can be highly misleading. There are investigators using antibodies to detect Nε-carboxymethyllysine in diabetic patients who claim a 10-fold increase in the epitope concentration, with respect to normal healthy human subjects. When we do the analysis with LCMS/MS we don’t see any significant difference in concentration of analyte in plasma protein of diabetic patients and healthy controls. The antibody is recognizing additional epitopes. Immunoassay data often need corroboration by LCMS/MS. Rao: Is there any information regarding the turnover rate of these proteins that have acetaldehyde adducts? Are they targeted more towards the proteosome? Worrall: In cells the proteins stay around longer. Outside cells they go away quicker. Their half lives halve, but in cells it can go up by about tenfold. Rao: This means in the cell, adducts can accumulate during the course of acetaldehyde exposure. So the longer the exposure means more adducts. Worrall: There seems to be a point where a steady state is reached. Niemelä: I suspect this is very much dependent on the half-life of the protein. Worrall: I’m talking in general terms, that is, in relation to the whole population of proteins. Preedy: There seems to be a relatively high basal level of adduct formation, due to endogenous acetaldehyde. In some of these dosing experiments where there is an increase in the amount of adducts, couldn’t it be that the treatment is not only increasing the adducts, but also impairing the rate at which they are being degraded? Worrall: If you look in animals over time, the adduct concentration goes up and then plateaus. This could be indicative of the generation of newly modified proteins but also of the previously modified proteins being turned over at a slower rate. Thus, it is likely that there will be old and newly modified proteins in the population. Niemelä: There are proteins such as collagens, which are prone to cross-linking. It might be that conditions involving high acetaldehyde levels favour protein crosslinking, and this could occur with connective tissue proteins in alcoholics and alter the half-lives of proteins and biophysical properties of the extracellular matrix. Apte: Did you try to correlate protein adducts and cytokine levels, or did you just measure the different cytokines present? Niemelä: Our idea was initially to examine the changes in the patterns of pro- and anti-inflammatory cytokines in the same population of patients in whom we measured antibodies and who were classified according to the presence or absence of liver disease. We wanted to see whether there was any association between the antibody responses, cytokines, alcohol consumption and inflammation in general. In alcoholic liver disease patients, we saw high levels of especially IgA antibodies together with high levels of pro- and anti-inflammatory cytokines. In the alcoholics without liver disease, high levels of IgA antibodies occurred together with high IL6 and anti-inflammatory IL10 levels. It appears that the IgA antibodies correlate most

196

NIEMELÄ

strikingly with the amount of alcohol consumed. Our more recent data have indicated these antibodies also correlate with IL6 and tumour necrosis factor (TNF)α levels (Hietala et al 2006). Apte: It made me think of endotoxin and endotoxinaemia, which is known to occur in alcoholics. This is thought to be due to increased mucosal permeability leading to an increased translocation of Escherichia coli from the gut into the blood. Do these adducts have any effect on tight junctions of the mucosa in the gut, which could explain the increased permeability that has been shown? Niemelä: That is an interesting point, which hasn’t been looked into yet. Gastric permeability is indeed increased, and there is also endotoxinaemia as a result of alcohol abuse. There are also autoantibodies generated against endotoxin and other gut-derived antigens that have been suggested to be due to the increased gastrointestinal permeability. It may be that this IgA response to adducts, which seems to be very specific for alcohol consumption, could be related to these events in the gastrointestinal tract. Their specific relationship to endotoxinaemia hasn’t, however, been studied. Rao: The acetaldehyde effect on gut permeability is mediated by perturbation of intracellular signal transduction. The mechanism involved in this process is the formation of acetaldehyde adduct with the signalling molecules. One of the signalling molecules that is quickly inactivated by acetaldehyde adduct formation is PTP1B. This activity is inhibited by acetaldehyde, which results in increased tyrosine phosphorylation of junctional proteins. Albano: Concerning the role of endotoxins in the development of alcohol hepatotoxicity, experiments performed in collaboration with Tom Badger’s group in Little Rock have shown that liver injury develops in alcohol-fed rats also in the absence of an appreciable increase in plasma endotoxin levels (Ronis et al 2004). Interestingly, in these animals there is a strict correlation between the titres of antibodies against the MDA adducts and the liver expression of TNFα mRNA. This suggests the possibility that immune responses triggered by oxidative stress might stimulate inflammation in the liver even in absence of endotoxin (Ronis et al 2005). Apte: So you don’t think LPS has any role to play. Albano: No, LPS is a great stimulus for macrophages and antigen presenting cells to present antigens to the helper T lymphocytes. So the presence of endotoxin would probably speed up the process and increase immune responses triggered by oxidative stress. Seitz: You referred to the work of Charles Halstead on folate deficiency in alcoholics (Halsted et al 2002). You showed that when CYP2E1 increases, its adducts also increase. I understand with a folate-deficient diet, homocysteine is present as is endoplasmic oxidative stress. There is also more fat in the liver. But why are there more acetaldehyde adducts? Is this because of more production of CYP2E1?

CIRCULATING ALDEHYDE ADDUCTS

197

Niemelä: It’s hard to say. Acetaldehyde adducts were increased in ethanol-containing diets, but it should be noted that although adduct levels in the folate-deficient ethanol diet were slightly higher than those in the folate-sufficient ethanol diet, the difference between those two was not significant. As discussed here previously, it appears that acetaldehyde may actually also be able to destroy folate. The CYP2E1 and adducts in those experiments were detected by immunohistochemistry, which is not quantitative enough for differentiation if there are only small differences between groups. We might also need protein level studies for confirmation. Worrall: We have a cell system with no acetaldehyde at all. We load cells up with unsaturated fatty acids and then expose them to iron. This can produce acetaldehyde adducts, by making lots of MDA, which decomposes to give some acetaldehyde. MAA adducts are also formed. References Halsted CH, Villanueva JA, Devlin AM et al 2002 Folate deficiency disturbs hepatic methionine metabolism and promotes liver injury in the ethanol-fed micropig. Proc Natl Acad Sci USA 99:10072–10077 Hietala J, Koivisto H, Latvala J, Anttila P, Niemela O 2006 IgAs against acetaldehyde-modified red cell protein as a marker of ethanol consumption in male alcoholic subjects, moderate drinkers, and abstainers. Alcohol Clin Exp Res 30:1693–1698 Lin RC, Shahidi S, Kelly TJ, Lumeng C, Lumeng L 1993 Measurement of hemoglobinacetaldehyde adduct in alcoholic patients. Alcohol Clin Exp Res 17:669–674 Ronis MJJ, Hakkak R, Korourian S et al 2004 Alcoholic liver disease in rats fed ethanol as part of oral or intragastric low-carbohydrate diets. Exp Biol Med 229:351–360 Ronis MJJ, Butura A, Sampey BP et al 2005 Effects of N-acetyl cysteine on ethanol-induced hepatotoxicity in rats fed via total enteral nutrition. Free Radic Biol Med 39:619–630

GENERAL DISCUSSION Worrall: We have looked at modification of plasma proteins in ethanol-fed rats which have been fed for 12 weeks on a Lieber-DeCarli liquid diet. We have clearly shown that albumin from the ethanol-fed animals carries a whole series of different adduct types which are the sort you would expect to find in the liver. It is quite likely that the albumin itself is actually being modified during its secretory progress through the liver. On the other hand, IgG, which is not synthesised by the liver, also carries modifications by acetaldehyde and others from malondialdehyde or hydroxyethyl radicals. Unless something like plasma cells are metabolizing ethanol and producing acetaldehyde, this suggests that this modification is occurring out in the circulation. We have also done some work on brain tissue from 12 week Lieber-DeCarli-fed rats using a battery of antibodies in ELISAs to compare control and ethanol pairfed animals. Adducts such as unreduced acetaldehyde adducts seem to be present in brain. We don’t see reduced acetaldehyde adducts. Malondialdehyde and malondialdehyde–acetaldehyde (MAA) adducts are also present. We have also done work on some highly characterized human brain material, from the Brisbane node of the NHMRC National Neural Tissue Resource Consortium, from people with alcoholic cerebellar degeneration. Using a similar panel of antibodies we can show that these people have an elevated concentration of unreduced acetaldehyde adducts in their cerebellum. They also have MAA adducts in their cerebellum. Finally, they have a new adduct (acetaldehyde-generated advanced glycation end-product; AcHAGE) we have just started working on, which seems to be a glycosylation adduct, but one that is made without normal glycosylation reactions occurring—you can just use acetaldehyde and do not need a sugar. There is one existing report on this in the literature (Takeuchi et al 2003). The general properties of glycosylation adducts include brown colouration, fluorescence and cross-linking. The adducts we are interested in can be made on a simple molecule, hexylamine. If the reaction mixture is treated with cyanoborohydride immediately after the addition of acetaldehyde then AcH-AGE adduct formation is stopped. This suggests that a Schiff base is initially formed followed by further reactions to generate the AcH-AGE adduct. We know that Dean Tuma uses hexylamine to make MAA adducts so we checked to see whether we were also forming a MAA derivative. However, we found that the adduct we had formed had very different fluorescence excitation and emission wavelengths, and a different absorption spectrum. In the literature, this adduct has been applied to rat cortical neurons in culture and it killed them all 198

GENERAL DISCUSSION

199

off (Takeuchi et al 2003). Thus, it seems that this adduct may be a potential neurotoxic insult. Emery: I’d like to ask about the brain data: are you confident that those adducted proteins were present when the people died? Worrall: Yes, I am. There are different postmortem delays, but between the groups this is averaged out. They are age- and sex-matched. We see a consistent difference between the controls and the alcoholics. Preedy: In vitro these adducts seem to be neurotoxic. How stable are they? Worrall: They are stable enough to be used as inhibitors in competition assays, to check the binding to the proteins. I don’t know how stable exactly. We are looking at this at present. Niemelä: This is most interesting because these are the first human data of this nature. A few years ago in experimental animals after life-long alcoholic exposure we saw adducts in cerebellum and some deeper layers of the frontal cortex. Were your findings specific for cerebellum? Worrall: So far we have only accessed the cerebellar tissues. We plan to look at other brain area such as prefrontal cortex (damaged in alcohol abuse) and motor cortex (spared in alcohol abuse). Morris: Is it possible that the adducts are the result of the disease process and not the cause? Worrall: I’m not saying that these are causative agents, but they seem to be a good marker of damage. Ren: So is there no sugar component in these adducts? Worrall: You can take hexylamine and acetaldehyde, incubate them in a test-tube and get this adduct from the reaction. Rao: Is there any information on the functional consequence of adduct formation in haemoglobin? Does it affect the oxygen binding kinetics? Niemelä: Not that I know of, at least as far as physiological concentrations of acetaldehyde are concerned. There is evidence for the potential role of erythrocytes as bioreactors in acetaldehyde removal. It is also possible that erythrocyte cell membranes exhibit altered physicochemical properties as a result of acetaldehyde binding. Thornalley: It is known that glycated haemoglobin is associated with a slightly increased affinity for oxygen, and decreased cooperativity in binding. In our studies where we have examined the modification of haemoglobin by dicarbonyls, which may be part of the alcoholic syndrome, there is a similar effect. Reference Takeuchi M, Watai T, Sasaki N 2003 Neurotoxicity of acetaldehyde-derived advanced glycation end products for cultured cortical neurons. J Neuropathol Exp Neurol 62:486–496

Pancreatic MAP kinase pathways and acetaldehyde M. Apte, J. McCarroll, R. Pirola and J. Wilson Pancreatic Research Group, South Western Sydney Clinical School, Liverpool Hospital and the University of New South Wales, Sydney, Australia

Abstract. Alcohol abuse is a major cause of pancreatitis, a condition that can manifest as both acute necroinflammation and chronic damage (acinar atrophy and fibrosis). It is generally accepted that alcohol-induced pancreatic injury is a consequence of the metabolism of alcohol by the pancreas (via the oxidative and non-oxidative pathways) producing the toxic metabolites acetaldehyde and fatty acid ethyl esters (FAEEs) respectively. Ethanol oxidation within the pancreas also leads to oxidant stress within the gland. Acetaldehyde, oxidant stress and FAEEs cause numerous molecular changes in pancreatic acinar cells which predispose the gland to autodigestion and necroinflammation. An important recent development relates to the identification of pancreatic stellate cells (PSCs) as the key mediators of alcohol-induced pancreatic fibrosis, when activated by ethanol, acetaldehyde or oxidant stress. Recent studies implicate the mitogen activated protein kinase (MAPK) pathway, a major signalling pathway in mammalian cells, as a critical regulator of the effects of ethanol and acetaldehyde on acinar cells as well as PSCs. Particularly important are the modulatory effects of ethanol and its metabolites on downstream transcription factors NF-κB and AP-1 (which regulate inflammatory responses via cytokine production) in acinar cells. In PSCs, additional signalling molecules identified as important to the process of ethanol and acetaldehyde-induced PSC activation include protein kinase C (PKC), phosphatidylinositol-3-kinase (PI3K) and peroxisome proliferator-activated receptor γ (PPARγ). Interestingly, cross-talk has been demonstrated between PI3K and MAPK in acetaldehyde-treated PSCs. The above advances in the identification of relevant signalling molecules may enable therapeutic targeting of these pathways so as to prevent/reduce alcohol-induced acute as well as chronic injury of the pancreas. 2007 Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Foundation Symposium 285) p 200–216

Alcoholic pancreatitis, a serious complication of alcohol abuse, has both acute ( pancreatic necroinflammation) and chronic (acinar atrophy and fibrosis) manifestations. Over the past two decades significant advances have been made in our understanding of the toxic effects of ethanol and its metabolites on the major functional cell of the exocrine pancreas (the acinar cell) and on the key nonparenchymal cell responsible for pancreatic fibrosis (the pancreatic stellate cell) 200

ACETALDEHYDE AND PANCREATIC MAPK PATHWAYS

201

(Apte et al 2005). It is now established that the pancreas can metabolize ethanol via both the oxidative pathway generating acetaldehyde and the non-oxidative pathway generating fatty acid ethyl esters (FAEEs) (Gukovskaya et al 2002, Haber et al 1998, 2004). It is also known that the effects of alcohol, its metabolites and its metabolic by-products (reactive oxygen species) lead to: (i) increased content of digestive and lysosomal enzymes (via increased synthesis and possibly decreased secretion in the case of digestive enzymes); and (ii) increased potential for contact between digestive and lysosomal enzymes (via increased organelle fragility mediated by compounds such as cholesteryl esters, FAEEs and reactive oxygen species) (Apte et al 2005). These changes may facilitate premature intracellular activation of digestive enzymes and predispose the gland to autodigestive injury and necroinflammation, in the presence of an appropriate (as yet unidentified) trigger factor. More recently, ethanol and acetaldehyde have been shown to activate pancreatic stellate cells thereby increasing the synthesis of extracellular matrix proteins leading to pancreatic fibrosis (Apte et al 2000). Given the demonstrated effects of ethanol and its metabolites on pancreatic acinar and stellate cells, researchers have now focussed their attention on the intracellular signalling pathways that may mediate these deleterious effects of ethanol on the pancreas. Obviously, the ultimate goal of such work is to identify and characterize the signalling molecules or pathways that could be therapeutically targeted in order to prevent or reverse the toxic effects of ethanol on the gland. Cells in the human body are in constant interaction with each other and their surroundings by sending and receiving signals or messages. Extracellular factors such as alcohol, growth factors and cytokines transmit signals to intracellular targets via many different signalling pathways. Each pathway comprises a network of interacting molecules that are responsible for regulating numerous cellular processes. In response to these extracellular signals, cells can modulate (up-regulate or down-regulate) specific sets of genes which control functions such as cell metabolism, protein synthesis, proliferation, migration, and differentiation. A major signalling pathway in mammalian cells is the mitogen activated protein kinase (MAPK) pathway (Pearson et al 2001). MAP kinases are serine/threonine kinases that are focal points for a diverse array of cellular functions. Four MAPK families have been described to date. These include: (i) extracellular signal-related kinase 1 and 2 (ERK1/2); (ii) c-jun N-terminal kinase 1 and 2 ( JNK1/2); (iii) p38 kinase; and (iv) ERK5/big MAPK 1 (BMK-1). However, the last named family is yet to be fully characterised. Typically, the MAPK pathway consists of three hierarchical protein kinases which include MAPK kinase kinase (MAPKKK/Raf ), MAPK kinase (MAPKK/MEK) and MAPK. An activating event (such as binding of agonists to G protein-coupled receptors, interaction of small GTP-binding proteins, or the presence of cellular stressors such as oxidant stress, UV radiation and aldehydes) sets off a cascade of phosphorylations of the tyrosine and threonine residues in

202

APTE ET AL

each kinase. Thus, MAPKKK phosphorylates and activates MAPKK, which in turn phosphorylates and activates MAPK. Activated MAP kinases modulate the activities of several downstream transcription factors, including c-jun, c-fos (components of activator protein 1, AP-1), NF-κB, Elk and c-myc, thereby controlling gene expression and cell behaviour (Fig. 1). NF-κB and AP-1 are of particular interest since they are known to modulate cytokine expression (which may play a role in pancreatic necroinflammation) in cells. AP-1 also regulates extracellular matrix protein synthesis (a function relevant to alcohol-induced pancreatic fibrosis). In addition to nuclear targets, ERK1/2 activation has been shown to regulate cytoplasmic targets such as cytoskeletal proteins and cell surface receptors such as EGF and NGF receptors (Dabrowski 2003). Upstream kinases that influence the MAPK cascade include phosphatidylinositol-3-kinase (PI3K) and protein kinase C (PKC). PI3K exerts its effect by phosphorylating the 3′-OH position of the inositol ring of phosphatidylinositol (a component of the lipid bilayer within the cell membrane) (Anderson & Jackson 2003). PKC represents a family of phospholipid-dependent serine/threonine kinases (Dempsey et al 2000). PKC isoforms are grouped into

Ligand Cellular Stressors: (Ethanol, acetaldehyde, oxidant stress)

G protein:

Ras

Rac, PAK, Cdc42

Cdc42

Raf

MEKK1,-2,-3,-4,-5, ASK-1, MLK-2,-3,TAK

MEKK1,-2,-3,-4,-5, ASK-1, MLK-2,-3,TAK

MAPKK:

MEK1&2

Mkk4&7

Mkk3,4&6

MAPK:

ERK1/2

MAPKKK:

JNKs

P38(a,b,g,s)

Nuclear Transcription Factors

Proliferation; Differentiation; Inflammation; Apoptosis; Protein synthesis FIG. 1. Schematic diagram of mitogen-activated protein kinase (MAPK) activation in mammalian cells. Depicted are the three MAPK cascades which are typically activated by small G proteins leading to a series of phosphorylations resulting in activated ERK1/2, JNK and p38K. Each of these influences the phosphorylation of downstream transcription factors either directly or through other kinases, resulting in the regulation of a variety of genetic programs within the cell.

ACETALDEHYDE AND PANCREATIC MAPK PATHWAYS

203

three categories based on their structure and substrate specificities: classical PKCs which are Ca2+-dependent and activated by binding to diacylglycerol (DAG) and phosphatidylserine (PS); novel PKCs which are Ca2+-independent but activated by DAG and PS and atypical PKCs which are Ca2+ and DAG independent but are PS sensitive. Activation of PKCs can lead to phosphorylation of Raf and activation of the MAPK cascade. This paper reviews current knowledge regarding the effects of ethanol, its oxidative metabolite acetaldehyde, its non-oxidative metabolite FAEE and the by-product of ethanol metabolism, oxidant stress, on the MAPK pathway and downstream transcription factors in pancreatic acinar cells and pancreatic stellate cells. Effect of ethanol and its metabolites on acinar cell signalling To date, limited information is available on MAPK signalling in pancreatic acinar cells in response to ethanol or its metabolites. However, important clues may be obtained from studies in the liver (the other major abdominal organ affected by alcohol) where MAPK activity in response to ethanol has been relatively well characterised (see for review Aroor & Shukla 2004). Lee et al (2002) have shown that cultured rat hepatocytes exhibit moderate activation of ERK1/2 upon acute exposure to ethanol for 1 hour and for 24 hours, while acetaldehyde induces ERK activation as early as at 10 minutes of incubation. JNK was found to exhibit a strong and sustained activation in response to ethanol. In contrast to ERK1/2 and JNK, p38K activity was not influenced by ethanol alone (Chen et al 1998). However, ethanol was found to potentiate the inductive effect of tumour necrosis factor (TNF)α on p38 kinase. Given that p38K signalling results in caspase 3 activation, mitochondrial depolarization and cell death, the authors postulated that this pathway may mediate the ethanol-induced sensitization of cultured hepatocytes to TNFα-induced damage. Contrary to the MAPK activation observed after acute ethanol exposure, chronic treatment with ethanol has been shown to inhibit ERK1/2 and p38K activation induced by partial hepatectomy or by other factors such as EGF or insulin (Chen et al 1998). Similarly, in contrast to the prolonged activation of JNK induced by acute ethanol exposure, JNK activity in rats chronically treated with ethanol was attenuated (Lee et al 2002). Endotoxin-induced JNK activation is also reduced in hepatocytes from ethanol-treated mice compared to mice treated with endotoxin alone (Koteish et al 2002). The reasons for this diversity of MAPK responses upon exposure to acute and chronic ethanol are not immediately apparent. Whether acute and chronic ethanol exposure affects the MAPK pathway in pancreatic acinar cells in a manner similar to hepatocytes is yet to be delineated and represents an important area of future research. Most of the studies characterising MAPK activation in pancreatic acinar cells relate to the effects of the secretagogue cholecystokinin (CCK) on isolated acini. These

204

APTE ET AL

have been recently well reviewed by Williams et al (2002) and Dabrowski (2003). It has been demonstrated that physiological concentrations (pM) of CCK rapidly activated ERK1/2 with maximum activation being reached at 5 minutes. JNK activation occurred at a slower pace (reaching a maximum at 30 minutes) but with 10-fold higher CCK concentrations. Supraphysiological levels of CCK used to induce experimental pancreatitis increased JNK activity by 25-fold; this is thought to represent a stress response in the cells. CCK was also found to rapidly activate p38K. Interestingly, inhibition of p38K using the specific inhibitor SB2030580 was found to block the CCK-induced changes in the actin cytoskeleton. Actin reorganization by CCK is thought to mediate the inhibition of acinar cell secretion characteristic of experimental pancreatitis induced by supramaximal CCK. This blockage of acinar secretion is thought to facilitate premature intracellular activation of digestive enzymes leading to autodigestive injury of the gland. The postulated role of p38K in actin reorganization is relevant to ethanol-induced pancreatic injury given that ethanol (at least in acute doses) is known to inhibit pancreatic acinar secretion. In this regard, Siegmund et al (2004) have demonstrated that ethanol in concentrations ranging from 25 mM (often seen with social drinking) to 346 mM, disrupted the actin cytoskeleton in pancreatic acini and that this effect was blocked by a PKC inhibitor ( p38K was not examined in this study). The authors speculated that ethanol occupies low-affinity CCKA receptors on acinar cells, since PKC is a major pathway activated by ligand binding to these receptors. However, given that ethanol is a freely diffusible compound, a direct effect of ethanol on PKC cannot be discounted. While little work has been done with ethanol and its metabolites on the MAPK pathway in pancreatic acinar cells, recent studies have examined the effects of these compounds on the downstream transcription factors NF-κB and AP-1. Gukovskaya et al (2002) have demonstrated that ethanol (albeit at a high concentration of 100 mM) and acetaldehyde (0.1–10 mM) inhibited NF-κB activation, but up-regulated AP-1 activity in isolated pancreatic acini. The inhibitory effect of ethanol on NF-κB was thought to be mediated, at least in part, by acetaldehyde accumulation. The authors found that inhibition of the acetaldehyde metabolizing enzyme ALDH by cyanamide (which inhibits metabolism of acetaldehyde to acetate) further enhanced the ethanol-induced inhibition of NF-κB. These findings concur with the negative regulation of NF-κB by acetaldehyde reported in the liver (Lindros et al 1999). In contrast to the differential effects of acetaldehyde on the above transcription factors, FAEEs, the non-oxidative metabolites of ethanol, up-regulate both NF-κB and AP-1 activities in acinar cells. The same group has also reported that long-term ethanol feeding down-regulates NF-κB activity in the pancreas (Pandol et al 1999), a finding that supports their in vitro observations. However, the authors also found that chronic ethanol feeding sensitized the animals to CCKinduced pancreatitis. This was associated with potentiation of the CCK-induced pancreatic NF-κB activation and cytokine expression. They postulate that down-

ACETALDEHYDE AND PANCREATIC MAPK PATHWAYS

205

regulation of NF-κB by chronic ethanol administration may prevent increased cytokine production, thereby protecting the gland from pancreatitis (since it is well established that chronic ethanol feeding alone does not cause overt pancreatitis in rodents). However, an additional insult such as CCK infusion may shift the balance of ethanol metabolism towards the non-oxidative pathway, thereby leading to up-regulation of both NF-κB and AP-1 resulting in induction of cytokines and pancreatic injury. As noted earlier, reactive oxygen species are generated as a by-product of ethanol oxidation. Oxidant stress has been shown to occur in the pancreas after both acute (Altomare et al 1996) and chronic (Norton et al 1998) ethanol exposure and is thought to play an important role in the pathogenesis of alcoholic pancreatitis. Dabrowski et al (2000) have assessed the effects of the reactive oxygen species donors H2O2 and MND (2-methyl-1,4-naphthoquinone) on the MAPK pathway in acinar cells. Incubation of pancreatic acini with increasing doses of H2O2 resulted in a rapid, dose-dependent activation of ERK1/2, JNK and p38K, with maximum activation at 30 minutes followed by a gradual decline. MND exposure also caused activation of all three MAPK pathways, albeit at a slower rate with maximum activation of JNK at 60 minutes and that of ERK1/2 and p38K at 120 minutes. Importantly, the activation of MAP kinases by the above compounds was prevented by pre-incubation of acini with the anti-oxidant N-acetyl cysteine, indicating that the observed effects were specific for oxidant stress. Upstream regulation of oxidant stress-induced MAPK activation was also examined by Dabrowski and colleagues (Dabrowski 2003) and was found to be independent of PKC activity. The demonstrated effect of oxidant stress on MAPK activation in acinar cells may be important in the pathogenesis of alcoholic pancreatitis via stimulation of cytokine production as well as via disruption of the actin cytoskeleton (impairing acinar cell secretion and facilitating premature zymogen activation) as described earlier. Effects of ethanol and its metabolites on pancreatic stellate cell signalling Pancreatic stellate cells (PSCs) are now established as key effector cells in pancreatic fibrosis and there is increasing evidence to support a role for PSCs in alcoholinduced pancreatic fibrosis (Apte et al 2005). PSCs have been shown to exhibit alcohol dehydrogenase (ADH, a major ethanol oxidising enzyme) activity suggesting that these cells have the capacity to oxidise ethanol to acetaldehyde (Apte et al 2000). Thus, during ethanol consumption, PSCs in vivo may be exposed to acetaldehyde not only in a paracrine manner (i.e. acetaldehyde generated in adjacent acinar cells), but also directly via oxidation of ethanol within the cells themselves. In vitro studies have shown that PSCs are directly activated (as indicated by increased expression of the cytoskeletal protein α smooth muscle actin and increased production of extracellular matrix [ECM] proteins collagen, fibronectin and laminin) by clinically

206

APTE ET AL

relevant concentrations (10 mM and 50 mM) of ethanol (Apte et al 2000). Using 4methylpyrazole, an inhibitor of ADH, it has been demonstrated that the observed ethanol-induced activation of PSCs is mediated via the oxidation of ethanol to acetaldehyde and the subsequent generation of oxidant stress within the cells. Most recently, the signalling pathways mediating PSC activation by ethanol, acetaldehyde and oxidant stress have been identified. These studies have reported that all three families of the MAPK pathway (ERK1/2, JNK and p38K) are induced in PSCs exposed to ethanol or acetaldehyde (Masamune et al 2002, McCarroll et al 2003). This induction was observed at both early (15 minutes) and late (24 hours) time points of incubation. These findings concur with reports of acetaldehydeinduced MAPK activation in hepatic stellate cells (the counterparts of PSCs in the liver) (Anania et al 1999). An interesting result was noted with respect to ERK1/2 activation in PSCs exposed to ethanol or acetaldehyde (McCarroll et al 2003). The early increase in ERK1/2 activity at 15 minutes was followed by a return to control levels at 30 and 60 minutes and a second increase at 24 hours (Fig. 2). This biphasic activation of ERK1/2, though atypical, has been observed in other cell types such as vascular smooth muscle cells (Gurjar et al 2001). It is postulated that the early

E50 p-Elk-1

A200 p-Elk-1

250

Densitometry 150 units (% of control)

#

*

200

**

#

100 50 0

C 15 min 30 min 60 min 24 h **p < 0.02; **p < 0.01; #p < 0.05; n = 3 PSC preparations

FIG. 2. Induction of ERK1/2 activity by ethanol and acetaldehyde. Representative western blots and densitometry analyses for phosphorylated Elk-1 (downstream substrate of ERK1/2) showing a biphasic increase in ERK1/2 activity in pancreatic stellate cells exposed to 50 mM ethanol or 200 µM acetaldehyde, compared to control cells incubated with culture medium alone.

ACETALDEHYDE AND PANCREATIC MAPK PATHWAYS

207

induction of ERK1/2 may be important for cell migration whereas sustained ERK1/2 activation may be critical for cell proliferation. Signalling molecules upstream of the MAPK cascade in PSCs have also been examined with respect to their response to ethanol and acetaldehyde ( J. McCarroll 2005 PhD thesis, University of New South Wales, Australia). The PI3K pathway was found to be induced as early as at 5 minutes after exposure to the above compounds, with the ethanol-induced PI3K activation being sustained over 30 minutes. Similarly, both ethanol and acetaldehyde were found to activate all three classes of PKCs at 5 minutes of incubation, with the effect being sustained for 30 minutes in ethanol-treated cells. In order to determine the functional role of the MAPK, PI3K and PKC pathways in mediating ethanol and acetaldehyde-mediated PSC activation, specific inhibitors targeting individual pathways have been employed. These studies report that ERK1/2 inhibition (using U0126) and JNK inhibition (using SP600125) had no effect on ethanol or acetaldehyde-induced αSMA expression in PSCs (McCarroll et al 2003). However, inhibition of p38K (using SB203580) significantly prevented the induction of αSMA by ethanol and acetaldehyde (Masamune et al 2003, McCarroll et al 2003) (Fig. 3). With regard to ECM proteins, inhibition of all three families of

E50 p-Akt

A200 p-Akt 350

*

300

**

250 200

*

Densitometry 150 units (% of control)

100 50 0 Cont

5 min

30 min 120 min

24 h

**p < 0.02 ; *p < 0.001; n = 3 PSC preparations

FIG. 3. Activation of PI3 kinase by acetaldehyde. A representative western blot and densitometry analysis for PI3K activity (phosphorylated Akt) showing a significant increase in PI3 kinase activity at 5 minutes in pancreatic stellate cells exposed to 200 µM acetaldehyde, compared to control cells incubated with culture medium alone.

208

APTE ET AL

MAPK, abolished the ethanol and acetaldehyde-induced increase in collagen and fibronectin expression. Inhibition of the upstream pathways, PI3K and PKC (using the specific inhibitors wortmannin and calphostin C respectively) prevented the ethanol and acetaldehyde-induced ECM protein expression in PSCs ( J. McCarroll 2005 PhD thesis, University of New South Wales, Australia). Cross-talk between a variety of signalling molecules is a known feature of intracellular signalling. In order to determine whether such cross-talk exists between ethanol or acetaldehyde-induced ERK1/2 activation and PI3K or PKC, in vitro studies assessing ERK1/2 activation after inhibition of PI3K or PKC in PSCs exposed to ethanol or acetaldehyde have been performed ( J. McCarroll 2005 PhD thesis, University of New South Wales, Australia). Inhibition of PI3K by wortmannin abolished ERK1/2 activation in ethanol or acetaldehyde treated PSCs, indicating a regulatory role for PI3K in this process (Fig. 4). In contrast, PKC inhibition

Phospho-ERK1/2 Total-ERK1/2 Cont

160 140 120 100 Densitometry 80 units 60 (% of control) 40 20 0

E50

*

A20 0

Wt

E50+ A200+ Wt Wt

**

† #

Cont

E50

A200

Wt

#

E50 + A200 + Wt WT

*p<0.04, **p<0.03, †p<0.03, #p<0.001; n = 5 PSC preparations

FIG. 4. Effect of PI3K inhibition on ethanol- and acetaldehyde-induced ERK1/2 activation. A representative western blot and densitometry analysis for ERK1/2 activation demonstrating a significant increase in ERK1/2 activation in pancreatic stellate cells exposed to ethanol (50 mM) or acetaldehyde (200 µM). Inhibition of PI3K by wortmannin decreased basal ERK1/2 activation and notably, prevented the ethanol- and acetaldehyde-induced increase in ERK1/2 activation, suggesting the existence of cross-talk between the PI3K and ERK pathways in pancreatic stellate cells.

ACETALDEHYDE AND PANCREATIC MAPK PATHWAYS

209

by calphostin C significantly enhanced the ethanol and acetaldehyde-induced ERK1/2 activation in PSCs. A possible explanation for this finding relates to the known ability of PKC to induce the activity of MAPK phosphatases (enzymes that degrade MAPKs) (Valledor et al 2000). Inhibition of PKC could lead to decreased MAPK phosphatase activity (with a consequent decrease in inactivation of MAPK), thereby allowing an accumulation of phosphorylated ERK1/2 in PSCs. Thus, PKC may normally act as a negative regulator of ERK1/2 in PSCs. As noted earlier, oxidant stress is a known consequence of ethanol oxidation. Evidence of oxidant stress, as indicated by increased levels of the lipid peroxidation product malondialdehyde (MDA) has been reported in PSCs exposed to ethanol and acetaldehyde (Apte et al 2000). Importantly, this increase in ethanol- or acetaldehyde-induced oxidant stress within PSCs was prevented by pre-treatment of the cells with the antioxidant vitamin E. Exposure of PSCs to the pro-oxidants FeSO4/ascorbic acid or H2O2 has been reported to induce activation of PSCs as assessed by αSMA expression and/or ECM protein synthesis (Apte et al 2000). Recent studies have demonstrated that all three MAPK families (ERK1/2, JNK and p38K) are activated in PSCs exposed to FeSO4/ascorbic acid ( J. McCarroll 2005 PhD thesis, University of New South Wales, Australia) or H2O2 (Kikuta et al 2006) or the lipid peroxidation product hydroxynonenal (HNE) (Kikuta et al 2004). McCaroll ( J. McCarroll 2005 PhD thesis, University of New South Wales, Australia) has also demonstrated induction of PKC in PSCs subjected to oxidant stress. Notably, activation of these pathways was prevented by pre-treatment of PSCs with vitamin E. Furthermore, inhibition of MAPK and PKC by specific inhibitors was found to prevent the oxidant stress-induced increase in ECM protein expression in PSCs indicating that these pathways played a regulatory role in oxidant stress-related PSC activation. Interestingly, ERK1/2 activation in PSCs subjected to oxidant stress showed a biphasic pattern (activation at 15 minutes, return to control levels at 30 and 60 minutes and a second rise at 24 hours) similar to that seen in PSCs exposed to ethanol or acetaldehyde. This observation strengthens the concept that ethanol- or acetaldehyde-induced MAPK activation in PSCs is mediated by oxidant stress. Downstream targets of the MAPK pathway in PSCs, such as the transcription factors NF-κB and AP-1, have been the subject of some attention in recent years. Ethanol, acetaldehyde and the pro-oxidant H2O2 have been shown to activate AP-1 but not NF-κB in PSCs (Kikuta et al 2006, Masamune et al 2002). Interestingly, the observed induction of AP-1 (but not NF-κB) by ethanol mirrors the findings in ethanol-treated acinar cells reported by Gukovskaya and colleagues (Gukovskaya et al 2004), described earlier. ECM protein synthesis is known to be regulated by AP-1 (Ahn et al 2004) and the induction of this transcription factor may explain the observed increase in collagen and fibronectin production by PSCs exposed to ethanol, acetaldehyde or oxidant stress. The selective activation of AP-1 but not

210

APTE ET AL

NF-κB in PSCs exposed to the above compounds is different from the activation of both AP-1 and NF-κB observed in PSCs exposed to proinflammatory cytokines. The mechanisms responsible for the differential response of NF-κB and AP-1 in PSCs exposed to ethanol and acetaldehyde need further investigation. In summary, considerable progress has been made in recent years in identifying and characterising the MAPK pathway as well as the upstream regulators and downstream transcription factors that may mediate the effects of ethanol and its metabolites on pancreatic parenchymal (acinar) cells as well as non-parenchymal (stellate) cells. These developments provide important information regarding specific signalling molecules or pathways that may be therapeutically targeted in order to prevent or reverse the deleterious effects of ethanol on the pancreas. However, given the significant cross-talk and inherent redundancy in intracellular signalling systems, the challenge remains to translate these predominantly in vitro findings into effective treatment strategies in the in vivo situation. References Ahn JD, Morishita R, Kaneda Y et al 2004 Transcription factor decoy for AP-1 reduces mesangial cell proliferation and extracellular matrix production in vitro and in vivo. Gene Ther 11:916–923 Altomare E, Grattagliano I, Vendemiale G, Palmieri V, Palasciano G 1996 Acute ethanol administration induces oxidative changes in rat pancreatic tissue. Gut 38:742–746 Anania FA, Womack L, Potter JJ, Mezey E 1999 Acetaldehyde enhances murine alpha2(I) collagen promoter activity by Ca2+-independent protein kinase C activation in cultured rat hepatic stellate cells. Alcohol Clin Exp Res 23:279–284 Anderson KE, Jackson SP 2003 Class I phosphoinositide 3-kinases. Int J Biochem Cell Biol 35:1028–1033 Apte MV, Phillips PA, Fahmy RG et al 2000 Does alcohol directly stimulate pancreatic fibrogenesis? Studies with rat pancreatic stellate cells. Gastroenterology 118:780–794 Apte MV, Pirola RC, Wilson JS 2005 Molecular mechanisms of alcoholic pancreatitis. Dig Dis 23:232–240 Aroor AR, Shukla SD 2004 MAP kinase signaling in diverse effects of ethanol. Life Sci 74:2339–2364 Chen J, Ishac EJ, Dent P, Kunos G, Gao B 1998 Effects of ethanol on mitogen-activated protein kinase and stress-activated protein kinase cascades in normal and regenerating liver. Biochem J 334:669–676 Dabrowski A 2003 Exocrine pancreas; molecular basis for intracellular signaling, damage and protection—Polish experience. J Physiol Pharmacol 54 Suppl 3:167–181 Dabrowski A, Boguslowicz C, Dabrowska M, Tribillo I, Gabryelewicz A 2000 Reactive oxygen species activate mitogen-activated protein kinases in pancreatic acinar cells. Pancreas 21:376–384 Dempsey EC, Newton AC, Mochly-Rosen D et al 2000 Protein kinase C isozymes and the regulation of diverse cell responses. Am J Physiol Lung Cell Mol Physiol 279:L429–438 Gukovskaya AS, Mouria M, Gukovsky I et al 2002 Ethanol metabolism and transcription factor activation in pancreatic acinar cells in rats. Gastroenterology 122:106–118 Gukovskaya AS, Hosseini S, Satoh A et al 2004 Ethanol differentially regulates NF-kappaB activation in pancreatic acinar cells through calcium and protein kinase C pathways. Am J Physiol Gastrointest Liver Physiol 286:G204–213

ACETALDEHYDE AND PANCREATIC MAPK PATHWAYS

211

Gurjar MV, Deleon J, Sharma RV, Bhalla RC 2001 Role of reactive oxygen species in IL-1 betastimulated sustained ERK activation and MMP-9 induction. Am J Physiol Heart Circ Physiol 281:H2568–2574 Haber PS, Apte MV, Applegate TL et al 1998 Metabolism of ethanol by rat pancreatic acinar cells. J Lab Clin Med 132:294–302 Haber PS, Apte MV, Moran C et al 2004 Non-oxidative metabolism of ethanol by rat pancreatic acini. Pancreatology 4:82–89 Kikuta K, Masamune A, Satoh M, Suzuki N, Shimosegawa T 2004 4-hydroxy-2, 3-nonenal activates activator protein-1 and mitogen-activated protein kinases in rat pancreatic stellate cells. World J Gastroenterol 10:2344–2351 Kikuta K, Masamune A, Satoh M et al 2006 Hydrogen peroxide activates activator protein-1 and mitogen-activated protein kinases in pancreatic stellate cells. Mol Cell Biochem 291: 11–20 Koteish A, Yang S, Lin H, Huang X, Diehl AM 2002 Chronic ethanol exposure potentiates lipopolysaccharide liver injury despite inhibiting Jun N-terminal kinase and caspase 3 activation. J Biol Chem 277:13037–13044 Lee YJ, Aroor AR, Shukla SD 2002 Temporal activation of p42/44 mitogen-activated protein kinase and c-Jun N-terminal kinase by acetaldehyde in rat hepatocytes and its loss after chronic ethanol exposure. J Pharmacol Exp Ther 301:908–914 Lindros KO, Jokelainen K, Nanji AA 1999 Acetaldehyde prevents nuclear factor-kappa B activation and hepatic inflammation in ethanol-fed rats. Lab Invest 79:799–806 Masamune A, Kikuta K, Satoh M, Satoh A, Shimosegawa T 2002 Alcohol activates activator protein-1 and mitogen-activated protein kinases in rat pancreatic stellate cells. J Pharmacol Exp Ther 302:36–42 Masamune A, Satoh M, Kikuta K et al 2003 Inhibition of p38 mitogen-activated protein kinase blocks activation of rat pancreatic stellate cells. J Pharmacol Exp Ther 304:8–14 McCarroll JA, Phillips PA, Park S et al 2003 Pancreatic stellate cell activation by ethanol and acetaldehyde: is it mediated by the mitogen-activated protein kinase signaling pathway? Pancreas 27:150–160 Norton ID, Apte MV, Lux O et al 1998 Chronic ethanol administration causes oxidative stress in the rat pancreas. J Lab Clin Med 131:442–446 Pandol SJ, Periskic S, Gukovsky I et al 1999 Ethanol diet increases the sensitivity of rats to pancreatitis induced by cholecystokinin octapeptide. Gastroenterology 117:706–716 Pearson G, Robinson F, Beers Gibson T et al 2001 Mitogen-activated protein (MAP) kinase pathways: regulation and physiological functions. Endocr Rev 22:153–183 Siegmund E, Luthen F, Kunert J, Weber H 2004 Ethanol modifies the actin cytoskeleton in rat pancreatic acinar cells—comparison with effects of CCK. Pancreatology 4:12–21 Valledor AF, Xaus J, Comalada M, Soler C, Celada A 2000 Protein kinase C epsilon is required for the induction of mitogen-activated protein kinase phosphatase-1 in lipopolysaccharidestimulated macrophages. J Immunol 164:29–37 Williams JA, Sans MD, Tashiro M et al 2002 Cholecystokinin activates a variety of intracellular signal transduction mechanisms in rodent pancreatic acinar cells. Pharmacol Toxicol 91: 297–303

DISCUSSION Albano: What are the signals upstream on the kinase cascade that you showed in your presentation? What is happening with ethanol and acetaldehyde? You drew some receptors in your scheme, and it is known that PI3K is coupled with the EGF

212

APTE ET AL

receptor. I wonder whether acetaldehyde or ethanol could mediate some of their effects in acinar cells by favouring the transduction of EGF-mediated signals. Apte: Acinar cells haven’t been studied, and for the stellate cells this is the first work we have done, so we will have to look at what happens when we combine growth factors with ethanol and acetaldehyde. Because ethanol and acetaldehyde are fairly easily diffusible, it may be just direct effects: it doesn’t have to be receptor mediated. Albano: Do you see apoptosis when you block the PI3K pathway? You still have ethanol activating both the MAP kinase and Jun kinase pathways. In many cell types the prolonged activation of these kinases leads to cell death. Apte: No, we don’t see apoptosis in these cells. These are cells incubated in ethanol for up to 24 h. The pancreatic stellate cell field is still very new compared with the hepatic stellate cell field. We still need to characterize all the apoptotic pathways and the receptors. Stellate cells are known to have FAS ligand receptors, so we know that they can undergo apoptosis. In this system we didn’t find any apoptotic cells. Ren: There was a paper a few years back (Lali et al 2000) indicating that the SB 203580 compound can inhibit PI3K and protein kinase B. Apte: We have used another PI3K inhibitor with similar results. Whether that other inhibitor also inhibits PI3K I am unsure. We haven’t revisited PI3K activity in the presence of this inhibitor in stellate cells. One of the problems with cell signalling studies is that the inhibitors may not be as specific as we think they are. Seitz: There are obviously similarities with respect to the liver, as well as some differences. The results with AP-1 in the pancreas are similar to what we observe in the liver. Is this related to retinoic acid metabolism? This would give a similar mechanism. The NF-κB is obviously different. Are there FFAs in the pancreas? We know from the liver that FFAs may switch on NF-κB and TNFα. This may be different in the pancreas. Apte: There are FFAs in the pancreas, particularly in alcohol-exposed cells. Fatty acid ethyl esters (FAEEs) actually produce FFAs. This is one of the other toxic mechanism of FAEEs. With the vitamin A question, we have looked at the effects on MAPK signalling, but we are yet to get to the transcription factor. Preedy: Clinically, in pancreatitis do you see apoptosis? Apte: It depends on the model and the stage. Obviously, we can’t get clinical specimens of human pancreatitis. If you are looking at acute pancreatitis then there is apoptosis. In a more severe model you get more necrosis and less apoptosis. I’m sure it occurs clinically. Eriksson: Could you use a model with chronic methylpyrazole, to try to get away from the alcohol metabolism and keeping high ethanol concentrations? In those conditions the fatty acids should increase a lot. We have a paper that just came out (Best et al 2006) which showed that 4-methylpyrazole in conjunction with alcohol increased the fatty acids levels in humans.

ACETALDEHYDE AND PANCREATIC MAPK PATHWAYS

213

Apte: There is a shift from the oxidative to the non-oxidative pathway. This has been suggested also by Werner’s group. He has looked at FAEEs and acetaldehyde production in isolated pancreatic acini. Inhibiting ADH by 4-methylpyrazole shifted it to the FAEE production. The only thing I wasn’t sure about is that although they used the inhibitors, they didn’t go back and measure ADH and CYP2E1 levels to see whether all those inhibitors were working. Eriksson: This can be seen by monitoring the alcohol concentration. Apte: They didn’t do this. This is why FAEEs may have more of an important role in the pancreas, given that the FAEE synthetic capacity in the pancreas seems much higher than in the liver, per gram or organ. Accumulation of these toxic products might be the cause of pancreatitis. Rao: A comment regarding the acinar data: actin staining didn’t show convincingly that the actin organization is disrupted by alcohol, because compared to control the actin stain was higher in alcohol-treated cells. The zymogen transport from the ER to Golgi to the apex of the cell mostly involves microtubules rather than actin. The actin plays a role just at the apical part of the cell. Perhaps you should look at microtubules because we know acetaldehyde can form adducts with tubulin. Apte: In fact, acetaldehyde-induced tubulin dysfunction has been described. My interest in the actin was only because it might be one of the mechanisms. Rao: In alcoholic pancreatitis, is the duct affected? Fibrosis leading to occlusion of the duct can also regurgitate the proteases. Apte: In late stages, when you get to see sections of alcoholic pancreatitis the fibrosis is periductal as well as periacinar. Is it the duct obstruction that causes backflow and then causes acinar damage, or is the initiation in the acinar compartment? I think both happen. Interestingly, we have a model of alcoholic pancreatitis in rats fed alcohol for 8 weeks and then injected with LPS intravenously. This results in acinar necrosis, and if we inject repeatedly the result is fibrosis. This is periacinar. Shukla: In relation to the communication between the acinar cells and stellate cells, how much is known about the function of these MAP kinases in stimulating a signal from one cell to another, to stimulate fibrosis? Apte: It is known that in response to alcohol and acetaldehyde, PSCs make more TGFβ. We’ve shown this, and others have shown that there is increased production of TNFα and interleukin (IL)6. It is possible that these then act on acinar cells. There is close apposition between the stellate cells and the basolateral side of acinar cells, which might indicate another function. We are hoping to show that these cells might play a role in acinar cell secretion, which is a complicated story. Shukla: Alcoholics don’t normally have pancreatic problems. There must be some additional trigger factor. Have you looked for this? Could it be something like LPS? Apte: That is what we are starting to think. If you feed a rat alcohol and give it endotoxin, this produces disease, whereas alcohol alone doesn’t.

214

APTE ET AL

Shukla: It would be interesting to see whether MAP kinase pathways are altered after that insult. Niemelä: In humans, hypertriglyceridaemia has been suggested as a risk factor for acute pancreatitis. Have you had a chance to manipulate the levels of fatty acid ethyl esters to see what happens? Apte: We have done a human study looking at alcoholics with pancreatitis and those without. We examined serum triglyceride levels in response to a dose of ethanol. The proper use of the controls is very important. If you use alcoholics with pancreatitis versus non-alcoholics you get a big difference in serum triglyceride levels in the alcoholics with pancreatitis compared with the non-drinkers. But as soon as we put in the right controls—alcoholics with the same cumulative consumption without pancreatitis—those serum triglyceride levels superimpose on each other. We don’t think that hypertriglyceridaemia is a risk factor for alcoholic pancreatitis in this setting. Niemelä: What about extra iron? Apte: Theoretically this is feasible. People who have haemochromatosis have extra iron, and alcohol is supposed to enhance the progression of this. In the pancreas, not much is known about iron. Crabb: With regard to the endotoxin experiments, are there macrophages that live like the Kupffer cells in the pancreas? Apte: That’s an important question. We were wondering whether the stellate cells are macrophages in the pancreas. They’ve been shown to have phagocytic functions. Crabb: People have fed rats alcohol and given them endotoxin. It is a model of endotoxin-induced alcoholic hepatitis. This experiment has been done inadvertently by other people: do you think there was pancreatitis in their animals which they didn’t look for? Apte: That is exactly what happened to me. When I did my Master’s thesis it was on an alcohol feeding model that was injected with i.v. endotoxin. The result was alcoholic hepatitis lesions. I never looked at the pancreas, but I could have done and saved myself some time. Models are just models, and all we can do is use them to try to figure out molecular mechanisms of disease. Our excitement is because we now have a temporal model. Albano: Ron Thurman’s group reported on an in vivo model of chronic pancreatitis by using their procedures of intragastric alcohol feeding together with a diet rich in unsaturated fatty acids (Kono et al 2001). This model might be useful to verify your observations in an in vivo setting. Seitz: This is a bit different to the real situation in life. There is either cirrhosis or pancreatitis. Everyone will ask this: why do some have pancreatitis but not cirrhosis, and the other way around. Apte: This is what we are trying to answer.

ACETALDEHYDE AND PANCREATIC MAPK PATHWAYS

215

Preedy: Could the difference between Ron Thurman’s model and yours be to do with the fact that you use Lieber-DeCarli whereas Ron uses intragastric administration of alcohol? Apte: The problem with the intragastric model is that it is completely unphysiological. No one drinks for 24 h a day. Albano: This is true, however, intragastric feeding is at the moment the only procedure that allows to attain in rodents blood alcohol levels high enough to cause appreciable liver injury. There are several differences between the intragastric feeding procedures used by the different groups. In addition the way the animal or the diet is handled might possibly justify why high levels of circulating endotoxins have been observed in alcohol-fed rat by the Thurman’s group, whereas Badger’s group reported a modest rise of plasma endotoxins in their animals. Even though liver injury is evident both in the presence or in the absence of endotoxaemia, it is possible that endotoxins might speed up the development of hepatic lesions by worsening inflammation. The use of intragastric feeding has also shown that the amount of carbohydrates (Ronis et al 2004) and the presence of unsaturated fatty acids in the diet are critical for the development of pathology (Nanji et al 1995). Apte: The Lieber-DeCarli diet controls for the high fat, and the carbohydrate is replaced by ethanol. Preedy: There is a Finnish group who used high fat in a Lieber-DeCarli-type diet. Instead of isocaloric carbohydrate they are using isocaloric fat. They claim to get the same lesions in the liver as you would with the more severe continuous models. It may be more to do with the diet than the route. Ren: Light to moderate drinking helps maintain a low sugar level in patients with diabetes. Is there any effect of acetaldehyde on insulin secretion? Apte: I don’t know. Seitz: This is probably because of a decrease in peripheral insulin resistance. Crabb: Do you see a link between the stress pathways you looked at and what ultimately seems to have to happen in acute pancreatitis, such as secretory vesicles getting in the wrong place and being autoactivated? I don’t see the connection, but I don’t work on the pathways. Apte: I’d have to look at the literature. People have looked at protein transport but not necessarily the signalling pathways that could interrupt or facilitate the vesicle movement from the Golgi. Aranda: Have you studied the role of Ca2+ in signal transduction in your system? Apte: We haven’t, but Ole Petersen’s group has looked at Ca2+ release in acinar cells exposed to FAEEs. There is a sustained increase of Ca2+ in these cells. They propose that the FAEEs get converted to fatty acids which then affect the mitochondria, preventing ATP synthesis. So the ATP synthase-dependent pumps are

216

APTE ET AL

not working, resulting in a sustained increase in Ca2+ which eventually causes mitochondrial depolarization and cell death. Aranda: Mutations in two Ca2+/calmodulin-dependent kinases NK1 and 2 have a role in acetaldehyde and ethanol resistance in yeast. Apte: There are many studies yet to be done. References Best CA, Sarkola T, Eriksson CJ, Cluette-Brown JE, Laposata M 2006 Increased plasma fatty acid ethyl ester levels following inhibition of oxidative metabolism of ethanol by 4-methylpyrazole treatment in human subjects. Alcohol Clin Exp Res 30:1126–1131 Kono H, Nakagami M, Rusyn I et al 2001 Development of an animal model of chronic alcoholinduced pancreatitis in the rat. Am J Physiol Gastrointest Liver Physiol 280:G1178–G1186 Lali FV, Hunt AE, Turner SJ, Foxwell BM. The pyridinyl imidazole inhibitor SB203580 blocks phosphoinositide-dependent protein kinase activity, protein kinase B phosphorylation, and retinoblastoma hyperphosphorylation in interleukin-2-stimulated T cells independently of p38 mitogen-activated protein kinase. J Biol Chem 275:7395–7402 Nanji AA, Sadrzadeh SMH, Yang EK, Fogt F, Maydani M, Dannenberg AJ 1995 Dietary saturated fatty acids: A novel treatment for alcoholic liver disease. Gastroenterology 109:547–554 Ronis MJJ, Hakkak R, Korourian S et al 2004 Alcoholic liver disease in rats fed ethanol as part of oral or intragastric low-carbohydrate diets. Exp Biol Med 229:351–360

Acetaldehyde alters MAP kinase signalling and epigenetic histone modifications in hepatocytes Shivendra D. Shukla, Youn Ju Lee, Pil-hoon Park and Annayya R. Aroor Department of Medical Pharmacology & Physiology, University of Missouri School of Medicine, Columbia, MO 65212, USA

Abstract. Although both oxidative and non-oxidative metabolites of ethanol are involved in generating ethanol metabolic stress (Emess), the oxidative metabolite acetaldehyde plays a critical role in the cellular actions of ethanol. We have investigated the effects of acetaldehyde on p42/44 MAP kinase, p46/p54 c-jun N-terminal kinase ( JNK1/JNK2) and p38 MAP kinase in hepatocytes. Acetaldehyde caused temporal activation of p42/44 MAPK followed by JNK, but the activation of the p42/44 MAPK was not a prerequisite for the JNK activation. Activation of JNK1 by acetaldehyde was greater than JNK2. Ethanol and acetaldehyde activated JNK have opposing roles; ethanol-induced JNK activation increased apoptosis whereas that by acetaldehyde decreased apoptosis. Acetaldehyde also caused histone H3 acetylation at Lys9 and phosphorylation of histone H3 at Ser10 and 28, the latter being dependent on p38 MAP kinase. Phosphorylation at Ser28 was higher than at Ser10. Thus acetaldehyde distinctively alters MAP kinase signalling and histone modifications, processes involved in transcriptional activation. 2007 Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Foundation Symposium 285) p 217–228

Cellular actions of ethanol are complex. Chronic ethanol causes injury to multiple organ systems and has serious medical and public health implications (Room et al 2005). Although the effects of ethanol are widely known, the mechanisms of ethanol action at the molecular level are poorly understood. Therapeutic tools to control or reverse the ethanol-induced cellular damages are also lacking. Actions of ethanol are unique in the sense that both ethanol and its metabolites have their own effects and can modulate the responses of other agents. This ‘double edge’ effect makes ethanol actions multifaceted. Ethanol is oxidatively metabolized by alcohol dehydrogenase (ADH) to acetaldehyde which is then metabolized by aldehyde dehydrogenase (ALDH) to acetate. Non-oxidatively, phosphatidylethanol 217

218

SHUKLA ET AL

(PEth) is exclusively produced from ethanol by a phospholipase D-mediated reaction (Shukla et al 2001). Another non-oxidative metabolite is fatty acid ethyl ester (FAEE). Ethanol also causes generation of reactive oxygen species (ROS) and superoxide dismutases (SOD) are considered key players in this response. Thus, by analogy, actions of ethanol are like ‘cluster bombs’ for the cell with acetaldehyde being the major damaging entity. It is a combination of these oxidative and nonoxidative stresses that causes injury to cells and has been termed as ‘Ethanol metabolic stress’ (Emess) (Shukla & Aroor 2006). Acetaldehyde is the major component of Emess. Liver is a major organ for ethanol metabolism and is therefore susceptible to the deleterious effects of ethanol and acetaldehyde leading to alcoholic liver injury. However, it is only recently that the molecular mode of action of acetaldehyde has begun to be investigated. Such mechanistic studies are essential to identify molecular targets for the development of therapeutic tools to prevent and control the damaging effects of acetaldehyde. In this context we have investigated the influence of acetaldehyde on the cellular MAP kinase signalling cascade (Aroor & Shukla 2004) and how this may relate to the epigenetic modifications (acetylation, phosphorylation) in histones (Strahl & Allis 2000) using primary cultures of rat hepatocytes as the model.

Results and discussion Acetaldehyde effects on p42/44 MAP kinase, JNK and p38 MAP kinase The true concentration of either ethanol or acetaldehyde in different organs in vivo after alcohol consumption is not well established. The in vivo peripheral blood concentration of ethanol in chronic alcoholics is normally <50 mM. But the in vivo concentration in situ in liver remains poorly understood. In pigs orally administered with ethanol, portal venous blood ethanol was twofold higher than in peripheral blood (Elmer et al 1982, Luca et al 1997). Furthermore, the hepatic venous blood is diluted about five times in the caval vein (Nuutinen et al 1984). In the liver, blood supply is high through the mesenteric portal venous system and acute administration of ethanol increases portal blood flow by 40–60% (Orrego et al 1988). Thus, liver cells are exposed to higher concentrations and higher amounts of ethanol than the levels in peripheral blood. Interestingly the upper limits of ethanol in chronic alcoholics can be 100–300 mM (Deitrich & Harris 1996). Thus in vitro use of ethanol in the range of 50–200 mM is not far from that observed in vivo in chronic alcohol abusers. A similar argument can be put forward for the concentrations of acetaldehyde used experimentally in vitro. We have previously shown that

ACETALDEHYDE, MAP KINASES AND HISTONE MODIFICATIONS

219

acetaldehyde increases hepatocyte JNK activity in a dose (0.2–5 mM) dependent manner (Lee et al 2002). Use of acetaldehyde at 1 or 5 mM can be viewed as a very high concentration. However, the following lines of evidence support the rationale that higher concentrations of acetaldehyde are pathophysiologically relevant. Chronic alcoholics exhibit acetaldehyde levels as high as 30 µM in peripheral vein (Salaspuro 1991). The acetaldehyde concentration in hepatic vein is 10–30 times higher than that in peripheral vein in humans (Nuutinen et al 1984). In animals acetaldehyde concentration in portal vein is also 10 times higher than in peripheral vein (Matysiak-Budnik et al 1996). Thus, the estimated concentration in liver can be close to 1 mM. Furthermore, liver can be exposed to additional acetaldehyde in the portal vein absorbed from the colon. In fact, the level of acetaldehyde in the colon can reach up to 3 mM after a 1.5 g/kg dose of ethanol in rats (Koivisto & Salaspuro 1997). In vitro, serum has the capacity to bind more than 447 mM acetaldehyde and the detected concentration of acetaldehyde represents free rather than total acetaldehyde including its adducts (Brecher et al 1997). Taken together, it is apparent that millimolar concentrations of acetaldehyde are relevant in in vivo in liver and justify use of these levels in vitro studies with hepatocyte cultures. In primary cultures of rat hepatocytes, acetaldehyde in the range of 0.2–5 mM causes activation of MAP kinases in a distinctive fashion. A rapid activation of p42/44 MAP kinase is followed by activation of the JNK and p38 MAP kinases. The activated levels of these kinases then decrease to the basal level at 2 h (Fig. 1). The protein levels of these kinases do not change under these conditions. The apoptotic changes, as monitored by caspase 3 activation, occur after 2 h. Inhibition of the p42/44 MAP kinase by U0126 had no inhibitory effect on the acetaldehyde-induced JNK activation. Thus p42/44 MAP kinase activation by acetaldehyde is not a pre-requisite for the JNK activation and these are likely to be independent events. One of the noticeable differences in the JNK1 versus JNK2 activation by acetaldehyde was the much higher level of JNK1 activation (about three–fourfold) over JNK2. It may be noted that JNK1 overactivation has been proposed to be associated with the development of steatohepatitis, insulin resistance and metabolic syndrome (Shattenberg et al 2006). Acetaldehyde also induced apoptotic changes in hepatocytes monitored by nuclear staining with Hoechst dye or by caspase 3 activation. Of interest is the observation that pretreatment of hepatocyes with SP600125 (a JNK inhibitor) enhanced the effect of the acetaldehyde-induced apoptosis. This suggests that JNK activation by acetaldehyde has an anti-apoptotic effect. This is in contrast to ethanol, where JNK activation appears to be pro-apoptotic (Lee & Shukla 2005). Other characteristics of the acetaldehyde-induced JNK activations are that PKC inhibitor, tyrosine kinase inhibitor or the antioxidant N-acetylcysteine had little

220

SHUKLA ET AL 16

P-p42/44MAPK P-JNK P-p38MAPK Caspase-3

Fold Increase

14 12 10 8 6 4 2 0 15 min

30 min

2h

Time FIG. 1. Pattern of MAP kinase activation and apoptosis by acetaldehyde. Hepatocytes were treated with 5 mM acetaldehyde for indicated times. The activations of p42/44 MAPK (P-p42/ 44MAPK), JNK (P-JNK), p38 MAPK (P-p38 MAPK) and caspase 3 were detected by western blot and quantitated by Quantity One software (Bio-Rad). Values represent fold increase over control (control value = 1). Data are representative of 3–4 separate experiments.

effect on its activation. But p42/44 MAPK activation by acetaldehyde was inhibited by inhibitors of PKC and tyrosine kinase (Lee & Shukla 2005). These findings indicate that the acetaldehyde effects are, in many ways, quite unique and are dependent on only selected pathways. This may prove to be advantageous in the targeting of acetaldehyde actions. Acetaldehyde modifies acetylation and phosphorylation of histone H3 Histone modifications constitute important aspects of the epigenetic alterations associated with cellular processes, particularly the opening of the DNA template for transcriptional activation (Strahl & Allis 2000). We have investigated the acetaldehyde-induced changes in histone H3 acetylation and phosphorylation in hepatocytes. Acetaldehyde caused acetylation of H3 at Lys9 and phosphorylation of H3 at Ser10 and Ser28 in a dose-dependent manner (Lee & Shukla 2006). When one compares the degree of changes in these two modifications by acetaldehyde, it was apparent that at 1 h of exposure of cells to acetaldehyde the acetylation of H3-Lys9 was negligible. On the other hand the Ser10 and Ser28 phosphorylations were dramatically higher at about two- and fourfold above control levels, respectively (Fig. 2). At 24 h of the exposure the phosphorylation levels decreased to almost basal level but the acetylation increased about twofold. These data suggest

ACETALDEHYDE, MAP KINASES AND HISTONE MODIFICATIONS

Fold Increase

6

221

P-H3 Ser10 P-H3 Ser28 Ac-H3 Lys9

5 4 3 2 1 0 1

2

4

24

Time (h) FIG. 2. Acetylation and phosphorylation of histone H3 by acetaldehyde. Hepatocytes were treated with 5 mM acetaldehyde for different time periods. Histone H3 phosphorylation at serine 10 (P-H3 Ser10), or serine-28 (P-H3 Ser28) and acetylation at lysine 9 (Ac-H3 Lys9) were monitored separately by western blotting. Values represent fold increase over control (control value = 1). Data are from 2–3 separate experiments with bars representing standard deviation.

that (a) acetaldehyde-induced phosphorylation is rapid compared to the acetylation and (b) that Ser28 phosphorylation is more prominent than Ser10, indicating a difference in their responses to acetaldehyde. It is of further interest to mention that, compared to ethanol-induced changes in histone phosphorylation at these serine residues, the acetaldehyde-induced changes are much greater. However, ethanol-induced acetylations are more pronounced than the acetaldehyde-induced acetylation of histone H3 at Lys9. Histone acetylation and phosphorylations were affected by pharmacological manipulations of MAP kinases. It was observed that inhibition of p42/44 MAP kinase by U0126 (MEK inhibitor), or JNK by SP600125, decreased histone acetylation by ethanol, but the inhibition of p38 MAP kinases was without effect. Intriguingly, the acetate-induced H3-Lys9 acetylations were not influenced by the inhibitors U0126 or SP600125 and raise the issue of separate mechanisms for the ethanol-induced and acetate-induced acetylations, as far as it relates to their sensitivities to MAP kinase pathways (Park et al 2005). However, inhibition of p38 MAP kinase blocked acetaldehyde-induced histone phosphorylations at H3 Ser10 and 28. Thus it appears that acetaldehyde-activated MAP kinase signalling pathways have link(s) to the histone modifications but in a MAP kinase ‘type’ selective manner. Other lines of evidence also strongly point towards the ethanol metabolic dependence of the histone acetylations. Treatment of hepatocytes with 4-methylpyrazole, an ADH inhibitor, substantially reduced the ethanol-induced histone H3-Lys9

222

SHUKLA ET AL

acetylations (Park et al 2003). Ethanol had small influence on H3 acetylation in HepG2 cells that are deficient in ADH. These observations indicate the involvement of acetaldehyde in histone acetylation. Furthermore, exposure of hepatocytes to acetate, a product generated by ALDH action on acetaldehyde, caused an increase in H3 acetylation quite similar to that observed with ethanol. Thus it is becoming apparent that ethanol-induced alterations in histone acetylation can be largely attributed to acetaldehyde and its metabolite. Ethanol-induced histone H3-Lys9 acetylation also occurs in vivo in rat liver and other organs (Kim & Shukla 2005, 2006). Modifications in histones by ethanol have recently been shown to be relevant to transcriptional activation. It was observed that ethanol treatment of rat hepatocytes increased the association of acetylated Lys9-histone H3 with the alcohol dehydrogenase 1 (ADH1) gene promoter as determined by chromatin immunoprecipitation assay. A concomitant increase in the mRNA expression of ADH1 was also noted (Park et al 2005). These data can be conceptually argued as follows. Ethanol increases expression of ADH1 via histone modifications, and this will generate more acetaldehyde and acetate. This in turn, will further increase histone acetylation and additional ADH1 expression. Thus, ‘an amplification cycle’ can be envisaged based on the histone modifications by acetaldehyde. Existence of such a positive feedback loop may magnify the effects of ethanol and acetaldehyde. Details of this scenario and its mechanistic insight will be interesting to tease out in future studies. Summary Acetaldehyde, the major active metabolite of ethanol metabolism, has differential effects on the activation of the three major members of the MAP kinase family. There are distinctive features of their activations by acetaldehyde. The JNK1 activation was greater than JNK2. The histone H3-Ser28 phosphorylation was much higher than H3-Ser10. Furthermore, the acetylations of histone H3 were at a slower rate than their phosphorylation. These differences highlight the unique aspects of acetaldehyde actions on liver cells and should be considered as important molecular targets for the therapeutic control of the actions of the acetaldehyde. It is also concluded that there is a link between acetaldehyde modulation of MAP kinases and the post-translational modifications of the nuclear histone (Fig. 3). These acetaldehyde-induced epigenetic alterations may have an important bearing upon the genetic expressions related to the pathogenesis of the liver.

Acknowledgements Work in authors laboratory is supported by grants # AA 11962 and AA 14852 from the National Institute on Alcohol Abuse & Alcoholism (NIAAA) of the NIH (USA). We thank Mr Daniel Jackson for technical help.

ACETALDEHYDE, MAP KINASES AND HISTONE MODIFICATIONS

223

Ethanol

Emess

Acetaldehyde

Oxidative stress

PEth FAEE

Acetate

MAPK JNK 1/2 signaling

p38 MAPK

p42/44 MAPK

MSK1 Epigenetic histone modifications

Histone acetylation

?

Histone phosphorylation ?

Gene expression Alcoholic liver Apoptosis injury Proliferation

Inflammation Insulin resistance Cirrhosis

FIG. 3. A schematic diagram of the possible interactions among Emess, MAP kinases and histone modifications. In this simplified diagram other histone modifications, e.g. methylations, are not shown. MSK1, mitogen and stress activated kinase 1.

References Aroor AR, Shukla SD 2004 MAP kinase signalling in diverse effects of ethanol. Life Sci 74:2339–2364 Brecher AS, Hellman K, Basista MH 1997 A perspective on acetaldehyde concentrations and toxicity in man and animals. Alcohol 14:493–496 Deitrich RA, Harris RA 1996 How much alcohol should I use in my experiments? Alcohol Clin Exp Res 20:1–2 Elmer O, Bengmark S, Göransson G, Sundqvist K, Söderström N 1982 Acute portal hypertension after gastric administration of ethanol in the pig. Eur Surg Res 14:298–308 Kim J, Shukla SD 2005 Histone H3 modifications in rat hepatic stellate cells by ethanol. Alcohol Alcohol 40:367–372 Kim J, Shukla SD 2006 Acute in vivo effect of ethanol (binge drinking) on histone H3 modifications in rat tissues. Alcohol Alcohol 41:126–132 Koivisto T, Salaspuro M 1997 Effects of acetaldehyde on brush border enzyme activities in human colon adenocarcinoma cell line caco-2. Alcohol Clin Exp Res 21:1599–1605 Lee YJ, Shukla SD 2005 Pro- & anti- apoptotic role of JNK in ethanol and acetaldehyde exposed rat hepatocytes. Eur J Pharm 508:31–45 Lee YJ, Shukla SD 2006 Nuclear activation of p38 MAPK and histone H3 phosphorylation by ethanol and acetaldehyde in rat hepatocytes. FASEB J 20:A1122

224

SHUKLA ET AL

Lee YJ, Aroor A, Shukla SD 2002 Temporal activation of p42/p44 MAP kinase and c-Jun Nterminal kinase by acetaldehyde in rat hepatocytes and its loss after chronic ethanol exposure. J Pharmacol Exp Ther 301:908–914 Luca A, Garcia-Pagan JC, Bosch J et al 1997 Effect of ethanol consumption on hepatic hemodynamics in patient with alcoholic cirrhosis. Gastroenterology 112:1284–1289 Matysiak-Budnik T, Jokelainen K, Karkkainen P, Makisalo H, Ohisalo J, Salaspuro M 1996 Hepatotoxicity and absorption of extrahepatic acetaldehyde in rats. J Pathol 178:469–474 Nuutinen HU, Salaspuro M, Valle M, Lindros KO 1984 Blood acetaldehyde concentration gradient between hepatic and antecubital venous blood in ethanol intoxicated alcoholics and control. Eur J Clin Invest 14:306–311 Orrego H, Carmichael FJ, Israel Y 1988 New insights on the mechanism of the alcohol-induced increase in portal blood flow. Can J Physiol Pharmacol 66:1–9 Park P, Miller R, Shukla SD 2003 Acetylation of histone H3 at lysine 9 by ethanol in rat hepatocytes. Biochem Biophys Res Commun 306:501–504 Park P, Lim RW, Shukla SD 2005 Involvement of histone acetyltransferase (HAT) in ethanol induced acetylation of histone H3 in hepatocytes: potential mechanism for gene expression. Am J Physiol 289:G1124–1136 Room R, Babor T, Rehm J 2005 Alcohol and public health. Lancet 365:519–530 Salaspuro M 1991 Epidemiological aspects of alcohol and alcoholic liver disease, ethanol metabolism, and pathogenesis of alcoholic liver injury. In: McIntyre N, Benhamou JP, Bircher J, Rizzeto M, Rodes J (eds) Oxford textbook of clinical hepatology. Oxford Press, p 791–810 Schattenberg JM, Singh R, Wang Y et al 2006 JNK1 but not JNK2 promotes the development of steatohepatitis in mice. Hepatology 43:163–172 Shukla SD, Aroor AR 2006 Epigenetic effects of ethanol in liver and GI injury. World J Gastroenterol 12:5265 –5271 Shukla SD, Sun G, Wood WG, Savolainen MJ, Alling C, Hoek JB 2001 Ethanol and lipid metabolic signalling. Alcohol Clin Exp Res 25:33S–39 Strahl BD, Allis CD 2000 The language of covalent histone modifications. Nature 403:41–45

DISCUSSION Preedy: I have a query about some of the methodology you used to show the transient effects of acetaldehyde. When I was doing a study using a hormone in an in vitro system, I lost more than 99% of it because it was absorbed onto the glass. Have you actually measured the acetaldehyde levels throughout the period of culture? When you add high loads of acetaldehyde, how much is being bound to the plastic and proteins, and how much is free? Shukla: Good question. In general, one third of the acetaldehyde is used after an hour in primary culture of rat hepatocytes. A very small fraction evaporates because parafilm is used to seal the dishes. Albano: You showed that acetaldehyde and ethanol have opposite effects in causing apoptosis in your cell system. How do you explain these effects? The differential modulation of MAP kinase pathways won’t explain this. Have you looked at other signalling pathways, to see whether they are activated differentially by ethanol or acetaldehyde? Shukla: We have been asking this question ourselves. It is possible that under some conditions acetaldehyde may have a transient protective effect. This is a

ACETALDEHYDE, MAP KINASES AND HISTONE MODIFICATIONS

225

hypothesis at the moment, although I have no evidence for it. The long-term sustained activation of MAPK may be damaging. In terms of the PI3K we haven’t monitored this, but other labs have shown that ethanol and acetaldehyde can modulate PI3K. Albano: You could look at the effect of acetaldehyde in the presence of the PI3K inhibitor wortmannin, to investigate whether the down-modulation of PI3K-dependent signals has a protective effect. Shukla: Wortmannin is not very specific. The interesting question is not what is upstream of the MAPKs, but rather the downstream effects, which relates to the fibrosis and other changes. Apte: Just for clarification, you talked about p38 MAP kinase being phosphorylated after being translocated to the nucleus. In my naïve way I thought that all kinases had to be phosphorylated before they are translocated. Is that the case? Shukla: No. There is evidence that you can get translocation without phosphorylation. Apte: What is the stimulus for translocation? Shukla: It is currently unknown. But proteins and MAPK family members can be translocated without phosphorylation. In our system, when we looked at the hepatocyte cytosolic fraction, we couldn’t see any phosphorylated p38 MAPK. Even after acetaldehyde and ethanol stimulation there was no phosophorylated p38 MAPK in the cytosol. It is likely that an upstream MAPK residing in the nucleus is phosphorylating the translocated p38 MAPK. Apte: You showed an effect of ethanol on histone at very low concentrations (5 mM), yet to get an effect on MAPK you had to go up to 200 mM. Could it just be a direct effect of ethanol? Shukla: We can see an effect of ethanol at 50 mM on the MAPK. p42/44 MAPK is rather insensitive to ethanol at lower concentrations, which is different from pancreas. But p38 MAPK and JNK are sensitive to ethanol. Seitz: I have a question about the meaning of 100 mM ethanol, which is a very high concentration. No cell is ever exposed to that level in vivo. So is this result relevant? I also have a question on the histone effects, which you see at low levels. Remember, when we looked at our gastrointestinal H3 histone expression, which was overexpressed, it could well be that the modification leads to hyperproliferation. This may link your experiments with these cells to what we found in the gastrointestinal tract. The alcohol levels you are using seem to be rather high. M Salaspuro: Have you had plans to continue with some other cell lines? With liver there are many problems with regard to the real acetaldehyde concentration. But in the pathogenesis of gastrointestinal tract cancer, acetaldehyde appears to be important. Have you thought of going to cell lines representing the colon or oesophagus and doing the same types of studies? Shukla: That’s an excellent suggestion. It would be good to examine this in other cells.

226

SHUKLA ET AL

Rao: I want to follow up on the question about JNK. It’s interesting that this can have the opposite effects in response to ethanol and acetaldehyde. There are two possible mechanisms. One could be that JNK1 and 2 are involved and have different effects. The other possibility is that you showed a nice kinetic difference in p38 activation by ethanol and acetaldehyde. Is there such a difference in terms of activation of JNK? There is evidence ( Jin et al 2006) suggesting that if JNK is constitutively active it is more apoptotic than if it is transiently activated. Shukla: JNK1 especially may be more involved in that process. We will have to use JNK1 knockouts to address some of these issues. Yin: With regard to your dosage, 100 or 200 mM ethanol and 1 mM acetaldehyde may be a bit high. The acetate concentration you use is much more physiological at 1 mM. In the 1960s a Danish group led by Frank Lundquist already showed that. Our data are in agreement. Regardless of the ALDH genotype, the acetate and ethanol levels are the same; only acetaldehyde differs. After social drinking 1 mM acetate is physiological. The other thing is the muscle removal capacity for acetate is quite limited, so acetate lasts a long time after ethanol ingestion. This would allow more time for your signalling pathway to be activated. Eriksson: I am a little bit concerned about the concentration of ethanol and acetaldehyde you used. In the liver we don’t see millimolar concentrations of acetaldehyde: it is closer to 100 µM. When you use concentrations like this, you always see fantastic effects. This means that we end up in a jungle where everyone finds all kinds of effects. It is difficult to find the pieces that are relevant for the aetiology. It makes sense to go to concentrations that are as low as possible to see which are most sensitive to low concentrations, so you can pinpoint these and concentrate on them. But you had some other data where you had used methylpyrazole in the histone acetylation. This was interesting, but you also had cyanamide present. What did the cyanamide really do? Shukla: It inhibited the histone acetylation. This is the same as the methylpyrazole effect. Eriksson: But they are doing totally the opposite things. One is increasing acetaldehyde while the other is decreasing it. Shukla: It means that acetaldehyde metabolism increases the acetylation of histones. Decreasing the acetate also decreases the acetylation. Eriksson: So that is further evidence for the role of the acetate. Shukla: Let me come back to the question of concentration. I agree that 1, 2 or 5 mM are very high levels of acetaldehyde. But we have also observed the same effects at 0.2 mM, which is more in the pathophysiological range. Morris: The statement that acetaldehyde is a rodent carcinogen is based on the inhalation bioassay which results in nasal tumours. The tissue concentrations in the nose are thought to be in the millimolar range (Morris 1997). We can’t have this both ways: we can’t say this is relevant and that millimolar concentrations are not.

ACETALDEHYDE, MAP KINASES AND HISTONE MODIFICATIONS

227

If these concentrations never occur in the liver, then the nasal carcinogenesis is an event that occurs at an irrelevantly high tissue concentration. Shukla: I believe that in situ concentrations of acetaldehyde may be higher than our measurements suggest. At the moment we can’t say that 1 mM is irrelevant. You have made a good point. Thornalley: I am interested in what actually activates these protein kinases. If we look at the precedent of small molecules that modify thiol groups, there is a precedent in the work on isothiocyanates. These activate MAPK and JNK as well. After much study the mechanism was found to be because isothiocyanates modify a particular protein phosphatase, which is then ubiquitinated once modification occurs and directed to the proteasome. It is the decreased protein phosphatase activity that activates JNK rather than an agonist effect. Have you looked at specific protein phosphatases? Acetaldehyde may well be causing modification of phosphatases and directing them to the proteasome too. Then you are getting agonist-free activation of these proteins. Shukla: That is a good comment. We haven’t looked at phosphatases. Rao: In my laboratory we have taken recombinant PTP1B, incubated it with acetaldehyde and shown a significant inhibition. There is a direct interaction of acetaldehyde with PTP1B. PTP1B could still be a trigger because most of these upstream signalling proteins are regulated by tyrosine-phosphorylation. Apte: We have been talking about acetaldehyde as the common factor explaining alcohol-related damage at various sites in the body. What about downstream of this? What if oxidant stress is actually the final common pathway causing the damage we have seen? Does oxidative stress have any effect on histone acetylation or phosphorylation? Shukla: That’s a good question. At this stage the answer would be ‘perhaps’. There may be a contribution from ROS in these modifications. Eriksson: I think this is a relative question. Oxidative stress is also caused by acetaldehyde metabolism. We have to go back to the 1960s and 1970s when redox was found. These are the types of reductive oxidative stress that are much neglected in modern research but which could be behind many of these things. Then there is the radical story. All these go together. Some effects are caused by direct molecular reactions with acetaldehyde and some are caused by oxidative stress, both radical formation and redox. Albano: There is a further loop of cell signalling that may be mediated simply by changes in cell volume. We have observed that in isolated hepatocytes incubated with ethanol the intracellular pH became acidic because of the accumulation of acetate. This activates the acid buffering capacity of the cells that leads to the increase of sodium influx through the Na+/H+ exchanges and Na+-HCO3− cotransporter (Carini et al 2000). The increase in cell hydration in response to Na+ promotes the release of ATP that in turn stimulates purinergic P2 receptors,

228

SHUKLA ET AL

activating a kinase pathways involving Src, ERK1/2 and p38MAPK (Carini et al 2006). This could explain the acetate effects seen in your experiments. Shukla: Oxidative stress may not be responsible for every effect of ethanol. Nonoxidative routes of ethanol metabolism also deserve investigation, especially the role of PEth (phosphatidylethanol) and FAEE (fatty acid ethyl esters). This field is developing fast, and in the next few years we are likely to learn more about these metabolites. M Salaspuro: The oxidative stress is an excellent theory, but the biggest problem is that when we come to intervention studies in which we try to prevent oxidative stress the results have been very disappointing. Apte: This could be because we are unable to inhibit oxidative stress effectively in tissues. References Carini R, De Cesaris MG, Splendore R, Albano E 2000 Ethanol potentiates hypoxic liver injury: role of hepatocyte Na+ overload. Biochim Biophys Acta 1502:508–514 Carini R, Alchera E, De Cesaris MG et al 2006 Purinergic P2Y2 receptors promote hepatocyte resistance to hypoxia. J Hepatol 45:236–245 Jin S, Ray RM, Johnson LR 2006 Rac1 mediates intestinal epithelial cell apoptosis via JNK. Am J Physiol Gastrointest Liver Physiol 291:G1137–1147 Morris JB 1997 Dosimetry, toxicity and carcinogenicity of inspired acetaldehyde in the rat. Mutat Res 380:113–124

Endogenous α-oxoaldehydes and formation of protein and nucleotide advanced glycation endproducts in tissue damage Paul J. Thornalley1 Disease Mechanisms and Therapeutics Research Group, Department of Biological Sciences, University of Essex, Wivenhoe Park, Colchester, Essex CO4 3SQ , UK Abstract. Human and other biological tissues face a continual threat of damage by αoxoaldehydes formed endogenously. Glyoxal, methylglyoxal and 3-deoxyglucosone are formed by the degradation of glycolytic intermediates, glycated proteins and lipid peroxidation. They are potent glycating agents of protein and nucleotides leading to the formation of advanced glycation endproducts (AGEs). With proteins, they are arginine residuedirected glycating agents forming mainly hydroimidazolones, found at 0.1–1% of total arginine residues in tissues (2–20% of proteins modified). With nucleotides, imidazopurinone- and N2-carboxyalkyl- derivatives of deoxyguanosine are formed, found at 0.1–0.8 per 106 nucleotides in DNA. Glycation occurs in all tissues and body fluids. Cellular proteolysis of AGE-modified proteins and DNA releases glycated amino acids and nucleosides. Glycated amino acids and nucleosides are released into plasma, undergo glomerular filtration and are excreted in urine. The damage to tissue protein and nucleotides by αoxoaldehydes is suppressed by the metabolism of α-oxoaldehyde glycating agents by the glutathione-dependent enzyme, glyoxalase I, and aldo-keto reductases. These enzymatic activities are part of the enzymatic defence against glycation. Tissue damage by α-oxoaldehyde glycation is implicated in diabetic and non-diabetic vascular disease, renal failure, cirrhosis, Alzheimer’s disease, arthritis and ageing. 2007 Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Foundation Symposium 285) p 229–246

Introduction: endogenous formation of α-oxoaldehydes α-Oxoaldehydes

are a type of reactive class of aldehyde that are highly electron deficient and bidentate, typically but not exclusively forming novel ring structures in proteins and DNA: arginine-derived hydroimidazolone residues and deoxy-

1

Current address: Protein Damage and Systems Biology Research Group, Warwick Medical School and Systems Biology Centre, Clinical Sciences Research Institute, University of Warwick, University Hospital, Coventry CV2 2DX, UK 229

230

THORNALLEY

guanosine-derived imidazopurinone residues, respectively. These modifications are linked to functional impairment of proteins and mutation in DNA (Thornalley 2003a, 2005). Some α-oxoaldehydes are formed endogenously and therefore there is a continuous risk of protein and nucleotide damage from within. Physiologically important α-oxoaldehydes are glyoxal, methylglyoxal and 3deoxyglucosone (3DG). Many others are formed by caramelization of glucose and other saccharides and derivatives during thermal sterilization of dialysis fluids for renal replacement therapy and during thermal processing of food (Nagao et al 1986, Thornalley 2005). Glyoxal is formed by lipid peroxidation, degradation of nucleotides and glycated proteins. Methylglyoxal is formed mainly by the spontaneous degradation of triosephosphates; typically, 0.1% of the flux of triosephosphates degrades to methylglyoxal but this may rise to 1% in hyperglycaemia associated with diabetes (Thornalley 1988). Other sources of methylglyoxal are cytochrome P450 2E1catalysed oxidation of acetone formed from ketone bodies and minor formation by the degradation of glycated proteins, lipid peroxidation and threonine catabolism (Thornalley 2005). 3DG is formed by the phosphorylation of fructosyl-lysine residues in glycated proteins and fructose by fructosamine 3-phosphokinase. The phosphorylated derivates thereby formed, fructosamine-3-phosphate residues and fructose-3-phosphate, degrade spontaneously to 3DG (Delpierre et al 2000, Szwergold et al 2001). 3DG is also formed by the non-enzymatic degradation of glycated proteins (Thornalley et al 1999) (Fig. 1a). The endogenous formation of α-oxoaldehydes occurs in all physiological systems. The flux of formation of methylglyoxal is ca. 120 nmol/g tissue/day and may be increased five–sixfold in hyperglycaemia associated with diabetes (Thornalley 1988). The fluxes of formation of glyoxal and 3DG are uncertain but they are probably 10–50-fold lower than this. The formation of glyoxal increases when lipid peroxidation is enhanced and the formation of 3DG is increased in hyperglycaemia associated with diabetes. Glycation of proteins by α-oxoaldehydes Glycation of proteins is a complex series of parallel and sequential reactions collectively called the Maillard reaction. Physiological glycation has been most extensively studied for glucose where the initial Schiff’s base adduct rearranges to the major early stage glycation adducts, fructosamines: Nε-1-deoxyfructosyl-lysine (FL) and N-terminal Nα-1-deoxyfructosyl-amino acid residues (Fig. 1b). Fructosamines degrade slowly to form stable end-stage adducts called advanced glycation endproducts (AGEs). α-Oxoaldehydes react with proteins to form AGEs directly. α-Oxoaldehydes react predominantly to form hydroimidazolone AGE residues of proteins (Ahmed et al 2005b, Dobler et al 2006). Hydroimidazolones derived from glyoxal, methylglyoxal and 3-DG are denoted by the acronyms G-H1, MG-H1 and 3DG-H (and related to structural isomers)—as defined previously (Thornalley et al 2003a). Other α-oxoaldehyde-derived AGEs include: Nω-carboxymethylarginine (CMA),

ENDOGENOUS α-OXOALDEHYDES IN TISSUE DAMAGE

231

FIG. 1. α-Oxoaldehydes and protein glycation. (a) Physiological α-oxoaldehydes. (b) Early glycation adducts—fructosamines. (c) Formation of advanced glycation endproducts (AGEs). (d) AGEs formed from α-oxoaldehydes.

232

THORNALLEY

TABLE 1 Methylglyoxal-derived hydroimidazolone residues of mammalian tissue and extracellular proteins Organism

Protein source

n

MG-H1 residue (mmol/mol arg)

Human

Plasma protein Red blood cells Peripheral lymphocytes Mesangial cells (in vitro) Lens protein Plasma protein Aortal collagen Heart Liver Skeletal muscle Brain Renal glomeruli Retina Sciatic nerve

10 10 3 3 55 13 13 7 13 7 7 7 7 7

0.31 ± 0.20 3.14 ± 0.72 (%Hb) 7.46 ± 1.13 0.60 ± 0.05 15.0 ± 1.7 1.29 ± 0.47 0.55 ± 0.30 3.43 ± 1.01 3.34 ± 0.32 1.70 ± 0.77 2.73 ± 0.33 2.30 ± 0.25 1.88 ± 0.51 4.74 ± 2.74

Rat

(From Ahmed et al 2003, 2005a, Babaei-Jadidi et al 2003, 2004, Dobler et al 2006, Thornalley et al 2003.)

Nε-carboxymethyl-lysine (CML) and the lysine dimer, 1,3-di(N ε-lysino)imidazolium (GOLD) (formed from glyoxal); argpyrimidine, Nε-carboxyethyl-lysine and the lysine dimer, 1,3-di(N ε-lysino)-4-methyl-imidazolium (MOLD) (formed from methylglyoxal); and pyrraline and the lysine dimer, 1,3-di(N ε-lysino)-4-(2,3,4-trihydroxybutyl)imidazolium (DOLD) (Fig. 1). α-Oxoaldehydes are up to 20 000 times more reactive than glucose in glycation reactions. The effect of this is countered in vivo by steadystate concentrations of α-oxoaldehydes up to 50 000 fold lower than that of glucose. Recent advances in the application of liquid chromatography with tandem mass spectrometric detection (LC-MS/MS) have shown hydroimidazolone AGE residues to be significant types of protein damage quantitatively in fluids and tissues of experimental animals and human subjects (Thornalley et al 2003) (Table 1). Modification of arginine residues to form hydroimidazolones is associated with increased molecular volume and loss of positive charge (Ahmed et al 2005b). It is also expected to break π-cation bonding interactions between arginine and tryptophan, phenylalanine and tyrosine residues (Gallivan & Dougherty 1999). The bioinformatics calculation of receptor binding domain predicting amino acid residues located in functional sites of protein–protein interaction suggests arginine residues have a frequency of 19.6%—the highest of any amino acid residue (Gallet et al 2000). Hence, modification of arginine residues by α-oxoaldehydes is likely to cause functional impairment. This was illustrated in three recent peptide mapping

ENDOGENOUS α-OXOALDEHYDES IN TISSUE DAMAGE

233

studies of methylglyoxal modified human serum albumin, type IV collagen and the corepressor protein Sin 3A. Modification of human serum albumin by methylglyoxal produced hotspot modification of Arg410 with hydroimidazolone MG-H1 residue formation. Modification of Arg410 by methylglyoxal was found in albumin glycated in vivo. Arg410 is located in drug binding site II and the active site of albumin-associated esterase activity. Hydroimidazolone formation at Arg410 inhibited drug binding and esterase activity. Molecular dynamics and modelling studies indicated that hydroimidazolone formation caused structural distortion leading to disruption of arginine-directed hydrogen bonding and loss of electrostatic interaction (Ahmed et al 2005b). MG-H1 residue formation in albumin and other plasma proteins was increased in clinical diabetes and end-stage renal disease (Agalou et al 2005, Ahmed et al 2005a) (Fig. 2a–c). Modification of vascular basement membrane type IV collagen by methylglyoxal formed MG-H1 residues at hotspot modification sites in RGD and GFOGER integrin-binding sites of collagen, causing endothelial cell detachment, anoikis and inhibition of angiogenesis (Fig. 2d,e). Endothelial cells incubated in model hyperglycaemia in vitro and experimental diabetes in vivo produced the same modifications of vascular collagen, inducing similar responses. Increased methylglyoxal, formed from triosephosphate degradation within the endothelial cell, crossed the plasma membrane and modified type IV collagen (Dobler et al 2006). This may contribute to increased shedding of vascular endothelial cells and increased numbers of circulating endothelial cells in diabetes and uraemia where plasma levels of methylglyoxal and other α-oxoaldehydes are abnormally high (Mann et al 1999, McLellan et al 1994). The number of circulating endothelial cells, indicative of damage to the endothelium, is prognostic for vascular disease (Segal et al 2002). m Sin 3A is a component of a large multiprotein corepressor complex with histone deacetylase activity. Physical interactions of m Sin 3A with many sequence-specific transcription factors has linked the m Sin 3A corepressor complex to the regulation of many signalling pathways (Dannenberg et al 2005). Modification of m Sin 3A by methylglyoxal produced increased recruitment of O-linked β-N-acetylglucosamine transferase (OGT) to complex of m Sin 3A with transcription factor Sp3, with consequent increased modification of Sp3 by O-linked N-acetylglucosamine. This produced decreased binding of the repressor complex to a glucose-responsive GC box in the angiopoietin 2 (Ang2) promoter, resulting in increased Ang2 expression (Fig. 3). When insufficient levels of vascular endothelial derived growth factor VEGF and other angiogenic signals are present, Ang2 causes endothelial cell death, acellular capillaries, capillary occlusion and ischaemia—implicated particularly in retinopathy. m Sin 3A paired amphipathic helix 4 (PAH4) domain was necessary and sufficient for increased association of methylglyoxal-modified m Sin 3A with OGT. Within the PAH4 domain, mutation of R925 markedly impaired this increased association, while mutation of R923 or K938 had smaller effects. Double mutation

FIG. 2. Methylglyoxal modification of human serum albumin and type IV collagen. (a) Molecular modelling of drug binding site II of human albumin and modification by methylglyoxal. Structure of native albumin showing domain 3A residues Asn319, Arg410 and Tyr411 (helices nomenclature also indicated). (b) Computed structure of domain 3A with MG-H1-410 residue. (c) Formation of hydroimidazolone residue MG-H1. (d) Interaction of type IV collagen GFOGER motifs and integrin α2-I domain: energy-minimized structure of the interaction of type IV collagen GFOGERMG-H1 motif and integrin α2-I domain. (e) Schematic representation of integrin ligation of basement membrane collagen by the endothelial cells and loss of integrin ligation and anoikis after glycation by methylglyoxal (Ahmed et al 2005b, Dobler et al 2006).

FIG. 3. Methylglyoxal modification of Sin 3A and activation of angiopoietin 2 expression. (a) Under basal conditions, complexes of Sp3 and m Sin 3A bind to the GC box in the Ang2 promoter and inhibit angiopoietin (Ang2) transcription. (b) Increased glucose-derived methylglyoxal (MG) causes increased Sin 3A modification. (c) Increased modification of m Sin 3A increases association of O-GlcNAc transferase (OGT) with the bound complex and increased glucose-derived UDPN-acetylglucosamine (UDP-GlcNAc) is transferred to Sp3 by bound OGT. (d) Increased Sp3 modification by O-GlcNAc causes decreased binding affinity to the GC box, and altered DNA partitioning of the Sp3-m Sin 3A-OGT complex and decreased bound Sp3 allows Sp1 to bind to the GC box, activating Ang2 transcription (Yao et al 2006).

236

THORNALLEY

of R925 and K938 completely prevented the increased association, while double mutation of R923 and R925 was no more effective than mutation of R925 alone. The arginine and lysine residues most reactive with methylglyoxal are those with proximate arginine and lysine residues that decrease the pKa of both interacting amino groups and a glutamate or aspartate residue carboxylate side chains on the alternate side in the sequence to act as a catalytic base. R923 has proximity to W927 in a helix configuration, with the formation of a strong cation-π bonding interaction (Yao et al 2006). Proteolysis of α-oxoaldehyde-modified proteins and formation of glycation free adducts Protein glycation was originally viewed as a post-translational modification that accumulated mostly on extracellular proteins. Specifically, AGEs were thought to be formed slowly throughout life and the concentrations of AGEs found represent a life-long accumulation of the glycation adduct. This applies to chemically stable AGEs formed on long-lived proteins. FL and some AGEs (hydroimidazolones, for example) have relatively short chemical half-lives under physiological conditions (2–6 weeks), however, and they may also be formed on cellular and shortlived extracellular proteins. The turnover of these proteins with degradation by cellular proteolysis releases AGE free adducts. AGE free adducts are exported from cells, leak into plasma and are excreted in the urine. Major quantitative AGEs, hydroimidazolones, CML and CEL, have high renal clearances of 35–93 ml/min in normal human subjects, but this declines in uraemia (Thornalley et al 2003). AGE free adducts are the major form by which glycation adducts are eliminated, taking into account AGEs in both urinary albumin and protein fragments. They are also the major form of glycation adduct eliminated in dialysate of end-stage renal disease (ESRD) patients on peritoneal dialysis (PD) and haemodialysis (HD) (Agalou et al 2005). Patients with mild uraemia and chronic renal failure (CRF) had plasma AGE free adduct concentrations increased up to fivefold associated with a concomitant decline in renal clearance but no increase in 24 h urinary elimination. In ESRD patients, plasma AGE free adduct concentrations were increased up to 18-fold on PD and 40-fold on HD. In PD patients, 24 h elimination of AGE free adducts was increased up to ninefold, suggesting there was increased formation of AGE free adducts; increased AGE free adduct formation may also contribute to the extraordinary high plasma concentrations of AGE free adducts in HD (Agalou et al 2005). The plasma concentrations and 24 h urinary excretion of some glycation free adducts was also increased markedly in diabetic patents with normal renal function and moderate glycaemic control: the plasma concentration of MG-H1 free adduct was increased 10-fold and the 24 h urinary excretion increased 15-fold, although there was moderate glycaemic control

ENDOGENOUS α-OXOALDEHYDES IN TISSUE DAMAGE

237

(glycated haemoglobin HbA1c was 7.8%) (Ahmed et al 2005a). The urinary excretion of MG-H1 free adduct of ESRD patients on PD therapy was approximately twice that of diabetic patients with normal renal function. This suggests that the diabetes produces profound increases in concentrations of some plasma AGE free adducts which are increased further in uraemia (with and without diabetes). These increases in AGE free adduct excretion are probably due to increased glycation of endogenous proteins with subsequent proteolysis. Saccharide-rich foods are a good source of α-oxoaldehyde-derived and other AGE residues, particularly bakery products. There is a low bioavailability of AGE residues in proteins of ingested foods, however, such that <10% is absorbed. Proteins highly glycated by FL and AGEs are resistant to proteolysis and some AGEs inhibit intestinal proteases. The highest concentration of absorbed food AGE is expected in portal venous plasma. There was no evidence for increased AGE residues of plasma protein and AGE free adducts in portal venous plasma, compared to peripheral venous plasma (although some AGE residues and AGE free adducts were increased in cirrhosis). There were, however, AGE-rich peptides with a content of MG-H1 residues approximately 10-fold higher than in plasma protein in portal venous plasma (Ahmed et al 2004). AGEs from food are therefore probably absorbed as both AGE free adducts and AGE-rich peptides; the latter appear to be degraded efficiently after absorption. Since AGE free adducts have high renal clearance and low plasma concentrations, AGEs absorbed from food are expected to have low toxicity in subjects with normal renal function. The contributions of AGEs absorbed from ingested food to AGE exposure are probably greatest when the endogenous formation of AGEs is normal and renal clearance is impaired, as occurs in non-diabetic subjects with mild uraemic CRF. Little is known about the effect of high concentrations of glycation free adducts on cell function. It is conceivable that high concentrations of glycation free adducts in blood plasma may back up into vascular cells and contribute to vascular cell dysfunction (Fig. 4). Glycation of nucleotides DNA is susceptible to glycation by α-oxoaldehydes, and the nucleotide most reactive under physiological conditions is deoxyguanosine. Under conditions of limiting α-oxoaldehyde, as applies in vivo, the major nucleotide AGEs are the imidazopurinone derivatives 3-(2′-deoxyribosyl)-6,7-dihydro-6,7-dihydroxyimidazo [2,3-b]purin-9(8)one (dG-G) and 6,7-dihydro-6,7-dihydroxy-6-methylimidazo[2,3-b]purine-9(8)one (dG-MG). Other adducts are also formed: N2-carboxymethyl-deoxyguanosine (CMdG) and 5-glycolyldeoxycytidine (gdC) from glyoxal, N2-(1-carboxyethyl)-deoxyguanosine (CEdG) from methylglyoxal, and N2-(1-oxo2,4,5,6-tetrahydroxyhexyl)deoxyguanosine (dG-3DG) derived from 3-deoxyglucosone (Fig. 5). There have been few quantitative estimates of nucleotide AGEs in

238

THORNALLEY

FIG. 4. Biodistribution scheme illustrating flows of formation and removal of protein glycation free adducts.

FIG. 5. Nucleotide advanced glycation endproducts. The dashed line indicates the 2′deoxyribosyl linkage to phosphodeoxyribosyl backbone.

ENDOGENOUS α-OXOALDEHYDES IN TISSUE DAMAGE

239

mammalian cells. Basal levels of dG-G in rat liver were in the range 0.6–0.9 per 106 nucleotides and of dG-MG in human peripheral lymphocytes 0.08 per 106 nucleotides. The CEdG and gdC adducts have be detected immunochemically. The level of dG-G is similar to the major oxidative adduct of DNA, 8-hydroxydeoxyguanosine (8-HO-dG), 0.1–0.8 per 106 nucleotides. Imidazopurinones have only moderate stability under physiological conditions but are stabilized by formation in single- and double-stranded DNA: half-lives for dG-G and residues in single- and double-stranded DNA were 14.8 h, 285 h and 595 h at pH 7.4 and 37 °C, respectively. The CMdG and CEdG adducts of dG are expected to be more stable. The formation of CEdG was associated with depurination of DNA (Thornalley 2003a). DNA glycation, mutagenesis and nucleotide excision repair The formation of nucleotide AGEs in DNA is associated with increased mutation frequency, DNA strand breaks and cytotoxicity. Glycation of the shuttle vector pMY189 containing the bacterial suppressor tRNA supF gene by glyoxal and transfection in COS-7 cells increased mutations and decreased DNA replication. Most mutations were single-base substitutions at C:G sites, predominantly G:C → T:A with some G:C → C:G transversions, G:C → A:T transitions and A:T → T:A transversions. For methylglyoxal, multi-base deletions were predominant with some base-pair substitutions, mostly at C:G sites. Most mutations were G:C → C:G and G:C → T:A transversions. For both glyoxal and methylglyoxal, there were hotspot mutation sites in the SupF gene (Murata-Kamiya et al 1997, 2000b). The prevention of α-oxoaldehyde-induced mutation by nucleotide excision repair (NER) was investigated by incubation of wild-type and NER-deficient Escherichia coli with glyoxal and methylglyoxal and the LacI gene analysed for mutations. Increasing concentration of α-oxoaldehyde led to increased mutations, decreased DNA replication and cell death. Base-pair mutations were higher in the NERdeficient strain relative to the wild-type strain. Most base-pair substitutions were at C:G sites where G:C → T:A transversions were predominant. There were also G:C → A:T transitions and A:T → T:A transversions with glyoxal and TGGC frameshift mutations for methylglyoxal (Murata-Kamiya et al 1998, 2000a). Histone proteins are susceptible to glycation. Glycation by glyoxal and methylglyoxal resulted in histone cross-linking in vitro (Roberts et al 2003). Methylglyoxal also produced DNA–protein cross-links, cross-linking the DNA template to the Klenow fragment of DNA polymerase. Glyoxal was much less effective than methylglyoxal in this cross-linking (Thornalley 2003a). Nucleotide glycation and related effects are expected to be most marked in diseases associated with the accumulation of α-oxoaldehydes to high concentrations,

240

THORNALLEY

such as occurs in diabetes and uraemia. Patients with ERSD have a high incidence of tumours. Increased nucleotide glycation by α-oxoaldehydes in diabetes has also been implicated in impaired growth of keratinocytes and fibroblasts in skin lesions (Roberts et al 2003) and teratogenicity associated with perinatal mortality (Eriksson et al 1998). The enzymatic defence against glycation: prevention of protein and nucleotide glycation, and potent antitumour agents from pharmacological inhibition The GSH-dependent enzyme glyoxalase I (Glo1) catalyses the detoxification of glyoxal and methylglyoxal in physiological systems. Aldoketo reductases catalyse the detoxification of 3DG. These enzymes are part of the enzymatic defence against glycation, suppressing α-oxoaldehyde-mediated protein and nucleotide damage (Thornalley 2003a,b). Overexpression of Glo1 counters effectively the increased glycation damage by α-oxoaldehydes in hyperglycaemia associated with diabetes (Shinohara et al 1998). Overexpression of Glo1 in the nematode Caenorhabditis elegans enhanced lifespan by prevention of mitochondrial glycation and increased formation of reactive oxygen species, preventing protein damage by glycation, oxidation and nitration (Morcos et al 2005). Surprisingly, overexpression of Glo1 also produced multidrug resistance (MDR) in tumours in vitro to clinical antitumour agents (Sakamoto et al 2000). Studies on MDR human monocytic leukaemia UK711 cells and erythroleukaemia K562/ADM cells, using a subtractive hybridization mRNA approach, revealed that Glo1 was overexpressed in both cell lines. Moreover, a stable transfectant of Jurkat cells acquired MDR when overexpression of Glo1 was produced. The Glo1 inhibitor S-p-bromobenzylglutathione cyclopentyl diester (SpBrGSHCp2) lifted MDR and the cells regained sensitivity to antitumour agents (Sakamoto et al 2000). Cellpermeable inhibitors of Glo1 produce an accumulation of methylglyoxal within tumour cells, increasing nucleotide glycation and inducing apoptosis (Thornalley et al 1996). Human lung cancer NCI-H522 and DMS114 cells with high expression and activity of Glo1 underwent apoptosis when treated with SpBrGSHCp2. The sensitivity to SpBrGSHCp2 correlated to the level of Glo1 expression. SpBrGSHCp2 induced the activation of the stress-activated protein kinases c-Jun NH2terminal kinase 1 and p38 mitogen-activated protein kinase, which led to caspase activation in GLO1-overexpressing tumour cells. SpBrGSHCp2 inhibited the growth of xenografted DMS114 and human prostate cancer DU-145; it was the most potent antitumour agent against human tumours overexpressing Glo1, particularly refractory lung and prostate carcinomas (Sakamoto et al 2001). These studies suggested that accumulation of Glo1 substrates is involved in the antitumour activity of important clinical antitumour agents (for example, doxorubicin), and Glo1

ENDOGENOUS α-OXOALDEHYDES IN TISSUE DAMAGE

241

inhibitors are active against refractory tumours associated with high incidence and mortality rates (lung, breast, prostate and colon carcinomas) (Thornalley 2003a). The link between overexpression of Glo1 and MDR in cancer chemotherapy may involve increased formation of methylglyoxal in response to DNA repair. The cellular response to antitumour drugs that modify DNA or disrupt DNA metabolism is to activate processes of DNA repair, including poly(ADP-ribose) polymerase. This depletes cells of NAD+ such that glyceraldehyde-3-phosphate dehydrogenase activity is depleted and triosephosphates, glyceraldehyde-3phosphate and dihydroxyacetonephosphate increase markedly. Methylglyoxal is formed mainly by triosephosphate degradation and, hence, a consequent dramatic increase in methylglyoxal formation is expected. The increased methylglyoxal concentrations may modify DNA and protein involved in activation of apoptosis, thereby potentiating the cytotoxic effect of the antitumour agent. This potentiatory effect of methylglyoxal will be blocked by overexpression of Glo1 in MDR (Thornalley 2003a). Concluding remarks Glycation by α-oxoaldehydes is a major type of physiological damage to proteins and nucleotides in physiological systems. It influences longevity and morbidity and mortality in disease and therapeutic interventions. Acknowledgements I thank the Wellcome Trust (UK), Diabetes UK (UK), Juvenile Diabetes Research Fund (USA) and Cancer Research UK for support for my research.

References Agalou S, Ahmed N, Babaei-Jadidi R, Dawnay A, Thornalley PJ 2005 Profound mishandling of protein glycation degradation products in uremia and dialysis. J Am Soc Nephrol 16: 1471–1485 Ahmed N, Thornalley PJ, Dawczynski J et al 2003 Methylglyoxal-derived hydroimidazolone advanced glycation endproducts of human lens proteins. Invest Ophthalmol Vis Sci 44: 5287–5292 Ahmed N, Thornalley PJ, Luthen R et al 2004 Processing of protein glycation, oxidation and nitrosation adducts in the liver and the effect of cirrhosis. J Hepatol 41:913–919 Ahmed N, Babaei-Jadidi R, Howell SK, Beisswenger PJ, Thornalley PJ 2005a Degradation products of proteins damaged by glycation, oxidation and nitration in clinical type 1 diabetes. Diabetologia 48:1590–1603 Ahmed N, Dobler D, Dean M, Thornalley PJ 2005b Peptide mapping identifies hotspot site of modification in human serum albumin by methylglyoxal involved in ligand binding and esterase activity. J Biol Chem 280:5724–5732 Babaei-Jadidi R, Karachalias N, Ahmed N, Battah S, Thornalley PJ 2003 Prevention of incipient diabetic nephropathy by high dose thiamine and Benfotiamine. Diabetes 52:2110–2120

242

THORNALLEY

Babaei-Jadidi R, Karachalias N, Kupich C, Ahmed N, Thornalley PJ 2004 High dose thiamine therapy counters dyslipidaemia in streptozotocin-induced diabetic rats. Diabetologia 47: 2235–2246 Dannenberg JH, David G, Zhong S, van der Torre J, Wong WH, DePinho RA 2005 m Sin 3A corepressor regulates diverse transcriptional networks governing normal and neoplastic growth and survival. Genes Dev 19:1581–1595 Delpierre G, Rider MH, Collard F et al 2000 Identification, cloning, and heterologous expression of a mammalian fructosamine-3-kinase. Diabetes 49:1627–1634 Dobler D, Ahmed N, Song LJ, Eboigbodin KE, Thornalley PJ 2006 Increased dicarbonyl metabolism in endothelial cells in hyperglycemia induces anoikis and impairs angiogenesis by RGD and GFOGER motif modification. Diabetes 55:1961–1969 Eriksson UJ, Wentzel P, Minhas HS, Thornalley PJ 1998 Teratogenicity of 3-deoxyglucosone and diabetic embryopathy. Diabetes 47:1960–1966 Gallet X, Charloteaux B, Thomas A, Braseur R 2000 A fast method to predict protein interaction sites from sequences. J Mol Biol 302:917–926 Gallivan JP, Dougherty DA 1999 Cation-pi interactions in structural biology. Proc Natl Acad Sci USA 96:9459–9464 Mann VM, Tucker B, Thornalley PJ, Dawnay A 1999 Elevated plasma methylglyoxal and glyoxal in uraemia: implications for advanced glycation endproduct formation. Kidney Int 55:2582 McLellan AC, Thornalley PJ, Benn J, Sonksen PH 1994 The glyoxalase system in clinical diabetes mellitus and correlation with diabetic complications. Clin Sci 87:21–29 Morcos M, Du XL, Hutter AASH et al 2005 Life extension in Caenorhabditis elegans by overexpression of glyoxalase I—The connection to protein damage by glycation, oxidation and nitration. Free Radic Res 39:S43–43 Murata-Kamiya N, Kamiya H, Kaji H, Kasai H 1997 Glyoxal, a major product of DNA oxidation, induces mutations at G:C sites on a shuttle vector plasmid replicated in mammalian cells. Nucleic Acids Res 25:1897–1902 Murata-Kamiya N, Kamiya H, Kaji H, Kasai H 1998 Nucleotide excision repair proteins may be involved in the fixation of glyoxal-induced mutagenesis in Escherichia Coli. Biochem Biophys Res Commun 248:412–417 Murata-Kamiya N, Kaji H, Kasai H 2000a Deficient nucleotide excision repair increases base-pair substitutions but decreases TGGC frameshifts induced by methylglyoxal in Escherichia coli. Mutat Res 442:19–28 Murata-Kamiya N, Kamiya H, Kaji H, Kasai H 2000b Methylglyoxal induces G:C to C:G and G:C to T:A transversions in the supF gene on a shuttle vector plasmid replicated in mammalian cells. Mutat Res 468:173–182 Nagao M, Fujita Y, Wakabayashi K, Nukaya H, Kosuge T, Sugimura T 1986 Mutagens in Coffee and Other Beverages. Environ Health Perspect 67:89–91 Roberts MJ, Wondrak GT, Cervantes-Laurean D, Jacobsen MK, Jacobsen EL 2003 DNA damage by carbonyl stress in human skin cells. Mutat Res 522:45–56 Sakamoto H, Mashima T, Kazaki A et al 2000 Glyoxalase I is involved in resistance of human leukemia cells to antitumour agent-induced apoptosis. Blood 95:3214–3218 Sakamoto H, Mashima T, Sato S, Hashimoto Y, Yamori T, Tsuruo T 2001 Selective activation of apoptosis program by S-p-bromobenzylglutathione cyclopentyl diester in glyoxalase Ioverexpressing human lung cancer cells. Clin Cancer Res 7: 2513–2518 Segal MS, Bihorac A, Koc M 2002 Circulating endothelial cells: tea leaves for renal disease. Am J Physiol Renal Physiol 283:F11–19 Shinohara M, Thornalley PJ, Giardino I et al 1998 Overexpression of glyoxalase I in bovine endothelial cells inhibits intracellular advanced glycation endproduct formation and

ENDOGENOUS α-OXOALDEHYDES IN TISSUE DAMAGE

243

prevents hyperglycaemia-induced increases in macromolecular endocytosis. J Clin Invest 101: 1142–1147 Szwergold BS, Howell S, Beisswenger PJ 2001 Human fructosamine-3-kinase. Purification, sequencing, substrate specificity, and evidence of activity in vivo. Diabetes 50:2139–2147 Thornalley PJ 1988 Modification of the glyoxalase system in human red blood cells by glucose in vitro. Biochem J 254:751–755 Thornalley PJ 2003a Protecting the genome: defence against nucleotide glycation and emerging role of glyoxalase I over expression in multidrug resistance in cancer chemotherapy. Biochem Soc Trans 31:1372–1377 Thornalley PJ 2003b The enzymatic defence against glycation in health, disease and therapeutics: a symposium to examine the concept. Biochem Soc Trans 31:1343–1348 Thornalley PJ 2005 Dicarbonyl intermediates in the Maillard reaction. Ann N Y Acad Sci 1043:111–117 Thornalley PJ, Edwards LG, Kang Y et al 1996 Antitumour activity of S-p-bromobenzylglutathione cyclopentyl diester in vitro and in vivo. Inhibition of glyoxalase I and induction of apoptosis. Biochem Pharmacol 51:1365–1372 Thornalley PJ, Langborg A, Minhas HS 1999 Formation of glyoxal, methylglyoxal and 3deoxyglucosone in the glycation of proteins by glucose. Biochem J 344:109–116 Thornalley PJ, Battah S, Ahmed N et al 2003 Quantitative screening of advanced glycation endproducts in cellular and extracellular proteins by tandem mass spectrometry. Biochem J 375:581–592 Yao D, Taguchi T, Matsumara T et al 2006 Methylglyoxal modification of m Sin 3A links glycolysis to angiopoietin-2 transcription. Cell 124:275–286

DISCUSSION Emery: We discussed the rate of degradation of these damaged proteins earlier in this meeting. There were some comments about acetaldehyde-modified proteins being longer lived in some circumstances. But as you indicated, one would expect them to be more rapidly degraded. Thornalley: People have looked at the proteosomal targeting of glycated and oxidized proteins, and they are targeted to proteosomal destruction. But if you highly glycate proteins they become resistant to cellular proteolysis. In cultured cells we can see evidence for the products of cell proteolysis of glycated proteins, ‘glycation free adducts’, being released from cells and accumulating in the media. This can be prevented with a proteasome inhibitor. Highly modified proteins are resistant to proteolysis, but we don’t find any evidence for highly modified proteins in cell systems. Rao: It appears that most of these adducts are targeting the basic amino acids. Histone is supposed to be a basic protein. Can this be modified and change gene expression? Thornalley: Yes, I think so. Certainly, people have studied glycation of histone proteins. Another feature that can affect gene expression is modification of transcription factors. We had a paper in Cell (Yao et al 2006) where we showed with our

244

THORNALLEY

collaborator Michael Brownlee that the gene co-repressor protein Sin 3A is a target for methylglyoxal modification. This can affect binding of the O-(N-acetylglucosa mine)transferase to that complex and affects enzymatic glycosylation of proteins. So some transcription factors are particular targets for these dicarbonyl modifications as well. Niemelä: You showed elevated adduct levels in cirrhotic patients. Apparently this was found in both alcoholics and non-alcoholics with cirrhosis. Is there any evidence for a specific effect of alcohol consumption in producing these types of adducts? Thornalley: We’ve not studied that, but it could be important to look at. The main objective in that study was to assess whether damaged proteins in the plasma are removed from circulation in the liver. A long-standing hypothesis in the glycation field was that glycated proteins are removed from circulation in the liver. This is why we were studying protein damage in bloodflow into and out of the liver. The conclusion was that there is no evidence for the liver being a site for the removal of damaged proteins. In fact, we know the kidney digests damaged albumin. We have studied many disease states and this is the only one in which we have found this selective profound increase in glyoxal adduct. Taken with other markers this could provide a signature for characteristic damage in cirrhosis. Perhaps there will be a signature for alcohol-induced effects also. Seitz: If I saw this correctly, most of the adducts are arginine bound. In the cirrhotics arginine is probably higher than normal; the uric cycle is activated because of the ammonia present. It depends on which stage of cirrhosis you look at: if it is activated, there is probably a high amount of arginine present. Thornalley: The arginine residue concentration is still much higher than the free arginine concentration in blood. So you’ve still got a far greater number of arginine residues that are potential targets for modification than free arginine. Apte: You showed data on free adducts measured in the plasma of diabetics. Even the moderately well controlled diabetics had high levels of free adducts. Does this mean that these may not be useful markers for diabetic control as such? Thornalley: Not necessarily. Some of the free adducts are responsive to both fasting and postprandial hypoglycaemia, whereas our conventional clinical measure is glycated haemoglobin, which is unresponsive to postprandial hyperglycaemia. Some of the free adducts look like they are even better markers of glycaemic control than current clinical markers. Apte: Do these markers change in any way in diabetics who drink? Thornalley: We have not studied this. Ren: Is the cross-linking a permissive step for acetaldehyde-induced cell injury? Could a cross-linking breaker be used to reverse or prevent this injury? Thornalley: There were some claims about thiazolium compounds that were called AGE breakers and could supposedly cleave dicarbonyl cross-links (Vasan et al

ENDOGENOUS α-OXOALDEHYDES IN TISSUE DAMAGE

245

1996). We published a paper refuting this work (Thornalley & Minhas 1999). It was flawed because the agent spontaneously degrades and acidifies cell culture media. Also, no one knows whether the cross-link that this agent was claimed to break actually exists in vivo. After our criticism, this group produced other compounds. M Salaspuro: With regard to your findings, what do you think is the real cause of the changes in plasma proteins? Is this a normal physiological or ageing process or is it somehow related to diabetes. Thornalley: There is damage occurring to plasma proteins continuously and this is increased in diabetes. In the diabetic state the half-life of albumin is not changed. In cells when proteins become damaged there is selective targeting of damaged proteins to the proteasome with release of damaged amino acids. Albano: Are the methylgyoxal-modified proteins able to interact with the scavenger receptors and do they have the capacity to trigger an immune response? Thornalley: There has been a lot of research on receptor-mediated responses to aldehyde-modified proteins. It very much depends on the extent of modification. If you highly modify albumin with a glucose-derived adduct or a chemically derived adduct (acetylation, succinylation) then the highly modified albumin derivative is recognized by the scavenger receptor. Scavenger receptor recognition depends on loss of charge of particularly important lysine residues. It doesn’t matter too much what the modification agent is. In vivo, albumin is minimally modified and is not competent to bind to scavenger receptors. There have also been studies proposing other receptors that these proteins bind to. One of the best characterized is RAGE, which is involved in inflammatory responses. Probably the agonist for this receptor is not a modified protein, but rather Ca2+-binding protein called calgra-nulins or S100 proteins, mediators of inflammatory processes. When proteins are modified to become competent agonists of this receptor it is necessary to modify them highly. It is the small amount of severely damaged albumin aggregate which is probably being recognized by this receptor. It will recognize fibrillar β-amyloid as well. Our current thought is that cell receptor-mediated responses to damaged proteins are not so important, but it may be that the accumulation of damaged proteins inside the cell is more important, or even the effects of the proteolytic products on the cell function. Albano: Did cells see these adducts as foreign compounds? Thornalley: If we incubate cultured cells with levels of these damaged amino acids, we can induce adverse effects. But these are at the levels found in renal failure, so they tend to be perhaps 10-fold higher than normal. Shukla: You looked at the effects of methylglyoxal on human artery endothelial cells. In the culture media how much glucose was present? Normal culture has 20 mM glucose so most of the proteins are already glycated. Thornalley: In those experiments the cells were not exposed to high glucose. It is the collagen that was exposed to the medium from cells that had been incubated

246

THORNALLEY

under hyperglycaemic conditions, or collagen modified by methylglyoxal. The cells cannot attach to the modified collagen even at normal glucose concentrations. References Thornalley PJ, Minhas HS 1999 Rapid hydrolysis and slow α,β-dicarbonyl cleavage of an agent proposed to cleave glucose-derived protein cross-links. Biochem Pharmacol 57:303–307 Vasan S, Zhang X, Kapurniotu A et al 1996 An agent cleaving glucose-derived protein crosslinks in vitro and in vivo. Nature 382:275–278 Yao D, Taguchi T, Matsumara T et al 2006 Methylglyoxal modification of m Sin 3A links glycolysis to angiopoietin-2 transcription. Cell 124:275–286

Measurement of acetaldehyde: what levels occur naturally and in response to alcohol? C. J. Peter Eriksson National Public Health Institute, Department of Mental Health and Alcohol Research, Helsinki, Finland

Abstract. The aim of the present paper is to update the status regarding human acetaldehyde levels in blood, breath and saliva during normal ethanol oxidation, i.e. without deficiency in, or inhibition of, aldehyde dehydrogenase activity. The previous conclusion according to which no detectable (<0.5 µM), adequately determined ‘free and/or loosely bound’ acetaldehyde has not yet been found in venous blood, more or less, still holds. The only new findings within this context consist of low venous blood acetaldehyde levels (1–3 µM on average) observed in some women during the use of oral contraceptives or during the high oestradiol phases of normal menstrual cycle. Breath acetaldehyde levels are about 10–20 and 20–40 nM at blood ethanol concentrations of about 10 and 20 mM, respectively. Theoretically calculated corresponding blood acetaldehyde levels in pulmonary blood would be about 2–4 and 4–8 µM. The acetaldehyde in the breath most likely reflects pulmonary blood acetaldehyde, microbial and tissue acetaldehyde production in the aerodigestive tract. As well as with breath acetaldehyde, salivary acetaldehyde levels also correlate positively with the blood ethanol concentrations. At blood ethanol concentrations of about 10 and 20 mM the average acetaldehyde concentration in saliva is about 15–25 and 20–40 µM, respectively. Saliva acetaldehyde represents mostly microbial acetaldehyde formation in the oral cavity, but also, to some extent, ethanol oxidation in nearby tissues. More studies are still needed to clarify the proportion of the underlying sources for blood, breath and salivary acetaldehyde at different ethanol concentrations. The problem with rapid acetaldehyde oxidation, which may markedly affect the recovery of low acetaldehyde levels, also needs to be solved. 2007 Acetaldehyde-related pathology: bridging the trans-disciplinary divide. Wiley, Chichester (Novartis Foundation Symposium 285) p 247–260

In a previous update on the status of human blood acetaldehyde (Eriksson & Fukunaga 1993) it was concluded that the concentration of ‘free’ and/or ‘loosely bound’ acetaldehyde is below detection (<0.5 µM) in venous blood during normal conditions, i.e. with no deficiency in, or inhibition of, aldehyde dehydrogenase activity. However, as was summarized, low concentrations (≤ 4 µM) in blood from 247

248

ERIKSSON

the right atrium and levels up to 68 µM in the hepatic vein had been determined with reliable methods during normal conditions, implying rapid disappearance of the acetaldehyde during the blood circulation. The aim of the present work is to briefly update the current status of ‘free’ and/or ‘loosely bound’ human blood acetaldehyde levels, as well as how these levels relate to new data regarding saliva and breath acetaldehyde concentrations. Both endogenous levels and concentrations during alcohol intoxication are included. The aspects of reduced aldehyde dehydrogenase capacity and increased alcohol oxidation by situational factors are excluded in the present work. Determination of acetaldehyde Blood acetaldehyde has traditionally been assayed by headspace gas chromatography and high-performance liquid chromatography (Eriksson & Fukunaga 1993). The detection limit has not been problematic. However, the problems and pitfalls have concerned artefactual acetaldehyde formation and disappearance during the procedures involving sampling, treatment of samples and even the assay processes (Eriksson 1980, 1983, Eriksson & Fukunaga 1993). Naturally, the main source of the artefactual acetaldehyde is the ethanol present in 1000–10 000 fold concentrations compared with the acetaldehyde levels. In particular, the denaturation of haemoglobin, possibly creating oxidative radicals, seems to be problematic. Depending on the treatment and assay procedures more or less acetaldehyde may also be released from bound sources and/or from chemical degradation reactions. Theoretically, since the transition from in vivo to in vitro takes some time with blood sampling, some remainder from possible in vivo ethanol oxidation in the blood may continue during the sampling procedure. Essential criteria for successful blood acetaldehyde determination are minimizing the artefactual acetaldehyde formation and to apply appropriate correction procedures for the remaining artefactual formation. ‘Appropriate correction’ means that different concentrations of ethanol are added to the control blood which is then treated exactly the same way as the test blood. The acetaldehyde values thus obtained after adding the ethanol to control blood are then subtracted from the test values obtained at corresponding ethanol concentrations. Examples of methods, which successfully have fulfilled the criteria of minimizing the artefactual formation (to about 1–3 µM at blood ethanol concentration of 10 mM) and which have applied the appropriate correction, are listed in Table 1. Rapid disappearance in human blood is a difficult issue regarding the accurate measurement of acetaldehyde. Such a reaction has been observed in several studies, in one of which it was demonstrated that the acetaldehyde was oxidized to acetate (Stowell et al 1980a). With a view to the rather long sampling time of human blood and the lack of knowledge how the transition occurs between the in vivo and in vitro

MEASURING ACETALDEHYDE

249

TABLE 1 Procedures and methods for human blood acetaldehyde determination with correction for minimal artefactual acetaldehyde formation Method

Reference

Semicarbazide trapping of acetaldehyde → protein precipitation of plasma → HSGC

Eriksson & Peachey (1980) (original without correction: Stowell 1979) Eriksson et al (1982) Lucas et al (1986) DeMaster et al (1983)

Perchloric acid/saline treatment → supernatant → HSGC (HPLC) Polyethylene glycol/Na-azide treatment → supernatant → HSGC

HSGC, headspace gas chromatography; HPLC, high-performance liquid chromatography.

states regarding the balance between possible acetaldehyde formation and elimination, there is a lot of uncertainty about the real recovery for low human blood acetaldehyde levels. The normally obtained high recoveries, obtained by adding the acetaldehyde to fresh blood and immediately stopping the biological reactions, doesn’t necessarily tell the whole story. To avoid the methodological problems regarding blood acetaldehyde determination, breath acetaldehyde measurements have been developed. An additional advantage with breath is that it should reflect the free and/or loosely bound acetaldehyde in the alveolar capillary membranes of the lungs. The methods are fairly easy, e.g. a portion of breath can directly be transferred into headspace bottles for further gas chromatography (GC) analyses (Sarkola et al 2002). Usually, the partition ratio of 1 : 190 has been used for the conversion of breath to blood concentrations ( Jones 1995). Saliva determinations have also proved to be useful in the estimation of acetaldehyde levels in the oral cavity. The most frequently used method has involved perchloric acid precipitation of a saliva sample followed by direct headspace GC measurement (Homann et al 1997). Blood acetaldehyde Although endogenous blood acetaldehyde levels usually below 1 µM, have been reported (Eriksson & Fukunaga 1993), the exact concentration (if any) of free endogenous acetaldehyde would be extremely difficult to assess due to the previously described technical difficulties (see section on determination of acetaldehyde). However, some estimation may be done based on endogenous breath acetaldehyde analyses as further discussed in the section on breath acetaldehyde.

250

ERIKSSON

Regarding the situation during normal ethanol intoxication, little has happened on the human blood acetaldehyde front since the previous update (Eriksson & Fukunaga 1993). However, the term ‘normal’ has to be well specified regarding to the context. Such a statement is supported by new findings demonstrating low acetaldehyde levels (1–3 µM on average) in some women during the use of oral contraceptives or during the high oestradiol phases of normal menstrual cycle (Fukunaga et al 1993, Eriksson et al 1996). In a more recent study with fewer test persons no significant blood acetaldehyde levels were detected in men, women using oral contraceptives, or in women not using oral contraceptives and close to the mid-cycle phase (Sarkola et al 2002). In this study, however, previous gender differences were supported by the fact that the alcohol dehydrogenase inhibitor 4-methylpyrazole decreased breath acetaldehyde levels exclusively in the women. The relativity of natural or normal conditions is also apparent with a view to genetic polymorphisms with such functional relevance for the regulation of ethanol and acetaldehyde oxidation which would affect blood acetaldehyde levels during alcohol intoxication. Best known examples are the polymorphisms of the aldehyde dehydrogenase genes ALDH1 and ALDH2 (see review by Eriksson 2001). Indirect evidence for an acetaldehyde elevating effect by the polymorphism of alcohol dehydrogenase genes ADH1B and ADH1C has also been reported (Eriksson 2001), although these polymorphisms have not yet been associated with corresponding blood acetaldehyde concentration differences. Acetaldehyde in breath In an excellent review by Jones (1995) on measuring and reporting human acetaldehyde concentrations in breath, it was concluded that the endogenous acetaldehyde concentrations vary up to about 0.6 nM in the breath. Data by Shaskan & Dolinsky (1985), and by Hesselbrock & Shaskan (1985), show that both smoking and alcoholism elevate endogenous acetaldehyde in breath. Thus, it seems that the real endogenous acetaldehyde in the breath of non-smokers and in combination with no history of excessive alcohol consumption varies between 0.02 to 0.25 nM (Dannecker et al 1981, Shaskan & Dolinsky 1985). These levels correspond to 0.004 to 0.050 µM in blood, using the conversion coefficient of 190. Such levels would be virtually impossible to detect accurately in blood considering the technical difficulties involved in pre-analytical stages. The relationship between blood ethanol and acetaldehyde concentration in breath during alcohol intoxication in white populations without history of alcoholism, smoking during drinking or known genetic deficiency of ALDH activity are listed in Table 2. In spite of separate investigators and populations there is a remarkable resemblance regarding the magnitude of acetaldehyde levels in breath during ethanol oxidation. Thus, at blood ethanol concentrations of about 10, 20 and 30 mM the

MEASURING ACETALDEHYDE

251

TABLE 2 Relationship between breath and blood acetaldehyde levels and blood ethanol concentration during alcohol intoxication in white populations without history of alcoholism and smoking during experiments Reference

Population

Stowell et al (1980a) Stowell et al (1980b) Couchman & Crow (1980) Stowell et al (1984) Jones et al (1984) Truitt et al (1987) Sarkola et al (2002) Preliminary new dataa

5 men + 4 women 5 men 6 men + 4 women 3 men 4 men 4 men + 7 women 13 men 22 women 13 men + 7 women

Blood ethanol (mM)

Breath acetaldehyde (nM)

Calculated blood acetaldehyde (mM)

22 14 29 24 20 10 22 9 20 10 12 8 12 11 34 10

26 19 42 34 20 10 39 14 39 26 26 11 28 23 57 21

5 4 8 6 4 2 7 3 7 5 5 2 5 4 11 4

Values represent mean estimations. Blood acetaldehyde is calculated on basis of a 1 : 190 partition ratio between breath and blood. a The experimental design is described in the section on human intervention study.

average acetaldehyde concentration in breath seems to be about 10–20, 20–40 and 40–60 nM, respectively. Calculated corresponding blood acetaldehyde levels in pulmonary blood would be about 2–4, 4–8 and 8–12 µM, respectively. That breath acetaldehyde concentrations correlate so well with the blood ethanol levels (see Table 2), the finding that 4-methylpyrazole only partly lowers breath levels (Sarkola et al 2002), that breath acetaldehyde during ‘normal’ conditions doesn’t correlate so well with blood acetaldehyde (Jones 1995) and that acetaldehyde levels in venous blood are below detection even at elevated ethanol concentrations, implies that other sources than the systemic acetaldehyde derived by the first-order hepatic alcohol oxidation contribute to the acetaldehyde in the breath. Microbial ethanol oxidation in the upper respiratory tract has been demonstrated to be such a source (Pikkarainen et al 1980, Jauhonen et al 1982). Acetaldehyde in saliva Salivary acetaldehyde levels in relation to blood ethanol concentrations during ethanol oxidation are listed in Table 3. As with the breath acetaldehyde, salivary

252

ERIKSSON

TABLE 3 Relationship between acetaldehyde levels in saliva and blood ethanol concentration during alcohol intoxication in subjects without history of alcoholism and smoking during experiments Reference

Subjects

Homann et al (1997) Väkeväinen et al (2000) Väkeväinen et al (2001) Salaspuro et al (2002) Visapää et al (2006)

6 men + 4 women 13 men + womena 6 men 9 men 4 men (ADH1C1*1)b 8 men (ADH1C1*2 + 2*2)

Preliminary new datac

13 men + 7 women

Blood Ethanol (mM)

Salivary acetaldehyde (mM)

18 8 18 10 8 4 43 19 40 20 10 40 20 10 34 10

35 26 34 21 21 13 41 32 85 41 21 40 19 10 46 16

Values represent mean estimations. a Subjects are Asians with the ALDH21*1 genotype. The rest of the subjects listed in the table represent Finnish populations. b Subjects with different alcohol dehydrogenase genotypes are compared. c The experimental design is described in the section on human intervention study.

acetaldehyde levels also correlate positively with the blood ethanol concentrations. Thus, at blood ethanol concentrations of about 10, 20 and 30–40 mM the average acetaldehyde concentration in saliva is reported to be about 15–25, 20–40 and 40–80 nM, respectively. In addition to the role of the ethanol concentration, it has been shown that a considerable part of the salivary acetaldehyde variation is also caused by the status of the microflora in the mouth (Homann 1997). Recent findings by Visapää et al (2006) support the possibility that genetic differences in the systemic ADH-dependent alcohol oxidation also may be reflected in the salivary acetaldehyde levels (Table 3). Previously it has been demonstrated that 4-methylpyrazole, which is not such an efficient inhibitor of microbial ethanol oxidation, doesn’t affect salivary acetaldehyde levels until these are elevated to such a degree that acetaldehyde concentration may also be detected in venous blood (Väkeväinen et al 2000). Human intervention study At present, we are conducting an ongoing alcohol intervention study in our laboratory. The study design is a crossover experiment. Each healthy volunteer partici-

MEASURING ACETALDEHYDE

253

100

Acetaldehyde in breath (nM)

80

60

40

20

0 0

20

40

60

80

100

Salivary acetaldehyde (µM)

FIG. 1. Correlation between acetaldehyde concentrations in saliva and breath. Spearman’s ρ = 0.764, P < 0.001. Experimental design is described in the section on human intervention study.

pates in four two-day (evening and following morning) sessions. The sessions include, in random order, placebo/placebo, placebo/alcohol, alcohol/placebo and alcohol/alcohol in the evening /morning. The evening alcohol dose is 1.5 and 1.4 g /kg (as 10%, v/v, in lingonberry juice, drunk in 3 hours, starting 7 pm) for men and women, respectively. The morning dose is 0.5 g /kg (drunk in 30 min, starting 11 am). Blood, breath and saliva samples for ethanol and acetaldehyde determination are taken before and up to 4 and 3 hours after start of drinking at evening and morning sessions, respectively. Acetaldehyde and ethanol in breath were determined according to Sarkola et al (2002) and in saliva and blood as described earlier for blood (Eriksson et al 1982). No significant artefactual acetaldehyde was observed with the saliva determinations. The preliminary data from the alcohol sessions of 13 men and 7 women (age range: 23–63 years) are depicted in Fig. 1. A significant positive correlation between salivary and breath acetaldehyde concentrations is displayed, which has not earlier been reported. Also, the partial correlation between salivary and breath acetaldehyde adjusted for blood ethanol concentrations was significant (r = 0.557, P < 0.001). Hence, it is quite clear that even acetaldehyde in the breath is to some degree affected by microflora in upper oral cavities, as well as in the upper respiratory tract. This may provide an additional reason for why breath acetaldehyde levels exceed the corresponding blood acetaldehyde levels.

254

ERIKSSON

Conclusions (1) Based on breath acetaldehyde data the endogenous venous blood acetaldehyde levels may be estimated to range between 0.00–0.05 µM in white subjects with no recent history of smoking or drinking alcohol. (2) Usually, during alcohol intoxication, no significant acetaldehyde concentrations can be detected in venous blood during conditions without deficiency in or inhibition of aldehyde dehydrogenase activity, and/or increased acetaldehyde production by elevated alcohol oxidation rate. Such conditional changes may, however, explain why in some women, during the use of oral contraceptives or during the high oestradiol phases of normal menstrual cycle, low acetaldehyde levels (about 1–3 µM) can be detected. (3) It remains to be solved how fast acetaldehyde disappears and how much this affects the recovery at human blood acetaldehyde measurements. (4) The acetaldehyde in the breath most likely reflects pulmonary blood acetaldehyde, microbial acetaldehyde production in the upper pathway of the respiratory and oral tracts, and possible local ethanol metabolism in the respiratory tracts. The higher the ethanol concentration, the higher the microbial contribution. (5) Saliva acetaldehyde represents mostly microbial acetaldehyde in the upper oral cavities, but also, to some extent, local ADH-dependent ethanol oxidation. References Couchman KG, Crow KE 1980 Breath acetaldehyde levels after ethanol consumption. In: Thurman RG (ed) Alcohol and aldehyde metabolizing systems –IV, Plenum Press, New York, p 451–457 Dannecker JR, Shaskan EG, Phillips M 1981 A new highly sensitive assay for breath acetaldehyde: detection of endogenous levels in humans. Anal Biochem 114:1–7 DeMaster EG, Redfern B, Weir EK, Pierpont GL, Crouse LJ 1983 Elimination of artifactual acetaldehyde in the measurement of human blood acetaldehyde by the use of polyethylene glycol and sodium azide: normal blood acetaldehyde levels in the dog and human after ethanol. Alcohol Clin Exp Res 7:436–442 Eriksson CJP 1980 Problems and pitfalls in acetaldehyde determinations. Alcohol Clin Exp Res 4:22–29 Eriksson CJP 1983 Human blood acetaldehyde concentration (update 1982). Pharmacol Biochem Behav 18 Suppl 1:141–150 Eriksson CJP 2001 The role of acetaldehyde in the actions of alcohol (update 2000). Alcohol Clin Exp Res 25 Suppl: 15S–32 Eriksson CJP, Peachey JE 1980 Lack of difference in blood acetaldehyde of alcoholics and controls after ethanol ingestion. Pharmacol Biochem Behav 13 Suppl 1.101–105 Eriksson CJP, Fukunaga T 1993 Human blood acetaldehyde (update 1992). Alcohol Alcohol Suppl 2:9–25 Eriksson CJP, Mizoi Y, Fukunaga T 1982 The determination of acetaldehyde in human blood by the perchloric acid precipitation method: the characterization and elimination of artefactual acetaldehyde formation. Anal Biochem 125:259–263

MEASURING ACETALDEHYDE

255

Eriksson CJP, Fukunaga T, Sarkola et al 1996 Estrogen-related acetaldehyde elevation in women during alcohol intoxication. Alcohol Clin Exp Res 20:1192–1195 Fukunaga T, Sillanaukee P, Eriksson CJP 1993 Occurrence of blood acetaldehyde in women during ethanol intoxication: preliminary findings. Alcohol Clin Exp Res 17:1198–1200 Hesselbrock VM, Shaskan EG 1985 Endogenous breath acetaldehyde levels among alcoholic and non-alcoholic probands: effect of alcohol use and smoking. Prog Neuropsychopharmacol Biol Psychiatry 9:259–265 Homann N, Jousimies-Somer H, Heine R, Salaspuro M 1997 High acetaldehyde levels in saliva after ethanol consumption: methodological aspects and pathogenetic implications. Carcinogenesis 18:1739–1743 Jauhonen P, Baraona E, Miyakawa H, Lieber CS 1982 Origin of breath acetaldehyde during ethanol oxidation. J Lab Clin Med 100:908–916 Jones AW 1995 Measuring and reporting the concentration of acetaldehyde in human breath. Alcohol Alcohol 30:271–285 Jones AW, Skagerberg S, Borg S, Änggàrd E 1984 Time course of breath acetaldehyde concentrations during intravenous infusions of ethanol in healthy men. Drug Alcohol Depend 14:113–119 Lucas D, Ménez JF, Berthou F et al 1986 Determination of free acetaldehyde in blood as the dinitrophenylhydrazone derivative by high-performance liquid chromatography. J Chromatogr 382:57–66 Pikkarainen P, Baraona E, Seitz H, Lieber CS 1980 Breath acetaldehyde: evidence of acetaldehyde production by oropharynx microflora and by lung microsomes. In: Thurman RG (ed) Alcohol and aldehyde metabolizing systems –IV, Plenum Press, New York, p 469–474 Salaspuro V, Hietala J, Kaihovaara P et al 2002 Removal of acetaldehyde from saliva by a slowrelease buccal tablet of L-cysteine. Int J Cancer 97:361–364 Sarkola T, Iles MR, Kohlenberg-Mueller K, Eriksson CJP 2002 Ethanol, acetaldehyde, acetate, and lactate levels after alcohol intake in White men and women: effect of 4-methylpyrazole. Alcohol Clin Exp Res 26:239–245 Shaskan EG, Dolinsky ZS 1985 Elevated endogenous breath acetaldehyde levels among abusers of alcohol and cigarettes. Prog Neuropsychopharmacol Biol Psychiatry 9:267–272 Stowell AR 1979 An improved method for the determination of acetaldehyde in human blood with minimal ethanol interference. Clin Chim Acta 98:201–205 Stowell AR, Crow KE, Couchman KG, Batt RD 1980a Acetaldehyde levels in peripheral venous blood and breath of human volunteers. Adv Exp Med Biol 126:425–438 Stowell A, Hillbom M, Salaspuro M, Lindros KO 1980b Low acetaldehyde levels in blood, breath and cerebrospinal fluid of intoxicated humans as assayed by improved methods. In: Thurman RG (ed) Alcohol and aldehyde metabolizing systems –IV, Plenum Press, New York, p 635–645 Stowell A, Johnsen J, Aune H et al 1984 A reinvestigation of the usefulness of breath analysis in the determination of blood acetaldehyde concentrations. Alcohol Clin Exp Res 8:442–447 Truitt EB, Gaynor CR, Mehl DL 1987 Aspirin attenuation of alcohol-induced flushing and intoxication in Oriental and Occidental subjects. Alcohol Alcohol Suppl 1:595–599 Visapää J-P, Götte K, Benesova M et al 2006 Increased cancer risk in heavy drinkers with the alcohol dehydrogenase 1C* allele, possibly due to salivary acetaldehyde. Gut 53: 871–876 Väkeväinen S, Tillonen J, Agarwal DP et al 2000 High salivary acetaldehyde after moderate dose of alcohol in ALDH2-deficient subjects: strong evidence for local carcinogenic action of acetaldehyde. Alcohol Clin Exp Res 24:873–877 Väkeväinen S, Tillonen J, Salaspuro M 2001 2-Methylpyrazole decreases salivary acetaldehyde levels in ALDH2-deficient subjects but not in subjects with normal ALDH2. Alcohol Clin Exp Res 25:829–834

256

ERIKSSON

DISCUSSION Preedy: You used the statement ‘no significant acetaldehyde’, but shouldn’t that rather read ‘no significant free acetaldehyde’? There may be substantial amounts bound. Eriksson: You are right. One researcher once made the correction in such a way that he took away all the minus values. After this he got a small plus value and reported this as the level. We come to a situation where after the subtraction you also have minus acetaldehyde levels, which is absurd in a way. But since you are doing an average correction you have to consider all forms of technical variation. Preedy: Have you taken blood and hydrolysed it or treated it to release any additional free acetaldehyde that you might capture by a vacuum tube, and then fed this into the gas chromatograph? Eriksson: We are interested in the more firmly bound acetaldehyde. We published one paper on this aspect a few years ago (Fukunaga et al 1993). It is a question of temperature. The more this increased the more you find. I would like to look at all the tissues and try to release the acetaldehyde. There are enormously complex biochemical reactions. But it would be nice to see if one could release molecular acetaldehyde. Morris: What is the normal dietary intake of acetaldehyde? It has been reported to be 40–80 mg/day (Morris et al 1996). Even people who smoke a pack of cigarettes a day only derived one-third of the acetaldehyde that they get from their diet through smoke. Eriksson: Mikko Salaspuro is in the process of investigating this. We found that smokers had a lot of acetaldehyde coming from their breath after smoking one cigarette, which suggests high levels of bound acetaldehyde in the lungs far in excess of levels reached during alcohol intoxication. Morris: The enigma is that if you measure the constituents in mainstream cigarette smoke there is less than a milligram of acetaldehyde. Eriksson: That may be, but the levels of acetaldehyde in the breath after smoking are far higher than in the breath after alcohol intake. Morris: Unless smoking is releasing bound acetaldehyde from a sink in some way. The fact that there is not much acetaldehyde in smoke would suggest that it is not derived directly from tobacco smoke. Eriksson: What is the other source? Morris: There are a lot of materials in tobacco smoke that will change the pH of your mucosa. Eriksson: That’s interesting, but there would need to be an initial source of acetaldehyde. If it is impossible that this amount could come from cigarettes, then it has to come from other sources.

MEASURING ACETALDEHYDE

257

M Salaspuro: When we studied acetaldehyde concentrations in the hepatic vein, we were only able to find acetaldehyde in volunteers who were very heavy drinkers. We must have a very efficient system in the liver for removing acetaldehyde. Eriksson: More than 99% is removed. M Salaspuro: In chronic drinkers, we find some leakage from the liver to the hepatic vein. As far as I remember 70 µM was the highest value we found and it was only a peak in one person. We have tested many microbes with a variety of different ADHs. Some microbial ADHs may be inhibited by rather high concentrations of 4-methylpyrazole but some others aren’t. Emery: What is the chemical form of bound acetaldehyde? Eriksson: Acetaldehyde forms a loose Schiff base with most amino acids, and especially with those that have more free amino groups. There is an abundance of different possibilities for binding acetaldehyde. One definition of loosely bound is that it is easily reversible under physiological conditions. Thornalley: We have done dicarbonyl measurements. From what we know of the reversible binding equilibria with plasma protein for methylglyoxal, greater than 95% of the 100 nM methylglyoxal in plasma of normal, healthy control human subjects is reversibly bound to plasma protein in situ. When we measure methylglyoxal in plasma, we trap free and reversibly bound forms by derivatization with diaminobenzene. Has anyone done stable isotope labelling studies? I don’t know how expensive [13C]ethanol is, but this could be helpful in trying to characterize or control the formation of acetaldehyde during sample processing. You would get isotopic dilution by the acetaldehyde that is being produced from other sources. Eriksson: No one has done this to my knowledge. Thornalley: This would help to control for artefactual production (but only from non-alcohol sources). Eriksson: How do you label it? The artefact comes from the alcohol you have labelled. Thornalley: You’d have [13C]ethanol and all the acetaldehyde coming from ethanol should retain that. If there is acetaldehyde coming from other sources it won’t be labelled. Eriksson: That’s true, but you can’t control for the artefact because it is the same alcohol that is oxidized. Seitz: What is the reason for the oestradiol story, where there was higher acetaldehyde? We have given alcohol to premenopausal women at different phases of their menstrual cycle. We found that oestradiol levels are increased. If we give small amounts leading to blood alcohol levels of 20 mg alcohol per 100 ml blood we have seen an increase in oestradiol of 30–40%. What is the explanation for higher acetaldehyde levels with higher oestradiol? Eriksson: There are two explanations. You could be increasing the alcohol oxidation rate or you could be inhibiting the ALDH activity. The problem is that because

258

ERIKSSON

of the kinetics of acetaldehyde levels, just decreasing the efficacy of ALDH by 0.5 or 1% will cause the acetaldehyde levels to increase noticeably. So a small effect can be responsible, and this could be hard to find. If you increase alcohol oxidation by 10% this has just a minor effect on acetaldehyde levels because of an overcapacity of ALDH. Seitz: This is interesting because three groups have shown that the ADH1C1 homozygosity in females is associated with breast cancer. This means, that in these patients acetaldehyde levels are probably higher and may contribute to breast cancer pathogenesis. Eriksson: We have looked carefully at the effect of alcohol on oestradiol, but couldn’t really see anything significant. It is in the postmenopausal women that more substantial effects are found. Oestrogens are transcription factors. Perhaps the ADH1C gene is a candidate for oestradiol-mediated transcription. It would be interesting to know. Morris: If components of tobacco smoke inhibited oral ALDH this would result in increased acetaldehyde levels. The endogenous acetaldehyde would not be rapidly destroyed by the ALDH in this case. Eriksson: Where is the acetaldehyde coming from if it isn’t there to start with? Morris: It could be coming from the oral bacterial sources. V Salaspuro: You can also have these kinds of acetaldehyde concentrations if you elute smoke through sterile water. Eriksson: There is also a rat model in which we could trap the acetaldehyde. Rao: I am curious about the rapid loss of acetaldehyde. You showed that acetaldehyde level in blood is rapidly decreased after blood withdrawal. Do you know the reason for this? Eriksson: Stowell did this very well in 1980/81. You can pick it up as the acetate. It is not evaporating and it isn’t bound. It is an acetaldehyde oxidation. I don’t know what the exact mechanism for this phenomenon is. Rao: Is it possible that you lose the source of acetaldehyde production in the endothelial cells when you remove blood? Eriksson: This problem has been dwelling in my mind for a while. What I showed was an average from four people. This is a fairly easy study so I don’t know why it has been neglected. If you want to do it really carefully you need another group in which you also add the alcohol. There may be ongoing alcohol and acetaldehyde metabolism in blood. The minute you shift from in vivo to in vitro, it changes this balance. The same may be true to some degree with the saliva. Rao: If you use an inhibitor of ALDH in the tube, we can get around this problem. Eriksson: That is a good idea. Worrall: How much blood do you need?

MEASURING ACETALDEHYDE

259

Eriksson: In the early days I used to drip blood into liquid nitrogen. In those days I didn’t correct in the right way. Apte: For those of us interested in tissue effects of alcohol, the challenge is to measure the tissue acetaldehyde level. Eriksson: When you measure tissue acetaldehyde you also have to eliminate the artefactual formation. In tissue with remaining blood you have in addition some of the blood problems. In the tissue itself, it seems that in all the tissues we have looked at (brain, liver and kidney), thiourea completely blocks it (Eriksson et al 1977). Apte: In my model of alcoholic pancreatitis, for example, what do I do with the pancreatic homogenates? Eriksson: You just try to inhibit the artefactual acetaldehyde formation by earlier described methods for other tissues (Eriksson et al 1977). Seitz: You can use the DHI method. We put the tissue or the blood directly into the solution which contains the adduct forming compound so that we homogenize the tissue with the compound. As soon as the adduct formation takes place, acetaldehyde is stable and can be measured by HPLC, using fluorescence detection. This gives nearly the same values as if you use the headspace gas chromatogram. Apte: The problem with the pancreas is the presence of digestive enzymes. Okamura: What is the long term stability of acetaldehyde in saliva or blood? We have kept samples for 5–10 years in the freezer. Eriksson: It is a good question. If you take blood and put it in the freezer, you can forget about measuring acetaldehyde after some time. During the freezing and thawing all kinds of things happen. They are very difficult to control. If you do the precipitation and save the supernatant, and you have controls, then you compare the recoveries. For saliva, the artefactual formation is much smaller. You might have a chance here of saving samples. It would probably be advisable to make the precipitate first. M Salaspuro: We have done some preliminary studies in order to get saliva samples from China to Finland. If we keep saliva at –70 °C for two weeks there will be about a 10% decrease in its acetaldehyde content. Eriksson: A good idea in this case would be to add a standard concentration to control saliva. Yin: Regarding the methodology of acetaldehyde determination, as in the blood the acetaldehyde concentration is three orders of magnitude lower than ethanol. This causes problems for headspace gas chromatography. Semicarbazide is a good method because it forms a stable derivative with acetaldehyde, but detecting it in the visible range is less sensitive than that by fluorescence detection method. Eriksson: The problem with the semicarbazide is that it may also cause artefactual acetaldehyde formation. There is something going on with the semicarbazide: when

260

ERIKSSON

it is stored even for a short time some artefactual formation occurs. We didn’t notice this at first. Yin: The method we have been using is a modification of the reported method with a fluorescent 1,3-cyclohexanedione derivative of acetaldehyde (Peng et al 1999). The sensitivity is high but we had problems with the background reading. Tim Peters and his co-workers have another fluorescent derivative method (Rideout et al 1986). He told me that the fluorescence is relatively stable. I haven’t tried this yet. What kind of method do you recommend for blood acetaldehyde? Eriksson: If you measure acetaldehyde from normal venous blood and you find some then you are probably doing it wrong! It’s an easy way of testing techniques. The early methods Peters and co-workers used had lots of artefactual formation. The real problem is before you come to the detection stage. Yin: Our experience is that acetaldehyde determination is quite tricky, because the physiological range is so low. This requires careful determination. References Eriksson CJP, Sippel HW, Forsander OA 1977 The determination of acetaldehyde in biological samples by head-space gas chromatography. Anal Biochem 80:116–124 Fukunaga T, Sillanaukee P, Eriksson CJP 1993 Problems involved in the determination of endogenous acetaldehyde in human blood. Alcohol Alcohol 28:535–541 Morris JB, Robinson DE, Vollmuth TA, Brown RP, Domeyer BE 1996 A parallelogram approach for safety evaluation of ingested acetaldehyde. Regul Toxicol Pharmacol 24:251–263 Peng GS, Wang MF, Chen CY et al 1999 Involvement of acetaldehyde for full protection against alcoholism by homozygosity of the variant allele of mitochondrial aldehyde dehydrogenase gene in Asians. Pharmacogenetics 9:463–476 Rideout JM, Lim CK, Peters TJ 1986 Assay of blood acetaldehyde by HPLC with fluorescence detection of its 2-diphenylacetyl-1,3-indandione-1-azine derivative. Clin Chim Acta 161: 29–35

FINAL DISCUSSION

Emery: We mentioned earlier this whole question of the extent to which there is oxidative stress induced by acetaldehyde. This seems slightly paradoxical. Thornalley: If you compare the mechanism of oxidative stress in vascular cells in hyperglycaemia, oxidative stress is driven by increased electron flow into mitochondria via the glycerophosphate shuttle, leading to excessive electron flow into complex III and increased superoxide and peroxide formation. Similar electron flows may occur in ethanol metabolism: increased reducing equivalents produced by ethanol and acetaldehyde oxidation may flow into mitochondria producing similar mitochondrial dysfunction and formation of reactive oxygen species (ROS). Another possibility is formation of glyoxal from ethanol metabolism may lead to glyoxal modification of mitochondrial proteins, mitochondrial coupling and increased ROS formation. Albano: I think that the impairment of mitochondrial respiration and inflammation might be a substantial source of ROS in the tissues exposed to acetaldehyde. However, it cannot be excluded that acetaldehyde metabolism by oxidase enzymes will probably contribute to oxidative stress in some tissues as the heart or the skeletal muscles. This deserves further investigations. Seitz: Then there is the acetaldehyde effect on the antioxidant defence system. Also, there may be an effect on fatty liver due to acetaldehyde damage of mitochondria which are the major site of fatty acid oxidation. Acetaldehyde may affect microtubules as well, and may inhibit by this mechanism secretion of fat as very low density lipoprotein (VLDL) from the liver to the blood. Free fatty acid by itself may have properties of oxidative stress. Apte: I know I was the one who set the cat among the pigeons about oxidative stress. I am not wedded to the idea, but we have to acknowledge that there are other pathways of alcohol-induced damage in addition to acetaldehyde, such as the nonoxidative pathway which cannot be ignored, especially in the pancreas. M Salaspuro: I became interested in what Peter Eriksson said about ‘acetaldehydism’. How much evidence do we have for that? We can establish animal models for alcohol dependence, but is there an animal model for acetaldehyde dependency? Acetaldehyde can be given to experimental animals by inhalation or in drinking water. But are there any reports of withdrawal symptoms caused by long-term acetaldehyde exposure or acetaldehyde dependency? 261

262

FINAL DISCUSSION

Quertemont: Animal studies have shown that acetaldehyde has reinforcing effects at certain concentrations. The problem with chronic acetaldehyde administration is its toxic effects. The question that remains about acetaldehyde is more whether it mediates or is involved in alcohol dependence. There is still a problem of acetaldehyde concentrations in the brain. At the moment we don’t know whether there is sufficient acetaldehyde in the brain to induce a reinforcing effect when we inject ethanol. Eriksson: There are two aspects to what Mikko Salaspuro just said. He was talking about withdrawal, which is physical dependence. If you are an alcoholic, perhaps you want to drink to avoid the withdrawal, but it is not the withdrawal that makes you dependent on alcohol. There is no evidence that pure alcohol without any metabolism could cause any addiction at all. But I would say that the level is not necessarily so important here. Within the brain the metabolism is responsible for its production. For acetaldehyde, the level is almost a silly concept: all the time something happens with the acetaldehyde. The acetaldehyde that is formed is directly bound or metabolized, and during that process something is affected. Emery: So you are saying that what matters is the acetaldehyde flux. Eriksson: Yes, but I would not disregard alcohol totally, because it is doing something irrespective of acetaldehyde. It is also causing redox reactions, and its metabolism is affecting other systems, causing local changes. I think pharmacologists have placed too much emphasis on looking for a specific receptor system for alcohol’s effects. Apte: Do you need a specific receptor for alcohol? It is so diffusible. Eriksson: I’m not at all sure what we need. The whole thing can be based on some other reactions that we don’t know about yet. Shukla: Acetaldehyde may be a minor player in many situations. Non-oxidative products of ethanol may have important roles and this remains to be explored. For example, the brain is loaded with phospholipid D; a good site for formation of phosphatidyl ethanol (PEth). The question therefore arises whether PEth has any relevance in addiction. Eriksson: You yourself have presented data that fatty acid alcohol esters may be related to the toxic and harmful effects. I don’t think it is so much related to the addiction. Theoretically, you can come up with whatever idea you like. We are usually putting the most emphasis on the most probable hypotheses. Deitrich: We need to be careful about the definition of a ‘receptor’. It’s true that we can’t measure ethanol binding to a specific ‘receptor’ in the brain, but it does have a specific receptor—alcohol dehydrogenase, which can easily be measured. In the brain the studies with the GABA receptor and ethanol’s binding to it, whether you call this an ethanol receptor or not, you can change this by changing a single amino acid and lose the effect (Mascia et al 2000, Ueno et al 2000, Israel et al 2006).

FINAL DISCUSSION

263

To my way of thinking, this is the best definition of a receptor. The problem is that the ethanol has such a low binding capacity, you need huge amounts. It doesn’t bind to its ‘receptor’ very well. Yin: With regard to the notion of acetaldehydism, the homozygous ALDH2deficient alcoholic I mentioned earlier would be an extreme example of that. He continuously sipped only three to five bottles of beer a day. His blood ethanol levels would have been well below 1 mM, but he still showed alcohol dependence by the criteria of DSM-III-R. Eriksson: One of the aldehyde dehydrogenase (ALDH) inhibitors was withdrawn from the market because it was known that some alcoholics were going for low amounts of alcohol coupled with ALDH inhibitors. Seitz: There was a similar story with cimetidine in American Indians. Eriksson: Many pieces of evidence for the role of acetaldehyde in alcohol addiction have been reported (Eriksson 2001). Perhaps the best evidence comes from a recent publication that shows that polymorphism in the ALDH2 gene explains differences in alcohol expectancy. Thus, the ALDH2*1/2 genotype (creating elevated acetaldehyde levels) is associated with more positive expectancy of drinking alcohol comparing with the ALDH2*1/1 genotype (Hahn et al 2006). Emery: In closing this meeting I’d like to comment briefly on a couple of things that have occurred to me during our discussions. A lot of this meeting has focused on the role of ethanol, and I’m sure this is still the major source of the acetaldehyde. It would be useful to try to put this into perspective in relation to other sources. Cigarette smoke and air pollution presented some difficulties: small amounts of acetaldehyde are coming in yet there is evidence that this is having more of an effect than we might expect. More work needs to be done here. Food is an unexplored area, including drink. We have covered metabolism: there are a variety of metabolic pathways. I take the point that the subsequent products such as acetate have an important role. How much acetaldehyde is present? This seems to be a difficult one to answer. We measure it where we can, but this isn’t necessarily where we need to be looking. With regard to mechanisms, we found that there are many more questions to answer than we even thought of. There’s a lot of work to do here. Much of the work that is beginning to reveal interesting molecular mechanisms is still in vitro, so it must be interpreted with caution. We have heard a fair amount about the adducts: not just about the adducts in general, but also which proteins are affected, and which residues within these proteins. Indeed, the whole kinetics of the formation and degradation of those adducts will be an issue we need to explore further. Lots of tissues are affected, and the acetaldehyde seems to be present pretty much everywhere we seek it. What are we going to do about it? We’ve heard some interesting ideas, but we need to think of what it is appropriate to try to achieve. A benefit of this sort of meeting is the contacts people have made and the possible collaborations that will result. I’d like to thank you all for your participation.

264

FINAL DISCUSSION

References Eriksson CJ 2001 The role of acetaldehyde in the actions of alcohol (update 2000). Alcohol Clin Exp Res 25(5 Suppl ISBRA):15S–32S Hahn CY, Huang SY, Ko HC et al 2006 Acetaldehyde involvement in positive and negative alcohol expectancies in Han Chinese persons with alcoholism. Arch Gen Psychiatry 63: 817–823 Israel Y, Quintanilla ME, Sapag A, Tampier L 2006 Autosomal and maternal genes influence alcohol intake in alcohol drinker and nondrinker rat lines: role of the ‘acetaldehyde burst’. Alcohol Clin Exp Res 30:276A Mascia MP, Trudell JR, Harris RA 2000 Specific binding sites for alcohols and anesthetics on ligand-gated ion channels. Proc Natl Acad Sci USA 97:9305–9310 Ueno S, Lin A, Nikolaeva N et al 2000 Tryptophan scanning mutagenesis in TM2 of the GABAA receptor a subunit: effects on channel gating and regulation by ethanol. Br J Pharmacacol 131: 296–302

Contributors Index

Non-participating co-authors are indicated by asterisks. Entries in bold indicate papers; other entries refer to discussion contributions.

A

F

Albano, E. 20, 41, 42, 43, 44, 76, 93, 123, 154, 177, 181, 196, 211, 212, 214, 215, 224, 225, 227, 245, 261 *Ameno, K. 137 Apte, M. V. 17, 18, 41, 42, 44, 46, 48, 49, 64, 76, 77, 78, 90, 93, 94, 106, 107, 108, 120, 122, 124, 135, 136, 143, 155, 178, 195, 196, 200, 212, 213, 214, 215, 216, 225, 227, 228, 244, 259, 261, 262 Aranda, A. 48, 156, 215, 216 *Aroor, A. R. 217 *Asai, S. 97

*Farrés, J. 158 *Fukushima, C. 97

C

*Kohno, S. 97 *Kumihashi, M. 137

Crabb, D. W. 4, 16, 17, 18, 19, 21, 43, 45, 46, 47, 63, 64, 92, 108, 119, 154, 155, 156, 158, 181, 214, 215

I *Ijiri, I. 137 J Jamal, M. 50, 137, 142, 143 K

L

D

*Lee, Y. J. 217 *Liangpunsakul, S. 4

Deitrich, R. A. 16, 19, 23, 40, 41, 43, 44, 45, 46, 47, 48, 49, 67, 107, 108, 121, 143, 154, 262

M

E Emery, P. W. 1, 18, 42, 44, 45, 48, 50, 64, 77, 79, 91, 92, 94, 95, 108, 119, 120, 121, 122, 123, 153, 154, 156, 158, 178, 179, 199, 243, 257, 261, 262, 263 Eriksson, C. J. P. 17, 18, 21, 40, 43, 44, 45, 46, 47, 49, 50, 64, 65, 66, 67, 78, 79, 91, 92, 93, 94, 107, 123, 133, 134, 135, 136, 142, 143, 154, 155, 156, 177, 212, 213, 226, 227, 247, 256, 257, 258, 259, 260, 262, 263

*McCarroll, J. 200 Matsuse, H. 97, 106, 107, 108 Morris, J. B. 41, 89, 91, 92, 93, 106, 107, 108, 135, 136, 199, 226, 256, 258 N Niemelä, O. 18, 49, 77, 107, 180, 183, 194, 195, 196, 197, 199, 214, 244 O Okamura, T. 48, 49, 67, 94, 122, 259 265

266 P *Park, P.-h. 217 *Peng, G.-S. 52 *Petersen, D. 23 *Pirola, R. 200 Preedy, V. R. 41, 42, 47, 64, 77, 78, 89, 94, 95, 123, 133, 142, 153, 156, 158, 177, 178, 179, 180, 181, 195, 199, 212, 215, 224, 256

CONTRIBUTORS INDEX 212, 214, 215, 225, 244, 257, 258, 259, 261, 263 *Shimoda, T. 97 Shukla, S. D. 20, 48, 49, 77, 120, 178, 179, 213, 214, 217, 224, 225, 226, 227, 228, 245, 262 T

Quertemont, E. 45, 65, 66, 67, 141, 143, 262

*Tardif, R. 125 Thornalley, P. J. 19, 45, 108, 119, 120, 135, 179, 180, 193, 194, 195, 199, 227, 229, 243, 244, 245, 257, 261

R

U

Rao, R. K. 47, 76, 77, 78, 90, 94, 107, 120, 123, 135, 154, 155, 180, 195, 196, 199, 213, 226, 227, 243, 258 Ren, J. 43, 69, 76, 77, 78, 95, 122, 136, 181, 199, 212, 215, 244

*Uekita, I. 137

Q

S Salaspuro, M. 16, 44, 47, 48, 65, 66, 80, 89, 90, 91, 92, 93, 94, 95, 106, 107, 121, 134, 155, 156, 177, 225, 228, 245, 257, 259, 261 Salaspuro, V. 92, 134, 145, 153, 154, 155, 156, 258 Seitz, H. K. 20, 40, 41, 48, 77, 78, 89, 90, 92, 93, 94, 95, 110, 119, 120, 121, 122, 123, 124, 134, 154, 155, 178, 180, 181, 196,

V *Vasiliou, V. 23 W *Wang, W. 137 *Wilson, J. 200 Worrall, S. 19, 42, 49, 50, 119, 123, 133, 134, 142, 154, 179, 180, 193, 194, 195, 197, 198, 199, 258 Y Yin, S.-J. 52, 63, 64, 65, 66, 67, 108, 121, 143, 226, 259, 260, 263

Subject Index

A α-oxoaldehydes 229–246 advanced glycation endproducts 229, 230–232, 236–239, 243–245 antitumour agents 240–241 DNA 239–240 endogenous formation 229–230 enzymatic defences 240–241 glycation free adducts 236–237, 243–246 mutagenesis 239–240 nucleotides 229, 237–241 proteins 229–237, 239–241, 243–246 proteolysis 236–237 α-tocopherol 41–42, 180 acetaldehyde dehydrogenase see aldehyde dehydrogenase acetoxymethyl-methyl-nitrosamine (AMMN) 117, 119 acetylation of histone 217, 220–223, 226–228 acetylcholine esterase (AChE) 137–138, 139–141, 143 acetylcysteine 153, 178, 205, 219–220 AChE see acetylcholine esterase achlorhydric atrophic gastritis 86–87 acinar atrophy/fibrosis 200–201, 203–210, 212 actins 73 acute alcoholic myopathy 159, 161, 167–168 acute pancreatitis 200 adducts 183–197, 198–199 alcohol induced myopathy 189–190 autoimmune responses 187–188, 190, 193–196 characteristics 184 erythrocytes 183, 185–186, 199 functional implications 184–185 glycation free 236–237, 243–246 plasma proteins 186–187

proteins 165–169, 179–180, 183–197, 198–199 Schiff bases 184 198, 230 tissue distribution 188–190, 199 ADH see alcohol dehydrogenase advanced glycation endproducts (AGEs) α-oxoaldehydes 229, 230–232, 236–239 adducts 193–195, 198 alcoholic myopathy 179 DNA 239–240 nucleotides 237–239 proteins 229, 230–232, 236–237, 243–245 AGE breakers 244 Agency for Research on Cancer (ARC) 112 AGEs see advanced glycation end products AIMD see alcohol induced muscle disease air pollution 1, 263 airway constriction 97–109 alcohol-induced asthma 97–103, 107 experiment design 98–99 genetic mechanism 100–102 in vitro mechanism 102–103 in vivo mechanism 99–100 inflammation 97, 103–104, 107–108 alcohol airway constriction 97–104, 106 breath acetaldehyde 125–131, 135–136 carcinogenicity 80–85, 87, 89–90 cardiomyopathy 69–71 cholinergic function 137–144 enzymology 4–5, 9–10, 16–17 gastrointestinal tract 110–124, 145–146, 148 human intervention studies 252–253 measurement of acetaldehyde 252–253, 256–257, 262–263 metabolism in liver 24–25 267

268 pancreatitis 200–216 removal of acetaldehyde 24–25, 32, 43–48, 49–50 salivary acetaldehyde 80–85, 89 synergy with smoking 1, 81–82, 84–85, 95 toxicity 165, 167, 185 withdrawal symptoms 262 alcohol dehydrogenase (ADH) 1, 4–11, 16–21 airway constriction 107 alcoholic myopathy 163–164, 171–172 allele frequencies 53–56 breath acetaldehyde 126 Ca2+ fluxes 72 carcinogenicity 83, 90–91, 95 cholinergic function 138 control of expression 9–10 expressed sequence tags 8, 11, 17, 19 gastrointestinal tract 110, 112–113, 119, 121–123, 151, 154–155 genetic variants 7–9, 53–56 hepatocytes 217, 222 measurement of acetaldehyde 254, 257–258, 262–263 pancreatitis 205–206, 213 pharmacodynamics 58–61 pharmacokinetics 56–58 polymorphisms 5, 6, 52–61, 65–67 post-translational modifications 7, 10 properties 5–7 removal of acetaldehyde 46 substrate and product concentrations 10 tissue distribution 5–8, 10–11, 19 transgenic models 69–74 alcohol induced muscle disease (AIMD) 159–163 see also alcoholic myopathy alcoholic liver disease (ALD) 187–188 alcoholic myopathy 158–182 acute 159, 161, 167–168 chronic 159–161, 166–167, 170, 177 enzymes responsible 163–165 exercise 177 fibre/muscle types 160, 161–163 models 161–163 protein adducts 165–169, 179–180, 189–190 protein degradation 169–170 protein synthesis 170–172, 177–178 reversibility 178 skeletal muscle 159–161, 172–173

SUBJECT INDEX alcoholic pancreatitis see pancreatitis alcoholism alcoholic myopathy 178 gastrointestinal tract 115, 119 polymorphisms 52–61, 63–67 withdrawal symptoms 262 ALD see alcoholic liver disease aldehyde dehydrogenase (ALDH) airway constriction 97–102, 106, 108–109 alcohol dehydrogenase 16–19 alcoholic myopathy 163–165, 171, 178 allele frequencies 54–55 breath acetaldehyde 125, 126–127, 131 Ca2+ fluxes 74 carcinogenicity 80–81, 85–86, 89, 91 cholinergic function 138, 143 deficiencies 26, 32 gastrointestinal tract 110, 113–117, 123, 151, 154–155 genetic variants 26, 46, 54–55, 58, 64 hepatocytes 217, 222, 226 4-hydroxy-2-nonenal 26–27, 32, 40–41, 43 inhibition 26–32, 40–41, 43, 65–66 measurement of acetaldehyde 247, 250, 254, 257–258, 263 PCR determination 98 pharmacodynamics 58–61 pharmacokinetics 56–58 polymorphisms 52–61, 64–67, 71 properties/classification 27–32 removal of acetaldehyde 23, 24–32, 40–50 tissue distribution 25–26 transgenic models 69–74, 78 aldehyde oxidases 33, 43 ALDH see aldehyde dehydrogenase allele frequencies 53–56 Alzheimer’s disease 47 AMMN see acetoxymethyl-methyl-nitrosamine angina pectoris 63 angiopoietin 2 (Ang2) 233, 235 Antabuse see disulfiram antibiotics 91, 151, 156–157 antibodies 187–188, 190, 193–196 antihistamines 65, 66 antioxidants 153, 178, 205, 219–220 antitumour agents 240–241 AP-1 204, 209–210, 212 apolipoproteins 185 apoptosis alcoholic myopathy 178 cardiomyopathy 73, 76–77

SUBJECT INDEX gastrointestinal tract 111–112, 114, 120 hepatocytes 219–220, 223–224 pancreatitis 212 asthma, alcohol-induced 97–103, 107 atmospheric exposure 1, 263 autoimmune responses 187–188, 190, 193–196 azelastine hydrochloride 98, 99–100 B β-carotene 124 beta blockers 65 betel nuts 93 blood acetaldehyde 247, 248–250, 256–260 blood ethanol 247, 251, 252 bound acetaldehyde 130–131, 133–134 brain cholinergic function 137–144 enzymology 19 protein adducts 190, 199 removal of acetaldehyde 49–50 breast cancer 3, 258 breath acetaldehyde 125–136 alcohol 125–131, 135–136 analysis 127–128 applications 127 bound acetaldehyde 130–131, 133–134 concentrations 128–130 ethanol-blended gasoline 131 experimental design 127–128 measurement 247, 250–251 modifying factors 130–131, 135 salivary acetaldehyde 133–134 smoking 125, 131, 135 C c-jun N-terminal kinases ( JNKs) cardiomyopathy 76 hepatocytes 217–221, 225–227 pancreatitis 201–207, 209 C/EBP see CCAAT-enhancer binding proteins Ca2+ flux α-oxoaldehydes 245 cardiomyopathy 71, 72, 74, 76, 78 pancreatitis 203, 215–216 cadherins 120–121 Caenorhabditis elegans 240 calvados 94, 113 Candida infections 89

269 carcinogenicity 80–96 acetaldehyde 111–112 alcohol 80–85, 87, 89–90 biochemical interrelationships 82–83, 85–86 epidemiological interrelationships 81–82, 84–85, 95 foodstuffs 94 gastric acetaldehyde 86–87, 89–90 gastrointestinal tract 110–124 microbial acetaldehyde 83–85 oral tobacco 93 salivary acetaldehyde 80–86, 89, 91–93, 95 smoking 80–82, 84–86, 91–93 cardiomyopathy 69–79 alcohol 69–71 alcohol dehydrogenase 69–74 aldehyde dehydrogenase 69–74, 78 Ca2+ fluxes 71, 72, 74, 76, 78 experiment design 77–78 FVB cardiomyocytes 69, 72–74 gene variations 71–72 protein adducts 190 smooth muscle cells 77–78 cardioprotective effects 178–179, 181 cardiovascular system pharmacodynamics 58–61 pharmacokinetics 56–58 polymorphisms 52–68 carnosinase 180 caspase 3 219–220 catalase 4, 12–13, 19 alcoholic myopathy 164 cardiomyopathy 70 cholinergic function 137, 139, 142–143 protein adducts 186 removal of acetaldehyde 25, 49–50 catecholamines 67, 181 CCAAT-enhancer binding proteins (C/EBP) 9–10 CCK see cholecystokinin ChAT see choline acetyltransferase chlorhexidine 145, 146–147, 152 cholecystokinin (CCK) 203–205 cholesterol 48 choline acetyltransferase (ChAT) 137–138, 139–141, 143 cholinergic function 137–144 catalase 137, 139, 142–143 experiment design 138 microdialysis 137, 138

270 RT-PCR and western blot 137, 138–141 chronic alcohol intake 9–10, 26–27 epigenetic histone modifications 217–219 gastrointestinal tract 114–117 chronic alcoholic myopathy 159–161, 166–167, 170, 177 chronic pancreatitis 200–201, 203–210 chronic renal failure (CRF) 236 cigarettes see smoking cimetidine 87 ciproloxacin 151, 152, 156–157 cirrhosis 214, 223, 244 COGA see Consortium of Genetics of Alcoholism collagens 185, 195 colorectal cancer 114, 117, 121–123, 146 Consortium of Genetics of Alcoholism (COGA) 19 coronary disease 63 CRF see chronic renal failure cross-linking see advanced glycation end products cyanamide alcoholic myopathy 158–159, 167–168, 170–171 brain 142–143 cardiomyopathy 72 gastrointestinal tract 117 hepatocytes 226 1,3-cyclohexanedione 260 CYP2E1 see cytochrome P450 2E1 cysteine 145–146, 147–150, 153, 155–156 cytochrome c 77 cytochrome P450 2E1 (CYP2E1) 4–5, 11–12, 17–20 α-oxoaldehydes 230 alcoholic myopathy 164, 172, 181 breath acetaldehyde 130, 131 carcinogenicity 92–93, 95 cardiomyopathy 70 gastrointestinal tract 124 4-methylpyrazole 172 pancreatitis 213 protein adducts 185, 189, 196–197 removal of acetaldehyde 25, 33–34, 40, 44, 47, 50 cytosolic xanthine oxidoreductase 14 D 3-deoxyglucosone (3DG) 230, 239 diabetes mellitus 195, 230, 244

SUBJECT INDEX dialysis 236 dietary exposure 1 digestive tract cancers 80–82, 85–86, 92, 95 pharmacological treatments 145–147, 151 disulfiram 46–47, 58, 65–66, 126 DNA glycation 239–240 E echocardiomyopathy 77–78 ELISA see enzyme-linked immunosorbent assay Emess see ethanol metabolic stress end-stage renal disease (ESRD) 236–237 endotoxins 196, 213, 215 enzyme-linked immunosorbent assay (ELISA) 166–169, 193–194, 198 enzymology alcohol dehydrogenase 1, 4–11, 16–21 catalase 4, 12–13, 19 cytochrome P450 2E1 4–5, 11–12, 17–20 cytosolic xanthine oxidoreductase 14 experiment design 19–21 nitric oxide synthases 13–14 epigenetic histone modifications acetylation 217, 220–223, 226–228 hepatocytes 217–228 mitogen activated protein kinase 217–228 phosphorylation 217, 220–223, 225, 227–228 ERKs see extracellular signal-related kinases erythrocytes 183, 185–186, 199 Escherichia coli 196, 239 ESRD see end-stage renal disease ESTs see expressed sequence tags ethanol see alcohol ethanol metabolic stress (Emess) 218, 223 ethanol patch tests 98, 101 ethanol-blended gasoline 131 ethnicity/race 7, 9 airway constriction 100–102 alcoholic myopathy 181 breath acetaldehyde 131 carcinogenicity 80–81, 85–86, 89, 91 cardiomyopathy 71–72 gastrointestinal tract 113, 114–116 polymorphisms 53–56, 58, 63–64 removal of acetaldehyde 26, 46 ethylene glycol 66 exercise 177

SUBJECT INDEX exhaled breath see breath acetaldehyde expressed sequence tags (ESTs) 8, 11, 17, 19 extracellular signal-related kinases (ERKs) 201, 202–209, 228 F fatty acid ethyl esters (FAEEs) cardiomyopathy 70 hepatocytes 218, 228 pancreatitis 201, 203, 204, 212–213, 215 fibrosis 77, 200–201, 203–210, 213 fixed coronary disease 63 folate 119, 122, 196–197 foodstuffs 1, 94 formaldehyde 48, 108 free acetaldehyde 247, 256 fructosamines 230 FVB cardiomyocytes 69, 72–74

271 nucleotides 237–239 proteins 230–236, 239–241, 243–246 see also α-oxoaldehydes; advanced glycation endproducts glycation free adducts 236–237, 243–246 glyceraldehyde-3-phosphate (G3P) dehydrogenase 32–33 glycosylation 194 glyoxal gastrointestinal tract 119–120 protein glycation 230, 239–240, 244 see also α-oxoaldehydes glyoxalase I (Glo1) 240–241 granulocyte macrophage colony stimulating factor (GM-CSF) 97, 103–104, 106 GRAS (generally recognised as safe) classification 80 growth hormone 9

G G3P see glyceraldehyde-3-phosphate gasoline 131 gastric cancer 86–87, 89–90 gastrointestinal tract (GIT) 2 acetaldehyde production 112–114 alcohol 110–124, 145–146, 148 carcinogenicity 110–124 ciproloxacin 151, 152, 156–157 colonic bacterial production 114, 117, 121–123 cysteine 145–146, 147–150, 153, 155–156 experiment design 114–116 genetic linkage studies 114–116 hepatocytes 225 lactulose 151–152, 154 large intestine 150–151 motility 123 pharmacological treatments 145–157 protein adducts 188, 190 removal of acetaldehyde 41–42 salivary acetaldehyde 113, 116 smoking 113, 145–146, 148–149, 156 see also upper digestive tract GIT see gastrointestinal tract Glo1 see glyoxalase I glutathione 154 glycaemic control 236, 244 glycation adducts 193–194, 199 DNA 239–240

H haemoglobin 185, 199, 248 headspace gas chromatography (HSGC) 248–249 heart disease see cardiomyopathy heat shock proteins 179 Helicobacter pylori 87, 90 hepatitis 214 hepatocyte nuclear factor 1 (HNF-1) 9, 11, 12 hepatocytes 217–228 heterozygotes 55–61, 64–66, 115 high-performance liquid chromatography (HPLC) 248–249, 259 histamine 97–99, 103, 106–107 histamine H1 receptor agonists 98 histones 2, 239, 243 see also epigenetic histone modifications 4-HNE see 4-hydroxy-2-nonenal HNF-1 see hepatocyte nuclear factor 1 homocysteinaemia 122 homozygotes 55–61, 64–66, 115 HPLC see high-performance liquid chromatography HSGC see headspace gas chromatography human intervention studies 252–253 hydrogen peroxide see reactive oxygen species hydroimidazolones 229–234, 236 hydroxyethyl radical–protein adducts 165, 198

272 4-hydroxy-2-nonenal (4-HNE) adducts 183, 184, 189 alcoholic myopathy 165 carcinogenicity 93 gastrointestinal tract 119–120 pancreatitis 209 removal of acetaldehyde 26–27, 32, 40–41, 43 hyperglycaemia 230, 244 hyperproliferation 41 hyperprolinaemia 32 hypochlorhydria 87 hypoglycaemia 244 I IARC see International Agency of Research on Cancer Ig see immunoglobulins IGF1 see insulin-like growth factor 1 IL see interleukins imidazolidinones 184 imidazopurinones 237–239 immunocytochemical studies 186, 188 immunoglobulins (Ig) 186, 188, 195–196 immunohistochemical studies 166–169 indole-3-aldehyde 24 inflammation airway constriction 97, 103–104, 107–108 hepatocytes 223 protein adducts 195–196 insulin resistance 215, 223 insulin-like growth factor 1 (IGF1) 171 interleukins (IL) 12, 188, 195–196 International Agency of Research on Cancer (IARC) 93–95 isopentanol 49 J JNKs see c-jun N-terminal kinases K ketosteroid reductase 185 Kuppfer cells 193, 214 L Lactobacillus bulgariensis 94 lactulose 151–152, 154 large intestine 150–151

SUBJECT INDEX LCMS/MS see liquid chromatography mass spectrometry/MS LDL see low density lipoprotein leukotrienes 107 lipid peroxidation α-oxoaldehydes 230 alcoholic myopathy 165 carcinogenicity 93 protein adducts 184 lipopolysaccharide (LPS) 177, 193–194, 196 lipoproteins 185, 186–187 liquid chromatography mass spectrometry (LCMS)/MS 193, 195, 232 liver metabolism of ethanol 24–25 protein adducts 187–190 loosely bound acetaldehyde 247, 256 low density lipoprotein (LDL) 186–187 LPS see lipopolysaccharide lungs 2, 93 Lys9 acetylation of histone 217, 220–223, 226–228 M MAA see malondialdehyde–acetaldehyde adducts Maillard reaction 230–236 malnutrition 177 malondialdehyde (MDA) adducts 183–184, 187, 189–190, 196–198 alcoholic myopathy 165, 180 carcinogenicity 93 pancreatitis 209 malondialdehyde–acetaldehyde adducts (MAA) 165, 179–180, 183–184, 187–189, 194, 197–199 MAPK see mitogen activated protein kinases mast cells 107–108 MDA see malondialdehyde MDR see multidrug resistance Mediterranean diet 180 membrane potential 77 Meniere disease 32 MEOS see microsomal ethanol-oxidizing system methanol 48, 66 methylglyoxal gastrointestinal tract 119–120 measurement of acetaldehyde 135, 257

SUBJECT INDEX protein glycation 230, 232–235, 239–241, 244–246 see also α-oxoaldehydes 4-methylpyrazole 17, 65–66 alcoholic myopathy 171–172 carcinogenicity 91, 95 hepatocytes 221, 226 measurement of acetaldehyde 250 pancreatitis 212–213 2-methyl-thiazolidine-4-carboxylic acid (MTCA) 146, 147–148, 154 microbial acetaldehyde 83–85 microdialysis 137, 138 microsomal ethanol-oxidizing system (MEOS) 70, 163 mitochondrial ALDH see aldehyde dehydrogenase mitogen activated protein kinases (MAPK) 76, 200–216 c-jun N-terminal kinases 76, 201–207, 209, 217–221, 225–227 epigenetic histone modifications 217–228 extracellular signal-related kinases 201, 202–209, 228 hepatocytes 217–228 p38 kinase 201, 202–209, 217–221, 225– 226, 228 p42/44 MAP kinase 217–221 motility 123 mouthwashes 95, 130, 145, 146–147, 152 mSin3A 233, 235, 244 MTCA see 2-methyl-thiazolidine-4-carboxylic acid multidrug resistance (MDR) 240–241 mutagenesis 239–240 myopathy see alcoholic myopathy

273 nitric oxide synthases (NOS) 13–14 nitrosamines carcinogenicity 92 gastrointestinal tract 117, 119, 121 nitroso-thiazolidine-4-carboxylic acid 147–148, 154 NOS see nitric oxide synthases nuclear factor κB (NF-κB) airway constriction 97, 99, 103–104, 106 carcinogenicity 111–112, 120 pancreatitis 202, 204–205, 209–210, 212 nucleotides α-oxoaldehydes 229, 237–241 excision repair 239–240 glycation 237–239 O O6-methylguanosyl transferase 111 oestradiol 254, 257–258 opium 92 oral acetaldehyde see salivary acetaldehyde oral contraceptives 254, 257–258 oral ethanol provocation tests 98, 101 oral hygiene breath acetaldehyde 130 carcinogenicity 85, 92, 95 see also mouthwashes oral tobacco 93 oxidases 33, 41, 42–43, 261 oxidative stress 2, 261 alcoholic myopathy 165 cardiomyopathy 70, 73–74 gastrointestinal tract 119–120 hepatocytes 227–228 protein adducts 183, 184, 196–197 removal of acetaldehyde 26, 33, 43–44 oxoaldehydes see α-oxoaldehydes

N N-acetylcysteine 153, 178, 205, 219–220 N-nitroso compounds 86 N 2-ethyl desoxyguanosine 111–112 N 2-propano-DG 112 Na+ flux 227 NAD/NADH see nicotinamide adenine dinucleotide necroinflammation 200 NF-κB see nuclear factor κB nicotinamide adenine dinucleotide (NAD/NADH) 23, 27, 33, 41 nitrefazole 58

P p38 76 p38 kinase hepatocytes 217–221, 225–226, 228 pancreatitis 201, 202–209 p42/44 MAP kinase 217–221, 225 p53 120 pancreatic stellate cells (PSCs) 203, 205–210, 212–213 pancreatitis 200–216 acinar atrophy/fibrosis 200–201, 203–210, 212–213

274 acute 200 chronic 200–201, 203–210 measurement of acetaldehyde 259 mitogen activated protein kinases 200–216 phosphatidylinositol-3-kinase 202, 207–209, 211–212 protein kinase C 202, 205, 207–209 stellate cells 203, 205–210, 212–213 penicillamine 154 perchloric acid/saline 249 peroxisome proliferator-activated receptor γ (PPARγ ) 200, 202 phosphatidylethanol (PEth) 218, 228, 262 phosphatidylinositol-3-kinase (PI3K) hepatocytes 225 pancreatitis 202, 207–209, 211–212 phosphorylation of histone 217, 220–223, 225, 227–228 PI3K see phosphatidylinositol-3-kinase PKA see protein kinase A PKC see protein kinase C plasma proteins 186–187 plastic bottles 107 polyethylene glycol/sodium azide 249 polymorphisms 52–68 alcohol dehydrogenase 5, 6, 52–61, 65–67 alcoholism 52–61, 63–67 aldehyde dehydrogenase 52–61, 64–67, 71 catalase 13 cytochrome P450 2E1 11, 20 experiment design 53 genetic variants 53–56, 58, 63–64 pharmacodynamics 58–61 pharmacokinetics 56–58 PPARγ see peroxisome proliferator-activated receptor γ prebiotics 151 pregnancy 183, 185–186 proteases 169–170, 213 protein kinase A (PKA) 12 protein kinase C (PKC) alcoholic myopathy 179 hepatocytes 219–220 pancreatitis 202, 205, 207–209 proteins 2 α-oxoaldehydes 229–237, 239–241, 243–246 adduct formation 165–169, 179–180, 183–197, 198–199 alcoholic myopathy 165–172 degradation 169–170

SUBJECT INDEX glycation 230–236, 239–241, 243–246 proteolysis 236–237 removal of acetaldehyde 34 synthesis 170–172, 177–178 proteolysis 236–237 PSCs see pancreatic stellate cells R race see ethnicity/race reactive oxygen species (ROS) 261 alcoholic myopathy 178 enzymology 11, 12–13 hepatocytes 218, 227 pancreatitis 205, 209 removal of acetaldehyde 41, 44 receptor for advanced glycation end products (RAGE) 193, 245 removal of acetaldehyde 23–51 aldehyde dehydrogenase 23, 24–32, 40–50 catalase 25, 49–50 cytochrome P450 25, 33–34, 40, 44, 47, 50 glyceraldehyde-3-phosphate dehydrogenase 32–33 4-hydroxy-2-nonenal 26–27, 32, 40–41, 43 oxidases 33, 41, 42–43 oxidative pathways 24–34 protein binding 34 reduction to ethanol 24 reserpine 71 resveratrol 179 retinol 124 reverse transcriptase polymerase chain reaction (RT-PCR) 137, 138–141 ROS see reactive oxygen species RT-PCR see reverse transcriptase polymerase chain reaction ryanodine receptors 76 S salivary acetaldehyde 2 alcohol 80–85, 89 breath acetaldehyde 133–134 carcinogenicity 80–86, 89, 91–93, 95 chlorhexidine 146–147, 152 cysteine 145–146, 147–150, 153, 155–156 gastrointestinal tract 113, 116 measurement 247, 251–252, 259–260 microbial acetaldehyde 83–85 pharmacological treatments 145–157 smoking 80–81, 84–86, 91–93

SUBJECT INDEX Schiff base adducts 184, 198, 230, 257 semicarbazide 249, 259–260 Ser10/Ser28 phosphorylation of histone 217, 220–223, 225, 227–228 SHAS see Subjective High Assessment Scale single nucleotide polymorphisms (SNPs) 13, 20, 55 skeletal muscle 159–161, 172–173, 189–190 see also alcoholic myopathy smoking airway constriction 104, 108 breath acetaldehyde 125, 131, 135 carcinogenicity 80–82, 84–86, 91–93 gastrointestinal tract 113, 145–146, 148–149, 156 measurement of acetaldehyde 256, 263 removal of acetaldehyde 48 salivary acetaldehyde 80–81, 84–86, 91–93 synergy with alcohol 1, 81–82, 84–85, 95 smooth muscle tissue 77–78, 97–109 SNPs see single nucleotide polymorphisms Sp3 233, 235 stomach cancer 81–82, 86–87 Streptococcus salivarius 83 Streptococcus thermophilus 94 Subjective High Assessment Scale (SHAS) 53 superoxide dismutases (SOD) 218 T tachycardia 63 thiamine 119 thiazolium compounds 244–245 tissue distribution 2–3 alcohol dehydrogenase 5–8, 10–11, 19 aldehyde dehydrogenase 25–26 catalase 13 cytochrome P450 2E1 12 enzymology 9–12 protein adducts 188–190 TNFα see tumour necrosis factor α tobacco see smoking

275 transcription factors 2 α-oxoaldehydes 233, 235, 243–244 enzymology 9–10, 12, 13 measurement of acetaldehyde 258 removal of acetaldehyde 45 triglycerides 214 tubulin 180, 185, 213 tumour necrosis factor α (TNFα) adducts 196 airway constriction 106 alcoholic myopathy 177 enzymology 18 pancreatitis 203, 212 tyrosine kinase 219–220 U ulcers 108 upper digestive tract cancers 80–82, 85–86, 92, 95 pharmacological treatments 145–147, 151 V very low density lipoprotein (VLDL) 261 vinyl acetate 93 vitamin B6 122 vitamin B12 122 vitamin E 41–42 VLDL see very low density lipoprotein volatile organic chemicals (VOCs) 125–126 W western blotting 137, 138–141, 164–165 withdrawal symptoms 64, 261–262 wortmannin 225 X xanthine oxidases 33, 41, 42–43 Y yoghurt 94