Copper (1997) - Department of Health and Ageing

N A T I O N A L E N V I R O N M E N T A L H E A L T H F O R U M

Copper

1 9 9 7

Metal Series No. 3

National Environmental Health Monographs

Copper

Report of an International Meeting20-21 June 1996

Brisbane

Edited by

Michael R Moore, Paula Imray, Charles Dameron, Phil Callan,

Andrew Langley and Sam Mangas

National Environmental Health Forum Monographs

Metal Series No. 3

N a t i o n a l E n v i r o n m e n t a l H e a l t h F o r u m

1997

C o p p e r

2

Copyright 1997 Department of Human ServicesPrinted by Openbook Publishers.Published by the National Environmental Health Forum.Prepared for publication by Sam MangasPublic and Environmental Health ServiceDepartment of Human Services

Price available on application.

Minor amendments have been made to pages 8 and 9.The list of 'Published Monographs' has been updatedand a disclaimer has been included.

National Library of Australia Cataloguing-in-Publication

Copper: report of an international meeting20-21 June 1996, Brisbane.

ISBN 0 642 28009 6.

ISSN 1327-4775

1. Copper - Physiological effect - Congresses.2. Copper - Environmental aspects - Congresses.3. Copper - Toxicology - Congresses.I. Moore, M. R. (Michael R.). II. NationalEnvironmental Health Forum. (Series: NationalEnvironmental Health Forum monographs. Metalseries ; no. 3).

572.518

C o p p e r

3

Contents

Preface 7Acknowledgements 7

Disclaimer 7

Published monographs 8

Copper - A Primer 91. History and background 9

2. Exposure 9

3. Homeostasis 10

4. Copper deficiency 10

5. Excess of copper 11

6. Copper and zinc - co-homeostasis 11

Health significance of copper 12

Genetic disorders of copper metabolism and the dual nature of copper in biology 181. Introduction 18

2. The essential nature of copper 19

2.1 Nutritional copper deficiency 19

2.2. Menkes disease and its variants 20

2.3 The mottled mice 22

3. The toxic nature of copper 23

3.1 Excess copper disposed of by biliary excretion 23

3.2 Wilson disease 23

3.3 Animal models of WD 24

4. Molecular and cellular basis of copper transport 24

4.1 Structural and function of MNK and WND 24

4.2 Copper resistance is acquired by amplification of MNK 26

4.3 Do metallothioneins protect against copper toxicity? 27

5. Conclusions 27

Copper in the Aquatic Environment 321. Analysis of copper 32

2. Chemical speciation 33

3. Copper bioavailability and toxicity 35

4. Water quality criteria and standards 37

5. Aquatic sediments 38

6. Conclusions 39

C o p p e r

4

Evaluation of copper guideline values for drinking-water 431. Introduction 43

2. Historical perspective 43

2.1 Aesthetics-based guideline values 44

2.2 Health-based guideline values 44

3. Derivation of health-based guideline values 44

3.1 US Environmental Protection Agency 44

3.2 World Health Organisation 45

3.3 Australia and New Zealand 46

3.4 European Commission 46

4. Overall evaluation 46

Population exposure to copper in drinking water 48Abstract 48

1. Introduction 49

2. Factors that affect the population exposure to copper in drinking water 49

2.1 Water composition 49

2.2 Stagnant contact time 49

2.3 Age of copper piping 49

2.4 The design and installation procedures 50

2.5 The use of copper in the distribution network 50

2.6 Drinking habits of the population 50

3 Analysis of water utility monitoring data of copper in drinking water in the USA 50

4. Conclusions 53

Risk assessment for essential trace elements: A proposed methodology 541. Introduction 54

2. Risk assessment and IPCS 55

2.1 IPCS goals in risk assessment methodology 55

2.2. Concepts and definitions 55

2.3 Development of tolerable intakes for non-essential chemicals 58

3. Development of recommended safe and adequate daily dietary intakes 58

3.1 Comparison of methodologies for derivation of RDAs and TIs 58

4. Principles and methods for assessing human health risks from exposure to an ETE 59

4.1 Scientific principles 59

4.2 The AROI concept in human health risk assessment - A proposed methodology 62

5. The Deficiency - Toxicity (DT) Model in Environmental Risk Assessment 65

6. Concluding remarks 65

C o p p e r

5

Assessment of the requirement of copper in the nutritional support of the severely ill patient 671. Copper deficiency and treatment 68

2. Biochemical assessment 68

2.1 Plasma copper 68

2.2 Monitoring of neonates 69

2.3 Acute phase plasma protein effect 69

2.4 Action limits 69

2.5 Contra-indication for copper supplementation 69

2.6 Copper isotope tracer studies 70

3. Summary and conclusions 70

Copper tailing impacts in coastal ecosystems of Northern Chile: From species to community responses 71

1. Introduction 71

2. Study sites and methods 72

3. Results and discussion 73

3.1 Copper concentration in sea water and Enteromorpha compressa 73

3.2 Rocky intertidal species richness 74

3.3 Enteromorpha compressa as a biological model: current status of knowledge and

future perspectives 77

Wilson’s Disease after cloning of the gene 811. Genetics 81

2. DNA-based diagnosis 82

3. Pathogenesis 82

4. Clinical features 83

5. Hepatic manifestations 83

6. Hepatic pathology 84

7. Laboratory diagnosis of Wilson's Disease 85

7.1 Serum ceruloplasmin 85

7.2 Urinary copper excretion 86

7.3 Hepatic copper concentration 86

7.4 Incorporation of orally administered radiocopper into ceruloplasmin 86

7.5 Abnormal imaging 87

8. Diagnostic screening 87

9. Treatment 87

9.1 Diet 87

9.2 Pharmacologic therapy 87

9.3 Long-term management 89

9.4 Liver transplantation 89

C o p p e r

6

Indian Childhood Cirrhosis (ICC) - Revisited 911. Background 91

2. Copper and ICC 91

3. Clinical features 92

4. Diagnosis 93

Workshop 1 - Copper and Health 941. Summary 94

2. Research recommendations 95

Workshop 2 - Distribution and Metabolism 961. Summary 96

Intracellular protection against copper toxicity in mammals 971. Introduction 97

2. General mechanisms for metal detoxification 97

3. Redundancy 98

4. Metalloregulation 98

5. The copper detoxification pathway 99

6. Copper detoxification mechanisms 100

6.1 Metallothioneins 100

6.2 Transcriptional regulation of metallothioneins 100

6.3 Cu-ATPases 101

6.4 Regulation of the human copper ATPases 102

7. Conclusion 103

Workshop 3 - Copper and the Environment 1051. Summary 105

2. Aquatic systems 105

3. Toxicity testing 105

4. Bioavailability 106

5. Terrestrial testing 106

6. Essentiality versus toxicity 106

7. Bioaccumulation 107

8. Conclusions and recommendations 107

Participants in the international workshop on copper 108

C o p p e r

7

Preface

The National Environmental Health Forum has been established by the Directors of EnvironmentalHealth from each State and Territory and the Commonwealth with a secretariat provided by theCommonwealth Department of Health and Family Services.

The National Environmental Health Forum is publishing a range of monographs to give expert adviceand guidance on a variety of important and topical environmental health matters. This publication isthe third in the metals series. A list of published monographs appears on page viii.

Acknowledgements

This publication has been made possible by technical assistance and funding provided by each Stateand Territory Environmental Health Branch and the Commonwealth Department of Health andFamily Services. Comment has been provided by reviewers from the Commonwealth, States andTerritories. Graphic design and layout assistance have been provided by Sandra Sowerby,Environmental Health Branch, South Australian Health Commission. The South Australian HealthCommission library has assisted research and cataloguing for the document.

Financial assistance from the International Copper Association is gratefully acknowledged.

Disclaimer

The papers in these proceedings do not necessarily represent the views of the National EnvironmentalHealth Forum, or the Health Departments represented on the Forum, or of the authors.

This document has been prepared in good faith exercising due care and attention. However, norepresentation or warranty, expressed or implied, is made as to the relevance, accuracy, completenessor fitness for purpose of this document in respect of any particular user’s circumstances. Users of thisdocument should satisfy themselves concerning its application to, and where necessary seek expertadvice about, their situation. The NEHF, its participants and the DHS shall not be liable to thepurchaser or any other person or entity with respect to any liability, loss or damage caused or allegedto have been caused directly or indirectly by this publication.

C o p p e r

8

Published monographs

Water Series

1. Guidance for the control of Legionella (1996)

2. Guidance on water quality for heated spas (1996)

3. Guidance on the use of rainwater tanks (1998)

Soil Series

1. Health-based soil investigation levels, 2nd edition (1998)

2. Exposure scenarios and exposure settings, 2nd edition (1998)

3. Composite sampling (1996)

Metal Series

1. Aluminium, 2nd edition (1998)

2. Zinc (1997)

3. Copper (1997)

Air Series

1. Ozone (1997)

2. Benzene (1997)

3. Nitrogen Dioxide (1997)

4. Sulfur dioxide (1999)

General Series

1. Pesticide use in schools and school grounds (1997)

2. Paint film components (1998)

3. Guidelines for the control of public health pests –Lice, fleas, scabies, bird mites, bedbugs and ticks (1999)

4. National Standard for licensing pest management technicians (1999)

Indigenous Environmental Health Series

1. Indigenous Environmental Health No. 1 (1999)

Exposure Series

1. Child activity patterns for environmental exposure assessment in the home (1999)

Counter Disaster series

1. Floods: An environmental health practitioner's emergency management guide (1999)

C o p p e r

9

Copper - A PrimerProfessor Michael R. MooreNHMRC National Research Centre for Environmental ToxicologyThe University of Queensland - Griffith University

1. History and background

Mankind's use of copper stretches into the mists of antiquity and it probably fair to say that theutilisation of this metal in the bronze age, around 3,000 B.C. heralded the relentless process ofdevelopment of civilisation as we know it. We continue to use this metal in very great quantities andbecause of its proximity to man there is the attendant possibility of toxic consequences associatedwith exposure to it. However, this is not necessarily the appropriate response in viewing theconsequences of exposure to this element. Copper is an essential trace element and a co-factor inmany enzymic reactions. It is for example, central to the operation of cytochrome oxidase, and haemsynthesis (Tephly et al., 1978). In these circumstances the appropriate evaluation of an element of thissort is to consider the balance between its essentiality and the potential for it to become toxic, that isthe mechanisms of homeostasis.

2. Exposure

Copper exposure in the environment is inevitable. It is estimated that in excess of 75,000 tonnes ofcopper is released into the atmosphere annually of which a quarter is thought to come from naturalsources, whilst the rest is of anthropogenic origin. Many soils contain copper and the copper contentscan be supplemented through human activities, industrial and agricultural processes. In terms ofbioavailability, it is likely that dissolved copper in water supplies present the most likely source ofhuman exposure to the metal. In general however, copper content of potable water supplies tends tobe low around 0.8 mg/L in Australia. Such low values can be greatly increased as the water suppliesare soft and acidic. Copper is in general a micro-nutrient, essential for normal growth of lowerorganisms and higher mammals. Various authorities have suggested dietary allowances of between 1and 3 mg/day in adults. The WHO has previously suggested that copper intake should not exceed 0.5mg/kg/body weight. In view of the very large differences in these values, it is hard to escape theconclusion that in certain cultural circumstances, copper intake may well be sub-optimal. (Refer totable 1).

Table 1: Mean dietary intake of copper in adults

Country Mean intake mg/day % of subjects lessthan 2mg/day

Australia: male, MB female, MB

Denmark, DD

Germany, DD

Netherlands, DD

New Zealand

Norway, DD

Sweden, MB

United Kingdom: male female,TD

USA: male, MB female, MB

1.92.2

1.2

0.9

1.5

-

1.0

1.2

1.61.2

1.20.9

79%

84%

Key: MB - Market basket, DD - Duplicate Diet, TD - Total Diet

C o p p e r

10

3. Homeostasis

Copper is an essential element that can be toxic when exposures exceed physiological needs. Thisrelationship is described by a U-shaped curve expressing risk of deficiency or excess with the centralportion of the curve between the arms expressing the range of exposure that is related to optimalfunction (health). The relationship between intake and health is affected by physiological factors formaintenance of homeostasis and extrinsic factors that affect availability of copper. (Refer to figure 1).

Figure 1: Dose -Response Relationships -Typical responses of an individual exposed to CopperCopper Homeostasis. Essential elements show a U shaped curve lying between deficiency -ultimately leading to death-and excess which is toxic and which may also lead to death. Therelationship differs from that of toxic elements which are non-essential and show a characteristicsigmoidal curve starting in a region of no-effect and rising to the region of toxicity.

The homeostatic model defines the principle of an acceptable range of exposures (acceptable range oforal intake, AROI) for an essential trace element like copper. In the acceptable range, it provides thesubstrates for expression of the genetic potential of the individual. Environmental levels of copper donot produce adverse effects among the general population or the environment. However, there areindividuals or groups with imbalances in relation to other trace elements, or with disorders inhomeostatic mechanisms that experience effects, of either deficiency or toxicity, from exposureswithin the acceptable range. These disorders may be acquired or of genetic origin.

4. Copper deficiency

Clinical copper deficiency in adults is rarely found in the general population. However recent dietarysurveys show that the mean population intake is suboptimal. In some regions of the world such asEurope and USA intakes are about 20% below the recommended levels. The health consequences ofbarely adequate intakes remain to be determined.

C o p p e r

11

Low birth weight infants are particularly at risk for deficiency. Frequent episodes of diarrhoea areanother risk factor leading to copper deficiency. Copper deficiency commonly occurs during therecovery from protein energy malnutrition, since these infants grow rapidly and are usually fed dietsthat supply inadequate copper. Individuals taking supplements of zinc and ascorbic acid are at risk ofdeveloping copper deficiency and there are malabsorption states associated with copper deficiency.Patients receiving prolonged intravenous nutrient mixtures which lack sufficient copper may developsymptomatic evidence of copper deficiency.

Menkes disease is a rare (~1:200,000) X-linked recessive disorder which results in a defect in theintestinal absorption of copper. This disorder leads to a severe, symptomatic, fatal deficiency stateeven at copper intakes above the AROI.

Copper deficiency has been implicated as a possible risk factor in the pathogenesis of cardiovasculardisease. When copper homeostatic control is defective and/or copper intake is excessive, coppertoxicity may occur - this is an infrequent occurrence.

5. Excess of copper

There is little doubt that acute toxicity associated with copper exposure is largely associated withinstance of attempted suicide or accidental oral exposure to the metal. The features most commonlyreported are metallic taste in the mouth, gastric pain, headache, nausea, dizziness and diarrhoea withmassive gastrointestinal bleeding. In addition to these gastrointestinal effects, tachycardia respiratorydifficulty, haematuria and liver and kidney failure have been reported. In general however, the morelikely situation are the problems associated with copper deficiency which has been linked to anaemianeutropenia and bone abnormality. Where there is upset of copper homeostasis such as the changesassociated with genetic conditions, there can be both resultant copper deficit and copper excess.

The most common reason for copper accumulation is chronic liver disease and choleostasis but inthese circumstances the accumulation of copper is an epiphenomenon associated with the inability ofthe liver to clear the metal and consequent accumulation in the tissues.

6. Copper and zinc - co-homeostasis

Another reason for alteration in copper homeostasis is related to the co-absorption of the metal withother essential metals. The most common of these are the inter-relationships between increased zincintake and copper deficiency. Disproportionate intake of zinc in relationship to copper can inducecopper deficiency in humans. In these circumstances, there are increased physiological copperrequirements. The early work in this area carried out by Reiser et al. (1987) involved copperdeficiency produced experimentally in a young man with a diet containing 0.8 mg of copper per day.An interesting change observed in the subject was that of increases in plasma cholesterol anddevelopment of ventricular tachycardia which was reversed when the subject was given 4 mg ofcopper daily. In subsequent experiments by Reiser et al, there was evidence of cardiac dysfunction in2 out of 24 men given an increased zinc diet which induced copper deficiency. There were somechanges in standard markers of copper status such as plasma-copper ceruloplasmine and erythrocytesuperoxide dismutase (Uauy et al., 1996).

All of these measures normalised after copper supplementation (Uauy & Olivares, 1996). Copperdeficiency induced by excessive quantities of zinc in the diet resembles deficiencies induced by dietlow on copper. Because of this inhibitory effect of zinc or copper utilisation and in that respect it isinteresting to note that there is a relationship between the mortality rate for coronary heart disease andthe ratio of zinc to copper in cows milk (Klevay, 1975). It is obvious from this that copperhomeostasis requires much clearer definition in respect both of its individual role in metabolism andthe concurrent role that might be taken in its interaction with other ions. In particular the ratiobetween zinc and copper requires further investigation.

Lifestyle issues in which the relationship between the internal dose of these metals in organs such asliver, should be related to the dietary ratios of these same metals to establish whether there areendpoints in the process which have clinical significance.

C o p p e r

12

Copper will continue to be used by man in the foreseeable future. It's physical properties and inparticular its cost-effective ability to conduct electricity guarantees a continuing need for its use in oursociety. Much of the research at the current time would seem to suggest that our principal concernsabout the metal should be directed towards the consequences of copper deficiency rather than to theeffects of copper excess. The copper-zinc ratio is known to induce dyslipidaemia and thus studies onthe cardiac consequences of this and indeed upon cardiac conduction need to be further investigated.

This book presents the results of discussions of a workshop held in the National Research Centre forEnvironmental Toxicology prior to the IPCS Task Group Meeting on Copper. The papers presentedare a synthesis of the current thinking on copper essentiality, toxicity and on the disease statesassociated with these.

References

Klevay, L. (1975), ‘Coronary heart disease: The zinc/copper hypothesis’, Am J Clin Nutr.; vol. 28, pp.764-774.

Reiser, S., Powell, A., Yang, C.Y. and Canary, J.J. (1987), ‘Effect of copper intake on bloodcholesterol and its lipoprotein deistribution in Men’, Nutr Rep In., vol. 36, pp. 641-9.

Tephly, T.R., Wagner, G., Sedman, R. and Piper, W. (1978), ‘Effects of metals on heme biosynthesisand metabolism’, Fed Proc., vol. 37, pp. 35-9.

Uauy, R. and Olivares M. (1996), ‘Copper nutrition in humans: essentiality and toxicity’, Amer J ClinNutr., vol. 63, no.5.

Health significance of copperDr Ricardo Uauy and Dr Manuel OlivaresInstitute of Nutrition and Food TechnologyUniversity of Chile

Criteria for essentiality of nutrients in humans include the relative composition of the element in theorganism, the effects of removing the nutrient from diet on health, and finally the effect of restoringthe nutrient in question. Following this approach we find that whole body copper content in the adultranges from about 80 mg with a range of 50 to 120 mg. The organs with the highest concentration areliver and brain with 5.1 mg and 6.3 mg per gram of wet weight respectively (Cartwright andWintrobe, 1964). The high liver content is interpreted to serve as a reserve, and the high content ofcopper in the basal ganglia of the brain is considered to be associated to its role in neurotransmittermetabolism. The other condition is deficiency disease associated with removing copper from the diet;this condition occurs in humans given copper-free parenteral nutrition. For example, parenterally fedsubjects exhibit within a few weeks, anaemia and neutropenia unless copper is provided. Theseabnormalities are reversed by a treatment with copper.

Copper is involved in the function of multiple important enzymes, most participate in redox reactions(Linder and Hazegh-Azam, 1996). Cytochrome-c-oxidase is a key enzyme for the electron transportchain in the mitochondria, superoxide dismutase has a role in scavenging of free radicals, lysyloxidase is crucial for the cross linking of collagen and ceruloplasmin acts as a ferroxidasetransforming iron from ferrous to ferric state, a step crucial for the transport of this mineral.Dopamine β-monooxygenase is necessary for dopamine and catecholamine metabolism, andtyrosinase for the production of melanin, the skin and hair pigment.

Copper is actively absorbed, primarily in the stomach and duodenum. Absorption ranges from 25-60%of intake depending on other dietary components and on the copper intake (Lönnerdal, 1996;Turnlund et al., 1989; Ehrenkranz et al., 1989). The percent of copper absorbed decreases as copperintake is increased (see Table 1), (Turnlund et al., 1989; Ehrenkranz et al., 1989). Zinc, iron, ascorbicacid, calcium and phosphorus, fibre, phytates, sucrose, fructose, molybdenum and cadmium have been

C o p p e r

13

demonstrated to inhibit copper absorption while animal protein, specifically the amino acid histidine,enhances it (Lönnerdal, 1996). Copper is transferred from the intestinal mucosa to portal bloodcomplexed principally with albumin (Linder and Hazegh-Azam, 1996). Within the hepatocyte, it isbound to metallothionein, superoxide dismutase and other binding proteins (Luza and Speisky, 1996).Copper is secreted from the liver into the blood predominantly bound to ceruloplasmin or complexedwith low molecular weight compounds. Cellular uptake in some tissues may be mediated by specificceruloplasmin receptors on their cell surface (Linder and Hazegh-Azam, 1996). Biliary excretion isthe main form of elimination in humans (Linder and Hazegh-Azam, 1996).

The main dietary sources of copper are shellfish, fish, liver, meats, nuts and chocolate. A lowerconcentration is found in legumes, grains, human milk, and especially cow milk (Pennington et al.,1995).

Table 1: Effect of dietary copper intake on percentage of copper absorption

Subjects Copper intake (mg/d) 65Cu absorption (%) Net Cu absorption(%)

Adults 0.785 55.6 ± 0.9 not available

1.68 36.3 ± 1.3 not available

7.53 12.4 ± 0.9 not available

LBW infants

Human milk fed 86.6 ± 28.2 67.2 ± 14.6 59.7 ± 13.6

Human milk fed 151.7 ± 34.4 57.4 ± 13.1 38.7 ± 10.2

Formula fed 188.9 ± 39.4 39.3 ± 21.9 15.4 ± 20.0

Formula fed 194.2 ± 27.5 26.5 ± 6.9 20.6 ± 24.1

LBW: low birth weight

Copper requirements have been established using different approaches. These include the clinicalsyndrome of copper deficiency requirement which in this case, is defined as the amount of coppernecessary to cure the deficiency. It is also established by evaluating the effect of experimental dietswith different copper intakes to determine the lowest intake required to prevent the development ofbiochemical or functional alterations. Another approach is based on epidemiological studies,measuring copper intake of healthy populations (Olivares and Uauy, 1996a). The large variability ofcopper intake in healthy populations makes it difficult to employ this methodology to define preciseintakes. However this approach provides a reasonable range of values from which to select levels tobe tested under controlled experimental studies.

The US National Academy of Sciences has recommended that all adults should receive a daily intakeof 1.5-3 mg of copper to satisfy physiological requirements (safe and adequate intake range). Thecorresponding values for infants, children and adolescents are shown in Table 2 (NRC, 1989). TheWHO recommendation for infants is 80 µg/kg per day (WHO, 1985). More recently theFAO/IAEA/WHO expert consultation has provided recommendations using new data from long termbalance studies validated by biochemical markers associated with copper status (FAO/IAEA/WHO,1996). These values are also included in Table 2. The FAO/IAEA/WHO recommendations for thefirst time suggest an upper value for acceptable dietary intake based on the very limited data available.The upper value of 0.2 mg per kg of body weight per day should be considered as a no observed effectlevel rather than a true NOAEL (No-Observed-Adverse-Effect-Level) (FAO/IAEA/WHO, 1996).

C o p p e r

14

Table 2: Estimated safe and adequate daily dietary copper intake (NRC, 1989) and safe minimum mean copper intake (FAO/IAEA/WHO, 1996)

Copper (mg/day)

Age range (y) or state Sex NRC FAO/IAEA/WHO

0-0.25 M & F 0.4-0.6 0.33-0.55

0.25-0.5 M & F 0.4-0.6 0.37-0.62

0.5-1 M & F 0.6-0.7 0.60

1-3 M & F 0.7-1.0 0.56

3-6 M & F 1.0-1.5 0.57

6-10 M & F 1.0-2.0 0.75

10-12 F 1.5-2.5 0.77

12-15 F 1.5-2.5 1.00

15-18 F 1.5-2.5 1.15

18-60+ F 1.5-3.0 1.15

10-12 M 1.5-2.5 0.73

12-15 M 1.5-2.5 1.00

15-18 M 1.5-2.5 1.33

18-60+ M 1.5-3.0 1.35

Pregnancy - 1.15

Lactation - 1.25

Various factors predispose to acquired copper deficiency. Deficit can be the consequence ofdecreased copper stores at birth, the best example of this is the low birth weight infant (Shaw, 1992).Most of copper accumulation in the fetus occurs during the third trimester of pregnancy, so a pretermnewborn will have a lower hepatic copper content compared to a full term infant. In addition, there isa greater need for rapid growth, thus they easily become copper deficient unless extra copper isprovided. Another cause of copper deficiency is an inadequate copper supply, this is a very commonsituation where cow milk and high carbohydrates feeds are given (Cordano et al., 1964). Thiscombination is very frequently found during the recovery from infant malnutrition. Copper deficiencyis also observed in cases of total parenteral nutrition with inadequate copper supplementation, or inmalabsorption syndromes. Increased requirements are observed in premature or malnourished infants;increased faecal losses are principally observed in cases of chronic or prolonged episodes of diarrhoea(Olivares and Uauy, 1996b).

The most frequent clinical manifestations of acquired copper deficiency are anaemia, neutropenia andbone abnormalities that include osteoporosis and fractures (Olivares and Uauy, 1996b). Less frequentsigns are hypopigmentation of the hair, hypotonia, impaired growth, increased incidence of infections,alterations of phagocytic capacity of neutrophils, and immune cellular abnormalities. Menkes diseaseis a rare (1:200,000) X-linked recessive genetic disorder in which deficiency is secondary to a defectin cellular copper uptake and transport. In this condition there is a defect in a membrane proteinnecessary to transport copper out of the basolateral pole of the intestinal cell (Harris and Gitlin,1996). This disease is characterised by progressive mental deterioration, hypothermia, skin and hairdepigmentation, growth retardation, bone and connective tissue abnormalities (Harris and Gitlin,1996).

Toxicity to copper may occur from eating food, drinking water or breathing air with excessive coppercontent. A small amount of copper may enter the body by skin contact with copper-containing

C o p p e r

15

substances. The oral route is the main pathway of exposure to the element. Food and water are thepredominant sources of copper intake. Food may account for over 90% of copper intake in adults ifwater has low copper content (<0.1 mg/L). If water copper content is higher (1-2 mg/L) it mayaccount for up to 50% of total intake. In infants consuming copper supplemented artificial formula,the contribution of water may be less than 10% whereas, if the formula is not fortified with copper,water may contribute over 50% of total copper intake, especially when water copper content is 1-2mg/L.

Acute copper toxicity is infrequent in man, and usually is a consequence of ingesting contaminatedfoodstuffs or beverages (including drinking water), and from accidental or voluntary ingestion of highquantities of copper salts. Acute symptoms include salivation, epigastric pain, nausea, vomiting anddiarrhoea (USEPA, 1985; USEPA, 1987; NIPHEP, 1989). Intravascular haemolytic anaemia, acuteliver failure, acute renal failure with tubular damage, shock, coma and death have been observed insevere copper poisoning. There are some reports in humans, suggesting that the consumption ofbeverages or drinking water contaminated with copper results in nausea, vomiting, and diarrhoea(Wyllie, 1957; Spitalny et al., 1984; Knobeloch et al., 1994).

The long term toxicity of copper has been less studied. Chronic toxicity in humans is observedprincipally in patients with Wilson’s disease and from the occurrence of infantile cirrhosis in areas ofIndia (ICC, Indian childhood cirrhosis), and isolated clusters of cases in Germany, Austria and othercountries (ICT, idiopathic copper toxicosis) that have been also related to excessive copper intake(Pandit and Bhave 1996, Weiss et al., 1989, Adamson et al., 1992, Horslen et al., 1994). Wilson’sdisease is an autosomal recessive genetic disorder (1:30,000) characterised by a defect in copperbiliary excretion. In this condition the accumulation of copper in liver and brain is associated withaltered structure and function of these organs (Harris and Gitlin, 1996; Scheinberg and Sternlieb,1996). Copper-associated infantile cirrhosis is an extremely rare condition; for ICT the estimatedincidence based on prospective data from Germany is 1:500,000 to 1:1,000,000. The data for ICCfrom India reveal a dramatic decline of this condition. Recent observations from Pune district basedon hospital admissions reveal a drop from 47 cases per year in 1980-1983 to 2 cases in 1992-1993(Pandit and Bhave, 1996).

ICC and ICT have been linked to the use of copper containers for storage or heating of infant formulaor to the high copper content of well water conducted through corroded copper pipes. In the casesreported in India, copper intake was over 900 µg/kg/day, which is over ten times the daily requirementrecommended by WHO (O’Neill and Tanner, 1989). However, some cases of cirrhosis reported inIndia, and in other countries have occurred in breast-fed infants or despite the virtual absence ofcopper in the infant’s drinking water suggesting an inherited metabolic defect as a cause for thiscondition (Horslen et al., 1994). Furthermore, an epidemiological study performed in seven towns ofMassachusetts, with copper concentrations in the drinking water ranging from 8.4 - 8.8 mg/L, did notshow any deaths due to liver disease (Scheinberg and Sternlieb, 1994). A recent publication whichsystematically evaluated the association of infantile liver disease and copper content of drinking waterin the UK did not reveal a connection between these variables (Fewtrell et al., 1996).

The familial occurrence and consanguinity in the parents of patients in some cases of infantilecirrhosis strongly suggest a genetic disorder as the aetiology of this disease. A recent report from theprovince of Tyrol, Austria, reviewed data from 138 infants and young children dying from infantile livercirrhosis during this century (Müller et al., 1996). Clinical features of the cases were indistinguishablefrom ICT or ICC, and pedigree analysis of the affected families indicated that the susceptibility to thisdisease was inherited as an autosomal recessive gene.

The sex ratio was 0.5 and parental consanguinity increased the risk. Segregation analysis based on a totalof 343 children and 112 cases of childhood cirrhosis favoured a genetic factor in the aetiology of thedisease. The evaluation of infant and childhood feeding practices indicated that untinned copper andbrass cooking utensils contributed to the development of this disease by providing extremely high dietarycopper intakes. The replacement of copper cooking utensils by modern non-copper pots occurring in thepast decades has eradicated the disease, such that no cases have been diagnosed after 1974.

C o p p e r

16

In conclusion, ICC/ICT is a disease of unknown aetiology. The most likely explanation for thiscondition appears to be a combination of a genetically determined defect in copper metabolism and ahigh copper intake. The relative contribution of each factor remains to be determined.

The range of acceptable intake to prevent copper deficiency and toxicity should be based on theprotection of healthy populations and should not be expected to meet requirements or prevent excessof special individuals. Disease conditions or genetic alterations in copper metabolism that determinespecial sensitivity for excess or deficit deserve the attention of public health authorities based on therelevance of these conditions within a given ecological setting. The upper and lower cut-off points forthe range of acceptable oral intakes should be defined using a population-based model for theassessment of health risks associated to deficiency or excess. The lower cut-off point should besufficient to meet the requirements of most individuals in the population. Similarly the upper endpoint should protect most individuals from the risk of toxicity (see figure 1). A detailed presentationof this risk assessment model as applicable to essential elements is presented in Dr. Becking’s chapter(Becking, 1996).

Figure 1: Model to illustrate distribution of individual copper requirements and toxicity in a normalpopulation and sub groups with genetic abnormalities in copper metabolism. ICC is IndianChildhood Cirrhosis, ICT is Idiopathic Copper Toxicosis. The upper and lower cut-off points for therange of acceptable oral intakes are defined using the population distributions of requirements andtoxicity. The lower cut-off point is sufficient to meet the requirements of most individuals in thepopulation while the upper limit prevents toxicit in most individuals.

C o p p e r

17

References

Adamsom, M., Reiner, B., Olson J.L. et al. (1992), ‘Indian childhood cirrhosis in an American child’,Gastroenterology, vol. 102, pp. 1771-1777.

Becking, G.C. (1996), ‘Risk Assessment for Essential Trace Elements: A Proposed Methodology’ inpress.

Cartwright, G.E. and Wintrobe, M.M. (1964), ‘Copper metabolism in normal subjects’, Am J ClinNutr., vol. 14, pp. 224-232.

Cordano, A., Baertl, J. and Graham, G.G. (1964), ‘Copper deficiency in infants’, Pediatrics, vol. 34, pp.324-336.

Ehrenkranz, R.A., Gettner, P.A., Nelli, C.M., Sherwonit, E.A., Williams, J.E., Ting, B.T.G. andJanghorbani, M. (1989), ‘Zinc and copper nutritional studies in very low birth weight infants:comparison of stable isotopic extrinsic tag and chemical balance methods’, Pediatr Res., vol. 26, pp.298-307.

FAO/IAEA/WHO. (1996), ‘Copper In: Trace elements in human nutrition and health’, WHO(Geneva), pp. 123-143.

Fewtrell, L., Kay, D., Jones, F., Baker, A. and Mowat, A. (1996), ‘Copper in drinking water - aninvestigation into possible health effects’, Public Health, vol. 110, pp. 175-177.

Harris, Z.L. and Gitlin J.D. (1996), ‘Genetic and molecular basis for copper toxicity’, Am J ClinNutr., vol. 63, pp. 836S-41S.

Horslen, S.P., Tanner, M.S., Lyon, T.D.B., Fell, G.S. and Lowry, M.F. (1994), ‘Copper associatedchildhood cirrhosis’, Gut, vol. 35, pp. 1497-1500.

Knobeloch, L., Ziarnik, M., Howard ,J., Theis, B., Farmer, D., Anderson, H. and Proctor, M. (1994),‘Gastrointestinal upsets associated with ingestion of copper-contaminated water’, Environ HealthPerspect ., vol. 102, pp. 958-961.

Linder, M.C. and Hazegh-Azam, M. (1996), ‘Copper biochemistry and molecular biology’, Am J ClinNutr., vol. 63, pp. 797S-811S.

Lönnerdal, B. (1996), ‘Bioavailability of copper’, Am J Clin Nutr., vol. 63, pp. 821S-9S.

Luza, S.C. and Speisky, H. (1996), ‘Liver copper storage and transport during development:implications for cytotoxicity’, Am J Clin Nutr, vol. 63, pp. 812S-20S.

Müller, T., Feichtinger, H., Berger, H. and Müller, W. (1996), ‘Endemic Tyrolean cirrhosis: anecogenetic disorder’, Lancet, vol. 347, pp. 877-80.

National Institute of Public Health and Environmental Protection (NIPHEP). (1989), ‘Integratedcriteria document copper’, Appendix to report Nº 758474009. The Netherlands, Bilthoven: NationalInstitute of Public Health and Environmental Protection.

National Research Council (NRC).(1989), ‘Recommended dietary allowances’,10th edn, Washington,DC, National Academy Press, pp. 224-30.

O'Neill, N.C. and Tanner, M.S. (1989), ‘Uptake of copper from brass vessels by bovine milk and itsrelevance to Indian childhood cirrhosis’, J Pediatr Gastroenterol Nutr., vol. 9, pp. 167-172.

Olivares, M. and Uauy, R. (1996a), ‘Limits of metabolic tolerance to copper and biological basis forpresent recommendations and regulations’, Am J Clin Nutr., vol. 63, pp. 846S-52S.

Olivares, M. and Uauy, R. (1996b), ‘Copper as an essential nutrient’, Am J Clin Nutr., vol. 63, pp.791S-6S.

Pandit, A. and Bhave, S. (1996), ‘Present interpretation of the role of copper in Indian childhoodcirrhosis’, Am J Clin Nutr., vol. 63, pp. 830S-5S.

C o p p e r

18

Pennington, J.A.T., Schoen, S.A., Salmon, G.D., Young, B, Johnson, R.D. and Marts, R.W. (1995),Composition of core foods of the US food supply, 1982-1991. III. Copper, manganese, selenium, andiodine’, J Food Compos Anal., vol. 8, pp. 171-217.

Scheinberg, I.H. and Sternlieb I. (1994), ‘Is non-Indian childhood cirrhosis caused by excess dietarycopper?’, Lancet, vol. 344, pp.1002-1004.

Scheinberg, I.H. and Sternlieb I.(1996), ‘Wilson disease and idiopathic copper toxicosis’, Am J ClinNutr., vol. 63, pp. 842S-5S.

Shaw, J.L.C. (1992), ‘Copper deficiency in term and preterm infants’. In: Fomon SJ, Zlotkin S, eds.Nutritional anemias. Nestlé Nutrition Workshop Series. Vol. 30. New York: Raven Press, pp. 105-19.

Spitalny, K.C., Brondum, J., Vogt, R.L., Sargent, H.E. and Kappel, S. (1984), ‘Drinking water-induced intoxication in a Vermont family’, Pediatr., vol.74, pp. 1103-1106.

Turnlund, J.R., Keyes, W.R, Anderson, H.L. and Acord, L.L. (1989), ‘Copper absorption andretention in young men at three levels of dietary copper by use of the stable isotope 65Cu’, Am J ClinNutr., vol. 49, pp. 870-8.

US Environmental Protection Agency (USEPA).(1985), ‘Drinking water criteria document forcopper’ (final draft). Environmental Criteria and Assessment Office. Cincinnati, OH: EnvironmentalProtection Agency, EPA 600/X-84/190-1.

US Environmental Protection Agency (USEPA). (1987), ‘Summary review of the health effectsassociated with copper’. Environmental Criteria and Assessment Office. Cincinnati, OH:Environmental Protection Agency, EPA 600/8-87/001.

Weiss, M., Müller-Höcker, J., Wiebecke, B. and Belohradsky, B.H. (1989), ‘First description ofIndian childhood cirrhosis in a non-Indian infant in Europe’, Acta Paediatr Scand ., vol. 79, pp. 152-156.

WHO Report of an Expert Committee. (1985), ‘Trace elements in human nutrition’. World HealthOrgan Tech Rep Ser 724.

Wyllie, J. (1957), ‘Copper poisoning at a cocktail party’, Am J Public Health, vol. 47, p. 617.

Genetic disorders of copper metabolism and the dual nature ofcopper in biology

Julian F.B. MercerMurdoch Institute,Royal Children’s Hospital, Victoria, Australia

1. Introduction

Copper is an essential element which is used as a cofactor for a number of important enzymesincluding cytochrome-c oxidase, the last enzyme in the electron transport chain. Excess copper is,however, toxic to cells and this dual nature of copper, essential yet toxic, must have required thedevelopment of a tightly regulated system of copper homeostasis. Since cytochrome-c oxidase is anenzyme that has been used by all aerobic organisms, the use of copper is an early event in evolution,and one might expect that some of the molecules involved in copper transport would be wellconserved.

The essentiality and toxicity of copper are graphically demonstrated in two genetic disorders ofcopper transport. Menkes disease (MD) is an X-linked recessive condition, which causes death inearly childhood from copper deficiency. Wilson disease (WD) on the other hand is a copper toxicosisdisorder characterised by massive copper accumulation in the liver, with consequent aberrant releaseof copper and damage to the central nervous system (Danks, 1995). Despite the differences in

C o p p e r

19

phenotype, these diseases are caused by mutations in closely related genes which encode copperATPases (Mercer et al., 1993; Chelly et al., 1993; Vulpe et al., 1993; Bull et al., 1993; Tanzi et al.,1993; Yamaguchi et al., 1993). From the structure of the predicted proteins and the pattern of copperdisturbance in the two diseases, it appears likely that both function as copper efflux molecules. Verysimilar molecules have been found to be copper transporters in yeast and bacteria (Fu et al., 1995;Rad et al., 1994; Odermatt et al., 1993).

Current research in this area is aimed at elucidating the role of these novel Cu-ATPases in coppertransport and the types of mutations which cause the MD and WD. The features shown by theseextreme phenotypes are yielding valuable insights into the normal processes of copper requirementsand protective mechanism against copper toxicity. This paper will review some of the informationcoming from these molecular studies and attempt to relate some of this information with older studieson the effects of copper deficiency and toxicity.

2. The essential nature of copper

2.1 Nutritional copper deficiency

Copper was first established as an essential nutrient in the 1920s and 1930s from studies of laboratoryand farm animals. Many of the features found in these copper deficient animals can be understood interms of the reduction in the activity of various copper dependent enzymes, but there are still aspectsof copper deficiency that are not fully understood.

One of the most sensitive indicators of copper deficiency in animals is the hypopigmentationassociated with reduced activity of tyrosinase, a copper dependent enzyme required for melaninformation.

In sheep, alternate bands of pigmentation in wool can be produced by adding or withdrawing copperfrom the diet (Underwood, 1977). Another effect of copper deficiency in sheep is the production of“steely” wool, that is, wool which has lost the natural crimp. This is important for wool producerssince steely wool has low tensile strength and reduced elasticity (Underwood, 1977). Steely hair is acharacteristic feature of children with Menkes disease (see below). Copper is required for the normalcross-linking of the sulphydryl groups of keratin, but no enzyme has been associated with thisprocess.

Neurological problems are a feature of copper deficiency in developing animals and are pronouncedfeature of MD. Neonatal ataxia is a defect of myelination found in copper deficient lambs and is alsoknown as “swayback”. The disease is directly due to deficient myelination in the spinal chord, andthere are also lesions in parts of the cerebral white matter and brain stem (Smith et al., 1977). Alikely cause of the demylenation is a deficiency in cytochrome-c oxidase.

Deficiency of the copper dependent enzyme, lysyl oxidase, causes defective cross linking of collagenand elastin leading to abnormalities of skin, cartilage, arteries and bone. Various degrees ofconnective tissue abnormalities are found in MD and its variants, and it appears that these geneticdefects can produce a more specific connective tissue phenotype than nutritional deficiency. Possiblereasons for this will be discussed below.

Iron-unresponsive anaemia and neutropenia occur in severe copper deficiency, and part of theexplanation is that the copper oxidase, ceruloplasmin is required for oxidation of Fe2+ to Fe3+ fortransferrin binding. For example, when plasma ceruloplasmin level fell below 1% of normal in copperdeficient pigs, the cell to plasma iron flow was impaired and cellular iron accumulated (Roeser et al.,1970). More recently patients with a genetic disorder aceruloplasminaemia, have a been shown todevelop an iron disorder, haemosiderosis (Harris et al., 1995).

It has been more difficult to establish the effects of Cu deficiency in humans than in animals, andindeed the very possibility of Cu deficiency in humans was debated in the 1950s and 1960s. Asdiscussed below, the severe effects of Cu deficiency brought about by the genetic defect in MDprovided decisive evidence of the effects of copper deficiency in humans. Severe copper deficiencyhas also been observed in patients receiving total parenteral nutrition with inadequate copper content

C o p p e r

20

(Karpel and Peden, 1972). The question of whether mild chronic copper deficiency is a problem inhumans is still unresolved, and is difficult to answer because of the lack of reliable indicators ofmarginal copper status (Milne, 1994).

2.2. Menkes disease and its variants

MD is the most severe form of a spectrum of genetic copper deficiency conditions, which all resultfrom mutations in the Menkes disease gene (MNK). Children with MD are copper deficient at birth,because of defective placental transfer of copper and continue to suffer from the effects of thedeficiency, since the transfer of copper across the small intestine is very inefficient. Thephysiological effects of the reduced copper intake are exacerbated by the fact that MNK is requiredfor the transport of copper across epithelial cellular layers and within cellular organelles of most celltypes. Hepatocytes are the exception, for in these cells the function of MNK is fulfilled by WND, theprotein affected in Wilson disease. As illustrated in Fig 1a, MNK is needed to pump copper from thesmall intestine into the body. Since this step is blocked, copper accumulates in the mucosal epithelialcells, bound to metallothioneins (MT). MNK is also required to transport copper across the bloodbrain barrier (Fig 1b), effectively imposing a second block to copper transport at a period ofdevelopment when the developing brain is especially sensitive to the lack of copper. It is likely thatthe resulting deficiency of cytochrome oxidase is instrumental in producing the profound neurologicaldefects in this disease; in neonatal mice deficiency of copper has long lasting effects on cytochromeoxidase and some other cuproenzymes in the brain (Prohaska and Bailey, 1993).

(In this chapter MNK is used for the gene affected in MD and MNK for its protein product. MNK isalso known as ATP7a. WND is used for the gene affected in Wilson disease and WND for the protein.WND is also known as ATP7b).

C o p p e r

21

Figure 1: Transport of copper out of most cells requires the Menkes gene product, MNK.

(a) In polarized epithelial cells, such as the intestinal enterocytes and the proximal tubules of the kidney, MNKpossibly effluxes copper across the basolateral surface of the cell. This process is blocked in MD and copperaccumulates in the cell bound to metallothioneins (MT).

(b) MNK is needed to efflux copper from the vascular endothelial cells, e.g. (1) into the brain and into capillaries(2) following absorption from the diet.

(c) Cultured fibroblasts from MD patients accumulate copper since efflux of copper is defective. Also MNK isrequired to transport copper into the Golgi for incorporation into the copper dependent enzyme lysyl oxidase,but cytoplasmic enzymes such as superoxide dismutase (SOD) can receive copper directly.

C o p p e r

22

Paradoxically, copper levels in some organs are elevated in MD patients, for example the kidneycopper concentrations are about 5 fold normal. The convoluted tubules in the kidney are involved incopper resorption from the urine and MNK is probably needed to pump copper across the basolateralsurface back into the circulation (Fig 1a). When MNK is absent, the copper continues to be absorbedacross the apical surface but is trapped in the cell. Fibroblasts cultured from MD patients alsoaccumulate copper and this has been shown to be due to defective efflux (Fig 1c). Measurement ofcopper efflux from amniotic cells is a reliable method of prenatal diagnosis of MD (Tonneson andHorn 1989).

The mildest mutant allele of MNK causes a disease termed occipital horn syndrome (OHS)alternatively known as cutis laxa. OHS is primarily a connective tissue disorder, due to low activityof the copper dependent enzyme, lysyl oxidase (Byers et al., 1980). The neurological defects that areso apparent in MD are only present to a mild degree in this disorder. The prominence of theconnective tissue defects suggests that lysyl oxidase is particularly sensitive to the lack of MNK. Anexplanation for this is that MNK is needed for the transport of copper across an intracellular organellein which lysyl oxidase receives copper (Das et al., 1995). This may occur in the trans Golgi networkand this is where MNK is found in copper resistant CHO cells (Fig 1c and see below). Therequirement for an additional MNK-dependent step may render lysyl oxidase very sensitive toreduction in MNK activity (Das et al., 1995). An intermediate phenotype, termed mild Menkes, hasmore severe mental retardation, but patients survive the critical neonatal period (Danks,1988).

The reason for the different phenotypes arising from mutations of MNK is not fully understood, butmolecular analysis of the mutations in the various diseases has shown that MD arises when there islittle or no activity of MNK. The types of mutations causing MD are deletions of the MNK gene,premature stop codons, major splice site mutations and serious missense mutations (Das et al., 1994).The milder disorders OHS and mild Menkes have been shown to result from less severe splice sitemutations which allow some normal MNK mRNA, and presumably protein, to be formed (Das et al.,1995; Kaler et al., 1994). It is possible as with a number of inborn errors of metabolism, that residualenzyme activity of greater than about 5% allows sufficient copper to be transported to permit nearnormal development.

2.3 The mottled mice

Mutations of the X-linked mottled (Mo) gene in mice produce a range of phenotypes which includeclose homologues to MD and OHS all with defects of copper transport (Lyon and Searle 1990). Thebrindled mouse (Mobr) has features closely resembling those seen in MD; the affected male dies atabout 14 days after birth, a similar developmental stage at which boys with MD usually die. Featuresinclude hypopigmented coat, curly whiskers (like the hair defect in Menkes), connective tissueabnormalities and severe neurological defects. The tissue copper levels are abnormal and like MDpatients, copper concentrations are elevated in the small intestine and kidney but the liver and brainare markedly copper deficient (Mercer et al., 1991; Camarakis 1979). The blotchy mutant (Moblo)resembles occipital horn syndrome patients since the principal defects are in the connective tissue,and the mouse survives to adult life but dies prematurely from aortic aneurysms, and has a markedlysyl oxidase deficiency (Rowe et al., 1977).

An interesting severe mottled variant is the dappled mutant (Modp) which dies during fetaldevelopment (Phillips, 1961). We have been characterising a similar severe allele Mo9H, in whichfetal death appears to be caused by connective tissue failure. Multiple fetal defects have been foundranging from abdominal wall rupture , vascular abnormalities and even the necrosis of the entire hindquarter of the fetus (Ambrosini and Mercer, unpublished data). There appears to be no humanequivalent of these fetal lethal mutants.

Despite the variable phenotypes, it appears likely that all the mottled mice arise from mutations of theMenkes gene homologue and show the range of defects that copper deficiency can produce in themouse (Levinson et al., 1994, Mercer et al 1994). Fetuses affected by the severe dappled mutationexpress little or no MNK mRNA, like patients with MD, but for unknown reasons the absence of

C o p p e r

23

MNK in the mouse results in a more severe phenotype (ie. foetal death) than in humans. The blotchymouse has a mild splice site mutation of exactly the type reported in OHS patients (Das et al., 1995),thus confirming that milder defects result in predominantly connective tissue abnormalities. We haverecently identified a two amino acid deletion in the brindled mouse MNK, in a highly conserved, butnot well studied region of MNK (Grimes et al., 1997). Such a mutation is likely to allow productionof normal amounts of protein with a greatly reduced activity, and Western blot data using anantiserum specific for MNK have demonstrated that the mutant does have normal amounts of MNK inthe kidney.

3. The toxic nature of copper

3.1 Excess copper disposed of by biliary excretion

Oral copper is relatively non-toxic for humans and most animals, although there are markeddifferences in copper homeostasis between species. Neonatal animals are thought to be more sensitiveto excess copper intake. The disposal of excess dietary copper is achieved by regulating the rate ofbiliary excretion, which is the principal way copper is removed from the body (Evans, 1973). Sheepstore large amounts of copper in the liver and are particularly sensitive to copper toxicosis since,unlike most other species, the rate of copper excretion in the bile is not influenced by theconcentration of copper in the liver (Danks, 1995). On normal pastures, without excess molybdenum,which is an effective antagonist of copper uptake, sheep will gradually accumulate copper. This issimilar to the hepatic accumulation of copper in Wilson disease (Danks, 1995). Some breeds of sheepare avid copper accumulators and will develop copper toxicosis even on normal pastures (MacLachlanand Johnston, 1982). Rodents represent the other extreme of copper adaptation; rats and miceefficiently excrete copper and very high dietary intakes are needed to elevate hepatic copper levels.For example rats need to consume diets containing about 300 times the standard level of copper forseveral months to raise hepatic concentrations to levels similar to those found in Wilson disease(Haywood 1985).

3.2 Wilson disease

In patients with Wilson disease (WD) the biliary excretion of copper is greatly reduced, and sincecopper absorption from the diet is unaffected, liver copper levels steadily rise until hepatocytes arekilled and the patient enters a phase of terminal liver failure (Danks, 1995). Interestingly a variantform of the disease can present as a neurological condition; copper is thought to be released fromnecrotic hepatocytes and deposits in the central nervous system, ultimately causing neurologicaldamage, presumably by the promotion of free radical damage to the neurones.

The molecular basis of WD has been shown to be a defective copper ATPase, WND, closely relatedto MNK (Bull et al., 1993; Tanzi et al., 1993). Both molecules have the characteristics of P-typeATPases and have the interesting cluster of copper binding sites in the N-terminal region (see below).

It is thought that WND will also be a copper efflux molecule, functioning to deliver copper to the bileperhaps across the biliary cannalicular membrane (see Fig. 3); Since ceruloplasmin levels are low inmost patients with WD, WND probably also pumps copper into an intracellular organ (possibly theGolgi apparatus) for incorporation into ceruloplasmin, although this conclusion has been recentlydisputed (Chowrimootoo 1996).

Neurological damage in two other human diseases may involve copper. The genetic disorder,amyotrophic lateral sclerosis, a neurodegenerative condition, has been shown in some cases to be dueto mutant forms of copper/zinc superoxide dismutase, which may actually increase free radicalformation (Wiedau-Pazos et al., 1996). More recently copper has been implicated as playing a role inthe development of Alzheimer’s disease (Multhaup, 1996), so there is clearly a lot more to learn aboutthe toxic potential of copper in biology.

C o p p e r

24

3.3 Animal models of WD

As noted above, sheep have very low biliary excretion of copper and accumulate very high levels inthe liver, much as is found in WD. This characteristic of sheep possibly represents an adaptation of aspecies to chronically low copper intakes, the accumulated hepatic copper acting as a reserve whensupplies are limited (Weber, 1980). This tendency can be a disadvantage under modern agriculturalconditions when copper supplies are adequate leading to copper toxicosis (MacLachlan and Johnston,1982). The molecular basis of this copper accumulation tendency has not been established.

Another animal model of WD is a breed of dog, the Bedlington Terrier, which commonly develops acopper toxicosis with many of the features of WD (Su et al., 1982). Brewer has studied the linkage ofDNA markers to copper toxicosis and has concluded that the affected gene is unlikely to be WND. Ifthis proves to be correct then it may suggest that another gene is involved in the hepatic metabolism ofcopper (Yuzbasiyan, 1993).

The mutant LEC rat, has been recently shown to have a partial deletion of the WND homologue, thusestablishing it as the first true animal WD model (Wu et al., 1994). Copper accumulates in the liverof the mutant rat due a defect of biliary excretion (Sugawara et al., 1993) and is associated mainlywith metallothioneins (Suzuki, 1995). In common with patients with WD, incorporation of copperinto ceruloplasmin in the Golgi apparatus is greatly reduced in the mutant (Murata, 1995). Themutant has the unusual feature of developing a high frequency of hepatitis and hepatomas (Fujimoto,1989), which has been attributed to the high copper concentration in the liver, but recently thetendency to develop hepatomas has been shown to segregate independently of the copperaccumulation phenotype (Hattori, 1995).

We have been studying a mouse model of WD which has been termed the toxic milk mouse, tx(Rauch, 1983). The mutant mouse accumulates high levels of hepatic copper associated withmetallothioneins and the pattern of accumulation is very similar to that seen in the LEC rat (Howelland Mercer, 1994). However, this mouse has the unusual feature that the pups of mutant dams areborn copper deficient and the milk produced by the tx dam is copper deficient. The combined effectsof this copper deficiency is often fatal to the neonates. Despite this copper deficiency, unexpected ina model of copper toxicosis, we have found a mutation in the WD homologue, showing that thismouse is also a true model of WD (Theophilos et al., 1996). The cause of the copper deficiency in thefetus and milk is not clear, but may be related to the disturbance of copper transport in the liver, but ifso then similar copper deficiency would be expected in the LEC rat and WD mothers, and so far thishas not been reported.

4. Molecular and cellular basis of copper transport

4.1 Structural and function of MNK and WND

The predicted structure of MNK and WND contains a number of interesting features. Firstly they areclearly a member of the family of cation transporters, P-type ATPases (Pederson and Carafoli, 1987).These are transmembrane proteins which move cations across the membrane by means ofconformation changes induced by the binding of cations and hydrolysis of ATP.

As part of the reaction cycle, a critical aspartic acid residue becomes phosphorylated, hence the nameP-type ATPase. MNK and WND are closely related ATPases and it has been suggested that theyform a subgroup of heavy metal transporting P-type ATPases termed CpX type ATPases, somewhatdistinct from the more commonly known Ca and Na/K ATPases (Solioz and Vulpe, 1996). They arein fact, more closely related to bacterial copper transporters (Bull and Cox, 1994).

It is predicted that most of the molecule is in the cytoplasm, with 6 or 8 transmembrane loops forminga channel through which copper passes following the hydrolysis of ATP (Vulpe et al., 1993; Bull andCox 1994). A two dimensional representation of MNK, which applies also to WND, is shown in Fig2. These models are based on the extensively studied calcium ATPases (Brandl et al., 1986). This

C o p p e r

25

model suggests that these molecules function in copper efflux from the cell and a role in effluxaccords with both the phenotype of MD and WD.

Figure 2: Hypothetical arrangement of MNK in the cell membrane. There are eight transmembraneloops (TM 1-8), six potential copper binding sites (Cu 1-6) in the N-terminal region , the amino acidscys.pro.cys (CPC) are found in TM 6 and possibly bind copper in the channel. The position of theaspartic acid which is phosphorylated during the reaction cycle is indicated as D-phos and the ATPbinding site is shown as ATP. The arrow heads indicate the regions encoded by each exon, numbered2 to 23.

As illustrated in Fig 1, it is likely that MNK functions as the main copper efflux molecule from mostcells except hepatocytes. Little or no MNK mRNA is found in the liver (Vulpe et al., 1993, Paynter etal., 1994). Uptake of copper is normal in cells from MD patients, but efflux is defective (Herd et al.,1987). Thus absorption of copper from the small intestine cannot occur because MNK is required toefflux copper across the basolateral surface of the enterocyte into the blood (Fig 1a); similarly copperaccumulates in the kidney proximal tubules of the brindled mouse (Yoshimura, 1994), because MNKis needed to efflux the resorbed copper across the basolateral surface back into the blood. As notedabove, the deficiency of copper in the brain is exacerbated since MNK is again required for the effluxof copper from the cytoplasm of the cells which constitute the blood brain barrier into the brain (Fig1b). Copper is found to accumulate in these vascular endothelial cells in the mottled mouse mutants(Kodama 1993).

WND is expressed mainly in the liver, but significant amounts of WND mRNA is found in othertissues where MNK is also expressed. It is not known whether the two proteins have distinct roles insuch a situation. As shown in Fig 3, WND probably functions to efflux copper from the hepatocyteinto the bile. Copper efflux may be via direct biliary excretion across the cannalicular membrane orvia a vesicular efflux system, eg. lysosomes. There is evidence for two modes of copper secretioninto the bile (Gross et al., 1989; Dijkstra et al., 1995; Nederbragt, 1989). WND most probably donatescopper to ceruloplasmin in the Golgi (Muruta et al., 1995). Ceruloplasmin in turn may function as acopper transporter to peripheral tissues (Lee et al., 1993; Harris and Percival 1989). Presumably theactivity of WND is regulated in some way by the copper status of the liver, such that the excretion ofcopper is increased when the concentration becomes too high. Little is known of such a mechanism,however, our work with MNK in cultured cells is suggesting some interesting possibilities, as isdiscussed in the next section.

C o p p e r

26

Figure 3: Efflux of copper from hepatocytes and incorporation of copper into ceruloplasmin (CP)requires WND. For efflux into the bile, WND may be located on the biliary cannalicular membraneor perhaps on the lysosome. CP probably receives copper in the Golgi apparatus, and hence WND isprobably also located in this organelle.

4.2 Copper resistance is acquired by amplification of MNK

Cells in culture can be selected for resistance to copper by growth in progressively increasedconcentrations of the metal ion. In the case of hepatocytes, resistant cells have elevated expression ofmetallothionein (Freedman et al., 1986).

In copper resistant Chinese hamster ovary (CHO) cells, however, metallothionein expression has notbeen found to contribute to copper resistance. Instead, the Menkes gene homologue is amplified,leading to production of up to 100 fold more MNK (Camakaris et al., 1995). The increased amount ofMNK allows the cells to efflux more copper and maintain low intracellular copper levels even inculture media containing high amounts of copper. This is the first direct evidence that MNK canfunction to efflux copper. Significantly, the increased efflux is only apparent at high concentrationsof copper in the medium. In normal media there is little difference between the efflux rates of themost resistant cell, CUR3 and the parental K1 (Camakaris et al., 1995), suggesting that the cell is ableto regulate the activity of MNK.

We have used these cell lines to determine the intracellular location of MNK, using an antibodyagainst the N-terminal one third of MNK. Our data suggest that the primary location of MNK is inthe transGolgi Network, and significantly this localisation may be altered by growth in high copperconcentrations, with more MNK appearing on the plasma membrane (Petris et al., 1996). Previousdata on the levels of MNK mRNA suggested that amounts of mRNA are not altered by copperdeficiency or copper loading (Paynter et al., 1994), so the regulation of cellular copper levels may bemediated by changes in the intracellular localisation of MNK. This may also be the case for WND inthe liver.

C o p p e r

27

4.3 Do metallothioneins protect against copper toxicity?

Copper resistant hepatocytes have elevated amounts of metallothionein (MT) in their cytoplasm(Freedman et al., 1986), but this is not the case in the copper resistant CHO cells. The lack of MTelevation in copper resistant CHO cells may suggest that MTs do not protect all cell types fromcopper toxicity, however, MT genes are not expressed in the parental CHO cell K1, due to genemethylation, and hence the cell may use amplification of MNK rather than expression of MT toachieve copper resistance.

More evidence for the protective effect of MT against copper is the invariant association of high MTlevels in cells in which copper has accumulated due to a genetic defect of copper transport. Forexample in the LEC rat and the toxic milk mouse, both models of WD, the excess hepatic copper isbound to MTs (Suzuki et al., 1993; Koropatnick and Cherian, 1993). Cultured cells from MD patientsaccumulate copper and this excess copper is bound to MTs (LaBadie et al., 1981). This association ofMT with high copper and the ability of copper to induce MT synthesis (Wake and Mercer 1985;Durnam and Palmiter 1981) strongly suggests that MT is protective against copper toxicity.

It was surprising, therefore, to find that mice which lack the MT-I and MT-II genes (MT minus mice),did not show increased sensitivity to high levels of copper in the drinking water, but were verysensitive to cadmium (Michalska and Choo 1993; J. Mercer, unpublished data). The non-toxicity ofcopper for the MT minus mice would be explained if both the normal and MT minus mice have a veryeffective biliary excretion mechanism and hepatic copper levels were never high enough to causetoxicity. Indeed, in contrast to rats on the same level of copper which accumulate large amounts ofhepatic copper (Gross et al., 1989), the hepatic copper in the mice was only marginally increased (J.Mercer, unpublished data).

A marked effect of copper toxicity in the absence of metallothionein is observed in a double mutantbrindled/ MT minus mouse. The double mutant mice die very early in development (less than 10 daysgestation). This is most likely due to copper toxicity, perhaps causing placental failure (Kelley andPalmiter, 1996). Certain tissues including the placenta of the brindled mouse accumulate copper dueto defective efflux, like in MD, without MT to complex the copper, cell death may occur.

5. Conclusions

The genetic defects of copper transport in humans, MD and Wilson disease have finally provided thekey to advancing our understanding of some of the key molecules involved in homeostasis of copper.Moreover the structure of these genes and future molecular and cellular investigations promise toestablish the molecular basis of copper homeostasis. Further understanding of the molecules used bymammals to handle copper is coming from the metallothionein minus mouse, a mutant generated byhomologous recombination. The power of modern molecular and cellular biological techniques, willprovide information which can be integrated with previous knowledge of copper nutrition andtoxicological studies to yield a depth of knowledge of copper transport and regulation which has beenimpossible to obtain previously.

Acknowledgements

I am grateful to Michelle Winsor for her help with the preparation of the figures.

C o p p e r

28

References

Brandl, C. J., Green, N. M., Korczak, B. and MacLennan, D. H. (1986), ‘Two Ca2+ ATPase genes:homologies and mechanistic implications of deduced amino acid sequences’, Cell, vol. 44, pp. 597-607.

Bull, P. C. and Cox, D. W. (1994), ‘Wilson’s disease and Menkes syndrome: new handles on heavy-metal transport’, Trends in Genetics, vol. 10, pp. 246-252.

Bull, P. C., Thomas, G. R., Rommens, J. M., Forbes, J. R. and Cox, D. C. (1993), ‘The Wilson’sdisease gene is a putative copper transporting P-type ATPase similar to the Menkes gene’, NatureGenet., vol. 5, pp. 327-337.

Byers, P. H., Siegel, R. C., Holbrook, K. A., Narayanan, A. S., Bornstein, P. and Hall, J. G. (1980),‘X-linked cutis laxa: defective cross-link formation in collagen due to decreased lysyl oxidaseactivity’, New Eng. J. Med., vol. 303, pp. 61-65.

Camakaris, J., Mann, J. R. and Danks, D. M. (1979), ‘Copper metabolism in mottled mouse mutants:copper concentrations in tissues during development’, Biochem J., vol. 180, pp. 597-604.

Camakaris, J., Petris, M., Bailey, L., Shen, P., Lockhart, P., Glover, T. W., Barcroft, C. L., Patton, J.and Mercer, J. F. B. (1995), ‘Gene amplification of the Menkes (MNK; ATP7A) P-type ATPase geneof CHO cells is associated with copper resistance and enhanced copper efflux’, Human Mol. Genet.,vol. 4, pp. 2117-2123.

Chelly, J., Turmer, Z., Tonnerson, T., Petterson, A., Ishikawa-Brush, Y., Tommerup, N., Horn, N.andMonaco, A. P. (1993), ‘Isolation of a candidate gene for Menkes syndrome that encodes apotential heavy metal binding protein’, Nature Genet., vol. 3, pp. 14-19.

Chowrimootoo, G. F. E., Ahmed, H. A. and Seymour, C. A. (1996), ‘New insights into thepathogenesis of copper toxicosis in Wilson’s disease: evidence for copper incorporation and defectivecanalicular tranport of ceruloplasmin’, Biochem. J., vol. 315, pp. 851-855.

Danks, D. M. (1995), ‘Disorders of copper transport. The metabolic and molecular basis of inheriteddisease’. Scriver, Beaudet, Sly and Valle eds. McGraw-Hill. New York.

Danks, D. M. (1988), ‘The mild form of Menkes disease: progress report on the original case’, J.Med. Genet., vol. 30, pp. 859-864.

Das, S., Levinson, B., Vulpe, C., Whitney, S., Gitschier, J. and Packman, S. (1995), ‘Similar splicingmutations of the Menkes/mottled copper-tranporting ATPase gene in occipital horn syndrome and theblotchy mouse’, Am. J. Hum. Genet., vol. 56, pp. 570-576.

Das, S., Levinson, B., Whitney, S., Vulpe, C., Packman, S. and Gitschier, J. (1994), ‘Diversemutations in patients with Menkes disease often lead to exon skipping’, Am. J. Hum. Genet., vol. 55,pp. 883-889.

Dijkstra, M., In ‘t Velt, G., van den Berg, G. J., Muller, M., Kuipers, F. and Vonk, R. J. (1995),‘Adenosine triphosphate-dependent copper transport in isolated rat liver plasma membranes’, J. Clin.,Invest, vol. 95, pp. 412-416.

Durnam, D. M. and Palmiter, R. D. (1981), ‘Transcriptional regulation of the mouse metallothionein-Igene by heavy metals’, J. Biol.Chem., vol. 256, pp. 5712-5716.

Evans, G. W. (1973), ‘Copper homeostasis in the mammalian system’, Physiological Reviews, vol.53, pp. 535-569.

Freedman, J. H., Weiner, R. J. and Peisach, J. (1986), ‘Resistance to copper toxicity of culturedhepatoma cells’, J. Biol. Chem., vol. 261, pp. 11840-8.

Fu, D., Beeler, T. J. and Dunn, T. M. (1995), ‘Sequence, mapping and disruption of CCC2, a genethat cross-complements the Ca2+ sensitive phenotype of csg1 mutants and encodes a P- type ATPasebelonging to the Cu2+ ATPase subfamily’, Yeast, vol. 11, pp. 283-292.

C o p p e r

29

Fujimoto, Y., Takahasi, H., Dempo, K., Mori, M. and al, e. (1989), ‘Hereditary hepatitis in LEC rats:accumulation of abnormally high ploid nuclei’, Cancer Detect. and Prevent., vol. 14, pp. 235-237.

Grimes, A., Hearn, C.J., Lockhart, P., Newgreen, D.F. and Mercer, J.F.B. (1997), ‘Molecular basis ofthe brindled Mouse mutant (Mobr): a murine model of Menkes disease’, Human Mol. Genet., vol. 6.In press.

Gross, J. B., Jr., Myers, B. M., Kost, L. J., Kuntz, S. M. and LaRusso, N. F. (1989), ‘Biliary copperexcretion by hepatocyte lysosomes in the rat. Major excretory pathway in experimental copperoverload’, J Clin Invest, vol. 83, pp. 30-9.

Harris, E. D. and Percival, S. S. (1989), ‘Copper transport: insights into a ceruloplasmin-baseddelivery system’, Adv. Exp. Medicine and Biology, vol. 258, pp. 95-102.

Harris, Z. L., Takahashi, Y., Miyajima, H., Serizawa, M., MacGillivray, R. T. and Gitlin, J. D. (1995),‘Aceruloplasminemia: molecular characterization of this disorder of iron metabolism’ Proc. Natl.Acad. Sci. USA, vol. 92, pp. 2539-2543.

Hattori, A., Sawaki, M., Enomoto, K., Tsuzuki, N., Isomura, H., Kojima, T., Kamibayashi, Y.,Sugawara, N., Sugiyama, T. and Mori, M. (1995), ‘The high hepatocarcinogen susceptibility of LECrats is genetically independent of abnormal copper accumulation in the liver’Carcinogenesis, vol. 16,pp. 491-494.

Haywood, S. (1985), ‘Copper toxicosis and tolerance in the rat. I—Changes in copper content of theliver and kidney’, J Pathol., vol. 145, pp. 149-58.

Herd, S. M., Camakaris, J., Christofferson, R., Wookey, P. and Danks, D. M. (1987), ‘Uptake andefflux of copper-64 in Menkes disease and normal continuous lymphoid cell lines’, Biochem J., vol.247, pp. 341-7.

Howell, J.McC. and Mercer, J. F. B. (1994), ‘The pathology and trace element status of the toxic milkmutant mouse’, J. Comp. Path., vol. 110, pp. 37-47.

Kaler, S. G., Gallo, L. K., Proud, V. K., Percy, A. K., Mark, Y., Segal, N. A., Goldstein, D. S.,Holmes, C. S. and Gahl, W. A. (1994), ‘Occipital horn syndrome and a mild Menkes phenotypeassociated with splice site mutations at the MNK locus’, Nature Genet., vol. 8, pp. 195-202.

Karpel, J. T. and Peden, V. H. (1972), ‘Copper deficiency in long-term parenteral nutrition’, J.Pediatr., vol. 80, pp. 32-36.

Kelley, E. J. and Palmiter, R. D. (1996), ‘A murine model of Menkes’ syndrome reveals aphysiological function of metallothionein’, Nature Genetics, vol. 13, pp.219-222.

Kodama, H. (1993), ‘Recent developments in Menkes disease’, J. Inher. Metab. Disease, vol. 16, pp.791-799.

Koropatnick, J. and Cherian, M. G. (1993), ‘A mutant mouse (tx) with increased hepaticmetallothionein stability and accumulation’, Biochem. J., vol. 296, pp. 442-449.

LaBadie, G. U., Beratis, N. G., Price, P. M. and Hirschhorn, K. (1981), ‘Studies of the copper-bindingproteins in Menkes and normal skin fibroblast lysates’, J. Cell Physiol., vol. 106, pp. 173-178.

Lee, S. H., Lancey, R., Montaser, A., Madani, N. and Linder, M. C. (1993), ‘Ceruloplasmin andcopper transport during the latter part of gestation in the rat’, Proc. Soc. Exp. Biol. Med., vol. 203, pp.428-439.

Levinson, B., Vulpe, C., Elder, B., Martin, C., Verley, F., Packman, S. and Gitschier, J. (1994), ‘Themottled gene is the mouse homologue of the Menkes disease gene’, Nature Genet.,vol. 6, pp. 369-373.

Lyon, M. F. and Searle, A. G. (1990), ‘Mo locus, Chromosome X’, Genetic Variants and Strains ofthe Laboratory Mouse, Lyon and Searle ed. Oxford University Press. Oxford.

MacLachlan, G. K. and Johnston, W. S. (1982), ‘Copper poisoning in sheep from North Ronaldsaymaintained on a diet of terrestrial herbage’, Vet Rec., vol. 111, pp. 299-301.

C o p p e r

30

Mercer, J. F. B., Grimes, A., Ambrosini, L., Lockhart, P., Paynter, J. A., Dierick, H. and Glover, T.W. (1994), ‘Mutations in the murine homologue of the Menkes syndrome gene in dappled and blotchymice’, Nature Genet, vol. 6, pp. 374-378.

Mercer, J. F. B., Livingston, J., Hall, B. K., Paynter, J. A., Begy, C., Chandrasekharappa, S.,Lockhart, P., Grimes, A., Bhave, M., Siemenack, D. and Glover, T. W. (1993), ‘Isolation of a partialcandidate gene for Menkes syndrome by positional cloning’, Nature Genet., vol. 3, pp. 20-25.

Mercer, J. F. B., Stevenson, T., Wake, S. A., Mitropoulis, G., Camakaris, J. & Danks, D. M. (1991),‘Developmental variation in copper, zinc metallothionein mRNA in brindled mutant and nutritionallycopper deficient mice’ Biochim. Biophys. Acta., vol. 1097, pp. 205-211.

Michalska, A. E. and Choo, K. H. A. (1993), ‘Targetting and germ line transmission of a nullmutation at the metallothionein I and II loci in mouse’, Proc. Natl. Acad. Sci. USA, vol. 90, pp. 8088-8092.

Milne, D. B. (1994), ‘Assessment of copper nutritional status’, Clin Chem., vol. 40, pp. 1479-84.

Multhaup, G., Schlicksupp, A., Hesse, L., Beher, D., Ruppert, T., Masters, C. L. and Beyreuther, K.(1996), ‘The amyloid precursor protein of Alzheimer’s disease in the reduction of copper (II) tocopper(I)’, Science, vol. 271: 1406-9.

Murata, Y., Yamakawa, E., Iizuka, T., Kodama, H., Abe, T., Seki, Y. & Kodama, M. (1995), ‘Failureof copper incorporation into ceruloplasmin in the golgi apparatus of LEC rat hepatocytes’, Biochem.Biophys. Res. Comm., vol. 209, pp. 349-355.

Nederbragt, H. (1989), ‘Effect of the glutathione-depleting agents diethylmaleate, phorone andbuthionine sulfoximine on biliary copper excretion in rats’, Biochem. Pharmacol., vol. 38, pp. 3399-3406.

Odermatt, A., Suter, H., Krapf, R. and Solioz, M. (1993), ‘Primary structure of two P-type ATPasesinvolved in copper homeostasis in Enterococcus hirae’, J. Biol. Chem., vol. 268, pp. 12775-12779.

Paynter, J. A., Grimes, A., Lockhart, P. and Mercer, J. F. B. (1994), ‘Expression of the Menkes genehomologue in mouse tissues: lack of effect of copper on the mRNA levels’, FEBS lett., vol. 351, pp.186-190.

Pederson, P. L. and Carafoli, E. (1987), ‘Ion motive ATPases. I. Ubiquity, propertes, and significanceto cell function’, TIBS, vol. 12, pp. 146-150.

Phillips, R. J. S. (1961), ‘Dappled, a new allele at the mottled locus in the house mouse’, Genet. Res.,vol. 2, pp. 209-295.

Petris, M.J., Mercer, J.F.B., Culvenor, J.G., Lockhart, P., Gleeson, P.A., Camakaris, J. (1996),‘Ligand-regulated transport of the Menkes copper P-Type ATPase efflux pump from the Golgiapparatus to the plasma membrane: a novel mechanism of regulated trafficking’, EMBO J., vol. 15,pp. 6084-6095.

Prohaska, J. R. and Bailey, W. R. (1993), ‘Persistent regional changes in brain copper, cuproenzymesand catecholamines following perinatal copper deficiency in mice’, J Nutr., vol. 123, pp. 1226-34.

Rad, M. R., Kirchrath, L. and Hollenberg, C. P. (1994), ‘A putative P-type Cu2+ transporting ATPaseon chromosome II of Saccharomyces cerevisiae’, Yeast, vol. 10, pp. 1217-1225.

Rauch, H. (1983), ‘Toxic milk, a new mutation affecting copper metabolism in the mouse’, J. Hered.,vol. 74, pp. 141-144.

Roeser, H. P., Lee, G. R., Nacht, S. and Cartwright, G. E. (1970), ‘The role of ceruloplasmin in ironmetabolism’, J. Clin. Invest., vol. 49, pp. 2408-2417.

Rowe, D. W., McGoodwin, E. B., Martin, G. R. and Grahn, D. (1977), ‘Decreased lysyl oxidaseactivity in the aneurism-prone mottled mouse’, J. Biol. Chem., vol. 252, pp. 939-942.

C o p p e r

31

Smith, R. M., Fraser, F. J. & Russell, G. (1977), ‘Enzootic ataxia in lambs: appearance of lesions inthe spinal chord during foetal development’, J. Comp. Path., vol. 87, pp. 119-128.

Solioz, M. and Vulpe, C. (1996), ‘CPX-Type ATPases: a class of P-type ATPases that pump heavymetals’, Trends in Biochem. Sci., vol. 21, pp. 237-241.

Su, L.-C., Owen, C. A., Zollman, P. E. and Hardy, R. M. (1982), ‘A defect of biliary excretion ofcopper in copper-laden Bedlington terriers’ Am. J. Physiol., vol. 243, pp. G231-G236.

Sugawara, N., Li, D., Sugawara, C. and Miyake, H. (1993), ‘Decrease in biliary excretion of copper inLong-Evans Cinnamon (LEC) rats causing spontaneous hepatitis due to a gross accumulation ofhepatic copper’, Res. Comm. Chem. Path. and Pharmacol., vol. 81, pp.45-52.

Suzuki, K. T. (1995), ‘Disordered copper metabolism in LEC rats, an animal model of Wilson’sdisease: role of metallothionein’, Res. Commun. Mol. Path. Pharmacol., vol. 82, pp. 221-239.

Suzuki, K. T., Yamamoto, K., Kanno, S., Aoki, Y. and Takeichi, N. (1993), ‘Selective removal ofcopper bound to metallothionein in the liver of LEC rats by tetrathiomolybdate’, Toxicology, vol. 83,pp. 149-158.

Tanzi, R. E., Petrukhin, K., Chernov, I., Pellequer, J. L., Wasco, W., Ross, B., Romano, D. M.,Parano, E., Pavone, L., Brzustowicz, L. M., Devoto, M., Peppercorn, J., Bush, A. I., Sternlieb, I.,Pirastu, M., Gusella, J. F., Evgrafov, O., Penchaszadeh, G. K., Honig, B., Edelman, I. S., Soares, M.B., Scheinberg, I. H. and Gilliam, T. C. (1993), ‘ The Wilson disease gene is a copper transportingATPase with homology to the Menkes disease gene’, Nature Genet., vol. 5, pp. 344-350.

Theophilos, M. B., Cox, D. W. and Mercer, J. F. B. (1996), ‘The toxic milk mouse is a murine modelof Wilson’s disease’, Hum. Mol. Genet., vol.5, pp. 1619-1624.

Tonneson, T. and Horn, N. (1989), ‘Prenatal and postnatal diagnosis of Menkes disease, an inheriteddisorder of copper metabolism’, J. Inher. Metab. Dis., vol. 12, pp. 207-214.

Underwood, E. J. ‘Copper.’Trace Elements in Human and Animal Nutrition. 1977 Academic Press.New York.

Vulpe, C., Levinson, B., Whitney, S., Packman, S. & Gitschier, J. (1993), ‘Isolation of a candidategene for Menke disease and evidence that it encodes a copper-transporting ATPase’, Nature Genet.,vol. 3, pp. 7-13.

Wake, S. A. & Mercer, J. F. B. (1985), ‘Induction of metallothionein mRNA in rat liver and kidneyafter copper chloride injection’, Biochem. J., vol. 228, pp. 425-432.

Weber, K. M., Boston, R. C. and Leaver, D. D. (1980), ‘A kinetic model of copper metabolism insheep’, Aust. J. Agric. Res., vol. 31, pp. 773-790.

Wiedau-Pazos, M., Goto, J. J., Rabizadeh, S., Gralla, E. B., Roe, J. A., Lee, M. K., Valentine, J. S.and Bredesen, D. E. (1996), ‘Altered reactivity of superoxide dismutase in familial amyotrophiclateral sclerosis’, Science, vol. 271, pp. 515-8.

Wu, J., Forbes, J. R., Chen, H. S. and Cox, D. W. (1994), ‘The LEC rat has a deletion in the coppertransporting ATPase homologous to the Wilson disease gene’, Nature Genet., vol. 7, pp. 541-545.

Yamaguchi, Y., Heiny, M. E. & Gitlin, J. D. (1993), ‘Isolation and characterization of a human livercDNA as a candidate gene for Wilson disease’, Biochem. Biophys. Res Comm., vol. 197, pp. 271-277.

Yoshimura, N. (1994), ‘Histochemical localization of copper in various organs of brindled mice’Pathol. Int., vol. 44, pp. 14-9.

Yuzbasiyan, G. V., Wagnitz, S., Blanton, S. H. and Brewer, G. J. (1993), ‘Linkage studies of theesterase D and retinoblastoma genes to canine copper toxicosis: a model for Wilson’s disease’,Genomics, vol. 15, pp. 86-90.

C o p p e r

32

Copper in the Aquatic EnvironmentDr Herbert E AllenDepartment of Civil and Environmental EngineeringUniversity of Delaware, Newark, Delaware U.S.A.

Copper is a ubiquitous trace metal that is required by aquatic organisms. At higher concentrations itis toxic. However, the bioavailability of copper varies greatly depending on the water chemistry.Thus, it is necessary to establish the physical and chemical form of copper in the environment beingstudied. Aquatic life criteria and standards must reflect these speciation effects to be predictive ofecosystem effects. To evaluate aquatic systems, it is necessary to obtain valid measurements of theconcentration of copper in the environmental samples and to compare these to criteria and standards.This paper provides a brief review of the most important factors involved in the evaluation ofpotential effects of copper in the aquatic environment

1. Analysis of copper

Copper concentrations in natural waters are frequently in the low µg/L concentration range. Toobtain valid data for such low concentrations it is necessary to take special precautions in samplingand analysis programs. The EPA (Prothro, 1993) has summarized the necessary components of themeasurement program:

1. Use of clean techniques during each step of the measurement process to avoid contamination.

2. Use of analytical methods that have sufficiently low detection limits.

3. Avoidance of interference in the quantification step.

4. Use of blanks to assess contamination.

5. Use of sample spikes and certified reference materials to assess the effects of contamination and interference on accuracy.

6. Use of replicates to assess precision.

Although the importance of none of these items should be underestimated, the first item is ofparticular importance. This includes such factors as acid washing sample bottles, use of “clean hands- dirty hands” operating protocols during sampling, and use of laminar flow hoods and benches andclean rooms for laboratory measurements. The analytical method should have a detection limit nogreater than one-tenth that of the sample and the concentration found for the blank should not exceedone-tenth that of the blank.

The need for these precautions in sampling and analysis for environmental studies was first pointedout by Patterson and Settle (1976) with regard to the analysis of lead. Since then it has becomeapparent that these considerations must be applied to the sampling and analysis of all trace metals(Bruland, 1983; Nriagu et al., 1993; U.S. EPA, 1995). Although the oceanographic researchcommunity quickly realized the importance of using these clean techniques, those concerned with theanalysis of freshwaters and effluents were slower to respond.

This attitude changed after the publication of Windom and his colleagues (1991) that demonstratedthe questionability of the data that had been collected for the U.S. Geological Survey’s NASQANnational stream quality network. They found copper concentrations about 3-fold lower, cadmiumconcentrations 30-fold lower, lead concentrations 100-fold lower and zinc concentrations almost 20-fold lower by using appropriate low level sampling and analysis techniques.

Significant regulatory and economic issues arise as a consequence of results that are biased high. Asummary of the results of analyses for waters of New York Harbour are presented in Table 1 (BattelleOcean Sciences, 1991).

C o p p e r

33

The harbour had been routinely sampled and analyzed by the New York City Department ofEnvironmental Protection. Of the six metals studied, for 1987 only for cadmium was the annualaverage below the value of the Water Quality Criteria. The annual average copper concentrationexceeded the average more than four-fold. In the following year, a series of cruises was carried out bythe U.S. EPA on the Research Vessel Anderson. Sampling and analysis was carried out by Battellescientists using low level sampling and analysis techniques. The concentrations of all six elementsdecreased greatly, as much as 40-fold in the case of cadmium, when the 1988 results are compared tothose of 1987. The copper concentrations of the earlier data are, on average, three-fold greater thanare the results that were obtained using the “clean” techniques. As a result of evaluation of the newerdata, the potential ecological impacts were reassessed and an unnecessary, costly upgrade of sewagetreatment plants discharging to the lower Hudson River and the New York Harbour was not instituted.

Table 1: Water Quality Criteria and annual average concentrations using conventional sampling andanalysis methods (1987) and low level sampling and analysis methods (1988) for New York Harbour(Battelle Ocean Sciences, 1991).

Metal

Water Quality Criteria1987µg/L

NYCDEP1987µg/L

Anderson1988µg/L

Cadmium 9.3 4.3 0.11

Copper 2.9 13 4.6

Lead 8.6 70 3.0

Mercury 0.025 0.2 0.015

Nickel 8.3 15.6 2.6

Zinc 86 264 10

The use of low level sampling and analysis methods may also be important for the evaluation ofindustrial discharges. Figure 1 shows the results for copper analyses of paired samples collected foran industrial waste discharge. Laboratory A had been doing the analyses of effluent samples, but didnot employ “clean techniques” in the laboratory. The results of Laboratory B indicated that the levelof copper in the discharge was not above the level specified in the discharge permit and that anexpensive new outfall was not necessary. The results of Laboratory A were as much as 9-fold greaterthan were those of laboratory B.

2. Chemical speciation

Frequently analyses of chemical species, rather than the total elemental concentration, are required.Both bioavailability and sorption of metals are strongly dependent on the metal species that arepresent (Allen, 1993). Speciation of trace metals involves determination of the physical and chemicalforms of the metal. This includes the evaluation of free metal ions, inorganic and organic complexes,and organometallic compounds. Although strictly not speciation, classification by size, includingfiltration is often considered speciation. Commonly speciation studies incorporate voltammetric andpotentiometric electrochemical methods and chromatography. Batley (1989) has reviewed many ofthe available methods.

Among the most important speciation measurements are determination of free metal ion andassessment of the interaction of metals with organic matter. Free metal ions can be quantified withion selective electrodes; the copper ion selective electrode can respond to free copper ionconcentrations as low as pCu = 19 (Avdeef et al., 1983). Samples are frequently titrated with metal,and the titration is monitored electroanalytically (Batley, 1989).

C o p p e r

34

Voltammetric techniques, particularly anodic and cathodic stripping voltammetry, have commonlybeen applied. The methods provide differentiation of weak from stronger ligands and the results ofthese titrations are usually expressed in terms of the complexation capacity of the sample (Neubeckerand Allen, 1983).

Cd2+, Cu2+, Pb2+ and other metals can form inner and outer sphere complexes with a number ofinorganic ligands, such as OH-, HCO3-, NH3, and organic ligands, such as oxalic acid and EDTA(Stumm and Morgan 1981; Pankow 1991). Stability constants for chemically defined ligands areavailable in a number of data bases. A particularly good source is the NIST Critical StabilityConstants of Metal Complexes Database (U.S. Department of Commerce, 1993). Stability constantsfor these reactions are well-known and the distribution of species can thus be easily computed usingthermodynamic principals.

201510500

4

8

12

16

20

Laboratory B

45° line

Figure 1: Paired samples of copper in industrial effluent. Laboratory A did not, and Laboratory Bdid, use “clean techniques.” Results in µg Cu/L. From Skrabal and Allen (unpub., 1993).

The soluble metal will be present as the free aquo ion and as metal contained in inorganic and organiccomplexes. The formation of complexes of a divalent metal ion M2+ with monodentate ligands isexpressed by the reaction

j=O=+ + åi− ↔ ji O=− å=( )+ (1)

If the ionic strength is low, the equilibrium is

β ML n=

ML2− n( )+

[ ]

M2+

[ ] L−[ ]

n

(2)

C o p p e r

35

where βMLn is the overall formation constant for the complex and the brackets indicate theconcentrations of the enclosed species. Typically, a number of complexes are present. Then the totalconcentration of metal, CM, is equal to the concentration of the free metal ion, plus the concentrationof metal contained in all complexed species

C M = M2+

[ ] 1 + β ML nL−

[ ]n

���

�� (3)

For ligands that are not monodentate, and for polynuclear reactions involving more than one metalion, similar reactions and stability constants can be formulated.

Chemical equilibrium computer programs are useful for computing the distribution of species insamples containing defined total concentrations of metal and ligands, if appropriate stability constantsare available (Nordstrom et al., 1979). Commonly used programs include MINTEQA2 (Allison et al.,1991) and MINEQL+ (Schecher and McAvoy 1992).

The description of metal complexation with natural organic matter (NOM) is much more complicated.NOM is an unresolvable mixture of a very large number of compounds varying in their properties,including their ability to bind metal ions. Several approaches have been proposed for the modeling ofmetal complexation by NOM and humic substances. These include gaussian distribution models(Perdue and Lytle, 1983) and multiple discrete sites (Fish et al., 1986). Recently, Tipping (1994) haspresented a model, using 5 sites, that is able to relatively accurately predict metal and proton bindingto naturally occurring organic matter.

3. Copper bioavailability and toxicity

The toxicity of copper and other metals is profoundly affected by their chemical form (see reviews byHodson et al., Luoma, 1983; O’Donnel et al., 1985; Flemming and Trevors, 1989). Steemann-Nielsenand Wium-Andersen (1970) recognized that free copper ions, at the level of copper found in naturalwaters, are very toxic to algae. They reported that the predominant forms of copper are not freecopper ions, but are organic complexes and they reported these to be “not poisonous to algae”.

In recently upwelled water, phytoplankton growth was limited and could be enhanced by the additionof a chelator. The understanding of the relationship between speciation and toxicity was greatlyenhanced by the work of Sunda and Guillard (1976) who determined copper ion activity using acopper ion selective electrode and found that the algal growth rate was related to the free copper ionactivity and not to the total copper concentration, which they had varied independently.

The amount of copper added to natural waters that is required to produce a given biological response,for example toxicity or reduction of growth, is different for different waters. An example is shown inFigure 2 for the growth of the alga Selanastrum capricornutum in waters collected in three locationsnear Chicago, Illinois. Based on total added copper, the amount required to produce a 50 percentreduction in growth rate varied by twenty-fold for the samples. This difference is interpreted as beingcaused by a difference in the bioavailability of the added copper (Benson et al., 1994).

Allen and Brisbin (1980) followed the labile copper concentration during a titration of a sample usinganodic stripping voltammetry and evaluated the sample’s conditional stability constant andcomplexation capacity. These values were used to compute the concentration of copper notcomplexed by organic matter. In a separate set of experiments, Selanastrum was grown in achemically defined medium and the concentration of copper ion required to inhibit the growth ratewas computed. They reported that they could predict the growth rate response of the algae in thenatural water samples. Verwiej et al. (1992) also found that growth inhibition for Scenedesmusquadricauda was highly correlated with copper detected by electrochemical methods and withcalculated free copper.

Because organism response can be correlated to the concentration of free copper ions, manyinvestigators have used bioassay methods to assess the concentration of free copper ions in naturalwaters (Gillespie and Vaccaro, 1978; Sunda and Gillespie, 1979; Allen et al., 1983). Hering et al.

C o p p e r

36

(1987) have shown that the free copper ion concentration, predicted from a bioassay, agrees with thatobtained by chemical measurement.

Some investigators have reported that chemical species other than free, ionic copper are toxic. Thishas recently been critically reviewed by Campbell (1995). Cowan et al. (1986) statistically analyzedavailable data in terms of the inorganic copper species that were toxic. Other investigators havefound that not all organic copper species are non-toxic. Florence and Stauber (1986) and Florence etal. (1992) reported that lipid soluble copper (II) complexes are highly toxic to the marine diatom,Nitzschia closterium. Guy and Kean (1980) and Borgmann and Ralph (1983) found that complexes ofcopper with synthetic ligands were toxic, but at reduced levels compared to free copper ion. BothMeador (1991) and Tubbing et al. (1994) reported that copper complexed with natural organic mattercontributed to toxicity.

A mechanistic approach to understanding the influence of solution-phase complexation of the toxicityof metals has been proposed by Pagenkopf (1983). He treated metal ion receptor sites on the gills offish as chemical entities that compete with ligands in the solution for metal ions such as copper. Thisallows the prediction of the effect of metal on the organism, based on equilibrium considerations suchas those predicted by computer programs such as MINTEQA2 or MINEQL+. This approach alsoallows prediction of the effects of Ca2+ and H+.

Figure 2 : Effect of added copper on the growth rate of alga Selanastrum capricornutum in samplesof three filtered, nutrient fortified waters.

The principal receptor site for toxic metals in freshwater fish are gills where Na+ and Ca2+ aretransported from the bulk water to the bloodstream by active, energy requiring “pumps”. The channelor carrier proteins associated with these pumps occur as specific, negatively charged ligands on thegill surface. Thus, the gills of freshwater fish have the two important physiological functions of gastransport (O2, CO2, NH3) and active uptake of ions (Na+, Ca2+) (Wood, 1992). Playle and co-workers(1992, 1993a and b) and MacRae et al. (1996) have treated the specific receptor sites on the gill ascompetitive ligands for the binding of copper and other metals. They have determined conditionalstability constants and site densities for these surface complexation reactions through competitivebinding experiments with ligands whose complexation constants are known. MacRae et al. (1996)showed that there is a strong relationship between the extent of saturation of the gill receptor sites and

C o p p e r

37

mortality. Experiments by these research groups have shown that gills indeed do compete for metalswith natural DOM.

An important conclusion from the gill complexation model is that given species should not be treatedas being bioavailable and other species as being not bioavailable. Rather, the presence of the gillcauses chemical re-equilibration in the system. It is the degree to which the gill complexation sitesare occupied by metal that determines whether toxicity will occur.

4. Water quality criteria and standards

The U.S. Environmental Protection Agency’s Water Quality Criteria for Copper (U.S. EPA, 1985)forms the basis for the standards for the protection of aquatic life that are promulgated by most states.The freshwater criteria for a four-day average concentration (in µg/L) of copper that is not to beexceeded more than once every three years on the average is given by the expressione(0.8545[ln(hardness)]-1.465). The acute criteria value (in µg/L), based on a one-hour averageconcentration should not exceed the numerical value e(0.9422[ln(hardness)]-1.464) more than once everythree years on the average. For hardnesses of 50, 100 and 200 mg/L as CaCO3 the four-day averageconcentrations of copper are 6.5, 12 and 21 µg/L, respectively, and the one-hour averageconcentrations are 9.2, 18, and 34 µg/L.

These criteria consider hardness as the only water quality parameter that is to be used in modificationof the numeric values. Erickson et al. (1996) determined the effects of various water chemistryparameters on the toxicity of copper to larval fathead minnows. Based on total copper concentrations,they found that increased pH, hardness, dissolved organic matter and suspended solids each causedtoxicity to decrease, but alkalinity had no observed effect. The effect of pH on toxicity was greaterthan that of hardness. They also determined cupric ion activity and found that the toxicity variedsignificantly when expressed on the basis of cupric ion activity, sometimes more than when thetoxicity was expressed on the basis of total copper. This study indicates that a number of waterquality parameters should be considered in setting criteria.

A single numerical value, even when modified by water hardness, does not provide the predictabilityof aquatic life effects that are desired for national criteria. Therefore, provision is made for site-specific modification of the criteria. The Water Effects Ratio (WER) has been recommended toprovide site-specific modified criteria (Stephan et al., 1985; U.S. EPA, 1992, 1994; Prothro, 1993). Toestablish a WER, toxicity tests are conducted in a site water and in a reference (“laboratory”) water.Reference water tests are used as surrogates for the laboratory tests that were used to derive nationalcriteria. The ratio of the toxicities (WER) is used as a multiplier to adjust the National Water QualityCriteria (NWQC) to account for differences in bioavailability, as measured by toxicity tests, thatwould be applicable to that site. For example:

Site - Specific WQC = NWQC × WER = NWQC × site - water LC50 (4)

reference - water LC50

Carlson et al. (1986) applied this procedure in a a series of stations in a stream receiving industrialand municipal treatment plant effluents and and found mean WERs of 3.9 to 7.0, reflective of reducedbioavailability of copper.

Allen and Hansen (1996) have analyzed the WER procedure in terms of the change in speciation thatoccurs in a sample as metal is added to a sample. They indicated that the fraction of metal that ispresent in forms having reduced bioavailability decreases as the total concentration of metalincreases.

This implies that the WER for sensitive organisms will be greater for more sensitive organisms than itwill be for less sensitive ones. This is in agreement with the information that has been presented byBrungs et al. (1991).

Allen and Hansen (1996) recommended that new Water Quality Criteria based on bioavailable metalbe developed. Such criteria would have universal applicability and would obviate the need for such

C o p p e r

38

site-specific modifications as WERs. A recent SETAC Pellston Conference (Bergman and Dorward-King, 1997) recommended that the gill complexation model be used as the basis for a WQC thatincorporated bioavailability by coupling biological site of action and aquatic speciation. Theyrecommended that Tipping’s WHAM model be used to compute chemical speciation in a receivingwater. This model should be coupled with MINTEQA2 or another speciation code that incorporatethe necessary constants for binding of metal at the gill.

5. Aquatic sediments

Metals and anthropogenic organic compounds are often present at elevated levels in sediments.Evaluation of sediment quality is frequently ascertained through bioassays. However, if toxicity isfound, the cause must be ascertained to ensure proper disposal or treatment to prevent furthercontaminant input and to allocate responsibility. This creates a difficult situation because highconcentrations of metals do not necessarily lead to toxicity. Figure 3 shows the results of a number oftoxicity tests in which cadmium, copper, nickel or zinc, or cadmium and nickel were added tosediments. When the concentration of metal is expressed as µmol/g dry weight of sediment there isno relationship between the concentration of added metal and the mortality of organisms. Non-toxicsediments were found to contain as much as 3 orders of magnitude more metal than did somesediments not exhibiting toxicity.

Figure 3 : Toxicity of metals in sediments. Metal concentrations expressed on a dry weightnormalized basis. Data courtesy of Dr. Dominic Di Toro, Manhattan College.

For example, Di Toro et al. (1990) added cadmium to samples of sediment that did not exhibittoxicity. The amount of cadmium that was required to be added before toxicity was observed differedby more than an order of magnitude for the two sediments and was greater than what isconventionally considered to represent a contaminated sediment. They found that the added cadmiumdisplaced iron in FeS to form CdS thus rendering it non-toxic. No toxicity was found unless theamount of added cadmium exceeded the concentration of available sulfide when both are expressedon a molar basis.

Analytical methods for the measurement of acid volatile sulfide (AVS) and the concentration ofsimultaneously extracted metals (SEM) have been described by Allen et al. (1993). Sulfide is evolvedby the addition of cold, dilute acid. It is trapped and quantified to provide the AVS value. The

C o p p e r

39

dissolved concentrations of the potentially toxic metal that can displace iron in FeS, which arecadmium, copper, lead, nickel and zinc, are determined in the acid.

The SEM value is the sum of these metal concentrations on a µmol/g dry weight of sediment basis.The SEM/AVS ratio is useful in prediction of those sediments for which toxicity will not beexhibited. Ratios less than one have not been found to be toxic to organisms. That is, there is nottoxicity if the amount of sulfide available to bind metals exceeds the concentration of metals. Itshould be noted that it is the absence of toxicity that is predicted. No prediction of toxicity isprovided by this procedure. This is shown in Figure 4 for the same data set shown in Figure 3. Notoxicity is seen for any sediments having SEM/AVS ratios less than 1.

Figure 4 : Toxicity of metals in sediments. Metal concentrations expressed on a SEM/AVSnormalized basis. Data are the same as those plotted in Figure 3. Data courtesy of Dr. Dominic DiToro, Manhattan College.

Ankley et al. (1993) found that there appeared to be copper binding in excess of the amount of sulfidepresent in sediment. This can be clearly seen in Figure 4 which demonstrates that, for somesediments, no toxicity was observed until the SEM/AVS ratio exceeded 5. Mahony et al. (1996) havedemonstrated that this additional binding phase for cadmium, copper and lead is the organic mattercontained in the sediments. Inclusion of this potential for binding of metals in addition to that of theAVS that is in excess of the SEM (i.e., SEM-AVS) should provide a good estimate of theconcentration of metals that could be present in a sediment before toxicity will be observed.

6. Conclusions

This review of the literature provides two important conclusions regarding the assessment of metalsand potential metal toxicity in aquatic systems:

Proper precautions must be taken in the sampling and analysis of environmental samples toensure accurate results. Much of the existing data is of questionable quality and these resultsare biased high.

Assessment of potential impacts of copper and other metals in water and sediment cannot bejudged on the basis of the total concentration of metals. Knowledge of metal speciation and ofthe total system chemistry is essential.

C o p p e r

40

It is possible to develop Water Quality Criteria and Sediment Quality Criteria that are predictive ofthe effects of metals on aquatic life. Such criteria, based on sound chemical and physiological bases,would have broad applicability. Present criteria and evaluations based on observations of effectssuffer from not providing the evaluator a technical basis for knowing their applicability to a newsituation. Site-specific modification is a less technically acceptable process than is the developmentof the proposed technically sound evaluative tools.

References

Allen, H.E. (1993), ‘The significance of trace metal speciation for water, sediment and soil qualitystandards’, Sci. Total Environ., vol. 134, Supplement Part 1, p. 23.

Allen, H.E., Blatchley, C. and Brisbin, T.D. (1983), ‘An algal assay method for determination ofcopper complexation capacity of natural waters’, Bull. Environ. Contam. Toxicol., vol. 30, p. 448.

Allen, H.E. and Brisbin, T.D. (1980), ‘Prediction of bioavailability of copper in natural waters’ Thal.Jugosl., vol. 16, p. 331.

Allen, H.E. G. Fu and Deng, B (1993), ‘Determination of acid-volatile sulfide (AVS) andsimultaneously extracted metals (SEM) for the estimation of potential toxicity in aquatic sediments’,Environ. Toxicol. Chem., vol. 12, p.1441.

Allen, H.E. and Hansen, D.J. (1996), ‘The importance of trace metal speciation to water qualitycriteria’, J. Water Environ. Res., vol. 68, p. 42.

Allison, J.D., Brown, D.S. and Novo-Gradac, K.J. (1991), ‘MINTEQA2/PRODEFA2, a geochemicalassessment model for environmental systems: version 3.0 user’s manual’. EPA/600/3-91/021. U.S.Environmental Protection Agency, Washington, DC.

Ankley, G.T., Matson,V., Leonard, E., West, C. and Bennett, J. (1993), ‘Predicting the acute toxicityof copper in freshwater sediments: evaluation of the role of acid volatile sulfide’, Environ. Toxicol.Chem., vol. 12, p. 312.

Avdeef, A., Zalironsky, J. and Stuting, H. (1983), ‘Calibration of copper ion selective electroderesponse to pCu 19’, Anal. Chem., vol. 55, p. 298.

Batley, G.E. (ed). (1989), ‘Trace Element Speciation: Analytical Methods and Problems’. CRCPress, Boca Raton, FL.

Battelle Ocean Sciences. (1991), ‘Results of the Ambient and Municipal Sample InterlaboratoryComparison Study in Ambient Waters and Discharges to New York/New Jersey Harbour’. Duxsbury,MA.

Benson, W.H., Alberts,J., Allen, H.E., Hunt, C.D.and Newman, M.C. (1994), ‘Synopsis ofdiscussion session on bioavailability of inorganic elements’ In A Mechanistic Understanding ofBioavailability: Physical, Chemical, and Biological Interactions. Chapter 2. J.L. Hamelink, W.H.Benson, H.L. Bergman and P.F. Landrum (eds) CRC Press, Boca Raton, FL.

Bergman, II.L and Dorward-King, E.J. (Eds). (1997), ‘Reassessment of Metals Criteria for AquaticLife Protection: Priorities for Research and Implementation’. Society for Environmental Toxicologyand Chemistry, Pensacola, FL.

Borgmann, U. and K.M. Ralph, K.M. (1983), ‘Complexation and toxicity of copper and the freemetal bioassay technique’, Water Res., vol. 17. p. 1697.

Bruland, K.W. (1983), ‘Trace elements in seawater.’, Chem. Oceanogr., vol. 8, p. 157.

Brungs, W.A. (1991), ‘Synopsis of Water-Effect Ratios for Heavy Metals as Derived for Site-Specific Water Quality Criteria’. U.S. EPA Contract 68-CO-0070.

Campbell, P.G.C. (1995) ‘Interactions between trace metals and aquatic organisms: A critique of thefree-ion activity model.’ In Metal Speciation and Bioavailability in Aquatic Systems.Tessier, A. andTurner, D.R. (eds). Wiley, New York.

C o p p e r

41

Carlson, A.R., Nelson, H. and Hammermeister, D. (1986), ‘Development and validation of site-specific water quality criteria for copper’, Environ. Toxicol. Chem., vol. 5, p. 997.

Cowan, C.E., Jenne, E.A. and Kinnison, R.R. (1986), ‘Methodology for determining the relationshipbetween toxicity and the aqueous speciation of a metal’. In Aquatic Toxicity and EnvironmentalFate: Ninth Volume. ASTM STP 921. T.M. Poston and R. Purdy, eds. American Society for Testingand Materials, Philadelphia p. 463.

Di Toro, D.M., Mahony, J.D., Hansen, D.J., Scott, K.J., Hicks, M.B., Mayr, S.M. and Redmond, M.S.(1990), ‘Toxicity of cadmium in sediments: The role of acid volatile sulfide’, Environ. Toxicol.Chem. vol. 9, p. 1487.

Erickson, R.J., Benoit, D.A., Mattson, V.R., Nelson, Jr, H.P. and Leonard, E.N. (1996), ‘The effectsof water chemistry on the toxicity of copper to fathead minnows’, Environ. Toxicol. Chem., vol. 15,p. 181.

Fish, W., Dzombak, D.A. and Morel, F.M.M. (1986), ‘Metal-humate interactions. 1. Application andcomparison of models’, Environ. Sci. Technol., vol. 20, p. 676.

Flemming, C.A. and Trevors, J.T. (1989), ‘Copper toxicity and chemistry in the environment: Areview’, Water, Air, Soil Pollut. , vol. 44, pp. 143.

Florence, T.M., Powell, H.K.J., Stauber, J.L. and R.M. Town. (1992), ‘Toxicity of lipid-solublecopper (II) complexes to the marine diatom Nitzschia closterium: Amelioration by humic substances’,Water Res., vol. 26, p. 1187.

Florence, T.M. and Stauber J.L. (1986), ‘Toxicity of copper complexes to the marine diatom,Nitzschia closterium’, Aquat. Toxicol., vol. 8, p. 11.

Gillespie, P.A. and. Vaccaro, R.F. (1978), ‘A bacterial bioassay for measuring the copper-chelationcapacity of seawater’, Limnol. Oceanogr., vol. 23, p. 543.

Guy, R.D. and Kean, A.R.. (1980), ‘Algae as a chemical speciation monitor - I. A comparison ofalgal growth and computer calculated speciation’, Water Res., vol. 14, p. 891.

Hering, J.G., Sunda, W.G., Ferguson, R.L. and F.M.M. Morel, F.M.M. (1987), ‘A field comparisonof two methods for the determination of copper complexation: bacterial bioassay and fixed-potentialamperometry; Mar. Chem., vol. 20, p. 299.

Hodson, P.V., Borgmann, U. and Shear, H. (1979), ‘Toxicity of Copper to Aquatic Biota’. InCopper in the Environment. Part II. Health Effects. J.O. Nriagu, ed. Wiley-Interscience, New York.

Luoma, S.N. (1983), ‘Bioavailability of trace metals to aquatic organisms - a review’, Sci. TotalEnviron., vol. 28, p. 1.

MacRae, R.K., D.E. Smith, D.E., Swoboda-Colberg, N., Meyer, J.S. and Bergman, H.L. (1996),‘Copper binding affinity of rainbow trout (Oncorhynchus mykiss) and brook trout (Salvelinusfontinalis) gills’, Environ. Toxicol. Chem., (submitted).

Mahony, J.D., Di Toro, D.M., Gonzalez, A.M., Curto, M., Dilg, M., De Rosa, L.D. and L.A.Sparrow, L.A. (1996), ‘Partitioning of metals to sediment organic carbon’, Environ. Toxicol. Chem.,vol. 15, p. 2187.

Meador, J.P. (1991), ‘The interaction of pH, dissolved organic carbon, and total copper in thedetermination of ionic copper and toxicity’, Aquat. Toxicol., vol. 19, p. 13.

Neubecker, T.A. and Allen, H.E. (1983), ‘The measurement of complexation capacity andconditional stability constants for ligands in natural waters--a review’, Water Res., vol. 17, p. 1.

Nordstrom, D.K., Plummer, L.N., Wigley, T.M.L., Wolery, T.J., Ball,W., Jenne, E.A.,. Bassett,R.L.,Crerar, D.A., Florence, T.M., Fritz, B., Hoffman, M., Holdren, Jr, G.R., Lafon, G.M.,Mattigod,S.V., McDuff, R.E., Morel, F., Reddy, M.M., Sposito, G. and Thraikill, J. (1979), ‘Comparison ofComputerized Chemical Models for Equilibrium Calculations in Aqueous Systems’. In Chemical

C o p p e r

42

Modeling in Aqueous Systems: Speciation, Sorption, Solubility and Kinetics. E.A. Jenne, ed. Symp.Ser. 93. American Chemical Society, Washington, DC.

Nriagu, J.O., Larson, G., Wong, H.K.T. and Azcue, J.M. (1993), ‘A protocol for minimizingcontamination in the analysis of trace metals in Great Lakes waters’, J. Great Lakes Res., vol. 19,p.175.

O’Donnel, J.R., Kaplan, B.M. and Allen, H.E. (1985), ‘Bioavailability of Trace Metals in NaturalWaters’. In Aquatic Toxicology and Hazard Assessment: Seventh Symposium, American Society forTesting and Materials, Philadelphia, PA.

Pagenkopf, G.K. (1983), ‘Gill surface interaction model for trace-metal toxicity to fishes: Role ofcomplexation, pH and water hardness’, Environ. Sci. Technol., vol. 17, p. 347.

Pankow, J. (1991), ‘Aquatic Chemistry Concepts’. Lewis Publ., Chelsea, MI.

Patterson, C.C. and D.M. Settle. (1976), ‘The reduction of orders of magnitude errors in leadanalyses of biological materials and natural waters by evaluating and controlling the extent andsources of industrial lead contamination introduced during sample collecting and analysis’. InAccuracy in Trace Analysis: Sampling, Sample Handling, Analysis. P.D. LaFleur (ed). Nat. Bur.Standards. NBS Spec. Pub. 422: 321.

Perdue, E.M. and Lytle, C.R. (1983), ‘Distribution model for binding of protons and metal ions byhumic substances’, Environ. Sci. Technol., vol. 17, p.654.

Playle, R.C., Dixon, D.G. and Burnison, D.G. (1993a), ‘Copper and cadmium binding to fish gills:Modification by dissolved organic carbon and synthetic ligands’, Can. J. Fish. Aquat. Sci., vol. 50, p.2667.

Playle, R.C., Dixon. D.G. and Burnison, D.G. (1993b), ‘Copper and cadmium binding to fish gills:Estimates of metal-gill stability constants and modelling of metal accumulation’, Can. J. Fish. Aquat.Sci., vol. 50, p. 2678.

Playle, R.C., Gensemer, R.W. and Dixon, D.G. (1992), ‘Copper accumulation on gills of fatheadminnows: Influence of water hardness, complexation and pH of the gill microenvironment’, Environ.Toxicol. Chem., vol. 1, p. 381.

Prothro, M.G. (1993), ‘Office of Water Policy and Technical Guidance on Interpretation andImplementation of Aquatic Life Metals Criteria’. U.S. Environmental Protection Agency,Washington, DC.

Schecher, W.D. and McAvoy, D.C. (1992), ‘MINEQL+: A software environment for chemicalequilibrium modelling’, Computers, Environment and Urban Systems, vol. 16, p. 65.

Skrabal, S. A. and Allen, H.E. (1993), Unpublished results.

Steemann-Nielsen, E. and Wium-Andersen, S. (1970), ‘Copper as poison in the sea and infreshwater’, Mar. Biol., vol. 6, p. 93.

Stephan, C.E., Mount, D.I., Hansen, D.J., Gentile, J.H., Chapman, G.A. and W.A. Brungs. (1985),‘Guidelines for Deriving Numerical National Water Quality Criteria for the Protection of AquaticOrganisms and Their Uses’. U.S. EPA, Office of Research and Development. PB85-227049. NTIS,Springfield, VA.

Stumm, W. and Morgan, J.J. (1981), Aquatic Chemistry. Wiley, New York.

Sunda, W. and R.R.L. Guillard, R.R.L. (1976), ‘The relationship between cupric ion activity and thethe toxicity of copper to phytoplankton’, J. Mar. Res., vol. 34, p. 511.

Sunda, W. and Gillespie, P.A. (1979), ‘The responses of a marine bacterium to cupric ion and its useto estimate cupric ion activity’, J. Mar. Res., vol. 37. p. 761.

C o p p e r

43

Tipping, E. (1994), ‘WHAM - A chemical equilibrium model and computer code for waters,sediments, and soils incorporating a discrete site/electrostatic model of ion-binding by humicsubstances’, Computers Geosci., vol. 20, p. 973.

Tubbing, D.M.J., Admiraal,W., Cleven, R.F.M.J., Iqbal, M., van de Meent, D. and Verweij, W.(1994), ‘The contribution of complexed copper to the metabolic inhibition of algae and bacteria insynthetic media and river water’, Water Res., vol. 28, p. 37.

U.S. Department of Commerce. (1993), ‘NIST Critical Stability Constants of Metal ComplexesDatabase’. Version 1.0 Washington, D.C.

U.S. EPA. (1985), ‘Ambient Water Quality Criteria for Copper - 1984’. Office of WaterRegulations and Standards, Washington, D.C.

U.S. EPA. (1992), ‘Interim Guidance on Interpretation and Implementation of Aquatic Life Criteriafor Metals’. Washington, D.C.

U.S. EPA. (1994), ‘Interim Guidance on Determination and Use of Water-Effect Ratios for Metals’.EPA 823-B-94-001. Washington, D.C.

U.S. EPA. (1995), Method 1669: Sampling Ambient Water for Trace Metals at EPA Water QualityCriteria Levels.

Verweij, Glazewski, R and De Haan, H. (1992), ‘Speciation of copper in relation to itsbioavailability’, Chem. Speciation Bioavail., vol. 4, p. 43.

Windom, H.L., Byrd, J.T., Smith, Jr, R.G. and Huan, F. (1991), ‘Inadequacy of NASQAN data forassessing metal trends in the nation’s rivers’, Environ. Sci. Technol., vol. 25, p. 1137.

Wood, C.M. (1992), ‘Flux measurements as indices of H+and metal effects of freshwater fish’,Aquatic Toxicology, vol. 22, pp. 239-264.

Evaluation of copper guideline values for drinking-waterDr D James FitzgeraldEnvironmental Health Branch, Public and Environmental Health ServiceSouth Australian Health Commission, Adelaide, South Australia

1. Introduction

Since 1925, when the US Public Health Service was the first regulatory agency to set a guidelinevalue for copper in drinking-water, various guidelines have been established based on aesthetic andhealth considerations. With time, as the essential nature of copper in the diet as well as the toxicity ofcopper became better understood, it became clear that there is a narrow margin between copper’sessentiality and its toxicity (Uauy & Olivares, 1996). How this margin can be better defined and aguideline set is currently an important issue for water regulatory authorities. This issue also becomesimportant because of the economic ramifications for the copper-pipe industry if a guideline isstringent and difficult to meet.

This chapter will examine how present guideline values have been derived, and will evaluate thevalidity of these derivations. Brief discussion only will be given to the aesthetics-based guidelinevalues before focussing on guideline values based upon health considerations.

2. Historical perspective

A compilation of some historical information on the guideline values for copper in drinking-water isshown in Table 1.

C o p p e r

44

2.1 Aesthetics-based guideline values

Due principally to consumer complaints of copper’s staining of laundry and sanitary ware, theaesthetics-based guideline value is now generally 1.0 mg/L. However, the set value in Europe is 3.0mg/L which should not be exceeded during 12 hours of stagnation (the Netherlands operates on 16hours). At one stage in the USA, the aesthetics-based value was also 3.0 mg/L but this was reduced tothe present-day 1.0 mg/L in 1942 on the basis of taste problems. In a range of experimental studies, ithas been demonstrated that copper salts dissolved in water can be tasted at levels of 1-2 mg Cu/L(Béguin-Bruhin et al., 1983; Cohen et al., 1960; Uauy, pers. comm.).

2.2 Health-based guideline values

The US Environmental Protection Agency (US-EPA) was the first regulatory agency to set a health-based guideline value for copper in potable water. Their 1985 proposal of 1.3 mg/L has since beenpromulgated and forms part of the Lead-Copper Rule (USEPA, 1988, 1991a). In 1993, the WorldHealth Organization (WHO) set a provisional guideline value of 2.0 mg/L, a figure which WHO hasrecently re-proposed (IPCS, 1996a). This same value has been adopted in Australia and New Zealand(NHMRC, 1996) and is being proposed in the European Commission (EC, 1996). Thus there is awider move to adopt health-based guidelines for copper in drinking-water.

Table 1: Copper guideline values for potable water

Agency/Regulator and Year Value

Aesthetics-based guidelines

US Public Health Service, 1925 0.2 mg/L

US Public Health Service, 1942 3.0 mg/L

US Public Health Service, 1962 1.0 mg/L

WHO, 1958, 1996 1.0 mg/L

EEC, 1980 3.0 mg/L

NHMRC - Australia, 1987a 1.0 mg/L

Health-based guidelines

US-EPA, 1985, 1988, 1991a 1.3 mg/L

WHO, 1993, 1996 2.0 mg/L

NHMRC - Australia, 1996b 2.0 mg/L

EC, proposed 2.0 mg/L

Modified from De Zuane (1990) and Fitzgerald (1997), with references therein and in Fitzgerald (1995),IPCS (1996a). aNHMRC (1987); bNHMRC (1996)

3. Derivation of health-based guideline values

3.1 US Environmental Protection Agency

The approach of the US-EPA has been to consider reports of acute effects of excess copper ingestion(nausea, vomiting, abdominal pains, diarrhoea). Several case reports of high copper levels inbeverages and of suicide attempts with copper sulphate were evaluated (USEPA, 1988). The studywas chosen in which the lowest acute oral dose was observed, this being a report of Dr John Wylliepublished in 1957. In his now much-discussed paper, Dr Wyllie recounted an incident involving asmall number of female nurses who had consumed various amounts of an alcohol cocktail prepared ina copper-containing vessel. Presumably, some of the copper had leached into the drink and resulted inacute effects in most of the nurses. A remake of the cocktail and subsequent chemical analysis formetals provided information on the copper exposures most likely experienced by the cases. Wyllie

C o p p e r

45

estimated that the smallest amount of copper ingested was 5.3 mg (Wyllie, 1957). The US-EPAdivided this figure by an uncertainty factor of 2, then again by 2 litres for average daily waterconsumption, to derive their drinking-water copper guideline value (Maximum Contaminant LevelGoal) of 1.3 mg/L (USEPA, 1988).

Close scrutiny of the brief Wyllie paper reveals a paucity of information, particularly regarding detailsof individuals’ intakes and symptoms. This, together with a number of ambiguities including doubtover the Lowest-Observed-Adverse-Effect Level, has resulted in the conclusion by some that thestudy is inadequate for the derivation of a guideline value (Fitzgerald, 1995, 1996a, 1996b; Fewtrelland Kay, 1995). Notwithstanding the recognised need for better data, 1.3 mg/L remains the USCongress-mandated guideline for copper in drinking-water under the Safe Drinking Water Act(USEPA, 1991a, 1991b).

A recent extensive survey of water utilities covering about half the population in the USA has shownthat probably no more than 1% of the population receives water containing copper at a mean valuegreater than 1.3 mg/L (G Lagos, pers. comm.). No mean value greater than 2 mg/L was recorded.

3.2 World Health Organisation

In contrast to the approach taken by the US-EPA, WHO did not consider acute effects of copper butinstead evaluated daily intakes and some chronic exposure circumstances. Firstly, a provisionalmaximum tolerable daily intake (PMTDI) was set at 0.5 mg/kg body weight, based on the belief that a10-fold excess of the daily intake of copper considered normal at the time, ie. 10 times 2-3 mg/day,would not elicit any health problems (Becking, 1996; WHO, 1967, 1982). Multiplication of thePMTDI by 60 kg and 0.1 for proportion of intake attributable to water, then division by 2L waterintake per day, yielded the guideline value of 1.5 mg/L; this has been rounded to 2 mg/L (Mercier,1996). WHO recognise the lack of useful data for setting a guideline, and so add the caveat that thisguideline value is provisional. [The process of establishing a guideline value for an essential elementusing dietary intake, multiplication of that value by an arbitrary number and attributing someproportion of intake to water is questionable.]

Secondly, WHO considered the results of an industry-conducted chronic copper-gluconate ingestionstudy in dogs, claiming that the above PMTDI for humans was based on the experiment’s No-Observed-Adverse-Effect Level (NOAEL) (WHO, 1982; IPCS, 1996a). This, however, is not thecase, but the erroneously reported and transcribed NOAEL (5 mgCu/kg/day), together with a safetyfactor of 10, seemed to corroborate the PMTDI of 0.5 mg/kg (Fitzgerald, 1995; Mercier, 1996). Wenow know that the supposed 5 mgCu/kg was actually 15 mg copper-gluconate/kg, being 2.1 mgCu/kg(Fitzgerald, 1995; Shanaman et al., 1972). In the IPCS Task Group (June, 1996) which met for theevaluation of health aspects of copper, toxicologists assessed this dog study and concluded that theliver enzyme effects, noted only in 2 of 12 dogs exposed to the highest dose (8.4 mgCu/kg/day), werereversible and not toxicologically significant (IPCS, 1996b).

Thirdly, WHO consider the liver pathologies associated with chronic copper exposure, and areespecially mindful of “.... concern regarding the possible involvement of copper from drinking-waterin early childhood liver cirrhosis in bottle-fed infants, .....” (WHO, 1993, p46). With no reference toquantitative or epidemiological data, the text continues: “.... a concentration of 2 mg/litre should alsocontain a sufficient margin of safety for bottle-fed infants, because their copper intake from othersources is usually low” (WHO, 1993).

Mention is made by WHO of acute gastric effects of copper, though there is some confusion over this.One document states: “Acute gastric irritation may be observed in some individuals at concentrationsin drinking-water above 3 mg/litre” (WHO, 1993, p46). A more recent publication states: “Theestimated concentration of copper(II) in drinking-water or beverages that can lead to symptoms ofthis type [ie. acute symptoms] is 30 mg/litre but may vary with the binding and chemical form ofcopper present....” (IPCS, 1996a, p222). It is likely that the former concentration of 3 mg/L is nearerthe mark for the threshold of gastric effects of copper in drinking-water (Spitalny et al., 1984;Knobeloch et al. 1994; Fitzgerald, 1996a; R Uauy, pers. comm.). For typical drinking-water, few

C o p p e r

46

components would be present that bind copper and reduce its gastric irritancy potential. In somebeverages, for example Tyrolean milk mixture prepared in copper vessels, copper’s irritancy isprevented even at levels of 30-60 mg/L (Müller et al., 1996; T Müller, pers. comm.). It is interestingthat Ramazzini in 1713 had noted the ameliorating properties of milk and whey in treatingoccupational disease in coppersmiths (Ramazzini, 1713).

3.3 Australia and New Zealand

The health-based drinking-water guideline value for copper in Australia and New Zealand has beenset at 2 mg/L. In deriving this value, essentially the WHO approach has been adopted; substituting 70kg for the average adult weight does not alter the outcome (NHMRC, 1996). Authorities in thesecountries have concluded that copper levels exceeding 2-3 mg/L will induce symptoms of acutetoxicity in some people. In South Australia, a Health Alert of 3 mgCu/L has been proposed (HAWQ,1994).

3.4 European Commission

Some water reticulation systems in Europe are comprised of lead piping, and consideration is beinggiven as to whether copper provides a safe alternative. The Scientific Advisory Committee of theEuropean Commission recently conducted an assessment of the evidence for the toxicity of copper indrinking-water. This included examination of WHO’s guideline development and of a range of animaland human studies. The main conclusions of this committee were: i) that animal data are insufficient,ii) that human experience suggests that 1-2 mg/L will be acceptable, and iii) that the proposed 2 mg/Lin the EC Draft Drinking Water Directive should be retained (EC, 1996).

4. Overall evaluation

Evaluation of the copper guidelines is an important process. A guideline that is too lax may beinsufficiently protective of human health, while one that is too conservative can impose heavycompliance costs. Already in the USA, hundreds of millions of dollars are spent on corrosion controlto reduce copper in water supplies to levels below 1.3 mg/L (G Lagos, pers. comm.). The need is fora scientifically-defensible guideline to ensure that large sums of money are not wasted.

Therefore, to progress the copper guideline debate, there is a definite requirement for further humandata since some of the key data employed for guideline derivation are clearly inadequate. There is aneed for prospective epidemiological studies that delineate copper exposure via drinking-water andacute effects of copper. However, such approaches will be plagued by the high background prevalenceof gastrointestinal symptoms in the community. Controlled studies with volunteers will also beextremely useful; at least one such project is in progress (R Uauy, pers. comm.). A betterunderstanding of the threshold for acute gastric effects of copper will contribute most significantly tothe honing of a health-based drinking-water guideline value for copper.

Acknowledgements

I thank Dr Ricardo Uauy, Professor Gustavo Lagos, Dr Thomas Müller and Dr Debdas Mukerjee fortheir sharing of information and unpublished data. Dr Andrew Langley is acknowledged for hiscomments on the manuscript.

References

Becking, G. (1996), In: Lagos GE & Cifuentes L.A (eds), Scientific Basis for the Regulation ofCopper in Potable Water. Catholic University of Chile, pp 64-67.

Béguin-Bruhin, Y., Escher, F., Solms, J. and Roth, H.R. (1983), ‘Threshold concentration of copperin drinking water’, Lebensmittel-Wissenschaft und Technologie, vol. 16, pp. 22-26.

C o p p e r

47

Cohen, J.M., Kamphake, L.J., Harris, E.K. and Woodward R.L (1960), ‘Taste thresholdconcentrations of metals in drinking water’, Journal of the American Water Works Association, vol.52, pp. 660-670.

De Zuane, J. (1990), Handbook of drinking water quality: standards and controls. Van NostrandReinhold, New York.

EC (1996), Opinion of the Scientific Advisory Committee concerning toxicologically acceptableparametric value for Copper in Drinking Water. Scientific Advisory Committee to examine thetoxicity and ecotoxicity of chemical compounds. European Commission, Brussels, Feb.20, 1996,CSTE/96/6/V.

Fewtrell, L and Kay, D (1995) ‘Copper in Drinking Water: An Appraisal of Health Effects andCurrent Standards’. Report of the Centre of Research into Environment and Health (CREH),University of Leeds, UK.

Fitzgerald, D.J (1995), ‘Copper guideline values for drinking water: Reviews in need of review?’,Regulatory Toxicology & Pharmacology., vol. 21, pp. 177-179.

Fitzgerald, D.J (1996a), ‘Copper regulatory level in drinking-water as proposed by Sidhu et al.’,Regulatory Toxicology & Pharmacology, vol. 23, pp. 173-175.

Fitzgerald, J (1996b), ‘A critical review of the copper standard in potable water’. In: Lagos GE &Cifuentes LA (eds.), Scientific Basis for the Regulation of Copper in Potable Water. CatholicUniversity of Chile, pp. 55-63, G1-G11.

Fitzgerald, J (1997), Safety guidelines for copper in water. In: Lonnerdal B and Uauy R (eds.),‘Essentiality and Toxicity of Copper’. American Journal of Clinical Nutrition Supplement, in press.

HAWQ (1994), ‘Health Alert for copper in drinking-water’. Governmental Standing Committee,Health Aspects of Water Quality. South Australian Health Commission, Adelaide.

IPCS (1996a), Guidelines for drinking-water quality, 2nd ed. Vol. 2. Health criteria and othersupporting information. International Programme on Chemical Safety, World Health Organization,Geneva.

IPCS (1996b), Environmental Health Criteria Document: Copper. International Programme onChemical Safety, World Health Organization, Geneva; in press.

Knobeloch, L., Ziarnik, M., Howard, J., Theis, B., Farmer, D., Anderson, H and Proctor, M. (1994),‘Gastrointestinal upsets associated with ingestion of copper-contaminated water’, EnvironmentalHealth Perspectives, vol.102, pp. 958-961.

Mercier, M. (1996), ‘Health risk assessment of chemicals with particular reference to copper indrinking water’. In: Lagos GE & Cifuentes LA (eds), Scientific Basis for the Regulation of Copper inPotable Water. Catholic University of Chile, pp 11-15, B1-B10.

Müller, T., Feichtinger, H., Berger, H. and Müller W (1996) Endemic Tyrolean infantile cirrhosis: anecogenetic disorder. Lancet 347, 877-880.

NHMRC (1987), Guidelines for drinking water quality in Australia. National Health & MedicalResearch Council, and Australian Water Resources Council. Australian Government PublishingService, Canberra.

NHMRC (1996), Australian drinking water guidelines. National Health & Medical ResearchCouncil, and Agricultural & Resource Management Council of Australia and New Zealand.Australian Government Publishing Service, Canberra.

Ramazzini B (1713), ‘De Morbis Artificum’. In: Wright WC, Diseases of Workers: The Latin text of1713. The University of Chicago Press, Chicago, 1940.

Shanaman, J.E., Wazeter, FX. and Goldenthal, E.I. (1972), ‘One year chronic oral toxicity of coppergluconate, W10219A, in Beagle dogs’. Research Report No. 955-0353. Warner-Lambert Res. Inst.,Morris Plains, New Jersey.

C o p p e r

48

Spitalny, K.C., Brondum, J., Vogt, R.L., Sergent, H.E. and Kappel S (1984), ‘Drinking-water-inducedcopper intoxication in a Vermont family, Pediatrics , vol. 74, pp. 1103-1106.

Uauy, R and Olivares, M (1996)(eds.), ‘Copper nutrition in humans: essentiality and toxicity’, TheAmerican Journal of Clinical Nutrition, vol. 63, pp. 791S-852S.

US EPA (1988), ‘Drinking Water Regulations: Maximum contaminant level goals and nationalprimary water regulations for lead and copper’, Federal Register , vol. 53, no.160, pp. 31516-31578.

US EPA (1991a), ‘Maximum contaminant level goals and national primary drinking waterregulations for lead and copper; final rule’, Federal Register, vol. 56, no.110, pp. 26460.

US EPA (1991b), ‘Drinking water regulations; maximum contaminant level goals and nationalprimary drinking water regulations for lead and copper’, Federal Register, vol. 56, no.135, pp.32112-32113.

WHO (1967), Specifications for the Identity and Purity of Food Additives and Their ToxicologicalEvaluation: Some Emulsifiers and Stabilizers and Certain Other Substances. Tenth Report of theJoint FAO/WHO Expert Committee on Food Additives. WHO Technical Report Series, No. 373.World Health Organization, Geneva.

WHO (1982), Evaluation of Certain Food Additives and Contaminants. Twenty-Sixth Report of theJoint FAO/WHO Expert Committee on Food Additives. WHO Technical Report Series, No. 683, pp31-32. World Health Organization, Geneva.

WHO (1993), Guidelines for Drinking-Water Quality, 2nd ed. Vol.1. Recommendations. WorldHealth Organization, Geneva.

Wyllie, J. (1957), ‘Copper poisoning at a cocktail party’, The American Journal of Public Healthvol. 47, p. 617.

Population exposure to copper in drinking waterGustavo LagosCatholic University of Chile

Abstract

Surface waters used for the production of drinking water contain a median level of 10 µg/L of copper.Values above this level (up to a few milligrams/litre) in drinking water reflect either the existence ofhigh copper content in the source water, due for instance to the presence of acidic soils, or theleaching of copper from plumbing fixtures and copper piping during transport and distribution.

The exposure of the population to copper from drinking water depends on several factors: the waterchemical composition, the stagnant contact time between water and the pipe, the age of the pipe, theinstallation procedures of the copper pipes, the use of copper pipes in the water distribution networks,and the drinking habits of the population.

All these factors, except the last one, are intrinsically considered in the copper monitoring dataprovided by the American Water Works Association, AWWA, and taken from house water taps by284 Water Utilities in the USA, which supply water to a population of 105.8 million people. A totalof 27,407 samples were taken by these utilities in order to comply with the USEPA regulation aboutcopper and lead in drinking water.

The mean concentration of the distribution of copper concentrations of the mentioned population is0.284 mg/L. It was assumed that the consumption of water is evenly distributed throughout the day,therefore the mean exposure of this population was 0.4 mg/day, assuming a consumption of 1.4 L peradult per day. In a worst case scenario, if all people drank all the water from the first draw in themorning, just after stagnation, then the average concentration ingested would be 0.987 mg/L, and thetotal mean exposure to copper from drinking water would be 1.38 mg/adult/day.

C o p p e r

49

1. Introduction

Surface waters used for the production of drinking water contain a wide range of copperconcentrations that vary between 0.5 and 1000 µg/L. A median level of 10 µg/L is reported by Daviesand Bennet (1985). Values above these levels (up to a few milligrams/litre) in drinking water usuallyreflect the by-product liberation of copper from plumbing fixtures and copper piping during watertransport and distribution.

This paper examines the main factors that produce liberation of copper by products to water fromcopper plumbing pipes. The calculation of the population exposure to copper in drinking water isbased on the monitoring data of the American Water Works Association of the USA (AWWA, 1996;Edwards et al., 1996a) generated since 1991.

2. Factors that affect the population exposure to copper in drinking water

The exposure of the population to copper from drinking water depends on several factors: the watercomposition, the stagnant contact time between water and the pipe, the age of the pipe, the installationprocedures of the copper pipes, the use of copper pipes in the water distribution networks, and thedrinking habits of the population.

2.1 Water composition

The concentration of copper after water transport and distribution in a copper pipe depends on thechemical composition of the water, mainly its pH and alkalinity (Schock and Neff, 1988; Dodrill andEdwards 1994; Schock et al., 1996). Other compositional variables such as the dissolved inorganiccarbon, organic substances and other compounds also play a role in copper by-product release (Cruseet al., 1988; Rehring et al., 1994)

Dodrill and Edwards (1994) conclude that copper in drinking water exceedance problems, withrespect to the United States Environmental Protection Agency, USEPA, action level (when the 90thpercentile copper concentration exceeds 1.3 mg/L), are confined mainly to two water characteristics:

1. when pH < 7.0 & alkalinity < 30 mg/L as CaCO3

2. when pH < 7.8 & alkalinity > 90 mg/L as CaCO3

Accordingly the main strategy for reducing copper by-product release should be to modify pH and/oralkalinity. Also phosphate inhibitors are used in the USA as corrosion inhibitors for iron, copper andlead, and can be effective in reducing copper release only below pH 7.8. Above this value its effectsare highly variable (Dodrill and Edwards, 1994).

It should be added that the 90th percentile concentration, according to the regulation of the USEPA, isdefined as the concentration of the 90th sample, out of a hundred, after the concentration of eachsample has been sorted in ascending order.

2.2 Stagnant contact time

The copper content of water that has been stagnant and in contact with a copper pipe increases itsconcentration to any value generally between 0.1 and 10 mg/L of copper, depending mainly on watercomposition and on the age of the pipe (Lagos, 1996b; Meyer, 1996). When water is flushed after astagnation period, a few liters of water can be collected at a high concentration. After flushing waterfor more than 3 to 5 minutes, the copper concentration falls to the source level, i.e., usually to a fewmicrograms per liter.

2.3 Age of copper piping

Studies conducted in Germany and in other countries show that for new copper pipes the copper byproduct release from the pipe to the water, also denominated cuprosolvency, is greater than for oldcopper pipes. Meyer (1996) showed that the reduction of copper concentration of waters can range

C o p p e r

50

from 2 to approximately 10 times, when a copper pipe ages. The reduction factor depends mainly onthe water composition.

The distinction between “new pipe” and “old pipe” is not quite that clear cut in terms of time, but dataavailable suggest that usually the concentration of copper in drinking water that has been stagnant in acopper pipe for a few hours, begins to fall after 12 months of use of the pipe and that this reductionafter 24 months of pipe use is quite substantial.

The reduction of copper by-product release in the case of “old” pipes can be explained by the build upof a protective layer at the inner surface of the pipe that protects the metal from corrosion. There arenumerous studies regarding the nature and properties of copper pipe films, but there are no conclusivefindings about the composition of these films and the way that they build up (Lagos, 1996b).

2.4 The design and installation procedures

The liberation of copper by-products from a copper pipe to the water can be influenced by networkdesign parameters such as high velocity flowing water, pipes in contact with different types of soil orconcrete, differences in temperature, excessive number of bends in the network, electrical groundingcontacts, stresses built up in the piping system, etc (Cruse et al., 1988).

Also, copper pipes can be joined by mechanical fittings, by welding, by brazing or by soft soldering(Cruse et al., 1988; Mattson, 1990). Fluxes are used in brazing and soft soldering.

Liberation of copper by-products to the water is produced when the materials that constitute the jointcome into contact with water and with copper. The result is usually to accelerate the copper pipecorrosion at the joint due to chemical attack by the jointing materials or by galvanic corrosion.

Jointing corrosion has been extensively studied and materials and methods are available in manycountries in order to prevent it. These developments have taken place mainly during the last decadeand they are still not applied in many developing countries.

2.5 The use of copper in the distribution network

The use of copper in plumbing pipes varies widely throughout the world (CDA, 1994). It is more than90% in the U.S.A. and the U.K., i.e., less than 10% of pipes of other materials are used in plumbingtubes in these countries. It is about 50% in Germany, between 50 and 60% in Spain and in France,close to 40% in Italy, and approximately 12% in Japan. Also, some countries like Germany, precludethe use of copper in plumbing tubes when the water composition is prone to copper by productrelease, and this may affect less than 5% of the country waters. In other countries, such as the U.S.A.,the regulation was required since 1991, that water utilities treat the water, when the copper actionlevel is exceeded, so as to prevent copper by-product release (Lagos, 1996a).

2.6 Drinking habits of the population

As well as the above factors, the drinking habits of the population, i.e., the time at which people drinkwater from the tap, the selection from the tap or from a kettle, the use of one versus several taps in thehouse, etc., affect the total copper ingested.

3 Analysis of water utility monitoring data of copper in drinking water in the USA

Dodrill and Edwards (1994) have examined data originally compiled during a 1991 AWWA survey,which included 366 utilities that distribute water to 159 million people and which represent more than60% of the US population. This study considered the effects of pH and of alkalinity, with or withoutthe addition of corrosion inhibitors on 90th percentile lead and copper concentrations and some of itsconclusions were described before. Parts of this database, added to information about population perutility, was made available for this study. Also some monitoring data before and after corrosiontreatment of the water for some utilities was obtained in order to assess the efficiency of thetreatment.

C o p p e r

51

The data base analysed (Edwards, 1996; AWWA 1996) contains the population data of 284 utilities(out of a total of 366), the number of samples taken by each utility, the 90th percentile, the lowest andthe highest copper concentrations measured by each utility, the pH, the alkalinity and also thecorrosion treatment status of the waters at the time of monitoring. The 284 utilities supply water to apopulation of 105.8 million people. A total of 27 407 samples was taken by these utilities in order tocomply with the US EPA regulation about copper and lead in drinking water (US EPA, 1991). The Ftest performed with the original sample (366 utilities) and with the reduced sample (284 utilities)showed that the value of F is 0.965, well outside the critical region defined by F<0.751 & F>1.338.There is 99% confidence that there is no difference between the variances of the two samples (σ366 =σ284) and hence, there is no significant error introduced by the sample reduction.

Figure 1 shows the copper concentration distributions for the 284 utilities: highest, 90th percentile,mean and lowest.

Figure 1 : Distribution of minimum, mean, 90% tile and maximum copper monitoring results for 286AWWA utilities, USA, 1991.

The mean concentration of each utility was calculated with the following equation:

cmean= (clowest +c90th% ) /2*0.9 + (c90th% +chighest) /2*0.1

The results of this calculation were tested with available utility distributions of copper concentrationsand it was found that the error was within ± 2.7%. The mean copper concentration of one utilityrepresents the mean exposure of the population supplied by that utility, provided that the waterconsumption is spread evenly during the day. The highest concentrations measured are weakly related(correlation coefficient R2 of 0.15) to the number of samples taken by each utility, within the range ofsamples required by the Lead-Copper Rule (US EPA, 1991). This means that the highest value wouldnot change substantially if more samples were taken.

The concentration distribution of utilities include therefore factors such as stagnant contact time, newand old pipes, water composition, and jointing factors, any one of which can cause high copper by-product release.

Figure 2 shows the distribution of mean copper concentrations for the 284 utilities, and also thedistribution of mean copper concentrations for 118 utilities (33.3 million people) whose water had notreceived corrosion treatment before monitoring. It should be added that the remaining 168 utilities hadreceived corrosion treatment in order to reduce iron corrosion. This treatment consisted of pHadjustment and/or addition of phosphate inhibitors. The sample of 118 utilites without corrosioncontrol is not statistically representative of the total sample (F test).

C o p p e r

52

This shows that iron corrosion control introduces a significant statistical difference between theconcentrations of copper present in the water, with and without treatment. Nevertheless, figure 2shows that the two distributions are very similar especially at the low end of the concentrations.

Figure 2: Mean population exposure to copper from drinking water (analysis of AWWA WITAFproject)

The mean concentration of the total distribution is 0.284 mg/l, and the maximum mean exposure is1.96 mg/l. This distribution signifies that 1.9% of the sampled population (total of 105.8 million) isexposed, to an average concentration over 1 mg/L of copper, 0.49% is exposed to more than 1.3 mg/Land none is exposed to more than 2 mg/L. The mean concentration of the no treatment distribution is0.367 mg/L and the maximum mean exposure is the same as in the total distribution. According to theno treatment distribution, 5.27% of the population is exposed to more than an average concentrationof 1 mg/L and 1.04% to more than 1.3 mg/L.

In the sample analysed, there are 15 utilities that had their highest reading above 3 mg/L of copperand two of them above 5 mg/L. The utility with highest value had 7.6 mg/L of copper. No estimationof the potential exposure has been made but it is apparent that a population in the thousands has beenpotentially exposed to concentrations above 3 mg/L at least during certain periods of the day. Itshould be stressed that these data points were included in the calculation of mean exposure discussedabove.

From the data of figure 2, and assuming a daily intake of drinking water of 1.4 litres (WHO, 1993),daily intakes of copper in adults will vary between less than 0.1 mg and 2.74 mg per day.

The data shown are no longer reproducible because after 1991 the utilities that exceeded the copperaction limit defined by EPA began treating the water in order to reduce copper and lead by-productrelease. Data after this treatment are scarce, but the example of two utilities is worthy of note: the firstone (see figure 3) supplies water to a population of 175 thousand people and its 90th percentileconcentration before treatment was 2.26 mg/L, with a highest concentration of 4.56 mg/L and aminimum close to zero. The pH before corrosion control was 7.1 and the alkalinity was 268 mg/L asCaCO3 .

To meet the action limit the utility raised pH to 7.4 and lowered alkalinity to 98 mg/L. Twosubsequent monitoring events conducted after this water quality change was implementeddemonstrated that the 90th percentile concentration dropped to 0.31 mg/L, and the highestconcentration to 0.6 mg/L. Release of copper by-products was mitigated at all houses. A second utilityachieved an overall reduction of 90 to 95% of the release of copper by-products after corrosion watertreatment.

C o p p e r

53

Figure 3: Copper concentration monitoring for one water utility in Colorado, USA, before and aftercorrosion treatment (AWWA WITAC project database).

4. Conclusions

The data base analysed contains the population data of 284 water utilities in the USA that supplywater to a population of 105.8 million people. A total of 27407 samples were analysed for copper bythese utilities in order to comply with the US EPA regulation about copper and lead in drinking water(US EPA, 1991).

The mean copper concentration exposure from drinking water of this population is 0.284 mg/Lprovided that water consumption is spread evenly during the day. The maximum mean exposure is1.96 mg/l. Of the sampled population, 1.9% is exposed to an average concentration over 1 mg/L ofcopper, 0.49% is exposed to more than 1.3 mg/L and none is exposed to more than 2 mg/L.

In the case of the waters that are not treated for iron corrosion control, which are supplied to 31.5% ofthe total population sampled, the mean copper concentration exposure is 0.367 mg/L and themaximum mean exposure is the same as in the total distribution. According to the no treatmentdistribution, 5.27% of the population is exposed to more than 1 mg/L as an average, and 1.04% tomore than 1.3 mg/L.

Assuming a daily intake of drinking water of 1.4 litres, daily intakes of copper in adults vary betweenless than 0.1 mg and 2.74 mg per day for the two copper concentration distributions considered.

References

American Water Works Association (1996), Database for AWWA WITAF Project: Initial MonitoringExperiences of Large Utilities under USEPA’s Lead Copper Rule. Version 2 including modificationsof Dodrill and Edwards, Denver CO.

CDA Annual data (1994), Copper Brass Bronze: Copper Supply & Consumption in the USA in theperiod 1973-1993. Copper Development Associaton, USA.

Cruse, H., Von Franque, O. and Pomeroy, R. (1988), Corrosion in Potable Water Systems, Chapter 5of Corrosion in Pipes, Published by the American Water Works Association, pp 317- 416.

Davies D.J.A., and Bennet, B.G. (1985), ‘Exposure of man to environmental copper: an exposurecommitment assessment’, Sci. Total Environ, vol. 46, p. 215-227.

Dodrill D. and Edwards M (1994), ‘Corrosion Control on the basis of utility experience’, submitted toJ. of Am. Waterworks Assoc., December.

C o p p e r

54

Edwards, M., Patel, S., Dodrill, D.M., Reiber, S. and Perry, S. (1996a), A general framework forcorrosion control based on utility experience, in press, AWWA, Denver, CO.

Edwards, M., Schock, M.R and Meyer, T.E. (1996b), ‘Alkalinity, pH and copper corrosion by-product release’, submitted to J. AWWA, January.

Lagos, G.E. (1996a), Regulatory aspects of copper in drinking water & corrosion of copper plumbingtubes: a review, in press. Catholic University of Chile, 93 pp, 1996.

Lagos, G.E. (1996b), ‘Introductory Remarks’ , published in Scientific basis for the regulation ofcopper in potable water. Edited by G.E. Lagos and L.A. Cifuentes, Published by the CatholicUniversity of Chile, January.

Mattson, E. (1990), Copper and Brass for Plumbing: a Guide for Corrosion Prevention, Procedings ofCopper 90 Refining, Fabrication, Market, October, The Institute of Metals.

Meyer, E. (1996), Determinants of Copper Intake from water, International Conference on Geneticand Environmental Determinants of Copper Metabolism, NIH, Washington D.C., March 18-20.

Rehring, J.P., and Edwards M, (1994), The Effects of NOM and Coagulation on Copper Corrosion,Proceedings 94, National Conference on Environmental Engineering",USA.

Schock, M.R., and Neff, C.H. (1988), ‘Trace metal contamination from brass fittings’, J. of Am.Waterworks Assoc., vol. 7, p. 47-56.

U.S. Environmental Protection Agency (US EPA) (1991) Maximum contaminant level goals andnational primary drinking water regulations for lead and copper; final rule. 40 CFR Parts 141 and 142.Fed. Reg. 56:110.

WHO (1993), ‘Guidelines for the quality of drinking water’, 2nd ed.. Volume 1: Recommendations.World Health Organisation, Geneva.

Risk assessment for essential trace elements: A proposedmethodology

Dr George C. BeckingInternational Programme on Chemical SafetyWorld Health Organization

1. Introduction

Risk assessment to protect human health and the environment from the adverse effects of chemicals isa scientific activity. It provides a framework for the critical examination of all available scientificdata in order to define and characterise dose-response relationships in humans in as quantitative amanner as possible. For non-essential chemicals it has a well developed methodology (NRC, 1983;NRC, 1994; IPCS, 1994) and has been used to develop protective health and environmental-basedguidance values within countries (Barnes and Dourson, 1988) and internationally (IPCS, 1987; WHO,1996). However, in the case of essential trace elements (ETEs) there needs to be a re-examination ofthe scientific principles and methods used in any risk assessment process for such elements. Thedevelopment of toxicologically- based guidance values for some ETEs (e.g. zinc) which are very near,or in conflict with, the recommended daily allowance serves to reinforce this view (Smith, 1994).

For ETEs we have what has been called a U-shaped dose-response curve, that is, adverse effects mayresult from deficiency associated with low intakes as well as be associated with high intakes (Goyer,1994; Bowman and Risher, 1994). There is, therefore, a need to define a tolerable intake (TI) toprevent toxicity as well as the level at which the risk of deficiency for the general population isminimised. As is done for non-essential chemicals, one cannot assume that a zero exposure is withoutadverse effects.

C o p p e r

55

It is hoped that this paper can present some ideas on how the scientific principles of toxicology andnutrition can be applied to the evaluation of ETEs. Before proposing a methodology it is importantfor the reader to know what is actually meant by risk assessment within the International Programmeon Chemical Safety (IPCS) and then summarise the proposed methodology considered usefulinternationally. IPCS in proposing such a methodology has drawn greatly upon earlier efforts toexamine this issue (Mertz et al., 1994; Nordberg and Skerfving, 1993). The methodology proposedwill be for non-carcinogenic ETEs.

2. Risk assessment and IPCS

2.1 IPCS goals in risk assessment methodology

The IPCS is a scientifically based response to problems caused to human health and the environmentby the expanded use of chemicals. It was initiated in 1980 by three cooperating organisations (UnitedNations Environment Programme, International Labour Organisation and the World HealthOrganization), to provide scientific support for the development of chemical safety programmes inMember States. Emphasis would be placed on assessing the risks from chemical exposures (eg. zinc,copper, dioxins, etc.) and the development and improvement of the methodology for risk assessment.

These are, in fact, two of the six objectives agreed upon by the Cooperating Organisations andMember States in 1980. It is within this framework that IPCS is concerned over the methodologyused to assess human health and environmental risks from exposure to ETEs. It is part of theinternational effort by IPCS to improve the scientific foundation for such assessments whichhopefully will result in an harmonisation of risk assessment methodologies worldwide.

2.2. Concepts and definitions

Within IPCS, the four step process first put forward by the US National Academy of Sciences (NRC1983) is considered an appropriate framework for the evaluation of health and environmental datarelated to effects of chemical exposures. An adaptation (NRC 1994) of this process is presented inFigure 1. The four steps are still hazard identification, dose-response assessment, exposureassessment, and risk characterisation. An important step included in Figure 1 is the identification ofresearch needs in order to lessen the uncertainty of future evaluations. For ETEs, it is hoped to showthis is an essential step for both nutritional and toxicological aspects. Much of the uncertainty withregards to evaluation of ETEs results from an inadequate database and a lack of understanding ofmammalian homeostatic mechanisms.

C o p p e r

56

Figure 1: NAS/NRC risk assessment/management paradigm (NRC 1994).

In describing the process in Figure 1, hazard identification and dose-response evaluation involvecomparing all available experimental animal and human data and the associated doses, routes andduration of exposure to determine if an agent causes the toxicity in question (neurotoxicity, generalsystemic toxicity, etc.) and the relevant exposure conditions under which this would occur. Exposureassessment ideally provides an estimate of human exposure from all sources and identifies the humanpopulations exposed and the magnitude and duration of such exposures.

Simply stated, risk characterisation is the summation of all the other steps and includes an estimationof the incidence and severity of the adverse effects that are liable to occur in a population orecosystem due to actual or predicted exposures. It includes a description of all default positions takenand the level of uncertainty within the overall assessment.

In the development of guideline values as described by Barnes and Dourson (1988), IPCS (1994) andIPCS (1987), the process is more of a safety assessment than a true probabilistic risk assessment.That is, a threshold is assumed for non-cancer end-points, a no-observable adverse affect level(NOAEL) (or lowest observed adverse effect level, LOAEL) identified and a guidance valuedeveloped by the use of uncertainty or modifying factors to reflect the uncertainties in the data baseand the severity of the critical effect identified. That is, exposure guidelines for a chemical aredeveloped which are assumed to pose no risk to the population (see Figure 2).

C o p p e r

57

Figure 2: Comparison of the safety assessment and probability models for assessment risks fromchemical exposures.

Many definitions and terms have been proposed for the health-based guidance values developed bythe process in Figure 2. These include: ADI-acceptable daily intake (IPCS 1987); TI or TDI -tolerable or tolerable daily intake (IPCS 1994); PTW(D)I - provisional tolerable weekly (daily) intake(IPCS 1987). These are all defined as estimates of exposure (intake) of a substance over a lifetimethat is considered without risk to most of the population. Although tolerable means "tolerated"; notnecessarily acceptable (IPCS 1994), the US Environmental Protection Agency prefers to utilise theterm RfD - daily oral reference dose (Barnes and Dourson 1988). This is an estimate of a dailyexposure (intake) to the human population (including sensitive subgroups) to be without appreciablerisk during a lifetime with uncertainty spanning an order of magnitude. All of these estimates aregiven as mg (µg) /kg body weight.

C o p p e r

58

2.3 Development of tolerable intakes for non-essential chemicals

All of the estimates described in section 2.2 are derived using similar default assumptions and thesafety assessment process in Figure 2. The NOAEL or LOAEL is usually determined from dataobtained from the most appropriate (to humans) species, or the most sensitive species of experimentalanimal. However, where human data of adequate scientific quality are available this is used to avoidthe uncertainties of interspecies extrapolation.

Most often there is much scientific uncertainty with the data available, for example, its relevance tohumans (interspecies extrapolation) and the variation in susceptibility of individual humans(intraspecies variability). Other factors leading to the uncertainty include adequacy of the pivotalstudy, nature of the toxicity (severity) and the adequacy of the overall database. Adjustments aretherefore made to the data through the use of uncertainty factors (UFs) which actually reduce the doserate to account for uncertainties or inadequacies in the scientific database.

Traditionally, UF of 10 for intraspecies variability and one of 10 for interspecies differences havebeen used, UF1 x UF2, for a total of 100 (IPCS 1987). Other UFs ranging from 1-10 have beensuggested as modifying factors (MF) for adequacy of data (eg. less than chronic exposure) andseverity of endpoint (eg. reproductive versus a general systemic effect which is often reversible). Toobtain the TI one then divides the NOAEL or LOAEL by the product of the uncertainty factors(NOAEL or LOAEL ÷ UF1 x UF2 x MF). When the product of the factors is greater than 5000 oneshould question whether the data base is of sufficient quality to make a reasonable estimate of the TI.Where data from humans are used to derive a TI, UFs used are usually not larger than 10 and, in fact,values of 2 or 3 are often used where the data are robust and a good dose-response relationship wasreported.

3. Development of recommended safe and adequate daily dietary intakes

It is beyond the scope of this paper to discuss in detail the development of Recommended DietaryAllowances (RDAs) - "levels of intake of essential nutrients that, on the basis of scientific knowledge,are judged to be adequate to meet the known nutrient needs of practically all healthy persons" (NRC1989) or an Estimated Safe and Adequate Dietary Intake (ESADDI). The latter forms interimguidance until adequate data become available to establish an RDA. However, there is merit inreviewing the basic principles for such estimations and compare these to the methodology forestimating an TI. More details on the derivation of RDAs and ESADDIs and a comparison of themethodologies used in estimating both RDAs and TDIs can be found in Mertz et al. (1994) andBowman and Risher (1994).

In summary, the first step is to determine the average requirement (x) assuming a normal distributionwithin the population, assume a coefficient of variation (usually 15%) and calculate an RDA (RDA =x +2 SD = 1.3 x ). Such data are usually obtained from studies of human populations, althoughextrapolation from animal experiments has been used to recommend an ESADDI for selenium of 50-200 µg/day (Mertz et al., 1994).

For details on the use of balance studies, human exposure studies and Factorial EstimatedRequirements in deriving RDAs the reader is referred to WHO (1996) and Mertz et al. (1994).

3.1 Comparison of methodologies for derivation of RDAs and TIs

A few important differences in the assumptions underlying the derivation of RDAs and TIs are: (1)RDAs assume all intake is from dietary sources, whereas, for TIs all sources of exposure areconsidered; (ii) in deriving RDAs bioavailability, homeostatic regulatory mechanisms, foodconsumption and dietary interactions are all considered, whereas, for a TI the toxicity measured israrely modified by similar considerations; (iii) the RDA is considered protective of 97.5 percent of thepopulation whereas TIs are usually considered more protective. In fact, RfDs (another term for a TI)derived by the US EPA are by definition considered protective of 100 percent of the population

C o p p e r

59

(Bowman and Risher, 1994); and (iv) the toxic endpoint considered adverse in deriving a TI,particularly from animal experiments, is often much less significant clinically in humans than theknown severe clinical manifestations found in human deficiency states for ETEs.

4. Principles and methods for assessing human health risks from exposure to an ETE

4.1 Scientific principles

4.1.1 The acceptable range of oral intake for an ETEFor an ETE by definition there are risks associated with both low (deficiency) and high (toxicity)intakes. The relationship between intake of ETEs and risk is best represented by the U-shaped curveas shown in Figure 3. There is a range of intake below which the risk from the adverse effects ofdeficiency increase in the population and at the upper end of this range the risk from adverse toxicresponses increases as exposures increase. The range between point A, the RDA or ESADDI andpoint B, the TI, can be considered an acceptable range of oral intake (AROI). A similar concept wasfirst proposed by Beaton (1988) and the range was termed "a safe range of intake". Within this rangenormal homeostasis is maintained. Neither the lower or upper boundary of the AROI should beconsidered an absolute value, below or above which adverse effects in a population or individual willbe initiated. In fact, RDAs and TIs are not guidelines for individuals but for populations.

Figure 3: Percent of population subjected to deficiency and toxic effects according to exposure orintake of the essential trace element (ETE). As intake drops below A (lower limit of the AROI) risk fordeficiency increases, reaching 100 percent at extremely low intakes. As intakes increase beyond B aprogressively larger proportion of subjects will exhibit effects of toxicity.

For ETE the oral route is by far the predominant exposure pathway for humans and for this reason theterm AROI is maintained to facilitate comparison of the RDA which only considers oral intakes andthe TI for toxicity. The concept of a range of intakes within which normal homeostasis is maintainedcan be used in situations where inhalation becomes a major exposure pathway, for example, inoccupational exposures. One then needs to consider an acceptable range of total intake (ARTI).However, for this paper the AROI will be used and the range suggested would apply to protection ofnormal, healthy, non-occupationally exposed populations.

4.1.2 Determination of boundaries of AROIIn deriving the upper bound (point B, Figure 3) of the AROI the use of UF/MFs to derive a TI from aNOAEL (see section 2.3) needs to be less rigidly applied and other factors such as bioavailability and

C o p p e r

60

nutrient interactions need to modify the choice of any UF used. Such an approach has been suggestedby Abernathy et al. (1993) in the derivation of a RfD for zinc where an UF of 3 rather than the usualfactor of 10 when a LOAEL from a human study is used. The lower UF took into account nutritionaland energy requirements but no consideration was given to the level of bioavailability or nutrientinteractions. It should be noted that this RfD (0.3 mg/kg/day in adults) does not provide the RDA forinfants, preadolescent children, or possibly lactating women (Abernathy et al., 1993; Sandstead,1993).

The conflict remaining between the RDA and the US EPA RfD for zinc raises other principles to beconsidered in assessing the risks from all ETEs. In using even a modified approach to setting TIs foran ETE, toxicologists so used to the principle that zero exposure for toxic materials is to be strivedfor, fail to consider bioavailability of the ETE and possible nutrient interactions in animal or humanstudies even in cases where levels well in excess of the RDA were consumed and in a form markedlydifferent than that found in normal diets.

Animal experiments and some human controlled studies use pharmacological doses of extremelybioavailable forms (eg. chlorides or gluconates) of the ETE when compared to the human dietarysituation. It is essential that toxicologists make better use of the available studies on the ETE intakein normal healthy populations to validate the TI derived from animal experiments. Clearly, from apublic health perspective, it would be unwise to define an AROI that is outside the range of customaryintakes of healthy populations worldwide.

4.1.3 Clinical equivalency of end-points for deficiency and excess exposureIn defining an AROI for an ETE, for example, copper, a key issue is the appropriate assessment of theeffects of deficiency and toxicity relevant to human health. The most sensitive indices of copperdeficiency or toxicosis are biochemical markers with no clear functional or health significance. On theother extreme, death associated with organ damage induced by excess copper or deficiency are clearlyof health significance but are not relevant as sensitive early indicators of health risks in preventivepublic health programmes. Examples of the type of effects to assess health risks associated withcopper exposure need to be described and their health significance (adversity) ascertained.Biochemical changes such as in erythrocyte superoxide dismutase (ESOD) activity as an index ofcopper deficiency (Uauy et al., 1985) or changes in the plasma Cu/Ceruloplasmin molar ratio as anindex of copper excess are sensitive but not significant indicators of health risks.

Biological indices of subclinical effects indicating potential adverse effects of copper deficiency orexcess on organ function need to be developed. For example, decreased white blood cell phagocyticcapacity in the case of deficiency or increased serum aminotransferase or transpeptidase hepaticenzymes in response to excess copper are often used in human studies. The important concept shownby these examples is the principle of comparable effects to define effects of excess and deficiency ofcopper. A set of theoretical curves developed by Dr G. Nordberg, University of Umea, Sweden,further illustrating this point is given in Figure 4. One should select effects of similar healthsignificance to define the upper and lower ranges of the AROI. In general, the range should bedefined by dose-effect (response) to intake/exposure levels that prevent the appearance of subclinicaladverse effects. The review of these indices and the corresponding studies should be done bytoxicologists and nutritionists familiar with health risk assessment. The combined effort should yieldclearly defined critical end-points for the upper and lower ranges of the AROI given the availableinformation. Even if the human information is limited, a starting point in the definition of AROIshould be the customary intake/exposure determined in "healthy" populations in various regions of theworld.

C o p p e r

61

Figure 4: Theoretical dose-response curves for various effects occurring in a population at variouslevels of intake (doses) of an essential trace element. While lethal effects and clinical disease mustalways be prevented, subclinical effects indicating impairment of organ function are often identifiedas critical effects. The lower end of the dose-response curve for such critical effects related todeficiency (curve 3) and toxicity (curve 6) define the range of acceptable daily oral intakes.Biochemical effects without functional significance (curves 4 and 5) are considered without healthimpact and should not be taken as critical effects.

4.1.4 Range of acceptable intake for healthy populationsAnother guiding principle in defining the AROI is that the acceptable exposure range should be safefor, but not expected to meet the requirements or prevent excess in, groups at special risk. Forexample, the copper needs of patients with chronic diarrhoea or of patients on chronic haemodialysismay fall outside the AROI. In both of these cases intakes that are of concern for the "healthy"population may be required to meet the special needs of these patients. Safe copper intake/exposurelevels for haemodialysis patients may lead to deficiency in healthy subjects while intakes required bypatients with chronic diarrhoea may be excessive for normal subjects.

The upper and lower cutoff points for the range of acceptable oral intakes should be defined forpopulation groups. The lower cutoff point should be sufficient to meet the requirements of mostindividuals in the population. Based on criteria used to define RDAs "most" usually implies 97.5% ofthe population; if the mean and standard deviation for requirements are known this point is defined bythe mean + 2 SD, if the SD is not known a coefficient of variation of 15% is customarily used toaccount for population variability in requirements for ETEs. Similarly, the upper boundary shouldprotect most individuals from the risk of toxicity. A statistical definition for "most" in this context islacking, but could be derived based on the mean and distribution of toxic effect doses.

Using the traditional toxicological approach, an ED50 can be defined and, based on known variability,or extrapolation of the dose-effect (response) relationship, an ED2.5 determined. Figure 5 summarisesthe concept of population-derived boundaries for the AROI. Special consideration of upper and lowercutoff points should be made in defining AROI for physiological conditions affecting normalpopulations, such as infancy, pregnancy, lactation, and aging. In the case of the lower boundary of the

C o p p e r

62

AROI these are usually considered in setting RDAs but, for the upper cut off point specific values forthese population categories is often lacking but should be determined as a research priority.

Figure 5: Ideal model for distribution of intakes to meet nutritional requirements of a healthpopulation and prevent toxicity. The lower limit to the AROI should cover the requirements of most(97.5%) of the population while the higher limit of the AROI should protect most of the populationfrom toxic effects.

In the case of essential elements it is clearly impossible to assure that all (100%) of any populationwill be protected from deficiency or excess levels of ETEs such as copper. The range of AROIdefined at the international level is not intended to address disease conditions or genetic alterations incopper metabolism which determine special sensitivity to excess or deficient exposures. This wouldrequire a change in philosophy of agencies using the US EPA RfD in their public health programmes.The AROI is clearly not intended to meet the special needs of population subgroups with geneticalterations of copper metabolism, such as Menkes’ syndrome or Wilson's disease, and as more databecome available probably idiopathic childhood cirrhosis which most likely has a genetic componentin its aetiology (Uauy, personal communication).

4.2 The AROI concept in human health risk assessment - A proposed methodology

A summary of the scientific principles supporting the use of an AROI in the assessment of humanhealth risks from ETEs is given in Table 1. Many of these were agreed upon at a workshop in 1992(Mertz 1993) and some have been applied in the derivation of an RfD for zinc (Abernathy et al.,1993).

C o p p e r

63

Table 1: Principles underlying use of the Homeostatic Model in human health risk assessment

(1) For all ETEs there is a "zone of safe and adequate exposure" a zone compatible with good health- an acceptable range of oral intake (AROI).

(2) Both nutritional scientists and toxicologists must be involved in developing an AROI.

(3) Data on toxicity and deficiency should receive equally critical evaluation.

(4) The concept of bioavailability (biologically effective dose) should be applied and nutrientinteractions considered when known.

(5) Chemical species studied and the route and mode of application should be fully described.

(6) Biological end-points used to define the lower (RDA) and upper (toxic) boundaries of the AROIshould have similar degrees of clinical significance.

(7) Safety margins and uncertainty factors are used to determine both boundaries of the range. Theyare usually higher for overexposure but need to be applied taking nutritional needs into account.

Adapted from Mertz, 1993.

When the principles in Table 1 are applied within the framework shown in Figure 6, it should bepossible to provide guidance to all countries on the exposure levels for any ETE which would provideadequate nutrition and be without risk from toxicity. The iterative nature of this proposed schememakes it possible to identify research needs and accommodate special groups within the populationwith advisories regarding exposures. It is essential that this scheme be applied by nutritionalscientists and toxicologists working together to address such aspects as data quality, critical effects,dose-response, bioavailability and nutrient interactions. It is essential that all scientists involvedpractice sound scientific judgement during each step and make it clear whenever default positions aretaken in lieu of scientific data.

C o p p e r

64

Figure 6: Application of principles for the assessment of risk from essential trace elements.

The scheme in Figure 6 should not be considered a new and novel way of assessing health risks. Itreally represents the thought process used by those scientifically competent and versed in theprocedures used to carry out evaluation of data on ETEs whether from studies on experimentalanimals or humans. The end result should be a range of recommended exposures fully protective ofhuman health, something a single number for an ETE can never do.

C o p p e r

65

5. The Deficiency - Toxicity (DT) Model in Environmental Risk Assessment

It is well beyond the scope of this paper, and the expertise of the author, to discuss in detail thescientific basis for assessing risks to the various compartments of the environment from chemicals.However, a recently proposed model (van Tilborg, 1995), the Deficiency Toxicity Model (DT) hasmany features of the AROI model described in section 4 of this paper.

Based on the dose-response curves for both deficiency and toxicity in various species within say theaquatic environment, a range termed the "Optimal Concentration band for Essential Elements"(OCEE) can be derived. This range, as for the AROI, has a lower boundary as the point below whichrisk from deficiency increase and an upper boundary based on increased risks from toxicity. Themodel, as applied to the fresh surface water ecosystem, is shown in Figure 7.

Figure 7: The DT-model for ecological risk assessment for zinc. The natural boundaries of theconcentration band defining optimum conditions for life in fresh surface water are shown and forcomparison the natural average zinc concentrations of other typical aqueous habitats are shown.

Many of the principles shown in Table 1 apply to the DT Model as well as to human health riskassessment models. In particular, the need for scientists with appropriate training and experience tocarry out the evaluations and the need for sound scientific decisions in assessing data quality, dose-response relationships and in determining the critical effect(s). Of critical importance in applying theDT Model is the relevance to the real environment of the testing procedures used to generate data.Test conditions need to be critically evaluated. Only data generated under conditions consideredappropriate to the environmental compartment of concern should be used to set the upper boundary ofthe OCEE.

6. Concluding remarks

The proposed methodology for assessing the health effects of human exposure to ETEs has built upondiscussions at national and international workshops and conferences over the last four or five years.

C o p p e r

66

A synthesis of these discussions into what is hopefully a useful set of principles which can be agreedupon worldwide is presented here.

Although the primary focus in this paper was on human health risk assessment, the recently proposedDT Model for ecological risk assessment is based on many of the same scientific principles, modifiedfor the complexity of each environmental compartment. Therefore, it may in the future be possible todevelop an integrated methodology for both human health and environmental risk assessments forETEs. However, the usefulness of such methodology will only be as good as the data available. Atpresent, the data on environmental effects needs to be strengthened and testing strategies developed toprovide more relevant data if the DT Model is to become widely utilised.

For both the DT Model, and that proposed for the development of an AROI for human health, thispaper is simply proposing a scheme where risk assessments will be carried out by scientists (or teamsof scientists) expert in risk assessment procedures as well as both the nutritional and toxicologicalsciences. Given the need to be concerned over the effects of deficiency as well as excess exposures,the proposed scheme is even more multidisciplinary in scope than the one used for non-essentialchemicals (Figs.1 and 2). In working together nutritional scientists and toxicologists need tounderstand the principles underlying the development of both RDAs and TIs in the default positionstaken in both cases, and how such assumptions can result in apparent conflicts between the upper andlower boundaries of the AROI.

By fostering a closer interaction between scientists in the areas of nutrition and toxicology it is hopedthat more adequate scientific data will be developed to address the boundaries of the AROI.However, unless both groups of scientists are willing to use all available scientific data, including thaton nutrient interactions and bioavailability, we will continue to have apparent conflicts between theTIs and RDAs. The rigid application of conservative UF/MFs in any evaluation process is nosubstitute for the use of scientific expert judgement on the complete database when developing publichealth strategies related to the health and environmental effects from exposure to ETEs.

Acknowledgements

The assistance in developing these concepts by the participants at a recent IPCS consultation isgratefully acknowledged, namely: Dr C. Boreiko, USA; Mr G. Ethier, Canada; Dr H. Gibb, USA; DrR. Goyer, USA; Dr W. Mertz, USA; Dr G. Nordberg, Sweden; Dr S. Olin, USA; and Dr R. Uauy,Chile.

References

Abernathy, C.O., Cantilli, R., Du, J.T. and Levander, O.A. (1993), Essentiality versus toxicity: someconsiderations in the risk assessment of essential trace elements. In Saxena, J. (ed.). Hazardassessment of chemicals, Vol. 8. Taylor and Francis, Washington, D.C., 81-113.

Barnes, D.G. and Dourson, M. (1988), ‘Reference dose (RfD): description and use in health riskassessments’, Reg. Toxicol. Pharmacol., vol. 8, pp. 471-486.

Beaton, G.H. (1988), ‘Nutrient requirements and population data’, Proc. Nutr. Soc., vol. 47, pp. 63-78.

Bowman, B.A. and Risher, J.F. (1994), Comparison of the methodological approaches used inderivation of recommended dietary allowances and oral reference doses for nutritionally essentialelements. In: Risk Assessment of Essential Elements. Mertz, W., Abernathy, C. and Olin, S. eds. ILSIPress, Washington, D.D., 63-73.

Goyer, R.A. (1994), Biology and nutrition of essential elements. In: Risk Assessment of EssentialElements. Mertz, W., Abernathy, C. and Olin, S. eds. ILSI Press, Washington, D.D., 13-19.

IPCS (1987) , Principles for the safety assessment of food additives and contaminants in food.International Programme on Chemical Safety, Environmental Health Criteria 70. World HealthOrganization, Geneva.

C o p p e r

67

IPCS (1994), Assessing human health risks of chemicals: derivation of exposure. InternationalProgramme on Chemical Safety, Environmental Health Criteria 170. World Health Organization,Geneva.

Mertz, W. (1993), ‘Risk assessment of essential elements’, Scand. J. Work Environ. Health, vol. 19,no. 1, p. 112.

Mertz, W., Abernathy, C.O. and Olin, S. (1994), Risk Assessment of Essential Elements, ILSI Press,Washington, D.C., 300.

Nordberg, G.F. and Skerfving, S. (1993), Editors. Biological monitoring, carcinogenicity and riskassessment of trace elements. Scand. J. Work Environ. Health, 19 Suppl 1: 140.

NRC (1983), US National Research Council. Risk Assessment in the Federal Government:Managing the Process. National Academy Press, Washington, D.C.

NRC (1989), US National Research Council. Recommended dietary allowances, 10th ed. NationalAcademy Press, Washington, D.C.

NRC (1994), US National Research Council. Science and Judgement in Risk Assessment. NationalAcademy Press, Washington, D.C.

Sanstead, H.H. (1993), ‘Zinc requirements, the recommended dietary allowance and the referencedose’, Scand. J. Work Environ. Health, vol. 19, issue 1, pp. 128-131.

Smith, J.C., Jr. (1994), Comparison of reference dose with recommended dietary allowances forzinc: methodologies and levels. In: Risk Assessment of Essential Elements. Mertz, W., Abernathy, C.and Olin, S. eds. ILSI Press, Washington, D.D., 127-143.

Uauy, R. Castillo-Duran, C., Fisberg, M., Fernandez, N and Valenzuela, A. (1985), ‘Red cellsuperoxide dismutase activity as an index of copper nutrition’, J. Nutr., vol. 115, pp. 1650-1655.

van Tilborg, W.J.M. (1995), Integrated Criteria Document Zinc (ICDZ) - Industry Addendum.Project groep Zink, BMRO-VNO. European Zinc Institute, Brussels., 50.

WHO (1996), Guidelines for drinking-water quality, 2nd ed. Volume 2: Health Criteria and othersupporting information. World Health Organization, Geneva.

Assessment of the requirement of copper in the nutritional supportof the severely ill patient

Dr Gordon S Fell, Dr TDB LyonScottish Trace Element Service, Macewen BuildingRoyal Infirmary University and NHS Trust Hospital, Glasgow.Dr W WatsonDepartment of Clinical Physics , Southern General Trust Hospitals, Glasgow.

The various stages of the metabolic responses to severe injury were investigated and classified by D PCuthbertson working in Glasgow over a 60 year period (Barton et al., 1990). The ebb or shock phaseof hypo-metabolism is followed by the flow or acute phase of hyper-metabolism during which patientshave an obligatory negative nitrogen balance. This is a measure of the catabolic loss of intracellularmass which if unchecked will cause muscle weakness, delayed wound healing and a reduced immuneresponse with increased risk of infection.

Increased urinary losses of cell contents also includes major inorganic elements such as K, S, P andessential trace elements such as Zn and Cu. As the patient recovers the anabolic phase of weightregain in adults, and renewed growth in children begins. This is characterised by a positive nitrogenbalance and requirement for nutritional support. This must include a supply of all nutrients needed torestore the lost intracellular contents. The severely injured patient may also have a variable period of

C o p p e r

68

inadequate food intake due to the nature of the injuries, and where intestinal surgery is involved havea degree of malabsorption of diet. For these and other reasons, appropriate nutritional support is nowan important part of patient care (Hill, 1994).

Complete nutritional support is usually instituted as soon as the patient has been stabilised and it isclear that normal diet cannot be taken at all or in insufficient quantity. Often it is necessary to supplynutrients by intravenous infusion (IVN). This requires suitable forms of energy substrates such asglucose and lipid emulsions, as well as amino acids, minerals and the essential trace elements andvitamins. Ideally all of these nutrients should be infused in the same chemical form and amount as thatabsorbed into portal blood from good quality oral diet. This is not feasible in practice and the essentialtrace elements, including copper are given in the form of simple salts. The amount required shouldideally be determined for each patient by assessment of tissue concentration and the detection of adeveloping deficiency by a functional assay of an appropriate intracellular metalloezyme system.Additionally the net input-output balance for each essential trace element should be determined byanalysis of all IVN fluids, and excreta.

In practice such detailed studies are not possible especially in critically ill patients. A compromisemulti-component additive mixture of containing several essential trace elements is used. This additiveoffers amounts estimated to meet most requirements but not thought liable to cause toxicity (Shenkinet al., 1987).

1. Copper deficiency and treatment

Failure to supply sufficient copper in the nutritional regimens some years ago, did result insymptomatic copper deficiency (Jeejbhoy, 1989). Symptoms take several months to develop. This islonger period than it takes for the equivalent zinc deficiency disease to develop, due the reserve storeof copper in the normal liver.

In adults copper deficiency can present as haematological abnormalities. There is a low neutrophilcount (neutropenia) and a mild iron resistant microcytic anaemia. A brisk increase in reticulocytecount is seen when copper is added to the IVN regimen with reversal of the neutropenia and anaemiaThis condition has also been recorded in a patient with chronic intestinal malabsorption (Bruce et al.,1995).

In children, especially low birth weight pre-term babies, there is a particular risk of the developmentof copper deficiency. (Aggett, 1994). This is caused by the failure of the fetal liver to obtain maternalcopper during the last trimester of pregnancy. As well as neutropenia these infants have brittle, easilyfractured bones. This is due to impairment of the copper-dependent enzymes required for bonecollagen formation. The condition can be prevented by supplementation of diet with adequateamounts of copper.

The amount of copper recommended for an adult on IVN is 20 µmol per day (1.3 mg Cu). This ismore than the 6-10 µmol/day needed to maintain overall balance, but is an amount needed to replacelosses and is not associated with toxic effects (Cruickshank et al., 1985). A safe and “adequate”dailyintake of 0.4-0.6 mg/day has been proposed for term neonates. Recommendations vary quite widely,and preterm babies may have higher needs ranging from 20-40 µg/kg/day (Tyrala, 1992).

2. Biochemical assessment

2.1 Plasma copper

The variable amount of copper stored in liver acts to maintain plasma copper concentrations even inthe face of low copper supply. Measurement of plasma copper or plasma ceruloplasmin althoughconvenient markers of copper nutritional status are a relatively insensitive indices of tissue depletion.

In a case study, a 42 year old man with severe intestinal malabsorption had been maintained on homeIVN for some 5 years. and had remained well and weight stable. He prepared his own IVN fluids andmade the trace element additions. However persistent IV line blockage occurred and led him to reducethe use of the trace element additive mixture. Regular monitoring of this patient revealed that his

C o p p e r

69

plasma Cu had declined to 6.9 µmol/L (reference interval (12-24 µmol/L). Oral copper supplementswere given (40 µmol, 2.5 mg) on alternate days which corrected his impending copper depletion butmay have triggered a later symptomatic zinc deficiency (de Caestecker 1985).

2.2 Monitoring of neonates

The limited amount of blood available from a low birth weight preterm baby poses considerabletechnical problems. However analytical methods have been developed able to determine theconcentration of some seven essential elements, including Cu, in only 1-2 ml of blood serum. In astudy comparing two different forms of trace element additives we were able to show that therecommended input of 20 µg/kg /day maintained satisfactory plasma copper concentration. Thechanging values of plasma copper in neonates with increasing gestational age (Agett, 1994) makes theinterpretation difficult, and requires repeated sampling and knowledge of gestational age.

2.3 Acute phase plasma protein effect

An additional difficulty in the use of plasma copper levels to monitor acutely ill patients, particularlywhere there is accompanying sepsis and inflammation, is the increased liver synthesis ofceruloplasmin. This copper containing protein accounts for some 80-90% of total plasma copper andincreases in plasma about 2-3 days after infection or injury as a positive “acute phase protein.”

Sequential measurements after cardiac surgery or choleocystectomy show the marked variation ofboth ceruloplasmin and plasma copper with time after surgical incision (Fraser et al., 1989). Suchchanges are independent of copper dietary intake.

A biochemical measure of the acute phase effect is plasma C-reactive protein. (CRP). This proteinwhich is normally present in plasma at low concentrations (<10 mg/L) increases 10-100 fold within48 hours of infection and/or tissue injury. A study of two groups of lung cancer patients showed thatthose who were catabolic and weight losing had higher CRP values and higher plasma Cu andceruloplasmin levels than an equivalent group of weight stable patients. Both groups had similarnutritional intakes.

2.4 Action limits

Although plasma Cu does not reflect short term changes in Cu intake the finding of a persistently lowconcentration of <5.0 µmol/L (30 µg/dL) is likely to indicate impending symptomatic Cu deficiencycaused by chronic copper depeletion. Since the acute phase response, present in most severely illpatients should normally increase plasma Cu well above the reference interval of 12-22 µmol Cu/L isprobable that values of < 12 µmol Cu/L in plasma reflect an incomplete acute phase response due tosome degree of Cu deficiency. However it has been shown that, even when Cu input exceeds Culosses, plasma Cu concentrations do not change, at least in the short term. Therefore the finding of avalue within the reference range for plasma Cu does not exclude some degree of tissue Cuaccumulation.

2.5 Contra-indication for copper supplementation

It has been shown by metabolic balance studies in animals and in humans, that the major regulator ofcopper homeostasis, which keeps tissue concentrations of copper constant in the face of widelyvarying dietary inputs, is the biliary excretion of excess copper. We have demonstrated by metabolicstudies (Cruickshank et al., 1988) that a patient with extensive bile fluid losses was in negative copperbalance -8.6 µmol (0.55mg) Cu /day in spite of receiving the standard amount of Cu in the IVNregimen. It follows that patients with obstructive liver disease and reduced bile excretion will beliable to accumulate copper. Dosage in such cases should reduced and the patient examined for signsof Cu toxicity. However the storage capacity of the liver for Cu is large, and at the daily inputsroutinely used it is unlikely that overt Cu toxicity would develop quickly.

C o p p e r

70

2.6 Copper isotope tracer studies

Marginal copper depletions will not be detected by determination of plasma Cu concentrations for thereasons given above. Direct determination of liver Cu concentration by analysis of needle biopsysamples is feasible but is an invasive process not undertaken unless other strong clinical indicationsare present. The kinetics of Cu intestinal absorption, blood transport, tissue uptake and excretion inbile and urine have been investigated in animals and in a few humans by use of radioactive Cuisotopes. These are relatively short-lived radioisotopes (T1/2 24-48 h) and not suitable foradministration to infants and women of child bearing age. Recently it has been possible to determinethe stable copper isotope 65Cu in blood plasma by an inductively-coupled plasma mass spectrometrymethod (Lyon and Fell 1990)

The ratio of 65Cu : 63Cu can be measured sequentially in blood plasma after oral dosage with 3mg65Cu and the appearance curve of the enriched isotope determined (Lyon et al., 1995). This togetherwith data taken from published “metabolic models” has allowed the construction of a mathematicalmodel of human copper metabolism.

The administration of the stable isotope 65Cu to various categories of patients is feasible and could beused to quantify the disturbance of copper metabolism caused by disease. An estimate of the percentintestinal absorption from a single dose (3mg 65Cu ) has shown that in healthy controls a mean of 50%is absorbed. Patients with residual malabsorption due to intestinal disease take up a more variableamount but usually less, at around 30% of the administered dose. By conducting quantitativemetabolic balance estimations and using the stable isotope as a tracer, we can fit the results into themathematical model and make an estimate of the concentration of copper in the liver compartmentsand in the “rest of body”.

3. Summary and conclusions

The routine provision of 1.2 mg/Cu per day during IVN will prevent overt copper deficiency diseasein most adult patients. Copper toxicity is not likely although patients with severe obstructive liverdisease need to monitored. In pre term infants an increased amount of copper /kg body weight may beneeded to allow for their low liver stores. This element is not always present in adequate amounts inoral feeds or IVN regimens. Monitoring for Cu deficiency in both adults and babies can be based onscreening for very low levels of plasma Cu. A finding of repeated values of (<5.0 µmol/l) indicates aneed for further copper supplementation. There are alterations in the metabolism of copper in the illpatient which are independent of dietary copper supply. Plasma copper concentrations at the lowerend of the reference range (<12 µmol Cu/L) may mask an underlying tissue copper depletion.

Marginal copper depletions may have an adverse effect on antioxidant defences and the immunesystem. Metabolic balance studies employing stable copper isotopes as a tracer may be required todemonstrate copper depletion in at risk groups.

References

Aggett, P.J. (1994), ‘Aspects of neonatal metabolism of trace elements’, Acta paediatr, Suppl 402, pp.75-82.

Barton, R.N., Frayn, K.N. and Little, R.A. (1990), In Chapter 33. Trauma, Burns and Surgery. Themetabolic and molecular basis of acquired disease Vol 1 Eds Cohen RD, Lewis B, Alberti KGMMand Denman AM, Baillere Tindall.

Bruce, A., Hayton, H., Broome, E and Lilenbaum, R.C. (1995), ‘Copper deficiency induced anemiaand neutropenia secondary to intestinal malabsorption’, Amer. J of Haematology, vol. 48, pp. 45-47.

Cruickshank, A.M., Rodgers, P, Dunbar P., Fell, G.S. and Shenkin, A (1988), ‘Copper balance inintravenously fed patients’, Proc. Nutr. Soc., vol.17.

de Caestecker, J.S., Shenkin, A., Fell, G.S., Heading, R.C. (1985), ‘Hazards and benefits in a patienton long term total parenteral nutrtion’, Proc Nutr Soc., vol. 45, p. 23 A.

C o p p e r

71

Fraser, W.D., Taggart, D.P., Fell, G.S., Lyon, T.D.B., Wheatly, D., Garden O.J. and Shenkin, A.(1989), ‘Changes in iron, zinc and copper concentrations in serum and in their binding to transportproteins after choleocystectomy and cardiac surgery’, Clincial Chemistry, vol. 35, pp. 2243-2247.

Hill, G.L. (1994), ‘The impact of nutritional support on the clinical outcome of the surgical patient’,Clinical Nutrition, vol. 13, pp. 331-340.

Jeejbhoy, K.N. (1990), ‘Trace elements in total parenteral nutrition’ in Trace elements in ClinicalMedicine, Ed Tomita, H. Springer -Verlag.

Lyon, T.D.B and Fell, G.S. (1990), ‘Isotopic composition of copper in serum by inductively coupledplasma mass spectrometry’, JAAS, vol. 5, pp. 135-137.

Lyon, T.D.B., Fell, G.S., Gaffney, D., McGaw, B.A., Russel, R.I., Park, R.H.R., Beattie, A.D., Curry,G., Crofton, R.J., Gunn, I., Sturniolo, G.S., D’Inca, R. and Patriarca, M. (1995), ‘Use of a stablecopper isotope ( 65 Cu) in the differential diagnosis of Wilson’s Disease’, Clinical Science, vol. 88,pp. 727- 732.

Shenkin, A., Fraser, W.D., McLelland, A.J.D., Fell, G.S. and Garden, O.J. (1987), ‘Maintenance ofvitamin and trace element status in intravenous nutrition using a complete nutritive mixture’, Journalof Parenteral and Enteral Nutrition, vol. 11, no. 3, pp. 238- 242.

Tyrala, E.E. (1992), ‘Trace elements’, in Chapter 9 of Intravenous Feeding of the Neonate ed. Yu VHand Macmahon R. publ. Edward Arnold

Copper tailing impacts in coastal ecosystems of Northern Chile:From species to community responses

Professor Juan C Castilla and Dr Juan A. CorreaDepartamento de Ecologia. Facultad de Ciencias Biologicas,Pontificia Universidad Catolica de Chile,Santiago, Chile.

1. Introduction

At the beginning of the 1980s, the anthropogenic inputs of copper into aquatic ecosystems variedbetween 34.7 and 190.5 tons/year. Copper mining, smelting, and refinery activities accounted forapproximately 14% (Pacyna et al., 1995). Mining copper activities take place in all six continents andsupport a demand of about 10 million tones of copper per year. Copper mines in South America,mainly in Chile and Peru, account for approximately one third of the total world demand. In Chile, theproduction of fine copper increased from 691,600 metric tones in 1970 (10.9% of the worldproduction) to 2,219,900 metric tones in 1994, representing 24% of the world production(Anonymous, 1995).

Chilean copper mine activities are based on 10 major open or underground mines located at highaltitude (usually over 2,000 m above sea level) in the Andes mountains, between 22°-34° S. In most ofthe cases the ore extraction, processing, smelting and tailing dumping take place around the mine pits.Nevertheless, in the past (1938-1990), El Salvador copper mine dumped all its untreated tailingsdirectly to coastal areas in northern Chile (this practice is presently banned according to Chileanenvironmental legislation). The El Salvador case has been reported in detail elsewhere (Castilla andNealler, 1978; Castilla, 1983; Paskoff and Petiot, 1990; Vermeer and Castilla, 1991; Castilla, 1996;Correa et al., 1996a).

El Salvador copper mine is located approximately 120km from the coastal city of Chanorol. From1938 to 1975 about 150 million tones of untreated copper tailings generated at this mine (andpreviously at Potrerillos) were conducted through the Salado River (Fig.1) and dumped directly intothe sandy beach of Chanorol Bay, causing notorious beach progradations (Castilla, 1983; Paskoff and

C o p p e r

72

Petiot, 1990). In 1976 the dumping site at Chanorol was abandoned and a new dumping site wasestablished in a rocky beach known as Caleta Palito. This site is located approximately 10 km northfrom Chanorol Bay and between 1976 and 1989 received approximately 130 million tones of solidsfrom the mine.

In 1990 the El Salvador company built a tailing sedimentation dam away from the coast, in thedesertic Pampa Austral. Since then, the so called "aguas claras" or sediment-free "clear waters", witha legal upper limit of 2,000 µg/L of total copper, have been continuously driven from the dam,through the El Salado River bed, to Caleta Palito (Castilla, 1996). The clear water is pumped from thedam at a rate of 250-500 L/sec and mixes with water discharges from local towns and small-scale(artisanal) mining operations occurring along the El Salado River bed.

The 1976-1989 copper tailing dumping caused severe changes to the ecosystem in Caleta Palito(sandy and rocky shores), characterized by major increase of copper concentration in the sea water,beach progradation and the total elimination of invertebrates and most of the algal species around thedumping site. This lead to a reduction of biodiveresity (Castilla, 1983,1996) and the disruption oftrophic chains. Among the most striking features of the impacted areas is the almost absolutedominance of primary substratum by the green alga Enteromorpha compressa (L.) Grev., whichextends from Caleta Palito to Caleta La Lancha (Fig. 1) and occupies the entire intertidal fringe. Thisdominance became apparent shortly after the untreated tailings from El Salvador began pouring intoCaleta Palito, and has persisted since then.

We are using the El Salvador-Caleta Palito study case as a model, and in this context, the objectivesof this paper are to review the previous research on the Palito study case, to up-date information andto describe the future lines of research. This study includes: (a) the present concentration of copper incoastal sea water polluted and unpolluted sites, and in the sentinel green alga E. compressa; (b)differences and trends in species richness at polluted and unpolluted sites and (c) the characteristicsand mechanisms by which the green algae E. compressa may be able to resist copper enrichments.

2. Study sites and methods

The study sites are located in the desertic region of northern Chile. The El Salvador mine, at 2600 mabove sea level (26o 14' S, 69o 37' W), and the El Salvador town (approximately 10,000 habitants), are120km east from the coastal city of Chanorol Bay. The Salado River bed extends from Los AndesMountains through the Salado Valley (120km long) and connects with the sea at Chanaral Bay. Thescarcity of rains characteristic of the region results in a very limited water flux, mainly from ice fieldsin the higher mountains, most of which either evaporates or is absorbed before reaching the sea.

Floods, known as Bolivian winters, are caused by rain episodes occurring during the summeraccompanied by ice melting in the Andes. These floods result in large volumes of water and mudbeing channelled through the Salado River bed which are discharged directly in to Chanorol Bay.Because Bolivian winters are infrequent and of short duration, the Salado River remains dry most ofthe time with intermittent, brief flood episodes.

The towns of Diego de Almagro (ca. 8,000 inhabitants) and El Salado (ca. 1,200 inhabitants), as wellas around 20 small-scale artisanal mining operations present along the Del Salado Valley dischargetheir waste waters to the Salado River bed. These waste waters join the clear waters from the damabout 10 km before reaching Chanorol city; an artificial canal of approximately 15 km diverts thestream from the the Salado Valley and conducts the water directly to Caleta Palito where it dischargesopenly on the shore (Fig. 1).

C o p p e r

73

Figure 1: Map of the study area.

Sea water and E. compressa samples were collected in 1994, 1995 and 1996 around the pollutionpoint source at Caleta Palito and at three control sites: a) Caleta Pan de Azucar, approximately 24 kmnorth from Caleta Palito; b) Caleta Guanillo, approximately 45 km north from Caleta Palito and c)Caleta Zenteno, about 68 km south from Caleta Palito.

Sampling procedures and analytical methods for sea water and E. compressa were the same as thosedescribed by Correa et al. (1996a). Dissolved copper in water samples was quantified bypotentiometric stripping analysis in stationary solution, using a computerised Radiometer ISS 820analyzer. Algal tissue was treated with nitric acid and copper was quantified in a GBC 909 PBTatomic absorption spectrophotometer. Transmission electron microscopy was done with E. compressacollected at Caleta Palito and Caleta Zenteno. Processing of algal tissue was done according to Correaand McLachlan (1994).

Intertidal rocky shore monitoring of sessile and mobile invertebrates and macroalgae was done duringlow tides (0.23-0.30 m) between June 1-5, 1996 at Caleta Palito, Caleta Pan de Azucar, CaletaZenteno, La Lancha beach, and Caleta Coquimbo (see Castilla, 1996 and Fig. 1). The latter is a muchless impacted site, approximately 12 km north from Caleta Palito. Field monitoring was conducted intwo rocky platforms, with a slope of 10 - 40° and 30-40m long, selected in each site. The platformswere divided in four intertidal fringes: low, mid-low, mid-upper and upper, as described by Castilla(1981, 1996). Two independent observers recorded the species present in each fringe by walkingslowly and counting the species seen in intervals of one minute, with a maximum of five minutes perfringe.

3. Results and discussion

3.1 Copper concentration in sea water and Enteromorpha compressa

Values of dissolved copper in the water (Table 1) varied greatly among the different sampling datesand localities. At Caleta Palito values ranged from 10 µg/L in May 1996 to slightly more than 40 µg/Lin July 1995. Major fluctuations can be detected even within the same year. At the discharge point,for example, values of almost 20 µg/L were recorded in April 1996, a fifth of the concentrationmeasured at the same spot two months later. These fluctuations should be expected based on the high

C o p p e r

74

number and diversity of pollution sources occurring along the El Salado river (see Introduction). Inspite of that, copper concentration in the water from those localities not directly impacted by the ElSalado river (mainly the control sites of Zenteno and Guanillo) were consistently lower than those inwater from Caleta Palito. The lowest value of dissolved copper recorded among the control localitieswas 0.5 µg/L in water from Caleta Pan de Azucar (May 1996), whereas the maximum was 6.5 µg/L inwater from Caleta Zenteno (April 1995). On the other hand, in water from Caleta Palito the minimumvalue was 10.0 µg/L (April 1995) and the maximum was 203.3 µg/L (July 1995). The latter differencelikely reflects temporal changes in the degree of pollution of the Salado river inputs into the disposalpoint.

Fluctuations also occurred in the content of copper quantified in the tissues of E. compressa. In spiteof that, copper content was consistently higher in tissues from individuals collected at Caleta Palito(range: 34-358 µg g-1, April 1995) than in tissues from algae growing in Caleta Zenteno (highestvalue: 5.7µg/g, April 1995) and Guanillo (highest value: 30.6µg/g, April 1995).

Table 1: Values of dissolved copper in coastal waters and in Enteromorpha compressa from northernChile. Copper values in algal tissue are based on dry weight. Replicate number is indicated inparenthesis. Standards from the National Research Council of Canada were run simultaneously towater and tissue samples and copper values in the standards never deviated more than 2% from thecertified values.

Locality Date Water (µg/L) E. compressa (µg/g)

Caleto Palito1 August 1994 26.8 - 31.8(2)

54.9 - 71.8(2)

Caleta Zenteno August 1994 2.9 - 3.4(2)

1.9 - 2.1(2)

Pan de Azucar April 1995 0.9 - 1.5(4)

-

Guanillo April 1995 1.5 - 2.3(4)

26.6 - 30.6(2)

Caleta Zenteno April 1995 4.0 - 6.5(4)

4.7 - 5.7(3)

Caleta Palito1 April 1995 10.0 - 12.8(4)

34.9 - 358.3(3)

Caleta Palito2 April 1995 13.6 - 19.3(4)

-

Caleta Palito1 July 1995 37.7 - 40.7(3)

-

Caleta Palito2 July 1995 162.9 - 203.3(2)

-

Pan de Azucar May 1996 0.5 - 0.7(2)

-

Caleta Palito1 May 1996 10.2 - 13.9(2)

-

Caleta Zenteno May 1996 1.9 - 2.0(2)

-

1. Replicates collected at 50 m south from the discharge point.2. Replicates collected in front of the discharge point, at the mixture zone between the water from the canal and the sea.

3.2 Rocky intertidal species richness

The rocky intertidal species richness in the horizontal platforms of the five sites monitored rangedbetween 10 and 64 (Table 2). Caleta La Lancha, the most polluted site, has four species of algae

C o p p e r

75

(including a bio-film of diatoms), five species of mobile invertebrates and one vertebrate. No sessileinvertebrates were recorded. Caleta Palito, also a polluted site, shares ten species with Caleta LaLancha, but additionally shows eight more species. Caleta Coquimbo represents a site withintermediate species richness (35), and the control sites, Caleta P. de Azucar and Caleta Zenteno, arethe richest with 53 and 64 species respectively.

Table 2: List of species present (+) and absent (-). Presence assessed at four tidal levels using amaximum of 5 minutes recording time and non-destructive sampling.

Sites

Species Caleta Pan Caleta Caleta La Caleta Caletade Azucar Coquimbo Lancha Palito Zenteno

AlgaeColpomenia sinuosa - - - - +Codium dimorphum + - - - +Glossophora kunthii + + - - +Petalonia fascia + + - - +Lessonia nigrescenes + - - - +Scytosiphon lomentaria - - + + -Ectocarpus sp. + + + + -Ralfsia sp. + + - - +Lingbya sp. - - - - +Ulva sp. + + - + +Rama novaezelandiae + - - - +Cladophoropsis sp. - - - - +Enteromorpha compressa + + + + +Enteromorpha linza + - - - +Schottera nicaeensis + - - - -Porphyra columbina + + - - +Hildebrandtia lecannellieri + + - - +Polysiphonia sp. + + - + +Centroceras clavulatum - - - - +Ceramium sp. - - - - +Rhodoglossum sp. - - - - +Corallina officinalis + - - - +Corallina sp. + - - - -Gymnogongrus sp. - - - - +Ahnfeltia sp. - + - - +"lithothamnioides" nd. + - - - +Gelidium chilense + + - - +Gelidium lingulatum - - - - +Halopteris hordacea + + - - +Bangia sp. - + - - -Diatoms (biofilm) - + + + -

InvertebratesSessilesJehlius cirratus + + - + +Notochthamalus scabrosus + - - - +Balanus laevis + - - - +Austromegabalanus psittacus - - - + +

C o p p e r

76

Balanus flosculus + - - - +Semimytilus algosus - - - + +Perumytilus purpuratus + + - - -Phymactis clematis + + - - +Anthothöe chilensis + + - - +Phymanthea pluvia + - - - -Anthozoa nd. + - - - -Porifera nd. + - - - +Pyura chilensis + - - - +Serpulidae nd. - - - - +MobilesNodilittorina peruviana + - + + +Nodilittorina araucana + + - - +Scurria viridula + + - - +Scurria d'orbigni + + - - +Scurria plana + - - - -Scurria ceciliana + + + + +Scurria araucana - - + + +Scurria sp. + - + + +Scurria parasitica - + - - +Siphonaria lessoni + + - - +Fissurella crassa + + - - +Fissurella limbata + + - - +Fissurella maxima - - - - +Chiton granosus + + - - +Acanthopleura echinata + + - - +Enoplochiton niger - - - - +Concholepas concholepas - + - - +Tegula atra + - - - +Gastropoda nd. + - - - -Leptograpsus variegatus + + + + +Acanthocyclus gayi - - - - +Petrolistes violaceus - - - - +Taliepus sp. - + - - -Betaeus sp. + - - - -Amphipoda nd. + + - + +Acarii nd. - - - + +Heliaster helianthus + - - - +Patiria chilensis - - - - +Stichaster striatus + - - - +Tetrapygus niger + - - - +Loxechinus albus + - - - +Nemertini nd. - - - + -

VertebratesSicyases sanguineus + + - - -Microlophus sp. + + + + +

The pattern of species richness in the rocky intertidal platforms described in this paper coincides withthat reported by Castilla (1996) for perpendicular intertidal walls (approximately 70-85o) in the same

C o p p e r

77

geographical area. Nevertheless, as expected, the number of species is much greater in horizontalplatforms (range between 10-64, this paper) than in perpendicular walls (range between 6-32),Castilla, 1996). This is particularly true for mobile species of invertebrates.

The chlorophycean E. compressa remains as the dominant species covering the primary substratum(rock) on intertidal horizontal platforms at Caleta Palito and Caleta La Lancha. This alga is also foundin unpolluted sites (Table 2), but its cover values are extremely low (see also Castilla, 1996). CaletaPalito arose as an interesting monitoring site in this study. In June 1996 the rocky platforms of thissite presented four species (at low densities) not recorded before: the alga Ulva sp. (Chorophyta); thebarnacles Jehlius cirratus and Austromegabalanus psittacus; and the mussel Semimytilus algosus.They were not found at Caleta La Lancha. These species are abundant in unpolluted sites andtherefore, their presence could represent a further step in the direction of natural ecologicalrestoration occurring at Caleta Palito. The same process could eventually follow at Caleta La Lancha.

3.3 Enteromorpha compressa as a biological model: current status of knowledge andfuture perspectives

Species of Enteromorpha, particularly E. compressa, have been reported as organisms able to resistheavy metal enrichments, including copper. The mechanisms involved in that resistance, however, arepoorly understood. One of the few experimental studies available (Reed and Moffat 1983) addressedthe hypothesis that copper-enriched environments could lead to the development of copper-tolerantecotypes. Reed and Moffat (1983) showed higher copper tolerance in plants from a copper-enrichedenvironment (vessel hulls with copper-based anti-fouling paint) than those used as controls, obtainedfrom a site with no history of copper pollution. The presence of the same parental pattern of coppertolerance in the progeny lead to the conclusion that tolerance was an inherited character, a feature thatsupported the original notion of ecotype development.

These ideas were recently re-assessed by Correa et al. (1996a), who used the same experimentalapproach to compare the in vitro responses to copper enrichments of two intertidal populations of E.compressa. The population growing at Caleta Palito has been exposed to a consistently higher copperconcentration than the population of the same species growing in the non-polluted Caleta Zenteno.Inconsistencies between the results from the two studies became apparent. Chilean E. compressaobtained from spores of parents collected in copper-polluted (ie. Caleta Palito) and copper-normal (ie.Caleta Zenteno) localities did not show different tolerances to copper in laboratory assays. Theseresults clearly questioned the generalization by Reed and Moffat (1983) and, together with the resultsfrom the parental generation (Correa et al., 1996a), suggested physiological flexibility as theresponsible for the observed tolerance responses to the copper by the Chilean species.

Further analysis of the experimental information available in the literature indicates the existence ofimportant gaps in our understanding of E. compressa as a biological model in the field of algal-heavymetal interactions. At least two areas of research can be foreseen as important to fully unveil the basisfor the successful colonization and persistence of E. compressa in heavy metal-enriched coastalwaters. One of them relates to the possible occurrence of species-specific mechanisms for metaltolerance. A number of potential mechanisms have been individualized by various authors (see reviewby Correa et al., 1996b), although most of them are passive in the sense that they include metalcomplexation by either non-specific exudates or charged groups, which are normal components of thealgal cell walls, like polysaccharide (Veroy et al., 1980). It is unlikely, however, that these general,non-specific processes provide advantages in tolerance only to E. compressa to the extent ofdetermining the characteristic, almost absolute, monopoly of the primary substratum by this alga inpolluted areas. (Castilla, 1996). There is, on the other hand, a much more specific mechanism whichallows organisms to bind and detoxify metals occurring at concentrations higher than normal and itinvolves the synthesis of proteins known as metallothioneins. These proteins, which are encoded bynuclear DNA, have been found in a number of animals and plants. Their occurrence in algae is still amatter of discussion (Correa et al., 1996b) and certainly no information regarding metallothioneininduction in E. compressa (or in any species of Enteromorpha) is currently available. Indirect cellular

C o p p e r

78

evidence suggests, however, that this physiological process may be operating in the Chilean E.compressa. Recent unpublished ultrastructure observations have shown the occurrence ofintracellular, electron dense deposits in cells of E. compressa from Celeta Palito (Figures 2 and 3),whereas similar deposits are absent in cells from plants collected at Celeta Zenteno. Highly magnifiedimages of the above deposits have discarded their crystalline structure, but rather indicate a finelyreticular to homogeneous appearance, suggesting that a proteinaceous matrix may be serving as ametal trap. This hypothesis is further supported by the occurrence of structurally similar, metalcontaining proteinaceous bodies in microalgae exposed to heavy metals (Silverberg et al., 1976;Daniel and Chamberlain, 1981; Wong et al., 1994).

Figure 2: Cross sections through Enteromorpha compressa cells from Caleta Palito. Large electrondense granules (Gr) clearly extravacuolar, with a darker center and a pale periphery.

Figure 3: Spheric granules, with a compacted material homogeneously electron dense, occurringapparently within the vacuole.

The other area in need of experimental research relates to the shift of the ecological relationshipsamong species once an opportunistic alga, like E. compressa, becomes the dominant species. In thiscontext, the situation developed at Caleta Palito and surroundings represents a unique opportunity tocharacterize the species interplay, starting with E. compressa dominated system and moving into one

C o p p e r

79

where the primary space is shared at different, and fluctuating, ratios with other algae andinvertebrates. It must be understood that E. compressa is a normal component of the algal flora of theregion, but remains at very low densities, with a patchy distribution, usually restricted to intertidalpools and shaded walls. In Caleta Palito, on the other hand, E. compressa extends almosthomogeneously throughout the intertidal fringes of vertical and horizontal platforms. What isimportant to consider is that the original, long standing, catastrophic event (solid disposal creatingsandy tailing beaches and suspended particulate matter abrading the rocks) that was probably the maincause of the disappearance of algal and invertebrate species from the Palito-La Lancha polluted area(Fig. 1) is no longer present. Thus, the question remaining to be addressed is what are the factorspermitting the current dominance by the opportunistic E. compressa and what is the role of thisspecies in the eventual restoration process likely to be operating in the area.

Within the above framework, our 3-year research program involves a multidisciplinary approachwhere molecular biologists, marine microbiologists, analytical chemists, and marine ecologists arejointly tackling the deficiency areas, based on the situation developed in Caleta Palito. Thus, thesearch for copper binding proteins similar to those reported in some marine microalgae and higherplants is underway. We are conducting our search using wild E. compressa from Caleta Palito andseveral control sites, as well as thalli incubated at various copper concentrations in order to establishcausality between the presence of the metal at concentrations higher than normal and the synthesis ofspecific proteins. At the same time, pioneering work is being attempted to individualizemetallothionein encoding genes in nuclear DNA of E. compressa. In the same context, acomplementary set of analyses is being conduced to fully characterise the metal content of theintracellular electron dense granules developed in E. compressa from Caleta Palito. In this case,energy dispersive X-ray microanalysis is being implemented to either detect and quantify copper inthe granules, or elucidate the cellular compartment where the metal is accumulated.

In order to explain the ecological dominance by E. compressa, several avenues are currently beingtested. One of them involves testing the hypothesis that there are no propagules or larvae from otherorganisms in the coastal waters of Caleta Palito and therefore, E. compressa has no competitors orconsumers that may recruit and exert an effect upon its pattern of abundance and distribution. To testthe hypothesis, water samples are being collected seasonally, and algae developed after a pre-established culture period are recorded. For primary producers, there is yet another possibility, whichis that algae other than E. compressa are present as a bank of microscopic forms, unable to developdue to the presence of E. compressa. To test this, rock samples from Caleta Palito, without E.compressa, are being incubated under laboratory conditions to facilitate the development of otherspecies potentially present. As a complementary set of experiments, we are assessing the level ofcopper tolerance of the most common algae species which, according to their distribution rangeshould be present in the area of Caleta Palito. In these species, the effect of copper is not only testedon adult individuals, but on all the different stages through their life cycles, including alternatephases.

Finally, the sequential colonisation of algae or invertebrate species (ie. other than E. compressa) isbeing monitored along the impacted coastal area and various field manipulative experiments such asin situ bioassay with mussels and plantonic larvae studies are being implemented to assess their rolein the natural restoration processes.

Acknowlegements

We thank Nelson Lagos, Manuel Varas, Marco Ramirez and Jose Miguel Farina for their help withthe field work. This study is part of the research grant Nº 1196303-1, funded by the InternationalCooper Association (ICA), through the Centro de Investigaciones Minero Metalurgicas (CIMM), toDr J.A. Correa, Pontificia Universidad Catolica de Chile. JCC appreciates the financial support fromthe Corporacion del Cobre de Chile (COCHILCO) which permitted his attendance to the InternationalWorkshop on Copper, National Research Centre for Environmental Toxicology (20-21 June 1996),Brisbane, Australia.

C o p p e r

80

References

Anonymous (1995). Chilean mining compendium. Edited by R. Cortés. Editec Ltda., Santiago, Chile.pp. 531.

Castilla, J.C. and Nealler, E. (1978), ‘Marine environmental impact due to mining activities of ElSalvador Copper Mine’, Chile Mar. Poll. Bull., vol. 9, no3, pp. 67-70.

Castilla, J.C. (1981), ‘Perspectivas de investigacion en estructura y dinamica de comunidadesintermareales rocosas de Chile Central. II Depredadores de alto nivel trofico’, Medio Ambiente, vol. 5,pp. 190-215.

Castilla, J.C. (1983), ‘Environmental impacts in sandy beaches of copper mine tailings atChanaral,’Chile. Mar. Poll. Bull., vol. 14, no.2, pp. 459-464.

Castilla, J.C. (1996), ‘Copper mine tailing disposal in Northern Chile rocky shores: Enteromorphacompressa (Chlorophyta) as a sentinel species’, Environ. Monitor. Assess., vol. 40, pp.41-54.

Correa, J.A. and McLachlan, J. (1994), ‘Endophytic algae of Chondrus crispus (Rhodophyta). V. Finestructure of the infection by Acrochaete operculata (Chlorophyta)’, Eur. J. Phycol., vol. 29, pp. 33-47.

Correa, J.A., Gonzalez, P., Sanchez, P., Munoz, J. and Orellana, M.C. (1996a), ‘Copper-algaeinteractions: inheritance or adaptation?’, Environ. Mon. Assess., vol. 39, pp. 41-54.

Correa, J.A. Ramirez, M.A., Fatigante, F.A. and Castilla, J.C. (1996b), ‘Copper, macroalgae and themarine environment. The Chanaral case in northern Chile’. In: Proceedings of the Symposium on theBiology of Microalgae, Macroalgae and Seagrasses in the western Indian Ocean. M. Björk (ed.). Inpress.

Daniel, G.F. and Chamberlain, A.H.L. (1981), ‘Copper immobilization in fouling diatoms’, Bot. Mar.,vol. 24, pp. 229-243.

Pacyna, J.M., Scholtz, M.T. and Li, Y.-F.A. (1995), ‘Global budget of trace metal sources’. Environ.Rev., vol. 3, pp. 145-159.

Paskoff, R. and Petiot, R. (1990), ‘Coastal progradation as a by-product of human activity: anexample from Chanaral Bay, Atacama Desert, Chile’, J. Coastal Res. Special, no. 6, pp.91-102.

Reed, R.H. and Moffat, L. (1983), ‘Copper toxicity and copper tolerance in Enteromorpha compressa(L.)’, Grev. J. Exp. Mar. Biol. Ecol., vol. 69, pp. 85-103.

Silverberg, B.A. Stokes, P.M. and Ferstenberg, L.B. (1976), ‘Intranuclear complexes in a copper-tolerant green alga’, J. Cell Biol., vol. 69, pp. 210-214.

Vermeer, K. and Castilla, J.C. (1991), ‘High cadmium residues observed during a pilot study inshorebirds and their prey downstream from El Salvador copper mine, Chile’, Bull. Environ. Contam.Toxicol., vol. 46, pp. 242-248.

Veroy, R.L., Monta o, N., Guzman, L., Laserda, E.C. and Cajipe, G.J. (1980), ‘ Studies on the bindingof heavy metals to algal polysaccharide from Philippine seaweeds. I. Carrageenan and the binding oflead and cadmium’, Bot. Mar., vol. 23, pp. 59-62.

Wong, S.L., Nakamoto, L. and Wainwright, J.F. (1994), ‘Identification of toxic metals in affectedalgal cells in assays of waste waters’, J. Appl. Phycol., vol. 6, pp. 405-414.

C o p p e r

81

Wilson’s Disease after cloning of the geneDr John L GollanDirector Gastroenterology DivisionBrigham and Women’s HospitalAssociate Professor of Medicine, Harvard Medical School, Boston, Massachusetts U.S.A.

It is more than 80 years since the London neurologist Samuel Alexander Kinnier Wilson defined thefamilial syndrome of progressive lenticular degeneration associated with cirrhosis of the liver.Considerable advances have been achieved in elucidating the clinical, biochemical, genetic, andhistologic features, as well as the management of patients with this disease (Zucker and Gollan 1996).With the recent cloning of the Wilson's Disease gene, our understanding of the disease hasaccelerated, although clarification of the pathogenetic defect and the application of genetic screeningare among the major problems yet to be resolved.

1. Genetics

The gene for Wilson's disease is distributed worldwide, having been demonstrated in virtually allraces. Current estimates indicate that the prevalence of the disease is approximately 1 in 30,000 livebirths, with prevalence rates ranging from 15 to 30 per million (per annum). The gene frequencyvaries between 0.3 and 0.7%, corresponding to a heterozygote carrier rate of slightly greater than 1 in100.

Genetic studies from a large Israeli-Arab kindred identified a linkage between the Wilson's diseaselocus and the red cell enzyme esterase D, thereby establishing that the gene mutation responsible forWilson's disease was located on chromosome 13 (Frydman et al., 1985). Using multipoint linkagetechniques, the abnormal gene for Wilson's disease was localised more specifically to 13q14-q21. In1993, a candidate gene for Wilson's disease (WND) was reported independently by several differentgroups of investigators, using slightly different strategies for positional cloning (Bull et al., 1993,Petrukhin et al., 1993; Tanzi et al., 1993). The WND gene consists of a transcript of approximately7.5 kilobases, which is expressed primarily in liver, kidney and placenta; although it also has beendetected in heart, brain, lung, muscle and pancreas, albeit at much lower levels. the full-length cDNAsequence of the WND gene (Bull et al., 1995; Tanzi et al., 1993) predicts a 1,411 amino acid protein,which is a member of the cation-transporting P-type ATPase subfamily, highly homologous to theMenkes syndrome gene product and the copper transporting ATPase (cop A) found in copper resistantstrains of Enterococcus hirae. From sequence analysis of the cDNA, the WND protein is predicted topossess a metal binding domain (containing 5 specific binding sites), an ATP binding domain, acation channel and phosphorylation region, and a transduction domain responsible for the conversionof the energy of ATP hydrolysis to cation transport.

To date, more than forty disease-specific mutations in the Wilson's disease gene have been identified.The wide spectrum of clinical manifestations in Wilson's disease raises the question as to whethervariability exists at the molecular level. The fact that Wilson's disease is linked to chromosome 13markers in all populations studied suggests that there is a single genetic locus for the disease, and ithas been postulated that different mutations at that locus may explain the clinical variability. Indeed,physiological studies employing an animal model of Wilson's disease reveal that a single genemutation may inhibit copper transport at multiple locations within the cell. Hence, the variety ofmutations identified in the Wilson's disease gene potentially may affect copper transport to varyingdegrees, and at different cellular sites (Schilsky, 1994).

However, detailed genetic and epidemiological studies suggest that the variability in clinicalexpression observed in Wilson's disease patients may not be solely a consequence of allelicheterogeneity, since marked differences in presentation, age of onset and disease course have beenobserved in family members who have inherited two identical mutant alleles.

C o p p e r

82

2. DNA-based diagnosis

Developments involving the molecular genetics of Wilson's disease have provided a means for carrierdetection and early diagnosis (Sternlieb, 1993). In fact, several studies using haplotype analysis ofrelatives with closely linked markers have permitted precise carrier detection with less than 1 to 2%error. There also is a report of prenatal exclusion of Wilson's disease by analysis of DNApolymorphism in a chorionic-villous biopsy performed at nine weeks' gestation (Cossu et al., 1993).Unfortunately, the use of genetic techniques in the diagnosis of Wilson's disease has significantlimitations. Currently, DNA marker studies can be performed only within families, and undercircumstances where the diagnosis already has been established definitively in at least one familymember by standard biochemical methods. The index patient's DNA is then used as a reference torecognise the disease-carrying chromosomes in other members of the family. However, spontaneouschromosomal rearrangements can cause such markers to be uninformative, thereby limiting thediagnostic reliability.

These findings indicate considerable potential difficulties for DNA-based genetic screening, sincemost patients will possess alleles with two different mutations of the Wilson's disease gene (Schilsky,1994). Moreover, evaluation for Wilson's disease by DNA market analysis is generally performedonly in a few specialised centers and there is a delay before results are obtained. Given the rapidityand accuracy of biochemical analyses in establishing the diagnosis of Wilson's disease, as well as theaforementioned limitations of genetic testing, standard biochemical methods should continue to beutilised in the evaluation of the vast majority of suspected cases (Schilsky, 1994). The most likelyapplication of genetic linkage testing will be in the uncommon situation where biochemical methodsdo not provide a definitive answer, particularly under circumstances where patients have receivedprior chelation therapy. In addition, genetic screening of young family members of patients afflictedwith the disorder would facilitate early diagnosis and permit initiation of therapy in thepresymptomatic state.

3. Pathogenesis

It is postulated that the harmful effects of excess copper are mediated by the generation of free-radicals, which deplete cellular stores of glutathione and oxidise lipids, enzymes, and cytoskeletalproteins. Indeed, it has been shown that a number of intracellular systems are disrupted by elevatedcopper concentrations, including organellar membranes, DNA, microtubules, and various enzymesand proteins, although the principal cellular target of copper toxicity is unknown. In the earliest stagesof hepatocellular injury, ultrastructural abnormalities involving the endoplasmic reticulum,mitochondria, peroxisomes and nuclei have been identified (Sternlieb, 1990). These changes, inconjunction with diminished mitochondrial enzyme activities, may be important steps in thepathophysiologic events leading to lipid peroxidation and triglyceride accumulation in the hepatocyte(Sternlieb, 199o).

Wilson's disease patients exhibit impaired biliary excretion of copper, which is believed to be thefundamental cause of copper overload. Moreover, the prompt reversal of abnormal copper metabolismin Wilson's disease patients following orthotopic liver transplantation confirms that the primary defectresides in the liver.

It has been proposed that the Wilson's disease gene product is responsible for copper secretion fromthe liver cell, either across the canalicular (apical) membrane of the hepatocyte or into a subcellularcompartment that communicates with the bile canaliculus (Tanzi et al., 1993). The latter is consistentwith a putative endoplasmic reticulum, Golgi or lysosomal defect underlying the diminished biliaryexcretion and systemic accumulation of copper observed in patients with Wilson's disease.

In addition, in an animal model of Wilson's disease, the Long-Evans Cinnamon rat, excessive hepaticcopper accumulation occurs in the setting of diminished biliary excretion. These rodents exhibitimpaired entry of copper into the lysosomes, with normal delivery of lysosomal copper to the bile(Schilsky et al., 1994a). The Long-Evans Cinnamon (LEC) rat is a mutant strain of the Long-Evans rat

C o p p e r

83

which spontaneously develops fulminant hepatitis at 3 to 4 months of age, resulting in a 40%mortality rate. Surviving animals manifest chronic hepatic disease, low serum ceruloplasmin levels,and increased copper concentrations in the liver. Thus, the LEC rat shares many important clinical,biochemical, and histologic features with Wilson's disease, and the recent availability of this animalmodel likely will provide new insight into the pathogenesis of the human disorder.

4. Clinical features

The biochemical defect which leads to the accumulation of copper in Wilson's disease is present atbirth; however, clinical symptoms rarely are observed before the age of 5 years. the initial signs ofWilson's disease generally are detected in older children, adolescents, and young adults, although casereports have documented the clinical onset as early as 4 years. Wilson's disease patients typicallypresent with hepatic and/or neurologic dysfunction. In a large series of patients (Sternlieb, 1985), theinitial clinical manifestations were hepatic in 42%, neurologic in 34%, psychiatric in 10%,haematologic in 12%, and renal in 1%. Less commonly, patients present with skeletal, cardiac,ophthalmologic, endocrinologic or dermatologic symptoms (Table 1). Approximately 25% of thepatients have involvement of two or more organ systems at initial evaluation, although, with theadvent of aggressive screening, there has been a significant increase in the number of asymptomaticpatients diagnosed.

Table 1: Clinical manifestations of Wilson's disease

Hepatic - cirrhosis, chronic active hepatitis, fulminant hepatic failure

Neurologic - bradykinesia, rigidity, tremor, ataxia, dyskinesia, dysarthria, seizures

Psychiatric - behavioural disturbances, cognitive impairment, affective disorder, psychosis

Orhthalmologic - Kayser-Fleischer rings, sunflow cataracts

Hematologic - haemolysis, coagulopathy

Renal - renal tubular defects, diminished glomerular filtration, nephrolithiasis

Cardiovascular - cardiomyopathy, arrhythmias, conduction disturbances, autonomic dysfunction

Musculoskeletal - osteomalacia, osteoporosis, degenerative joint disease

Gastrointestinal - cholelithiasis, pancreatitis, spontaneous bacterial peritonitis

Endocrinologic - amenorrhoea, spontaneous abortion, delayed puberty, gynecomastia

Dermatologic - azure lunulae, hyperpigmentation, acanthosis nigricans

Adapted from Zucker SD, Gollan JL. In: Prieto J, Rodés J, Shafritz DA, eds. Hepatobiliary Diseases. Springer-Verlag, 1992:809, withpermission.

5. Hepatic manifestations

Hepatic involvement in Wilson's disease tends to manifest at a younger age (mean of 8 to 12 years)than does neurologic dysfunction, and is nonspecific, mimicking the features of a variety of acute andchronic liver diseases. Three major clinical patterns of liver disease are observed: cirrhosis, chronicactive hepatitis and fulminant hepatic failure. In the early asymptomatic phase of Wilson's disease, orin the presence of inactive cirrhosis, liver function tests may be normal or only minimally elevated. Inthe majority of cases, hepatic injury develops insidiously and, if untreated, pursues a chronic andrelentless course to cirrhosis.

As estimated 5 to 30% of patients with Wilson's disease exhibit clinical, biochemical and histologicfeatures similar to those observed in chronic active hepatitis (Scott et al., 1978; Schilsky et al., 1991).The diagnosis may be overlooked in these patients, since a significant percentage, almost 50% in oneseries (Scott et al., 1978), have no evidence of neurologic dysfunction or Kayser-Fleischer rings onophthalmologic examination. Serum cerulophasmin levels also may be normal in the setting of severehepatic inflammation. It has been estimated that Wilson's disease represents the underlying aetiologyin 5% of patients with idiopathic chronic active hepatitis who are under 35 years of age (Schilsky et

C o p p e r

84

al., 1991). A distinctive feature of Wilsonian chronic active hepatitis is the relatively modestelevations of serum aminotransferase levels in the presence of severe hepatocellular necrosis andinflammation (Schilsky et al., 1991).

More dramatically, Wilson's disease occasionally manifests as fulminant hepatic failure. Thesepatients may be indistinguishable from individuals with viral-induced hepatic necrosis, and many ofthe biochemical tests used to establish the diagnosis of Wilson's disease are abnormal in patients withother forms of fulminant hepatic failure (McCullough et al., 1983). The clinical features mostsuggestive of fulminant Wilsonian hepatitis include the presence of intravascular haemolysis,splenomegaly, and Kayser-Fleicher rings. Biochemical markers indicative of Wilson's disease includerelatively mild elevations in serum transaminases despite massive hepatic necrosis,hyperbilirubinemia with normal or low alkaline phosphatase levels, and a markedly elevated serumcopper concentration (McCullough et al., 1983).

The serum level of aspartate aminotransferase (AST) typically is higher than that of alanineaminotransferase (ALT), as a result of the associated haemolysis. Although uncommonly observed inWilsonian fulminant hepatic failure, Kayser-Fleischer rings are not pathognomonic, since theyoccasionally are seen in patients with other cholestatic hepatic diseases. The presence of severecoagulopathy and hypercupriuria are not useful in distinguishing Wilsonian from non-Wilsonianhepatic failure (McCullough et al., 1983). Liver biopsy with measurement of quantitative copper maybe helpful, although deranged clotting function may preclude this procedure, or necessitate thetransjugular approach. If a biopsy specimen is obtained, histologic evidence of cirrhosis(predominantly micronodular) in a young patient with fulminant hepatitis is suggestive of Wilson'sdisease, as is an elevated hepatic copper content.

Wilson's disease patients with acute hepatic failure tend to be young and to have a fulminant clinicalcourse, with survival generally no longer than days to weeks unless hepatic transplantation isperformed (McCullough et al., 1983; Mowat 1987). Medical treatment frequently is unsuccessful,particularly when the disorder is associated with haemolysis and renal insufficiency. Whentransplantation is unavailable for patients it remains imperative to make the diagnosis of Wilson'sdisease for the purpose of aggressive medical therapy and family screening.

6. Hepatic pathology

Abnormal liver histology is evident in biopsy specimens from asymptomatic Wilson's disease patientswithin the first decade of life. The earliest changes detectable on light microscopy include glycogendeposition in the nuclei of periportal hepatocytes, and moderate fatty infiltration. the lipid droplets,which are composed of triglycerides, progressively increase in number and size, in some casesresembling the steatosis induced by ethanol.

The rate of progression of the liver histology from fatty infiltration to cirrhosis is variable, although ittends to occur by one of two general processes, either with or without hepatic inflammation. SomeWilson's disease patients develop a histologic picture that is indistinguishable from chronic activehepatitis (Scott et al., 1978; Schilsky et al., 1991). Pathologic features include mononuclear cellinfiltrates, which consist mainly of lymphocytes and plasma cells, piecemeal necrosis extendingbeyond the limiting plate, parenchymal collapse, bridging hepatic necrosis and fibrosis. If untreated,this may evolve into macronodular cirrhosis or progress rapidly to fulminant hepatitis. Thedevelopment of cirrhosis also may occur in the absence of significant parenchymal inflammatoryinfiltrate or necrosis. It is notable that the vast majority of Wilson's disease patients have evidence offibrosis on liver biopsy, despite widely varying levels and patterns of hepatic inflammation and injury.

Hepatocellular carcinoma is uncommonly associated with Wilson's disease, in contrast tohemochromatosis. It has been proposed that the diminished cancer risk is due to the relative dearth ofan inflammatory component in the development of most cases of Wilsonian cirrhosis. Indeed, animalstudies suggest that copper may exert a protective effect against the development of malignancy. Onthe other hand, LEC rats exhibit a high incidence of hepatocellular carcinoma in the setting of markedhepatic copper accumulation (Sokol, 1994), and this neoplastic potential is effectively abrogated by

C o p p e r

85

the administration of D-penicillamine. These observations have led to speculation that the lowincidence of hepatocellular carcinoma in Wilson's disease patients is attributable to chelation therapy,as long-term survival is rare in untreated patients.

The histochemical staining of liver biopsy specimens for copper is of little diagnostic value in patientswith Wilson's disease. This is due to the fact that during the initial stages of copper accumulation, themetal is distributed diffusely in the cytoplasm and frequently is undetectable by rhodamine orrubeanic acid staining.

Orcein, which is believed to stain polymerised metallothionein sequestered in lysosomes, exhibits acharacteristic granular pattern in only half of patients with early Wilson's disease. As the diseasebecomes more advanced, copper is sequestered within hepatocyte lysosomes and is detectable byroutine histochemical techniques, even though tissue concentrations actually are lower than in earlierstages of the disorder. In contradistinction to Wilson's disease, other conditions in which hepaticcopper is elevated (eg., primary biliary cirrhosis, sclerosing cholangitis, biliary atresia, intrahepaticcholestasis of childhood, Indian childhood cirrhosis, and the normal neonate) are nearly alwaysassociated with stainable copper. Due to the insensitivity of copper staining techniques, time-dependent changes in the distribution of copper within the liver cell, heterogeneity of hepatic copperdeposition, and the lack of specificity of hepatocyte copper granules, histochemical staining forcopper is unreliable in establishing the diagnosis of Wilson's disease.

7. Laboratory diagnosis of Wilson's Disease

The simplest screening procedure includes a slit-lamp examination of the eyes, and measurement ofserum ceruloplasmin and transaminase (ALT, AST) levels. If Kayser-Fleischer rings are present onophthalmologic examination and ceruloplasmin levels are below 20 mg/dL in a patient withneurologic signs or symptoms, the diagnosis of Wilson's disease is established. If a patient isasymptomatic, exhibits isolated liver disease, or lacks corneal rings, the coexistence of a hepaticcopper concentration above 250 µg/g dry weight and a low serum ceruloplasmin level also issufficient to make the diagnosis.

7.1 Serum ceruloplasmin

The normal serum concentration of ceruloplasmin is 20 to 40 mg/dL and, while levels are low in thehuman newborn, they gradually rise during the first two years of life, coincident with the postnataldecline in hepatic copper concentration. Although a decreased ceruloplasmin level per se is notdiagnostic of Wilson's disease, approximately 90% of all patients, and 85% of individuals presentingwith hepatic manifestations of the disease have levels of this glycoprotein that are below the normalrange. Hypoceruloplasminemia occasionally may occur in other hepatic conditions, such as fulminantnon-Wilsonian hepatitis, as a consequence of diminished hepatic synthetic function. Patients withnephrotic syndrome, protein-losing enteropathy, malabsorption, or severe malnutrition also maymanifest low serum ceruloplasmin levels, although there usually is no diagnostic difficulty in thesecases.

Difficulty may arise with regard to the 10% of heterozygous carriers of the gene for Wilson's diseasewho manifest diminished serum levels of ceruloplasmin, yet never develop clinical symptoms or signsof the disease. These individuals, who represent approximately 1 in 2,000 persons in the generalpopulation, may present a difficult diagnostic dilemma if they fortuitously develop chronic activehepatitis or cirrhosis (or another aetiology), thereby mimicking the clinical, biochemical andhistological features of Wilson's disease.

Normal ceruloplasmin concentrations are found in up to 15% of patients with Wilson's disease andactive liver involvement (Scott et al., 1978). This presumably is due to increased hepatic synthesis andrelease of the glycoprotein in response to hepatic inflammation, as ceruloplasmin is an "acute phasereactant". The ceruloplasmin concentration declines to the low levels typically associated withWilson's disease as the inflammatory activity in the liver abates. A further reduction in theceruloplasmin level generally follows the initiation of chelation therapy. Plasma concentrations also

C o p p e r

86

are influenced by a variety of humoral and hormonal agents, and elevated estrogen levels, secondaryto pregnancy or exogenous administration, occasionally may elevate previously low ceruloplasminlevels into the normal range.

7.2 Urinary copper excretion

The urinary excretion of copper is greater than 100 µg/24 hours (normal: <40 µg/24 hours) in mostpatients with symptomatic Wilson's disease, reflecting increased serum levels of the readily filterablefraction of non-ceruloplasmin copper. In patients with fulminant hepatic necrosis due to Wilson'sdisease, urinary excretion of the metal may exceed 1,000 µg/24 hours, as hepatic copper stores arereleased into the systemic circulation. Unfortunately, the measurement of urinary copper often ismisleading due to inaccuracies in collection and laboratory analysis, and care must be taken to usecopper-free containers for storage of the urine samples. It also should be noted that asymptomaticWilson's disease patients do not necessarily exhibit elevated urinary copper concentrations.

Moreover, urinary copper levels may be elevated in a variety of other hepatic disorders includingcirrhosis, chronic active hepatitis, and cholestatic disorders such as primary biliary cirrhosis. Thus,the quantification of urinary copper is of little value as a screening test for Wilson's disease; althoughit may be useful as a means of confirming the diagnosis and in evaluating compliance and theresponse to chelation therapy.

7.3 Hepatic copper concentration

If Kayser-Fleischer rings or neurologic abnormalities are absent, a liver biopsy for quantitative copperdetermination is essential to establish the diagnosis of Wilson's disease. Care must be taken to insurethat the biopsy needle and specimen container are free from coper contamination. The normal hepaticcopper concentration varies from 15 to 55 µg/g (0.24 - 0.87µmol/g) dry liver. Virtually all untreatedpatients with Wilson's disease have elevated hepatic copper levels, ranging from 250 to as high as3,000 µg/g dry liver. Values below 250 µg/g usually are attributable to the irregular distribution ofcopper in the liver, particularly in the presence of cirrhosis, when small fragmented biopsy samplesare obtained.

The finding of a normal hepatic copper concentration effectively excludes the diagnosis of untreatedWilson's disease. However, an elevated liver copper level alone is insufficient to establish thediagnosis of Wilson's disease, since concentrations above 250 µg/g may be found in other chronichepatic disorders (mostly cholestatic), including primary biliary cirrhosis, primary sclerosingcholangitis, extrahepatic biliary obstruction or atresia, chronic active hepatitis, intrahepaticcholestasis of childhood and Indian childhood cirrhosis. These patients are readily distinguished fromthose with Wilson's disease on the basis of history, physical findings and biochemical testing.Moreover, in the great majority of individuals with prolonged cholestasis, serum ceruloplasminconcentrations are either normal or increased.

7.4 Incorporation of orally administered radiocopper into ceruloplasmin

Rarely, when a diagnostic dilemma remains or liver biopsy is contraindicated (eg. severecoagulopathy), the radiocopper loading test may be useful (Sternlieb and Scheinberg 1979). Serumradioactivity is measured at 1, 2, 4 and 48 hours after oral administration of the radionuclide (2 mgcupric acetate containing 0.3 to 0.5 mCi of 64Cu, mixed in 100 to 150 mL fruit juice or ginger ale). Inhealthy subjects and in patients with hepatic disorders that mimic Wilson's disease, the plasmaconcentrations of radiocopper rise rapidly, are maximal within one to two hours, and then fall and riseagain over the ensuing 48 hours, as the non-ceruloplasmin bound radiocopper is incorporated intonewly synthesised ceruloplasmin in the liver and released into the circulation. Wilson's diseasepatients, on the other hand, incorporate little or no radiocopper into nascent ceruloplasmin, even in thepresence of normal ceruloplasmin concentrations. Heterozygotes have a pattern of incorporation thatis intermediate between that of Wilson's disease patients and healthy individuals.

C o p p e r

87

7.5 Abnormal imaging

Despite advances in computed tomography (CT) and magnetic resonance imaging (MRI), theseradiologic modalities have little to offer in the diagnosis or evaluation of hepatic involvement inWilson's disease. While evidence of chronic liver disease (eg., splenomegaly, heterogeneous liverparenchyma, varices) may be identified in patients with advanced disease, these findings are neitherspecific nor sensitive for Wilson's disease.

8. Diagnostic screening

The diagnostic approach to Wilson's disease must be tailored according to the clinical presentation. Itmust be emphasised that, in the absence of definitive DNA haplotype analysis, the diagnosis ofWilson's disease should not be based on the results of an individual laboratory test, and can only beestablished in the setting of confirmatory clinical and biochemical data. Patients with neurologic orpsychiatric manifestations should undergo slit-lamp examination of the eyes and serum ceruloplasmindetermination. The documentation of Kayser-Fleischer rings and a low serum ceruloplasminconcentration is sufficient to establish the diagnosis, which can be confirmed by the presence ofincreased 24 hour urinary copper excretion. A liver biopsy with quantification of hepatic copper isessential if either (1) Kayser-Fleischer rings are absent (in order to exclude the possibility that thepatient is heterozygous for the gene), or (2) if ceruloplasmin levels are normal (as occurs in up to 15%of cases).

It is imperative that all first-degree relatives be screened, particularly siblings. Wilson's disease maybe clinically silent even in the presence of significant organ damage; hence, a delay in the diagnosis orin the initiation of therapy may lead to irreversible hepatic and/or neurologic injury. Biochemicalscreening of children should not be performed prior to 3 or 4 years of age. The evaluation shouldconsist of a careful history and physical examination, serologic tests of liver function, a slit-lampexamination of the eyes, and a serum ceruloplasmin level, with liver biopsy and quantitative hepaticcopper determination reserved for diagnostic dilemmas. Once the diagnosis of presymptomaticWilson's disease is established, lifelong chelation therapy should be commenced immediately.

9. Treatment

9.1 Diet

The ubiquitous presence of copper in most foodstuffs makes stringent dietary copper restrictionimpractical, although it is suggested that patients avoid eating foods with a high copper content, suchas liver, chocolate, nuts, mushrooms, legumes, and shellfish. Some authors also recommend the use ofdeionised or distilled water if the copper content of the patient's home drinking water exceeds 0.2ppm, particularly during initial therapy. The use of domestic water softeners should be avoided, sincethese may substantially increase copper concentrations.

9.2 Pharmacologic therapy

9.2.1 PenicillamineOver the past three decades, it has been well documented that oral D-penicillamine results in completereversal or alleviation of hepatic, neurologic and psychiatric abnormalities in most patients withWilson's disease, and this drug remains the "gold-standard" therapy for this disorder. The key to asuccessful outcome is early diagnosis and treatment, and clinical disease can be prevented indefinitelyin asymptomatic patients, provided that they adhere to continuous maintenance therapy. Someindividuals demonstrate a dramatic response within weeks of initiating D-penicillamine, while othersmay exhibit no clinical improvement, or even temporary neurologic deterioration, for several months.

The precise mechanisms of action of D-penicillamine remains controversial. Although the logic forthe use of this drug in the treatment of Wilson's disease is based on its in vitro copper chelatingproperties, there is conflicting evidence regarding the ability of penicillamine to "decopper" the liverand other organs. It has been proposed that D-penicillamine also detoxifies the liver by sequestering

C o p p e r

88

intracellular copper in an innocuous state, either through the direct formation of copper complexes orby the induction of metallothionein synthesis. The fulminant decompensation observed in previouslycompliant patients who discontinue D-penicillamine therapy for a relatively brief period of time offerssupport for this hypothesis. Other postulated mechanisms of action of D-penicillamine includeinhibition of collagen cross-linking, enhancement of intracellular levels of reduced glutathione, andsuppression of inflammation via effects on leukotriene and prostaglandin metabolism.

The standard dose of D-penicillamine is 1 to 2 g daily given orally in four divided doses 30 minbefore meals, although as much as 4 g/day can be administered to critically ill patients for briefperiods of time. It is best taken on an empty stomach since food reduces its absorption. Most of theexcess liver copper appears to be mobilised within the initial year of therapy, with urinary copperexcretion approaching 2 to 5 mg/day during this period. After several years of treatment, urinarycopper exretion levels decline to approximately 0.5 to 1.0 mg daily, as hepatic copper concentrationsapproach normal levels. When symptoms have largely abated and a stable clinical course has beenachieved, the maintenance dose of D-penicillamine may be reduced to 1 g daily.

A variety of adverse effects of D-penicillamine have been recognised (Zucker and Gollan 1996;Sternlieb and Scheinberg, 1985), although serious complications necessitating discontinuation of thedrug are infrequent. Thus, D-penicillamine has been demonstrated to be effective and safe for use inthe treatment of Wilson's disease, and remains the first-line drug in this disorder.

9.2.2 TrientineTrientine (triethylene tetramine dihydrochloride) was introduced in 1969 as an alternative chelatingagent for cases in which serious toxic reactions to D-penicillamine occur. It has been well-establishedthat 1 to 2 g administered orally in 3 divided doses induces negative copper balance and effectsclinical improvements in patients with Wilson's disease. As with penicillamine, trientine should beadministered prior to meals, since food interferes with absorption. The exact mechanism of action ofthis drug remains unknown, although it has been shown both to enhance urinary copper excretion andto decrease intestinal copper absorption. Unlike D-penicillamine, trientine causes the serum copperconcentration to rise during cupriuresis, suggesting that the two agents may mobilise copper fromdifferent systemic pools. Most of the toxic side-effects necessitating conversion from D-penicillamineto trientine typically subside while on trientine.

The exception is elastosis perforans serpiginosa, which may progress in some patients. Symptoms ofskin rash, gastrointestinal distress and rhabdomyolysis have been reported in patients with Wilson'sdisease or primary biliary cirrhosis treated with trientine. Sideroblastic anaemia is the only majorside-effect attributed to this medication. Although trientine appears to cause minimal toxicity, it hasless of a cupriuretic effect than D-penicillamine, and hence, is currently not recommended as primarytherapy.

9.2.3 ZincThe principal mode of action of zinc is postulated to be via the induction of intestinal metallothioneinsynthesis, which results in the sequestration of copper in intestinal epithelial cells, thereby preventingabsorption into the portal circulation and enhancing faecal copper excretion. Zinc also may directlyexhibit a protective effect on hepatocytes by inducing the synthesis of metallothionein in these cells.A minimum of 75 mg of zinc sulfate or zinc acetate per day, administered in 2 divided doses betweenmeals, can maintain neutral or negative coper balance, although most studies have administered 50 mgthrice daily. Common side-effects of oral zinc include headaches and gastrointestinal upset, theincidence of which may be reduced by the use of zinc acetate, rather than zinc sulfate. Although long-term followup is limited, major complications have not been reported with the use of zinc for thetreatment of Wilson's disease.

Due to the slower onset of action as compared with other chelating agents, zinc is not recommendedfor the initial treatment of symptomatic Wilson's disease. In fact, most of the published experiencewith zinc therapy has been in patients who previously had been decoppered with D-penicillamine. Ithas been suggested that zinc may be useful for presymptomatic or pregnant patients, as well as for

C o p p e r

89

maintenance therapy in individuals who have previously been decoppered with chelating drugs.However, there are data indicating that hepatic copper may continue to accrue in some Wilson'sdisease patients treated with zinc sulfate alone (Zucker and Gollan, 1996). Additionally, long-termfollowup studies are needed to determine whether the de-coppered state is sustained with zinctherapy, with monitoring for untoward side-effects. Based on the information currently available, oralzinc should be used as a third-line therapy in the rare patients who develop intolerance to both D-penicillamine and trientine. There appears to be no synergistic effect of zinc in combination with anadditional chelating agent; hence the concomitant administration of penicillamine or trientine and zincis not recommended.

9.2.4 ThiomolybdatesThiomolybdates appears to lower systemic copper levels by complexing lumenal copper, and therebyinhibiting intestinal absorption. In addition, the portion of the drug that is systemically absorbed maybind excessive serum copper and render it less available for cellular uptake, ultimately resulting in theremoval of copper from intracellular stores. Moreover, in contrast to penicillamine and trientine,thiomolybdates exhibit a higher affinity for copper than metallothionein in vitro, suggesting thatpotentially this drug may be able to more effectively remove copper from the cell. Limited studies ofammonium tetrathiomolybdate (60 to 100 mg daily in two divided doses) in Wilson's disease patientsin whom D-penicillamine and/or trientine was poorly tolerated or ineffective demonstrated the drug tobe highly successful in lowering hepatic copper concentrations (Zucker and Gollan, 1996). Additionaltrials in a small number of patients have supported the efficacy and safety of ammoniumtetrathiomolybdate in the initial treatment of Wilson's disease, and have further suggested that thismedication is less prone to precipitate the neurologic decompensation observed with other chelatingagents. Although these results appear promising, thiomolybdates have caused bone marrowsuppression. Thus, further investigation is required before the routine use of this drug can berecommended.

9.3 Long-term management

Lifelong chelation therapy, without interruption, is necessary in all Wilson's disease patients.Cessation of therapy may result in rapid and irreversible hepatic and neurologic deterioration. In astudy of 11 patients who discontinued treatment, 8 patients died of fulminant hepatitis after anaverage survival of only 2.6 years. Thus, it is imperative that an alternative agent be administered toany patient who is unable to continue D-penicillamine due to adverse effects (Scheinberg 1987).

9.4 Liver transplantation

Despite advances in medical therapy, significant mortality rates are still observed in specific subsetsof patients with Wilson's disease. These individuals, in whom orthotopic liver transplantation hasproven most successful, generally present with acute fulminant hepatic failure associated withhemolysis and hypercupremia (either as the initial presentation or due to poor compliance withmedical therapy), or with advanced cirrhosis and hepatic insufficiency, unresponsive to an adequatetrial of chelation therapy and supportive measures. In the absence of severe hepatic disease, livertransplantation generally is not recommended for the management of refractory extrahepaticmanifestations, such as neurologic deterioration. In a series of 55 patients with Wilson's disease whounderwent hepatic transplantation, a 79% one-year survival was observed, with an overall survivalrate of 72%, at 3 months to 20 years (Schilsky et al., 1994b). Transplant recipients uniformly manifestcomplete reversal of the underlying defects in copper metabolism, and demonstrate significantimprovement in a variety of symptoms and signs of the disease.

C o p p e r

90

References

Bull, P.C., Thomas, G.R., Rommens, J.M., et al. (1993), ‘The Wilson's disease gene is a putativecopper transporting P-type ATPase similar to the Menkes gene’, Nature Genet, vol.5, p. 327.

Cossu, P., Pirastu M., Nucaro, A., et al. (1993), ‘Prenatal diagnosis of Wilson's disease by analysis ofDNA polymorphism’, N Engl J Med., vol. 327, p. 57.

Frydman, M., Bonne-Tamir, B., Farrer, L.A., et al. (1985), ‘Assignment of the gene for Wilson'sdisease to chromosome 13: linkage to the esterase D locus’, Pro Natl Acad Sci. USA., vol. 82, p. 1819.

McCullough, A.J., Fleming, R., Thistle, J.L., et al. (1983), ‘Diagnosis of Wilson's disease presentingas fulminant hepatic failure’, Gastroenterology, vol. 84, p. 161.

Mowat, A.P. (1987), ‘Liver disorders in children: the indications for liver replacement inparenchymal and metabolic diseases’, Transplant Proc., vol. 19, p. 3236.

Petrukhin, K., Fischer S.G., Pirastu, M., et al. (1993), ‘Mapping, cloning and genetic characterizationof the region containing the Wilson disease gene’, Nature Genet., vol. 5, p. 338.

Scheinberg, I.H., Jaffe, M.E and Sternlieb, I. (1987), ‘The use of trientine in preventing the effects ofinterrupting penicillamine therapy in Wilson's disease’, N Engl J Med., vol. 317, p. 209.

Schilsky, M.L., Scheinberg, I.H and Sternlieb, I. (1991), ‘Prognosis of Wilsonian chronic activehepatitis’, Gastroenterol., vol. 100, p. 762.

Schilsky, M.L., (1994), ‘Identification of the Wilson's disease gene: clues for disease pathogenesisand the potential for molecular diagnosis’, Hepatology, vol. 20, p. 529.

Schilsky, M.L., Stockert, R.J and Sternlieb, I. (1994a), ‘Pleiotropic effect of the LEC mutation: arodent model of Wilson's disease’, Am J Physiol., vol. 266, p. G907.

Schilsky, M.L., Scheinberg, I.H and Sternlieb I. (1994b), ‘Liver transplantation for Wilson's disease:indications and outcome’, Hepatology, vol. 19, p. 583.

Scott, J., Gollan, J.L., Samourian, S., et al. (1978), ‘Wilson's disease, presenting as chronic activehepatitis’, Gastroenterol. vol. 74, p. 645.

Sokol, R. J. (1994), ‘At long last: an animal model of Wilson's disease’, Hepatology, vol. 20, p. 533.

Sternlieb, I., Scheinberg, I.H (1979), ‘The role of radiocopper in the diagnosis of Wilson's disease’,Gastroenterol, vol. 77, p. 138.

Sternlieb, I. and Scheinberg, I.H. (1985), ‘Wilson's disease’. In: Wright R, Millward-Sadler GH,Alberti KGMM, et al, eds. Liver and Biliary Disease. London, W.B. Saunders Company : 949.

Sternlieb, I. (1990), ‘Perspectives on Wilson's disease’, Hepatology, vol. 12, p. 1234.

Sternlieb, I. (1993), ‘The outlook for the diagnosis of Wilson's disease’, J. Hepatol. vol.17, p. 263.

Tanzi, R.E., Petrukhin, K., Chernov, I., et al. (1993), ‘The Wilson disease gene is a coppertransporting ATPase with homology to the Menkes syndrome gene’, Nature Genet. vol. 5, p. 344.

Zucker, S.D, Gollan, J.L. (1996), ‘Wilson's disease and hepatic copper toxicosis’. In: Zakim D, BoyerTD, editors. Hepatology: A Textbook of Liver Disease. Philadelphia: Saunders: 1405.

C o p p e r

91

Indian Childhood Cirrhosis (ICC) - RevisitedDr Seema SethiDepartment of PathologyLady Hardinge Medical College & Smt.Sucheta Kripalani Hospital, New Delhi, India

1. Background

Indian childhood cirrhosis (ICC) was once the fourth most common cause of death in preschoolchildren in India (Sub-committee of the Indian Council of Medical Research, 1955). Despite itsrecognition for over a century now, its etiopathogenesis is still an enigma. Much work has been doneon ICC but every study seems to be but a drop in the vast ocean of this disease. True is the remarkabout ICC, "I see sea". It is unusual to find a disease confined to one geographical area. ICC is onesuch disease. Whether cases described from other countries are similar aetiologically to ICC is as yetunclear.

A disease of childhood, most cases occur between 6 months and 3 years. However, it can occur up to5 years of age (Bhave et al., 1982; Sethi et al., 1993). A male:female ratio of 3:2 has been reported(Sethi et al 1993). A high rate of parental consanguinity and up to 22% of incidence in siblings isknown (Pandit and Bhave, 1996). Although various aetiological factors have been proposed in ICCnone has been confirmed so far. Toxins, nutritional factors, infections, alpha-1-fetoprotein, alpha-antitrypsin, disturbances in disaccharide tolerance, abnormality of tryptophan - niacin pathway andvarious trace metals like zinc selenium, cadmium, manganese, magnesium and copper have all beenimplicated as causes of ICC (Sub-commitee of the Indian Council of Medical Research, 1955).

2. Copper and ICC

The high hepatic copper content in ICC and early introduction of copper-contaminated animal milkled to the hypothesis that copper is an aetiological agent in ICC (Portmann et al., 1978; Tanner et al.,1979). Traditionally, milk and water are boiled and stored in brass (an alloy of 70% copper and 30%zinc) vessels in India. Milk thus boiled and stored in brass vessels has a sixty fold increase in copperconcentration and water about six times (Tanner et al., 1983). Not all patients with ICC received milkwhich had been stored in brass vessels (Sharda and Bhandari, 1984). In one prospective study nohistory of use of brass vessels was seen in 56% of cases (Table 1) (Sethi et al., 1993). Occurrence ofICC has been reported in bottle rather than breast-fed children (Tanner et al.,1983). However, ICC hasbeen noted in children breast-fed 6-9 months (Sethi et al., 1993). Other family members and siblingsreceiving milk from the same source as ICC cases were found to have normal serum and urinarycopper levels (Sharda and Bhandari, 1984).

Whilst ingestion of large amounts of copper in early infancy may be a factor in the aetiology, it cannotfully explain the disease. In a study conducted in India on a group of 32 children who developed ICC,all had a significantly higher mean serum copper values in comparison with the age-matched controlgroup (Table 2) (Sethi et al., 1993). In another 82 children suffering from ICC liver biopsies revealedraised liver copper concentrations. The liver copper concentrations increased with the severity of thedisease (Table 3) (Sethi et al.,1993). Variable serum and hepatic content in the same stage of thedisease has been explained on the basis of genetic heterogeneity (Sethi et al.,1993). On the basis offamilial occurrence and high consanguinity a genetic aetiology is ICC had been suspected (Agarwal etal., 1979). A pedigree analysis compatible with an autosomal recessive inheritance has been reported(Chandra, 1976). It has been hypothesised that ICC is a genetic alteration in the copper metabolismperhaps related to the binding of copper with metallothionein (Sethi et al., 1993).

C o p p e r

92

Table 1 : Use of brass utensils in families of 32 ICC cases

Only brass andcopper utensils Mixed use

Number use ofbrass utensils

Number of cases 5 9 18

Percentage 16 28 56

Mean Serum coppervalues (µg/dL)

142 148.2 123.5

Table 2 : Serum copper levels in ICC

Serum Copper No. Mean (µg/dL) Range (µg/dL)

Control 10 96* 64-120

ICC 32 137.27* 70-240

Stage I 7 123.2 88-172

Stage II 17 146.5 70-240

Stage II 8 187 164-210

* p<0.01.

Table 3 : Correlation of histological grades and orcein grades in ICC

Orcein Grade

Histological Grade 1 2 3 4 Total

I 4 4 5 1 14

II 2 20 20 4 46

III 1 3 12 6 22

Total 7 27 37 11 82

3. Clinical features

ICC is generally insidious in onset. The clinical presentation of the disease is divided into 3 stages(Sub-committee of the Indian Council of Medical Research, 1955). In the early stage the symptomsare non-specific like abdominal distension, irregular fever, excessive crying and altered appetite.Jaundice is generally a late feature but may rarely be seen early in the course of the disease. Mostpatients present in the intermediate or "classical" stage of ICC with jaundice, loss of appetite,distended abdomen with a characteristic firm to hard liver with a "leafy" edge. The child is prone tosecondary infections at this stage. The progress is generally relentless and within a few months thechild develops the terminal stage with oedema, ascites, splenomegaly and haemorrhagiccomplications. The liver at this stage may be less palpable than the second stage.

The disease generally runs this course and within 6-18 months the patient dies due to hepato-cellularfailure leading to haemorrhage and coma or due to intercurrent infections. Spontaneous recoveryvariously reported from 13% (Sur and Bhatti, 1978) to 30% (Sub-committee of the Indian Council ofMedical Research, 1955) in early stages of ICC has been reported but the cause is not known.

C o p p e r

93

Anaemia is commonly seen in patients of ICC. The standard liver function tests are usually derangedbut not specific for differentiation of early ICC from other childhood liver disorders. Serum copper israised significantly in ICC. Mean copper values rise with the clinical progression of the disease (Sethiet al., 1993; Tanner et al., 1979; Sharda and Bhandari 1984). Serum ceruloplasmin levels are however,normal or raised in contrast to Wilson's disease. Hepatic copper levels are raised. A hepatic copperlevel >800 µg/g dry weight helps to distinguish ICC from other liver disorders occurring at this stage(Pandit and Bhave 1996).

4. Diagnosis

Histopathology remains the corner stone of definitive diagnosis of ICC (Pandit and Bhave 1996). Thetwo diagnostic features are : (1) widespread, coarse, dark brown orcein deposits and (2) intralobular,pericellular fibrosis (Pradhan et al 1983). Hepatocytic necrosis (97%) and hyaline (66%) arediagnostic though late features. Portal fibrosis, inflammation and disruption of the limiting plate areseen in most cases but are not specific for ICC. Parenchymal fat is usually absent and cholestasis is alate feature (Pandit and Bhave 1996). The intensity of orcein staining representing copper correlatessignificantly with the histological grades of the disease (Sethi et al 1993).

The histological changes in ICC have been divided in three grades (Sub-committee of the IndianCouncil of Medical Research, 1955). In the early stage there is cloudy swelling of hepatocytes withportal inflammation. In the second stage there is increased fibrosis and in around the portal tractsleading to pseudo-lobule formation and degeneration of hepatocytes. The last stage is characterised byextensive fibrosis with coarse fibrous bands leading to complete loss of liver architecture and hepaticdegeneration.

Steroids have been used to allay symptoms in ICC. The copper chelater, D-penicillamine has beenclaimed to lead to clinical and histological remission in up to 65% of patients (Tanner et al., 1987).This is a single study performed on 29 patients with early ICC. More studies need to be done todefinitely determine the role of D-penicillamine in the treatment of ICC.

A reduction in the incidence of ICC, has been explained on the basis of the reduction in the use ofbrass utensils in India (Pandit and Bhave 1996). The possibility of dilution of the genetic pool due tointer-caste marriages in India has not been ruled out. A similar reduction seen in fatal infantile livercirrhosis in certain regions of Austria has been reported (Müller et al.,1996). Further studies need tobe done to determine the possible genetic defect in ICC. With the increasing rarity of the disease,genetic studies on the animal model, the "bedlington terrier" dog, would be useful.

References

Agarwal, S.S., Lahori, U.C., Mehta, S.K., Smith, D.G. and Bajai, P.C (1979), ‘Inheritance of Indianchildhood cirrhosis’, Hum Hered., vol. 29, pp. 82-9.

Bhave, S., Pandit, A.N., Pradhan, A.N., Sidhaye, D.G., Kantarjian, A., Williams, A., Tablot, I.C. andTanner, M.S. (1982), ‘Paediatric liver disease in India’, Arch Dis Child., vol. 57, pp. 922-8.

Chandra, R.K.(1976), ‘ICC geneologic data, alpha-foetoprotein MbsAg & circulating immunecomplexes’, Trans Roy Soc Trop Med Hyg., vol. 70, pp. 296-301.

Liver Diseases Sub-committee of the Indian Council of Medical Research. (1955), ‘Infantile cirrhosisof the liver in India’, Indian J Med Res., vol. 43, pp.723-47.

Müller, T., Feichtinger, H., Berger, H. and Müller, W. (1996), ‘Endemic Tyrolean infantile cirrhosis:an ecogenetic disorder’, Lancet, vol. 347, pp. 877-80.

Pandit, A., and Bhave, S. (1996), ‘Present interpretation of the role of copper in Indian childhoodcirrhosis’, Am J Clin Nutr., vol. 63, pp. 830S-5S.

Portmann, B., Tanner, M.S., Mowat, A.P and Williams, R (1978), ‘Orcein positive liver deposits inIndian childhood cirrhosis’, Lancet, vol. 1, pp. 1338-40.

C o p p e r

94

Pradhan. A.M., Talbot, I.C and Tanner, M.S. (1983), ‘Indian childhood cirrhosis and other cirrhosis ofIndian children’, Pediatr Res., vol.17, pp. 435-8.

Sethi, S., Grover, S. and Khodaskar, M.B. (1993), ‘Role of copper in Indian childhood cirrhosis.Annals of Trop.Paed.,’vol. 13, pp. 3-6.

Sharda, B and Bhandari, B (1984), Copper concentration in plasma, cells, liver, urine, hair and nailsin hepatobiliary disorders in children’, Indian Pediatr., vol. 21, pp. 167-71.

Sur, A. and Bhatti, A. (1978), ‘ICC : An inherited disorder of tryptophan metabolism’, Brit. Med J.,vol. 2, pp. 529-31.

Tanner, M.S., Portmann, B., Williams, R., Mowat, A.P., Pandit, A.N., Mills, C.F. and Brenner, I.(1979), ‘Increased hepatic copper concentration in Indian childhood cirrhosis’, Lancet, vol. 1, pp.1203-5.

Tanner, M.S., Bhave, S.A., Kantarjian, A.H and Pandit, A.N. (1983), ‘Early introduction of coppercontaminated animal milk feeds as possible cause of ICC’, Lancet, vol. 2, pp. 992-5.

Tanner, M.S., Bhave, S.A., Pradhan, A.M. and Pandit, A.N. (1987), ‘Clinical trials of penicillaminein Indian childhood cirrhosis’, Arch Dis Child, vol. 62, pp.118-24.

Workshop 1 - Copper and Health

Chairman: Dr K Bentley

Rapporteur: Dr L Tomaska

Participants:Dr G Becking Dr T M Florence Dr E V Ohanian

Dr K Buckett Dr R Gaunt Prof A Oskarsson

Dr S Churches Mr M T Gerschel Prof S Sethi

Prof B D Culver Prof R Goyer Dr M Taylor

Prof H H Dieter Dr P Imray Dr R Uauy

Prof G Fell Dr B Markey Mr J Williamson

Dr D J Fitzgerald Prof M Moore

1. Summary

In considering the exposure of people to copper, the human population was divided into three separategroups:

• populations at risk of toxicity arising from exposure to copper;

• populations which were at risk of copper deficiency; and

• general population group.

The population at risk to copper was considered to include special groups of people with clinicalconditions such as ICC (unknown aetiology), ICT (genetic defect), G-6-PD (GSH-T defect leading tohaemolysis) and Wilson’s disease (defective ATPase), and groups of people on dialysis who havesuffered liver disease.

The Workshop participants gave consideration to the inclusion of other groups in this category,including pregnant women, normal infants and Wilson disease heterozygotes but have concluded thatthere was no scientific evidence which indicates that any of these groups are routinely at risk fromcopper toxicity.

C o p p e r

95

The population at risk from deficiency was considered to include low birth weight infants, peoplewith malnutrition, people who suffered from chronic enterolosses, people who supplement their dietswith zinc or vitamin C, people with Menkes’ syndrome (genetic disorder) and people who consumesome, but not all, high fibre diets.

The general population was identified as people who fell into neither of the first two groups. Indiscussion regarding the levels of copper exposure to the normal population, the Workshopconsidered a WHO report (Trace Elements in Human Nutrition and Health, 1996) which estimatedlevels of copper exposure from 136 studies to be (at the 50th percentile):

• adults 1.5 mg copper/day

• children 1.3 mg copper/day

It was also noted that, in no case did exposure exceed 6 mg copper/day for adults and 4 mgcopper/day in children.

The report recommended a copper intake of 1.3 mg copper/day for adults and 0.6 mg copper/dayfor children. It was noted that the upper limit of intake was recommended to be 13 mg copper/dayfor adults and 1 mg copper/day for children. However, the Workshop noted that this recommendedupper limit of intake was not based on any scientifically-derived NOAEL.

Professor Dieter gave a short presentation of a report on childhood cirrhosis in Germany, based on103 cases collected between 1982 and 1994. Of these, three could be directly attributed to copperexposure.

2. Research recommendations

The Workshop on Copper and Health made the following research recommendations which would beconsidered to advance the information base for copper:

• To determine the concentration of copper and quality of drinking water that produce acute toxicity(such as gastrointestinal effects) from single and chronic exposures;

• To undertake studies on ICC populations to determine:

- genetic component;

- relationship to ICT and sporadic cases of childhood cirrhosis;

- possible mechanisms related to basic defect;

- environmental factors which influence copper exposure; and

- methods for early diagnosis.

• To determine the incidence of genetic disorders of copper homeostasis;

• To determine factors and quantitative influences which determine the bioavailability of dietarycopper including vegetarian diets, levels of iron, and zinc;

• To determine methodologies for identifying copper excess or deficiency in human populations;

• To study methods for applying stable isotope technology to define bioavailability and stores

of copper;

• To determine the effect of marginal intake of copper on the prevalence of chronic disease;

• To determine mechanisms that influence copper homeostasis of absorption, storage and excretionin different age groups; and

• To determine dietary intakes of copper in populations living in developing countries to evaluateadequacy of those diets relative to copper.

C o p p e r

96

Workshop 2 - Distribution and Metabolism

Facilitator: Dr C Dameron

Rapporteur: Mr P Callan

Participants:Dr G Anderson Mr M Harrison Dr R Sadler

Prof HH Dieter Dr JFB Mercer Dr KH Summer

Dr R Gooneratne Dr S Moghaddas

1. Summary

Participants of Workshop 2 - Distribution and Metabolism noted that copper is an essential elementwhich is required for several critical metabolic processes in humans. However, it was also recognisedthat adverse health effects may result where copper deficiency or chronic copper excess occurs inhumans.

Manifestations of copper deficiency include anaemia, neutropenia and bone abnormalities, but clinicaldeficiency is relatively infrequent in humans. Adverse effects of acute copper excess includeepigastric pain, tachycardia, haemolytic anaemia, haematuria, liver and kidney failure, andgastrointestinal bleeding. Respiratory effects, including metal fume fever, may result fromoccupational exposure to high concentrations of copper in the air.

Some disorders associated with copper deficit or excess have a well defined genetic basis. Thesediseases include Menkes disease, which is characterised by deficiencies of copper; and Wilson'sdisease, which leads to progressive accumulation of copper. Indian Childhood Cirrhosis (ICC) andIdiopathic Copper Toxicosis (ICT), which are also considered to be associated with genetically-basedcopper sensitivity, are fatal liver conditions in early childhood which relate to copper excess.

While there was considerable discussion on the distribution and metabolism of copper in humans,principle discussion of the Workshop focussed on consideration of those biological markers whichwould be useful in determining copper deficiency or excess in humans. In particular whether it waspossible to use biomarkers for identification of marginal adverse and chronic excess effects insusceptible individuals, sub-populations and the general population.

It was noted that superoxide dismutase and cytochrome oxidase activity, and the levels of low densitylipoprotein and ceruloplasmin in blood changed during copper deficiency and excess. It was agreedthat while serum copper and ceruloplasmin levels are useful indicators of moderate to severe copperdeficiency, they were not sensitive measures of marginal copper deficiency.

Similarly, tyrosinase deficiency which would be expected to lead to hypopigmentation of the skin andhair was not considered as a useful indicator of early exposure as microscopy analysis of the tissuewas a time consuming method and that tyrosinase, while being an excellent marker, was very difficultto measure.

Overall, the Workshop Group noted that while there were currently several biomarkers to indicatemoderate to severe copper deficiency and excess, the Workshop Group was unable to identifybiomarkers which could be used to efficiently identify effects, marginal copper deficiencies orexcesses.

C o p p e r

97

Intracellular protection against copper toxicity in mammalsCharles T. Dameron, Shadi Moghaddas and Mark D. HarrisonNHMRC National Research Centre for Environmental Toxicology,The University of Queensland and Griffith University.

(Paper presented to Workshop 2 Group)

1. Introduction

Transition metals such as copper, molybdenum and zinc are essential to life because of the catalyticand structural roles they play in proteins and other biomolecules. Excessive concentrations ofessential and non-essential metal ions like cadmium, mercury and lead can induce toxicities at thecellular, tissue and organ levels. Most organisms utilise a redundant array of cellular mechanisms tolimit the toxicity of metal ions. The purpose of this review is to discuss intracellular mechanismsused to metabolise excessive concentrations of copper in mammalian cells.

2. General mechanisms for metal detoxification

A selection of the cellular mechanisms utilised by mammalian cells to detoxify metals are illustratedin Figure 1 to demonstrate the scope of available mechanisms. Simplistically, the detoxificationsystems can be subdivided into mechanisms to reduce metal uptake, and to enhance metalsequestration and export.

The reduction of metal importation to limit toxicity can operate through inhibition of the importmachinery for the metal or by making the extracellular metal unavailable for absorption. Forexample, the level of iron taken up by mammalian cells is partially regulated by controlling at atranslational level the membrane concentration of the receptor for the iron transport protein ferritin(Klausner and Dancis, 1994; Basilion et al., 1994). The regulation is mediated by intracellularconcentration of iron atoms. The utilisation of these extracellular mechanisms will not be discussed inthis review.

The intracellular chelation or sequestration of metals into relatively innocuous complexes ororganelles is a commonly used mechanism to limit their toxicity. In many cases the chelating agentsare peptides or proteins that form stable complexes which limit the element's reactivity and/or aid inits excretion. In addition to providing a means of limiting the reactivity and toxicity of essentialmetals, some complexes appear to serve as storage sites for the metal ions. Sequestration of metalscan also be an initial step in a pathway that ultimately leads to exportation of the metal or metal-complex through a pump mechanism. Alternatively, a sequestered ion may be pumped into a vesiclefor storage or extrusion by the vesicle. Metallothioneins (MTs), involved in the sequestration ofcopper, will be described in detail below.

C o p p e r

98

Figure 1: Mammalian Metal Detoxification Mechanisms M+ is any metal-ion, essential or non-essential.

Exportation of excess metal ions to limit their intracellular toxicity is a ubiquitous process. Thecation translocating P-type ATPases are among the more common pumping mechanisms used totransport metal ions out of cells or into organelles. The basic pump design is conserved across all lifeforms from bacteria to humans and is used for a range of elements, drugs, toxins and proteins. Thesequence, presumably the structure, is frequently modified to increase the specificity and efficacy fora given element. In addition to the ATPase pumps some mammalian cells transport transition metal-ions, like zinc by utilising non-ATPase pumps (Palmiter and Findley, 1995).

3. Redundancy

An essential feature of metal ion detoxification pathways is their redundancy. Many detoxificationmechanisms are not entirely specific and are utilised against a number of different metal ions. Anexample of a non-specific detoxification pathway is metallothionein and its function in thesequestration of cadmium, zinc and copper ions. In conjunction with non-specific detoxification therecan exist a set of detoxification mechanisms that are entirely specific for a particular metal ion. TheMenkes protein, a copper-specific ATPase, is a good example of metal-specific detoxificationmechanism. Specialisation of cells within an organism or tissue often leads to different patterns andlevels of expression of the detoxification mechanisms available to the cells. Few of these mechanismshave been described in detail and in most cases specificity or lack of specificity is not understood at amolecular level.

4. Metalloregulation

Metalloregulation, broadly defined as a cellular response to an intracellular metal ion concentration, iswell documented in bacterial, fungal and animal systems. Metalloregulation can be accomplished attranscriptional, translational or enzymatic levels. At a molecular level these mechanisms operatethrough metal binding proteins. Metalloregulatory proteins serve as conformational switches with themetal cations organising or stabilising an active conformation of the protein (O'Halloran, 1993;Dameron et al., 1991; Dameron et al., 1993). These proteins allow cells to respond biochemically toincreases and decreases in the intracellular copper concentration.

C o p p e r

99

5. The copper detoxification pathway

The mammalian copper detoxification pathway contains at least two mechanisms to detoxify excesscopper. Firstly, copper can be sequestered by the copper-binding protein metallothionein into aninnocuous complex , described below. Secondly, excess copper can be transported out of the cellwith the aid of an export pump (Dameron and Harrison, in press). An outline of the metabolicpathway for copper in a mammalian liver cell is shown in Figure 2. The abbreviated pathwaydescription highlights some of the known mechanisms used by mammalian cells. The pathway reflectspublished mechanisms and should not be regarded as complete or exact since this area of research isevolving rapidly. A hepatic (liver) cell was chosen to highlight the pathway because of the liver'scentral role in copper metabolism. Not all cells would be expected to metabolise copper in the way theliver cell does and some cells are expected to have pathways that liver cells do not have. In contrastto the specific induction of metallothionein by copper in lower eukaryotes the mammalianmetallothionein is induced by copper, cadmium, zinc and other transition metals (Hamer, 1986).

Figure 2: The hepatic copper metabolic pathway in mammals

All mammalian metallothioneins are homologous, sharing common metal-binding motifs, andinduction mechanisms. Metallothionein's function and structure are described below. The disruptionof the metallothionein gene does not lead to a marked increase in copper sensitivity in transgenic mice(Michalaska and Choo, 1993), possibly because the cells can use the murine homolog of the MenkesATPase pump to detoxify (export) the excess copper (Levinson et al., 1994). Transgenic mice thathave had their metallothionein genes deleted are, however, sensitive to cadmium (Michalaska andChoo, 1993; Masters et al., 1994). It is plausible that the mice are cadmium sensitive because theircadmium detoxification pathway, being less redundant, lacks a specific export mechanism and musttherefore rely wholly on sequestration by metallothionein.

Metals, especially cadmium, can accumulate in mammalian cells bound to metallothionein and/or besequestered into vesicles. The retention of metals, including copper, is tissue-specific and can inducecellular and tissue damage in cases of extreme overload. The primary route of copper excretion inhumans is through the bile and may involve metallothionein (Hamer, 1986). The exact rolemetallothionein plays in the metabolism and detoxification of metals is still being elucidated.

C o p p e r

100

6. Copper detoxification mechanisms

6.1 Metallothioneins

The principle sequestration molecule for transition metal ions in groups 11 and 12, including Cu(I)and Cd(II), are the cysteine-rich metallothioneins (MT). The cysteinyl sulfurs in these proteinsfunction as ligands for the metals. Ultimately, the metals are bound in polynuclear metal-thiolateclusters (Dameron et al., 1991; Winge et al., 1993; Pickering et al., 1993; Winge et al., 1994). Thecysteines in metallothioneins typically account for 20-30% of the amino acids and are arranged inrepetitive Cys-Cys, Cys-Xaa-Cys and Cys-Xaa-Xaa-Cys motifs (where Xaa is any amino acid). Thestructural features of the MTs have been reviewed recently (Kille et al., 1994). Metallothioneins aresmall, 25-62 amino acid, cysteine-rich proteins that contain very few hydrophobic residues (fourexamples are highlighted in Table 1)

Table 1 : Amino acid sequence of selected metallothioneins

S. cervisiaeMTMFSELINFQNEGHECQCQCGSCKNNEQCQKSCSCPTGCNSDDKCPCGNKSEETKKSCCSGK

N. crassa MTGDCGCSGASSCNCGSGCSCSNCGSK

MouseMT2aMDPNCSCAAGDSCTCAGSCKCKECKCTSCKKSCCSCCPVGCAKCAQGCICKGASDKCSCCA

HumanMT2aMDPNCSCAAGDSCTCAGSCKCKECKCTSCKKSCCSCCPVGCAKCAQGCICKGASDKCSCCA

Metallothioneins are found in a variety of forms in selected bacteria (Silver, 1994), fungi (Tohoyamaet al, 1992; Mehra et al., 1990; Galli et al., 1994; Cervantes and Gutierrez-Corona, 1994), plants(Foley and Singh, 1994; Ledger and Gardner, 1994; Zhou and Goldsbrough, 1994), and animals(Hamer, 1986). The mammalian proteins are very homologous and have 62 amino acids with 20cysteine residues. The protein is organised into two domains when metals are bound. Themammalian MT will bind a total of seven tetrahedral four coordinate atoms, like Cd(II) and Zn(II), ortwelve three coordinate trigonal planar atoms, like Cu(I). The Cu(I) and Cd(II) containing forms ofthe protein have distinct structures (Kille et al., 1994). Ligation is accomplished entirely by cysteinylthiolates.

As seen in the structures of the rabbit 5Cd2Zn metallothionein solved by NMR (Arseniev et al, 1988)and X-ray crystallography, (Robbins and Stout, 1992) the lack of core hydrophobic residues tostabilise the tertiary structure of MT is compensated for by the formation of a metal thiolate core(Kille et al., 1994; Robbins and Stout, 1992). In the absence of metals to organise their core structurethe metallothioneins adopt random coil configurations. The number and type of metal species in theprotein dictates the tertiary structure of metallothionein. The structure of the copper form of themammalian metallothionein has not been determined but it is known to adopt a two domain structurewith six trigonally bound Cu(I) atoms in each domain.

6.2 Transcriptional regulation of metallothioneins

Metallothionein synthesis is transcriptionally regulated by metals in all higher eukaryotes (Hamer,1986), excluding the recently discovered brain MT III (Palmiter et al., 1992) and epithelial MT IVs(Quaife et al., 1994), which have unknown functions. In a given organism or cell type the extent ofthe metallothionein induction depends on the concentration and species of metal ion and the geneticlocus. Some cell types respond more strongly to cadmium and zinc (compared to copper) while othersare induced primarily by copper ions.

C o p p e r

101

The transcriptional regulation of MT synthesis in mammals by copper is analogous to the well studiedyeast system but is mechanistically distinct. In animals metalloregulation is accomplished through aseries of 5' metal regulatory elements (MREs) that are arranged in tandem repeats. The MREs are nothomologous to the 5' regulatory up-stream activation sequences (UAS) in yeast (examples arehighlighted in Table 2).

Table 2: MRE and UAS DNA Sequences

S.cervisiae Upstream Activation Sequence

5'GATGCGTCTTTTCCGCTGAACC3'

Mammalian Metal Regulatory Element

5'CTCTGCACTCCGCC3'

MREs are conserved in higher mammals (Hamer, 1986). Partial purification, sequential cloningstudies (Palmiter, 1994) and overlap of the MREs with basal control elements, especially GC richSpl-like binding sites (Hamer, 1986), suggest the mammalian mechanism relies on multiple proteins,only one of which is a metal sensor. The yeast regulatory mechanism functions through a singlemetal-sensing transcription factor. If the mammalian mechanism was analogous to the yeastmechanism described above, the sensing protein would bind copper, or other metals, to form an activetranscription factor-complex with an increased affinity for the MRE. Homology-based searches andprobes for transcription factors homologous to the yeast factor in higher eukaryotes throughconventional molecular biology techniques have not been successful.

6.3 Cu-ATPases

ATPase pumps are ubiquitous, being involved in the movement or translocation of ions such as H+,Na+, K+, Ca++ and a variety of metal ions. The ATPase pumps that translocate transition metal ionsare in the P-type family of ATPases (Tsai et al., 1992). The "P" designation stems from the covalentphosphorylation of a conserved aspartic acid residue (single letter amino-acid code P) that is part oftheir reaction cycle. The P-type cation ATPases, including the copper ATPases, are highly conservedfrom bacteria to humans (Silver, 1994; Solioz et al., 1994; Silver et al., 1993).

The human P-type Cu-ATPases are intimately involved in the metabolism and detoxification ofcopper ions as evidenced by the effects that defective pumps have on copper metabolism (Camakariset al., 1995). Defects in one of the human Cu-ATPases (MNK) leads to a fatal copper deficiencydisease called Menkes disease. The defect blocks transport of copper across the serosal membrane.Copper accumulates in the intestinal cell as a metallothionein complex and is ultimately lost throughthe normal sloughing of the intestinal cells. Copper transfer across membranes is also blocked in anumber of other tissues. Defects in another human Cu-ATPase (WLD) that is primarily expressed inthe liver leads to a copper excretory disorder that manifests as a chronic liver disease called Wilsonsdisease. Wilsons disease can be treated with chelation therapy.

The Menkes and Wilsons proteins have the characteristic elements of a P-type ATPase; aphosphorylation domain, a phosphatase domain, an ATP binding site and a transmembrane cationchannel. WLD, the Wilsons ATPase, is 57% identical to the Menkes protein; the homology increasesto 79% or greater in the transduction, ion channel, phosphorylation and ATP-binding regions. Theputative metal binding sites in MNK and WLD have the characteristic -Cys-Xaa-Xaa-Cys- motifsfound in a variety of metal binding proteins including some zinc fingers, the metallothioneins, thecopper-regulated yeast transcription factors (ACE1 and AMT) and other metal-binding sites intransition metal ATP translocases. Outside of the metal-binding motif there is no homology betweenthe metal ATP translocase proteins and other metal-binding proteins.

The putative metal-binding subdomains in the N-terminus of the human P-type Cu-ATPases havesignificant homology to the N-terminal metal-binding domains of prokaryotic P-type ATPases utilisedfor copper, cadmium, mercury and calcium transport. Both MNK and WLD have strong homology to

C o p p e r

102

CopA, a bacterial copper ATPase (Dameron and Harrison, in press; Vulpe et al., 1993). The closehomology between MNK and WLD and CopA suggests that CopA is responsible for the active exportof copper from bacterial cells. MNK and WLD contains six repetitive homologous sequences(MNKr1-6) of approximately 70 amino acids, each of which contains a single conserved -Met-Xaa-Cys-Xaa-Xaa-Cys- motif as found in the bacterial ATPases. The regions between these putativesubdomains are not conserved in the MNK protein, the Wilsons gene or in the bacterial ATPases.These six putative sub-domains are individually homologous in both sequence and predictedsecondary structure to a bacterial mercury detoxification protein, merP. The three-dimensionalstructure of merP has been determined by NMR spectroscopy (Eriksson and Sahlman, 1993).

6.4 Regulation of the human copper ATPases

The Menkes protein functions to export excess intracellular copper and is postulated to be reversiblymetalloregulated through the specialised copper-binding subdomains in the amino terminus of theprotein. The metalloregulation couples the cellular export of copper to the intracellular concentrationof copper ions. In Chinese Hamster Ovary (CHO) cells amplification of the MNK gene andsubsequently of the protein, mRNA and protein levels correlate with the cellular copper resistance(Camakaris et al., 1995). Moreover, the amount of message transcribed and translated is not copperdependent but the activity increases with increasing copper. Thus, the regulation appears to takeplace at an enzymatic level rather than at a transcriptional or translational level. Activation of theMNK pump is hypothesised to occur through a conformational mechanism analogous to that proposedfor the mammalian P-type Ca2+- ATPase. The binding of calmodulin to the regulatory domain of theCa2+-ATPase induces a conformational change that appears to uncover a sterically-protected asparticacid residue (Falchetto et al., 1992). Phosphorylation of the protected aspartic acid in the P-typeATPases is required for the export of metal cations. Were MNKr to contain a series of structures orsub-domains with folds analogous to merP, the copper metalloregulation could be afforded byrearrangements of the subdomains, Figure 3. The organisation of MNKr into sub-domains that arerearranged via Cu(I) during activation of MNK would document a new regulatory mechanism.

Figure 3: Activation of the Menkes Protein by Copper-Binding to MNKr1-6 Panel A. depicts theMenkes protein in its inactive state. This corresponds to normal intracellular levels of copper. Thesix subdomains contain no bound copper. Panel B. depicts the Menkes protein in its active state.

C o p p e r

103

Increasing intracellular copper levels leads to copper-binding to MNKr1-6 and activation of theATPase with concomitant copper translocation.

7. Conclusion

Living systems have developed an array of mechanisms to maintain intracellular copper levels withinthe "window" between deficiency and excess. This review has focussed on the means by which cellsprotect themselves from the potentially cytotoxic effects of excess copper. Mammalian cells possessmultiple, redundant mechanisms for detoxifying excess intracellular copper. The recent discovery ofthe genes responsible for Menkes and Wilsons disease has focussed international interest on theunderstanding of copper homeostasis in mammalian cells. The role of these proteins in copperhomeostasis in vivo is still poorly understood and remains an exciting area of biochemical research.

References

Arseniev, A., Schultze, P., Worgotter. E., Braun, W., Wagner, G., Vasek, M., Kagi, J.H.R. andWuthrich, K. (1988), ‘Three-dimensional structure of rabbit liver [Cd7] Metallothionein-2a inaqueous solution determined by nuclear magnetic resonance’, J Mol Biol., vol. 201, pp. 637-657.

Basilion, J.P., Rouault, T.A., Massinople, C.M., Klausner, R.D. and Burgess, W.H. (1994), ‘The iron-responsive element binding protein: Localisation of the RNA-binding site to the aconitase active-site’,Proc Natl Acad Sci USA, vol. 91, pp. 574-8.

Camakaris, J., Petris, M.J., Bailey, L., Shen, P., Lockhart, P., Glover, TW., Barcroft, C.L., Patton, J.and Mercer, J.F.B.(1995), ‘Gene amplification of the Menkes (MNK;ATP7A) P-type ATPase gene ofCHO cells is associated with copper resistance and enhanced copper efflux’, Hum Mole Genetics, vol.4, pp. 2117-23.

Cervantes, C. and Gutierrez-Corona, F. (1994), ‘Copper resistance mechanisms in bacteria and fungi’,FEMS Microbiol Rev., vol. 14, pp.121-37.

Dameron, C.T., Winge, D.R., George, G.N., Sansone, M., Hu, S. and Hamer, D. (1991), ‘A copper-thiolate polynuclear cluster in the ACE1 transcription factor’, Proc Natl Acad Sci USA, vol. 88, pp.6127-31.

Dameron, C.T., Arnold, P., Santhanagopalan, V., George, G. and Winge, D.R. (1993), ‘Distinct MetalBinding Configurations in ACE’, Biochem., vol. 32, pp. 7294-301.

Dameron CT, Harrison MD. Mechanisms for protection against copper toxicity. Am J Clin Nutr. InPress.

Eriksson, P.O. and Sahlman, L. (1993), ‘1H NMR studies of the mercuric ion binding protein MerP:Sequential assignment, secondary structure and global fold of oxidised MerP’, Journal ofBiomolecular NMR, vol.3, pp. 613-626.

Falchetto, R., Vorherr, T. and Carafoli, E. (1992), ‘The calmodulin-binding site of the plasmamembrane Ca2+ pump interacts with the transduction domain of the enzyme’, Protein Sci., vol. 1,pp.1613-21.

Foley, R.C. and Singh, K.B. (1994), ‘Isolation of a vicia faba metallothionein-like gene - expressionin foliar trichomes’, Plant Mole Bio., vol. 26, pp. 435-44.

Galli, U., Schuepp, H., and Brunold, C. (1994), ‘Heavy metal binding by mycorrhizal fungi’,Physiologia Plantarum, vol. 92, pp. 364-8.

Hamer, D.H. (1986), ‘Metallothionein’, Annu Rev Biochem., vol. 55, pp. 913-51.

Kille, P., Hemmings, A and Lunney, E.A. (1994), ‘Memories of Metallothionein’, Biochim BiophysActa., vol.1205, pp. 151-61.

Klausner, R.D and Dancis, A. (1994), ‘A genetic approach to elucidating eukaryotic iron metabolism’,FEBS Lett., vol. 355, pp. 109-13.

C o p p e r

104

Ledger, S.E and Gardner, R.C. (1994), ‘Cloning and characterization of five cDNAs for genesdifferentially expressed during fruit development of kiwifruit (actinidia deliciosa var deliciosa)’,Plant Mole Bio., vol. 25, pp. 877-86.

Levinson, B., Vulpe, C., Elder, B., Martin, C., Verley, F., Packman, S. and Gitschier, J. (1994), ‘Themottled gene is the murine homolog of the Menkes disease gene’, Nat. Genet., vol. 6 (4, April), pp.369-373.

Masters, B.A., Kelly, E.J., Quaife, C.J., Brinster, R.L. and Palmiter, R.D. (1994), ‘Targeted disruptionof metallothionein I and II genes increases sensitivity to cadmium’, Proc Natl Acad Sci USA, vol. 91,pp. 584-8.

Mehra, R.K., Garey, J.R. and Winge, D.R. (1990), ‘Selective and tandem amplification of a memberof the metallothionein gene family in Candida glabrata’, J Biol Chem., vol. 265, pp. 6369-75.

Michalska, A.E and Choo, K.H. (1993), ‘Targeting and germ-line transmission of a null mutation atthe metallothionein I and II loci in mouse’, Proc Natl Acad Sci USA., vol. 90, pp. 8088-92.

O Halloran, T.V. (1993), Transition metals in control of gene expression’, Science, vol. 261, pp. 715-25.

Palmiter, R.D., Findley, S.D., Whitmore, T.E. and Durnam, D.M. (1992), ‘MT III, a brain-specificmember of the metallothionein gene family’, Proc Natl Acad Sci USA, vol.89, pp. 6333-7.

Palmiter, R.D. (1994), ‘Regulation of metallothionein genes by heavy metals appears to be mediatedby a zinc-sensitive inhibitor that interacts with a constitutively active transcription factor, MTF-1’,Proc Natl Acad Sci USA, vol. 91, pp.1219-23.

Palmiter, R.D. and Findley, S.D. (1995), ‘Clonning and functional characterisation of a mammalianzinc transporter that confers resistane to zinc’, EMBO, vol.14:639-49.

Pickering, I.J., George, G.N., Dameron, C.T., Kurtz, B., Winge, D.R. and Dance, I.G. (1993), ‘X-rayAbsorption Spectroscopy of Cuprous-Thiolate Multinuclear Clusters in Proteins and Model Systems’,J Am Chem Soc., vol. 115, pp. 9498-505.

Quaife, C.J., Findley, S.D., Erickson, J.C. et al. (1994), ‘Induction of a new metallothionein isoform(MT-IV) occurs during differentiation of stratified squamous epithelia’, Biochem., vol. 33, pp. 7250-9.

Robbins, A.H and Stout, C.D. (1992), ‘Crystal Structure of Metallothionein’. In: Stillman M.J., Shaw,C.F. and Suzuki, K.T, eds. Metallothioneins: synthesis, structure and properties of metallothioneins,phytochelatins and metal thiolate complexes. New York, New York: VCH Publishers, Inc., pp. 31-54.

Silver, S., Nucifora, G. and Phung, L.T. (1993), ‘Human Menkes X-chromosome disease and thestaphylococcal cadmium-resistance ATPase: a remarkable similarity in protein sequences’, MolecularMicro., vol. 10, pp. 7-12.

Silver, S. and Ji, G. (1994), ‘Newer systems for bacterial resistances to toxic heavy metals’, EnvironHealth Perspect., vol. 102, Suppl. 3, pp. 107-13.

Solioz, M., Odermatt, A. and Krapf, R. (1994), ‘Copper pumping ATPases: common concepts inbacteria and man’, FEBS Lett., vol. 346, pp. 44-7.

Tohoyama, H., Tomoyasu, T., Inouhe, M., Joho, M. and Murayama, T. (1992), ‘The gene forcadmium metallothionein from a cadmium-resistant yeast appears to be identical to CUP1 in a copper-resistant strain’, Curr Genet., vol. 21, pp.275-80.

Tsai, K., Yoon, K.P. and Lynn, A.R. (1992), ‘ATP-dependent cadmium transport by the cadAcadmium resistance determinant in everted membrane vesicles of bacillus subtilis’, J Bacteriol.,vol.174, pp. 116-21.

Vulpe, C., Levinson, B., Whitney, S., Packman, S. and Gitschier, J. (1993), ‘Isolation of a candidategene for Menkes disease and evidence that it encodes a copper-transporting ATPase’, publishederratum appears in Nat Genet., vol. 3, (March 3), p. 273. See comments,. Nat Genet., vol. 3, pp. 7-13.

C o p p e r

105

Winge, D.R., Dameron, C.T., George, G.N., Pickering, I.J and Dance, I.G. (1993), ‘Cuprous-ThiolatePolymetallic Clusters’, in Biology’. In: Karlin, K.D and Tyeklar, Z, eds. Bioinorganic Chemistry ofCopper. New York, NY: Chapman & Hall :110-23.

Winge, D.R., Dameron, C.T. and George, G.N. (1994), ‘The Metallothionein Structural Motif in GeneExpression’. In: Eichhorn, G.L, Marzilli, L.G, eds. Advances in Inorganic Biochemistry 10. NewJersey: Prentice-Hall, pp. 1-48.

Zhou, J. and Goldsbrough, P.B. (1994), ‘Functional homologs of fungal metallothionein genes fromArabidopsis’, Plant Cell, vol. 6, pp. 875-84.

Workshop 3 - Copper and the Environment

Chairman: Dr HE Allen

Rapporteur: Mr DM Wagner

Participants: Dr WJ Adams Dr M Hallman Dr C Lee

Dr G Batley Dr D Harkess Mr R Smith

Prof JC Castilla Dr P Howe Dr J Stauber

Dr R Erickson Prof T Hutchinson Dr JHM Temmink

Dr P Glazebrook Prof GE Lagos Dr JM Weeks

1. Summary

To set the scene for issues relating to copper and the environment, Dr Paul Howe, Institute ofTerrestrial Ecology, UK, outlined his perceptions of the issues that are the most important topics withregard to an environmental risk assessment of copper. These included:

• environmental sources of exposure;

• environmental fate;

• environmental bioavailability, including analytical techniques, uptake and biomonitoring; and

• deficiency versus toxicity, including essentiality, homeostatic mechanisms, adaptation, tolerance,and toxicity testing.

2. Aquatic systems

Considerable group discussion took place on copper in the aquatic system, including whether it wasbest to use total or dissolved copper as the measure of exposure, and what criteria should be used onthe issue of bioavailability. It was noted that the US EPA has given a clear direction on some of theseissues by recommending dissolved copper as a better measure for water bodies, while some totalcopper is recommended for discharges, though these recommendations still require implementation byindividual States in the US. The group also discussed definition for discharge zones, mixing zones,and whole river measurements by area, from both an Australian and North American perspective.

3. Toxicity testing

An extensive discussion followed on toxicity testing, including bioassays versus physicochemicalscreening assays, and it was noted that some regulatory authorities were establishing a compendiumof toxicity data, with minimum data set requirements, and that it was possible to set a maximumcriterion based on acute toxicity data, and that chronic toxicity data can also go through a similar

C o p p e r

106

procedure. Finally, it was felt that there was a strong need for a separate set of requirements forplants, as existing requirements did not adequately address this area.

4. Bioavailability

Speciation is reasonably understood for copper, with binding to natural organic matter being animportant control. The assessment of speciation is very important in understanding the bioavailabilityof copper in the environment, and necessary for conducting a risk assessment for copper.

Dissolved copper should be used to evaluate potential copper bioavailability, while still recognisingthat it generally overestimated the real value, but is still more appropriate than using total copper as ameasure.

Procedures are being developed, and should be refined and progressed to give more accurate estimatesof copper bioavailability.

Copper budget and distribution in the environment needs to be addressed, as does accumulation.

The receptor site model approach to the prediction of copper speciation was discussed. This is basedupon measured concentrations of dissolved copper and other water quality parameters and the knownbinding constants for copper at an organism’s receptor site, e.g. a gill membrane, which mightcompete for the copper and therefore will control the bioavailable fraction..

5. Terrestrial testing

Terrestrial testing generally stems from field observations, and currently lacks a framework whichreflects the mechanism of interaction between the organism and the environment. There is a need tounderstand the interactions between metals and soils, soils and water, etc.

There is difficulty in setting an acceptable background level for copper, given the variability innaturally occurring copper levels. Speciation and bioavailability are also needed for terrestrialtoxicity testing. For example, pH data is needed to allow for partitioning between soil and water tomake a model useful.

As with aquatic testing, it is rare to see single contaminants, and there is a need to take all compoundsinto account, as the possibility of synergism and/or antagonism arises with multiple sources andmultiple compounds.

The EU is publishing protocols for soil toxicity testing, covering several soil types representative of anumber of countries, a range of compounds including copper, and a range of test species.

There is a need to come to a basic mechanistic understanding on bioavailability in terrestrial soils forenvironmental risk assessment. The correct chemical measurements are important to give links tobiological effects of copper in soils. Given that the range of availabilities and concentrations ofcopper in soils and sediments is greater than in aquatic systems, it is more important to improve theunderstanding of bioavailability in the terrestrial environment than in water.

6. Essentiality versus toxicity

In terrestrial systems, deficiency exists, but copper can still be found in excess, and hence lead totoxicity. It is important to establish whether such a potential for excess and deficiency to copperexists in aquatic systems.

New recovery techniques for copper now makes extraction from older sites economically viable, thusincreasing the potential for new environmental exposure to copper. In terrestrial systems, exposurecan arise from a variety of sources, including from new and previous mine sites, sewerage wastedisposal, water courses and agricultural soil.

A range of organisms in the environment can uptake and regulate copper, and due to the range ofspecies and sensitivity to copper, there is in sufficient information available to decide what range ofcopper is acceptable in the environment.

C o p p e r

107

There is inadequate understanding of the effects that adaptation can have upon test results. Largeincreases in tolerance to copper are seen in some species in reproduction toxicity tests, and adaptationof populations can occur over a long time period.

There is a need for guidelines for copper and other trace elements present in testing environments thatensure that demands for essential levels are met.

7. Bioaccumulation

There are well established tests and surrogates for bioaccumulation of organics, but is there such arange of tests for inorganics? Without information on the mechanisms and biochemistry ofbioaccumulation across phyla, the use of overconservative safety factors can lead to regulatorystandards below normal environmental levels, or lead to deficiency.

8. Conclusions and recommendations

Measures of "total" copper usually overestimate bioavailability considerably in aquatic ecosystems."Dissolved" copper is a better estimate of bioavailable copper in aquatic systems than total copper,but is still conservative.

Chemical speciation of copper is, perhaps, the best way to measure or estimate the fraction of totalcopper that is bioavailable. State-of-the-art techniques are evolving which allow for measurements ofcopper speciation in water.

In a regulatory context, techniques which demonstrate reduced bioavailability in water, sediments, orsoil should be used to set site-specific protective levels.

It was agreed that an integrative approach should be used in assessing environmental risk associatedwith the use and disposal of copper. This integrative approach should consider toxicity, fate andtransport, dissolution and transformation rate, and extent (as well as potential) for bioaccumulationand background levels.

The currently used tests for evaluating the toxicity of metals to terrestrial species are very poor. Thebasic mechanisms controlling bioavailability of metals in soils have not been demonstrated. There isa need for tests which include different trophic levels and the tests need to consider factors controllingbioavailability.

Bioaccumulation of metals does not follow the partitioning laws that govern the accumulation ofnonpolar organics. It is recognised that as the concentrations of copper in water increase, thebioaccumulation factor decreases, and vice versa. Because of the balance between toxicity andessentiality, the philosophy underlying the regulation of copper in the environment should be toprotect against adverse effects on biota.

Where possible, local/indigenous organisms should be used for site-specific risk assessment purposes.

C o p p e r

108

Participants in the international workshop on copper

Addresses and titles are those current at the time of the workshop. Apologies are extended to thosewho may have been omitted from the following list.

Dr William J. AdamsICA Kennecott,Utah Copper,8315 West 3595 South, PO box 6001,MAGNA, UTAH USA

Professor Herb AllenUniversity of DelawareDepartment of Civil and Environmental Engineering,NEWARK DE 19716, USA

Dr Greg AndersonQIMRP.O. Royal Brisbane Hospital,HERSTON 4029, QLD AUSTRALIA

Dr Graeme BatleyCSIRO Centre for Advanced Analytical Chemistry,Private Mail Bag 7MENAI 2234, NSW AUSTRALIA

Dr George BeckingIPCS ICCUPO Box 12233, MD EC-10,Research Triangle Park,NORTH CAROLINA 27709, U.S.A.

Dr Keith BentleyCommonwealth DepT. of Health and Family ServicesGPO Box 9848,CANBERRA 2601, ACT AUSTRALIA

Dr Kevin J. BuckettHealth Dept. Western AustraliaEnvironmental Health Service,PO Box 8172,Stirling St, PERTH 6849, WA AUSTRALIA

Mr Phil CallanCommonwealth Dept. of Health & Family ServicesHealthy Public Policy Unit,GPO Box 9848, CANBERRA 2601, ACT AUSTRALIA

Professor J. Carlos CastillaP. Universidad Católica de ChileDepartamento de Ecologia, Ecologia MarinaFacultad de Ciencas Biologicas,Casilla 114-D, SANTIAGO, CHILE

Dr Sarah ChurchesEPA Canberra, Hazardous Wate Section,Waste Management Branch,40 Blackall St, BARTON 2600, ACT AUSTRALIA

Mr Aaron CosierNRCET, 39 Kessels RoadCoopers Plains, BRISBANE QLD 4108 AUSTRALIA

Professor B. Dwight CulverUniversity of California, Irvine14 Mendel CourtIRVINE, CALIFORNIA 92715, U.S.A.

Dr Charles DameronNRCET 39 Kessels RoadCoopers Plains, BRISBANE QLD 4108 AUSTRALIA

Professor Hermann H. DieterUmweltbundesamt Institut für Wasser, BodenLufthygiene BundesgesundheitsamtsPO Box 33 00 22, D-14191 BERLIN, GERMANY

Dr Melissa Haswell-ElkinsNRCET, 39 Kessels RoadCoopers Plains, BRISBANE 4108, QLD AUSTRALIA

Dr Russell EricksonUS EPA 6201 Congdon Boulevard, Deluth,MINNESOTA 55804, USA

Professor Gordon S. FellUniversity of GlasgowDepartment of Pathological BiochemistryGlasgow Royal InfirmaryCastle St, GLASGOW G4 0SF SCOTLAND

Dr Jim FitzgeraldSouth Australian Health CommissionPublic and Environmental Health ServiceEnvironmental Health BranchPO Box 6, RUNDLE MALLADELAIDE 5001, SA AUSTRALIA

Dr T. Mark FlorenceCentre for Environmental Health Sciences112 Georges River Crescent,OYSTER BAY 2225, NSW AUSTRALIA

Dr Richard GauntRTZ Ltd Occupational Health Physician6 St James Square, LONDON SW1Y 4LD,UNITED KINGDOM

Mr M. Thierry GerschelTrefimetaux - Dev/ Standardisation Manager11 bis rue de l’Hotel de Ville, F- 92400 Courbevoie,CEDEX FRANCE

Dr Peter GlazebrookCRA Limited55 Collins St GPO Box 384D,MELBOURNE 3001 VIC AUSTRALIA

Dr Ravi GooneratneAVSG Lincoln UniversityAnimal & Veterinary Sciences GroupLincoln University,PO box 84, CANTERBURY NEW ZEALAND

Professor Robert A. GoyerUniversity of Western Ontario 6405 Huntingridge Rd.,CHAPEL HILL,NC 27514, U.S.A.

Dr Mal HallmanOsborne MinesPO Box 5170 MCTOWNSVILLE 4810, QLD AUSTRALIA

Dr Donna HarkessDepartment of Primary Industries & EnergyEdmund Barton BuildingGPO Box 858,Barton, CANBERRA 2601,ACT AUSTRALIA

Mr Mark HarrisonNRCET 39 Kessels RoadCoopers Plains, BRISBANE QLD 4108 AUSTRALIA

Dr Paul D. HoweInstitute of Terrestrial EcologyMonks Woods Abbots RiptonHuntingdon, CAMBRIDGESHIRE 2LS,UNITED KINGDOM

C o p p e r

ii

Dr Paul D. HoweInstitute of Terrestrial EcologyMonks Woods Abbots RiptonHuntingdon, CAMBRIDGESHIRE 2LS,UNITED KINGDOM

Dr Paula ImrayQueensland Health, Scientific Adviser (Toxicology)Environmental Health BranchGPO Box 48 Charlotte St.,BRISBANE 4001 AUSTRALIA

Ms Sharon KratzmannNRCET, 39 Kessels RoadCoopers Plains, BRISBANE QLD 4108 AUSTRALIA

Professor Gustavo E. LagosPontificia Universidad Católica de Chile Mining CentreCasilla 306, Cod. 105, SANTIAGO 22, CHILE

Dr Christopher M. LeeInternational Copper Association260 Madison Ave.- 16th Floor,New York , NY 10016, USA

Dr Bruce MarkeyEnvironment Protection Authority (NSW)Citadel Towers, 799 Pacific HighwayPO Box 1135, CHATSWOOD 2057, NSWAUSTRALIA

Dr Julian F. B. MercerMurdoch InstituteRoyal Children’s HospitalFlemington Rd, PARKEVILLE 3052, VICAUSTRALIA

Dr Shadi MoghaddasNRCET 39 Kessels RoadCoopers Plains, BRISBANE 4108, QLDAUSTRALIA

Professor Michael R MooreDirector, NRCET39 Kessels Road,Coopers Plains, BRISBANE QLD 4108 AUSTRALIA

Mr Jack C. NgNRCET, 39 Kessels RoadCoopers Plains, BRISBANE QLD 4108 AUSTRALIA

Dr Edward V. OhanianOffice of Water,US EPA Health & Ecological CriteriaDivision (4304),401 M st. SW, WASHINGTON DC ,USA

Professor Agneta OskarssonSwedish University of Agricultural SciencesFaculty of Veterinary Medicine,Dept. of Food Hygiene,PO Box 7009, S-750 07 UPPSALA, SWEDEN

Dr Arun PrakashNRCET, 39 Kessels RoadCoopers Plains, BRISBANE QLD 4108 AUSTRALIA

Ms Lixia QiNRCET, 39 Kessels RoadCoopers Plains, BRISBANE QLD 4108 AUSTRALIA

Dr Ross SadlerQHSS GCL,39 Kessels Rd,Coopers Plains BRISBANE 4108, QLD AUSTRALIA

Dr Seema SethiLady Hardenge Medical College & S.M.T.Sucheta Kripalani Hospital,Department of Pathology,NEW DEHLI 11001, INDIA

Mr Ross SmithBHP R & D Environment,41 Goldieslie Rd,INDOORPILLY 4108, QLD AUSTRALIA

Dr Jenny StauberCSIRO Division of Coal and Energy TechnologyCentre for Advanced Analytical Chemistry,Private Mail Bag 7BANGOR 2234, NSW AUSTRALIA

Dr Karl H. SummerGSF-National Research Centre for Environment andHealthInstitute of Toxicology,Ingolstädter Landstraße 1, D-85758 NEUHERBERG,GERMANY

Dr Michael TaylorNZ Ministry of Health133 Molesworth StPO bx 5013, WELLINGTON NEW ZEALAND

Dr J. Hans M. TemminkLandbouwuniversiteitAgricultural University,Department of Toxicology,PO Box 8000, 6700 EA WACHENINGEN,THE NETHERLANDS

Dr Luba TomaskaCommonwealth Dept. of Health & Family ServicesPO Box 9848,CANBERRA 2601, ACT AUSTRALIA

Dr Ricardo UauyUniversidad de Chile (INTA)Instituto de Nutrición yTecnología de los AlimentosCasilla 138-11,SANTIAGO DE CHILE, CHILE

Mr Andrew M. WagnerCommonwealth Department of Health & Family ServicesChemicalsPolicy & Assessment UnitMail Drop point 88, GPO Box 9848,CANBERRA 2601 ACT AUSTRALIA

Dr Jason M. WeeksInstitute of Terrestrial EcologyMonks Woods Abots RiptonHuntingdon, CAMBRIDGESHIRE PE17 2LS,UNITED KINGDOM

Mr John WilliamsonCopper Development Association of AustraliaM M Kembla CopperSystems, Gloucester BoulevardePORT KEMBLA 2505, NSW AUSTRALIA