INTERNATIONAL PROGRAMME ON CHEMICAL SAFETY ENVIRONMENTAL HEALTH CRITERIA 101 METHYLMERCURY This report contains the collective views of an international group of experts and does not necessarily represent the decisions or the stated policy of the United Nations Environment Programme, the International Labour Organisation, or the World Health Organization. Published under the joint sponsorship of the United Nations Environment Programme, the International Labour Organisation, and the World Health Organization World Health Organization Geneva, 1990 The International Programme on Chemical Safety (IPCS) is a joint venture of the United Nations Environment Programme, the International Labour Organisation, and the World Health Organization. The main objective of the IPCS is to carry out and disseminate evaluations of the effects of chemicals on human health and the quality of the environment. Supporting activities include the development of epidemiological, experimental laboratory, and risk-assessment methods that could produce internationally comparable results, and the development of manpower in the field of toxicology. Other activities carried out by the IPCS include the development of know-how for coping with chemical accidents, coordination of laboratory testing and epidemiological studies, and promotion of research on the mechanisms of the biological action of chemicals. WHO Library Cataloguing in Publication Data Methylmercury. (Environmental health criteria ; 101) 1.Methylmercury compounds 2. Mercury poisoning I.Series ISBN 92 4 157101 2 (NLM Classification: QV 293) ISSN 0250-863X The World Health Organization welcomes requests for permission to reproduce or translate its publications, in part or in full. Applications and enquiries should be addressed to the Office of Publications, World Health Organization, Geneva, Switzerland, which will be glad to provide the latest information on any changes made to the text, plans for new editions, and reprints and translations already available. (c) World Health Organization 1990 Publications of the World Health Organization enjoy copyright protection in accordance with the provisions of Protocol 2 of the Universal Copyright Convention. All rights reserved. The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of the Secretariat of the World Health Organization concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. The mention of specific companies or of certain manufacturers' products does not imply that they are endorsed or recommended by the World Health Organization in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters. CONTENTS ENVIRONMENTAL HEALTH CRITERIA FOR METHYLMERCURY 1. SUMMARY AND CONCLUSIONS 1.1. Identity, physical and chemical properties, analytical methods 1.2. Sources of human and environmental exposure 1.3. Environmental transport, distribution, and transformation 1.4. Environmental levels and human exposure 1.5. Kinetics and metabolism 1.6. Effects on experimental animals and in vitro systems 1.7. Effects on man - mechanism of action 1.8. Conclusions 2. IDENTITY, PHYSICAL AND CHEMICAL PROPERTIES, ANALYTICAL METHODS 2.1. Identity 2.2. Physical and chemical properties 2.3. Conversion factors 2.4. Analytical methods 2.4.1. Sampling 2.4.2. Analytical procedures 2.4.3. Quality control and quality assurance 3. SOURCES OF HUMAN AND ENVIRONMENTAL EXPOSURE 3.1. Natural occurrence 3.2. Man-made sources 3.3. Uses 4. ENVIRONMENTAL TRANSPORT, DISTRIBUTION, AND TRANSFORMATION 4.1. Transport and distribution between media 4.2. Biotransformation 4.3. Interaction with other physical, chemical, or biological factors 5. ENVIRONMENTAL LEVELS AND HUMAN EXPOSURE 5.1. Environmental levels 5.1.1. Air 5.1.2. Water 5.1.3. Food 5.2. General population exposure 5.2.1. Estimated daily intakes 6. KINETICS AND METABOLISM 6.1. Absorption 6.2. Distribution 6.3. Metabolic transformation 6.4. Elimination and excretion 6.5. Retention and turnover 6.6. Reference or normal levels in indicator media 6.7. Reaction with body components 7. EFFECTS ON ORGANISMS IN THE ENVIRONMENT 8. EFFECTS ON EXPERIMENTAL ANIMALS AND IN VITRO TEST SYSTEMS 8.1. Neurotoxicity and nephrotoxicity 8.2. Reproduction, embryotoxicity, and teratogenicity 8.3. Mutagenicity and related end-points 8.4. Carcinogenicity 8.5. Special studies 8.6. Factors modifying toxicity; toxicity of metabolites 9. EFFECTS ON MAN 9.1. General population exposure 9.1.1. Effects on adults 184.108.40.206 Effects on the nervous system 220.127.116.11 Effects on non-nervous tissue 9.1.2. Effects on developing tissues 18.104.22.168 Effects on the nervous system 9.2. Occupational exposure 9.3. Mechanisms of toxicity 9.3.1. The mature organism 22.214.171.124 Mechanism of selective damage 126.96.36.199 The latent period 188.8.131.52 Cellular and molecular mechanisms 9.3.2. Developing tissues 9.3.3. Summary 9.4. Dose-effect and dose-response relationships in human beings 9.4.1. Adult exposure 184.108.40.206 The Minamata and Niigata outbreaks 220.127.116.11 The Iraqi outbreak 18.104.22.168 Exposed populations in Canada 22.214.171.124 Other fish-eating populations 126.96.36.199 Special groups 188.8.131.52 Summary 9.4.2. Prenatal exposure 184.108.40.206 Iraq 220.127.116.11 Canada 18.104.22.168 New Zealand 22.214.171.124 Summary 10. EVALUATION OF HUMAN HEALTH RISKS 10.1. Exposure levels and routes 10.2. Toxic effects 10.2.1. Adults 10.2.2. Prenatal exposure 10.3. Conclusions 11. RECOMMENDATIONS 11.1. Gaps in knowledge 11.2. Preventive measures 12. PREVIOUS EVALUATIONS BY INTERNATIONAL BODIES REFERENCES APPENDIX WHO TASK GROUP ON ENVIRONMENTAL HEALTH CRITERIA FOR METHYLMERCURY Dr L. Albert, Centro de Ecodesarrollo, Xalapa, Vera Cruz, Mexico Dr L. Amin-Zaki, Al-Damluji Clinic, Al-Nasr Street, Abu Dhabi, United Arab Emirates Professor S. Araki, Kumamoto University Medical School, First Department of Internal Medicine, Kumamoto, Japan Professor M. Berlin, Department of Environmental Medicine, University of Lund, Lund, Sweden ( Chairman ) Dr P.M. Bolger, Food and Drug Administration, Public Health Service, Center for Food Safety and Applied Nutrition, Division of Toxicological Review, Washington, DC, USA Dr T. Clarkson, The University of Rochester, Environmental Health Sciences Center, Rochester, New Yorka Dr D. Dimitroff, Health and Welfare Canada, Environmental Health Services, Medical Services Branch, Ottawa, Ontario, Canada Dr L. Magos, Medical Research Council, MRC Toxicology Unit, Carshalton, Surrey, United Kingdom ( Rapporteur ) Dr D. Marsh, The University of Rochester, Department of Neurology, Rochester, New York, USA Dr J. Piotrowski, Medical Academy in Lodz, Institute of Environmental Research and Bioanalysis, Lodz, Poland Professor A. Renzoni, Department of Environmental Biology, University of Siena, Siena, Italy Dr C. Shamlaye, Ministry of Health & Social Services, Botanical Gardens, Republic of Seychellesa Dr P. Stegnar, "Josef Stefan" Institute, Department of Nuclear Chemistry, Ljubljana, Yugoslavia Professor S. Yamaguchi, Institute of Community Medicine, University of Tsukuba, Tsukuba City, Japan ( Vice- Chairman ) Observers Dr M. Ancora, Centro Italiano Studi e Indagini, Piazza Capranica, Rome, Italy Professor M. Fujiki, Institute of Community Medicine, University of Tsukuba, Tsukuba City, Japan Miss M. Horvat, "Josef Stefan" Institute, Department of Nuclear Chemistry, Ljubljana, Yugoslavia Professor A. Igata, Kagoshima University, Kagoshima City, Japana Professor C. Maltoni, Institute of Oncology, Bologna, Italy Professor A.A.G. Tomlinson, IBE-Rome Research Area, Rome, Italy Secretariat Dr G.C. Becking, International Programme on Chemical Safety, Interregional Research Unit, World Health Organization, Research Triangle Park, North Carolina, USA ( Secretary ) Dr L.J. Saliba, WHO/EURO Project Office, Mediterranean Action Plan, Athens, Greece a Invited but unable to attend. NOTE TO READERS OF THE CRITERIA DOCUMENTS Every effort has been made to present information in the criteria documents as accurately as possible without unduly delaying their publication. In the interest of all users of the environmental health criteria documents, readers are kindly requested to communicate any errors that may have occurred to the Manager of the International Programme on Chemical Safety, World Health Organization, Geneva, Switzerland, in order that they may be included in corrigenda, which will appear in subsequent volumes. * * * A detailed data profile and a legal file can be obtained from the International Register of Potentially Toxic Chemicals, Palais des Nations, 1211 Geneva 10, Switzerland (Telephone No. 7988400 - 7985850). ENVIRONMENTAL HEALTH CRITERIA FOR METHYLMERCURY A WHO Task Group on Environmental Health Criteria for Methylmercury met in Bologna, Italy, at the Provincia from 5 to 9 June 1989. The meeting was sponsored by the Italian Ministry of the Environment and organized locally by the Institute of Oncology and Environmental Sciences with the assistance of the Provincial Government. Dr C. Maltoni, Director of the Institute, welcomed the participants on behalf of the host institution and the local governments, and Dr M. Ancora, C.I.S.I., spoke on behalf of the Minis- try of the Environment. Dr M. Mercier, Manager, IPCS, addressed the meeting on behalf of the three cooperating organizations of the IPCS (ILO/UNEP/WHO), reviewing the accomplishments of the Programme over the last few years. The Task Group made minor revisions to the draft document and made an evaluation of the human health risks from exposure to methylmercury. The efforts of DR T. CLARKSON, University of Rochester, Rochester, New York, USA, who prepared the first two drafts of this document, and all others who helped in its preparation and finalization are gratefully acknowledged. Dr G. Becking and Dr P.G. Jenkins, both members of the IPCS Central Unit, were responsible for the overall scientific content and technical editing, respectively. * * * Financial support for the meeting was provided by the Ministry of the Environment of Italy, and the Centro Italiano Studi e Indagini and the Institute of Oncology, Bologna, contributed to the organization and provision of meeting facilities. 1. SUMMARY AND CONCLUSIONS This monograph focuses on the risks to human health from compounds of monomethylmercury and examines the data that have become available since the publication of En- vironmental Health Criteria 1: Mercury (WHO, 1976b). The environmental effects of mercury are discussed in Environ- mental Health Criteria 86: Mercury - Environmental Aspects (WHO, 1989a). 1.1 Identity, Physical and Chemical Properties, Analytical Methods The solubility of methylmercury compounds in water varies greatly and depends on the nature of the anion. Most are soluble in water but much less soluble in non- polar solvents. They generally have appreciable vapour pressures at room temperature. Mercurials, including alkylmercurials, exhibit high affinities for sulfhydryl groups. Blood samples for analysis should be taken by veni- puncture, avoiding devices using mercury-containing pre- servatives. Current methods are capable of measuring mer- cury in 1- to 5-ml samples of whole blood, even in the case of non-exposed individuals. Hair is useful in assessing exposure to methylmercury in the diet and may be sampled as single or bunched strands. The single-strand procedure requires both sensitive analytical methods and the determination of the growth phase of the hair. The method of choice for determining total mercury in environmental and biological samples is flameless atomic absorption spectroscopy (detection limits, 0.5-4.0 ng/g). Neutron activation analysis serves as a sensitive refer- ence method. Gas chromatography is used to determine meth- ylmercury directly (detection limit, 1.0 ng/g sample). 1.2 Sources of Human and Environmental Exposure Environmental methylmercury arises largely, if not solely, from the methylation of inorganic mercury. The major source of atmospheric mercury is the natural degassing of the earth's crust, amounting to 2700-6000 tonnes per year. Deposition of atmospheric mercury, leaching from rocks, and anthropogenic sources all add to the mercury burden in bodies of water, but the exact contribution of each source is indeterminable. About 10 000 tonnes of mercury per year are mined, subject to considerable year-to-year variation. Other im- portant man-made sources are fossil fuels combustion, smelting of sulfide ores, production of cement, and refuse incineration. The total man-made global release of mercury to the atmosphere is approximately 2000-3000 tonnes per year, i.e., less than the natural emissions. Man-made emissions pose the greatest risk when they are released in confined areas. Mercury continues to be used in the production of caustic soda and chlorine, and it is widely used in the electrical industry for lamps, controls, rectifiers, bat- teries and switches, as well as in the dental profession. Environmental losses can also occur from its continued use in antifouling and mildew-proofing paints, in seed dress- ings, and in the extraction of gold. 1.3 Environmental Transport, Distribution, and Transformation There is a well recognized global cycle for mercury, whereby emitted mercury vapour is converted to soluble forms (e.g., Hg++) and deposited by rain onto soil and water. Mercury vapour has an atmospheric residence time of between 0.4 and 3 years, whereas soluble forms have resi- dence times of a few weeks. Transport in soil and water is thus limited, and it is likely that deposition will occur within a short distance. The change in speciation of mercury from inorganic to methylated forms is the first step in the aquatic bioac- cumulation process. Methylation can occur non-enzymically or through microbial action. Once methylmercury is released, it enters the food chain by rapid diffusion and tight binding to proteins. As a result of food-chain bio- magnification, highest levels are found in the tissues of such predatory species as freshwater trout, pike, walleye, bass and ocean tuna, swordfish, and shark. The bioconcen- tration factor, i.e., the ratio of the concentration of methylmercury in fish tissue to that in water, is usually between 10 000 and 100 000. Levels of selenium in the water may affect the availability of mercury for uptake into aquatic biota. Reports from Sweden and Canada suggest that methylmercury concentrations in fish may increase following the construction of artificial water reser- voirs. 1.4 Environmental Levels and Human Exposure The general population is primarily exposed to methyl- mercury through the diet. However, air and water, de- pending upon the level of contamination, can contribute significantly to the daily intake of total mercury. In most foodstuffs, mercury is largely in the inorganic form and below the limit of detection (20 µg mercury/kg fresh weight). However, fish and fish products are the dominant source of methylmercury in the diet, and levels greater than 1200 µg/kg have been found in the edible portions of shark, swordfish, and Mediterranean tuna. Similar levels have been found in pike, walleye, and bass taken from polluted fresh waters. It has been estimated that humans have a daily intake of about 2.4 µg methylmercury from all sources, and a daily uptake of approximately 2.3 µg. The total daily intake of all forms of mercury from all sources has been estimated to be 6.7 µg, with an added burden of 3.8 to 21 µg of mercury vapour from dental amalgams, if present. The level of mercury in fish, even for humans consuming only small amounts (10-20 g of fish/day), can markedly affect the intake of methylmercury. The consumption of 200 g of fish containing 500 µg mercury/kg will result in the intake of 100 µg mercury (predominantly methyl- mercury). This amount is one-half of the recommended pro- visional tolerable weekly intake (WHO 1989b). 1.5 Kinetics and Metabolism Methylmercury in the human diet is almost completely absorbed into the bloodstream and distributed to all tis- sues within about 4 days. However, maximum levels in the brain are only reached after 5-6 days. In humans, blood to hair ratios are about 1:250, with appreciable individual variation. Similarly, large individual differences are seen in cord to maternal blood mercury ratios, the levels generally being higher in cord blood. Species differences exist in the distribution of methylmercury between red blood cells and plasma (about 20:1 in humans, monkeys, and guinea-pigs, 7:1 in mice, and >100:1 in rats). Methylmercury is converted to inorganic mercury in experimental animals and humans. The duration of the ex- posure and the interval after its cessation, determine the fraction of total mercury present in tissues in the Hg++ form. In humans, after high oral intakes of methyl- mercury for 2 months, the following values were reported (percentage of total mercury in tissues as inorganic mer- cury): whole blood, 7%; plasma, 22%; breast milk, 39%; urine, 73%; liver, 16-40%. The rate of excretion of mercury in both laboratory animals and humans is directly proportional to the simul- taneous body burden and can be described by a single- compartment model with a biological half-time, in fish- eating humans, of 39-70 days (average approximately 50 days). Lactating females have significantly shorter half- times for mercury excretion than non-lactating ones. Mercury half-times in hair closely follow those in blood but with wider variation (35-100 days, average 65 days). Suckling mice are incapable of excreting methylmer- cury, but they abruptly change to the adult rate of ex- cretion at the end of the suckling period. In the case of continuous exposure, a single-compart- ment model with a 70-day half-time predicts that the whole-body steady state (where intake equals excretion) will be attained within approximately one year and that the maximum amount accumulated will be 100 times the aver- age daily intake. The validity of the single-compartment model is supported by the reasonable agreement between predicted and observed blood concentrations of methylmer- cury in single-dose tracer studies, single-dose fish intake experiments, and studies involving the extended controlled intake of methylmercury from fish. It is also supported by results from the longitudinal hair analysis of individuals with very high intakes of methylmercury. Mean reference values for total mercury in commonly used indicator media are: whole blood, 8 µg/litre; hair, 2 µg/g; urine, 4 µg/litre, and placenta, 10 µg/kg wet weight. Long-term fish consumption is the major determi- nant of methylmercury and, usually, total mercury levels in blood. For example, in communities in which there is a long-term daily consumption of 200 µg mercury/day from fish, blood mercury levels are approximately 200 µg/litre and corresponding hair levels about 250 times higher (50 µg/g hair). 1.6 Effects on Experimental Animals and In Vitro Systems In every animal species studied, the nervous system is a target of methylmercury, fetuses appearing to be at higher risk than adults. Concerning effects on the nervous system, animal studies reported since 1976 provide further support to the mechanistic models used to evaluate the available data in humans (summarized in section 1.7). Methylmercury is fetoxic in mice (single dose of 2.5-7.5 mg/kg); teratogenic in rats, and adversely affects the behaviour of monkey offspring (mercury doses of 50-70 µg/kg per day before and during pregnancy). It also affects spermatogenesis in mice (1 mg mercury/kg as methylmercury). 1.7 Effects on Man - Mechanism of Action The effects of methylmercury on adults differ both qualitatively and quantitatively from the effects seen after prenatal or, possibly, early postnatal exposures. Thus, effects on the mature human being will be considered separately from the effects on developing tissues. The clinical and epidemiological evidence indicates that prenatal life is more sensitive to the toxic effects of methylmercury than is adult life. The inhibition of protein synthesis is one of the earliest detectable bio- chemical effects in the adult brain, though the sequence of events leading to overt damage is not yet understood. Methylmercury can also react directly with important receptors in the nervous system, as shown by its effect on acetylcholine receptors in the peripheral nerves. In the case of prenatal exposure, the effects of methylmercury seem to be quite different and of a much more general basic nature. It affects normal neuronal development, leading to altered brain architecture, heterotopic cells, and decreased brain size. Methylmercury may also be exerting an effect, perhaps through inhibition of the microtubular system, on cell division during critical stages in the formation of the central nervous system. Since 1976, a wealth of data has been reported on dose-effect and dose-response relationships in humans. It has been derived from in-depth studies on populations exposed to methylmercury through mass poisonings or through the consumption of fish containing varying levels of methylmercury. Again, prenatal and adult data will be considered separately in view of the differences, both qualitative and quantitative, in effects and dose-response relationships. In adults, the reported relationships between response and body burden (hair or blood mercury concen- trations) are essentially the same as those reported in Environmental Health Criteria 1: Mercury (WHO, 1976b). No adverse effects have been detected with long-term daily methylmercury intakes of 3-7 µg/kg body weight (hair mercury concentrations of approximately 50-125 µg/g). Pregnant women may suffer effects at lower methylmercury exposure levels than non-pregnant adults, suggesting a greater risk for pregnant women. Severe derangement of the developing central nervous system can be caused by prenatal exposure to methylmer- cury. The lowest level (maximum maternal hair mercury concentration during pregnancy) at which severe effects were observed was 404 µg/g in the Iraqi outbreak and the highest no-observed-effect level for severe effects was 399 µg/g. Fish-eating populations in Canada and New Zealand have also been studied for prenatal effects, but exposure levels were far below those that caused effects in Iraq and no severe cases were seen. Evidence of psychomotor retardation (delayed achieve- ment of developmental milestones, a history of seizures, abnormal reflexes) was seen in the Iraqi outbreak at maternal hair levels below those associated with severe effects. The extrapolation of data suggested that one of these effects (motor retardation) rose above the back- ground frequency at maternal hair levels of 10-20 µg/g. The Canadian study reported that abnormal muscle tone or reflexes were positively associated with maternal hair levels in boys but not in girls (the highest maternal hair level during pregnancy was 23.9 µg/g). The New Zealand study reported evidence of developmental retardation (according to the Denver Test) in 4-year-old children at average maternal hair mercury levels during pregnancy within the range of 6-86 µg/g (the second highest value was 19.6 µg/g). The New Zealand mercury values should be multiplied by 1.5 to convert to maximum maternal hair levels in pregnancy. 1.8 Conclusions The general population does not face a significant health risk from methylmercury. Certain groups with a high fish consumption may attain a blood methylmercury level (about 200 µg/litre, corresponding to 50 µg/g of hair) associated with a low (5%) risk of neurological damage to adults. The fetus is at particular risk. Recent evidence shows that at peak maternal hair mercury levels above 70 µg/g there is a high risk (more than 30%) of neurological dis- order in the offspring. A prudent interpretation of the Iraqi data implies that a 5% risk may be associated with a peak mercury level of 10-20 µg/g in maternal hair. There is a need for epidemiological studies on chil- dren exposed in utero to levels of methylmercury that re- sult in peak maternal hair mercury levels below 20 µg/g, in order to screen for those effects only detectable by available psychological and behavioural tests. 2. IDENTITY, PHYSICAL AND CHEMICAL PROPERTIES, ANALYTICAL METHODS 2.1 Identity The primary constituent is the element mercury (CAS registry number 7439-97-6), which has a relative atomic mass of 200.59. In the inorganic form, mercury exists in three oxidation states: Hg° (metallic); Hg2++ (mercu- rous); and Hg++ (mercuric). The mercurous and mercuric states can form numerous inorganic and organic chemical compounds. The organic forms are those in which mercury is attached covalently to at least one carbon atom. This monograph focuses on the risk to human health of the compounds of monomethylmercury. The generic term "methylmercury" is used throughout this text to rep- resent monomethylmercury compounds. In many cases the exact identity of these compounds is not known except that the methylmercury cation, CH3Hg+, is associated either with a simple anion, like chloride, or a large molecule (e.g., a protein) with negative and positive charges. Other physical and chemical forms of mercury are dis- cussed in this monograph where they are relevant to the full evaluation of the risks to human health of methylmer- cury: for example, the atmospheric transport of elemental mercury vapour (Hg°), its deposition and oxidation in natural waters, and the subsequent methylation of inor- ganic mercury (Hg++). 2.2 Physical and Chemical Properties In its elemental form, mercury at room temperature is a heavy silvery liquid. At 20 °C the specific gravity of the metal is 13.456 and the vapour pressure is 0.16 Pa (0.0012 mmHg). Thus the saturated atmosphere at 20 °C con- tains mercury vapour at a concentration of approximately 15 mg/m3. This concentration is over 200 fold greater than the currently accepted concentrations for occu- pational exposure. It is of interest that certain forms of mercury, such as the methyl and ethyl derivatives, have appreciable vapour pressure at room temperature. Thus, the vapour pressure of methylmercuric chloride is 1.13 Pa (0.0085 mmHg) and the vapour pressure of dimethylmercury is sev- eral times greater. Mercurials differ greatly in their solubilities. Solubility in water increases in the order: mercurous chloride; elemental mercury; methylmercuric chloride; mercuric chloride. Certain species of mercury are soluble in non-polar solvents. These include elemen- tal mercury and the halide compounds of alkylmercurials. From the biochemical point of view the most important chemical property of mercuric mercury and alkylmercurials is their high affinity for sulfhydryl groups. 2.3 Conversion Factors 1 ppm = 1 mg/kg = 1 µg/g = 1 ng/mg 1 ppb = 1 µg/kg = 1 ng/g 1 µmol mercury • 1 µmol methylmercury • 200 µg mercury 2.4 Analytical Methods 2.4.1 Sampling Many different sampling procedures are used in the measurement of mercury. Procedures for environmental sampling in air, water, soil, and aquatic and animals species are beyond the scope of this monograph. Since its purpose is to evaluate the risks to human health, only the sampling of human indicator media and tissues will be con- sidered. Blood samples should be taken by venipuncture, the most convenient method being the use of heparinized "Vacutainers"a. Some commercial containers may contain a mercury compound added as a preservative. It is wise to analyse each commercial batch for mercury before use. The sample should be refrigerated but not frozen, as it is sometimes useful to measure mercury in plasma and red cells separately. The analysis should be carried out as soon as possible to avoid haemolysis of the sample. If the sample has clotted or if extensive haemolysis has occurred, the sample should be homogenized before aliquots are taken for analysis. Current methods are capable of measuring mercury in 1- to 5-ml samples of whole blood even in the case of non-exposed individuals. Urine sampling is not useful for individuals exposed to methylmercury, because little is excreted by this route. Hair samples are important in assessing exposure to methylmercury in the diet. Methylmercury in non-occu- pationally exposed individuals is incorporated into hair at the time the hair is formed, the methylmercury concen- tration in newly formed hair being proportional to its simultaneous concentration in blood. Once incorporated into the hair strand, its concentration remains unchanged. Thus, longitudinal analysis along a strand of hair pro- vides a recapitulation of previous blood levels. Since hair grows at about 1 cm per month, recapitulation is possible over several months or years, depending on the length of the hair sample. a Trade name of heparinized test-tube manufactured by Becton & Dickinson, USA, and used for blood sample collection. There are two sampling methods, single strands and bunched strands. The former requires a more sensitive method and the determination of the growth phase (anaphase and telophase) of each strand by the microscopic examin- ation of the hair root. However, most methods require at least 1 mg hair and, preferably, about 10 mg. Thus, if the hair is measured in 1-cm lengths, it is necessary to have about 50 strands. The best sampling procedure is to locate 50 strands of the longest hair on the head, hold them in place with a haemostat, and cut them as close to the scalp as possible with surgical scissors. The strands are tied with a cotton thread before the haemostat is released to ensure that the individual strands remain in the same alignment. The tied bunch of hair may be stored in a plas- tic bag or envelope until it is analysed. Bunch analysis tends to underestimate peak concentrations due to the different growth rates of individual hairs and to mechan- ical displacement of individual strands during collection and subsequent handling (Giovanoli-Jakubczak & Berg, 1974; Cox et al., in press). Single-strand analysis can give a more precise temporal recapitulation and avoids certain artifacts found in bunched-strand analysis. Agreement between concentrations of mercury in individual hair strands collected from the same person at the same time is within 10%. Nevertheless it is wise to collect more than one strand to guard against accidental contamination or breakage. 2.4.2 Analytical procedures The methods summarized in Table 1 have been selected from a large number of publications. They are typical of the various methods available for analysis of total mer- cury and its inorganic or organic species. All represent a considerable improvement on the orig- inal "dithizone" method. This method was widely used up to the introduction of atomic absorption in the late 1960s. Basically it involved the formation of a coloured complex with dithizone after all the mercury in the sample had been converted to Hg++ compounds by oxidation in strong acids. After neutralization of excess oxidant with a reducing agent, usually hydroxylamine, the coloured com- plex was extracted into a non-polar solvent. After washing the extract, the colour intensity was measured on a spec- trophotometer and the amount of mercury estimated from a standard curve. The limit of detection was of the order of 1-10 µg mercury so that large quantities of sample were required for such media as blood and hair. The neutron activation procedure is regarded as the most accurate and sensitive procedure and is usually used as the reference method (WHO, 1976b). The "Magos" selec- tive atomic-absorption method (Magos, 1971; Magos and Clarkson, 1972) has found wide application. It can deter- mine both total and inorganic mercury and, by difference, organic mercury. The apparatus is inexpensive, portable, and does not require sophisticated facilities. The gas chromatography method is usually used when there is a need to selectively measure methylmercury or other organic species. It has been widely used for the measurement of methylmercury in fish tissues. An alterna- tive approach is the separation of methylmercury from in- organic mercury by volatilization (Zelenko & Kosta, 1973), ion exchange (May et al., 1987), or distillation (Horvat et al., 1988a), and the estimation of the separated methylmercury by non-selective methods (e.g., atomic ab- sorption). Table 1. Analytical methods for the determination of total inorganic and methylmercury --------------------------------------------------------------------------------------------------------- Media Speciation Analytical Method Detection Comments References Limit (ng mercury/ g) --------------------------------------------------------------------------------------------------------- Food, tissues total mercury atomic absorption 2.0 method has many Hatch & Ott adaptations (1969) (see Peter & Strunc, 1984) Blood, urine, total mercury atomic absorption 0.5 also estimates Magos & hair, tissues inorganic mercury organic mercury Clarkson (1972) as difference between total and inorganic Blood, urine, total mercury atomic absorption 2.5 automated form of Farant et al. hair, tissues inorganic mercury the Magos Method (1981) Blood, urine, total mercury atomic absorption 4.0 automated form of Coyle & Hartley hair, tissues inorganic mercury the Magos Method (1981) Food, tissues, methylmercury gas chromatography 1.0 based on the Von Burg et al. biological fluids electron-capture original method (1974) of Westoo (1968) Cappon & Smith (1978) All media total mercury neutron activation 0.1 reference method Kosta & Byrne (1969) Byrne & Kosta (1974) --------------------------------------------------------------------------------------------------------------------------------------------- The estimation of total mercury in a single strand of hair by X-ray fluorescence has been described by Jaklevic et al. (1978). Emulsion autoradiography has been widely used in ex- perimental studies of tissue deposition of the radioiso- topes of mercury. However it should be noted that photo- graphic emulsions are also sensitive to non-radioactive inorganic forms of mercury (Rodier & Kates, 1988). It is necessary to note that, since methylmercury is not a sample contaminant, external contamination does not interfere with methylmercury-specific methods. Greater care is required when the method is sensitive to inorganic mercury contamination (Mushak, 1987). 2.4.3 Quality control and quality assurance The analysis of most samples of hair or blood involves very small quantities of mercury (in the ng or even sub-ng ranges). Therefore, considerable attention should be paid to procedures that will ensure accurate analytical data. The general considerations of quality control and quality assurance have been discussed at a recent WHO-sponsored conference on Biological Monitoring of Toxic Metals (Friberg, 1988). A Global Environmental Monitoring System (GEMS) programme has been described in which a new ap- proach to interlaboratory comparisons has been success- fully introduced on an international basis. Specific quality-control programmes for mercury using the GEMS approach have been described by Friberg (1983) and Lind et al. (1988a). 3. SOURCES OF HUMAN AND ENVIRONMENTAL EXPOSURE 3.1 Natural Occurrence As the predominant, if not only, source of environ- mental methylmercury is the methylation of inorganic mer- cury, we need to examine the environmental movement of the inorganic species if we are to understand the origins of human exposure to methylmercury. Thus, this section deals largely with the environmental aspects of elemental mer- cury vapour and inorganic compounds of mercury. The major natural sources of mercury (Fig. 1) are degassing of the earth's crust, emissions from volcanoes, and evaporation from natural bodies of water (National Academy of Sciences, 1978; Nriagu, 1979; Lindqvist et al., 1984). The most recent estimates indicate that natural emissions amount to 2700-6000 tonnes per year (Lindberg et al., 1987). The earth's crust is also an important source of mer- cury for bodies of natural water. Some of this mercury is undoubtedly of natural origin, but some may have been deposited from the atmosphere and may, ultimately, have been generated by human activities (Lindqvist et al., 1984). Thus it is difficult to assess quantitatively the relative contributions of natural and anthropogenic mer- cury to run-off from land to natural bodies of water. 3.2 Man-Made Sources The world-wide mining of mercury is estimated to yield about 10 000 tonnes/year, but this figure varies consider- ably from year to year, depending on the commercial value of the metal. Mining activities also result in losses of mercury through the dumping of mine tailings and direct discharges to the atmosphere. The Almaden mercury mine in Spain, which accounts for 90% of the total output of the European Community, was expected to produce 1380 tonnes in 1987 (Seco, 1987). Other important man-made sources are the combustion of fossil fuels, the smelting of metal sulfide ores, the production of cement, and refuse incin- eration. Using Sweden as a specific example (Swedish Environmental Protection Board, 1986), the mercury emissions to the atmosphere in 1984 were (in kg/year): incineration of household waste (3300); smelting (900); chloralkali industry (400); crematories (300); mining (200); combustion of coal and peat (200); other sources (200). The total man-made global release to the atmosphere has been estimated to be 2000-3000 tonnes/year (Lindberg et al., 1987; Pacyna, 1987). It should be stressed that there are considerable uncertainties in the estimated fluxes of mercury in the environment and in its speciation. Concentrations in the unpolluted atmosphere and in natural bodies of water are so low as to be near the limit of detection of current analytical methods, even for the determination of total mercury. Anthropogenic releases of mercury into confined areas can be the source of high toxicity risk even though these releases may be small relative to global emissions. The point is relevant to the contamination of Minamata Bay and the Agano River in Niigata, Japan, as well as to inadver- tent poisoning via contaminated bread in Iraq. 3.3 Uses A major use of mercury is as a cathode in the elec- trolysis of sodium chloride solution to produce caustic soda and chlorine gas. Quantities of the order of 10 tonnes of liquid metal are used in each manufacturing plant. In most industrialized nations, stringent pro- cedures have been taken to reduce losses of mercury. Mer- cury is widely used in the electrical industry (lamps, arc rectifiers, and mercury battery cells), in control instru- ments in the home and industry (switches, thermostats, barometers), and in other laboratory and medical instru- ments. It is also widely used in the dental profession for tooth amalgam fillings. In certain countries, liquid metallic mercury is still used in gold extraction. Mercury compounds continue to be used in anti-fouling and mildew- proofing paints and to control fungal infections of seeds, bulb plants, and vegetation. WHO has warned against the use of alkylmercury compounds in seed dressing (WHO, 1976a). Methylmercury compounds are still used in labora- tory-based research, and so the possibility of occu- pational exposures remains (Junghans, 1983). 4. ENVIRONMENTAL TRANSPORT, DISTRIBUTION, AND TRANSFORMATION 4.1 Transport and Distribution Between Media Human exposure to mercury should first be viewed in the context of the world-wide circulation of this highly mobile metal (Fig. 1). The vapour of metallic mercury, hereinafter referred to as mercury vapour or Hg°, is released into the atmosphere from a number of natural sources (section 3.1). Man-made emissions, mainly from the combustion of fossil fuels, form about 25% of the total emissions to the atmosphere. However, the anthropogenic contribution is greater in the northern than in the southern hemisphere and becomes the major form of emission in heavily industrialized areas, such as western Europe. The distribution constants of various mercury compounds between air and water are given in Table 2. Clearly, Hg° and dimethylmercury ((CH3)2Hg), as a result of their air/water distribution coefficients, are most likely to be found in the atmosphere. Table 2. Experimentally determined distribution constants for some compounds of relevance for the mercury cyclea ---------------------------------------------------------- Compound HgX (air)/ Temperature Cl- HgX (water) (°C) ionic strength (v/v) (mol) ---------------------------------------------------------- Hg° 0.29 20 0 (CH3)2Hg 0.31 25 0 (CH3)2Hg 0.15 0 0 CH3HgCl 1.9 x 10-5 25 0.7 CH3HgCl 1.6 x 10-5 15 1 CH3HgCl 0.9 x 10-5 10 0.2 x 10-3 Hg(OH)2 3.2 x 10-6 25 0.2 x 10-3 Hg(OH)2 1.6 x 10-6 10 0.2 x 10-3 HgCl2 2.9 x 10-8 25 0.2 x 10-3 HgCl2 1.2 x 10-8 10 0.2 x 10-3 ---------------------------------------------------------- a Adapted from: Lindqvist et al. (1984). The solubility of mercury vapour in water is not high enough to account for the concentrations of mercury found in rain water. Thus, Lindqvist et al. (1984) suggested that a small fraction of mercury vapour is converted to a water-soluble species, probably Hg++, which is deposited on land and water in rain. However, the putative water- soluble forms have yet to be positively identified. Par- ticulate forms account for less than 1% of total mercury in the atmosphere but may make an important contribution to mercury in rain water. The residence time of mercury vapour is estimated to be between 0.4 and 3 years, and as a consequence, mercury vapour is globally distributed. The soluble form is assumed to have a residence time of the order of weeks, and therefore the distance over which it may be transported is limited. The extremely low concen- trations in the atmosphere (section 5.1.1) present formi- dable difficulties both in the analysis of total mercury and in the identification and measurement of chemical and physical species. For example, methylmercury compounds have been reported in the air above polluted areas in the USA (WHO, 1976b), but their presence in unpolluted air still needs to be confirmed. Shimojo et al. (1976) found methyl donors in car exhaust gases, but not methylmercury in the ambient air. Mercury deposited on land and open water is, in part, re-emitted to the atmosphere as Hg°. This emission, deposition, and re-emission ("ping-pong" effect) creates difficulties in tracing the movement of mercury to its source. The bottom sediment of the oceans is thought to be the ultimate sink where mercury is deposited in the form of the highly insoluble mercuric sulfide. Recently, an expert group suggested that atmospheric mercury vapour could be taken up directly by plant foliage and that this might be an important pathway to watersheds in highly forested areas (Lindberg et al., 1987). 4.2 Biotransformation Despite the uncertainties concerning speciation, the global cycle of mercury is believed to involve almost exclusively the inorganic forms. These forms do not ac- cumulate in human food chains except in uncommon items, such as mushrooms (Minagawa et al., 1980). The change in speciation from inorganic to methylated forms is the first crucial step in the aquatic bioaccumulation process. Methylation takes place mostly on sediments in fresh and ocean waters but also in columns of fresh and sea waters (Lindberg et al., 1987). Fish intestinal contents (Rudd et al., 1980) and the outer slime of fish have also been found to methylate inorganic mercury (McKone et al., 1971; Jernelov, 1972; Rudd et al., 1980). The mechanism of synthesis of methylmercury compounds (both CH3Hg+ and (CH3)2Hg) is now well understood (Wood & Wang, 1983). Methylation of inorganic mercury involves the non-enzymic methylation of Hg++ by methyl cobalamine compounds (analogues of vitamin B12) that are produced as a result of bacterial synthesis. However, other pathways, both enzymic and non-enzymic, may play a role (Beijer & Jernelov, 1979). Factors affecting the aquatic methylation of mercury have been described by Fujiki & Tajuma (1975). Microorganisms have also been isolated that carry out the reverse reactions: CH3Hg+ -> Hg++ -> Hg° The enzymology of CH3Hg+ hydrolysis and mercuric ion reduction is now understood in some detail (Silver, 1984; Begley et al., 1986), as is the oxidation of mercury vapour to Hg++ by an enzyme that is critical to the oxy- gen cycle (catalase). These oxidation-reduction and methylation-demethylation reactions are assumed to be widespread in the environment, and each ecosystem attains its own steady state with respect to the individual species of mercury. However, owing to the bioaccumulation of methylmercury, methylation is more prevalent in the aquatic environment than demethylation. Once methylmercury is released from microorganisms, it enters the food chain by rapid diffusion and tight binding to proteins in aquatic biota. The results of a field study on the entry of methylmercury to the tuna food chain in the Mediterranean Sea fits the diffusion model (Bernhard et al., 1982). Methylmercury is rapidly accumulated by most aquatic biota and attains its highest concentration in the tissues of fish at the top of the aquatic food chain (Bernhard et al., 1982). Thus, large predatory species, such as trout, pike, walleye, and bass in fresh water and tuna, sword- fish, and shark in ocean water, contain considerably higher levels than non-predatory species (Table 3). The ratio of the concentration of methylmercury in fish tissue to that in water can be extremely large, usually of the order of 10 000 to 100 000 (US EPA, 1980). However, it should be noted that these bioconcentration ratios are not the result of partition between water and tissue but of biomagnification through the food chain. In addition to the influence of trophic level or species, factors such as the age of the fish, microbial activity and mercury in sediment (upper layer), dissolved organic content (humic content), salinity, pH, and redox potential all affect the levels of methylmercury in fish (WHO, 1989a). Methylmer- cury in freshwater fish is also affected by the catchment area of the lake and by recent flooding or diversion of rivers (see section 4.3). 4.3 Interaction with Other Physical, Chemical, or Biological Factors Following the identification of point sources of mer- cury pollution in the 1960s (Swedish Expert Group, 1971), it was discovered in the early 1970s that numerous lakes in Sweden had increased levels of methylmercury in pike, even though these lakes had not been subjected to any direct discharge of mercury. It was suggested by Hultberg & Hasselrot (1981) that three explanations should be considered: - mercury discharged into the atmosphere is washed down by precipitation or is deposited (in the dry form) in the lake; - acid precipitation causes the release of natural mer- cury or mercury deposited earlier by air that had been trapped; - acidity in lakes induces a change in the biological dynamics of the lakes, which results in a re-distri- bution of mercury in the ecologic system. The long-distance transport of mercury and the poten- tial role of acidification have become major factors con- cerning future human exposure to methylmercury. Low pH favours both the direct uptake of methylmercury through the gills of fish and dietary uptake due to increased mercury accumulation by organisms in lower trophic levels (Wiener, 1987; Xun & Campbell, 1987). According to Hultberg and Hasselrot (1981), an increase in acidity of one pH unit in a lake increases the mercury content in pike by approximately 0.14 mg/kg wet weight. Wiener (1987) reported that a change of pH from 6.1 to 5.6 increased the mercury concentration in 1-year-old yellow perch from 0.11 ± 0.002 (SEM) mg/g to 0.138 ± 0.003 mg/kg within one calendar year. The causal relationship between reduction in pH and elevated mercury levels in edible tissues of fish has not been established. Possible mechanisms include: - changes in population dynamics (a switch by pike from consumption of roach to consumption of perch); - a reduction in the total biomass where most of the methylmercury is found (the growth of fish may be retarded and, for a given size, the mercury concen- tration will be higher); - a low pH favours monomethyl versus dimethyl mercury; the latter is less avidly accumulated by fish; - a low pH may elute more mercury from sediments or soils; - as pH falls, the ratio of methylation to demethyl- ation reactions increases, thus favouring an increase in the net production of methylmercury (Ramlal et al., 1986); - Bjornberg et al. (1988) proposed that the concen- tration of the sulfide ion in water determines the bioavailability of inorganic mercury (Hg++) and, therefore, the extent of methylation and uptake by aquatic organisms. A reduction in pH will reduce the concentration of the sulfide ion making more Hg++ available for methylation. Table 3. The range of published average values of methylmercury (mg mercury/kg wet weight) in the muscle tissue of various species of fisha,b ---------------------------------------------------------------------------------- Species Atlantic Pacific Indian Mediterranean Ocean Ocean Ocean Sea ---------------------------------------------------------------------------------- Non-predators Mackerel 0.07 - 0.2 0.16 - 0.25 0.005 0.24 Sardine 0.03 - 0.06 0.03 0.006 0.15 Unspecified number of edible species 0.08 - 0.27 0.07 - 0.09 0.02 - 0.16 0.1 - 0.3 Predators Tuna 0.3 - 0.8 0.3 0.064 - 0.4 1.2 Swordfish 0.8 - 1.3 1.6 - 1.8 Shark, dogfish, ray 1.0 0.7 - 1.1 0.004 - 1.5 1.8 ---------------------------------------------------------------------------------- a Data from: US Department of Commerce (1978). b Where an analysis of methylmercury was not available, the data on total mercury has been used instead. Extensive investigations have been made in Canada in recent years to explain why methylmercury levels increase in fish when bodies of fresh water are relocated or redirected (Ramlal et al., 1985; Stokes & Wren, 1987). It is proposed that the redirecting of rivers and the forma- tion of reservoirs for hydroelectric production results in large quantities of organic material in the water, which serves as a food source for microorganisms. The resulting increase in microbial activity leads to an increase in the production of methylmercury from inorganic mercury nat- urally present in the sediment (Furutani & Rudd, 1980; Ramlal et al., 1986). This process is sustained by the repeated raising and lowering of water levels to maintain hydroelectric production, because the shorelines continue to be eroded and more vegetation enters the water. It is likely that future environmental impact statements will have to take into account this newly discovered source of methylmercury when hydroelectric schemes are planned. As noted by Bjornberg et al. (1988), "many biologi- cal, chemical and physical factors are linked to each other in the limnic ecosystem" and "many of these factors seem to be of importance for the Hg content of fish". Thus "it is not difficult to understand why it has been considered hard to find simple mechanisms explaining why certain lakes have a high mercury content in fish and others have not". They propose that the "central piece in the puzzle" is the critical influence of the sulfide ion, which forms the highly insoluble mer- curic sulfide with Hg++ (Ks = l0-52). The solubility product of mercury selenide, HgSe, is even lower (Ks = 10-58). Thus, studies made on a Canadian lake that had received a large discharge of inor- ganic mercury from a paper pulp factory suggest that the addition of selenite can reduce the availability of mer- cury for uptake into aquatic biota (Turner & Rudd, 1983). Studies on Swedish lakes confirm these findings (Björnberg et al., 1988). In these studies the selenium level was raised artificially from 0.4 to 2.4 µg/litre over a 1- to 2-year period, and the mean levels of mercury fell from 1.5 to 0.70 mg/kg in pike and from 0.56 to 0.16 mg/kg in perch. Such levels of selenium are below drinking-water standards. 5. ENVIRONMENTAL LEVELS AND HUMAN EXPOSURE 5.1 Environmental Levels There is considerable variation in mercury levels in those media that are the source of human exposure and, consequently, in their contribution to the toxicity risk. Non-occupational groups are primarily exposed through the diet. Although intake of the methylated form is of primary interest, levels of other species are summarized so as to provide a measure of total mercury intake. 5.1.1 Air Concentrations of total mercury in the atmosphere of the northern hemisphere have recently been estimated at 2 ng/m3, those in the southern hemisphere being half this value. Values in urban areas are usually higher (e.g., 10 ng/m3) (Lindqvist et al., 1984). Schroeder & Jackson (1987) found values in the range 3-27 ng/m3 (mean, 9 ng/m3) in rural areas of Canada and 5-15 ng/m3 (mean, 11 ng/m3) in urban areas. In Sweden, urban levels appear to be slightly lower (range, 0.8-13.2 ng/m3; mean, 4 ng/m3). Dental mercury fillings are reported to release mer- cury vapour into the oral cavity (Clarkson et al., 1988). The resulting concentrations in intra-oral air can sub- stantially exceed those found in the ambient atmosphere, especially after of period of chewing. Estimates of pul- monary absorption indicate that approximately 3000-17 000 ng mercury vapour enter the systemic circulation daily from this exposure. As tobacco leaves contain mercury, smoking may also contribute to inhalation exposure (Suzuki et al., 1976). As discussed in section 4.1, the major form of mercury in air is believed to be elemental mercury vapour. How- ever, the presence of methylmercury compounds in the ambi- ent atmosphere has been reported (Johnson & Braman, 1974). Recent data from the vicinity of Toronto, Canada, indi- cated the following average composition (as percentage of total mercury): Hg°, 75%; Hg++, 5%; and CH3Hg+, 20% (Schroeder & Jackson, 1987). The particulate fraction of mercury in air (as a percentage of total mercury) is usually 4% or less (Lindqvist et al., 1984). The way in which the "soluble fraction" of mercury in air (section 4.1) relates to these recent findings on individual chemi- cal species is still unclear. 5.1.2 Water Concentrations of total mercury in natural water are so low that accurate analysis is still a major problem. Values for rain water are usually within the range 5-100 ng/litre, but mean values as low as 1 ng/litre have been reported. The most recent data (Fig. 1) indicate lower values than those previously recorded (WHO, 1976b). Representative values for dissolved total mercury are: open ocean, 0.5-3 ng/litre; coastal sea water, 2-15 ng/litre; freshwater rivers and lakes, 1-3 ng/litre. The concentration range for mercury in drinking-water is the same as in rain, with an average of about 25 ng/litre (Lindqvist et al., 1984). The chemical speciation of mercury in water is still not completely defined. Mercury in ocean waters exists mainly in the form of Hg++ complexed with chloride ions. Speciation in fresh water is poorly understood. In a con- taminated lake system in Canada, methylmercury was found to constitute a varying proportion of total mercury, depending on the lake that was being tested, but, overall, accounted for approximately 1-6% of the total mercury (Canada-Ontario Steering Committee, 1983). 5.1.3 Food Concentrations of mercury in most foodstuffs (WHO, 1976b; US EPA, 1984; Piotrowski & Inskip, 1981) are often below the reported limit of detection (usually 20 µg/kg fresh weight). Fish and fish products are the dominant source of methylmercury in food. The highest concen- trations are found in both freshwater and marine fish at the highest trophic levels (Table 4). For example, shark and swordfish have average values of total mercury in edible tissues above 1200 µg/kg, whereas anchovies and smelt have average values below 85 µg/kg. Most other foodstuffs have average values below 20 µg/kg, with mercury mainly in the inorganic form (Cappon, 1981; Gartrell et al., 1985a,b, 1986). Cappon (1987) reported mercury levels in vegetables. 5.2 General Population Exposure 5.2.1 Estimated daily intakes The human intake of the three major forms of mercury present in the environment is summarized in Table 4. The intake of mercury from the ambient atmosphere has been estimated by assuming that the concentration of total mercury is 2 ng/m3 and that 75% is present as elemental mercury vapour, 5% as inorganic mercury compounds, and 20% as methylmercury. The daily intake of each form of mercury was estimated by assuming a daily ventilation of 20 m3, and the amount absorbed was estimated by assuming that 80% of the inhaled elemental vapour, 50% of the inorganic mer- cury compounds, and 80% of the methylmercury was absorbed across the pulmonary membranes (WHO, 1976b). Mercury intake from drinking-water was estimated by assuming a daily water intake of 2 litres, an average concentration of 25 ng/litre, and that all the mercury is in the inorganic form. Methylmercury has been found in a few samples taken from bodies of natural water, but there have been no reports of methylmercury in drinking-water. The intake of species of mercury in the diet was the most difficult to estimate. Total mercury intake from all foodstuffs in Belgium was 13 µg/day, compared with an intake from fish alone of 2.9 µg/day (Fouassin & Fondu, 1978). Also in Belgium, Buchet et al. (1983) measured a daily intake from all foodstuffs of 6.5 µg mercury. The intake of total dietary mercury (µg/day) measured during a market basket survey (1984-1986) of the Food and Drug Administration (FDA) in the USA (Shibko, 1988), according to age group was: 0.31 (6-11 months); 0.90 (2 years); 1.76 (16 years, females); 1.84 (14-16 years, males); 2.32 (25-30 years, females); 3.01 (25-30 years, males); 2.29 (60-65 years, females) and 2.52 (60-65 years, males). It is of interest that when these intake rates are converted to µg/day per kg body weight, the values fall in a much more narrow range from 0.04 to 0.09. In fact values for all the age groups except the two-year-olds fall between 0.044 and 0.054 µg/day per kg. Table 4. Estimated average daily intake and retention (ug/day) of total mercury and mercury compounds in the general population not occupationally exposed to mercurya -------------------------------------------------------------------------- Exposure Elemental Inorganic mercury Methylmercury mercury vapour compounds -------------------------------------------------------------------------- Air 0.030 (0.024) 0.002 (0.001) 0.008 (0.0064) Food Fish 0 0.600 (0.042) 2.4 (2.3) Non-fish 0 3.6 (0.25) 0 Drinking-water 0 0.050 (0.0035) 0 Dental amalgams 3.8-21 (3 - 17) 0 0 Total 3.9-21 (3.1 - 17) 4.3 (0.3) 2.41 (2.31) -------------------------------------------------------------------------- a See text for assumptions underlying the calculations of average daily intake and retention. Values given are the estimated average daily intake; the figures in parentheses represent the estimated amount retained in the body of an adult. Values are quoted to 2 significant figures. In Poland, the average daily dietary intake of mercury (estimated in 2134 duplicate portions) was 5.08 µg/day in the age group l-6 years, 5.43 µg/day in the age group 6-18 years, and 15.8 µg/day in adults (Szprengier- Juszkiewicz, 1988). Owing to the low fish consumption (6.76 kg/year) and low mercury concentration in market fish (65 µg/kg), only 7% of the dietary intake derived from fish (Nabrzyski and Gajewska, 1984). Bernhard & Andreae (1984) estimated the world-wide mercury intake from seafood to be 2 µg/day, which is equivalent to a daily intake of 20 g seafood with a mercury concentration of 0.1 mg/kg. This agrees with estimates by a United Nations expert group (GESAMP, 1986). It should be pointed out that the individual variation in intake is large and that significant proportions of national populations have a mercury intake via seafood many times higher than the average (GESAMP, 1986). For the purpose of estimating the average daily intake of total mercury and various mercury compounds (Table 4), it was assumed that the daily intake of total mercury from fish and fish products is 3 µg and that 20% of this is in the form of inorganic mercury compounds (i.e., 0.6 µg/day) and 80% is methylmercury (i.e., 2.4 µg/day). The intake of total mercury from non-fish sources was calculated as the difference between the average total dietary intake and the intake from fish. The average total dietary intake in the Belgium studies was (6.5 + 13)/2 = 9.75 µg/day, whereas the corresponding value for a 70-kg adult in the USA can be estimated from the FDA market basket survey as 3.5 µg. Taking the average of the Belgian and USA figures, the dietary intake of total mercury is estimated as (9.75 + 3.5)/2 = 6.6 µg/day. By subtracting from this figure the intake of methylmercury from fish (2.4 µg/day), the estimated total dietary intake of inorganic mercury is 4.2 µg/day. All the mercury from non-fish sources was assumed to be in the inorganic form. The amounts absorbed across the gastrointestinal tract were estimated on the assumption that 7% of the inorganic and 95% of the methyl species were absorbed (section 6). The estimated dietary intake of inorganic mercury of 4.3 µg/day is the least reliable of the estimates in Table 4. Data are not available on the species of mercury in most foodstuffs. In addition, the figures for dietary intake of total mercury come from only two countries - Belgium and the USA. Table 4 portrays the relative magnitude of the contri- butions from various media. It is clear that fish and fish products are the dominant source of human exposure to methylmercury, even when low fish consumption is assumed (as in Table 4). Daily methylmercury intake can vary over a wide range, depending on the amount of fish consumed and the methylmercury concentration in the fish (Table 5). A number of communities have been identified where individu- al intakes exceeded 200 µg mercury/day (WHO, 1976b, 1980; Turner et al., 1980; GESAMP, 1986). As it is assumed that 80% of this mercury is methylmercury and that 95% of the methylmercury is absorbed, the absorbed amount of methylmercury (>153 µg/day) will, in these cases, domi- nate the daily mercury exposure (Table 4). On the basis of general population surveys of fish consumption, it was estimated that in Australia 0.9% of the population eat more than 1000 g fish/week and that this corresponded to about 20 µg mercury/day (WGMF, 1980). In the USA, surveys of fish consumption (US Dept. Commerce, 1978) were used to estimate that, with no regulatory control of the mercury content of marketed fish, 99.81% of all respondents had an upper limit mercury intake lower than their personal allowable daily intake (based on 30 µg mercury/day for a 70-kg person) at a 95% level. An action level of 1 mg mercury/kg in fish for regulatory control would increase this percentage to 99.87% and an action level of 0.5 mg mercury/kg would increase it to 99.89%. Dental mercury amalgams account for the major back- ground intake of mercury vapour (Clarkson et al., 1988). It is possible that mercury liberated from the amalgam can dissolve in the saliva as inorganic mercury, but there are no published reports on this possibility. A detailed dis- cussion of the release of mercury from dental amalgams will be found in the Environmental Health Criteria mono- graph on Inorganic Mercury, which is due to be published in 1990. Table 5. Intake of methylmercury (ug/day) from fish with various methylmercury levels and at various rates of fish consumptiona ------------------------------------------------ Consumption Level of methylmercury in fish of fish (ug/kg fresh weight)b (g/day) 200 500 1000 2000 5000 ------------------------------------------------ 5 1 2.5 5 10 25 20 4 10 20 40 100 100 20 50 100 200 500 300 60 150 300 600 1000 1000c 200 500 1000 2000 5000 ------------------------------------------------ a Adapted from: WHO (1980). b For methylmercury concentration in fish see Table 3. c Data from GESAMP (1986) indicate that maximum intakes may equal 1000 g/day. 6. KINETICS AND METABOLISM A considerable amount of information was available on the metabolism of methylmercury at the time when Environ- mental Health Criteria 1: Mercury was published (WHO, 1976b). This section will briefly review the information in that document and quote more recent data where appro- priate. 6.1 Absorption Methylmercury in the diet is almost completely ab- sorbed into the bloodstream (WHO, 1976b). Animals studies (Walsh, 1982) indicate that age, including neonatal stage, has no effect on the efficiency of gastrointestinal ab- sorption, which is usually in excess of 90% of the oral intake. Data on rats indicate rapid and virtually complete absorption of inhaled methylmercury vapour into the blood- stream (Fang, 1980). 6.2 Distribution Methylmercury is distributed in the bloodstream to all tissues. Distribution is completed within about 4 days in human beings (Kershaw et al., 1980), but the time after a single dose for maximum levels to be reached in the brain is one or two days longer than for other tissues (Berlin, 1986). At this time, the total brain contains approxi- mately 6% of the dose (Kershaw et al., 1980), which is very near to 10% of the body burden (WHO, 1976b). These blood and brain values correspond to a six times higher concentration in the brain than in blood (Berlin, 1986). There are significant species differences in brain-to- blood ratios. After the prolonged administration of methylmercury, brain-to-blood ratios are between 3 and 6 in squirrel monkeys (Berlin, 1986) but somewhat lower in macaque monkeys (Evans et al., 1977). The ratio is gener- ally low in non-primate animals, except in pigs (where it is 3.3); it is 1.5 in guinea-pigs, 1.2 in mice, and 0.06 in rats (Magos, 1987). Sex differences in distribution and retention have been reported in rats dosed with methylmer- cury (Magos et al., 1981; Thomas et al., 1986) and in both adult (Hirayama & Yasutake, 1986) and prenatally exposed mice (Inouye et al., 1986). There are also species differences in the distribution of methylmercury between erythrocytes and plasma. After the ingestion by human volunteers of fish containing methylmercury, the background-corrected erythrocyte-to- plasma methylmercury concentration ratio was about 20 (Kershaw et al., 1980). The ratio is approximately the same in monkeys and guinea-pigs, 7 in mice, and more than 100 in rats (Magos, 1987). The blood-to-hair ratio in humans is about 1 to 250, but appreciable individual differences have been found (Table 6). Similarly, large individual differences exist in the ratio of cord blood to maternal blood concen- tration. Cord blood usually has somewhat higher methylmer- cury concentration than maternal blood (WHO, 1976b). Thus, in a group of Japanese women the average ratio of cord blood to maternal blood methylmercury concentration ranged from 0.8 to 2.8, with a mean of 1.65 (Suzuki et al., 1984b). The results of studies on rats (Ohsawa et al., 1981) and pigs (Kelman et al., 1980, 1982) indicate that placental transport of methylmercury into the fetus increases dramatically towards the end of pregnancy. 6.3 Metabolic Transformation Methylmercury is converted to inorganic mercury, as- sumed to be Hg++, in mammals (WHO, 1976b). The fraction of total mercury present in the tissues as Hg++ depends on the duration of exposure to methylmercury and the time after cessation of exposure. The percentage of total mercury present as Hg++ in the tissues and body fluids of people exposed to high oral daily intakes of methylmercury for about 2 months in the Iraqi outbreak were: whole blood, 7%; plasma, 22%; breast milk, 39%; and urine, 73% (Amin-Zaki et al., 1976; Magos et al., 1976; WHO, 1976b). Measurements of liver tissues from fatalities in Iraq revealed that 16-40% was present as inorganic mercury. Unfortunately, no other tissues were available for analysis. There is a possibility that exposure to other mercury compounds may have occurred in some members of the Iraqi population. Table 6. Relationship between mercury concentrations in the blood and hair of people with long-term exposure to methylmercury from fish ----------------------------------------------------------------------------------------------- Country Number of Whole blood Hair Linear Reference subjects (x) (y) regression (µg/litre) (mg/kg) ----------------------------------------------------------------------------------------------- Canada 339 1 - 60 1 - 150 y = 0.30x + 0.5 Phelps et al. (1980) Japan 45 2 - 800 20 - 325 y = 0.25x + 0 WHO (1976b) Netherlands 47 1 - 40 0 - 13 y = 0.26x + 0 Den Tonkelaar et al. (1974) Sweden 12 4 - 650 1 - 180 y = 0.28x - 1.3 WHO (1976b) 51 4 - 110 1 - 30 y = 0.23x + 0.6 WHO (1976b) 50 5 - 270 1 - 56 y = 0.14x + 1.5 WHO (1976b) 60 44 - 550 1 - 142 y = 0.23x - 3.6 WHO (1976b) United Kingdom 173 0.4 - 26 0.1 - 11 y = 0.25x + 0.6 Haxton et al. (1979) 98 1.1 - 42 0.2 - 21 y = 0.37x + 0.7 Sherlock et al. (1982) Yugoslavia 38 1.2 - 9.6 0.4 - 3.0 y = 0.34x - 22 Horvat et al. (1986b) ----------------------------------------------------------------------------------------------- In Canadian Indians repeatedly exposed to methylmer- cury in fish during the summer season every year, inor- ganic mercury accounted for about 5% of total mercury in whole blood and about 20% in samples of head hair (Phelps et al., 1980). Brain mercury levels were measured in one Indian who had died of natural causes 2 years after having a high blood level (approximately 600 µg/litre). Most of the mercury in the brain tissue was in the inorganic form, but, at the time of his death, the total mercury in the brain had fallen to near background levels (Wheatley et al., 1979). Following the outbreak in Minamata, Japan, in 1956, tissues from a number of early fatalities were analysed (Tsubaki & Takahashi, 1986). Death occurred between 19 and 100 days after the onset of symptoms. Tissues were also analysed from people who died from 1 to 17 years after the onset of symptoms. Samples were analysed initially by the dithizone colorimetric procedure in 1956-1960 and again in 1973-1983 by atomic absorption for total mercury and by gas chromatography for methylmercury. In this study, atomic absorption generally gave higher values for total mercury than the dithizone method. The methylmercury concentration was always less than that of total mercury, usually less than 50%, and in a few cases less than 10%. The chemical nature of the mercury not accounted for as methylmercury was not determined. It may, in whole or in part, have been methylmercury that could not be extracted in the gas chromatographic procedure, or it may have been inorganic mercury. Speciation of mercury in human brain has been studied by Friberg et al. (1986) and Nylander et al. (1987). An average of 80% of the mercury in the occipital lobe cortex of autopsy cases in Sweden was found to be inorganic mercury (3-22 ng/g wet weight). Exposure to mercury from dental fillings could explain the high proportion of inorganic mercury in some cases but not in all. There is considerable evidence indicating the presence of inorganic mercury in the tissues of animals dosed with methylmercury (WHO, 1976b). Magos & Butler (1972) showed that during long-term daily dosing, the fraction of inor- ganic mercury in rat tissues tended to approach a constant value, which was different for each tissue. The kidney and liver had the highest fractions, while the brain had one of the lowest. Speciation of mercury in the brain of monkeys exposed to methylmercury for several years was studied by Lind et al. (1988b). At the end of the exposure period, 10-30% of the brain mercury was in the inorganic form while in monkeys sacrificed 0.5-2 years after the same treatment, about 90% was in the inorganic form. Simi- lar observation was reported by Kawasaki et al. (1986), but in the cerebrum a substantially higher proportion of the total mercury was methylmercury than in the cerebellum (See also WHO, 1976b). It is clear that the proportion of inorganic mercury found at any time in a particular tissue will be determined by a number of processes, e.g., the relative rates of uptake and loss of inorganic mercury and methylmercury and the extent of biotransformation (if any) in that tissue. Studies by Suda & Takahashi (1986) indi- cate that macrophage cells, such as those present in the spleen, are capable of converting methylmercury to inor- ganic mercury. The reaction may involve the production of oxygen free-radicals. At present there are no definitive data that prove that demethylation actually takes place in brain tissue, but persuasive arguments have been presented by Lind et al. (1988b). The conversion of methylmercury to Hg++ may be a key step in the processes of excretion. The faecal pathway accounts for about 90% of the total elimination of mercury in man and other mammals after exposure to methylmercury (WHO, 1976b). Virtually all the mercury in human faeces is in the inorganic form (Turner et al., 1975). The pro- cess of faecal elimination begins with the biliary secretion of both methylmercury and Hg++, complexed mainly, if not entirely, with glutathione (GHS) (Refsvik & Norseth, 1975) or other sulfhydryl peptides (Norseth & Clarkson, 1971; Ohsawa & Magos, 1974). Inorganic mercury is poorly absorbed across the intestinal wall (WHO, 1976b) so that most (approximately 90%) of the inorganic mercury secreted in bile passes directly into the faeces. Methyl- mercury secreted into the intestinal contents is in large part reabsorbed into the bloodstream and may subsequently contribute to biliary secretion, thereby forming a secretion-reabsorption cycle (Norseth & Clarkson, 1971). This cycle (also called enterohepatic circulation) in- creases the amount of methylmercury passing through the intestinal contents and thus provides a continuous supply of methylmercury to serve as a substrate for the intesti- nal microflora. These microorganisms are capable of con- verting methylmercury to inorganic mercury, which then becomes the major contributor to total faecal elimination in the rat (Rowland et al., 1980). Presumably about 10% of the inorganic mercury produced by the intestinal micro- flora is absorbed into the bloodstream and contributes to the inorganic mercury concentrations in tissues, plasma, bile, breast milk, and urine. This intestinal microbio- logical activity may explain the influence of diet on methylmercury elimination rates in rats (Rowland et al., 1984, 1986), and the absence of demethylating intestinal microflora may be the reason for the low rate of faecal elimination of mercury in suckling mice (Rowland et al, 1983). To what extent this model of enterohepatic circulation and intestinal conversion to inorganic mercury applies to humans is not yet known. Considerable species differences exist in rates of biliary excretion (Naganuma & Imura, 1984). Though the species variation in the secretion of methylmercury does not entirely correspond to the biliary excretion of GSH, high GSH secretion (rat, mice, and ham- ster) is associated (on a group basis) with high methyl- mercury secretion and low GSH secretion (guinea-pig and rabbit) with low methylmercury secretion (Stein et al., 1988). Animal studies suggest that multigeneration exposure to methylmercury may change tissue distribution and metab- olism (Yamamoto et al., 1986). 6.4 Elimination and Excretion The rate of excretion of mercury in both humans and laboratory animals dosed with methylmercury is directly proportional to the simultaneous body burden, and there- fore may be described by a single biological half-time (WHO, 1976b). The reason is that methylmercury is so mobile in the body that the excretion process is the rate- limiting step. Data on biological half-times in human beings were summarized in Environmental Health Criteria 1: Mercury (WHO 1976b). Kershaw et al. (1980) and Sherlock et al. (1984) reported half-times of 52 (39-67) and 50 (42-70) days in blood, close to the valves found earlier in people who ate fish or had consumed contaminated bread (WHO, 1976b). The whole-body half-times, determined in volunteers given a single tracer dose, have an average value of about 70 days and a range of 52-93 days. Only 20 subjects have been studied to date. Biological half-times in blood and hair have been measured both in volunteers given carefully measured doses and, after cessation of exposure, in individuals exposed as a result of accidental intake or high fish consumption. Observations on volun- teers reveal values for blood half-time close to 50 days and a range of 39-70 days. Results from single tracer doses agree well with those from volunteers given measured doses in fish. It is clear that the blood half-times over- lap those for the whole body, but the average value is lower. A shorter blood half-time would account for the observation that the amount of methylmercury in blood constitutes a decreasing fraction of the body burden with time after a single tracer dose (Miettinen, 1973). Lac- tating women have significantly shorter half-times (aver- age value, 42 days) than non-lactating women (average value, 79 days), an observation confirmed by animals studies (Greenwood et al., 1978). Observations on both volunteers and environmentally exposed people indicate that half-times in hair closely follow those in blood (Amin-Zaki et al., 1976; Kershaw et al., 1980; Hislop et al., 1983). However, hair half-times tend to have a wider range; for example, Al-Shahristani & Shihab (1974) reported a bimodal distribution in 48 Iraqi subjects, 90% having a half-time of 35-100 days and the other 10% a half-time of 110-120 days. It is possible, but not proven, that analytical artifacts may contribute to the wider range seen with hair (WHO, 1980). In any case, data from animals point to the importance of sex, age, and genetically determined individual differences (Hirayama & Yasutake, 1986). Animal data indicate major ontogenic effects on bio- logical half-times (Doherty et al., 1977). Suckling mice are completely unable to excrete methylmercury. At the end of the suckling period, excretion abruptly switches on at the adult rate. Observations on infant monkeys confirm this finding (Lok, 1983). Likewise, biliary secretion of methylmercury in suckling animals is virtually absent and assumes the adult rate after weaning. It is of interest that biliary secretion of glutathione (GHS) shows parallel ontogenic changes (Ballatori & Clarkson, 1985). Microflora also have greatly diminished capacity to demethylate methylmercury during the suckling period (Rowland et al., 1983). In view of the failure of infant animals to excrete methylmercury, human infants may also have a diminished excretion. Unfortunately, no direct observations have yet been reported. 6.5 Retention and Turnover The evidence summarized in section 6.4 indicates that the accumulation and excretion of methylmercury in humans, measured in terms of hair or blood levels, can be rep- resented by a single-compartment model. The accumulation phase in the whole body or in a tissue compartment is described by the equation: A = ( a / b )(1-exp(- b x t )) (equation 1) where A = the accumulated amount a = the amount taken up by the body (or organ) daily b = the elimination constant t = time The elimination constant is related to the biological half-time ( T´) by the expression: T´ = ln2/ b (equation 2) and a is related to the daily dietary intake ( d ) by the expression: a = f x d (equation 3) where f is the fraction of the daily intake taken up by the body (or organ). At a steady state, the accumulated amount ( A ) is given by: A = a / b (equation 4) while the steady-state mercury concentration in blood (C) in µg/litre is related to the average daily dietary intake (in µg mercury) as follows: 0.95 x 0.05 x d C = f x d / b = ---------------- = 0.95 x d (equation 5) 0.01 days-1 x 5 litres assuming that 0.95 of the intake is absorbed, that 0.05 of the absorbed amount goes to the blood compartment, that the blood volume is 5 litres, and that the elimination constant is 0.01 days-1. Sherlock et al. (1984) tested the validity of equation 1 by measuring blood mercury concentrations during a 100- day period of methylmercury intake and a 100-day period after intake ceased in 20 volunteers who consumed measured daily amounts of methylmercury in fish. Close agreement was found between predicted and observed values. Equation 1 predicts that a steady state in which in- take equals excretion will be attained in about 5 half- times. Thus, in adult humans, the whole body would attain a steady state in about one year (5 x 70 days = 350 days). Thus, an important prediction of the single-compartment model is that constant dietary exposure to methylmercury for a period of several years should not result in any greater accumulation than after one year of exposure. Equation 4 predicts that the maximum amount accumu- lated in the whole body of adult humans will be 100 times the average daily intake. In fact, steady blood levels may also be calculated from equations 2, 3, and 4, using the kinetic parameters to the single-compartment model listed in Table 7. It is of interest to compare this predicted relation- ship with those observed in field studies on populations believed to have attained a steady state from long-term dietary exposure to methylmercury in fish (Table 7). The coefficients relating long-term dietary intake to steady- state blood concentration are all lower than the predicted value of 0.95 calculated in equation 5 above. The reasons for this discrepancy are not yet fully understood. The measurement of dietary intake in populations with uncon- trolled intakes is liable to considerable error (Turner et al., 1980). However, this would not explain the consist- ently lower values from field studies. It is more likely that these populations were not in true steady state, since intake is frequently seasonal in fish-eating popu- lations. The fact that close agreement is seen between single-dose tracer studies, single-dose methylmercury in- take from fish, extended controlled intake from fish, and longitudinal hair analysis of individuals with very high intakes lends support to the validity of the single-com- partment model and the values of the kinetic parameters listed in Table 8. Sherlock and Quinn (1988) presented a more detailed discussion of the differences between con- trolled and uncontrolled studies on the relationship between blood concentration and intake of methylmercury. Table 7. Relationship between steady-state blood concentrations and average daily intake of methylmercury in fish consumers and predicted relationships from experimental data --------------------------------------------------------------------------------------------- Number of Duration of Average mercury Steady-state blood Reference subjects exposure intake (µg/day per concentration 70 kg body weight) (ug mercury/litre) (x) (y) --------------------------------------------------------------------------------------------- Observed relationship 32 years 0 - 800 y = 0.7x + 1 WHO (1976b) 165 years 0 - 400 y = 0.3x + 5 WHO (1976b) 20 years 0 - 800 y = 0.8x + 1 WHO (1976b) 725 years 0 - 800 y = 0.5x + 4 WHO (1976b) 22 years 0 - 800 y = 0.5x + 10 WHO (1976b) Predicted relationship 15 1 dose tracer y = 1x WHO (1976b) 30 1 - 2 months 0 - 2340 y = 0.8x WHO (1976b) 5 1 dose 1400 y = 1x Kershaw et al. (1980) 20 100 days 0 - 230 y = 0.8x Sherlock et al. (1984) --------------------------------------------------------------------------------------------- It should be emphasized that this model refers to the "average" adult human with a body weight of 70 kg. The gastrointestinal absorption rate for methylmercury is high (about 95%) and is not known to vary with age, but the energy intake varies greatly with age and this tends to make children and teenagers more vulnerable to high in- takes of methylmercury. The one compartment model is a useful working model for comparing blood or hair levels to daily intakes of methylmercury. Clearly this model is only an approximation to the more complex kinetics of mercury distribution and metabolism. For example, determination of mercury in hair and blood will not produce information concerning small compartments in the body. Methylmercury is slowly trans- formed to inorganic mercury, a process which is known to follow multiphasic kinetics (Berlin, 1986). 6.6 Reference or Normal Levels in Indicator Media Reference values in non-exposed populations for con- centrations of total mercury in commonly used indicator media are given in Table 9. The mean concentration in whole blood is probably about 8 µg/litre, in hair about 2 µg/g, in urine 4 µg/litre, and in the placenta about 10 µg/g wet weight. Table 8. Principal kinetic parameters in the single-compartment model for methylmercury in adult human beings ---------------------------------------------------------------------------- Number Type of Dose Number Compartment Reference of subject (µg of ---------------------------- subjects mercury doses Whole body Blood /kg) ---------------------------- f T´ f T´ (days) (days) ---------------------------------------------------------------------------- 3 adult tracer 1 0.95 72 - - WHO (1976b) 15 adult tracer 0.94 76 0.07 50 WHO (1976b) 5 adult 20 1 - - 0.05a 52 Kershaw et al. (1980) 5 adult 3.3 100 - - 0.05a 53 Sherlock et al. (1984) 5 adult 1.5 100 - - 0.055a 51 Sherlock et al. (1984) 4 adult 100 100 - - 0.057a 48 Sherlock et al. (1984) 5 adult 0.6 100 - - 0.064a 46 Sherlock et al. (1984) ---------------------------------------------------------------------------- a Calculations were made from concentrations in blood. The value of f (fraction of dose which reached the compartment) was calculated assuming blood volume of 5 litres in a 70-kg adult. Table 9. Concentrations of mercury in indicator media in non-exposed populationsa ----------------------------------------------------------------------------------------- Country Indicator No. of Mercury concentrationb Reference media Subjects Mean Range ----------------------------------------------------------------------------------------- Belgium placenta 474 15 1.1 - 103 Roels et al. (1978)c whole blood 497 13d 0.1 - 47d Lauwerys et al. (1978) Italy whole blood 80 20 0 - 46 Pallotti et al. (1979) Japan maternal blood 11 6.6 2.0 - 16.4 Suzuki et al. (1984b) umbilical blood 7 8.9 3.1 - 20.6 Suzuki et al. (1984b) New Guinea hair 40 4.5 0.9 - 12.1 Suzuki et al. (1988) Norway urine 103 4d 0.6 - 24d Lie et al. (1982) Poland whole blood 270 11.3d 2.5 - 24d Szucki & Kurys (1982) hair 505 2.2 0.02-10 Szucki & Kurys (1982) Maternal hair (scalp) 141 1.9 0.02 - 41 Sikorski et al. (1986) (pubic) 141 1.0 ND-32 Sikorski et al. (1986) Neonatal hair 141 0.11 ND-0.62 Sikorski et al. (1986) Swedene hair 18 0.4 Ohlander et al. (1985) hair 41 0.53 Forhammer et al. (1984) United whole blood 88 8.8d,f 1.1 - 42d Sherlock et al. (1982) Kingdom urine 77 2.4g ND - 8g Taylor & Marks (1973) USA whole blood 210 8.1h 0 - 50h Gowdy et al. (1977) whole blood 25 3.4 0 - 7i Kuhnert et al. (1981) placenta 25 6.7j 0 - 13i Kuhnert et al. (1981) Yugoslavia hair 34 1.5 0.4 - 3.3 Horvat et al. (1988b) maternal blood 34 3.7d 1.2 - 9.6d Horvat et al. (1988b) umbilical blood 34 7.7d 1.2 - 21d Horvat et al. (1988b) placenta 34 13 2.8 - 37 Horvat et al. (1988b) Northern hair 312 3.1 0 - 9 Airey (1983) hemispherek Northern hair 4603 2.3 0 - 5 Airey (1983) hemisphere Southern hair 1449 1.7 0.8 - 2.5 Airey (1983) hemisphere ----------------------------------------------------------------------------------------- a No known occupational exposure; fish consumption usually less than one meal per week. b The units for mercury concentrations are: µg/kg for placenta, mg/kg for hair, and µg/kg for blood and urine, unless otherwise stated. ND = not detectable. c This reference contains data on levels published prior to 1976. d µg/litre e Pregnant women. f Values are for adults g µg/g creatinine h Values after exclusion of 9 samples >50µg/kg as outliers; without the exclusion the mean was 14.2 and range 0-298 µg/kg. i Range estimated as twice the standard deviation. j The placentae were perfused to remove blood before analysis. k North of 22° latitude. Additional data on levels in indicator media in dif- ferent populations are given in the following references: Belgium (Buchet et al., 1978), Canada (Galster, 1976; Kershaw et al., 1980; Phelps et al., 1980; McKeown-Eyssen & Ruedy, 1983a; McKeown-Eyssen et al., 1983; Valciukas et al., 1986), Federal Republic of Germany (Lommel et al., 1985), Finland (Mykkanen et al., 1986), Greenland (Hansen et al., 1984), Iceland (Johannesson et al.,1981), Italy (Capelli et al., 1986), the East Pacific area (Yamaguchi et al., 1977), Japan (Suzuki et al., 1984a), New Guinea (Kyle & Ghani, 1982a,b; 1983), New Zealand (Kjellstrom et al., 1982), Seychelles (Matthews, 1983), Spain (Gonzalez et al., 1985), and Sweden (Skerfving, 1974). Intake of methylmercury is reflected in elevated levels in whole blood and in erythrocytes (approximately 95% of blood mercury is in the erythrocytes). Animal studies indicate that, at non-toxic levels, blood methylmercury concen- tration is a good index of brain mercury concentration (Berlin, 1976). Urine and blood concentrations correlate with mercury vapour levels only after long-term exposures (Smith et al., 1970). Blood levels rise and fall sharply during and after short-term exposures (Cherian et al., 1978). Hair levels may be increased as a result of direct adsorption of mercury vapour onto the hair strands. Airey (1983) reported that the average hair mercury levels in the northern hemisphere are higher than in the southern hemisphere (Table 9). Long-term fish consumption determines almost com- pletely the concentrations of methylmercury and, usually, total mercury in blood. Thus, reference values must take into account fish consumption. In communities with high fish consumption rates, individuals with long-term intakes of 200 µg mercury/day will have blood levels in the range of 200 µg mercury/litre. Hair concentrations of methylmercury are proportional to blood concentrations at the time of formation of the hair strand (Table 6). In general, the concentration in hair is 250 times the simultaneous concentration in blood. Once mercury is incorporated into a hair strand, that hair mercury concentration remains unchanged. Thus, longitudi- nal measurement of mercury in hair provides a recapitu- lation of methylmercury levels in blood. Airey (1983) presented a comprehensive evaluation of mercury levels in hair. The author found that mean hair mercury concen- trations corresponded to fish consumption patterns as follows: once or less a month, 1.4 µg/g; once every 2 weeks, 1.9 µg/g; once a week, 2.5 µg/g; and once or more a day, 11.6 µg/g. Owing to their higher-than- average fish consumption, fisherman may have higher-than- average methylmercury concentrations in their hair. For example, in three Mediterranean countries, (Greece, Italy, and Yugoslavia) 33% of 212 fishermen but only 0.33% of 918 other residents had methylmercury levels in hair above 10 µg/g (WHO/FAO/UNEP, 1989). These data support the conclusion that long-term fish intake determines methyl- mercury levels in hair and also in blood. 6.7 Reaction with Body Components Information on the binding of methylmercury to tissue ligands other than haemoglobin is sparse. Methylmercury is believed to bind to cystinyl residues in the haemoglobin molecule. The number and position of these residues in the amino acid chains differ in haemoglobin from different species (Doi & Kobayashi, 1982, Doi & Tagawa, 1983). Methylmercury is complexed to glutathione (GHS) in both human and animal erythrocytes (Naganuma et al., 1980). The only known exception is rat erythrocytes where practi- cally all methylmercury is bound to haemoglobin (Doi & Tagawa, 1983). Methylmercury complexes may also be involved in the urinary excretion of methylmercury (Mulder & Kostyniak, 1985a,b). Animal data indicate that methyl- mercury complexes also exist in brain tissue (Thomas & Smith, 1979; Berlin et al., 1975), bile (Refsvik & Norseth, 1975), liver (Omata et al., 1978), and probably in kidney tissue (Richardson & Murphy, 1975). Complexes of methylmercury with GHS and possibly other low molecular- weight thiols play a role in blood transport and tissue distribution (Hirayama, 1980; Thomas & Smith, 1982) and in biliary secretion (Ballatori & Clarkson, 1985; Urano et al., 1988). Glutathione-S-transferase (ligandin) may be involved in the biliary secretion process (Refsvik, 1984a,b; Magos et al., 1985b), but evidence is still equivocal (Gregus & Varga, 1985). The activity of hepatic y -glutamyltransferase may also affect methylmercury secretion in bile (Stein et al., 1988). According to in vitro studies, the transport of methylmercury across cell membranes appears to be a diffusional process involving an unchanged complex of methylmercury chloride (Lakowicz & Anderson, 1980; Bienvenue et al., 1984). However, the relevance of these findings to in vivo transport across membranes is not clear. Due to the high affinity of the methylmercury cation for sulfhydryl groups, it is unlikely that methylmercury chloride will be present in significant amounts in plasma or other biological fluids. Amino acid complexes may be involved in membrane transport of methylmercury (Hirayama, 1980, 1985; Aschner & Clarkson, 1987; Watanabe et al., 1988). The administration of selenium compounds to animals protects against the toxic effects of methylmercury (Ganther et al., 1972; Iwata et al., 1973). It alters the tissue distribution and excretion of methylmercury (Ganther, 1978, Prohaska & Ganther, 1977) and also the inorganic-to-methyl mercury ratio in tissues (Komsta- Szumska & Miller, 1984; Brzeznicka & Chmielnicka, 1985a). In spite of the protective effect, selenite increases the brain concentration of methylmercury (Magos & Webb, 1977; Brzeznicka & Chmielnicka 1985b). The methylmercury cation has a high affinity for selenides and diselenides (Sugiura et al., 1978), the latter being formed by the reductive metabolism of selenite (Hsieh & Ganther, 1975; Ganther, 1979). It has been reported that (CH3Hg)2Se can be formed both in vitro and in vivo (Magos et al., 1979; Naganuma & Imura, 1980; Masukawa et al., 1982). To what extent the formation of this compound explains the altered tissue distribution of methylmercury is not yet clear. Selenium may also divert mercury from its endogenous binding sites, but Thomas & Smith (1984) were not able to find evidence for this. Maternal administration of selenium to mice causes a specific alteration in the form of selenium in fetal liver, as indicated by gel filtration chromatography. This change was not found in maternal liver or kidney or in the placenta (Nishikido et al., 1988b). 7. EFFECTS ON ORGANISMS IN THE ENVIRONMENT Data concerning the effects of methylmercury on organ- isms in the environment are discussed in Environmental Health Criteria 86: Mercury - Environmental Aspects (WHO, 1989a). 8. EFFECTS ON EXPERIMENTAL ANIMALS AND IN VITRO TEST SYSTEMS Methylmercury is a systemic poison and, depending on the dose and the length of exposure period, can affect various organ systems and functions. However, in every species the main target is the nervous system and one of the earliest objective clinical signs is ataxia. The sig- nificance of effects on animals is confounded by well- established species differences in both the localization of nervous system damage and in accompanying clinical and pathological changes (Berlin, 1986). Another common target is the fetus. As methylmercury is capable of corrosive action, it can damage any tissue (skin, eye, upper part of the digestive tract) if presented in sufficiently high concentrations (WHO, 1976b). 8.1 Neurotoxicity and Nephrotoxicity The effects of methylmercury given in a single lethal dose is uncharacteristic. After intraperitoneal adminis- tration, rats showed respiratory and vascular disorders and hamsters became comatose (Hoskins & Hupp, 1978), whereas after oral administration to pigs, central nervous system depression, ending in coma, was preceded by diar- rhoea and vomiting (Piper et al., 1971). Rats surviving an LD50 dose showed general debilitation with weight loss, but did not develop specific motorial changes, while a squirrel monkey that had survived the severe acute effects of a single dose (6.4 mg/kg) became uncoordinated by 22 days and blind at 24 days (Hoskins and Hupp, 1978). The cause of weight loss is anorexia. The anorexic effect shows significant species differences. Anorexia precedes the clinical signs of nervous system injury in rodents treated daily with methylmercury (Magos, 1982). In cats (Davies & Nielsen, 1977) and dogs (Davies et al., 1977), anorexia and gait disorder ("bunny-hopping") occur simultaneously. In the squirrel monkey only severe methylmercury intoxication is associated with weight loss (Evans et al., 1977). Studies on the neurotoxicity of methylmercury were reviewed by WHO (1976b). This review called attention to species differences in blindness and the involvement of peripheral nerves. Blindness may be caused in man, monkeys, and pigs, but not in cats. In rats, the initial damage appears in the dorsal root ganglions and associated peripheral nerves, while in monkeys the cerebral cortex is the first target. One of the most common lesions of the central nervous system is in the granular layer of the cerebellum. This type of damage has been observed in man (Takeuchi & Eto, 1975) and also in rats, cats, hens (Chang, 1977), dogs (Davies et al., 1977), calves (Herigstad et al., 1972), guinea-pigs (Falk et al., 1974), and rabbits (Jacobs et al., 1977), but not in pigs (Davies et al., 1976) or monkeys (Chang, 1977; Mottet et al., 1987). Female rats, which accumulate higher concentrations of methylmercury in their brain than males, also develop more severe cerebellar lesions (Magos et al., 1981). These experimental studies (see also Mitsumori et al., 1984 and Munro et al., 1980) confirmed clinical findings in human beings of irreversible damage to the nervous system. Other experimental studies carried out since the publication of Environmental Health Criteria 1: Mercury (WHO, 1976b), which focused on the mechanism of toxicity, are reported in section 126.96.36.199. Renal damage is one of the most frequently described non-neural effect of methylmercury. This damage may be caused by inorganic mercury split from methylmercury. In rats, treatment with methylmercury caused renal damage ranging from ultrastructural changes to the degeneration of the distal convoluted tubules (see WHO, 1976b). Male rats are more sensitive to the renotoxic effect of methyl- mercury than females (Munro et al., 1980). Renal damage was also observed in other experimental species. In most of the dogs that showed clinical and histological signs of methylmercury-induced neurotoxicity, there were also signs of renal necrosis, desquamation, and regeneration (Davies et al., 1977). In the kidneys of methylmercury-intoxicated guinea-pigs, only swelled epithelial cells in the proximal tubules were reported (Falk et al., 1974). In cats (Davies & Nielsen, 1977) and pigs (Davies et al., 1976), only hyalin and cellular casts were seen. Though treatment of six monkeys ( Macaca mulatta ) with daily doses of methyl- mercury (80-120 µg mercury/kg in apple juice) for 3.5-12 months did not affect the general health status adversely, it caused ultrastructural changes in the kidneys (Chen et al., 1983). These changes included intracytoplasmic vacu- oles and electron-dense inclusion bodies. In the same studies, degenerative changes in the Paneth cells of intestines were also observed. These changes were most pronounced in animals killed immediately after exposure (see also Mottet et al., 1987). 8.2 Reproduction, Embryotoxicity, and Teratogenicity Methylmercury added in vitro to a suspension of sperm from untreated monkeys ( Macaca fasicularis ) at 9-15 µg/ml decreased sperm motility but did not decrease oxygen con- sumption (Mohamed et al. 1986a,b). In fact, oxygen con- sumption was increased at the 15 µg/ml concentration when sperm motility was almost zero. Further studies with specific inhibitors revealed that mitochondrial energy production was not affected by mercury. The authors suggested that the primary effect was on the dynein/micro- tubule sliding assembly. Lee & Dixon (1975) reported damage to spermatogenesis in mice given a methylmercury dose of 1 mg mercury/kg, much lower doses giving rise to neurological effects. No special susceptibility to sterility, resulting from pre- natal exposure, could be detected in mice (Gates et al., 1986). When female mice were given a single intraperitoneal injection of methylmercury chloride (2.5, 5, or 7.5 mg/kg body weight) prior to mating, dose-related increases in pre- and early post-implantation fetal losses were recorded (Verschaeve & Leonard, 1984). This observation could have a genetic cause or result from physiological effects on the mother. Gunderson et al. (1986) treated 11 monkeys ( Macaca fasicularis ) with daily oral doses of methylmercury in apple juice (50-70 µg/kg per day) before and during preg- nancy. The mean blood levels during pregnancy, measured in each trimester and at delivery, were within the range of 1080-1330 µg/litre, with maximum values within the range of 1510-1840 µg/litre. The mean blood levels of the offspring at birth were 1690 µg/litre (range, 880- 2460 µg/litre). When tested 190 days post-conception, the mean blood levels had fallen to 1040 µg/litre. The ex- posed animals showed recognition deficits (compared with 10 untreated controls) when administered an adaptation of a standardized test of visual recognition memory. The same blood mercury concentration (600-2000 µg/litre) in preg- nant squirrel monkeys exposed to methylmercury resulted in a 22.5% (mean of six results) reduction in the cerebral weight of fetuses (Logdberg et al, 1988). Three months treatment with daily oral doses of methylmercuric hydrox- ide (50 or 90 µg/kg) increased the frequency of repro- ductive failure (i.e., non-conception, abortion) in non- human primates and decreased the birth weight of their offspring (Burbacher et al., 1984). Offspring from treated animals directed significantly less attention to novel stimuli than did controls. At doses which are not toxic to the rat dam, prenatal exposure produced hydrocephalus, decreased thickness of the cerebral cortex in the parietal section, increased thickness of the hippocampus in the occipital region (Kutscher et al., l985), and delayed ossification (Chmielnicka et al., 1985). A variety of structural changes, detectable at both the light and electron micro- scopic levels, were also observed by Reuhl et al., (1981a,b). Similar effects have been noted in prenatally exposed mice where the development of communicating hydrocephalus was associated causally with aqueductal stenosis (Choi et al., 1988). Prenatal exposure at doses not affecting the mother is known to produce abnormal behaviour in the offspring of several animal species (Spyker et al., 1972; Bornhausen et al., 1980; Zimmer et al., 1980; Shimai & Satoh, 1985; Elsner et al., 1988). The behavioural effects may be the consequence of an effect on neurotransmitters in the brain. Thus a single dose of 5.0 mg mercury/kg, given as methylmercury on postnatal day 2, resulted in increased serotonin concentration and movement and postural dis- orders by day 22-24 (O'Kusky et al., 1988). In addition, Bartolome et al. (1982) showed both acute and long-lasting effects on the maturation of central catecholamine neuro- transmitter systems following early postnatal exposure. Eccles & Annau (1982a,b) demonstrated altered behavioural sensitivity to amphetamine in adult offspring, and Cuomo et al. (1984) showed alterations in response to apomorph- ine. Prenatal exposure of rodents can produce a variety of effects on non-nervous tissues. It is well known that the administration of large doses of methylmercury to pregnant rodents produces cleft palate (e.g., Lee et al., 1979; Harper et al., 1981). Prenatal exposure of rats can pro- duce renal functional abnormalities detectable in off- spring at 42 days of age (Smith et al., 1983; Slotkin et al. 1986). 8.3 Mutagenicity and Related End-Points Methylmercury is capable of causing chromosome damage in cell cultures (Morimoto et al., 1982; Curle et al., 1983), in the golden hamster (Watanabe et al., 1982; Gilbert et al., l983), and in ovulating Syrian hamsters (Mailhes, 1983). It can induce histone protein pertur- bations (Gruenwedel & Diaham, 1982) and influence factors regulating the nucleolus-organizing activity (Verschaeve et al., 1983). The mutagenic response of V79 Chinese ham- ster cells to methylnitrosourea is enhanced by methylmer- cury (Onfelt & Jenssen, 1982). Methylmercury has been reported to interfere with gene expression in in vitro cultures of glioma cells at low concentrations (0.05- 0.1 µmol/litre) (Ramanujam & Prasad, 1979). The induc- tion of non-disjunction and sex-linked recessive lethal mutations was found in Drosophila melanogaster treated with methylmercury. Tolerance to methylmercury was corre- lated with the uptake of mercury and not with the rate of excretion (Magnusson & Ramel, 1986). 8.4 Carcinogenicity Methylmercury has been reported to produce renal car- cinomas in mice given diets containing methylmercury chloride (15 mg/kg) for about one year (Mitsumori et al., 1981). Animals given 30 mg/kg died from neurotoxicity after 6 months. Nixon et al. (1979) found that prenatal exposure to methylmercury increased the incidence in rats of neural tumours caused by sodium nitrite and ethylurea. The mothers had been exposed since weaning to 10 mg methylmercury/kg diet. These are the only reports of potential carcinogenicity. Blakley (1984) administered methylmercury chloride to Swiss mice (0.2, 0.5, or 2 mg/litre drinking-water) for 15 weeks and gave them a single dose (1.5 mg/kg) of urethane by intraperitoneal injection at the end of the third week. Methylmercury produced a dose-related increase in the size of lung aden- omas induced by urethane. However, only the highest dose (2 mg/kg) increased the incidence of adenomas. No effects on weight gain or water consumption related to methylmer- cury treatment were seen. 8.5 Special Studies Kato et al. (1981) reported that the detection of electro-oculographic changes in monkeys is one of the most sensitive indices of methylmercury effects. Using high doses of methylmercury, Wassick & Yonovitz (1985) demon- strated auditory deficits over a 4- to 78-kHz frequency range in mice. Methylmercury is known to affect auditory performance in human beings (section 9.1). Methylmercury has been found to depress both primary and secondary immune responses in rodents (Koller & Roan, 1980; Koller et al., 1980). Other effects reported in ani- mal studies include impairment of adrenal and testicular function in rats (Burton & Meikle, 1980), impairment of thyroid function in mice (Kawada et al., 1980), and im- pairment of sleep-walking rhythms in rats (Arito et al., 1983). Effects of methylmercury on glucose transport, glucose metabolism, and blood flow in the central nervous system were described by Hargreaves et al. (1986). Long-term treatment of Brown-Norway rats produces proteinuria (Bernaudin et al., 1981). This strain is genetically susceptible to immune-mediated renal damage produced by inorganic mercury. 8.6 Factors Modifying Toxicity; Toxicity of Metabolites Several substances have been found to affect the chronic toxicity of methylmercury. Of these, selenium, usually administered to animals as sodium selenite, has been the most widely studied, following the first report by Ganther et al. (1972). In general, selenium has a protective action in that it delays the onset of methylmercury toxicity or reduces the severity of its effects (Chang & Suber, 1982). Simultaneous administration of sodium selenite to mice during pregnancy was found to protect the offspring from effects on developmental reflexes caused by a 6 mg/kg dose of methylmercury given on day 9 (Satoh et al., 1985). In mice given selenium (11.6 nmol/ml in the drinking-water) before and after gestation, the incidence of methylmercury-induced resorp- tion was increased. The incidence of cleft palate in mice was not affected by excess selenium (Nobunaga et al., 1979) nor by selenium deficiency (Nishikido et al., 1988a), but selenium deficiency enhanced the fetal tox- icity of methylmercury (Nishikido et al., 1987). Another antioxidant, vitamin E, has also been found to be protective against methylmercury toxicity in both normal and selenium-deficient rats (Chang et al., 1978; Welsh, 1979). The antioxidant N,N'-diphenyl-p-phenyly- lenediamine, however, was more protective than vitamin E. The latter also protected against the in vitro damage caused by methylmercury to chromosomes in human blood cells (Morimoto et al., 1982), hamster fibroblast cells (Gilbert et al., 1983), and glioma cells (Prasad & Ramanujam, 1980). Phospholipids have been reported to protect against the in vitro action of methylmercury on rat liver enzymes (Magnaval & Batti, 1980). The signifi- cance of these findings to human health is not clear, as high doses were used and methylmercury poisoning in rodents may not be the same as in human beings. Moreover selenite only delays the onset of methylmercury intoxi- cation (Chang & Suber, 1982) and is not the form of selenium found in human diets. Though selenium in marine food may have the same protective action as selenite (Newberne et al., 1972; Ohi et al., 1980; Thrower & Andrewartha, 1981), the bioavailability of biological selenium for reaction with mercury is less than that of selenite (Magos et al., 1984) and its potential, free from other dietary effects, to delay the onset of methylmercury intoxication has not been proved. Ethanol has been found to potentiate the effects of methylmercury in rats (Turner et al., 1981). This single finding should be investigated further, preferentially in primates, as it could have considerable public health sig- nificance. Yamini & Sleight (1984) found that guinea-pigs on a diet deficient in vitamin C suffered more severe neuro- logical damage when exposed to 44 mg methylmercury/kg in their diet for 20 days than controls fed a diet with adequate vitamin C. It has been postulated that methylmercury might pro- duce its effects via cleavage of the mercury-carbon bond (Ganther, 1978). This could produce free radicals, causing lipid peroxidation. This might explain the protective action of vitamin E and selenium. Inorganic mercury released from methylmercury might be the actual toxic species. However ethylmercury, which decomposes faster and raises inorganic mercury concentration in the brain to a higher level than does methylmercury, produces less brain damage at equal doses (Magos et al., 1985a). 9. EFFECTS ON MAN 9.1 General Population Exposure The effects of methylmercury on the adult differ both quantitatively and qualitatively from effects seen after prenatal and, possibly, postnatal exposure. Thus, effects on adults will be treated separately from effects on developing tissues. 9.1.1 Effects on adults The effects of methylmercury on adult human beings have been thoroughly described in Environmental Health Criteria 1: Mercury (WHO, 1976b). They will be summarized here with relevant new material added as appropriate. 188.8.131.52 Effects on the nervous system The nervous system is the principal target tissue for the effects of methylmercury on adult human beings. The sensory, visual, and auditory functions, together with those of the brain areas, especially the cerebellum, con- cerned with coordination, are the most common functions to be affected. The earliest effects are non-specific symp- toms, such as complaints of paraesthesia, malaise, and blurred vision. Subsequently, signs appear such as concen- tric constriction of the visual field, deafness, dysar- thria, and ataxia. In the worst cases, the patient may go into a coma and ultimately die. In less severe cases, some degree of recovery in each symptom occurs; this is believed to be a functional recovery that depends on the compensatory function of the central nervous system. The subjective complaint of paraesthesia was found to be a permanent symptom in patients exposed in the Japanese outbreak, whereas in the Iraqi outbreak, paraesthesia was transient in many cases. The reason for this difference is not known. At high doses, methylmercury affects the peripheral nervous system (Rustam et al., 1975). In Iraq, patients had symptoms of neuromuscular weakness that could be ameliorated by treatment with acetylcholinesterase inhibi- tors. Methylmercury poisoning has several important fea- tures: - a long latent period usually lasting several months; - damage almost exclusively limited to the nervous sys- tem, especially the central nervous system; - areas of damage to the brain are highly localized (focal), e.g., in the visual cortex and the granular layer of the cerebellum, especially in the infolded regions (sulci); - effects in severe cases are irreversible due to de- struction of neuronal cells; - the earliest effects are non-specific subjective com- plaints, such as paraesthesia, blurred vision, and malaise. 184.108.40.206 Effects on non-nervous tissue The only effect on human beings not involving the ner- vous system is the claim that chromosome damage is associ- ated with long-term exposure to methylmercury (Wulf et al., 1986). No further reports have appeared on this sub- ject since the review in Environmental Health Criteria 1: Mercury (WHO, 1976b). 9.1.2 Effects on developing tissues 220.127.116.11 Effects on the nervous system Observations on both human subjects and animals indi- cate that the developing central nervous system is more sensitive to damage from methylmercury than the adult ner- vous system. The first indications arose from the outbreak of methylmercury poisoning in Minamata, Japan, in the 1950s, when it was found that mothers who were slightly poisoned gave birth to infants with severe cerebral palsy. Subsequent studies on experimental animals confirmed the increased sensitivity of the fetus. The Iraqi outbreak in 1971-1972 resulted in cases of severe damage to the cen- tral nervous system in infants prenatally exposed. More recent follow-up studies in Iraq indicated a milder syn- drome at lower dose levels (Marsh et al., 1980). In fact, it has been possible to demonstrate a relationship between the maximum hair level in the mothers during pregnancy and the frequency of abnormalities in their infants (Marsh et al., 1981). The clinical picture is dose dependent. In those in- fants who have been exposed to high maternal blood levels of methylmercury, the picture is of cerebral palsy indis- tinguishable from that caused by other factors. Microceph- aly, hyperreflexia, and gross motor and mental impairment, sometimes associated with blindness or deafness, is the main pattern (for review, see WHO, 1976b). Milder degrees of the affliction are not easy to diagnose during the first few months of life, but they later become clear. Patients show mainly psychomotor impairment and persist- ence of pathological reflexes (Marsh et al., 1977, 1980, 1981; McKeown-Eyssen et al., 1983). Milder cases have findings quite similar to the findings in the minimal brain damage syndrome. Post-mortem observations in Japan indicated that damage is generalized throughout the brain in the case of prenatal exposure, in contrast to adult exposure where focal lesions are predominant. The Japanese cases of pre- natal poisoning indicated disturbed development in the cytoarchitecture of the brain and the brain size was dim- inished in severe cases. Similar pathological findings were reported on autopsies of two prenatally exposed Iraqi infants (Choi et al., 1978). The pathological findings in these studies were attributed to incomplete and abnormal migration of neuronal cells to the cerebellar and cerebral cortices (section 9.3.2). 9.2 Occupational Exposure This type of exposure was reviewed in Environmental Health Criteria 1: Mercury (WHO, 1976b). No new infor- mation has become available. In fact, occupational ex- posure results in effects similar to those reviewed in section 9.1 (e.g., Hunter & Russell, 1954). 9.3 Mechanisms of Toxicity Section 9 is concerned with effects on human beings. However, in discussing mechanisms of methylmercury tox- icity, animal and other experimental data are used where they throw light on how damage is inflicted on the human organism. 9.3.1 The mature organism 18.104.22.168 Mechanism of selective damage The mechanism of selective damage is not well under- stood. In a review of publications, Syversen (1982) claimed that the selective effects relate to the ability of certain cells in the central nervous system to repair the damage initially inflicted by methylmercury. Thus, those cells capable of repairing injury survive, whereas those cells that lack the facility to repair the damage are the ones that are destroyed. For example, the small granule cells in the cerebellum lack the repair capacity and are the first cells in that area of the brain to succumb. Jacobs et al. (1977) noted that the small neurons in the central nervous system appear to be especially vul- nerable. Such cells have very little cytoplasm and only limited protein synthetic machinery to carry out repair. On the other hand, Berlin (1986) has proposed that selec- tive damage results from the inter-neuronal axonal trans- port of methylmercury. Thus, the sensory centres, e.g., in the visual cortex, are affected because axonal transport is in the afferent direction leading to a local accumu- lation of methylmercury. The motor systems are relatively unaffected by methylmercury because axonal transport is in the efferent direction leading to removal of mercury from the motor areas. 22.214.171.124 The latent period The reason for the long latent period is not under- stood. The mean latent period ranged from 16 to 38 days in Iraq, and, in many cases, initial symptoms appeared after the cessation of intake of contaminated bread. In the Japanese outbreaks, it was difficult to determine the exact latent period because in many cases the starting point of intake was unclear. However, in some cases in Japan, a very long latent period (up to several years) was reported (WHO, 1980). Included in these cases were some patients who showed only slight signs and symptoms but who later developed the clinical features of severe poisoning after they had stopped eating the polluted fish. In the same period of time, the other patients showed relative improvement or progression. A latent period of several years may be partially explained by psychogenic overlay, which modifies the symptoms, or by sub-clinical lesions, which may be revealed by the aging factor. However, slow accumulation in the brain of methylmercury (or of inor- ganic mercury split off from the methylmercury) cannot be the explanation. 126.96.36.199 Cellular and molecular mechanisms Since the publication of Environmental Health Criteria 1: Mercury (WHO, 1976b), numerous studies on the mechanism of action at the cellular and molecular levels have been reported (Clarkson, 1983; Berlin, 1986). Inhibition of protein synthesis in target nerve cells is a well docu- mented effect in animals that appears before the first clinical signs of intoxication (Yoshino et al., 1966; Carmichael et al., 1975; Omata et al., 1980, 1982; Syversen, 1982; Fair et al., 1987). It occurs also in vitro before other cellular functions are affected (Nakada et al., 1980; Sarafian et al., 1984). Verity and his colleagues (Cheung & Verity, 1985; Sarafian & Verity, 1985, 1986) have identified the step in protein synthesis that is most sensitive to methylmercury. The peptide-elongation process can be affected at high levels of methylmercury, but the first stage of synthesis, associated with transfer RNA, may be the most sensitive. It appears that the inhibition of protein synthesis is general; there is no selective inhibition of formation of any special proteins or group of proteins. The reason for the special sensitivity of protein synthesis to methylmercury is not known. Jacobs et al. (1977) noted that the mammalian ribosome contains 120 sulfhydryl groups, of which about half are exposed and reactive during peptide formation, and that the chain- initiation factor is strongly inhibited by sulfhydryl reagents, at least in the case of bacteria. They suggested that the sulfhydryl groups of active ribosomes are more vulnerable than those in other proteins, where disulfide bridge formation may predominate. Methylmercury also interferes with lipids (Nakada & Imura, 1983; Ando et al., 1985), myelin (Ganser & Kirschner, 1985), mitochondrial DNA synthesis (Miller et al., 1985), and glutathione per- oxidase (Hirota et al., 1980). Effects on neurotransmitters and receptors (Kobayashi et al., 1979, 1981; Concas et al., 1983), lipids (Macfarlane, 1981; Rebel et al., 1983; Leblanc et al., 1984), adenyl cyclase activity (Spuhler & Prasad, 1980), membrane structure (Kasuya, 1980), and on the integrity of microtubules have been reported in a variety of experimen- tal systems (Araki et al., 1981; Nakada & Imura, 1982; Sager et al., 1981a,b). Methylmercury inhibits amino acid transport in rat brain microvessels at concentrations similar to those known to cause toxicity in humans (Tayarani et al., 1988). Changes in glucose transport across the blood-brain barrier in rats have been observed in the latent period before overt signs of methylmercury intoxication appear (Hargreaves et al., 1986). When given systemically, methylmercury accelerates axonal transport of proteins in the optic nerve (Aschner, 1986). However, when it is given by intra-ocular injection, methylmercury inhibits protein transport along the optic nerve (Aschner et al., 1986). The relevance of the above findings to the pathogenesis of methylmercury poisoning is still a matter of speculation. Perhaps more firmly established is the connection be- tween muscular weakness seen in severe cases of poisoning in the Iraqi outbreak and the inhibition of acetylcholine transmission at the neuromuscular junction (Von Burg & Landry, 1976; Shamoo et al., 1976; Atchison & Narahashi, 1982; Quandt et al., 1982; Atchison, 1986). The stimu- latory action of methylmercury on the miniature endplate potentials of the neuromuscular junction appears to result from the loss of calcium ions from the nerve terminal mitochondria (Levesque & Atchison, 1988). 9.3.2 Developing tissues Post-mortem observations derived from the Japanese and Iraqi outbreaks suggested that the severe prenatal effects seen resulted from incomplete and abnormal migration of neuronal cells to the cerebellar and cerebral cortices. In support of their autopsy findings indicating abnormal neuronal migration, Choi et al. (1979) noted that methyl- mercury inhibited the in vitro migration and movement of cultured human cells. Changes in astrocyte membranes and in motility were also observed in cultures when methylmer- cury was added (Choi & Lapham, 1980). The ability of methylmercury to damage astrocytes in vitro was confirmed by findings of decreased DNA synthesis (Choi et al., 1980; Choi & Kim, 1984). The toxic effect on astrocytes may be relevant to the pathological picture, since these cells are believed to play a role in supporting normal neuronal migration in the developing brain (Choi & Lapham, 1976). More recently, Peckham & Choi (1986) showed that methyl- mercury alters the anionic surface charge on cultured fetal mouse astrocytes. Cell-to-cell recognition processes were found to be affected in aggregates of mouse cerebel- lar cells (Jacobs et al., 1986). The authors suggested that the mechanism involved depressed synthesis of specific proteins followed by alterations in microtu- bules. A second general mechanism by which brain development could be impaired is the inhibition of cell division (Chen et al., 1979; Sager et al., 1982, 1983; Rodier et al., 1984; Slotkin et al., 1985; Howard & Mottet, 1986; Vogel et al., 1986). Inhibition of cell proliferation probably explains the production of cleft palates in rats (Lee et al., 1979; Olson & Massaro, 1980), although this effect has not been seen in humans. Methylmercury is known to inhibit cell division by causing metaphase arrest, similar to that observed with colchicine, presumably by disruption of the mitotic spindle (Onfelt, 1983). Both spindle micro- tubules (Miura et al., 1978) and cytoplasmic microtubules in cultured rat glioma cells (Imura et al., 1980; Miura & Imura, 1987) and human fibroblasts (Sager et al., 1983) are disrupted by methylmercury. Damage to the microtubule system appears to underly the toxic effects of methylmer- cury on lymphocytes (Brown et al., 1988). Sager & Matheson (1988) have shown that disassembly of microtubules pre- cedes changes in other elements. In vitro polymerization of microtubules is also inhibited by methylmercury (Abe et al., 1975; Imura et al., 1980; Sager et al., 1983; De Saint-Georges et al., 1984; Miura et al., 1984). The effect on microtubules appears to be selective and does not involve other components of the cytoskeleton (e.g., vimentin or actin filaments) (Sager, 1988). Vogel et al. (1985) suggested that methylmercury binds to free sulf- hydryl groups on the ends and on the surface of microtu- bules. Sager et al. (1982, 1984) hypothesized that methylmer- cury might arrest the division of immature neurons at critical stages of brain development. They administered methylmercury to newborn mice at a time when the external granule layer (EGL) of the cerebellar cortex was rapidly dividing. They found fewer granule cells in the treated animals as well as a decrease in the percentage of late mitotic figures. This incomplete mitosis may have been responsible for the decreased cell numbers. Destruction or arrest of neuron growth during the development of the central nervous system may be an im- portant general mechanism in the pathobiology of prenatal damage (Rodier, 1977). The deranged cell migration reported by Choi et al. (1978) and the arrested cell div- ision found by Sager et al. (1984) might both be an expression of the action of methylmercury on microtubules and thus be consistent with the hypothesis that microtu- bule protein is an important molecular target for methyl- mercury in the developing brain. Enzymes associated with myelin formation have been found to be affected in the early postnatal period in rats (Grundt & Neskovic, 1985). Morphological de-differen- tiation of cultured brain cells has been shown to occur after addition of methylmercury (Grundt et al., 1981, 1982). These effects occur at lower methylmercury levels than those affecting energy metabolism (Grundt & Bakken, 1986). Choi et al. (1981) noted incomplete arborization of the dendritic tree of Purkinje cells in mice dosed with methylmercury in the early postnatal period. 9.3.3 Summary In summary, the clinical and epidemiological evidence indicates that prenatal life is more sensitive to the toxic effects of methylmercury than is adult life. The inhibition of protein synthesis is one of the earliest detectable biochemical effects in the adult brain, though the sequence of events leading to overt damage is not yet understood. Methylmercury can also react directly with important receptors in the nervous system, as shown by its effect on the acetylcholine receptor in the peripheral nerves. Concerning prenatal exposure, the effects of methylmercury seem to be quite different and of a much more general basic nature. It affects normal neuronal development and leads to altered brain architecture, hetero- topic cells, and decreased brain size. Methylmercury may also be exerting an effect, perhaps through inhibition of the micro- tubular system, on cell division during critical stages of formation of the central nervous system. 9.4 Dose-Effect and Dose-Response Relationships in Human Beings The relationships between concentrations in indicator media (e.g., blood and hair) or body burdens and the mag- nitude of an effect or frequency of an effect (response) were discussed in Environmental Health Criteria 1: Mercury (WHO, 1976b) using data from the Japanese and Iraqi out- breaks. They will be summarized here and considered in conjunction with studies reported subsequently. Other extensive and detailed reviews have been published since 1976 (Tsubaki & Irukayama, 1977; Inskip & Piotrowski, 1985; Tsubaki & Takahashi, 1986). Prenatal and adult exposures will be treated separ- ately in view of the differences, both qualitative and quantitative, in effects and dose-response relationships. 9.4.1 Adult exposure This section will examine new data published since 1976 and include a re-analysis of samples of hair, brain, and other tissues obtained from patients in Minamata and Niigata, Japan, the clinical follow-up of individuals exposed to methylmercury in Niigata, a new analysis of the Iraqi data, and reports on high fish consumers in Canada and elsewhere. 188.8.131.52 The Minamata and Niigata outbreaks Many of the original samples collected in Minamata and Niigata, Japan, had been analysed by the dithizone pro- cedure. Previous risk estimates were based on blood and hair mercury levels measured by this procedure. Tsubaki et al. (1978) reported on the repeat analysis of a hair sample from the patient in Niigata with the lowest hair mercury concentration at the onset of symptoms (52 µg/g). Re-analysis by atomic absorption yielded a value of 82.6 µg/g. Other hair samples collected from Niigata yielded values of about 100 µg/g for the onset of symptoms. As for blood samples, none were available for re-analysis by atomic absorption. The patient with the lowest blood level had a concentration of just below 200 µg/litre when extrapolated to time of onset of symptoms. However, the extrapolation used in Environmental Health Criteria 1: Mercury (WHO, 1976b) was based on a few points only, and the statistical uncertainty in the extrapolated value was high. Furthermore, the hair samples from the same patient indicated that the blood concen- tration was probably higher. Other blood samples extrapo- lated to values above 300 µg/litre. Further evidence by Tsubaki et al. (1978) indicates that hair and blood mer- cury levels at the onset of symptoms may not have been the true maximum values in these patients. Analyses of hair samples from Niigata indicate that the mercury concen- trations may have attained peak values about 2 months before the onset of symptoms. A stringent government warning against the consumption of contaminated fish was issued by June 1965; in many patients, symptoms appeared about 2 months later (WHO, 1980). Such evidence indicates that previous evaluations of the earlier data from Niigata may have underestimated the blood and hair concentrations associated with poisoning in the most sensitive patient and, therefore, overestimated the risk of poisoning. However, it should be noted that the atomic absorption method does need not always yield higher results than the dithizone method. Tsubaki et al. (1978) quote data from two hair samples in which agreement between the two results was excellent. Furthermore, the brain and other tissues were preserved for many years before the atomic absorption measurements were made (1973), whereas the dithizone method was used on fresh tissue (1956-1960). Information from the clinical follow-up from Niigata (Tsubaki & Irukayama, 1977; Tsubaki et al., 1978) suggests that there may be a latent period of several years between peak mercury concentrations and onset of symptoms. Such a latent period was reported in four patients whose maximum hair concentrations were in the range of 50-300 mg/kg, as measured by atomic absorption (Tsubaki et al., 1978). The results of studies on primates indicate latent periods of up to 1 year under conditions of continuous exposure. These latent periods are apparently dose dependent in ani- mals (Evans et al., 1977). A relationship between length of latent period and maximum hair concentrations was not apparent in the few Niigata cases. The delayed cases were mild, showing non-specific symptoms, so that cases of methylmercury poisoning could not be diagnosed with com- plete certainty. Follow-up studies examining mortality patterns in the Minamata population, including both poisoning cases and controls, did not reveal any clear pattern of mercury-related deaths (Tamashiro et al., 1984, 1986, 1987). 184.108.40.206 The Iraqi outbreak This outbreak of mass poisoning took place in the winter of 1971-1972 (Bakir et al., 1973; Kazantzis et al., 1976a,b; Al-Mufti et al., 1976; Greenwood, 1985). Seed grain treated with a methylmercury fungicide was used to prepare homemade bread in rural communities throughout the country. Consumption probably began in October-November, 1971, and the first cases of severe poisoning were admit- ted to hospital at the end of December, 1971. Total hospi- tal admissions rose to just over 6000, with most of these occurring in January, 1972. Over 400 deaths attributed to methylmercury were recorded in hospital. Both sexes and all ages were affected. Individual exposure ranged from a low non-toxic intake (when a few contaminated loaves were consumed) to prolonged daily intake (1-2 months), which in some cases produced severe signs of poisoning. The first effects were complaints of paraesthesia or malaise, followed by signs of ataxia, constriction of visual fields, and hearing loss. Some people experienced muscular weakness, which improved after treatment with an acetylcholinesterase inhibitor. Changes in peripheral nerve velocity were recorded in some severe cases. How- ever, most of the signs and symptoms were attributed to damage to the central nervous system. Effects on non- nervous tissue appeared to be absent or negligible. Early in the outbreak, it was noted that some pre- natally exposed infants showed signs of severe cerebral palsy similar to the cases reported in the Minamata out- break (Amin-Zaki et al., 1974). Later, more subtle effects on the developing nervous system were detected (Marsh et al., 1980). Fig. 2, reproduced from Environmental Health Criteria 1: Mercury (WHO, 1976b), demonstrates both dose- effect and dose-response relationships. For any given sign or symptom, e.g., paraesthesia, there is a background frequency indicated by the line parallel to the horizontal axis. Two scales are given for the horizontal axis because two different methods were used to estimate the maximum body burden. At higher values of the body burden, the frequency of response rises in proportion to the logarithm of the body burden (the horizontal axis has a logarithmic scale). The two lines (the horizontal and sloped line) form the shape of a "hockey stick", and this type of dose-response analysis is referred to by this name. The body burden corresponding to the point of inter- section of the two lines in the "hockey stick" was referred to as a "practical threshold" by the authors (Bakir et al., 1973). This threshold increases with increasing severity of the effects. Thus, using the upper scale in Fig. 2, the threshold for paraesthesia occurs at a body burden of about 25 mg, for ataxia at about 50 mg, for dysarthria at about 90 mg, for hearing loss at about 180 mg, and for death at over 200 mg. The dose-response relationship is illustrated by the "hockey-stick" line for each sign and symptom. Thus, the increase in frequency of paraesthesia increases in pro- portion to the log of the maximum body burden above the practical threshold. This increase is assumed to be caused by methylmercury. In fact, the only proof that methylmer- cury produced certain effects in this population is that: (a) these effects followed a known high exposure to methylmercury; (b) the frequency and severity of these effects increased with increasing exposure to methylmer- cury; (c) these effects are similar to those seen in other outbreaks of methylmercury poisoning; and (d) the major signs have been reproduced in some animal models. This cause-effect relationship is most difficult to establish at body burdens close to threshold levels. Here, the only effect may be the patient's complaint of paraes- thesia. This is a non-specific end-point that has a variety of causes other than methylmercury. The overall conclusion depends on a group-based statistical associ- ation between paraesthesia and exposure to methylmercury. Fig. 2 shows that the practical threshold value for par- aesthesia is a body burden of 25-40 mg mercury. Using the metabolic model discussed in section 6, the body burden of 25-40 mg mercury is equivalent to a blood level of 250- 400 µg/litre. This range of blood values compares favour- ably with the lowest-observed-effect level in Niigata. The "hockey-stick" analysis depicted in Fig. 2 implies the existence of a population threshold for the neurological effects of methylmercury in adults. Though the true popu- lation threshold cannot be determined, "the practical threshold" serves as an estimate of the population threshold. A re-analysis of Iraqi data has been published by Nordberg & Strangert (1976, 1978, 1982). This analysis assumed a continuous distribution of individual thresholds superimposed on a background frequency for such symptoms as paraesthesia. The analysis also took into account the inter-individual variation in whole-body biological half- times. The half-time was used together with other par- ameters of the metabolic model for methylmercury to esti- mate blood concentrations that would result from long-term daily intake of methylmercury. In turn, the blood levels can be related to the maximum body burdens by the distri- bution parameters presented in section 6. Thus, these two distributions of thresholds and biological half-times were combined to give an overall estimate of the risk of par- aesthesia for a given steady-state daily intake of methyl- mercury. The results are presented in Fig. 3. The calcu- lations indicate that an intake of 50 µg/day in an adult would involve a risk of about 0.3% of the symptoms of par- aesthesia, whereas an intake of 200 µg/day would involve a risk of about 8%. As pointed out by Nordberg & Strangert (1976, 1978, 1982), the background frequency of these non- specific symptoms, such as paraesthesia, plays a key role in determining the accuracy of the estimates of response of low frequencies. From the same Iraqi data, the authors estimated the background frequency of paraesthesia to be 6.3%. However, there is considerable uncertainty in deter- mining the precise value of the background frequency, and this uncertainty becomes the dominant cause of error at low rates of exposure. The re-analysis by Nordberg & Strangert is in agree- ment with the conclusions of Environmental Health Criteria 1: Mercury (WHO, 1976b) that the prevalence of the earliest effects could be expected to be approximately 5% in the adult population following a long-term daily methylmercury intake of 3-7 µg mercury/kg body weight. Such a long-term daily intake should give rise to blood concentrations of approximately 200 µg mercury/litre and maximum hair concentrations of about 50 µg mercury/g. It should be noted that estimates of the frequency of paraes- thesia below daily intakes of about 200 µg are extrapol- ations beyond the observed data and assume the absence of a population threshold. 220.127.116.11 Exposed populations in Canada More recently, clinical and epidemiological assess- ments have become available from studies of Canadian Indian population groups exposed seasonally over a long period of time to methylmercury through fish consumption. The levels of exposure, as determined by the analysis of blood and hair samples (or both), were generally lower than those in the diagnosed cases of poisoning studied in Iraq and in the Niigata epidemic in Japan. The highest blood level of mercury recorded in Canada was 660 µg/litre (Wheatly, 1979). Harada et al. (1976) clinically examined 89 residents of two Indian reservations who had been exposed to methyl- mercury by ingestion of contaminated fish. They found a number of signs and symptoms that have been associated with methylmercury intoxication. However, as the authors pointed out, the signs and symptoms were relatively mild and many of them were thought to be due to other factors. In the absence of controls, it is difficult to evaluate the possible role of methylmercury in the clinical findings reported in this study. A report of the Medical Services Branch of the Depart- ment of National Health and Welfare, Canada, recorded the clinical examination of 84 subjects who had a history of blood mercury levels above 100 µg/litre (Wheatley, 1979). Mild symptoms and signs were found that could be attributed to possible methylmercury exposure, but the causal relationship between exposure and effects was un- certain. However, of the 84 subjects examined, 11 cases were found where such an association could not be excluded. A major epidemiological study was carried out on Cree Indians from northwestern Quebec, Canada, exposed to methylmercury in fish (McKeown-Eyssen & Ruedy, 1983a,b). The authors claimed to find an association in adult men and women between a set of neurological abnormalities and the estimated exposure to methylmercury. However, it should be pointed out that this association was seen by only four of seven observers who reviewed videotaped recordings of the neurological screening tests. The severity of these neurological abnormalities is described as mild or questionable. It was not possible to estimate any threshold body burden or hair levels because this population had been exposed possibly for most of their lives and, therefore, peak values in previous years were unknown. However, observations on this population over several years indicated maximum blood concentrations below 600 µg/litre. On examining the reports from these studies, a WHO expert group (WHO, 1980) pointed to the potential importance of the long duration of exposure in the Canadian Indians and raised the possibility that this might be the first example of an endemic disease due to exposure to methylmercury. 18.104.22.168 Other fish-eating populations In addition to the extensive Canadian studies men- tioned above, some other reports have been published since 1976 on the blood or hair mercury levels in populations exposed to methylmercury through fish (Bacci et al., 1976; WHO, 1976b; Riolfatti, 1977 ; Haxton et al., 1979; Turner et al., 1980; Valciukas et al., 1986). Taking all reports into consideration, it seems that about 100 adults, who were exposed to methylmercury in fish, have been ident- ified outside Japan or Iraq as having had blood levels above 200 µg/litre. In none of these cases has a diag- nosis of Minamata disease been made, but it is possible that some may have suffered mild methylmercury poisoning. Even if it is assumed that none of these people suffered any adverse effects from the exposure, such a negative finding is still consistent (95% confidence level) with the maximum risk for paraesthesia of about 3%. 22.214.171.124 Special groups The above-mentioned risk estimates may not apply to pregnant women. Some severe cases of poisoning among preg- nant women exposed to high doses were reported in Iraq (WHO, 1980). At lower doses, transient paraesthesia and other mild symptoms have been reported (Tsubaki & Irukayama, 1977; Marsh et al., 1977, 1980, 1981). Maternal paraesthesia coincided with peak hair concentrations of methylmercury (Marsh et al., 1987). These observations suggest a greater risk for pregnant women than for non- pregnant women. 126.96.36.199 Summary The overall conclusion is that the reported relationships between response and body burden, hair, or blood mercury concentrations are essentially the same as those reported in Environmental Health Criteria 1: Mercury (WHO, 1976b). It is possible that the latent period after cessation of exposure may extend to one year or thereabouts. Pregnant women may exhibit paraesthesia at lower methylmercury exposure levels than non-pregnant women, suggesting a greater risk for pregnant women. 9.4.2 Prenatal exposure In contrast to the adult exposure situation, a con- siderable amount of new data has been published on dose- response relationships for human prenatal exposure. In 1976, when Environmental Health Criteria 1: Mercury (WHO, 1976b) was published, it was known that prenatal exposure could cause fetotoxic effects in human beings. In the Minamata outbreak, 23 children believed to be exposed in utero had severe cerebral involvement (palsy and retar- dation), whereas their mothers had mild manifestations or none at all (Takeuchi, 1977). Mercury levels in the mothers during pregnancy were not recorded (WHO, 1976b). There were no reports of prenatal poisonings in the Niigata outbreak. Psychological studies carried out in the Minamata area with children from elementary schools and junior high schools did not reveal major defects of IQ, compared with children from a control area (Harada & Moriyama, 1977). However, data on maternal exposure were not available. In another study from Minamata there was a correlation between mercury levels in umbilical blood and the occurrence of mental retardation in children (Harada et al., 1977). Studies of prenatal exposure to methylmercury based on populations in Canada, Iraq, and New Zealand have now been published. 188.8.131.52 Iraq Since the publication of WHO (1976b), results have been obtained from a clinical follow-up study on 29 infant-mother pairs in Iraq (Marsh et al., 1977, 1980). These reports described psychomotor retardation in infants caused by prenatal exposure (social bias excluded). A re- lationship was noted between maximum hair concentrations, measured in 1-cm segments during pregnancy, and the fre- quency of neurological effects in the infants. These ef- fects included delayed achievement of developmental mile- stones with or without neurological signs. The infants were 4´-5 years of age at the time of last examination. At hair mercury levels below 180 mg/kg, the infants showed minimal clinical neurological signs, but there was clear evidence of effects on psychomotor function, such as delayed walking or talking (Marsh et al., 1977). The fol- lowing criteria were adopted for developmental abnormali- ties: "motor retardation if the child was not walking at 18 months, speech retardation if not talking by 24 months, mental retardation or seizures (or convulsive- like attacks) according to the history provided by the mother, and neurological signs by agreement of the two examiners. No standards are available for head circum- ference or height of Iraqi children, so these factors were evaluated in terms of standard deviations below the mean for the group". Subsequently, a more complete report (Marsh et al., 1981) became available on 84 infant-mother pairs, includ- ing the 29 pairs described above. The peak maternal hair levels ranged from 0.4 to 640 mg/kg. Severe neurological deficits were observed in five children. These severe ef- fects are illustrated by the case report of one of these children. At the age of 4 years and 9 months, the child was blind and deaf and was unable to stand, walk, or talk. Tonic neck responses were present. All limbs showed an increase in tone and deep tendon reflexes with extensor plantar responses and abnormal posture of the wrist. Microcephaly was present, with a head circumference of 43 cm. The boy's height was 98 cm (Marsh et al., 1977). These severely affected children had been exposed to peak maternal levels during the second trimester of preg- nancy. These findings agree with a histopathological re- port of Choi et al. (1978), who found evidence of abnormal neuronal migration in the brain of Iraqi victims exposed maximally in the third and fourth months of pregnancy. This is known to be the critical period for neuronal mi- gration (Sidman & Rakic, 1973). These reports (Marsh et al., 1977, 1981) were based on the analysis of 1-cm segments of hair bundles. Peak hair concentrations probably underestimated the actual peak blood concentrations due to misalignment of hair strands during collection and to the differences in growth rates of individual strands (for further discussion, see Giovanoli-Jakubczak & Berg, 1974). Furthermore, the ana- lytical methods used in the hair analysis had a recovery of 73 ± 10% (Wigfield et al., 1981). Analysis of these Iraqi hair samples has been repeated using single-strand sampling and X-ray fluorescence giving complete recovery (Jaklevic et al., 1978). Detailed re- sults from the Iraqi outbreak have been reported by Marsh et al. (1987) are given in the Appendix. The effects in mothers during pregnancy were mild and transient, the most frequent symptom being paraesthesia. Symptoms were re- ported by the mother more frequently as exposure level increased. The severity of effects in the mother was much less than that in her offspring. Two of the mothers of the four most severely affected infants (pair numbers 45, 56, 68, and 70) recalled no symptoms, and the others com- plained only of transient paraesthesia during pregnancy. This confirmed the findings of Harada (1968) on infant- mother pairs in the Minamata outbreak. According to Marsh et al. (1987), the physical examin- ation of the children included: "observation, measurement of head circumference and body length, cranial nerve signs, speech, involuntary movements, strength, deep tendon reflexes, plantar responses, coordination, dexterity, primitive re- flexes, sensation, posture, and ability to sit, stand, walk, and run. A scoring system was adopted. When the neurological examination result was absolutely normal, the score was 0. In attempting to identify minimal signs, points were awarded for borderline findings such as possibly increased reflexes. Scores of 0-3 indicated no definite abnormality. The highest score in the most severely affected child was 11. The neuro- logical score was limited to signs found on examin- ation, and points were not awarded for features of retardation reported by the mother". Four cases received a neurological score of 11. The evidence is strong that these were severe cases of pre- natal methylmercury poisoning. The lowest-observed maxi- mum maternal hair level for these severe cases was 404 mg mercury/kg (mother-infant pair number 56). However, none of these observed symptoms were specific to prenatal methylmercury poisoning. The possibility of confounding factors was considered by the authors: "Maternal alcohol consumption was not a problem. They followed the Moslem precept to avoid alcohol, so the fetal alcohol syndrome was not a consideration. None of them smoked. The absence of antenatal medical supervision was uniform, so that no prescription medi- cation were taken and non-prescription medications were rarely available. There were no drug-induced fetotoxic effects to account for. These were agricul- tural communities with little socio-economic variation and no evidence of malnutrition". The evidence that such non-specific symptoms were caused by methylmercury is based on a statistical corre- lation of the frequency of these symptoms with methylmer- cury exposure and the absence of confounding factors. This was the first report of a milder syndrome of prenatal methylmercury poisoning, as opposed to the severe cases discussed above. An example of one such statistical correlation is given in Fig. 4. The frequency of a symptom of motor retardation (delayed onset of walking) is plotted against the logarithm of the maximum maternal hair concentration during pregnancy. The continuous line in Fig. 4a gives the best fit to the data calculated according to a non- parametric model. The frequency of response increases smoothly from virtually zero at a maternal hair mercury level of 5 mg/kg to approximately 70% at the highest con- centration. The shaded area is made up of individual 95% confidence limits for the individual response frequencies. The narrow confidence limits at the lowest hair levels (about 1 mg/kg) indicate a low background response fre- quency of less than 5%. Marsh et al. (1987) noted that, in five European countries, 10% of infants were not walking by the age of 16 months and 5% of a sample of infants in Paris, France, were not walking by 18 months, this being the criterium used in their study. Fig. 4. The relationship between the maximal maternal hair concentrations during pregnancy and the frequency of cases of motor retardation in offspring. Calculated from data in the Appendix according to the method of Cox et al. (1989). The non-parametric 95% confidence limits are superim- posed on two parametric models (the "hockey-stick" and logit models) in Fig. 4b. The figure shows that both parametric models are consistent with the data and with each other. The two curves are close to each other and both lie entirely within the non-parametric confidence limits. The collection of non-simultaneous confidence inter- vals, as depicted in the shaded area, cannot be used to obtain confidence intervals for such parameters as hair concentration for a 10% or 50% risk. Instead, the para- metric models were used. The result for the hockey-stick model are given in Table 10. The best estimate of a predicted threshold with the hockey-stick model is a maximum maternal hair concen- tration during pregnancy of 7.3 µg mercury/g with an upper confidence limit of 13.6 µg mercury/g. This best fit corresponds to a background of zero. This probably is the outcome of the small number of infant-mother pairs in the low exposure region. An assumed background frequency of 2% or 4% does not greatly change the estimated pre- dicted threshold values (8 and 9%, respectively). However, an assumed background of 4% greatly increases the upper 95% confidence limit, and an assumed background of 8% dramatically increases the estimate of the threshold to 119 µg mercury/g. These changes with increased values for the background frequency are due in part to the distribution of the data (the four abnormal values in the hair mercury concentration range from 10 to 50 µg/g) and in part to the assumed higher background values being further away from the best fit (0%). Table 10. The dependence of "practical" threshold values of hair mercury concentration and upper confidence limits on background frequency of responsesa ------------------------------------------------------ "Hockey-stick" model ------------------------------- Response Background Practical Upper 95% (%) threshold limit (mg mercury/kg) (best fit) ------------------------------------------------------ Retarded 0b 7.3 13.6 walking 2c 8 17 4c 9 190 8c 119 230 Central 2c 7.8 24 nervous 4c 8.4 32 system 9b 10 287 signs ------------------------------------------------------ a Data in Appendix according to Cox et al. (1989). b The best fit of the background frequency from data. c Assumed background frequency. However, a population threshold might not exist and the hockey-stick model will not then be applicable to these data. Thus the assumption that there is zero risk at the threshold value estimated by the hockey-stick model would be in error. The logit model provides a continuous relationship between dose and response and will there- fore give an estimate of the error in assuming zero risk at the threshold dose estimated by the hockey-stick model. Thus according to logit analysis, the excess risk over background is 5% at the threshold hair mercury level of 7.3 mg/g determined by the hockey-stick model. In short, if the hockey-stick model is not applicable, the estimated threshold of 7.3 mg/g will underestimate the risk by 5% (Fig. 4b). The logit model, in agreement with the hockey-stick model, gives a background frequency value of 0% according to best fit of the data. The logit and hockey-stick curves for abnormal central nervous system signs are depicted in Fig 5. The curves are superimposed on the 95% confidence limits estimated non- parametrically as described for Fig. 4. The two parametric models are consistent with each other and with the non-parametric confidence limits. With all three models, a statistically significant dose- response relationship exists. The hockey-stick model gives the best estimates of the practical threshold at 10 µg mercury/g but with a very high confidence limit (287 µg mercury/g) (Table 10). Lower assumed values of the background frequency (2% and 4%) give roughly the same practical threshold value, but with lower upper confidence limits (24 and 32 µg mercury/g, respectively). The lower upper limit values are due to the fact that a definite assumed value for the background frequency was used, whereas 9% is the best-fit estimate of background and consequently the overall uncer- tainty is greater. The logit model estimates a similar background fre- quency (9.3%). The excess risk over background is about 5% (Fig. 4). Thus, the data in the Appendix can be used to demon- strate a statistically significant dose-response relation- ship for signs and symptoms of prenatal poisoning. Esti- mates of a "threshold" or highest no-effect concen- tration were made with the hockey-stick model. These esti- mates are subject to considerable uncertainty due to the small number of infant-mother pairs. As in the adult dose- response relationships (see Fig. 3 and section 184.108.40.206), the background frequency can greatly influence estimates of risk at low (close to background) response rates, yet cannot be estimated accurately due to the small number of data points at the lowest hair mercury levels. 220.127.116.11 Canada McKeown-Eyssen et al. (1983) examined the relationship between prenatal exposure to methylmercury and neurologi- cal and developmental abnormalities among 234 Cree Indian children aged 12-30 months from four communities in north- ern Quebec, Canada. The authors described their study as follows: "A medical team visited each community and examined 95% of the eligible children and their mothers, `blinded' to their methylmercury exposure. One of the four pediatric neurologists documented each child's height, weight, and head circumference, assessed dysmorphic and congenital features, and reported the presence or absence of acquired disease. A neurologi- cal examination was also conducted and included an assessment of special senses, cranial nerves, sensory function, muscle tone, stretch reflexes, coordination, and persistence of the Babinski response which was judged to be abnormal for the child's age, as well as a summary of the presence or absence of neurological abnormality. Finally, the neurologist assessed the child's development by use of the Denver developmental scale; for each child, the results were expressed as the percentage of total test items that were passed, separately for gross and fine motor development, language development, and personal/social skills. "Each mother was interviewed about her alcohol and tobacco consumption both during the relevant pregnancy and at the time of the interview. Because of the un- certainty about the accuracy of the reporting, women were classified simply as users of alcohol or ab- stainers, and as smokers or nonsmokers according to whether they reported ever drinking or smoking. Caf- feine intake was calculated from answers to questions on tea and coffee consumption. "Information on pregnancy, labour, and delivery were sought from the medical certificate of childbirth, a standard form that reports the major characteristics of pregnancy and delivery for all births in Quebec. Because some deliveries occurred in the bush, these certificates could be obtained for only 85% of the births. "Methylmercury concentrations of the hair were measured in alternate one centimeter segments, begin- ning with the scalp-end segment. The maximum concen- tration in the segment of hair corresponding to the period from one month before conception to one month after delivery was used as an index of prenatal exposure. "A search was conducted to establish which measures (if any) of neurologic function and development were associated with methylmercury exposure. This was achieved by use of a regression analysis of the relationship between the methylmercury exposure index and the results of four tests of neurologic function (coordination, cranial nerves, muscle tone or re- flexes, and an overall neurologic assessment) and measures of four aspects of the Denver developmental scale (gross and fine motor development, language development, and personal/social skills). Once the measure of neurologic function most closely associated with exposure was identified, the odds ratio was esti- mated from a discriminant analysis in which children were classified as cases or controls depending on the presence or absence of abnormality of the relevant neurologic function. The analysis distinguished between the cases and controls first on the basis of confounding variables potentially associated with the neurologic abnormality (child's age, duration of breast-feeding, mother's age, and mother's smoking habit and consumption of beverages containing alcohol and caffeine), and then on the basis of the prenatal indices of methylmercury exposure". The ages of the mothers covered a wide range: 13% were below 20 years of age and 15% were at least 35 years old. About two-thirds said they consumed alcohol and slightly more were smokers. The percentages of high risk preg- nancies, complications at delivery, and duration of breast feeding were similar for boys and girls. The birth weight of the children tended to be above normal (34% weighed over 4 kg). The mean index of mercury exposure (maximum maternal hair concentration during pregnancy) was the same (6 µg/g), for both boys and girls, and only 6% of values were above 20 µg/g. None of the children showed abnor- mal physical development, but "abnormality" of muscle tone or reflexes was positively associated with the pre- natal index of methylmercury exposure (P <0.05, 2-tailed). The highest maternal hair level in this study group of 97 males was 23.9 µg/g (Table 10). No other measure of neurological function or development was significantly associated with methylmercury exposure either before or after adjustment for confounding variables. In girls, no adverse effects were associated with mercury exposure. In fact, a negative association was found between one neuro- logical abnormality (incoordination) and prenatal mercury exposure. McKeown-Eyssen et al. (1983) expressed some reasons to doubt the "importance" of the finding of methylmercury effects in boys. They noted that the "ab- normality of muscle tone or reflexes . . . was . . . of doubtful clinical importance", that previously reported prenatal effects were at higher exposures, that a consist- ent dose-response relationship was not seen in the boys (Table 11), and that the effect was only seen in boys and not in girls. Consequently, in their interpretation of the data, the authors did not exclude the possibility that the positive findings were chance observations. Table 11. Prevalence of abnormality of muscle tone or reflexes according to maternal mercury levels during pregnancya -------------------------------------- Prenatal Number % abnormal exposureb of boys index (mg/kg) -------------------------------------- 0 - 1.9 19 15.8 2 - 2.9 18 5.6 3 - 4.9 19 26.3 5 - 6.9 14 0 7 - 12.9 14 7.1 13 - 23.9 13 38.5 Total 97 15.5 -------------------------------------- a Adapted from: McKeown-Eyssen et al. (1983). b Maximum maternal hair concentration during pregnancy. 18.104.22.168 New Zealand Kjellstrom et al. (1986) reported preliminary tests carried out on prenatally exposed children in a fish- eating group in New Zealand. The study started with a cohort of 11 000 recent mothers and their offspring. Approximately 1000 of these mothers reported that they had consumed fish more than three times per week during preg- nancy. Analysis of samples of maternal head hair revealed that 73 of the mothers had levels above 6 mg/kg. People of Pacific Island descent accounted for 62%, Maoris for 27%, and Europeans 11% of this group of 73. Only 31 offspring from this group of 73 mothers could be contacted by the time they were 4 years old. They were matched according to ethnic group, maternal age, birth- place, and birth date with offspring having low prenatal exposure to methylmercury (maternal hair mercury below 6 mg/kg). The Denver Development Test, carried out on a double-blind basis, was used to assess the effects of methylmercury. Abnormal or questionable results were found in 17% of the controls, compared with 50% in the children exposed to high mercury levels (maternal hair mercury levels above 6 mg/kg). The difference was statistically significant. A statistically significant dose-response relationship was found between mean maternal hair mercury levels during pregnancy and the frequency of deficient Denver Test results. No influence of socio-economic factors (based on place of residence), maternal health status, or smoking habits was seen. However, other confounding factors inherent in these studies make it difficult to draw final conclusions. In 1985, it was possible to locate 61 of the original 73 high-exposure children and to conduct detailed psycho- logical and scholastic tests (carried out on a double- blind basis) at the age of 6-7 years (Kjellstrom et al., 1989). At this stage, the children had completed at least one year at school. These tests included, among others, the revised Wechsler Intelligence Scale for Children (WISC-R) and the Test of Language Development (TOLD). The high-exposure children (maternal hair mercury levels within the range of 6-86 mg/kg, with the second highest value being 19.6 mg/kg) were compared with three matched groups: one group with maternal hair mercury levels of 3-6 mg/kg and two groups with levels below 3 mg/kg (one group with high fish consumption and one with low fish consumption). The mothers were matched for child's sex and maternal ethnic group, age, smoking habits, residence area, and residence time in New Zealand. The results of the different tests were correlated, and showed that individual children with low scores in the TOLD or WISC-R tests also had low scores in the other tests. For those children who had been tested both at age 4 (Kjellstrom et al., 1986) and at age 6-7 (Kjellstrom et al., 1989), there was also a correlation between the Denver Test and the IQ scores (WISC-R scale). The sub- groups were small, but the data indicated that a child who had poor Denver Test results was highly likely to score very poorly in the IQ test at school age. According to the authors' summary, although "methylmercury exposure con- tributes only a small part of the variation in tests results" and "results of the psychological test vari- ables are influenced by the child's ethnic background", the study suggests that "an average hair mercury level during pregnancy of 13-15 mg/kg (equivalent to about 25 mg/kg peak mercury level) may be associated with a decreased test performance" (Kjellstrom et al., 1989). It should be noted that the studies in Canada and Iraq used maximum maternal hair concentrations during pregnancy based on 1-cm segments (roughly one month's hair growth). Kjellstrom et al. (1986, 1989) state that their mean maternal hair values should be multiplied by a factor of 1.5 to obtain the maximum 1-cm value during pregnancy. 22.214.171.124 Summary Severe derangement of the developing central nervous system can be caused by prenatal exposure to methylmercury. The lowest level (maximum maternal hair mercury concentration during preg- nancy) at which severe effects were observed was 404 µg/g in the Iraqi outbreak. The highest no-observed-effect-level (NOEL) for severe effect was 399 µg/g. Fish-eating populations in Canada and New Zealand have also been studied for prenatal effects, but exposure levels were far below the highest NOEL for severe effects in Iraq and no severe effects were seen. Evidence of psychomotor retardation (delayed achievement of developmental milestones, a history of seizures, abnormal reflexes) were seen in the Iraqi population at maternal hair levels well below those associated with severe effects. A statistical analysis revealed that one of these effects (motor retardation) rose above the background frequency at maternal hair mercury levels (maximum level during pregnancy) of 10-20 µg/g. This range of values in maternal hair is consist- ent with all available evidence and can be accepted as the range of critical concentrations. The Canadian study found that maternal hair levels were positively associated with abnormal muscle tone or reflexes in boys, but not in girls (the highest maximum maternal hair level during pregnancy was 23.9 µg/g). The New Zealand study found evidence of developmental retardation (according to the Denver Test) in 4-year-old children at average maternal hair mercury levels during pregnancy within the range of 6-86 µg/g (the second highest value was 20 µg/g). The New Zealand mercury values should be multiplied by 1.5 to convert to maximum maternal hair levels in pregnancy. 10. EVALUATION OF HUMAN HEALTH RISKS 10.1 Exposure Levels and Routes In view of the restrictions placed upon the use of methylmercury in most countries, occupational exposure will be low. The major source of human exposure to methylmercury is through the diet, more specifically from the consumption of fish and fish products. In most countries, the import- ant food fishes have methylmercury levels in their edible portion not exceeding 200-300 µg/kg. However, levels in such predatory species as ocean tuna, shark, and swordfish (even from non-polluted areas), as well as freshwater pike, walleye, and bass, may contain methylmercury levels in excess of 1000 µg/kg. In view of the worldwide vari- ation in dietary patterns and extent of pollution, it is difficult to calculate a general exposure level for methylmercury. However, assuming an average daily con- sumption of 20 g of non-predatory species containing 200 µg/kg, the daily methylmercury intake would be 4 µg. It has been estimated that long-term intake at this level would raise the blood methylmercury levels by 4 µg/litre, and hair levels by 1 µg/g. However, in some countries the average consumption can be as high as 100 g/day and may consist mainly of predatory species. In these cases, methylmercury intakes can exceed 100 µg/day. 10.2 Toxic Effects 10.2.1 Adults Concerning the risks in adults exposed to methylmer- cury, the conclusions reached in Environmental Health Criteria 1: Mercury (WHO, 1976) and the 1980 interim evaluation remain unchanged. A daily methylmercury con- sumption of 0.48 µg/kg body weight (WHO, 1989b) will not result in any detectable adverse effects. However, a daily intake of 3-7 µg/kg body weight would cause adverse effects on the nervous system, manifested as an approxi- mately 5% increase in the incidence of paraesthesia. Hair concentrations would be approximately 50-125 µg/g at this level of intake. Clinical observations in Iraq suggested that women are more sensitive to the toxic effects of methylmercury during pregnancy. 10.2.2 Prenatal exposure The report (WHO, 1980) that evaluated the health hazards from exposure to methylmercury through the con- sumption of fish underlined that damage to the fetal brain caused by prenatal exposure to methylmercury could be the critical effect, and that more information was needed for a proper risk assessment. Since 1980, experimental evi- dence to elucidate the mechanisms involved in the neuro- toxic action of methylmercury on the fetal brain has ac- cumulated and has been reviewed in this monograph. Further analysis of data from the Iraqi outbreak has extracted additional information relating to the effects of prenatal methylmercury exposure. From these sets of evidence, a pattern has emerged that permits the construction of a biological model for the neurotoxic action of methylmer- cury on the fetal brain. Methylmercury inhibits the growth of the fetal brain and the migration of neurons from the embryological gener- ation layer to the final destination in the brain cortex. This has been demonstrated in clinical cases in Japan and Iraq. An inhibition of brain growth is indicated by a decrease in brain size and weight, as was observed in studies on monkeys and humans. The inhibition of fetal brain development caused by methylmercury exposure results in the behavioural changes and reduced cognitive and motor ability found in clinical cases. It has been demonstrated that methylmercury interferes with microtubule formation, cell division, and neuronal protein synthesis, all of which could explain the effects described above. The model emerging to explain the neurological effects of methylmercury is a continuous dose-effect relationship, with a range from subtle changes in brain function (indi- cated by psychological tests) at low dose levels to a severe neurological syndrome of cerebral palsy with pro- nounced changes in the organization of brain structure at high exposure levels. However, the possibility of detecting and characterizing the methylmercury level at which the subtle and early adverse effects on the fetal brain may arise is limited by the availability of sensi- tive test procedures. At present, some effects can only be detected and adequately characterized in the epidemi- ological studies of fairly large populations. The crucial question is the actual exposure level (or body burden) of methylmercury in humans which can lead to subtle changes in the offspring. The actual exposure levels and patterns are usually unknown, but the effect can be related to the hair mercury level and an approxi- mate daily exposure calculated from the known kinetic parameters for methylmercury accumulation, distribution, and excretion. The statistical analysis of data on 84 infant-mother pairs (maternal peak hair mercury levels during pregnancy of 0.4-640 µg/g) showed that, at maternal hair concen- trations above 70 µg/g, children exhibited evidence of abnormal neurological signs, e.g., increased muscle tone in the leg and exaggerated deep tendon reflexes, often accompanied by ataxia together with a history of develop- mental delay. This statistical analysis indicated a 30% risk of these abnormal findings at maternal hair mercury concentrations around 70 µg/g. The data from the Iraqi outbreak do not permit firm conclusions to be drawn con- cerning the risk of adverse effects below that level. However, by applying the biological model described above, the extrapolation method of Cox et al. (1989), and the evaluation of other currently available data, it can be calculated that a maternal hair mercury concentration of 10-20 µg/g implies a 5% risk. The possibility cannot be excluded that effects detectable by psychological and behavioural testing or subclinical effects might occur at even lower levels of exposure, but evidence is lacking. 10.3 Conclusions The general population does not face a significant health risk from methylmercury. Certain groups with a high fish consumption may attain a blood methylmercury level (about 200 µg/litre, corresponding to 50 µg/g of hair) associated with a low (5%) risk of neurological damage to adults. The fetus is at particular risk. Recent evidence shows that at peak maternal hair mercury levels above 70 µg/g there is a high risk (more than 30%) of neurological disorder in the offspring. A prudent interpretation of the Iraqi data implies that a 5% risk may be associated with a peak mercury level of 10-20 µg/g in the maternal hair. There is a need for epidemiological studies on chil- dren exposed in utero to levels of methylmercury that re- sult in peak maternal hair mercury levels below 20 µg/g, in order to screen for those effects only detectable by available psychological and behavioural tests. 11. RECOMMENDATIONS 11.1 Gaps in Knowledge In spite of significant advances in our understanding of the toxicity and potential hazard of methylmercury, there remain areas in which there is an urgent need for additional studies. The most important of these areas is the lower end of the dose-response relationship for prenatal exposures. This will require well coordinated and designed, inter- national epidemiological studies that consider all rel- evant confounding factors (e.g., drugs, alcohol, smoking). As part of these studies, there is a need to develop objective measurements of clinical manifestations. The potential ability of selenium and other dietary components (e.g., antioxidants) to alter the toxic responses elicited by methylmercury should be investigated. These studies will not only need to describe this interaction, but also to provide data that can be used quantitatively in a risk assessment of methylmercury, particularly for the fetus. The mechanisms of damage to both the mature and developing nervous system remain to be elucidated. As no information on the relative vulnerability of the brain during differ- ent periods of pregnancy is available, more experimental work is needed to shed light on this aspect, which is important for risk assessment and clinical judgement. The selective damage to the nervous system and to specific areas in the brain, the long period in the case of adult poisoning, and the high vulnerability of the developing nervous system (including sex differences in suscepti- bility) are still unexplained. 11.2 Preventive Measures In populations that consume large amounts of fish (e.g., 100 g/day), the hair levels of methylmercury in women of child-bearing age should be monitored. If the results of these monitoring activities indicate excessive exposure to methylmercury, appropriate and practical measures, such as dietary recommendations, should be taken to reduce the possibility of long-term exposure during pregnancy and to keep it below internationally recommended allowable intakes. Measures to reduce methylmercury exposure via the con- sumption of fish will need to consider the impact of these measures on the overall dietary requirements of these individuals. 12. PREVIOUS EVALUATIONS BY INTERNATIONAL BODIES The human health risks from exposure to methylmercury were previously evaluated in Environmental Health Criteria 1: Mercury (WHO, 1976b). This was followed by a brief up- date (WHO, 1980), based upon a technical report prepared by the Monitoring and Assessment Research Centre (MARC, 1981). In the thirty-third report of the Joint FAO/WHO Expert Committee on Food Additives (JECFA), it was rec- ommended that the permissible tolerable weekly intake (PTWI) for methylmercury in adults be maintained at 200 µg (3.3 µg/kg body weight) (WHO, 1978; WHO, 1989b). However, the Committee noted that pregnant women and nursing mothers are likely to be at greater risk, although the available data were insufficient to recommend a specific mercury intake for these population groups. Regulatory standards established by some national bodies in different countries and the EEC are summarized in the data profile of the International Register of Potentially Toxic Chemicals (IRPTC, 1987). REFERENCES ABE, T., HAGA, T., & KUROKAWA, M. (1975) Blockage of axoplasmic transport and depolymerization of reassembled microtubules by methylmercury. Brain Res., 86: 504-508. AIREY, D. (1983) Total mercury concentrations in human hair from 13 countries in relation to fish consumption and location. Sci. total Environ., 31: 157-180. AL-MUFTI, A.W., COPPLESTONE, J.F., KAZANTZIS, G., MAHMOUD, R.M., & MAJID, M.A. (1976) Epidemiology of organomercury poisoning in Iraq: I. Incidence in a defined area and relationship to the eating of contaminated bread. Bull. World Health Organ., 53(Suppl.): 23-36. AL-SHAHRISTANI, H. & SHIHAB, K. (1974) Variation of biological half- life of methylmercury in man. Arch. environ. Health, 28: 342-344. AMIN-ZAKI, L., ELHASSANI, S., MAJEED, M.A., CLARKSON, T.W., DOHERTY, R.A. & GREENWOOD, M. (1974) Intra-uterine methylmercury poisoning in Iraq. Pediatrics, 54(5): 587-595. AMIN-ZAKI, L., ELHASSANI, S., MAJEED, M.A., CLARKSON, T.W., DOHERTY, R.A., GREENWOOD, M., & GIOVANOLI-JAKUBCZAK, T. (1976) Perinatal methylmercury poisoning in Iraq. Am. J. dis. Child., 130: 1070-1076. ANDO, S., TOYODA, Y., NAGAI, Y., & IKUTA, F. (1985) Abnormalities in gangliosides and other lipids of monkey, rabbit and human brains with chronic organic mercury intoxication. Jpn. J. exp. Med., 55(1): 1-6. ARAKI, K., WAKABAYASHI, M., SAKIMURA, K., KUSHIYA, E., OZAWA, H., KUMAMOTO, T., & TAKAHASHI, Y. (1981) Decreased uptake of GABA by dorsal ganglia in methylmercury treated rats. Neurotoxicity, 2: 557-566. ARITO, H., NOBORU, H., & SHITZU, T. (1983) Effect of methylmercury chloride on sleep-waking rhythms in rats. Toxicology, 28: 335-345. ASCHNER, M. (1986) Changes in axonally transported proteins in the rat visual system following systemic methylmercury exposure. Acta Pharmacol. toxicol., 59(2): 151-157. ASCHNER, M. & CLARKSON, T.W. (1987) Mercury 203 distribution in pregnant and non-pregnant rats following systemic infusions with thiol- containing amino acid. Teratology, 36: 321-328. ASCHNER, M., RODIER, P.M., & FINKELSTEIN, J.N. (1986) Reduction of axonal transport in the rat optic system after direct application of methylmercury. Brain Res., 381(2): 244-250. ATCHISON, W.D. (1986) Extracellular calcium-dependent and independent effects of methylmercury on spontaneous and potassium-evoked release of acetylcholine at the neuromuscular junction. J. Pharmacol. exp. Ther., 237(2): 672-680. ATCHISON, W.D. & NARAHASHI, T. (1982) Methylmercury-induced depression of neuromuscular transmission in the rat. Neurotoxicology, 3: 37-50. BACCI, E., ANGOTZI, G., BRALIA, A., LAMPARIELLO, L., & ZANETTE, E. (1976) Etude sur une population humaine exposée au méthylmercure par la consommation de poisson. Rev. int. Océanogr. méd., 41/42: 127-141. BAKIR, F., DAMLUJI, S.F., AMIN-ZAKI, L., MURTADHA, M., KHALIDI, A., AL-RAWI, N.Y., TIKRITI, S., DHAHIR, H.I., CLARKSON, T.W., SMITH, J.C., & DOHERTY, R.A. (1973) Methylmercury poisoning in Iraq. Science, 181: 230-241. BALLATORI, N. & CLARKSON, T. (1985) Biliary secretion of glutathione and of glutathione-metal complexes. Fundam. appl. Toxicol., 5: 816- 831. BARTOLOME, J., TREPANIER, P., CHAIT, E.A., SEIDLER, F.J., DESKIN, R., & SLOTKIN, T.A. (1982) Neonatal methylmercury poisoning in the rat: effects on development of central catecholamine neurotransmitter systems. Toxicol. appl. Pharmacol., 65: 92-99. BEGLEY, T.P., WALTS, A.E., & WALSH, C.T. (1986) Bacterial organomercurial lyase: overproduction, isolation, and characterization. Biochemistry, 25: 7186-7192. BEIJER, J. & JERNELOV, A. (1979) Methylation of mercury in aquatic environments. In: Nriagu, J.O., ed. The biogeochemistry of mercury in the environment, Amsterdam, Oxford, New York, Elsevier Science Publishers, pp. 203-210. BERLIN, M. (1976) Dose-response relations and diagnostic indices of mercury concentrations in critical organs upon exposure to mercury and mercurials. In: Nordberg, G.F., ed. Effects and dose-response relationships of toxic metals, Amsterdam, Oxford, New York, Elsevier Science Publishers, pp. 235-245. BERLIN, M. (1986) Mercury. In: Friberg, L., Nordberg, G.F., & Voulk, V., ed. Handbook on the toxicology of metals, 2nd ed., Amsterdam, Oxford, New York, Elsevier Science Publishers, pp. 387-445. BERLIN, M., BLOMSTRAND, C., GRANT, C.A., HAMBERGER, A., & TROFAST, J. (1975) Tritiated methylmercury in the brain of squirrel monkeys. Arch. environ. Health, 30: 591-597. BERNAUDIN, J.F., DRUET, E., DRUET, P., & MASSE, R. (1981) Inhalation or ingestion of organic or inorganic mercurials produces auto-immune disease in rats. Clin. Immunol. Immunopathol., 20: 129- 135. BERNHARD, M. & ANDREAE, M.O. (1984) Transport of trace metals in marine food chains. In: Nriagu, J.O., ed. Changing metal cycles and human health, Berlin, Heidelberg, New York, Springer-Verlag, pp. 143- 168. BERNHARD, M., BUFFONI, R., & RENZONI, A. (1982) Mercury in Mediterranean tuna. Why is their level higher than in Atlantic tuna? Thalassia (Yugosl.), 18: 231-243. BIENVENUE, E., BOUDOU, A., DASMAZES, J.P., GAVACH, C., GEORGESCAULD, D., SANDEAUX, J., SANDEAUX, R., & SETA, P. (1984) Transport of mercury compounds across biomolecular lipid membranes: effect of lipid composition, pH and chloride concentration. Chem.-biol. Interact., 48: 91-101. BJORNBERG, A.A., HAKANSON, L., & LUNDBERGH, K., (1988) A theory on the mechanisms regulating the bioavailability of mercury in natural water. Environ. Pollut., 49: 53-61. BLAKLEY, B.R. (1984) Enhancement of urethan-induced adenoma formation in Swiss mice exposed to methylmercury. Can. J. comp. Med., 48(3): 299- 302. BORNHAUSEN, M., MUSCH, H.R., & GREIM, H. (1980) Operant behaviour performance changes in rats after prenatal methyl-mercury exposure. Toxicol. appl. Pharmacol., 56: 305-310. BROWN, D., RUEHL, K., BORMANN, S., LITTLE, J. (1988). Effects of methylmercury on the microtubule system of mouse lymphocytes. Toxicol. appl. Pharmacol., 94: 66-75. BRZEZNICKA, E.A. & CHMIELNICKA, J. (1985a) Interaction of alkylmercury compounds with sodium selenite. III. Biotransformation, levels of metallothioneinlike proteins and endogenous copper in some tissues of rats exposed to methyl or ethylmercuric chloride with and without sodium selenite. Environ. Health Perspect., 60: 423-431. BRZEZNICKA, E.A. & CHMIELNICKA, J. (1985b) Interaction of alkylmercuric compounds with sodium selenite. II. Metabolism of methylmercuric chloride administered alone and in combination with sodium selenite in rats. Environ. Health Perspect., 60: 411-421. BUCHET, J.P., ROLES, H., HUBERMONT, G., & LAUWERYS, R. (1978) Placental transfer of lead, mercury, cadmium, and carbon monoxide in women. Environ. Res., 15: 494-503. BUCHET, J.P., LAUWERYS, R., VANDEVOORDE, A., & PYCKE, J.M. (1983) Oral daily intake of cadmium, lead, manganese, chromium, mercury, calcium, zinc, and arsenic in Blegium. A duplicate meal study. Food chem. Toxicol., 21: 19-24. BURBACHER, T.M., MONNETT, C., GRANT, K.S., & MOTTET, N.K. (1984) Methylmercury exposure and reproductive dysfunction in the nonhuman primate. Toxicol. appl. Pharmacol., 75(1): 18-24. BURTON, G.V. & MEIKLE, A.W. (1980) Acute and chronic methylmercury poisoning impairs rat adrenal and testicular function. J. Toxicol. environ. Health, 6: 597-606. BYRNE, A.R. & KOSTA, L. (1974) Simultaneous neutron activation determination of selenium and mercury in biological samples by volatilization. Talanta, 21: 1083-1090. CANADA-ONTARIO STEERING COMMITTEE (1983) Mercury pollution in the Wabigon-English river system of Northwestern Ontario, and possible remedial measures. Summary of a technical report, Toronto, Canada-Ontario Steering Committee, Provincial Ministry of the Environment. CAPELLI, R., MINGANTI, V., SEMINO, G., & BERTARINI, W. (1986) The presence of mercury (total and organic) and selenium in human placentae. Sci. total Environ., 48(1/2): 69-79. CAPPON, C.J. (1981) Mercury and selenium content and chemical form in vegetable crops grown on sludge-amended soil. Arch. environ. Contam. Toxicol., 10: 673-689. CAPPON, C.J. (1987) Uptake and speciation of mercury and selenium in vegetable crops grown on compost-treated soil. Water Air Soil Pollut., 3: 353-361. CAPPON, C.J. & SMITH, J.C. (1978) A simple and rapid procedure for the gas chromatographic determination of methylmercury in biological samples. Bull. environ. Contam. Toxicol., 19: 600-607. CARMICHAEL, N., CAVANAGH, J.B., & RODDA, R.A. (1975) Some effects of methylmercury salts on the rabbit nervous system. Acta neuropathol., 32: 112-125. CHANG, L.W. (1977) Neurotoxic effects of mercury - a review. Environ. Res., 14: 329-373. CHANG, L.W. & SUBER, R. (1982) Protective effect of selenium on methylmercury toxicity: a possible mechanism. Bull. environ. Contam. Toxicol., 29: 285-289. CHANG, L.W., GILBERT, M., & SPRECHLER, J. (1978) Modification of methylmercury neurotoxicity by vitamin E. Environ. Res., 17: 356-366. CHEN, W.J., BODY, R.L., & MOTTET, N.K. (1979) Some effects of continuous low-dose congenital exposure to methylmercury on organ growth in the rat fetus. Teratology, 20: 31-36. CHEN, W.J., BODY, R.L., & MOTTET, N.K. (1983) Biochemical and morphological studies of monkeys chronically exposed to methylmercury. J. Toxicol. environ. Health, 12: 407-416. CHERIAN, M.G., HURSH, J.B., CLARKSON, T.W., & ALLEN, J. (1978) Radioactive mercury distribution in biological fluids and excretion in human subjects after inhalation of mercury vapor. Arch. environ. Health, 33: 109-114. CHEUNG, M.K. & VERITY, M.A. (1985) Experimental methylmercury neurotoxicity: locus of mercurial inhibition of brain protein synthesis in vivo and in vitro. J. Neurochem., 44(6): 1799-1808. CHMIELNICKA, J., BRZEZNICKA, B., BARANSKI, B., & SITAREK, K. (1985) The effect of methylmercury on prenatal development and trace metal distribution in pregnant and fetal rats. Biol. Trace Elem. Res., 8: 191-201. CHOI, B.H. & LAPHAM, L.W. (1976) Interactions of neurons and astrocytes during growth and development of human fetal brain in vitro. Exp. Mol. Pathol., 24: 110-125. CHOI, B.H. & LAPHAM, L.W. (1980) Effects of meso-2-3, dimercaptosuccinic acid on methylmercury injured human fetal astrocytes in vitro: a lapse cinematographic phase and electron microscopic study. Fed. Proc., 39: 396. CHOI, B.H. & KIM, R.C. (1984) The comparative effects of methylmercuric chloride and mercuric chloride upon DNA synthesis in mouse fetal astrocytes in vitro. Exp. Mol. Pathol., 41(3): 371-376. CHOI, B.H., LAPHAM, L.W., AMIN-ZAKI, L., & SALEEM, T. (1978) Abnormal neuronal migration, deranged cerebral cortical organization and diffuse white matter astrocytosis of human fetal brain. A major effect of methylmercury poisoning in utero. J. Neuropathol. exp. Neurol., 37: 719-733. CHOI, B.H., CHO, K.H. & LAPHAM, L.W. (1979) Effects of methylmercury on human fetal brain cells in culture: a time lapse cinematographic phase and electron microscopic study. J. Neuropathol. exp. Neurol., 33: 307. CHOI, B.H., CHO, K.H., & LAPHAM, L.W. (1980) Effects of methylmercury on DNA synthesis of human fetal astrocytes: a radioautographic study. Brain Res., 202: 238-242. CHOI, B.H., KUDO, M., & LAPHAM, L.W. (1981) A golgi and electron- microscopic study of cerebellum in methylmercury poisoned neonatal mice. Acta neuropathol., 54: 233-237. CHOI, B.H., KIM, R.C., & PECKHAM, N.H. (1988) Hydrocephalus following prenatal methylmercury poisoning. Acta neuropathol. 75: 325-330. CLARKSON, T.W. (1983) Methylmercury toxicity in the mature and developing nervous system: possible mechanism. In: Sarker, B. ed. Biological aspects of metals and metal related diseases, New York, Raven Press, pp. 183-192. CLARKSON, T.W., FRIBERG, L., HURSH, J.B., & NYLANDER, M. (1988) The prediction of intake of mercury vapor from amalgams. In: Clarkson, T.W., Friberg, L., Nordberg, G.F., & Sager, P.R., ed. Biological monitoring of toxic metals, New York, London, Plenum Press, pp. 247-264. CONCAS, A., CORDA, M.G., SALIS, M., MULAS, M.L., MILIA, A., CORONGIU, F.P., & BIGGIO, G. (1983) Biochemical changes in the rat cerebellar cortex elicited by chronic treatment with methylmercury. Toxicol. Lett., 18: 27-33. COX, C., CLARKSON, T.W., MARSH, D.O., AMIN-ZAKI, L., TIKRITI, S., & MYERS, G.G. (1989). Dose-response analysis of infants prenatally exposed to methylmercury. An application of a single compartment model to single-strand hair analysis. Environ. Res., 49: 318-332. COYLE, P. & HARTLEY, T. (1981) Automated determination of mercury in urine and blood by the Magos reagent and cold vapor atomic absorption spectrometry. Anal. Chem., 53: 354-356. CUOMO, V., AMBROSI, L., ANNAU, Z., CAGIANO, R., BRUNELLO, N., & RACAGNI, G. (1984) Behavioural and neurochemical changes in offspring of rats exposed to methylmercury during gestation. Neurobehav. Toxicol. Teratol., 6(3): 249-254. CURLE, D.C., RAY, M., & PERSAUD, T.V.N. (1983) Methymercury toxicity: in vivo evaluation of teratogenesis and cytogenic changes. Anat. Anz. (Jena), 153: 69-82. DAVIES, T.W. & NIELSEN, S.W. (1977) Pathology of subacute methylmercurialism in cats. Am. J. Vet. Med., 38: 59-67. DAVIES, T.W., NIELSEN, S.W., & KIRCHNER, C.H. (1976) The pathology of subacute methylmercurialism in swine. Cornell Vet., 66: 32-55. DAVIES, T.W., NIELSEN, S.W., & JORTNER, B.S. (1977) Pathology of chronic and subacute canine methylmercurialism. J. Am. Anim. Hosp. Assoc., 13: 369-381. DE SAINT-GEORGES, I., VERSHAEVE, L., & LEONARD, A. (1984) The inhibitory effect of methylmercury chloride and mercuric chloride on microtubule polymerization in vitro. C.R. Séances Soc. Biol. (Paris), 178: 562-566. DEN TONKELAAR, E.M., VAN ESCH, G.J., HOFMAN, B., SCHULLER, P.L., & ZWIERS, J.H.L. (1974) Mercury and other elements in blood of the Dutch population. In: Proceedings of an International Symposium on Recent Advances in the Assessment of the Health Effects of Environmen- tal Pollution, Paris, 24-28 June, Luxembourg, Commission of the European Communities, Vol. 2, pp. 1017-1027. DOHERTY, R.A., GATES, A.H., & LANDRY, T. (1977) Methylmercury excretion: developmental changes in mouse and man. Pediatr. Res., 11: 416. DOI, R. & KOBAYASHI, T. (1982) Organ distribution and biological half- time of methylmercury in four strains of mice. Jpn. J. exp. Med., 52: 307-314. DOI, R. & TAGAWA, M. (1983) A study on the biochemical behavior of methylmercury. Toxicol. appl. Pharmacol., 69: 407-416. ECCLES, C.U. & ANNAU, Z. (1982a) Prenatal methylmercury exposure: alterations in neonatal activity. Neurobehav. Toxicol., 4: 371-376. ECCLES, C.U. & ANNAU, Z. (1982b) Prenatal methylmercury exposure. II. Alterations in learning and psychotropic drug sensitivity in adult offspring. Neurobehav. Toxicol. Teratol., 4: 377-382. ELSNER, J., HODEL, B., SUTER, K.E., OEKLKE, D., ULBRICH, B., SCHREINER, G., CUOMO, V., CAGIONO, R.A., ROSENGREN, L.E., KARLSSON, J.E., & HAGLID, K.G. (1988) Detection limits of different approaches in behavioral teratology, and correlation of effects with neurochemical parameters. Neurotoxicol. Teratol., 10: 155-167. EVANS, H.L., GARMAN, R.H., & WEISS, B. (1977) Methylmercury: exposure duration and regional distribution as determinants of neurotoxicity in non-human primates. Toxicol. appl. Pharmacol., 41: 15-33. FAIR, P.H., BALTHROP, J.E., WADE, J.L., & BRADDON-GALLOWAY, S. (1987). In vivo incorporation of 14C leucine into brain protein of mice treated with methylmercury and thiol complexes of methylmercury. Toxicol. Lett., 36: 213-220. FALK, S.A., KLEIN, R., HASEMAN, J.K., SANDERS, G.M., & TALLEY, F.A. (1974) Acute methylmercury intoxication and ototoxicity in guinea pigs. Arch. Pathol., 97: 297-305. FANG, S.C. (1980) Comparative study of uptake and tissue distribution of methylmercury in female rats by inhalation and oral routes of administration. Bull. environ. Contam. Toxicol., 24: 65-72. FARANT, J.P., BRISSETTE, D., MONCION, L., BIGRAS, L., & CHARTRAND, A. (1981) Improved cold-vapor atomic absorption technique for the microdetermination of total and inorganic mercury in biological samples. J. anal. Toxicol., 5: 47-51. FORHAMMER, M., ALBANUS, L., BRUCE, A., MATTSON, P., & OHLIN, B. (1984) Fish consumption and mercury in hair in pregnant women. Var Föda, 36: 2-13. FOUASSIN, A. & FONDU, M. (1978) Evaluation de la teneur moyenne en mercure de la ration alimentaire en Belgique. Arch. Belg. Med. Soc. Hyg. Med Trav. Med Leg., 36: 481-490. FRIBERG, L. (1983) Quality control in laboratories testing for environmental pollution. In: Clarkson, T.W., Nordberg, G.F., & Sager, P.R., ed. Reproductive and developmental toxicity of metals, New York, London, Plenum Press, pp. 811-829. FRIBERG, L. (1988) Quality assurance. In: Clarkson, T.W., Friberg, L., Nordberg, G.F., & Sager, P.R., eds. Biological monitoring of toxic metals, New York, London, Plenum Press, pp. 103-126. FRIBERG, L., KULLMAN, L., LIND, B., & NYLANDER, M. (1986) Mercury in the central nervous system in relation to amalgam fillings. Lakartidningen, 83: 519-522. FUJIKI, M. & TAJUMA, S. (1975) Studies on the methylmercury formation in the environment. In: Studies on the health effects of alkylmercury in Japan, Tokyo, Environment Agency, pp. 11-19. FURUTANI, A. & RUDD, J.W. (1980) Measurement of mercury methylation in lake water and sediment samples. Appl. environ. Microbiol., 40: 770- 776. GALSTER, W.A. (1976) Mercury in Alaskan Eskimo mothers and infants. Environ. Health Perspect., 15: 135-140. GANSER, A.L. & KIRSCHNER, D.A. (1985) The interaction of mercurials with myelin: comparison of in vitro and in vivo effects. Neurotoxicology, 6(1): 63-77. GANTHER, H.E. (1978) Modification of methylmercury toxicity and metabolism by selenium and vitamin E: possible mechanisms. Environ. Health Perspect., 25: 71-76. GANTHER, H.E. (1979) Metabolism of hydrogen selenide and methylated selenide. Adv. Nutr. Rev., 2: 107-128. GANTHER, H.E., GOUDIE, C., SUNDE, M.L., KOPECKY, M.J., WAGNER, P., OH, S.H., & HOEKSTRA, W.G. (1972) Selenium: relation to decreased toxicity of methylmercury added to diets containing tuna. Science, 175: 1122-1124. GARTRELL, M.J., CRAUN, J.C., PODREBARAC, D.S., & GUNDERSON, E.L. (1985a) Pesticides, selected elements and other chemicals in adult total diet samples: October 1978-September 1979. J. Chem. Assoc. Off. Anal. Chem., 68: 862-875. GARTRELL, M.J., CRAUN, J.C., PODREBARAC, D.S., & GUNDERSON, E.L. (1985b) Pesticides, selected elements and other chemicals in adult total diet samples: October 1979-September 1980. J. Chem. Assoc. Off. Anal. Chem., 68: 1184-1197. GARTRELL, M.J., CRAUN, J.C., PODREBARAC, D.S., & GUNDERSON, E.L. (1986) Pesticides, selected elements and other chemicals in adult total diet samples: October 1980-March 1982. J. Chem. Assoc. Off. Anal. Chem., 68: 146-161. GATES, A.H., DOHERTY, R.A., & COX, C. (1986) Reproduction and growth following prenatal methylmercuric chloride exposure in mice. Fundam. appl. Toxicol., 7: 486-493. GESAMP (1986) Review of potentially harmful substances: arsenic, mercury and selenium, Geneva. IMO/FAO/UNESCO/WMO/WHO/IAEA/ UN/UNEP Joint Group of Experts on the Scientific Aspects of Marine Pollution, World Health Organization (Reports and Studies No. 28). GILBERT, M.M., SPRECHER, J., CHANG, L.W., & MEISNER, L.F. (1983) Protective effect of vitamin E on genotoxicity of methylmercury. J. Toxicol. environ. Health, 12: 767-773. GIOVANOLI-JAKUBCZAK, T. & BERG, G. (1974) Measurement of mercury in human hair. Arch. Environ. Health, 28: 139-144. GONZALEZ, M.J., RICO, M.C., HERNANDEZ, L.M., & BALUJA, G. (1985) Mercury in human hair: a study of residents in Madrid, Spain. Arch. environ. Health, 40: 225-228. GOWDY, J.M., YATES, R., & DEMERS, F.X. (1977) Blood mercury concentration in an urban population. Sci. total Environ., 8: 247-251. GREENWOOD, M.R. (1985) Methylmercury poisoning in Iraq. An epidemiological study of the 1971-72 outbreak. J. appl. Toxicol., 5(3): 148-159. GREENWOOD, M.R., CLARKSON, T.W., DOHERTY, R.A., GATES, A.H., AMIN-ZAKI, L., ELHASSANI, S., & MAJEED, M.A. (1978) Blood clearance half-times in lactating and nonlactating members of a population exposed to methylmercury. Environ. Res., 16: 48-54. GREGUS, Z. & VARGA, F. (1985) Role of glutathione and hepatic glutathione S-transferase in the biliary excretion of methylmercury, cadmium, and zinc: a study with enzyme inducers and glutathione depletors. Acta pharmacol. toxicol., 56: 398-403. GRUENWEDEL, D.W. & DIAHAM, B. (1982) Methylmercury(II)-induced histone perturbations in nuclei isolated from calf thymus. Mol. Pharmacol., 22: 121-126. GRUNDT, I.K. & NESKOVIC, N.M. (1985) UDP-Galactose: ceramide galactosyltransferase and 2',3'-cyclic-nucleotide-3'-phosphodiesterase activities in rat brain after long-term exposure to methylmercury or triethyl-lead. Exp. Neurol., 88(3): 580-589. GRUNDT, I.K. & BAKKEN, A.M. (1986) Adenine nucleotides in cultured brain cells after exposure to methylmercury and triethyl lead. Acta pharmacol. toxicol., 59: 11-16. GRUNDT, I.K., AMMITZBOLL, T., & CLAUSEN, J. (1981) Triethyl-lead treatments of cultured brain cells. Effect on accumulation of radioactive precursors in callactolipids. Neurochem. Res., 6: 193-201. GRUNDT, I.K., ROUX, F., TREICH, I., LORIETTE, C., RAULIN, J., & FOURNIER, E. (1982) Effects of methylmercury and triethyllead on Na+K+ATPase and Pyruvate dehydrogenase activities in glioma C6 cells. Acta pharmacol. toxicol., 51: 6-11. GUNDERSON, V.M., GRANT, K.S., BURBACHER, T.M., FAGAN, J.F., & MOTTET, N.K. (1986) The effect of low-level prenatal methylmercury exposure on visual recognition memory of infant carb-eating macaques. Child Dev., 57: 1076-1083. HANSEN, J.C., KROMANN, N., WULF, H.C., & ALBOGE, K. (1984) Selenium and its interrelation with mercury in whole blood and hair in an east Greenlandic population. Sci. total Environ., 38: 33-40. HARADA, M., FUJINO, T., AKAGI, T., & NISHIGAKI, S. (1976) Epidemiological and clinical study and historical background of mercury pollution on Indian reservations in northwestern Ontario. Canada Bull. Inst. Const. Med., 26: 169-184. HARADA, M., FUJINO, T., & KABASHIMA, K. (1977) A study of methylmercury concentration in the umbilical cords of the inhabitants born in the Minamata area. Brain Dev., 9: 79-84 (In Japanese). HARADA, Y. (1968) Congenital (or fetal) Minamata Bay disease. In: Minamata disease, Kumamoto, Study Group of Minamata Disease, Kumamoto University. HARADA, Y. & MORIYAMA, H. (1977) [Study on the motor ability of school children in the Minamata district.] Environ. Health Rep., 40: 97-99 (in Japanese). HARGREAVES, R.J., MOORHOUSE, S.R., GANGOLI, S.D., & PELLING, D. (1986) The effects of methylmercury on glucose transport, glucose metabolism and blood flow in the central nervous system of the rat. In: Suckling, A.J., Rumsbury, M.G., & Bradbury, M.W.B., ed. The blood-brain barrier in health and disease, Chichester, Ellis Horwood, pp. 165-176. HARPER, K., BURNS, R., & ERICKSON, R.P. (1981) Genetic aspects of the effects of methylmercury in mice: the incidence of cleft palate and concentrations of adenosine 3':5' cyclic monophosphate in tongue and palatal shelf. Teratology, 23: 397-401. HATCH, R.W. & OTT, W.L. (1969) Determination of sub-microgram quantities of mercury by atomic absorption spectrophotometry. Anal. Chem., 40: 2085-2087. HAXTON, J., LINDSAY, D.G., HISLOP, J.S., SALMON, L., DIXON, E.J., EVANS, W.H., REID, J.R., HEWITT, C.J., & JEFFRIES, D.F. (1979) Duplicate diet study on fishing communities in the United Kingdom: mercury exposure in a "critical group". Environ. Res., 18: 351-368. HERIGSTAD, R.R., WHITEHAIR, C.K., BEYER, N., MICKELSEN, O., & ZABIK, M.J. (1972) Chronic methylmercury toxicosis in calves. J. Am. Vet. Med. Assoc., 160: 173-182. HIRANO, M., MITSUMORI, K., MAITA, K., & SHIRASU, Y. (1986) Further carcinogenicity study on methylmercury chloride in ICR mice. Nippon Juigaku Zasshi, 48: 127-135. HIRAYAMA, K. (1980) Effect of amino acids on brain uptake of methylmercury. Toxicol. appl. Pharmacol., 55: 318-323. HIRAYAMA, K. (1985) Effects of combined administration of thiol compounds and methylmercury chloride on mercury distribution in rats. Biochem. Pharmacol., 34: 3030-3032. HIRAYAMA, K. & YASUTAKE, A. (1986) Sex and age differences in mercury distribution and excretion in methylmercury-administered mice. J. Toxicol. environ. Health, 18: 49-60. HIROTA, Y., YAMAGUCHI, S., SHIMOJOH, N., & SANO, K. (1980) Inhibitory effect of methylmercury on the activity of glutathione peroxidase. Toxicol. appl. Pharmacol., 53: 174-176. HISLOP, J.S., COLLIER, T.R., WHITE, G.F., KHATHING, D.T., & FRENCH, E. (1983) The use of keratinised tissues to monitor the detailed exposure of man to methyl mercury from fish. In: Brown, S.S. & Savory, J., ed. Chemical toxicology and clinical chemistry of metals, New York, Academic Press, pp. 145-148. HORVAT, K., MAY, M., STOEPPLER, A., & BYRNE, R. (1988) Comparative studies of methylmercury determination in biological and environmental samples. Appl. organomet. Chem., 2: 515-524. HORVAT, M., STEGNAR, A., BYRNE, R., DERMELJ, M., & BRANICA, A. (1988) A study of trace elements in human placenta, blood and hair from the Yugoslav central Adriatic. In: Braetter, P. & Schramel, P., ed. Trace elements-analytical chemistry in medicine and biology, Berlin, W. de Gruyter & Co., pp. 243-250. HOSKINS, B.B. & HUPP, E.W. (1978) Methylmercury effects in rat, hamster, and squirrel monkey. Environ. Res., 15: 5-19. HOWARD, J.D. & MOTTET, N.K. (1986) Effects of methylmercury on the morphogenesis of the rat cerebellum. Teratology, 34: 89-95. HSIEH, H.S. & GANTHER, H.E. (1975) Acid-volatile selenium formation catalyzed by glutathione reductase. Biochemistry 14(8): 1632-1636. HULTBERG, H. & HASSELROT, B. (1981) Mercury in the ecosystem. In: Project coal, health and environment, Vallingby, Swedish State Power Board, pp. 33-51. HUNTER, D. & RUSSELL, D.S. (1954) Focal cerebral and cerebellar atrophy in a human subject due to organic mercury compounds. J. Neurol. Neurosurg. Psychiatr., 17: 235-241. IMURA, N., MIURA, K., INOKAWA, M., & NAKADA, S. (1980) Mechanism of methylmercury cytotoxicity: by biochemical and morphological experiments using cultured cells. Toxicology, 17: 241- 254. INOUYE, M., KAJIWARA, Y., & HIRAYAMA, K. (1986) Dose and sex-dependent alterations in mercury distribution in fetal mice following methylmercury exposure. J. Toxicol. environ. Health, 19: 425- 435. INSKIP, M.J. & PIOTROWSKI, J.K. (1985) Review of the health effects of methylmercury. J. appl. Toxicol., 5: 113-133. IRPTC (1987) IRPTC Legal file 1986, Geneva, International Register of Potentially Toxic Chemicals, United Nations Environment Programme, Vol. II. IWATA, H., OKAMOTO, H., & OHSAWA, Y. (1973) Effect of selenium on methylmercury poisoning. Res. Commun. chem. Pathol. Pharmacol., 5: 673. JACOBS, A.J., MANISCALCO, W.M., & FINKELSTEIN, J.N. (1986) Effects of methylmercuric chloride, cyclohexamide and colchicine on the reaggregation of dissociated mouse cerebellar cells. Toxicol. appl. Pharmacol., 86: 362-371. JACOBS, J.M., CARMICHAEL, N., & CAVANAGH, J.B. (1977) Ultrastructural changes in the nervous system of rabbits poisoned with methylmercury. Toxicol. appl. Pharmacol., 39: 249-261. JAKLEVIC, J.M., FRENCH, W.R., CLARKSON, T.W., & GREENWOOD, M.R. (1978) X-ray fluorescence analysis applied to small samples. In: Barrett, C.S., Leyden, D.E., Newkirk, J.B., & Rudd, C.O., ed. Advances in X-ray analysis, New York, London, Plenum Press, Vol. 21, pp. 171- 185. JERNELOV, A. (1972) Mercury and food chains. In: Hartung, R. & Binman, B.D., ed. Environmental mercury contamination, Ann Arbor, Michigan, Ann Arbor Science Publishers, pp. 174-177. JOHANNESSON, T., LUND, G., & STEINNES, E. (1981) Mercury, arsenic, cadmium, selenium and zinc in human hair and salmon fries in Iceland. Acta pharmacol. toxicol., 48: 185-189. JOHNSON, D.C. & BRAMAN, R.S. (1974) Distribution of atmospheric mercury species near ground. Environ. Sci. Technol., 8: 1003-1009. JUNGHANS, R.P. (1983) A review of the toxicity of methylmercury compounds with application to occupational exposures associated with laboratory uses. Environ. Res. 31: 1-31. KASUYA, M. (1980) The effect of methylcobalamin on the toxicity of methylmercury and mercuric chloride on nervous tissue in culture. Toxicol. Lett., 7: 87-93. KATO, I., AOYAGI, M., SATO, Y., MIZUKOSHI, K., & KAWASAKI, T. (1981) Electrooculographic evaluation of methylmercury intoxication in monkeys. Exp. Neurol., 72: 51-62. KAWADA, J., NISHIDA, M., YOSHIMURA, Y., & MITANI, K. (1980) Effects of organic and inorganic mercurials on thyroidal functions. J. Pharm. Dyn., 3: 149-159. KAWASAKI, Y., IKEDA, Y., YAMAMOTO, T., & IKEDA, K. (1986) Long-term toxicity study of methylmercury chloride in monkeys. J. Food Hyg. Soc. Jpn, 27(5): 528-552. KAZANTZIS, G., AL-MUFTI, A.W., AL-JAWAD, A., AL-SHAHWANI, Y., MAJID, M.A., MAHMOUD, R.M., SOUFI, M., TAWFIQ, K., IBRAHIM, M.A., & DABAGH, H. (1976a) Epidemiology of organomercury poisoning in Iraq: II. Relationship of mercury levels in blood and hair to exposure and to clinical findings. Bull. World Health Organ., 53(Suppl.): 37-48. KAZANTZIS, G., AL-MUFTI, A.W., COPPLESTONE, J.F., MAJID, M.A., & MAHMOUD, R.M. (1976b) Epidemiology of organomercury poisoning in Iraq: III. Clinical features and their changes with time. Bull. World Health Organ., 53(Suppl.): 49-60. KELMAN, B.J., WALTER, B.K., & SASSER, L.B. (1982) Fetal distribution of mercury following introduction of methylmercury into porcine maternal circulation. J. Toxicol. environ. Health, 10: 191-200. KELMAN, B.J., STEINMETZ ,S.E., WALTER, B.K., & SASSER, L.B. (1980) Absorption of methylmercury by the fetal guinea pig during mid to late gestation. Teratology, 21: 161-165. KERSHAW, T.G., CLARKSON, T.W., & DHAHIR, P.H. (1980) The relationship between blood levels and dose of methylmercury in man. Arch. environ. Health, 35(1): 28-36. KHM (1981) Summary situation report. In: Project coal, health and environment, Valingby, Swedish State Power Board. KJELLSTROM, T., REEVES, R., & MITCHELL, J. (1982) Comparison of mercury in hair with fish-eating habits of children in Auckland. Commun. Health Stud. (Adelaide), 6: 57-63. KJELLSTROM, T., KENNEDY, P., WALLIS, & MANTELL, C. (1986) Physical and mental development of children with prenatal exposure to mercury from fish. Stage 1. Preliminary tests at age 4, Solna, National Swedish Environmental Board, 96 pp (Report No. 3080). KJELLSTROM, T., KENNEDY, P., WALLIS, S., STEWART, A., FRIBERG, L., LIND, B., WUTHERSPOON, P., & MANTELL, C. (1989) Physical and mental development of children with prenatal exposure to mercury from fish. Stage 2. Interviews and psychological tests at age 6, Solna, National Swedish Environmental Board, 112 pp (Report no. 3642). KOBAYASHI, H., YUYAMA, A., MATSUSAKA, N., TAKENO, K., & YANAGIVA, I. (1979) Effects of methylmercury chloride on various cholinergic parameters in vitro. J. toxicol Sci., 4: 351-362. KOBAYASHI, H., YUYAMA, A., MATSUSUAKA, N., TAKENO, K., & YANAGIVA, I. (1981) Neuropharmacological effect of methylmercury in mice with special reference to the central cholinergic system. Jpn. J. Pharmacol., 31: 711-718. KOLLER, L.D. & ROAN, J.G. (1980) Effects of lead, cadmium and methylmercury on immunological memory. J. environ. Pathol. Toxicol., 4: 47-52. KOLLER, L.D., ROAN, J.G., & BRAUNER, J.A. (1980) Methylmercury: Effects of B-lymphocyte receptors and phagocytosis of macrophages. J. environ. Pathol. Toxicol., 3: 407-411. KOMSTA-SZUMSKA, E. & MILLER, D.R. (1984) A kinetic analysis of the interaction between methylmercury and selenium. Toxicology, 33: 247- 257. KOSTA, L. & BYRNE, R. (1969) Activation analysis for mercury in biological samples at nanogram level. Talanta, 16: 1297-1302. KUHNERT, P.M., KUHNERT, B.R., & ERHARD, P. (1981) Comparison of mercury levels in maternal blood, fetal cord blood, and placental tissues. Am. J. Obstet. Gynecol., 139: 209-213. KUTSCHER, C.L., SEMBRAT, M., KUTSCHER, C.S., & KUTSCHER, N.L. (1985) Effects of the high methylmercury dose used in the collaborative behavioral teratology study on brain anatomy. Neurobehav. Toxicol. Teratol., 7(6): 775-777. KYLE, J.H. & GHANI, N. (1982a) Methylmercury in human hair: a study of a Papua New Guinean population exposed to methylmercury through fish consumption. Arch. environ. Health, 37: 266-270. KYLE, J.H. & GHANI, N. (1982b) Elevated mercury levels in people from Lake Murray, Western Province. Papua New Guinea med. J., 25: 81-88. KYLE, J.H. & GHANI, N. (1983) Mercury concentrations in canned and fresh fish and its accumulation in a population of Port Moresby residents. Sci. total Environ., 26: 157-162. LAKOWICZ, J.R. & ANDERSON, C.J. (1980) Permeability of lipid bilayers to methylmercuric chloride: quantification by fluorescence quenching of a carbazole-labelled phospholipid. Chem.-biol. Interact., 30: 309-323. LAUWERYS, R., BUCHET, J.P., ROELD, H., & HUBERMONT, G. (1978) Placental transfer of lead, mercury, cadmium, and carbon monoxide in women. I. Comparison of the frequency distributions of the biological indices in maternal and umbilical cord blood. Environ. Res., 15: 278- 289. LEBLANC, R.M., JOLY, L.P., & PAIEMENT, J. (1984) pH-Dependent interaction between methylmercury chloride and some membrane phospholipids. Chem.-biol. Interact., 48: 237-241. LEE, I.D. & DIXON, R.L. (1975) Effects of mercury on spermato-genesis studied by velocity sedimentation, cell separation and serial mating. J. Pharmacol. exp. Ther., 194: 171-181. LEE, M., CHAN, K.S., SAIRENJI, E., & NIIKUNI, T. (1979) Effect of sodium selenite on methylmercury-induced cleft palate in the mouse. Environ. Res., 19: 39-48. LEVESQUE, P. & ATCHISON, W. (1988) Effect of alteration of nerve terminal Ca2+ regulation on increased spontaneous quantal release of acetylcholine by methylmercury. Toxicol. appl. Pharmacol., 94: 55-65. LIE, A., GUNDERSEN, N.M., & KORSGAARD, K.J. (1982) Mercury in urine-sex, age and geographic differences in a reference population. Scand. J. environ. Health, 8: 129-133. LIND, B., BIGRAS, L., CERNICHIARI, E., CLARKSON, T.W., FRIBERG, L., HELLMAN, M., KENNEDY, P., KIRKBRIDE, J., KJELLSTROM, T., & OHLIN, B. (1988a) Quality control of analyses of mercury in hair. Fresenius Z. anal. Chem., 332: 630-632. LIND, B., FRIBERG, L., & NYLANDER, M. (1988b) II. Demethylation of mercury in brain. J. Trace Elem. exp. Med., 1: 49-56. LINDBERG, S., STOKES, P., GOLDBERG, E., & WREN, C. (1987) Group report: Mercury. In: Hutchinson, T.W. & Meema, K.M., ed. Lead, mercury, cadmium and arsenic in the environment, New York, Chichester, Brisbane, Toronto, John Wiley & Sons, pp. 17-34. LINDQVIST, O., JERNELOV, A., JOHANSSON, K., & RODHE, R. (1984) Mercury in the Swedish environment: global and local sources, Solna, National Swedish Environment Protection Board, 105 pp (Report No. 1816). LOGDBERG, B., BRUN, A., BERLIN, M., & SCHUTZ, A. (1988) Congenital lead encephalopathy in monkeys. Acta neuropathol., 77: 120-127. LOK, E. (1983) The effect of weaning on blood, hair, fecal and urinary mercury after chronic ingestion of methylmercury chloride by infant monkeys. Toxicol. Lett., 15: 147-152. LOMMEL, A., KRUSE, H., & WASSERMANN, O. (1985) Organochloride and mercury in blood of a fish-eating population at the River Elbe in Schleswig-Hostein, FRG. Arch. Toxicol., 8(Suppl.): 264-268. MACFARLANE, D.E. (1981) The effects of methylmercury on platelets. Mol. Pharmacol., 19: 470-476. MCKEOWN-EYSSEN, G.E. & RUEDY, J. (1983a) Methylmercury exposure in northern Quebec. I. Neurological findings in adults. Am. J. Epidemiol., 118: 461-469. MCKEOWN-EYSSEN, G.E. & RUEDY, J. (1983b) Prevalence of neurological abnormality in Cree Indians exposed to methylmercury in northern Quebec. Clin. invest. Med., 6: 161-169. MCKEOWN-EYSSEN, G.E., RUEDY, J., & NEIMS, A. (1983) Methylmercury exposure in northern Quebec. II. Neurological findings in children. Am. J. Epidemiol., 118: 470-479. MCKONE, C.E., YOUNG, C.A., BACHE, C.A., & LISK, D.J. (1971) Rapid uptake of mercuric ion by goldfish. Environ. Sci. Technol., 5: 1138- 1139. MAGNAVAL, R. & BATTI, R. (1980) Protective effect of phospho-lipids in methylmercury inhibition of hydroxybutyrate dehydrogenase. Toxicol. Lett., 5: 353-356. MAGNUSSON, J. & RAMEL, C. (1986) Genetic variation in the susceptibility to mercury and other metal compounds in Drosophila melanogaster. Teratog. Carcinog. Mutagen., 6: 4. MAGOS, L. (1971) Selective atomic-absorption determination of inorganic mercury and methylmercury in undigested biological samples. Analyst, 96: 847-853. MAGOS, L. (1982) Neurotoxicity, anorexia and the preferential choice of antidote in methylmercury intoxicated rats. Neurobehav. Toxicol. Teratol., 4: 643-646. MAGOS, L. (1987) The absorption, distribution, and excretion of methyl mercury. In: Edccles, C.U. & Annau, Z., ed. The toxicity of methyl- mercury, Baltimore, John Hopkins University Press, pp. 24-44. MAGOS, L. & BUTLER, W.H. (1972) Cumulative effects of methylmercury dicyandiamide given orally to rats. Food Cosmet. Toxicol. 10: 513-517. MAGOS, L. & CLARKSON, T.W. (1972) Atomic absorption determination of total, inorganic and organic mercury in blood. J. Assoc. Off. Anal. Chem., 55: 966-971. MAGOS, L. & WEBB, M. (1977) The effect of selenium on the brain uptake of methylmercury. Arch. Toxicol., 38: 201-207. MAGOS, L., BAKIR, F., CLARKSON, T.W., AL-JAWAD, A.M., & AL-SOFFI, M.H. (1976) Tissue levels of mercury in autopsy specimens of liver and kidney. Bull. World Health Organ., 53: 93-96. MAGOS, L., WEBB, M., & HUDSON, A.R. (1979) Complex formation between selenium and methylmercury. Chem.-biol. Interact., 28: 201-207. MAGOS, L., PERISTIANIS, C., CLARKSON, T.W., BROWN, A., PRESTON, S., & SNOWDEN, R.T. (1981) Comparative study of the sensitivity of male and female rats to methylmercury. Arch. Toxicol., 48: 11-20. MAGOS, L., CLARKSON, T.W., & HUDSON, A.R. (1984) Differences in the effects of selenite and biological selenium on the chemical form and distribution of mercury after the simultaneous administration of HgCl2 and selenium to rats. J. Pharmacol. exp. Therap. 228(2): 478-483. MAGOS, L., BROWN, A.W., SPARROW, S., BAILEY, E., SNOWDEN, R.T., & SKIPP, W.R. (1985a) The comparative toxicology of ethyl- and methylmercury. Arch. Toxicol., 57: 260-267. MAGOS, L., CIKRT, M., & SNOWDEN, R. (1985b) The dependence of biliary methylmercury secretion on liver GSH and ligandin. Biochem. Pharmacol., 34: 301-305. MAILHES, J.B. (1983) Methylmercury effects on Syrian hamster metaphase II oocyte chromosomes. Environ. Mutagen., 5: 679-686. MARC (1981) Health effects of methylmercury, London, Monitoring and Assessment Research Centre, University of London, 82 pp (MARC Report Number 24). MARSH, D.O., MYERS, G.J., CLARKSON, T.W., AMIN-ZAKI, L., & TIKRITI, S. (1977) Fetal methylmercury poisoning. New data on clinical and toxicological aspects. Trans. Am. Neurol. Assoc., 102: 69-71. MARSH, D.O., MYERS, G.J., CLARKSON, T.W., AMIN-ZAKI, L., TIKRITI, S., & MAJEED, M.A. (1980) Fetal methylmercury poisoning: clinical and toxicological data on 29 cases. Ann. Neurol., 7: 348-355. MARSH, D.O., MYERS, G.J., CLARKSON, T.W., AMIN-ZAKI, L. TIKRITI, S, MAJEED, M.A., & DABBAGH, A.R. (1981) Dose-response relationship for human fetal exposure to methylmercury. Clin. Toxicol., 18: 1311-1318. MARSH, D.O., CLARKSON, T.W., COX, C., MYERS, G.J., AMIN-ZAKI, L., & AL-TIKRITI, S. (1987) Fetal methylmercury poisoning. Arch. Neurol., 44: 1017-1022. MASUKAWA, T., KITO, H., HAYASHI, M., & IWATA, H. (1982) Formation and possible role of bis(methylmercuric) selenide in rats treated with methylmercury and selenite. Biochem. Pharmacol., 31: 75-78. MATTHEWS, A.D. (1983) Mercury content of commercially important fish of the Seychelles, and hair mercury levels of a selected part of the population. Environ. Res., 30: 305-312. MATTSSON, J.L., MILLER, E., ALLIGOOD, J.P., KOERING, J.E., & LEVIN, S.G. (1981) Early effects of methylmercury on the visual evoked response of the dog. Neurotoxicity, 2: 499-514. MAY, K., STOEPPLER, M., & REISINGER, K. (1987) Studies in the ratio total mercury/methylmercury in the aquatic food chain. Toxicol. environ. Chem., 13: 153-159. MIETTINEN, J.K. (1973) Absorption and elimination of dietary mercury (Hg++) and methylmercury in man. In: Miller, M.W. & Clarkson, T.W., ed. Mercury, mercurials and mercaptans, Springfield, Illinois, Charles C. Thomas, pp. 233-246. MILLER, C.T., KREWSKI, D., & TRYPHONAS, L. (1985) Methylmercury-induced mitochondrial DNA synthesis in neural tissue of cats. Fundam. appl. Toxicol., 5(2): 251-264. MINAGAWA, K., SASAKI, T., TAKIZAWA, K., TAMURA, R., & OSHINA, T. (1980) Accumulation route and chemical form of mercury in mushroom species. Bull. environ. Contam. Toxicol., 25: 382-388. MITSUMORI, K., MAITA, M., SAITO, T., TSUDA, S., & SHIRASU, Y. (1981) Carcinogenicity of methylmercury chloride in ICR mice: preliminary note on renal carcinogenesis. Cancer Lett., 12: 305-310. MITSUMORI, K., MAITA, K., & SHIRASU, Y. (1984) Chronic toxicity of methylmercury chloride in rats: pathological study. Jpn. J. vet. Sci., 46: 549-557. MIURA, K. & IMURA, N. (1987) Mechanism of methylmercury cytotoxicity. Crit. Rev. in Toxicol., 18(3): 161-188. MIURA, K., SUZUKI, K., & IMURA, N. (1978) Effects of methylmercury on mitotic mouse glioma cells. Environ. Res., 17: 453-471. MIURA, K., INOKAWA, M., & IMURA, N. (1984) Effects of methylmercury and some metal ions on microtubule networks in mouse glioma cells and in vitro tubulin polymerization. Toxicol. appl. Pharmacol., 73: 218- 231. MOHAMED, M.K., EVANS, T.C., MOTTET, N.K., & BURBACHER, T.M. (1986a) effects of methylmercury on sperm oxygen consumption. Acta pharmacol. toxicol., 58: 219-224. MOHAMED, M.K., LEE, W.I., MOTTET, N.K., & BURBACHER, T.M. (1986b) Laser ligh-scattering study of the toxic effects of methylmercury on sperm motility. J. Androl., 5(1): 11-15. MORIMOTO, K., IIJIMA, S., & KOIZUMI, A. (1982) Selenite prevents the induction of sister-chromatid exchanges by methylmercury and mercuric chloride in human whole-blood cultures. Mut. Res., 102: 183-192. MOTTET, N.K., SHAW, C.M., & BURBACHER, T.M. (1987) The pathological lesions of methyl mercury intoxications in monkeys. In: Eccles, C.U. & Annau, Z., ed. The toxicity of methyl mercury, Baltimore, John Hopkins University Press, pp. 73-103. MULDER, K.M. & KOSTYNIAK, P.J. (1985a) Effects of L-(alpha) S,5S)- alpha-3-chloro-4,5-dihydro-5-isoxazoleacetic acid on urinary excretion of methylmercury in the mouse. J. Pharmacol. exp. Ther., 234: 156-160. MULDER, K.M. & KOSTYNIAK, P.J. (1985b) Involvement of glutathione in the enhanced renal excretion of methylmercury in CFW Swiss mice. Toxicol. appl. Pharmacol., 78: 451-457. MUNRO, I.C., NERA, E.A., CHARBONNEAU, S.M., JUNKINS, B., & ZAWIDZKA, Z. (1980) Chronic toxicity of methylmercury in the rat. J. environ. Pathol. Toxicol., 3: 347-447. MUSHAK, P. (1987) The quantitative measurement of methylmercury. In: Eccles, C.U. & Annau, Z., ed. The Toxicity of methyl mercury, Baltimore, Johns Hopkins University Press, pp. 1-12. MYKKANEN, H., RASANEN, L., AHOLA, M., & KIMPPA, S. (1986) Dietary intakes of mercury, lead, cadmium and arsenic by Finnish children. Hum. Nutr. appl. Nutr., 40: 32-39. NABRZYSKI, M. & GAJEWSKA, R. (1984) Determination of mercury, cadmium, and lead in food. Rocz. PZH, 35(1): 1-11 (in Polish). NAGANUMA, A. & IMURA, N. (1980) Bis(methylmercuric) selenide as a reaction product from methylmercury and selenite in rabbit blood. Res. Commun. chem. Pathol. Pharmacol., 27: 163-173. NAGANUMA, A. & IMURA, N. (1984) Species differences in biliary excretion of methylmercury. Biochem. Pharmcol., 33: 679-682. NAGANUMA, A., KOYAMA, Y., & IMURA, N. (1980) Behavior of methylmercury in mammalian erythrocytes. Toxicol. appl. Pharmacol., 54: 405-410. NAKADA, S. & IMURA, N. (1982) Uptake of methylmercury and inorganic mercury by mouse glioma and mouse neuroblastoma cells. Neurotoxicology 3: 249-258. NAKADA, S. & IMURA, N. (1983) Susceptibility of lipids to mercurials. J. appl. Toxicol., 3: 131-135. NAKADA, S., NOMOTO, A., & IMURA, N. (1980) Effect of methylmercury and inorganic mercury on protein synthesis in mammalian cells. Ecotoxicol. environ. Saf., 4: 184-190. NATIONAL ACADEMY OF SCIENCES (1978) An assessment of mercury in the environment, Washington, DC, National Academy of Sciences, National Research Council. NEWBERNE, P.M., GLASER, O., FRIEDMAN, L., & STILLINGS, B.R. (1972) Chronic exposure of rats to methyl mercury in fish protein. Nature (Lond.), 237: 40-41. NISHIKIDO, N., FURUYASHIKI, K., NAGANUMA, A., SUZUKI, T., & IMURA, M. (1987) Maternal selenium deficiency enhances the fetolethal toxicity of methylmercury. Toxicol. appl. Pharmacol., 88: 322-328. NISHIKIDO, N., SATOH, Y., NAGANUMA, A., & IMURA, N. (1988a) Effect of maternal selenium deficiency on the teratogenicity of methylmercury. Toxicol. Lett., 40: 153-157. NISHIKIDO, N., SATOH, Y., IWAI, I., ISHII, M., NAGANUMA, A., & IMURA, A. (1988b) Specific alteration of the form of selenium in fetal liver on maternal methylmercury treatment. Toxicol. appl. Pharmacol., 92(3): 497-499. NIXON, J.E., KOLLER, L.D., & EXON, J.H. (1979) Effect of methylmercury chloride on transplacental tumors induced by sodium nitrites and ethylurea in rats. J. Natl Cancer Inst., 63: 1057-1063. NOBUNAGA, T., SATOH, H., & SUZUKI, T. (1979) Effects of sodium selenite on methylmercury embryotoxicity and teratogenicity in mice. Toxicol. appl. Pharmacol., 47: 79-88. NORDBERG, G.F. & STRANGERT, P. (1976) Estimations of a dose-response curve for long-term exposure to methylmercury compounds in human beings taking into account variability of critical organ concentrations and biological half-times. A preliminary communication. In: Nordberg, G.F., ed. Effects and dose-response relationships, Amsterdam, Oxford, New York, Elsevier Science Publishers, pp. 273-282. NORDBERG, G.F. & STRANGERT, P. (1978) Fundamental aspects of dose- response relationships and their extrapolation for non-carcinogenic effects of metals. Environ. Health Perspect., 22: 97-108. NORDBERG, G.F. & STRANGERT, P. (1982) Risk estimation models derived from metabolic and damage parameter variation in the population. In: Vouk, V.B., Butler, G.C., Hoel, D.G., & Peakall, D.B., ed. Methods for estimating risk of chemical injury: Human and non-human biota and ecosystems, Chichester, West Sussex, John Wiley and Sons, pp. 477-491 (SGOMSEC 2). NORSETH, T. & CLARKSON, T.W. (1971) Intestinal transport of 203Hg- labeled methyl mercury chloride. Role of biotransformation in rats. Arch. environ. Health, 22: 668-577. NRIAGU, J.O. (1979) The biogeochemistry of mercury in the environment, Amsterdam, Oxford, New York, Elsevier Science Publishers. NYLANDER, N., FRIBERG, L., & LIND, B. (1987) Mercury concentration in the human brian and kidneys in relation to exposure from dental amalgams. Swed. dent. J., 11: 179-187. O'KUSKY, J.R., BOYES, B.E., & MCGEER, E.G. (1988) Methylmercury-induced movement and postural disorders in developing rat: regional analysis of brain catecholamines and indoleamines. Brain Res., 439(1/2): 138-146. OHI, G., NISHIGAKI, S., SEKI, H., TAMURA, Y., MAKI, T., MINOWA, K., SHIMAMURA, Y., MIZOGUCHI, I., INABA, Y., TAZIKAWA, Y., & KAWANISHI, Y. (1980) The protective potency of marine animal meat against the neurotoxicity of methylmercury: its relationship with the organ distribution of mercury and selenium. Food Cosmet. Toxicol., 18: 139-145. OHLANDER, E.H., OHLIN, B., ALBANUS, L., & BRUCE, A. (1985) Mercury levels in the hair of pregnant women eating Swedish fresh water fish. Var Föda, 8: 393-394. OHSAWA, M. & MAGOS, L. (1974) The chemical form of methylmercury complex in rat bile. Biochem. Pharmacol., 23: 1903-1906. OHSAWA, M., FUKUDA, K., & KAWAI, K. (1981) Accelerated accumulation of methylmercury in the rat fetus at the late pregnant stage. Ind. Health, 19: 219-221. OLSON, F.C. & MASSARO, E.J. (1980) Developmental pattern of cAMP, adenyl cyclase, and cAMP phosphdiesterase in the palate, lung, and liver of the fetal mouse: alterations resulting from exposure to methylmercury at levels inhibiting palate closure. Teratology, 22: 155-166. OMATA, S., SAKIMURA, K., ISHII, T., & SUGANO, H. (1978) Chemical nature of methylmercury complex with a low molecular weight in the liver cytosol of rats exposed to methylmercury chloride. Biochem. Pharmacol., 27: 1700-1701. OMATA, S., HORIGOME, T., MOMOSE, Y., KAMBAYASHI, M., MOCHIZUKI, M., & SUGANO, H. (1980) Effects of methylmercury chloride on the in vivo rate of protein synthesis in the brain of the rat: in vivo rate of protein synthesis in the brain of the rat: examination with the injection of a large quantity of 14C- valine. Toxicol. Appl. Pharmacol., 56: 207-215. OMATA, S., MOMOSE, Y., UEKI, H., & SUGANO, H. (1982) In vivo effect of methylmercury on protein synthesis in peripheral nervous tissues of the rat. Arch. Toxicol., 49: 203-214. ONFELT, A. (1983) Spindle disturbances in mammalian cells. I. Changes in the quantity of free sulfhydryl groups in relation to survival and C-mitosis in V79 Chinese hamster cells after treatment with colcemid, diamide, carbaryl and methylmercury. Chem.-biol. Interact., 46: 201- 217. ONFELT, A. & JENSSEN, D. (1982) Enhanced mutagenic response of MNU by post-treatment with methylmercury, caffeine or thymidine in V79 Chinese hamster cells. Mutat. Res., 106: 297-303. PACYNA, J.M. (1987) Atmospheric emissions of arsenic, cadmium, lead and mercury from high temperature processes in power generation and industry. In: Hutchinson, T.C. & Meema, K.M., ed. Lead, Mercury, Cadmium and Arsenic in the environment, New York, Chichester, Brisbane, Toronto, John Wiley & Sons, pp. 69-88 (SCOPE 31). PALLOTTI, G., BENCIVENGA, B., & SIMONETTI, T. (1979) Total mercury levels in whole blood, hair and fingernails for a population group from Rome and its surroundings. Sci. total Environ., 2: 69-72. PECKHAM, N.H. & CHOI, B.H. (1986) Surface change alterations in mouse fetal astrocytes due to methylmercury: an ultra-structural study with cationized ferritin. Exp. mol. Pathol., 44: 230-234. PETER, F. & STRUNC, G. (1984) Semiautomated analysis for mercury in whole blood, urine and hair by on-stream generation of cold vapor. Clin. Chem., 30(6): 893-895. PHELPS, R.W., CLARKSON, T.W., KERSHAW, T.G., & WHEATLEY, B. (1980) Interrelationships of blood and hair mercury concentrations in a North American population exposed to methylmercury. Arch. environ. Health, 35: 161-168. PIPER, R.C., MILLER, V.L., & DICKINSON, E.O. (1971) Toxicity and distribution of mercury in pigs with acute methylmercurialism. Am. J. vet. Res., 32: 263-273. PRASAD, K.N. & RAMANUJAM, S. (1980) Vitamin E and vitamin C alter the effect of methylmercuric chloride on neuroblastoma and glioma cells in culture. Environ. Res., 21: 343-349. PROHASKA, J.R. & GANTHER, H.E. (1977) Interactions between selenium and methylmercury in rat brain. Chem.-biol. Interact. 16: 155-167. QUANDT, F.N., KATO, E., & NARAHASHI, T. (1982) Effects of methylmercury on electrical responses of neuroblastoma cells. Neurotoxicology, 3: 205-220. RAMANUJAM, M. & PRASAD, K.N. (1979) Alterations in gene expression after chronic treatment of glioma cells in culture with methylmercuric chloride. Biochem. Pharmacol., 28: 2979-2984. RAMLAL, P.S., RUDD, J.W.M., FURUTANI, A., & XUN, L. (1985) The effect of pH on methylmercury production and decomposition in lake sediments. Can. J. Fish. aquat. Sci., 42: 685-692. RAMLAL, P.S., RUDD, J.W.M., & HECKY, R.E. (1986) Methods for measuring specific rates of mercury methylation and degradation and their uses in determining factors controlling net rates of mercury methylation. Appl. environ. Microbiol., 51: 110-114. REBEL, G., GUERIN, P., & PRASAD, K.N. (1983) Effect of methylmercuric chloride on gangliosides of mouse neuroblastoma cells in culture. Lipids, 18: 664-667. REFSVIK, T. (1984a) N-Acetylpenicillamine potentiated excretion of methylmercury in rat bile: influence of S-methyl-cysteine. Acta pharmacol. toxicol., 55: 121-125. REFSVIK, T. (1984b) N-Acetylpenicillamine potentiation of biliary excretion of methylmercury: influence of glutathione depletors. Acta pharmacol. toxicol., 55: 58-64. REFSVIK, T. & NORSETH, T. (1975) Methylmercuric compounds in rat bile. Acta pharmacol. toxicol., 36: 67-78. REUHL, K.R., CHANG, L.W., & TOWNSEND, J.W. (1981a) Pathological effects of in utero methylmercury exposure on the cerebellum of methylmercury exposure on the cerebellum of the golden hamster. I. Early effects upon the neonatal cerebellar cortex. Environ. Res., 26: 281-306. REUHL, K.R., CHANG, L.W., & TOWNSEND, J.W. (1981b) Pathological effects of in utero methylmercury exposure on the cerebellum of methylmercury exposure on the cerebellum of the golden hamster. II. Residual effects on the adult cerebellum. Environ. Res., 26: 307-327. RICHARDSON, R.J. & MURPHY, S.D. (1975) Effect of glutathione depletion on tissue deposition of methylmercury in rats. Toxicol. appl. Pharmacol., 31: 505-519. RIOLFATTI, M. (1977) [Further epidemiological studies on mercury concentrations in edible fish and in human blood and hair.] Ig. mod., 70: 169-186 (in Italian). RODIER, P.M. (1977) Correlation between prenatally-induced alterations in CNS cell populations and postnatal function. Teratology, 16: 235- 246. RODIER, P.M., ASHNER, M., & SAGER, P.R. (1984) Mitotic arrest in the developing CNS after prenatal exposure to methylmercury. Neurobehav. Toxicol. Teratol., 6: 379-385. RODIER, P.M. & KATES, B. (1988) Histological localization of methylmercury in mouse brain and kidney by emulsion autoradiography 203Hg. Toxicol. appl. Pharmacol., 92: 224-234. ROELS, H., HUBERMONT, G., BUCHET, J.P., & LAUWERYS, R. (1978) Placental transfer of lead, mercury, cadmium and carbon monoxide in women. III. Factors influencing the accumulation of heavy metals in the placenta and the relationship between metal concentration and the placenta and in maternal and cord blood. Environ. Res., 16: 236-247. ROWLAND, I.R., DAVIES, M.J., & EVANS, J.G. (1980) Tissue content of mercury in rats given methylmercuric chloride orally: influence of intestinal flora. Arch. environ. Health, 35: 155-160. ROWLAND, I.R., ROBINSON, R.D., DOHERTY, R.A., & LANDRY, T.D. (1983) Are developmental changes in methylmercury metabolism and excretion mediated by the intestinal microflora? In: Clarkson, T.W., Nordberg, G.F., & Sager, P.R., ed. Reproductive and developmental toxicity of metals, New York, London, Plenum Press, pp. 745-758. ROWLAND, I.R., ROBINSON, R.D., & DOHERTY, R.A. (1984) Effects of diet on mercury metabolism and excretion in mice given methylmercury: role of gut flora. Arch. environ. Health, 39: 401-418. ROWLAND, I.R., MALLETT, A.K., FLYNN, J., & HARGREAVES, R.J. (1986) The effect of various dietary fibres on tissue concentration and chemical form of mercury after methylmercury exposure in mice. Arch. Toxicol., 59: 94-98. RUDD, J.W.M., FURUTANI, A., & TURNER, M.A. (1980) Mercury methylation by fish intestinal contents. Appl. environ. Microbiol., 40: 777-782. RUSTAM, H., VON BURG, R., AMIN-ZAKI, L., & ELHASSANI, S. (1975) Evidence for a neuromuscular disorder in methylmercury poisoning. Arch. environ. Health, 30: 190-195. SAGER, P.R. (1988) Selectivity of methyl mercury effects on cytoskeleton and mitotic progression in cultured cells. Toxicol appl. Pharmacol., 94: 473-486. SAGER, P.R. & MATHESON, D. (1988) Mechanisms of neurotoxicity related to selective disruption of microtubules and intermediate filaments. Toxicology, 49(2-3): 479-492. SAGER, P.R., DOHERTY, R.A., & OLMSTEAD, J.B. (1981a) Interaction of methylmercury with microtubules. Toxicologist, 1: 118-119. SAGER, P.R., DOHERTY, R.A., & OLMSTEAD, J.B. (1981b) Interactions of methylmercury with microtubules. J. cell Biol., 91: 330a. SAGER, P.R., DOHERTY, R.A., & RODIER, P.M. (1982) Morphometric analysis of the effect of methylmercury on developing mouse cerebellar cortex. Toxicologist, 2: 116. SAGER, P.R., DOHERTY, R.A., & OLMSTEAD, J.B. (1983) Interaction of methylmercury with microtubules in cultured cells and in vitro. Exp. cell Res., 146: 127-137. SAGER, P.R., ASHNER, M., & RODIER, P.M. (1984) Persistent differential alterations in developing cerebellar cortex of male and female mice after methylmercury exposure. Brain Res., 12: 1-11. SARAFIAN, T. & VERITY, M.A. (1985) Inhibition of RNA and protein synthesis in isolated cerebellar cells by in vitro and in vivo methyl- mercury. Neurochem. Pathol., 3(1): 27-39. SARAFIAN, T. & VERITY, M.A. (1986) Mechanism of apparent transcription inhibition by methylmercury in cerebellar neurons. J. Neurochem., 47(2): 625-631. SARAFIAN, T.A., CHEUNG, M.K., & VERITY, M.A. (1984) In vitro methylmercury inhibition of protein synthesis in neonatal cerebellar perikarya. Neuropathol. appl. Neurobiol., 10: 85-100. SATOH, H., YASUDA, N., & SHIMAI, S. (1985) Development of reflexes in neonatal mice prenatally exposed to methylmercury and selenite. Toxicol. Lett., 25: 199-203. SCHROEDER, W.H. & JACKSON, R.A. (1987) Environmental measurements with an atmospheric mercury monitor having specific capabilties. Chemosphere, 16: 183-199. SECO, J.M. (1987) Big mercury deposit found in Spain. Financial Times, 30 October 1987. SHAMOO, A.E., MACLENNAN, D.H., & ELDEFRAWI, M.E. (1976) Differential effects of mercurial compounds on excitable tissues. Chem.-biol. Interact., 12: 41-52. SHERLOCK, J.C., LINDSAY, D.G., HISLOP, J.E., EVANS, W.H., & COLLIER, T.R. (1982) Duplication diet study on mercury intake by fish consumption in the United Kingdom. Arch. environ. Health, 37: 271-278. SHERLOCK, J.C. & QUINN, M. (1988) Underestimation of dose-response relationship with particular reference to the relationship between the dietary intake of mercury and its concentration in blood. Hum. Toxicol. 7: 129-132. SHERLOCK, J.C., HISLOP, J., NEWTON, D., TOPPING, G., & WHITTLE, K. (1984) Elevation of mercury in human blood from controlled chronic ingestion of methylmercury in fish. Hum. Toxicol., 3: 117-131. SHIMAI, S. & SATOH, H. (1985) Behavioral teratology of methylmercury. J. toxicol. Sci., 10: 199-216. SHIMOJO, N., YAMAGUCHI, S., & SANO-KENICHI (1976) Methylmercury derivatives in engine exhausts and its synthesis. Jpn. J. ind. Health, 18: 36-37. SIDMAN, R.L. & RAKIC, P. (1973) Neuronal migration, with special reference to developing human brain: a review. Brain Res., 63: 1-35. SIKORSKI, R., PASZKOWSKI, T., & SZPRENGIER-JUSZKIEWICZ, T. (1986) Mercury in neonatal scalp hair. Sci. total Environ., 57: 105- 110. SILVER, S. (1984) Bacterial transformations of and resistance to heavy metals. In: Nriagu, J.O., ed. Changing metal cycles and human health, Berlin, Heidelberg, New York, Springer-Verlag, pp. 199-225. SKERFVING, S. (1974) Methylmercury exposure, mercury levels in blood and hair, and health status in Swedes consuming contaminated fish. Toxicology, 2: 3-23. SLOTKIN, T.A., KAVLOCK, R.J., COWDERY, T., ORBAND, L., BARTOLOME, M., GRAY, J.A., REHNBERG, B.F., & BARTOLOME, J. (1986) Functional consequences of prenatal methylmercury exposure: effects on renal and hepatic repsonses to trophic stimuli and on renal excretory mechanisms. Toxicol. Lett., 34: 231-245. SLOTKIN, T.A., PACHMAN, S., KAVLOCK, R.J., & BARTOLOME, J. (1985) Effects of neonatal methylmercury exposure on development of nucleic acids and proteins in rat brain: regional specificity. Brain Res. Bull., 14: 397-400. SMITH, J.H., MCCORMACK, K.M., BRASELTON, W.E., Jr, & HOOK, J.B. (1983) The effect of prenatal methylmercury administration on postnatal renal functional development. Environ. Res., 30: 63-71. SMITH, R.G., VORWALD, A.J., PATIL, L.S., & MOONEY, T.F. (1970) Effects of exposure to mercury in the manufacture of chlorine. Am. Ind. Hyg. Assoc. J., 31: 687-700. SPUHLER, K. & PRASAD, K.N. (1980) Inhibition by methylmercuric chloride of prostaglandin E1-sensitive adenylate cyclase activity in glioma but not in neuroblastoma cells in culture. Biochem. Pharmacol., 29: 201- 203. SPYKER, J.M., SPARBER, S.B., & GOLDBERG, A.M. (1972) Subtle consequences of methylmercury exposure: behavioural deviations in offspring of treated mothers. Science, 177: 621-623. STEIN, A., GREGUS, Z., & KLAASSEN, C. (1988) Species variations in biliary excretion of glutathione-related thiols and methylmercury. Toxicol. appl. Pharmacol., 93: 351-359. STOKES, P.M. & WREN, C.D. (1987) Bioaccumulation of mercury by aquatic biota in hydroelectric reservoirs: a review and consideration of mechanisms. In: Hutchinson, T.W. & Meema, K.M., ed. Lead, mercury, cadmium and arsenic in the environment, New York, Chichester, Brisbane, Toronto, John Wiley & Sons, pp. 255-277. SUDA, I. & TAKAHASHI, H. (1986) Enhanced and inhibited bio-transfor- mation of methylmercury in the rat spleen. Toxicol. appl. Pharmacol., 82: 45-52. SUGIURA, Y., TAMI, Y., & TANAKA, H. (1978) Selenium protection against mercury toxicity: high binding affinity of methylmercury by selenium- containing ligands in comparison with sulfur-containing ligands. Bioinorg. Chem. 9(2): 167-180. SUZUKI, T., SHISHIDO, S., & URUSHIYAMA, K. (1976) Mercury in cigarettes. Tohoku J. exp. Med., 119: 353-356. SUZUKI, T., HONGO, T., & YAMAMOTO, R.A (1984a) Hair mercury levels of Japanese women during the period 1881-1968. J. appl. Toxicol. 4:101- 104. SUZUKI, T, YONEMOTO, J., SATOH, H., NAGANUMA, A., IMURA, N., & KIGAWA, T. (1984b) Normal organic and inorganic mercury levels in the human feto-placental system. J. appl. Toxicol., 4: 249-252. SUZUKI, T., WATANABE, S., HONGO, T., KAWABE, T., INAOKA, T., OHTSUKA, R., & AKIMICHI, T. (1988) Mercury in scalp hair of Papuans in the Fly Estuary, Papua New Guinea. Asia-Pacific J. public Health, 2: 39-47. SWEDISH ENVIRONMENTAL PROTECTION BOARD (1986) Mercury: Occurrence and turnover of mercury in the environment, Stockholm, National Environmental Protection Board (Mercury Report No. 3). SWEDISH EXPERT GROUP (1971) Methylmercury in fish. A toxicological- epidemiological evaluation of risks. Nord. Hyg. Tidskr., 4(Suppl.): 19-364. SYVERSEN, T.L.M. (1982) Changes in protein and RNA synthesis in rat brain neurons after a single dose of methylmercury. Toxicol. Lett., 10: 31-34. SZPRENGIER-JUSZKIEWICZ, T. (1988) [Evaluation of daily intake of mercury with food stuffs in Poland.] Bromatol. Chem. Toksykol., 21: 228-232 (in Polish). SZUCKI, B. & KURYS, H. (1982) [Mercury levels in the blood and hairs in the general population.] Rocz. PZH, 33(3): 143-148 (in Polish). TAKEUCHI, T. (1977) Pathology of fetal Minamata disease. The effect of methylmercury on intrauterine life of human being. Pediatrician, 6: 69-87. TAKEUCHI, T. & ETO, K. (1975) Minamata disease. Chronic occurrence from pathological viewpoints. In: Studies on the health effects of alkylmercury in Japan, Tokyo, Japan Environment Agency, pp. 28-62. TAMASHIRO, H., AKAGI, H., ARAKAKI, M., FUTATSUKA, M., & ROHT, L.H. (1984) Causes of death in Minamata disease: analysis of death certificates. Int. Arch. occup. environ. Health, 54: 135-146. TAMASHIRO, H., ARAKAKI, M., FUTATSUKA, M., & LEE, S. (1986) Methylmercury exposure and mortality in southern Japan: a close look at causes of death. J. Epidemiol. community Health, 40: 181-185. TAMASHIRO, H., FUKUTOMI, K., & LEE, E. (1987) Methylmercury exposure and mortality in Japan: A life table analysis. Arch. environ. Health, 42: 100-107. TAYARANI, I., LEFAUCONNIER, J., & BOURRE, J. (1988) The effect of mercurials on amino acid transport and rubidium uptake by isolated rat brain microvessels. Neurotoxicology, 8(4): 543-552. Taylor A, Marks V (1973) Measurement of urinary mercury excretion by atomic absorption in health and disease. Br J Ind Med 30: 293-296. THOMAS, D.J. & SMITH, J.C. (1979) Partial characterization of a low- molecular weight methylmercury complex in rat cerebrum. Toxicol. appl. Pharmacol., 47: 547-556. THOMAS, D.J. & SMITH, J.C. (1982) Effects of co-administered low- molecular-weight thiol compounds on short-term distribution of methyl- mercury in the rat. Toxicol. appl. Pharmacol., 62: 104-110. THOMAS, D.J. & SMITH, J.C. (1984) Effects of co-administered sodium selenite on short-term distribution of methylmercury in the rat. Environ. Res., 34: 287-294. THOMAS, D.J., FISHER, H.L., SUMLER, M.R., MARCUS, A.H., MUSHAK, P., & HALL, L.L. (1986) Sexual differences in the distribution and retention of organic and inorganic mercury in methylmercury-treated rats. Environ. Res., 41: 219-234. THROWER, S.J. & ANDREWARTHA, K.A. (1981) Glutathione peroxidase response in tissues of rats fed diets containing fish protein concentrate prepared from shark flesh of known mercury and selenium content. Bull. environ. Contam. Toxicol., 26: 77-84. TOLLEFSON, L. & CORDLE, F. (1986) Methylmercury in fish: a review of residue levels, fish consumption and regulatory action in the United States. Environ. Health Perspect., 68: 203-208. TSUBAKI, T. & IRUKAYAMA, K., ed. (1977) Minamata disease, Amsterdam, Oxford, New York, Elsevier Science Publishers. TSUBAKI, T. & TAKAHASHI, H. (1986) Recent advances in Minamata disease studies, Tokyo, Kodansha. TSUBAKI, T., HIROTA, K., SHIRAKAWA, K., KONDO, K., & SATO, T. (1978) Clinical, epidemiological and toxicological studies on methylmercury poisoning. In: Plaa, G.L. & Duncan, W.A.M., ed. Proceedings of the First International Congress on Toxicology, New York, London, San Francisco, Academic Press, pp. 339-357. TURNER, C.J., BHATNAGER, M.K., & YAMASHIRO, S. (1981) Ethanol potentiation of methyl mercury toxicity: a preliminary report. J. Toxicol. environ. Health, 7: 665-668. TURNER, M.A. & RUDD, J.W.M. (1983) The English-Wabigoon river system. III. Selenium in lake enclosures: its geochemistry, bioaccumulation and ability to reduce mercury bioaccumulation. Can. J. Fish. aquat. Sci., 40: 2228-2250. TURNER, M.D., KILPPER, R.W., SMITH, J.C., MARSH, D.O., & CLARKSON, T.W. (1975) Studies on volunteers consuming methylmercury in tuna fish. Clin. Res., 23: 2. TURNER, M.D., MARSH, D.O., SMITH, J.C., CLARKSON, T.W., INGLIS, J.B., RUBINO, E.C., CHIRIBOGA, J., & CHIRIBOGA, C.C. (1980) Methylmercury in populations eating large quantities of marine fish. Arch. environ. Health, 35: 367-378. URANO, T., NAGANUMA, A., & IMURA, N. (1988) Methylmercury- cysteinylglycine constitutes the main form of methylmercury in rat bile. Res. Commun. chem. Pathol. Pharmacol., 60: 197-210. US DEPARTMENT OF COMMERCE (1978) Report on the chance of US seafood consumers exceeding the current acceptable daily intake for mercury and on recommended regulatory controls, Washington, DC, US Department of Commerce, National Oceanic and Atmospheric Administration, National Marine Fisheries Service, pp. 1-198. US EPA (1980) Ambient water quality criteria for mercury, Washington, DC, US Environmental Protection Agency, Criteria and Standards Division (EPA-600/479-049). US EPA (1984) Mercury health effects update: health issue assessment, Washington, DC, US Environmental Protection Agency (EPA-600/8-84- 019F). VALCIUKAS, J.A., LEVIN, S.M., NICHOLSON, W.J., & SELIKOFF, I.J. (1986) Neurobehavioural assessment of Mohawk Indians for sub-clinical indications of methyl mercury neurotoxicity. Arch. environ. Health, 41: 269-272. VERSCHAEVE, L. & LEONARD, A. (1984) Dominant lethal test in female mice treated with methylmercury chloride. Mutat. Res., 136: 131-136. VERSCHAEVE, L., KIRSCH-VOLDERS, M., & SUSANNE, C. (1983) Mercury chloride- and methyl mercury chloride-induced inhibition of NOR activity. Teratog. Carcinog. Mutagen., 3: 447-456. VOGEL, D.G., MARGOLIS, R.L., & MOTTET, N.K. (1985) The effects of methylmercury binding to microtubules. Toxicol. appl. Pharmacol., 80: 473-486. VOGEL, D.G., RABINOVITCH, P.S., & MOTTET, N.K. (1986) Methyl-mercury effects on cell cycle kinetics. Cell Tissue Kinet., 19: 227-242. VON BURG, R. & LANDRY, T. (1976) Methylmercury and the skeletal muscle receptor. J. Pharm. Pharmacol., 28: 548-551. VON BURG, R., FARRIS, F., & SMITH, J.C. (1974) Determinations of methylmercury in blood by gas chromatography. J. Chromatogr., 97: 65- 70. WALSH, C.T. (1982) The influence of age on the gastrointestinal absorption of mercuric chloride and methylmercury chloride in the rat. Environ. Res., 27: 412-420. WASSICK, K.H. & YONOVITZ, A. (1985) Methylmercury ototoxicity in mice determined by auditory brain stem responses. Acta otolarynol. (Stockholm), 99: 35-45. WATANABE, H., SHIMOJO, N., SANO, K., & YAMAGUCHI, S. (1988) The distribution of total mercury in the brian after the lateral ventricular single injection of methylmercury and glutathione. Res. Commun. chem. Pathol. Pharmacol. 60(1): 57-69. WATANABE, T., SHIMADA, T., & ENDO, A. (1982) Effects of mercury compounds on ovulation and meiotic and mitotic chromosomes in female golden hamsters. Teratology, 25: 381-384. WELSH, S.O. (1979) The protective effect of vitamin E and N,N'- diphenyl-p-phenylenediamine (DPPD) against methyl mercury toxicity in the rat. J. Nutr., 109: 1673-1681. WESTOO, G. (1968) Determination of methylmercury salts in various kinds of biological material. Acta chem. Scand. 22: 2277-2280. WGMF (1980) Report on mercury in fish and fish products. Canberra Working Group on Mercury in Fish, Department of Primary Industry, p. 371. WHEATLEY, B. (1979) Methylmercury in Canada: exposure of Indian and Inuit residents to methylmercury in the Canadian environment, Ottawa, Department of National Health and Welfare, Medical Services Branch, p. 200. WHEATLEY, B., BARBEAU, A., CLARKSON, T.W., & LAPHAM, L. (1979) Methylmercury poisoning in Canadian Indians - the elusive diagnosis. Can. J. neurol. Sci., 6(4): 417-422. WHO (1976a) Conference on intoxication due to alkylmercury-treated seed, Baghdad, Iraq, 9-13 September 1974, Bull. World Health Organ., 53(Suppl.): 105-112. WHO (1976b) Environmental Health Criteria 1: Mercury, Geneva, World Health Organization, 132 pp. WHO (1978) WHO Technical Report Series, No. 631 (Evaluation of certain food additives and contaminants. Twenty-second report of the Joint FAO/WHO Expert Committee on Food Additives). WHO (1980) Consultation to re-examine the WHO Environmental Health Criteria for mercury, Geneva, 21-25 April 1980, Geneva, World Health Organization (Unpublished report EHE/EHC/80.22). WHO (1989a) Environmental Health Criteria 86: Mercury - Environmental aspects, Geneva, World Health Organization, 150 pp. WHO (1989b) Evaluation of certain food additives and contaminants. Thirty-third report of the Joint FAO/WHO Expert Committee on Food Additives, Geneva, World Health Organization (WHO Technical Report Series 776). WIENER, J.G. (1987) Metal contamination of fish in low-pH lakes and potential implications for piscivorous wildlife. Trans. North Am. Wildl. Natl. Resour. Conf., 52: 645-657. WIGFIELD, D.C., CROTEAU, S.M., & PERKINS, S.L. (1981) Elimination of the matrix effect in the cold-vapor atomic absorption analysis of mercury in human hair samples. J. anal. Toxicol., 5: 52-55. WOOD, J.M. & WANG, H.K. (1983) Microbial resistance to heavy metals. Environ. Sci. Technol., 17: 82a-90a. WULF, H.C., KROMANN, N., KOUSGAARD, N., HANSEN, J.C., NIEBUHR, E., & ALBOGE, K. (1986) Sister chromatid exchange (SCE) in Greenlandic Eskimos. Dose-respone relationship between SCE and seal diet, smoking and blood cadmium and mercury concentrations. Sci. total Environ., 48: 81-94. XUN, L. & CAMPBELL, N.E.R. (1987) Measurements of specific rates of net methyl mercury production in the water column and surface sediments of acidified and circumneutral lakes. Can. J. Fish. aquat. Sci., 44: 750- 757. YAMAGUCHI, S., FUJIKI, M., SHIMOJO, N., KAKU, S., HIROTA, Y., MORI, Y., PHAS, B., & SANO, K. (1977) A background of geographical pathology on mercury in the east Pacific area. J. occup. Med., 19: 502- 504. YAMAMOTO, R., SUZUKI, T., SATOH, H., & KAWAI, K. (1986) Generation and dose as modifying factors of inorganic mercury accumulation in brain, liver and kidneys of rats fed methylmercury. Environ. Res., 41: 309-318. YAMINI, B. & SLEIGHT, S.D. (1984) Effects of ascorbic acid deficiency on methylmercury dicyandiamide toxicosis in guinea pigs. J. environ. Pathol. Toxicol. Oncol., 5: 139-150. YOSHINO, Y., MOZAI, T., & NAKAO, K. (1966) Biochemical changes in the brain in rats poisoned with an alkyl mercury compound, with special reference to the inhibition of protein synthesis in brain cortex slices. J. Neurochem., 13: 1223-1230. ZELENKO, V. & KOSTA, L. (1973) A new method for the isolation of methylmercury from biological tissue and its determination at the parts-per-million level by gas chromatography. Talanta, 20: 115-123. ZIMMER, L., JOHNSON, P.C., & CARTER, D.E. (1980) A comparison of biochemical, pathological or behavioral methods for the assessment of the effectiveness of treatment for methylmercury poisoning in rats. Neurobehav. Toxicol., 2: 323-330. APPENDIX Comparison between maximum hair levels of mercury during pregnancy and symptoms and signs in the mother and her offspringa ----------------------------------------------------------------------------------------------- Mother Child ------------------------- ---------------------------------------------------------- Symptoms Symptoms ----------------------- ------------------------------- Mother- Paraes- Other Mercury Sex Walked Talked Mental Seizures Neurological infant thesia level (month) (month) score number (mg/kg) ----------------------------------------------------------------------------------------------- 2 0 0 1 female 12 24 0 0 2 3 0 0 1 male - - 0 0 2 5 0 0 1 male 14 12 0 0 0 6 0 0 1 male 18 18 0 0 0 7 0 0 1 male 18 24 0 0 3 9 0 0 1 male 13 26 + 0 3 13 0 1 1 female 12 12 0 0 4 14 0 0 1 female 18 26 0 0 2 18 0 0 1 female 16 18 0 0 0 19 0 0 1 male 12 12 0 0 0 26 0 0 1 male 12 14 0 0 0 31 0 0 1 male 11 12 0 0 1 33 0 0 1 female 12 20 0 0 2 39 0 0 1 female - - 0 0 4 41 0 0 1 male 11 18 0 0 3 69 0 0 1 male 18 20 0 0 4 4 0 0 2 male 18 - 0 0 0 8 0 0 2 female 13 18 0 0 0 10 0 0 2 male 16 18 0 0 3 11 0 0 2 male 12 24 0 0 3 12 0 0 2 male 18 18 0 0 3 16 0 0 2 female 18 18 0 0 2 27 0 0 2 male 12 10 0 0 0 32 0 0 2 male 14 - 0 0 0 47 0 0 2 male 18 24 0 0 3 1 0 0 3 female 14 18 0 0 0 21 0 0 3 male 18 - 0 0 1 23 0 0 5 male 12 12 0 0 1 34 0 0 6 male 12 10 0 0 0 38 0 0 6 female 18 10 0 0 2 35 0 0 7 male 12 18 0 0 0 29 0 0 8 male 18 19 0 0 0 40 0 0 9 female 12 - 0 0 0 20 0 0 10 male 12 12 0 0 2 24 0 + 10 male 14 - 0 0 3 22 0 0 12 male 18 14 0 0 1 28 0 0 12 female 14 18 0 0 2 17 + + 14 female 20b 18 0 0 1 ----------------------------------------------------------------------------------------------- Appendix (contd). ----------------------------------------------------------------------------------------------- Mother Child ------------------------ -------------------------------------------------- Symptoms Symptoms ------------------------ ------------------------------- Mother- Paraes- Other Mercury Sex Walked Talked Mental Seizures Neurological infant thesia level (month) (month) score number (mg/kg) ----------------------------------------------------------------------------------------------- 36 0 0 16 male 18 16 0 0 5 42 0 0 18 female 36b 30b 0 0 0 25 0 0 19 male 12 18 0 0 2 30 + + 23 female 12 12 0 0 1 37 0 0 26 male 12 12 0 0 1 15 0 0 38 male 20b 18 0 0 6 49 0 0 45 male - - 0 0 4 59 0 0 46 male 12 12 0 0 2 53 0 0 52 female 18 18 0 0 1 48 + + 59 female 18 26b 0 0 0 43 0 0 60 male 20b 18 0 0 5 44 0 0 62 male 18 24 0 0 0 54 + + 74 male 18 26b 0 0 2 46 + 0 75 female 12 12 0 0 0 52 + 0 78 female 14 28b + + 4 50 0 0 86 male 11 26b 0 0 6 58 0 + 98 female 12 12 0 0 2 57 0 0 104 female 15 18 0 0 4 72 0 0 114 male 14 48b + 0 0 64 0 0 118 female 18 18 0 0 0 51 + + 154 male 24b 36b 0 + 3 76 0 0 196 male 12 15 0 0 0 77 0 0 202 male 24b 20 0 0 3 61 0 + 242 female 15 18 + 0 0 74 0 0 263 female 18 18 0 0 5 66 0 0 269 male 12 12 0 0 1 63 + 0 294 male 20b 34b + + 0 62 + + 336 male 24b 36b 0 + 0 67 + 0 339 female 20b 26b 0 0 4 60 + 0 357 female 20b 30b 0 0 2 65 + 0 362 male 14 16 0 0 4 71 + + 376 female 20b 30b 0 + 2 75 0 0 399 male 15 24 0 0 3 56 + + 404 male 60b 60b + + 11 70 + 0 405 female 60b 36b + 0 11 68 0 0 418 male 72b 72b + + 11 45 0 0 443 male 36b 30b + 0 11 73 0 0 468 female 14 20 0 0 6 ----------------------------------------------------------------------------------------------- Appendix (contd). ----------------------------------------------------------------------------------------------- Mother Child ----------------------- ---------------------------------------------- Symptoms Symptoms ----------------------- ------------------------------- Mother- Paraes- Other Mercury Sex Walked Talked Mental Seizures Neurological infant thesia level (month) (month) score number (mg/kg) ----------------------------------------------------------------------------------------------- 80 0 + 557 female 22b 22b 0 0 2 79 0 0 568 female 20b 26b 0 0 4 78 0 0 598 male 24b 23 0 0 7 81 + + 674 male 24b 26b 0 0 2 ----------------------------------------------------------------------------------------------- a From: Marsh et al. (1987) o = absence of abnormality. + = presence of abnormality. - = no observations were made. b Abnormal value. For further details, see section 126.96.36.199.