IPCS INCHEM Home


    INTERNATIONAL PREGRAMME ON CHEMICAL SAFETY


    ENVIRONMENTAL HEALTH CRITERIA 60





    PRINCIPLES AND METHODS FOR THE
    ASSESSMENT OF NEUROTOXICITY
    ASSOCIATED WITH EXPOSURE TO CHEMICALS




    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, 1986

        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 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.

    ISBN 92 4 154260 8

    (c) World Health Organization 1986

        Publications of the World Health Organization enjoy copyright
    protection in accordance with the provisions of Protocol 2 of the
    Universal Copyright Convention. For rights of reproduction or
    translation of WHO publications, in part or  in toto, application
    should be made to the Office of Publications, World Health
    Organization, Geneva, Switzerland. The World Health Organization
    welcomes such applications.

        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.

    ISSN 0250-863X

    CONTENTS

    PRINCIPLES AND METHODS FOR THE ASSESSMENT OF NEUROTOXICITY ASSOCIATED
    WITH EXPOSURE TO CHEMICALS

    PREFACE

    1. INTRODUCTION

         1.1. The importance of studying the nervous system
              1.1.1. Methylmercury
              1.1.2. Carbon disulfide (CS2)
         1.2. Need to establish a comprehensive strategy for neurotoxicity
              testing
         1.3. Scope of the book
         1.4. Purpose of the publication

    2. GENERAL PRINCIPLES FOR THE ASSESSMENT OF TOXIC EFFECTS OF
         CHEMICALS ON THE NERVOUS SYSTEM

         2.1. Factors to be considered in the design of neurotoxicity
              studies
              2.1.1. General considerations
              2.1.2. Objectives
              2.1.3. Choice of animals
              2.1.4. Dosing regimen
              2.1.5. Functional reserve and adaptation
              2.1.6. Other factors
         2.2. Statistical analysis
              2.2.1. Type I and Type II errors
              2.2.2. Selection of the appropriate statistical test(s)

    3. TEST METHODS IN BEHAVIOURAL TOXICOLOGY

         3.1. Introduction
         3.2. Classes of behaviour
              3.2.1. Respondent behaviour
              3.2.2. Operant behaviour
         3.3. Test methods
              3.3.1. General attributes of behavioural methods
                     3.3.1.1   Sensitivity and specificity
                     3.3.1.2   Validity
                     3.3.1.3   Replicability
                     3.3.1.4   Costs
              3.3.2. Primary tests
                     3.3.2.1   Functional observation battery
                     3.3.2.2   Motor activity
              3.3.3. Secondary tests
                     3.3.3.1   Intermittent schedules of reinforcement
                     3.3.3.2   Motor function

                     3.3.3.3   Sensory function
                     3.3.3.4   Cognitive function
                     3.3.3.5   Eating and drinking behaviour
                     3.3.3.6   Social behaviour
              3.3.4. Strengths and weaknesses of various methods
         3.4. Research needs
              3.4.1. Compensatory mechanisms
              3.4.2. Method development and refinement

    4. NEUROPHYSIOLOGICAL METHODS IN NEUROTOXICOLOGY

         4.1. Introduction
         4.2. Methods for evaluation of the peripheral nervous system
              4.2.1. Conduction velocity
              4.2.2. Peripheral nerve terminal function
              4.2.3. Electromyography (EMG)
              4.2.4. Spinal reflex excitability
         4.3. Methods for evaluation of the autonomic nervous system
              4.3.1. Electrocardiography (EKG)
              4.3.2. Blood pressure
         4.4. Methods for the evaluation of the central nervous system
              4.4.1. Spontaneous activity - electroencephalography (EEG)
              4.4.2. Sensory systems
              4.4.3. General excitability
                     4.4.3.1   Convulsive phenomena
                     4.4.3.2   Stimulation of the cerebral motor cortex
                     4.4.3.3   Recovery functions
              4.4.4. Cognitive function
              4.4.5. Synaptic and membrane activity
         4.5. Interpretation issues
         4.6. Summary and conclusions

    5. MORPHOLOGICAL METHODS

         5.1. Introduction
              5.1.1. Role of morphology
              5.1.2. Basis for morphological assessment
         5.2. The nervous system and toxic injuries
              5.2.1. The nervous system
              5.2.2. Cellular structure of the nervous system
              5.2.3. Neurocellular reaction to injury
                     5.2.3.1   Biological principles
                     5.2.3.2   Neurons
                     5.2.3.3   Myelinating cells
         5.3. Experimental design and execution
              5.3.1. General principles and procedure
              5.3.2. Gross morphology
              5.3.3. The role of histology
                     5.3.3.1   Biological principles dictating tissue
                               response

              5.3.4. Use of controls
              5.3.5. Pattern of response
              5.3.6. Data acquired
         5.4. Principles, limitations, and pitfalls of the morphological
              approach
              5.4.1. Tissue state
              5.4.2. Principles of fixation
              5.4.3. Principles of tissue sampling
              5.4.4. Preparation of tissue for examination
              5.4.5. Recognition of artefact
              5.4.6. Recognition of normal structural variations
              5.4.7. Qualitative versus quantitative approaches
         5.5. Specific procedures
              5.5.1. Introduction
              5.5.2. Primary methods
                     5.5.2.1   Formaldehyde/paraffin method
                     5.5.2.2   Glutaraldehyde/epoxy method
              5.5.3. Special methods
                     5.5.3.1   Peripheral nerve microdissection
                     5.5.3.2   Frozen sections
                     5.5.3.3   Histochemical methods
                     5.5.3.4   Golgi method
                     5.5.3.5   Transmission electron microscopy
                     5.5.3.6   Other anatomical methods
         5.6. Conclusions

    6. BIOCHEMICAL AND NEUROENDOCRINOLOGICAL METHODS

         6.1. Introduction
         6.2. Fractionation methods
              6.2.1. Brain dissection
              6.2.2. Isolation of specific cell types
              6.2.3. Subcellular fractionation
         6.3. DNA, RNA, and protein synthesis
         6.4. Lipids, glycolipids, and glycoproteins
         6.5. Neurotransmitters
              6.5.1. Synthesis/degradation
              6.5.2. Transport/release
              6.5.3. Binding
              6.5.4. Ion channels
              6.5.5. Cyclic nucleotides
              6.5.6. Summary of nerve terminal function
         6.6. Energy metabolism
         6.7. Biochemical correlates of axonal degeneration
         6.8. Neuroendocrine assessments
              6.8.1. Anterior pituitary hormones
              6.8.2. Disruption of neuroendocrine function
                     6.8.2.1   Direct pituitary effects
                     6.8.2.2   Peripheral target effects

                     6.8.2.3   Disruption of hypothalamic control of
                               pituitary secretions
                     6.8.2.4   Other sites of action
              6.8.3. Sex differences
         6.9. Recommendations for future research

    7. CONCLUSIONS AND RECOMMENDATIONS

    REFERENCES
    
    WHO TASK GROUP ON PRINCIPLES AND METHODS FOR THE ASSESSMENT OF
    NEUROTOXICITY ASSOCIATED WITH EXPOSURE TO CHEMICALS

    Members

    b         Dr M.B. Abou-Donia, Department of Pharmacology, Duke
                   University Medical Center, Durham, North Carolina,
                   USA

    b,c       Dr W.K. Anger, Neurobehavioral Research Section,
                   Division of Biomedical and Behavioral Science,
                   National Institute for Occupational Safety and
                   Health, Cincinatti, Ohio, USA

    b,c       Dr G. Bignami, Section of Neurobehavioural
                   Pathophysiology, Laboratory of Organ and System
                   Pathophysiology, High Institute of Health, Rome,
                   Italy

    b         Dr T.J. Bonashevskaya, Sysin Institute of General and
                   Community Hygiene, Moscow, USSR

    b         Dr E. Bonilla, Institute of Clinical Research, Faculty
                   of Medicine, University of Zulia, Maracaibo,
                   Venezuela

    b,c       Professor J. Cavanagh, West Norwood, London, England
                   (Rapporteurb,c)

    b         Dr V.A. Colotla, National Autonomous University, City
                   University, Coyoacan, Mexico

    b         Dr I. Desi, Division of Toxicology, National Institute
                   of Hygiene, Budapest, Hungary

    a         Dr L. Di Giamberardino, Dpartement de Biologie,
                   Commissariat  l'Energie Atomique, Centre d'Etudes
                   Nuclares de Saclay, Gif-sur-Yvette, France

    b         Dr S. Frankova, Institute of Psychology, Czechoslovak
                   Academy of Sciences, Prague, Czechoslovakia

    b         Dr E. Frantik, Institute of Hygiene and Epidemiology,
                   Prague, Czechoslovakia

    c         Dr. I. Goto, Department of Neurology, Neurological
                   Institute, Faculty of Medicine, Kyushu University,
                   Higashiku, Fukuoka, Japan

    b,c       Professor Dr M. Hasan, Brain Research Centre, J.N.
                   Medical College, Aligharh Muslim University,
                   Aligharh, India

    b         Dr L. Hinkova, Institute of Hygiene and Occupational
                   Health, Sofia, Bulgaria

    a,b,c     Associate Professor Dr M. Horvath, Institute of Hygiene
                   and Epidemiology, Prague, Czechoslovakia
                   (Vice-Chairmanc)

    c         Professor A. Korczyn, Neurology Department, Tel Aviv
                   Medical Centre, Tel Aviv University, Tel Aviv,
                   Israel

    b         Dr N.N. Litvinov, Sysin Institute of General and
                   Community Hygiene, Moscow, USSR

    a,b       Professor R.V. Merkureva, Department of Medical
                   Biological Research, Sysin Institute of General and
                   Community Hygiene, Moscow, USSR
                   (Vice-Chairmana)

    a,b,c     Dr C. Mitchell, Laboratory of Behavioural and
                   Neurological Toxicology, National Institute of
                   Environmental Health Sciences, Research Triangle
                   Park, North Carolina, USA (Chairmana,b,c)

    b,c       Professor J.E. Murad, Medical School and Chairman, Al.
                   Ezequiel Dins, 275, Drugs Orientation Centre, Belo
                   Horizonte, Brazil

    b,c       Professor O.B. Osuntokun, Department of Medicine,
                   University of Ibadan, WHO Collaborating Centre for
                   Research and Training in Neurosciences, University
                   of Ibadan, Nigeria

    a,b,c     Dr L. Reiter, Division of Neurotoxicology (MD-74B), US
                   Environmental Protection Agency, Health Effects
                   Research Laboratory, Research Triangle Park, North
                   Carolina, USA

    b         Dr D.C. Rice, Toxicology Research Division, Food
                   Directorate, Health Protection Branch, Department
                   of National Health and Welfare, Tunney's Pasture,
                   Ottawa, Ontario, Canada

    b,c       Dr M. Rudnev, Kiev Research Institute of General and
                   Communal Hygiene, Kiev, USSR

    c         Dr M. Ruscak, Centre of Physiological Sciences, Slovak
                   Academy of Sciences, Bratislava, Czechoslovakia

    a         Dr H. Savolainen, Institute of Occupational Health,
                   Helsinki, Finland (Rapporteur)

    b,c       Professor V. Schreiber, Laboratory for Endodrinology,
                   Faculty of Medicine, Charles University, Prague,
                   Czechoslovakia

    b,c       Professor P. Spencer, Institute of Neurotoxicology,
                   Albert Einstein College of Medicine, New York, USA

    b         Dr T. Tanimura, Department of Anatomy, Kinki University
                   School of Medicine, Osaka, Japan (invited, but
                   could not attend)

    b,c       Dr H. Tilson, Laboratory of Behavioural and
                   Neurological Toxicology, National Institute of
                   Environmental Health Sciences, Research Triangle
                   Park, North Carolina, USA

    b,c       Dr L. Uphouse, Department of Biology, Texas Woman's
                   University, Denton, Texas, USA

    b,c       Dr T. Vergieva, Department of Toxicology, Institute of
                   Hygiene and Occupational Health, Sofia, Bulgaria

    c         Professor D. Warburton, Psychopharmacology Group, Early,
                   Reading, United Kingdom

    a,b       Dr G. Winneke, Institute of Environmental Hygiene,
                   University of Dsseldorf, Dsseldorf, Federal
                   Republic of Germany

    b         Dr V.G. Zilov, I.M. Sechnov Medical Institute,
                   Department of Normal Physiology, Moscow, USSR
                   (Vice-Chairmanb)

    Secretariat

    c         Dr G. Becking, International Programme on Chemical
                   Safety, World Health Organization, Research
                   Triangle Park, North Carolina, USA

    a         Dr C.L. Bolis, Neurosciences Programme, World Health
                   Organization, Geneva, Switzerland

    a         Dr A. David, Office of Occupational Health, World Health
                   Organization, Geneva, Switzerland

    a,b,c     Dr M. Draper, International Programme on Chemical
                   Safety, World Health Organization, Geneva,
                   Switzerland

    b         Dr M. Gounar, International Programme on Chemical
                   Safety, World Health Organization, Geneva,
                   Switzerland

    a         Mr E. Hellen, Occupational Safety and Health Branch,
                   International Labour Organisation, Geneva,
                   Switzerland

    a         Dr A.I. Koutcherenko, International Register for
                   Potentially Toxic Chemicals, United Nations
                   Environment Programme, Geneva, Switzerland

    a         Dr M. Mercier, International Programme on Chemical
                   Safety, World Health Organization, Geneva,
                   Switzerland

    b         Dr L.A. Moustafa, International Programme on Chemical
                   Safety, World Health Organization, Research
                   Triangle Park, North Carolina, USA

    a         Dr J. Parizek, International Programme on Chemical
                   Safety, World Health Organization, Geneva,
                   Switzerland (Secretary)

    a         Dr J. Purswell, Section on Hygiene, International Labour
                   Organisation, Geneva, Switzerland

    a         Dr C. Satkunananthan, International Programme on
                   Chemical Safety, World Health Organization, Geneva,
                   Switzerland (Temporary Adviser)

    a         Dr C. Xintaras, Office of Occupational Health, World
                   Health Organization, Geneva, Switzerland

              

    a  Preparatory consultation, Geneva, 14-16 September 1981.
    b  First Task Group meeting, Moscow, 1-7 June 1983.
    c  Second Task Group meeting, Prague, 17-21 September 1984.

    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.

                                     * * *

    PREFACE

        The need for generally accepted scientific principles and
    requirements in all areas of toxicology particularly applies to the
    newly developed field of neurotoxicology. Methods continue to be
    developed in isolation, and the comparability of results is often in
    doubt. Furthermore, until scientific principles have been agreed on,
    internationally accepted strategies to test the effects of chemicals
    on the many functions of the mammalian nervous system will not be
    developed. At the suggestion of the two participating institutions
    (Sysin Institute of General and Community Hygiene, USSR, and the
    National Institute of Environmental Health Sciences, USA), the
    International Programme on Chemical Safety (IPCS) undertook a study of
    the principles and methods used to study the effects of chemicals on
    the nervous system (neurotoxicology) in order to lay the foundations
    for further developments in this important area of toxicology. The
    publication of the results of the study in a monograph seemed the most
    effective means of achieving this goal.

        Members of an International Task Group of experts from 18
    countries generously devoted much time to the preparation of the
    monograph, and the IPCS wishes to express its deep gratitude to all
    the members of the Task Group for their efforts.

        The scope and plans for the development of the monograph were
    discussed at a consultation, chaired by Dr C.L. Mitchell with Dr R.V.
    Merkureva as Vice-Chairman. Representatives from the IPCS institutions
    that had expressed a particular interest in the assessment of
    neurotoxic effects of chemicals attended the meeting at the invitation
    of the Manager, IPCS, Dr M. Mercier. The following institutions were
    represented: National Institute of Environmental Health Sciences
    (NIEHS), Research Triangle Park, North Carolina, USA; Sysin Institute,
    Academy of Medical Sciences, Moscow, USSR; Institute of Hygiene and
    Epidemiology, Czechoslovak Academy of Sciences, Prague,
    Czechoslovakia; US Environmental Protection Agency (US EPA), Health
    Effects Research Laboratories, Research Triangle Park, North Carolina,
    USA; Atomic Energy Commission, Department of Biology, Gif-sur-Yvette,
    France; Institute of Occupational Health, Helsinki, Finland; and
    Institute of Environmental Hygiene, University of Dsseldorf, Federal
    Republic of Germany. It was agreed that nine background papers would
    be used as the basis for the preparation of sections in a monograph
    reviewing the principles and methods for the evaluation of effects on
    the nervous system associated with exposure to chemicals.

        On 1-7 June 1983, the first meeting of the Task Group was convened
    in Moscow, hosted by the USSR Ministry of Health, the USSR Commission
    for UNEP, and the Centre for International Projects, GKNT. The Sysin
    Institute of General and Community Hygiene collaborated in the
    preparations for this meeting. Dr S.N. Bajbakov, Director, Centre of
    International Projects, GKNT, formally opened the meeting and Dr A.M.

    Pisarev, USSR Ministry of Health and Dr N.N. Litvinov of the Sysin
    Institute added greetings to the Task Group. It was agreed that
    Dr C.L. Mitchell would be Chairman and Dr V.G. Zilov, Vice-Chairman.
    Dr J.B. Cavanagh was asked to be Rapporteur of the meeting.

        The scope and content of ten background papers made available by
    the Secretariat were discussed thoroughly by the experts, and
    agreement was reached on the appropriate sections needed. After a
    detailed discussion on the issues to be addressed in each section, the
    meeting divided into five working groups to draft the texts for the
    following sections: General Principles, Morphology/Pathology,
    Biochemistry, Neuroendocrinology, Neurophysiology, and Behaviour. At
    the end of the meeting, most of the text was completed, and Dr
    Mitchell was asked to act as an overall editor to assist the Working
    Group leaders in completing their sections.

        A revised text was submitted to the Task Group before its second
    meeting, held in Prague on 17-21 September 1984. Dr C.L. Mitchell was
    again asked to be Chairman and Dr M. Horvath, Vice-chairman.
    Dr J. Cavanagh was appointed Rapporteur. The meeting was hosted by the
    Ministry of Health of the Czech Socialist Republic, Prague, and
    Professor Dr D. Zuskova welcomed the experts on behalf of the Minister
    of Health. The final text for the document was agreed upon by the end
    of the meeting. Dr Mitchell was asked to continue as overall editor to
    ensure the incorporation of all material discussed at the meeting,
    including the appropriate references.

        The International Programme on Chemical Safety wishes to
    acknowledge the valuable help provided by many scientists who were not
    members of the Task Group, and, in particular, that of Dr R. Dyer
    (Neurophysiology), Dr R. McPhail and the late Dr P. Ruppert
    (Behaviour) of the US EPA, and Dr T. Damstra (Neurochemistry) of the
    National Institute of Environmental Health Sciences.

                                     * * *

        Partial financial support for the publication of this criteria
    document was kindly provided by the United States Department of Health
    and Human Services, through a contract from the National Institute of
    Environmental Health Sciences, Research Triangle Park, North Carolina,
    USA -- a WHO Collaborating Centre for Environmental Health Effects.
    The United Kingdom Department of Health and Social Security generously
    covered the costs of printing.

    1.  INTRODUCTION

        Interest in nervous system toxicology has been growing in recent
    years, not only because of increased public concern over the impact of
    toxic agents on human health and the quality of life, but also because
    the nervous system has been shown to be particularly vulnerable to
    chemical insult. Thus, there has been an increased demand for improved
    methods for the detection of neurotoxic effects and the assessment of
    health risks within the field of occupational and environmental health
    and safety research related to the science known as "Neurotoxicology"
    (Spencer & Schaumburg, 1980; O'Donoghue, 1985).

        Neurotoxicology includes studies on the actions of chemical,
    biological, and certain physical agents that produce adverse effects
    on the nervous system and/or behaviour during development and at
    maturity. Toxic disorders of the nervous system of human beings and
    animals may occur following abuse of such substances as ethanol,
    inhalants, narcotics, therapeutic drugs, products or components of
    living organisms (e.g., bacteria, fungi, plants, animals), chemicals
    designed to affect certain organisms (e.g., pest-control products),
    industrial chemicals, chemical warfare agents, additives and natural
    components of food, raw materials for perfumes, and certain other
    types of chemicals encountered in the environment.

        This document is intended to aid in the design and assessment of
    studies concerned with exploring the association between exposure to
    chemicals and the development of adverse neurobehavioural changes. The
    emphasis is on animals as systems to model and predict adverse
    reactions of the human nervous system to exogenous chemicals.

    1.1  The Importance of Studying the Nervous System

        The brain is an extremely complex organ, the function of which is
    to receive and integrate signals and then to respond appropriately, to
    maintain bodily functions. It supports a diversity of complex
    processes including cognition, awareness, memory, and language. Sexual
    behaviour, locomotion, and the use of a vast array of tools ranging
    from the slingshot to the microcomputer, suggest the range of
    responses available to the human organism. Moreover, the nervous
    system is influenced by the functioning of other organ systems (e.g.,
    hepatic, cardiovascular, and endocrine systems). Thus, toxicant-
    induced alterations in any of these organ systems can be reflected in
    changes in neurobehavioural output. This fact alone suggests that
    nervous system function should be among the first to be thoroughly
    assessed in cases of exposure to known or potentially hazardous
    agents.

        Major outbreaks of neurotoxicity in human populations of various
    sizes have emphasized the importance of neurotoxicology as an
    independent discipline. Poisoning episodes have resulted from

    exposures in the environment (e.g., methylmercury, lead), at the work-
    place (e.g., hexanes, carbon disulfide, leptophos, chlordecone), as
    well as from food toxins (cassava) and food contaminants
    (triorthocresyl phosphate (TOCP)). There are numerous sources of
    reference dealing with chemicals reported to produce behavioural and
    neurological effects in human beings and laboratory animals including
    Horvath (1973), Xintaras et al. (1974), Weiss & Laties (1975), Spencer
    & Schaumburg (1980), and O'Donoghue (1985). These references should be
    consulted by the reader unfamiliar with this area of toxicology.

        At present, it is not possible to give a precise estimate of the
    number of chemicals that exert behavioural or neurotoxic effects.
    Anger & Johnson (1985) list more than 850 chemicals that have been
    reported to produce such effects. Anger (1984) states that, of the 588
    chemicals listed in the 1982 edition of the American Conference of
    Governmental Industrial Hygienists publication "Threshold Limit Values
    for Chemical Substances and Physical Agents in the Workroom
    Environment", 167 (29%) "have threshold limit values (recommended
    exposure maxima) based, at least in part, on direct neurological or
    behavioural effects, or on factors associated with the nervous system
    (viz., cholinesterase inhibition)." The neurotoxic properties of
    chemicals have been identified because of the conspicuous nature of
    severe signs and symptoms. It is not known how often insidious
    problems of neurotoxicity may lie undetected because the effects are
    incorrectly attributed to other conditions (e.g., advancing age, mood
    disorders) or misdiagnosed. The early and incipient stages of
    intoxications produced by environmental agents are frequently marked
    by vagueness and ambiguity (Mello, 1975; US NAS, 1975), and many
    complaints are subjective (e.g., fatigue, anxiety, irritability,
    lethargy, headache, weakness, depression). Thus, the potential is
    large for the occurrence of subtle, undetected effects, which
    nonetheless have an important bearing on the quality of life.

        The following examples illustrate the types of neurotoxic
    outbreaks that have occurred and exemplify the usefulness of animal
    models to further characterize this neurotoxicity.

    1.1.1  Methylmercury

        The neurotoxicity of mercury compounds has received world-wide
    attention for centuries (Hunter et al., 1940). However, it was not
    until the outbreak of methylmercury poisoning in Minamata, Japan in
    the 1950s that the extreme toxicological consequences of human
    exposure were fully recognized (Takeuchi, 1968).  During this outbreak
    of poisoning, thousands of inhabitants were exposed to methylmercury.
    The source of exposure was found to be industrial effluent that
    contained large amounts of mercury. The mercury made its way into
    Minamata Bay where it was converted to methylmercury by marine biota.
    This accumulated in fish and shellfish, which were eventually consumed
    by the local residents. This consumption of contaminated seafood

    continued for several years and resulted in hundreds of reported cases
    of methylmercury poisoning. Since that time, methylmercury poisoning
    has been referred to as "Minamata Disease".

        A second major outbreak of methylmercury poisoning occurred in
    Iraq during 1971-72 (Bakir et al., 1973). In this case, the source of
    exposure was the ingestion of seed grain treated with a methylmercury
    fungicide; the grain had been ground into flour to make bread.
    Approximately 6500 people were hospitalized due to poisoning, and at
    least 400 died.

        The clinical manifestations of methylmercury poisoning are quite
    extensive and include disturbances in sensory, motor, and cognitive
    functions.  The earliest complaints are associated with sensory loss
    in the extremities, perioral numbness and concentric constriction of
    the visual fields, followed by the development of ataxic gait,
    dyarthria, incoordination, intention tremor, hearing problems, and
    muscle weakness (Takeuchi et al., 1979). Mental disturbances, as well
    as alterations in taste, smell, and salivation have also been reported
    (Takeuchi, 1968; Chang, 1977, 1980; Reuhl & Chang, 1979, Takeuchi et
    al., 1979).

        The pattern of sensory neural damage in the monkey resembles that
    of man. Visual system deficits in primates include constriction of
    visual fields, deficit in scotopic (low light) vision (Evans et al.,
    1975), and deficit in ability to detect flicker (Merigan, 1980).
    Spatial visual function was impaired in monkeys dosed from birth (Rice
    & Gilbert, 1982). Hypoesthesia (impaired sense of touch) has been
    reported in the monkey (Evans et al., 1975) as well as man (Hunter et
    al., 1940). Moreover, the body burden at which signs are observed in
    primates is similar to that reported for human beings.

        In rodents, neurobehavioural research on methylmercury has
    concentrated primarily on an evaluation of the later motor effects,
    but some of the sensory deficits have not been found (Shaw et al.,
    1975, 1980; Evans et al., 1977). Several investigators have observed a
    progressive weakness of the hindlimbs followed by decrease in forelimb
    function (Diamond & Sleight, 1972; Klein et al., 1972; Herman et al.,
    1973; Snyder & Braun, 1977; Ohi et al., 1978). Grip strength was
    reported to be reduced in rats following long-term dosing with
    methylmercury (Pryor et al., 1983). Decreased motor activity in mice
    exposed to methylmercury in the drinking-water was reported by
    MacDonald & Harbison (1977), but horizontally-directed motor activity
    did not appear to be markedly affected by repeated exposure to
    methylmercury (Salvaterra et al., 1973; Morganti et al., 1976).
    Responsiveness to noxious stimuli was reportedly intact in
    methylmercury-exposed rats, even in the presence of gross neuromotor
    deficits (Herman et al., 1973; Salvaterra et al., 1973). Similarly,
    Pryor et al. (1983) did not observe any significant alteration in
    startle responsiveness to an acoustic stimulus in methylmercury-
    exposed rats. Finally, in an attempt to investigate the effects of

    mercury on development, prepubescent rats were exposed to a single
    dose of methylmercury. Ability to learn an active avoidance response
    at 70 days of age was impaired (Reuhl & Chang, 1979).

    1.1.2  Carbon disulfide (CS2)

        Carbon disulfide (CS2) is a volatile solvent that has been used
    for a variety of industrial purposes. Since its discovery in 1776,
    there have been numerous examples of CS2-induced neurotoxicity. 
    Many cases of human CS2 poisoning occurred in the viscose industry
    during and after the second world war and consisted of various
    neurological and behavioural effects. According to Braceland (1942),
    the psychological effects consisted of personality changes,
    irritability, memory deficits, insomnia, bad dreams, decreased libido,
    and constant fatigue.

        Paraesthesia and dysesthesia from CS2 exposure tend to occur in
    a "stocking and glove" distribution characteristic of peripheral nerve
    injury and indeed Seppalainen et al. (1972) reported a slowing of
    motor conduction in peripheral nerves. Various sensory alterations
    have been reported following CS2 exposure, particularly in vision. 
    Central field loss, disturbances in colour vision, enlargement of the
    blind spot, and reduction in peripheral vision have been reported.
    Alterations in auditory, vestibular, and olfactory senses have also
    been reported (Wood, 1981).

        While there have been few studies that have attempted to quantify
    the sensory and psychological effects of CS2 in laboratory animals
    (Wood, 1981), the neuropathological changes and effects on motor
    behaviour have been extensively studied. CS2 produces filamentous
    inclusions in axons in both the central and peripheral nervous
    systems. Experimental CS2 neuropathy consisting of weakness or loss
    of power in the limbs, depressed reflexes, and general behavioural
    suppression has been reported in dogs (Lewey, 1941), rabbits
    (Seppalainen & Linnoila, 1975), and rats (Frantik, 1970; Lukas, 1970).
    Tilson et al. (1979b) reported decreases in grip strength and motor
    activity in rats following CS2 exposure; Horvath (1973) found
    decreases in motor activity in rats following exposure to high
    concentrations of CS2, but motor activity increased following
    exposure to low concentrations.

        The cases illustrated above are unfortunately typical in that the
    neurotoxicity of the chemicals was first discovered when human beings
    exposed to them became ill. Later, research using animal models in a
    controlled setting provided experimental evidence that the chemical
    believed to have caused the illness in the human population produced
    similar effects in animals. One purpose of this book is to provide the
    background through which this process can be reversed; i.e., the
    chemical can be tested in an animal model before human beings become
    overexposed, seriously ill, or irreversibly harmed by it.

        There are a few examples where this principle has been put into
    practice. Some compounds have been discovered to be neurotoxic in
    animals and, as a result, further human exposure has been
    discontinued. Other chemicals, shown to be neurotoxic for animals,
    have only later been discovered to produce comparable disorders in
    human beings.

        The former is illustrated by Musk Tetralin, a synthetic musk
    introduced in 1955 as a raw material and as an artificial flavouring
    substance. Apparently, the minimal acute toxicity testing in use at
    that time suggested that the compound was acceptable for human
    consumption. After 20 years of widespread use in the domestic
    environment, a new set of long-term animal toxicity studies showed
    that the substance was a potent neurotoxin inducing behavioural
    changes and irreversible structural damage throughout the nervous
    system of dermally- or orally-exposed animals (Spencer et al., 1980).

        Two examples illustrate that experimental animal studies may
    reveal the neurotoxic properties of substances before they are
    discovered in human beings. One concerns aluminium, a substance that
    was shown to induce central nervous system toxicity in animals as
    early as 1886. Subsequent studies confirmed this, before the first
    case of suspected human aluminum encephalopathy was reported in 1962.
    With the introduction of aluminium-containing phosphate-binding gels
    for the haemodialysis of individuals with kidney dysfunction, large
    numbers of treated patients developed a progressive dementing illness
    that often proved fatal (Crapper & DeBoni, 1980). Another recent
    example of this phenomenon concerns the peripheral neurotoxicity of
    megavitamin doses of pyridoxine phosphate (vitamin B6), recently
    reported in human beings taking the drug in prescribed amounts for
    therapeutic purposes (Schaumburg, 1982). The clinical neurological
    syndrome had been reproduced in animals some 40 years previously and
    would not have been recognized in human beings but for more recent
    experimental animal work that had highlighted the surprising
    neurotoxic potential of this essential nutrient (Krinke et al., 1981).

        Both examples demonstrate the power of animal studies in
    predicting human neurotoxicity and both illustrate the need for the
    medical world to pay close attention to discoveries in experimental
    neurotoxicology. With the further refinement and validation of methods
    used in this discipline, it should be possible to discover many other
    potentially harmful agents, presently in use, and to prevent new
    compounds with neurotoxic properties from reaching the human
    environment. Compounds that have been identified, through various
    types of research, are listed in Table 1 according to important
    nervous system targets.

    Table 1.  Illustrative manifestations of human neurotoxicity
                                                                        

    Function affected      Manifestation              Chemical example
                                                                        

    Sensorium change       irritability               carbon disulfide
                           apathy/lethargy            carbon monoxide
                           attention difficulty       anticholinesterases
                           illusions, delusions,      ergot
                           hallucinations
                           dementia                   aluminium
                           depression, euphoria       ozocerite
                           stupor, coma               dicyclopentadiene

    Sensory special        Abnormalities of:
                           smell                      cadmium
                           vision                     methanol
                           taste                      selenium
                           hearing                    toluene
                           balance                    methyl nitrite

    Somatosensory          skin senses (e.g.,         trichloroethylene
                           numbness, pain)            thallium
                           proprioception             acrylamide

    Motor                  muscle weakness:
                           paralysis,                 organophosphates
                           spasticity,                Lathyrus sativus
                           rigidity                   methyl phenyl
                                                      tetrahydropyridine
                           tremor                     chlordecone
                           dystonia                   manganese
                           incoordination             organomercury
                           hyperactivity              lead
                           myoclonus                  toluene
                           fasciculatton              anticholinesterases
                           cramps                     styrene
                           seizures, convulsions      acetonitrile

    Autonomic              Abnormalities of:
                           sweating                   acrylamide
                           temperature control        chlordane
                           gastrointestinal           lead
                             function
                           appetite, body-weight,     dinitrobenzene
                           cardiovascular control,    vocor
                           urination                  dimethylamino-
                                                        propionitrile
                           sexual function              -chloroprene
                                                                        

    1.2  Need to Establish A Comprehensive Strategy for Neurotoxicity
         Testing

        The need to establish a comprehensive strategy for neurotoxicity
    testing is made clear by estimates recently provided by the US EPA
    Office of Toxic Substances. In the USA alone, between 40 000 and
    60 000 chemicals are currently in commercial use; furthermore,
    approximately 1000 new chemicals are introduced into commerce each
    year (Reiter, 1980). It is not surprising that few or no toxicological
    data exist for many of these chemicals. Therefore, any toxicity
    testing strategy intended for general use must be flexible enough to
    evaluate a wide variety of chemicals and chemical classes and must be
    able to take into account the potential for chemical interaction, as
    most people are exposed to combinations of chemicals. The
    toxicological data that are available for a given chemical will
    influence testing requirements for that chemical. For example, the
    desired approach will depend on whether the investigation has been
    initiated to evaluate the toxicity of the chemical prior to its
    commercial use or to confirm reports of chemically-induced disease in
    man.

        Risk from exposure to a toxic substance is a function of both the
    intrinsic toxicity of the chemical and the human exposure pattern.
    These factors will influence where it enters the testing scheme; its
    potency and the human exposure pattern will influence the extent to
    which testing must be pursued. In both cases, steps must be
    incorporated that facilitate decision-making about acceptance,
    rejection, or continuation of testing. However, the eventual goal is
    to understand the mechanism(s) by which chemicals adversely alter
    nervous system function.

    1.3  Scope of the Book

        Emphasis in this book is placed on animal test data. An important
    role of the toxicologist, of course, is to provide data that can be
    used for quantitative estimation of the risks associated with exposure
    of human populations to toxic chemicals. Even in extensively explored
    areas such as carcinogenesis, risk assessment is an extremely
    difficult undertaking for which there is no concise research strategy.
    A major problem in dealing with most chemicals is that the available
    toxicological data base concerning precise information on human health
    effects is relatively sparse. The neurotoxicological data base is no
    exception in this respect; in particular, adequate human behavioural
    data are available for only a few chemicals (Anger & Johnson, 1985).
    Furthermore, a number of experimental problems inherent in most
    published human studies cloud interpretation of data from these
    studies. Perhaps the most serious problem is that of adequately
    defining the exposure level (dose) in human populations, a critical
    factor in risk assessment. Other factors that influence behavioural
    measures include sex, age, cultural variables, disease states, and

    possible exposure to additional toxic substances (Johnson & Anger,
    1983). As these critical variables cannot be adequately controlled in
    human field studies, in all or even most cases, the neurotoxicologist
    must rely to a great extent on animal test data to establish accurate
    dose-effect relationships.

        A word should be said about related and important areas excluded
    from this publication. Developmental and human neurotoxicology are not
    discussed in detail, since they have already been addressed in other
    publications of the World Health Organization (WHO, 1984; WHO, in
    preparation). The usefulness of tissue culture techniques will be
    considered in a future publication. The closely related area of
    psychopharmacology, which assesses the effects of drugs on the nervous
    system, is not within the scope of this document and has been
    addressed in other publications (CIOMS, 1983). However, many of the
    methods identified in this manuscript have been used in this area.

        The very important role played by the autonomic nervous system in
    the regulation of many physiological processes is well recognized
    (Dyck et al., 1984). However, it has not been possible to discuss in
    any detail the principles of the methods available to study the
    effects of chemicals on this system. A detailed discussion of such
    methods will be presented in a WHO publication that is in preparation
    and deals with methods for the assessment of toxicity from exposure to
    chemicals.

    1.4  Purpose of the Publication

        Given the magnitude of the problem and the potential for subtle
    damage to the nervous system, there is a pressing need to assess the
    behavioural and neurological effects of the mass of chemicals found in
    the work-place and the ambient environment.

        The purpose of this book is to provide an overview of the
    principles of neurobehavioural assessment and to identify methods that
    have been successfully applied to the study of neurotoxicity in the
    past. These methods, which may eventually be modified or supplemented
    by other better methods, have been generally established and can be
    relied on to provide an assessment of chemicals for their
    neurotoxicity.  The references for each method can be consulted for
    further details.

        The book is divided into five sections. The first deals with
    principles of assessment, and the remaining sections deal with methods
    in the four major research disciplines that contribute substantially
    to the assessment of the neurotoxicity of chemicals found in the
    occupational or community environment. Each discipline is not in a
    comparable stage of development as it relates to neurotoxic assessment
    of environmental chemicals. Behavioural and neurophysiological methods
    have been extensively applied, but there is limited agreement on which

    of the many methods described are the most appropriate for the initial
    screening of an unknown chemical and, thus, the approach of tiered
    testing is suggested or general guidelines are given for selecting
    methods. On the other hand, the neuromorphological section presents a
    relatively more methodological approach to the neuropathological
    assessment of nervous tissue that has been validated in other areas of
    research, recognizing that it is the experience of the
    neuropathologist that is critical to an adequate assessment rather
    than the test methods, as in the sections mentioned above. Finally,
    the section on endocrinological and biochemical methods reflects the
    fact that these methods have been applied far less to the neurotoxic
    assessment of environmental contaminants. HoweVer, their inclusion is
    important because of the critical role they play in exposure
    monitoring and the elucidation of mechanisms.

    2.  GENERAL PRINCIPLES FOR THE ASSESSMENT OF TOXIC EFFECTS OF CHEMICALS
        ON THE NERVOUS SYSTEM

    2.1  Factors to be Considered in the Design of Neurotoxicity Studies

    2.1.1  General considerations

        Many factors must be taken into consideration with regard to any
    toxicology study. These include the choice and number of animals,
    dosage, route and duration of administration, metabolism and
    pharmacokinetics, and testing procedures. These have been discussed in
    detail elsewhere (US NAS/NRC, 1970; US NAS, 1975; WHO, 1978) and the
    various reference sources should be consulted, as only aspects of
    special relevance to neurotoxicology will be emphasized here.

        The nervous system is protected from undesirable external
    influences by both physical and chemical barriers. This protection,
    however, is not complete. The blood-brain barrier has an important
    function in preserving the chemical constitution of the nervous
    system, but some noxious substances, particularly those that are lipid
    soluble, may still cross it. Another mode of entry is by uptake into
    the peripheral terminals of nerves, which may then transfer the
    substances into their cell bodies in the central nervous system
    through retrograde axonal flow. Such a mechanism operates for
    substances as remote as tetanus toxin and some viruses. The peripheral
    nervous system is, of course, more likely to be exposed to
    neurotoxicity. The neurons of the autonomic nervous system and the
    sensory ganglia are outside the blood-brain barrier, as are small
    regions of the CNS, circumventricular organs (e.g., area postrema)
    and, to a limited extent, the retina.

        As might be expected, the nervous system may be particularly
    vulnerable either during development or in senescence. Some aspects of
    this have been alluded to elsewhere (WHO, 1984). Physical changes or
    the presence of toxins may also disrupt the blood-brain barrier and,
    thus, allow substances normally excluded from the brain to reach and
    affect it adversely.

    2.1.2  Objectives

        The objectives of neurotoxicity testing are to:

        (a) identify whether the nervous system is altered by the toxicant
            (detection);

        (b) characterize the nervous system alterations associated with
            exposure;

        (c) ascertain whether the nervous system is the primary target for
            the chemical; and

        (d) determine dose- and time-effect relationships aimed at
            establishing a no-observed-adverse-effect level.

    In a sense, these objectives translate into a series of questions
    about the toxicity of a chemical, and achieving them requires
    behavioural, neurophysiological, biochemical, and neuropathological
    information.

        When faced with a chemical for which no toxicological data are
    available, the first question is whether the nervous system is or is
    not affected by the chemical. This represents the most fundamental
    level of investigation and entails procedures that "screen" for
    neurotoxicity. Such tests must not only forecast the potential of a
    substance to produce adverse effects, but must also be simple, rapid,
    and economical to administer. Once a chemical is known to produce
    neurotoxic effects, further studies must be performed in order to
    characterize the nature and mechanism of the alterations. These
    studies explore the consequences of toxicant exposure and give an
    indication of whether or not the nervous system is the primary target
    organ.

        Many functions are mediated by unique neural substrates, and
    chemicals may produce selective effects. Thus, it is important to use
    a variety of tests that measure different functions, in order to
    maximize the probability of detecting a toxic effect. It is clear that
    the methods used may differ depending on the following factors:

        (a) the objective of the study;

        (b) the age of the animal; and

        (c) the species examined.

    If the objective of the study is to provide an initial evaluation of
    the effect of a new substance on the nervous system, the methods used
    may differ considerably from those used when a great deal is known
    about a substance, and its mechanism of action is being investigated,
    or environmental or occupational standards are being set for
    acceptable levels in the biosphere. If the objective is to contribute
    substantively to the overall toxic risk assessment for the chemical,
    the methods used should attempt to model human disease states. Thus,
    it is important that the purpose of the study is clear to both the
    investigator and the evaluator.  In many situations, the evaluation of
    reference substances in the same protocols will help determine the
    specificity and validity of the observed changes. It also makes it
    possible to evaluate the relative potency of different chemicals
    (Horvath & Frantik, 1973), and this is always essential for the novice
    investigator or for the investigator using a new technique.

        Although certain chemicals produce selective damage in the nervous
    system, a more common finding is one of widespread damage and
    disruption of a variety of functions. Ideally, characterization of
    such generalized neurotoxicity by a variety of methods will establish
    a profile of the disrupted functions.

        Once a chemical has been identified as neurotoxic, the next
    objective is to determine dose-effect and time-effect relationships.
    One aim of these studies is to establish no-observed-adverse-effect
    levels, but to prove that a certain dose produces no effect may
    require a very large number of experimental animals (Dews, 1982). To
    be useful in risk assessment, threshold determinations must be
    obtained by the most sensitive tests available. For example, as the
    toxic effects of methylmercury became known, studies using subhuman
    primates began to focus on its effects on the visual system.
    Alterations in visual function in monkeys following methylmercury
    exposure has now been well documented using sophisticated visual
    psychophysical techniques (Evans et al., 1975).

        The question of how to define toxicity is of critical importance
    for the ultimate goal of risk assessment and the establishment of
    hygienic standards. Considerable controversy exists concerning what
    constitutes an adverse effect in toxicology. According to one view,
    any evidence of a behavioural or biological change is considered to be
    an adverse effect. According to others, evidence is required of both
    an irreversible decrement in the ability of the organism to maintain
    homeostasis and/or an enhanced susceptibility to the deleterious
    effects of other environmental influences. In this latter view,
    differentiation between "nonadversive" and "adverse" effects requires
    considerable knowledge of the importance of reversible changes and
    subtle departures from "normal" behaviour, physiology, biochemistry,
    and morphology in terms of the organism's overall economy of life,
    ability to adapt to other stresses, and their possible effects on life
    span (WHO, 1978). Real or potential risks to the nervous system are
    difficult to assess because of its complexity. Some of the problems in
    assessment are associated with the wide variations that can occur but
    are still considered to be within the "normal" range. Some are
    associated with the plasticity of the nervous system. Other problems
    in assessment are related to incomplete understanding of what is being
    measured by certain tests. It is clear, therefore, that no single test
    will suffice to examine the functional capacity of the nervous system.

        The above comments suggest tiered testing approaches, such as
    those recommended in the section on behavioural methods, where a
    variety of testing schemes, ranging from simple to complex, have been
    proposed (Pavlenko, 1975; US NAS, 1975; Tilson & Cabe, 1978). Such
    schemes typically begin with simple, rapid, inexpensive tests for
    detecting the presence of neurobehavioural effects. The tests in
    successive stages are designed to answer increasingly specific
    questions about the toxicity of the chemical. Each stage should also

    incorporate decision points as to whether the available information is
    sufficient for determining the toxicity of the chemical. When combined
    with estimates of potential exposure of human populations, this
    information can provide a basis for evaluating the justification for
    proceeding to the next level of testing. The advantage of the tiered
    approach is that decisions are made at each level of testing and,
    therefore, scarce resources are directed towards chemicals for which
    the greatest hazard or risk potential exists.

        Obviously, the amount of available information about the substance
    will determine the level at which the chemical will enter this testing
    scheme. Another inherent assumption is that a chemical's pattern of
    use, in combination with its toxic potency, will be considered in any
    decision about further testing. Testing requirements for a chemical
    that is indigenous the environment and to which large segments of the
    population are exposed will require extensive investigation leading,
    ideally, to determination of the no-observed-adverse-effect level.
    This information is extremely useful to governmental agencies
    responsible for setting exposure or hygienic standards. On the other
    hand, compounds that are being introduced into commerce and/or for
    which the projected exposure is limited may require less testing.

    2.1.3  Choice of animals

        For obvious reasons of safety and ethics, it is necessary to use
    animals in toxicity assessments. However, the extrapolation of animal
    toxicological data to human beings is always tenuous and should be
    carried out with caution. In preliminary mass screening of known or
    suspected environmental toxicants, there are economic factors that
    must be taken into account. It is also important that there be
    adequate anatomical, physiological, pharmacological, and toxicological
    data bases on the species chosen for study, so that meaningful
    interpretations of effects can be made and appropriate hypotheses
    about mechanisms and loci of action can be framed. For these reasons,
    the mouse or rat is usually preferred in a preliminary screen, though
    the rodent differs from man in many significant ways. For more
    detailed studies, other species may provide a more appropriate model.
    For example, the adult chicken is the animal of choice to test
    organophosphate-induced delayed neurotoxicity (Abou-Donia, 1981).

        Other variables, besides species, must also be considered. One of
    these is the strain of animal used. For example, it has been
    demonstrated that rats inbred from the Fischer strain are
    behaviourally different from Zivic-Miller rats, which are derived from
    the Sprague Dawley strain (Barrett et al., 1973; Ray & Barrett, 1975).
    Rats of the Fischer strain rapidly learn both where and when to run in
    a discriminated Y-maze avoidance task, whereas Sprague Dawley rats
    eventually learn where to escape but not when to avoid shock. The
    administration of amphetamine produces a dramatic improvement in the
    avoidance response of Sprague Dawley rats, whereas little or no
    behavioural facilitation is observed in the Fischer strain of rat

    (Barrett et al., 1974). Festing (1979) has reviewed the properties of
    isogenic and nonisogenic stocks and their relation to toxicity
    testing.

        Since an environmental agent may have a selective effect on either
    the male or female, gender cannot be ignored when assessing
    neurotoxicity. For example, sex differences have been seen for the
    toxic effects of polychlorinated hydrocarbons (Lamartiniere et al.,
    1979); gonadal hormones influence the biotransformation and toxicity
    of DDT (Durham et al., 1956) and parathion (Agrawal et al., 1982).

        Another important factor is the age of the animal. It is well
    known that the effects of a toxic agent may vary dramatically
    depending on the stage of maturation of the animal (Damstra & Bondy,
    1982; Hunt et al., 1982). For example, it has been established that
    young animals of otherwise sensitive species are not susceptible to
    organophosphate-induced delayed neurotoxicity (Abou-Donia, 1981). It
    has been suggested that, under conditions where exposure may occur
    pre- or perinatally, animals of both sexes should be tested at all
    stages of maturation (Spyker, 1975).

        Each of these considerations relates to extrapolation, a subject
    discussed in detail in WHO (1978). Quantitative and qualitative
    differences in sensitivity to, and body distribution of, chemicals
    affect extrapolation significantly. A better understanding of
    structure-activity relationships, pharmacokinetics, and mechanisms of
    toxicity will facilitate cross-species extrapolation (Dixon, 1976).

    2.1.4  Dosing regimen

        Some compounds produce toxic effects following a single exposure
    (e.g., trimethyltin, organophosphates); for others, cumulative effects
    follow prolonged or repeated exposure (e.g., acrylamide).  In
    environmental toxicology, the detection of cumulative toxicity
    following continued (or intermittent) exposure is a major goal. Thus,
    a multiple-dosing regimen is most frequently used. It is important to
    assess the toxicity at various intervals, since both quantitative and
    qualitative changes in the response to environmental factors can occur
    on repeated exposure, or even with time following a single exposure
    (Evans & Weiss, 1978). Assessments should be made for some time
    following cessation of the dosing regimen, since it is of interest to
    determine the reversibility of any effects noted during the dosing
    phase and to note any post-dosing effects.

    2.1.5  Functional reserve and adaptation

        Functional reserve is the excess capacity possessed by the nervous
    system. Thus, a portion of the nervous system can be damaged, and this
    damage can go undetected by the usual functional tests. The situation

    in which a change in function was observed at one time, but can no
    longer be detected by the usual functional tests, is referred to as
    adaptation and presumably reflects compensatory processes.

        If a part of a redundant system is damaged, it is reasonable to
    assume that the reserve potential has been reduced. If compensatory
    changes have occurred, the ability of a system to make further
    compensatory changes may also have been reduced. One way to assess
    such changes is to incorporate in the test procedures one or more
    conditions in which the system(s) or organism(s) are placed under
    stress. The combination of the test substance plus stress may result
    in a greater deficit in performance than can be seen in animals
    receiving either the stress or the toxicant only. Examples of
    stressors that have been used are pharmacological changes such as
    ethanol, muscular or work stress, exposure to cold, or auditory and
    electrical stimuli (Lehrer, 1974; Pavlenko, 1975); it is expected that
    other powerful stressors include those referred to as psychological or
    psychic.

    2.1.6  Other factors

        Several additional factors should be carefully considered in
    designing neurotoxicological tests. One condition that may affect
    toxicity is the nutritional state of the animal. Changes attributed to
    exposure to toxicants might be due to relatively nonspecific effects
    related to inhibition of growth or decreases in food or water
    consumption. This is particularly true in studies involving developing
    organisms.

        Another variable is the housing conditions of the experimental
    animal. In some cases, animals are housed individually in home cages
    during pharmacological or toxicological studies. This arrangement can
    alter the responsiveness of the subjects to drugs. Pirch & Rech (1968)
    found that alpha-methyltyrosine, a depletor of brain norepinephrine
    and dopamine, produced less depression of motor activity and rotorod
    performance in rats isolated for 34 days than in grouped rats. The
    potentiating effects of crowding on the toxicity of amphetamine in
    mice are also well known (Chance, 1946).

        It has been well established that numerous biological systems,
    ranging from metabolic pathways to behaviour of the whole organism,
    exhibit rhythmic changes in amplitude (Scheving et al., 1974). Classes
    of behaviour that show circadian rhythms include feeding, drinking,
    sleeping, motor activity, and mating (Rusak & Zucker, 1975). There is
    growing literature on how biological rhythms influence the
    pharmacological and toxicological response to chemicals (Reiter &
    MacPhail, 1982). These biological rhythms cannot be ignored and must
    be either controlled for in the study or studied explicitly.

    2.2  Statistical Analysis

    2.2.1  Type I and Type II errors

        All fields of biological research have at least one feature in
    common: inherent variability in their data. When there is considerable
    variation in the experimental material and when it is not feasible to
    examine the entire population, the research worker is forced to give a
    probability statement concerning any treatment differences observed.
    In order to do this, it is necessary for the study to be designed in
    such a way that statistical analysis of the data will yield a valid
    answer to the question, "What is the likelihood that the differences
    observed could have occurred by chance?" Thus, a null hypothesis is
    set up, i.e., a statement that there is no difference between the
    parameters or the distributions being estimated by the samples. Taking
    the simplest case (e.g., Student's t-test), the null hypothesis is
    that there is no difference between the mean values for 2 populations
    (e.g., control versus treated). When the null hypothesis is accepted,
    this may be either right or wrong. A Type I error (false positive) is
    made when the null hypothesis is rejected and is, in fact, true. A
    Type II error (false negative) is made when the null hypothesis is
    accepted and is, in fact, false.

        The probability of making a Type I error is called alpha and is
    fixed before the study is carried out. If something is statistically
    significant at P = 0.05, this means that the probability that these
    results could have occurred by chance is 0.05. The inference is that
    there is a "real" difference, but this can be wrong, and, in fact,
    there is a one in twenty chance that it is.

        There are two characteristics about neurotoxicology that make it
    highly likely that one or more Type I errors may occur in any given
    study. These are (a) the use of multiple tests (measuring multiple
    parameters), and (b) repeated measurements using the same animal in
    the same test. Multiple tests are used because of the complexity of
    the nervous system and the need to assess sensory and motor function
    as well as more complex behaviour such as discrimination and learning
    processes. Repeated testing is done because changes in the response to
    agents can occur on repeated exposure, or even with time following a
    single exposure (Evans & Weiss, 1978). Thus, in any given study, there
    are a number of statistical tests of hypotheses. The greater the
    number of statistical tests of hypotheses, the higher the probability
    of obtaining Type I errors (false positives). Although problems
    created by multiple comparisons can be dealt with statistically, to a
    certain extent, it is imperative to look at the internal consistency
    of the data and not simply at the presence or absence of statistical
    significance. Nothing, of course, can take the place of a well-
    designed study with a clear statement of its purpose. In any case,
    when in doubt, repetition of the study is in order, if it is
    sufficiently important.

        Variable data can increase the probability of a Type II error
    (false negative). A Type II error is made when the null hypothesis is
    accepted and is, in fact, false. The probability of making such an
    error is called Beta. The value of Beta is seldom, if ever, known. Its
    relative magnitude can be approximated and depends on:

        (a) the distance between the population parameters being estimated
            by the samples (population means in the case of Student's t);

        (b) the value selected for alpha (the probability of making a
            Type I error or rejecting the null hypothesis when it is, in
            fact, true); and

        (c) the sample size.

    The smaller the distance between the population parameters, the larger
    will be beta. Beta varies inversely with alpha, and both decrease as
    sample size increases. Thus, if the data are highly variable, a large
    sample size is needed to detect a small effect. In selecting sample
    size, these factors must be taken into consideration. When the sample
    size is determined, the size of the difference that is detectable has
    also been determined. The smaller the sample size, the larger the
    change has to be in order to be statistically significant. Techniques
    are available that identify the sample size needed to detect either a
    given incidence of occurrence (power function) or a given change in
    magnitude on the basis of an estimate of the variability in the
    population(s). Their use is strongly urged. Too many studies have been
    conducted with sample sizes that were not adequate to detect any real
    but subtle effect. This topic has been discussed in detail by Dews
    (1982).

        Unfortunately, it is not possible to increase the sample size at
    will. Therefore, other means must be used in an attempt to reveal
    individual sensitivity to a given toxicant. In addition to statistical
    analysis, the raw data should be examined (this is always true,
    regardless of the results). If any trends exist, it may be necessary
    to repeat the study. Alternatively, the data may be examined for a
    change in variability, since a common observation of near-threshold
    doses of environmental toxicants seems to be an increase in the
    variability of the data (Evans & Weiss, 1978). This increase in
    variability is a clue that certain individuals in the population are
    being affected differentially. Moreover, statistical procedures exist
    for comparing the degree of variability between control and treated
    groups (e.g., Steel & Torrie, 1960).

    2.2.2  Selection of the appropriate statistical test(s)

        In the selection of statistical test(s), it is essential for the
    investigator to:

        (a) be able to choose a technique that is appropriate for the data
            and hypothesis;

        (b) understand the assumptions being made when carrying out the
            statistical test;

        (c) be able to execute the procedure correctly;

        (d) be able to interpret the results correctly; and

        (e) recognize the existence of controversies or differences of
            opinion in any of the above areas (Bennett & Bowers, 1976).

    Clearly, it is beyond the scope of this section to discuss these at
    length. Similarly, it should be clear that to abide by these
    principles, frequent and close consultation with a biostatistican may
    be necessary. Moreover, such consultation should occur before the
    study is conducted, not afterwards. Consulted beforehand, the
    statistician can give the guidance necessary for the proper design of
    the study. Once the study is completed, statistical manipulation
    cannot compensate for an ill-conceived experimental design.

        Different statistical techniques are based on different
    assumptions, either with respect to the nature of the data, their
    distribution, or both. Also, the power of different techniques
    differs. In statistics, power refers to the ability of a test to
    detect the alternative hypothesis (i.e., that there is a difference)
    when it is true. Given two tests and a particular level of alpha
    (e.g., P = 0.05), the test with the greater power will detect a
    significant difference with a smaller sample size compared with the
    test having lesser power, providing that the assumptions underlying
    the tests are true. Non-parametric alternatives are available in most
    cases where the assumptions of parametric tests are not valid. Gad &
    Wiel (1982) present a "statistical testing decision tree" for
    selecting the most appropriate test based on whether or not the data
    are quantitative (continuous) or qualitative (discontinuous) in nature
    and/or normally distributed. Gad (1982) has also examined the
    statistical tests most commonly used in behavioural toxicology for
    different types of observations (data) and suggests procedures that
    are more appropriate.

        Although much more could be said about statistics, it is hoped
    that the above comments will serve as a warning that the prudent
    investigator should become facile with experimental design and
    statistics or work together with a biostatistician, or both.

    3.  TEST METHODS IN BEHAVIOURAL TOXICOLOGY

    3.1  Introduction

        Behaviour has been used to study the adverse effects of chemical
    and/or physical agents on intact organisms. Behavioural toxicology
    draws on the fields of experimental psychology, behavioural
    pharmacology, and behavioural brain research.

        Behavioural toxicology plays an important role within the broader
    field of neurotoxicology for two reasons. The first is that the
    behaviour of an organism is important in itself. As mentioned in the
    introductory chapter, the nervous system (and consequently behaviour)
    is influenced by the functioning of other organ systems. Thus,
    regardless of whether the site of action is the nervous system or some
    other organ system, toxicant-induced changes in performance,
    sensorimotor function, or cognitive function adversely affect the
    organism and its ability to interact with its environment. The second
    reason is that behaviour is the final product of nervous system
    activity and, therefore, toxicant-induced changes in the nervous
    system may be reflected by behavioural changes. Thus, behavioural
    analysis serves as a useful tool for measuring neurotoxicity (i.e.,
    the direct action of a chemical on neural tissue). This approach can
    be compared to measuring blood flow or cardiac output as an index of
    cardiovascular toxicity (i.e., the use of functional measures to
    evaluate the status of a target organ).

        This section deals with the use of animal behavioural testing to
    estimate neurotoxicity in human beings. It should be noted that
    Citovic (1930), a student of Pavlov, reported the use of conditioned
    reflexes to study the neurotoxic effects of gasoline and acetone.
    Soviet and Eastern European toxicologists have commonly incorporated
    behavioural testing in their studies. However, scientists in Canada,
    western Europe, and the USA placed heavy emphasis in chemical toxicity
    studies on defining pathological changes following exposure, and it
    was not until 1969 that the Annual Review of Pharmacology in the USA
    included a section on behavioural toxicology (Weiss & Laties, 1969).
    Since then, behavioural toxicology has been the subject of numerous
    books, symposia, and reviews (Anger & Johnson, 1985). Thus, it is a
    relatively new discipline, and its specific methodology continues to
    evolve rapidly. In this section, the major emphasis is placed on the
    basic principles of behavioural toxicity testing. General approaches
    are discussed in relation to their use in a comprehensive test
    strategy for evaluating behavioural toxicity. As discussed later, any
    such strategy should include tests that adequately evaluate each of
    the 5 main functional categories listed in Table 1 (p. 17). Strengths
    and weaknesses of existing methods are also considered, together with
    future methodological directions and needs.

    3.2  Classes of Behaviour

        In the experimental analysis of behaviour, the primary focus is on
    defining the functional relationships between an organism's behaviour
    and its environment, i.e., in this context, everything that has an
    effect on the organism. Behaviour is a dynamic process, since it
    reflects changes in the interaction of an organism with its
    environment. Thus, an important feature of behavioural toxicology is
    that the effects of a toxic agent may depend largely on environmental
    circumstances. In other words, with a given toxic effect on the
    nervous system, the observed behavioural effects may (and probably
    will) depend on environmental factors.

        The basic units of behaviour are termed responses. Aspects of the
    external or internal environment that affect behaviour are termed
    stimuli. Behavioural responses have been typically divided into two
    classes based on the functional relations that control their
    occurrence. One class of behaviour is controlled mainly or exclusively
    by the prior occurrence of an event (stimulus) in the environment.
    Such responses are referred to as elicited or respondent behaviour.
    The events are called eliciting stimuli, and the responses are called
    respondents. The other class of behaviour is controlled mainly or
    exclusively by its consequences and is referred to as operant or
    emitted behaviour. Behaviour may be either unconditioned (unlearned)
    or conditioned (learned). Conditioning refers to the modification of a
    response that results from an organism's interaction with its
    environment. In general, however, when a response is said to have been
    conditioned, this usually implies that the conditioning was done
    explicitly as part of an experimental procedure rather than as a
    result of some other experience the organism encountered.

        Descriptions of conditioning as either "operant" or "respondent"
    can be confusing, in that behaviour is also characterized by these
    terms. The issue becomes clearer if it is realized that operant and
    respondent conditioning are operationally defined procedures of
    behavioural modification, whereas operant and respondent responses are
    descriptions of two classes of behaviours, both of which are
    modifiable through conditioning.

    3.2.1  Respondent behaviour

        Respondent behaviours are those that are reliably elicited by a
    specific observable stimulus. Two major features of a respondent
    behaviour are: (a) its occurrence depends on the frequency of
    occurrence of the eliciting stimulus; and (b) its consequences do not
    affect its frequency, or affects it only to a minor extent.

        Respondents frequently take the form of simple or complex reflexes
    and typically involve smooth muscles, glandular secretion, autonomic
    responses, or environmentally-elicited effector responses. Examples

    are the auditory startle response (Hoffman & Fleshler, 1963),
    olfactory responses, such as homing behaviour (Gregory & Pfaff, 1971),
    visually guided responses, such as optokinetic nystagmus (King &
    Vestal, 1974), and responses elicited by somaesthetic cues, such as
    negative geotaxis (Alder & Zbinden, 1977).

        Respondent behaviour can be quantified in a number of ways,
    including: (a) latency of the response from onset of the stimulus;
    (b) stimulus intensity required to elicit the response (threshold);
    and (c) magnitude of the response. Response magnitude can be measured
    either directly (e.g., force, duration) or by the use of rating scales
    (Irwin, 1968). A respondent generally occurs in close temporal
    contiguity with its eliciting stimulus, and its occurrence is
    independent of stimulus parameters such as duration, intensity, and
    frequency.

        Various types of respondent behaviour, both unconditioned and
    conditioned, have been used in behavioural toxicology (Irwin, 1968;
    Pavlenko, 1975). Several features of unconditioned respondent
    behaviour argue for its use. Perhaps the most important features are
    that: (a) stimulus-response relations are well defined and easily
    controlled; and (b) such methods require no prior training of the
    animal and, therefore, are easily and rapidly administered.

        The use of unconditioned respondent behaviour has generally been
    focused on 2 response classes:

        (a) reflexes, in which the response is limited to a specific
            effector system, such as skeletal muscle (motor response) or
            smooth muscle (autonomic response); and

        (b) taxis, in which the whole animal orients itself towards or
            away from a particular stimulus.

    Reflexes have been more extensively studied. An example is the
    acoustic startle response, which occurs following an intense auditory
    stimulus. A principle motor component of this response-forelimb
    extension can be readily measured. The acoustic startle response has
    been used widely in the study of drugs (Davis, 1980) and more recently
    in the study of other neurotoxic substances (Reiter et al., 1980;
    Squibb & Tilson, 1982).

        Unconditioned respondent behaviour has also found specific
    application in the study of effects of toxic substances on the
    developing nervous system. Developmental profiles for many types of
    unconditioned respondent behaviour have been well described for the
    rat (Altman & Sudarshan, 1975) and mouse (Fox, 1965). Because normal
    development of these responses is very closely timed, measurements of
    their developmental time-course have been widely used as an index of

    nervous system development. A number of investigators have used
    measurement of this behaviour as an index of toxicity (Rodier, 1978;
    Butcher & Vorhees, 1979).

        Respondent behaviour can also be conditioned using the classical
    techniques initially described by Pavlov (1927). This involves the
    pairing of a previously neutral stimulus (one that does not normally
    elicit a response) with an eliciting stimulus. Through repeated
    pairings, the neutral (conditioned) stimulus comes to elicit a
    conditioned response. Respondent conditioning is illustrated in the
    following example: when food (unconditioned stimulus) is placed in a
    dog's mouth, it elicits salivation (unconditioned response). If a tone
    (conditioned stimulus) is sounded just before food is placed in the
    dog's mouth, the tone itself eventually comes to elicit salivation
    (conditioned response) in the absence of food presentation. Although a
    variety of different controlling stimuli can be used in classical
    conditioning, responses are limited to those for which there is an
    initial unconditioned eliciting stimulus. The conditioned response can
    be quantified in terms of its latency, magnitude, and frequency of
    occurrence. If the conditioned stimulus occurs repeatedly in the
    absence of the unconditioned stimulus, the conditioned response
    becomes progressively weaker and eventually disappears; this process
    is called extinction.

        Since movement is the basic measurement of behaviour, the majority
    of conditioned reflex procedures evaluate motor responses. The
    acquisition of a conditioned response can be studied by gradually
    producing it. This requires careful control over external conditions,
    particularly the intensity and timing of both the conditioned and
    unconditioned stimuli. Since the rate of acquisition of a conditioned
    response depends on the functional status of the nervous system, such
    procedures are useful in the study of behavioural toxicity.

        Pavlenko (1975) reviewed the extensive use of conditioning by
    behavioural toxicologists in the USSR. Classical conditioning methods
    were divided into 2 groups. The first group, "defence-motor reflexes",
    generally involved a motor response to an electrical stimulation
    applied to the skin. The second group, "alimentary-motor reflexes",
    generally involved the response of a hungry animal to the presentation
    of food. Soviet investigators have frequently studied the rate of
    acquisition of such conditioned responses in the presence of toxic
    substances. According to Pavlenko (1975), small dosages of toxic
    agents more readily affect the acquisition of conditioned responses
    than the performance of firmly established conditioned responses
    (Shandala et al., 1980).

    3.2.2  Operant behaviour

        Behaviour that appears to occur in the absence of an eliciting
    stimulus is referred to as an emitted or operant response. Operant
    responses are movements of the organism that operate on (or change)

    the environment. Although these responses may occur in the presence of
    many environmental stimuli, they are not readily associated with an
    identifiable eliciting stimulus, and their occurrence is controlled
    mainly by their consequences. However, some responses are known to
    include both respondent and operant components. The best-known example
    is provided by bird pecks, which are controlled partly by eliciting
    stimuli and partly by response consequences, apparently in relation to
    their consummatory and non-consummatory functions, respectively.

        Emitted behaviour generally occurs with a close temporal
    relationship to the deprivation and presentation of particular
    environmental conditions, regardless of whether the deprivation
    produces an obvious physiological change. For example, an animal given
    access to a novel environment will show a characteristic temporal
    pattern of "exploratory" activity, with initial high levels of
    activity diminishing to low levels. Availability of the novel
    environment is associated with motor activity, but the novel
    environment is not an eliciting stimulus. Under these conditions,
    operant behaviour can be studied by observing all or part of the
    animal's behaviour during a specified period of time.

        Various rating scales have been used to quantify the frequency of
    occurrence and/or magnitude of selected operant responses (Irwin,
    1968). A more detailed approach has been to quantify the frequency,
    duration, and temporal patterning of selected emitted responses using
    time-lapse photographic analysis (Norton, 1973).

        Operant or instrumental conditioning refers to the modification of
    an operant behaviour by the control of its consequences. The following
    example illustrates operant conditioning: a food-deprived rat is
    placed in a chamber equipped with a food dispenser and with a lever
    projecting from a wall. If depression of the lever is followed by
    presentation of food, there is an increased likelihood that this
    response (lever press) will occur again. That is, the consequences of
    behaviour (e.g., receipt of food) come to control the occurrence of
    the response (pressing the lever).

        In contrast to classical conditioning, in operant conditioning,
    mere temporal contiguity between stimulus events is not sufficient for
    learning to take place and it is the consequences of behaviour that
    control the learning process. This concept was first introduced by
    Thorndike (1932) in his "Law of Effect," which dealt with the nature
    of events that can control operant behaviour. When the occurrence of
    an event following a response increases the probability that the
    response will occur again, the event is termed a reinforcing stimulus
    or reinforcer. The presentation of the reinforcing stimulus is termed
    reinforcement. Events that serve as reinforcers when presented (e.g.,
    food) are termed positive reinforcers; events that reinforce when
    terminated (e.g., electric shock) are termed negative reinforcers. If,
    on the other hand, the probability of occurrence of a response is

    decreased by its consequences, the consequences are termed punishment.
    An operant is defined as the properties of behaviour upon which
    reinforcement (or punishment) is contingent. In the previous example,
    for instance, pressing the lever is the operant. There are, of course,
    many ways that the rat can press the lever during operant
    conditioning. Nevertheless, each press is considered to belong to a
    single response class: the operant.

        Some response outputs, such as the bird pecks mentioned above, can
    be modified by experience as a function both of stimulus-response
    contingencies (so-called autoshaping) and response-reinforcement
    contingencies. In other words, both classical and operant conditioning
    can contribute to the changes observed in a particular response
    output, and this may have to be taken into account when assessing
    treatment effects.

        One of the strong features of operant conditioning is the broad
    range of behaviour that can be controlled and the new responses that
    can be generated. In practice, the responses most commonly selected
    for study meet four basic criteria: (a) they are easily identified and
    readily counted; (b) they are easily recorded with automated
    equipment; (c) their emission requires little time; and (d) they are
    readily repeatable (Kelleher & Morse, 1968).

        Reinforcement need not accompany every response in order to
    maintain that response. More commonly, reinforcement occurs
    intermittently and according to a schedule defining a sequential
    and/or temporal relationship between the response and its
    reinforcement. Schedules of reinforcement are of critical importance
    in determining both an organism's rate and pattern of responding, and
    the effects of a chemical (Kelleher & Morse, 1968). Different
    schedules of reinforcement have been shown to generate characteristic
    patterns of behaviour between species, even when using a wide variety
    of response topographies and reinforcing stimuli. Through the
    systematic development of schedules of reinforcement (Ferster &
    Skinner, 1957), operant conditioning has become an important tool in
    behavioural pharmacology and, more recently, in behavioural toxicology
    (Laties, 1982).

    3.3  Test Methods

    3.3.1  General attributes of behavioural methods

        Behavioural toxicity test methods vary in sensitivity,
    specificity, validity, replicability, and cost.  These factors, in
    combination with the "question(s) to be asked" at a particular stage
    of testing, will strongly influence the choice of method. It is also
    important to take into account the extent to which the results of such
    tests may be applicable to human beings.

    3.3.1.1  Sensitivity and specificity

        Sensitivity refers to the ability of a test method to detect the
    occurrence of a toxicant-induced behavioural change. If the objective
    of a study is to detect whether a chemical produces behavioural
    toxicity, a test with moderate sensitivity may be sufficient,
    particularly when there is no limitation on dose range. However, if
    the intention is to define threshold levels of exposure associated
    with behavioural effects, sensitivity becomes extremely important.

        Specificity has been used in several contexts. In one, it refers
    to the ability to give a negative finding when no behavioural effect
    has occurred. In another, it refers to the number of nervous system
    functions it reflects. Some tests are relatively specific. Many are of
    limited specificity in that they reflect changes in a number of
    nervous system functions (which, in turn, may reflect changes in other
    organ systems). In this context, specificity is interrelated with
    validity (section 3.3.1.2). Since behavioural tests rely on motor
    performance, all lack specificity to a certain extent.

    3.3.1.2  Validity

        Validity is concerned with whether a chemically-induced change in
    a behavioural response reflects only the change(s) in the behaviour to
    be measured. If, for example, a test is used to assess learning, the
    possibility must be considered that factors other than learning (e.g.,
    motivational levels, motor function) may influence the test results.
    Some understanding of test validity is essential for the proper
    interpretation of behavioural toxicity test results. The use of a
    variety of behavioural tests, thereby determining a behavioural
    profile, can be useful in establishing the validity of a particular
    test.

    3.3.1.3  Replicability

        Replicability refers to the ability of a test method to give
    consistent results in repeated studies (both within and across
    laboratories); it is somewhat analogous to precision. Precise
    protocols and strict attention to experimental detail will lessen
    variability and, therefore, increase replicability.

    3.3.1.4  Costs

        The costs of performing different types of behavioural analyses
    will have some influence on test selection. During the early stages of
    testing, when the focus is on detecting the presence of behavioural
    toxicity, cost will considerably influence the choice of method. It is
    unreasonable to use a highly sophisticated procedure requiring
    expensive equipment to detect whether a compound has potential

    behavioural toxicity whenever a cheaper alternative is available. In
    contrast, studies to determine the no-observed-adverse-effect levels
    of a chemical may justify the use of expensive, time-consuming
    methods.

    3.3.2  Primary tests

        Tests can be divided into primary and secondary categories.
    Primary tests are used for screening neurotoxicity, but such tests
    must forecast the potential of such chemicals to produce effects.
    Secondary tests are used to further characterize the nature of these
    effects. Since many functions are mediated by unique neural substrates
    and many chemicals produce rather selective effects, it is important
    to employ a variety of tests that measure different behavioural
    functions.

    3.3.2.1  Functional observation battery

        Functional observation batteries are designed to detect major
    overt neurotoxic effects. A number of investigators have proposed
    series of tests that are generally intended to evaluate various
    aspects of behavioural, neurological, and autonomic status (Irwin,
    1968; Pavlenko, 1975; Pryor et al., 1983). These batteries consist of
    series of semiquantitative measurements appropriate for the initial
    level of behavioural assessment (e.g., tremor, convulsions, ataxia,
    autonomic signs, paralysis). The tests are, in effect, rating scales
    concerning the presence or absence (and, in some cases, the relative
    degree of presence) of certain reflexes. In addition, eating and
    drinking behaviour and body weight should also be considered in the
    context of primary behavioural assessment.

        The major advantages of these tests are that they can be easily
    administered and can provide some indication of the possible
    functional alterations produced by exposure. Potential problems
    include insufficient interobserver reliability, difficulty in defining
    certain measures (e.g., stupor), and the tendency towards subjective
    bias. As a consequence, it is essential that observations be carried
    out by individuals who are blind to the groups.

        Many types of screening tests are currently used to assess the
    effects of neurotoxic agents on motor and reflex function. The
    simplest of these include observational assessments of body posture,
    muscle tone, equilibrium and gait, and righting reflexes (Irwin, 1968;
    Snyder & Braun, 1977). These tests are quantal or categorical at best,
    and are generally subjective. Larger animals permit a conventional
    neurological examination similar to that used with human beings
    (Abou-Donia et at., 1983).

    3.3.2.2  Motor activity

        Spontaneous motor activity in rodents has been extensively used in
    behavioural toxicology (Reiter, 1978; Reiter & MacPhail, 1979).
    Movement within the living space or environment is a high-probability
    response in animals and can be easily manipulated by environmental
    changes, including exposure to neurotoxic agents. Although seemingly
    simple, locomotor activity is very complex behaviour comprising a
    variety of motor acts, such as horizontally- and vertically-directed
    movement, sniffing, and grooming. Rating scales have been developed to
    fractionate locomotor activity into its relative components (Draper,
    1967). The measures used most often in behavioural toxicology are
    horizontally- and vertically-directed activity (Reiter & MacPhail,
    1979). A large variety of devices, automated and nonautomated, have
    been invented to measure motor activity. Following exposure to a
    neurotoxic agent, various qualitative and quantitative changes can be
    observed, depending on the apparatus that is used. For example, the
    figure-eight maze has been used extensively and successfully to detect
    effects produced by a number of chemicals (Reiter, 1983).

        Although the figure-eight maze has been used as a residential maze
    to measure toxic effects on diurnal activity patterns, recent research
    has almost exclusively employed shorter time intervals (Reiter, 1983).
    Elsner et al. (1979) have reported a method for the continuous
    monitoring of spontaneous locomotor patterns in rats. Using computer-
    assisted techniques, these investigators found that methylmercury
    treatment lowered activity during the night portion of the diurnal
    cycle.

        The complexity of motor activity is emphasized by the finding that
    low-level exposure to volatile organic solvents increases activity,
    whereas high level exposure decreases it (Horvath & Frantik, 1973).
    Positive results in a motor activity test usually require further
    testing to identify the precise function affected. Activity is not a
    unitary measure and a change in the frequency of this behaviour can
    reflect toxicant-induced changes in one or more sensory or motor
    functions, alterations in reactivity (excitability) or motivational
    states, or perturbations of a variety of regulatory states (e.g.,
    diurnal cycles, energy balance of the animal). For example, a decrease
    in activity might mean that the animal is paralysed or, perhaps, that
    it suffers from "general malaise". Thus, if a change in motor activity
    is observed, additional tests are needed to determine the cause.

    3.3.3  Secondary tests

    3.3.3.1  Intermittent schedules of reinforcement

        Performance generated by intermittent schedules of reinforcement
    (Ferster & Skinner, 1957; Reynolds, 1958; Kelleher & Morse, 1968;

    Schoenfeld, 1970) has played an important role in behavioural
    pharmacology and is proving a useful tool in behavioural toxicology
    (Laties, 1982).

        Most intermittent schedules involve reinforcement as a function of
    the number of responses emitted, some temporal requirements for
    emission of responses, or both. Ratio schedules require the animal to
    emit a fixed number of responses (fixed ratio or FR) or a number
    distributed around some average (variable ratio or VR) in order to be
    reinforced. As the ratio requirement is increased, the latency of the
    first response increases; however, once responding begins, it
    typically proceeds at a high and constant rate. Interval schedules
    require that a certain length of time should elapse before the
    response is reinforced. This may be a fixed time (fixed interval or
    FI) or time distributed around an average (variable interval, VI).
    Although only one response need be emitted at the end of the interval
    to cause reinforcement, the organism typically emits many responses
    during the interval. Interval schedules usually generate lower rates
    of responding than ratio schedules. The FI schedule generates a
    characteristic pattern of responding for which a variety of parameters
    potentially sensitive to disruption by neurotoxic agents can be
    analysed (Kelleher & Morse, 1968). Other commonly-used intermittent
    schedules specify the temporal spacing of responses. In the
    differential reinforcement of low rate (DRL) schedule, the organism is
    required to wait a specific time between responses in order to be
    reinforced. In the differential reinforcement of high rate (DRH)
    schedule, the organism is required to emit a specified number of
    responses within a specified (short) time, and thus responds at a high
    rate.

        Intermittent schedules of reinforcement can be combined to form
    more complicated "multiple" schedules of reinforcement. A classic
    example is the combination of FR and FI schedules presented in
    succession during a single test session; the resulting multiple
    schedule is termed a multiple FR-FI schedule. Each component of the
    multiple schedule is independent and occurs in the presence of a
    different external discriminative stimulus, which signals it. The
    schedule components are typically presented in alternating fashion,
    allowing the investigator to collect data on both types of behaviour
    almost simultaneously. Individual schedule components can be combined
    in other ways with different levels of complexity (Tilson & Harry,
    1982). Such schedules are not yet in general use in behavioural
    toxicology.

        The most common measure of performance with intermittent schedules
    is response rate (responses per unit time). This measure may be
    sufficient to establish whether there is an effect in a particular
    schedule or schedule component before grossly toxic levels are
    reached. A measure that may be more sensitive to the effects of toxic
    agents is the inter-response time (IRT) distribution (Schoenfeld,

    1970). For example, IRT distribution was a sensitive indicator of lead
    toxicity in rats performing on a multiple FI-FR schedule (Angell &
    Weiss, 1982).

        Another measure that may be a sensitive indicator of behavioural
    toxicity is variability in performance, both within and across
    sessions (Schoenfeld, 1970). Near the no-observed-adverse-effect
    level, variability between animals may be increased in the exposed
    group(s); variability between groups may be a sensitive indicator of
    toxicity. For example, Laties (1982) studied the effects of
    methylmercury exposure in the pigeon using a fixed consecutive number
    (FCN) procedure, in which the animal was required to respond on one
    key a specified number of times consecutively before responding on
    another key to be reinforced. Methylmercury exposure decreased the
    mean number of times pigeons responded on the first key before
    switching (run length) and increased the standard deviation of the run
    length within a session. Rice (in press) found increases in both
    within- and between-session variability in FI response rate in monkeys
    exposed developmentally to lead at doses insufficient to produce
    changes in rate.

        Marked differences in individual susceptibility to the effects of
    lead exposure on development have been observed using FI performance
    in the rat (Cory-Slechta & Thompson, 1979) and monkey (Rice et al.,
    1979). In general, lower doses increased, and higher doses decreased,
    the response rate. In the monkey, the IRT distribution remained
    different between treated and control groups, even though changes in
    rate disappeared after about 40 min. Remarkable differences in
    individual susceptibility were also observed on a response duration
    schedule (Cory-Slechta et al., 1981). In this schedule, the animal was
    required to depress a lever for a specified time (several seconds)
    before releasing it. Some of the exposed animals gave performances
    indistinguishable from those of the controls, while others performed
    much more poorly than controls. Differences in individual
    susceptibility may be a common phenomenon, particularly in behavioural
    toxicology, and should be considered in the determination of
    no-observed-adverse-effect levels (Dews & Wenger, 1979). An increase in
    group variability may signal the presence of toxicity in some portion
    of the population of responders (Good, 1979).

        Simple schedules, such as FR, FI, VI, DRL, and continuous
    avoidance, have been used to detect effects produced by a number of
    industrial and environmental toxic agents (Padich & Zenick, 1977;
    Dietz et al., 1978; Geller et al., 1979; Zenick et al., 1979; Leander
    & MacPhail, 1980; Alfano & Petit, 1981; McMillan, 1982). Perinatal
    lead exposure resulting in a relatively low body burden of lead
    produced effects in a DRH schedule in the absence of changes in motor
    activity (Gross-Selbeck & Gross-Selbeck, 1981).

        Multiple schedules offer the opportunity to study behaviour
    controlled by different variables, which may be differentially
    sensitive to the effects of toxic agents (this is known to be true for
    pharmacological agents). For example, toluene decreased the response
    rate in the FR component and increased the rate in the DRL component
    of a multiple schedule (Colotla et al., 1979); furthermore, the
    relative sensitivities of the two components were different. The
    multiple FI-FR schedule has proved particularly useful in detecting
    behavioural toxicity (Levine, 1976; Dews & Wenger, 1979; Leander &
    MacPhail, 1980; Angell & Weiss, 1982; McMillan, 1982).

        Intermittent schedules of reinforcement can be used to monitor
    effects other than, or in addition to, direct effects on the central
    nervous system. These may include damage to the peripheral nervous
    system or to some other organ system that results in general malaise
    (Laties, 1982). For example, acrylamide, an organic solvent that
    produces a "dying back" axonopathy, produced decreases in FR response
    rate (Tilson et al., 1980). The schedules typically produce high
    response rates and thus may be sensitive to impairment in motor
    function. Exposure to ozone resulted in decreased responding on an FI
    schedule; this was interpreted as resulting from general discomfort
    produced by ozone (Weiss et al., 1981).

        In addition to schedules that maintain behaviour by positive
    reinforcement, responding can be maintained by negative reinforcement.
    Avoidance schedules are either continuous (i.e., each response
    postpones shock by a fixed amount of time) or discrete-trial (each
    shock is preceded by a warning signal during which a response will
    prevent punishment). Avoidance schedules have been used extensively in
    the study of anticholinesterase compounds, including pesticides, in
    order to assess dose-response relationships, the time course of
    behavioural depression during acute intoxication, and the effects of
    repeated exposure (Bignami et al., 1975). Their usefulness in the
    assessment of treatment effects is shown in the extensive drug
    literature. In fact, the confounding of treatment effects with
    "motivational" changes is postulated to be less of a problem with
    avoidance tasks than with appetitive tasks (positive reinforcement).
    Furthermore, once acquired, avoidance responses can be maintained at
    fairly stable levels without the precautions necessary in the study of
    appetitive responses (e.g., daily control of food or fluid intake
    and/or body weight), though induction of stress responses and changes
    in pain sensitivity must be considered.

    3.3.3.2  Motor function

        A variety of techniques developed to evaluate motor function have
    been used; these include performance on a rotating rod or treadmill
    (Frantik, 1970), swimming to exhaustion (Bhagat & Wheeler, 1973), or
    suspension from a horizontal rod (Molinengo & Orsetti, 1976). One
    increasingly used technique is the quantification of hindlimb splay.

    Edwards & Parker (1977) inked the feet of rats, dropped the animals
    from a specified height, and measured the distance between the digit
    marks. Schallert et al. (1978) used a similar technique of inking 
    paws to evaluate abnormal gait in rats treated centrally with
    6-hydroxydopamine.

        A negative geotaxis procedure was used by Pryor et al. (1983) to
    evaluate neurotoxic agent-induced alteration in motor coordination. 
    Reduction in grip strength is a frequently reported neurological sign
    in human beings; fore and hindlimb grip strength in rats and mice has
    been quantified using commercially available strain gauges (Meyer et
    al., 1979).

        Tremor is a common neurotoxic effect. A number of rating scales
    and semiquantitative procedures to measure tremor are available
    (Gerhart et al., 1982). A simple but expensive spectral analysis
    technique that permits rapid evaluation of tremor in freely moving
    animals has been reported by Gerhart et al. (1982).

        Other more complicated techniques have been devised to measure
    motor deficits in laboratory animals. For example, Falk (1970) trained
    animals to press a lever within a designated range of force for a
    given period. Falk and others (Fowler & Price, 1978) used this
    procedure to study the effects of toxic agents on fine motor control.

    3.3.3.3  Sensory function

        Exposure to toxic chemicals can cause a wide range of sensory
    effects. Alterations in sensory processes, such as paraesthesia or
    visual or auditory impairment, are frequently among the first signs of
    toxicity in human beings exposed to toxic agents (Damstra, 1978). In
    animals, "psychophysical" methods are used to arrive at some
    estimation of differential response in the presence of a stimulus
    varied across some physical dimension (Stebbins, 1970). The great
    majority of psychophysical studies have been carried out on non-human
    primates and birds; ideally, such studies should be conducted on
    species in which sensory function closely resembles that of human
    beings. Psychophysical methods range from those that assess a gross
    loss of sensation to those that provide a sensitive and precise
    analysis of changes in threshold levels and other ancillary or complex
    sensory phenomena.

        One of the least complex approaches to the study of sensory
    deficits is based on the localization or orientation response.
    Marshall and colleagues (Marshall & Teitelbaum, 1974; Marshall et al.,
    1974; Marshall, 1975) have described a battery of observational tests
    in which a visual, auditory, olfactory, or dermal stimulus is
    delivered to the organism. The presence or absence of a localization
    or orientation response to the source of this stimulus is then
    recorded. Such techniques have been used to demonstrate sensory

    inattention as well as hyperexcitability in rats having lesions in
    various regions of the brain. Pavlenko (1975) has described a variety
    of stimulus-elicited orientation reflexes used in the USSR. Despite
    the fact that observational tests are simple to perform, they are
    labour intensive, especially if the necessary inter-rater reliability
    scales are used. However, the scoring of the tests is frequently
    subjective and necessitates testing under "blind conditions." Finally,
    the data are usually quantal (i.e., the response is scored as either
    present or absent) or categorical (scored on a rating scale). Thus,
    interpretation of the results is difficult, particularly in repeated-
    measure designs.

        Several attempts have been made to develop simple yet objective
    tests for sensory dysfunction in rodents. Some investigators have used
    measurement of the acoustic startle reflex (i.e., measurement of the
    presence (and magnitude) or absence of response to a novel sound or
    tone) as a screen for auditory dysfunction. Pain sensitivity can be
    assessed using standard psychopharmacological techniques measuring
    reaction times to a noxious stimulus (Pryor et al., 1983). Electrical
    stimulation of the tooth pulp is a sensitive method to detect change
    in pain sensitivity (Costa & Murad, 1969). The flinch-jump technique
    also is used extensively to determine changes in pain threshold
    (Evans, 1962). Taste reactivity has been assessed using taste aversion
    procedures (Kodama et al., 1978). Depth perception has been assessed
    using a visual cliff procedure, which measures whether or not an
    animal chooses to step onto a nearby platform or floor ("shallow"
    floor) in preference to one that may be perceived as more distant
    ("deep" floor) (Sloane et al., 1978). Another simple test of visual
    function is the optokinetic drum, which relies on the optokinetic
    nystagmus or optomotor response (i.e., tracking a moving object with
    the eyes and head for a certain distance until the head is
    repositioned back into the frontal plane). On the basis of this
    measure, a procedure was developed that was believed to assess visual
    acuity as well as colour vision (Wallman, 1975).

        Changes in reflex response have been monitored in a variety of
    ways. The ability of a stimulus to inhibit a reflex response has been
    used to detect changes in auditory threshold in the rat (Young &
    Fechter, 1983). Reflex responses (e.g., front-paw withdrawal or ear
    flexion) following electric shock have been used extensively in the
    USSR to determine changes in pain threshold or detection of
    electromagnetic fields (Speranskij, 1965).

        More complicated paradigms have been used to assess sensory
    dysfunction in more precise ways. Mazes and maze-like apparatus appear
    to have some use for evaluating certain sensory deficits (e.g., visual
    or somesthetic) in rats (Overmann, 1977; Post et al., 1973). However,
    as Evans (1978) pointed out, the relative contributions of motor and
    higher-level functions should be carefully distinguished when
    interpreting results of such studies.

        Some of the more precise methods for evaluating subtle sensory
    deficits involve operant techniques.  In these studies, a response
    (e.g., pressing a lever) is maintained by food or electric shock. Once
    the animal has learned to make the response only under certain
    stimulus conditions, the intensity of the stimulus can be varied and
    the response determined as a function of the intensity. Such
    techniques have been used to show auditory loss following exposure to
    kanamycin (Harpur & d'Arcy, 1975; Chiba & Ando, 1976). Merigan (1979)
    used reinforcement of the identification of spots of light by a monkey
    with a fixed gaze to demonstrate the presence of scotomas following
    methylmercury exposure.

        Olfactory (Wood, 1978), shock or pain (Weiss & Laties, 1961), and
    vibration (Maurissen, 1979) thresholds have been determined using
    operant techniques. Operant methods also have been used to study the
    effects of toxic exposure on more complex sensory phenomena, including
    light flicker discrimination (Schechter & Winter, 1971), critical
    flicker frequency (Merigan, 1979), and discrimination of the duration
    of a visual or auditory stimulus (Johnson et al., 1975).

        Chemical agents can act as reinforcing and internal discriminative
    stimuli and gain control of a variety of behavioural responses. In
    fact, Wood (1979) demonstrated the abuse potential of toluene in an
    operant paradigm analogous to drug self-administration.

        Pryor et al. (1983) reported on the use of a relatively simple
    psychophysical technique to assess 3 sensory modalities concurrently
    in the same animal. In this procedure, rats learned to climb or pull a
    rope to avoid a noxious electric footshock. Eventually, the response
    was brought under the control of three conditioned stimuli; a tone, a
    low-intensity, nonaversive current on the floor, and a change in
    intensity of the chamber house light. Various intensities of each
    stimulus were presented, permitting generation of a quasipsycho-
    physical response function. Once the animals were trained, they were
    exposed to toxic agents and the changes in responding were measured.

        Another procedure that shows promise in the assessment of sensory
    function is the use of prepulse inhibition of the acoustic startle
    reflex. By varying the intensity of the prepulse stimulus, treatment
    effects on sensory acuity can be assessed. Such a procedure has been
    used to study the effects of triethyltin on auditory functions in
    experimental animals (Young & Fechter, 1983).

    3.3.3.4  Cognitive function

        Within general psychology, "cognition" refers to the processes by
    which knowledge of surroundings is gained; perception, thinking,
    learning, and memory refer to different aspects of cognitive
    processes. The attribute "cognitive" has been extended by analogy to
    classify animal behaviour, as well. Tolman (1948) used the concept of

    "cognitive maps" to account for the fact that rats are able to acquire
    and retain information about spatial relationships in experimental
    mazes, and to use this "latent" information for subsequent learning.
    Another example of the study of cognitive processes in animals are the
    "insight" studies on chimpanzees (Kohler, 1976), in which behaviour
    resembling human problem-solving has been demonstrated. Thus,
    cognitive functions cover a much broader field than just approach-
    avoidance learning (the prevailing paradigm in today's behavioural
    toxicology). In practice, however, learning and memory are the
    cognitive functions that have received particular attention in animal
    studies, because they are amenable to quantification and because the
    abilities to learn and to remember have obvious adaptive value for an
    organism. The capacity to learn permits an organism to escape or avoid
    situations, approach desirable objects, and store these contingencies
    for future use.

        Behavioural toxicologists have used a variety of experimental
    paradigms to assess learning and memory in laboratory animals. A few
    procedures have been developed to measure the ability to adjust to a
    new contingency once an initial task has been learned.  Studies have
    involved conditioned respondent as well as conditioned operant
    behaviours.

    (a) Procedures using negative reinforcement

        In passive avoidance techniques, the animal is trained to withhold
    a response, to avoid punishment. A standard procedure is to put the
    animal on the lighted side of a shuttle box and to shock it as it
    enters the dark compartment (Wolf, 1976). After a given interval of
    time, the animal is returned to the apparatus and is given no shock.
    Dependent variables include the initial response latency, subsequent
    response latencies, number of positive responses, number of
    compartment crossings, and total time spent in either compartment on
    retesting for a fixed interval of time. Walsh et al. (1982b) used
    one-way passive avoidance to demonstrate trimethyl tin-induced memory
    deficits in adult rats in the absence of changes in sensitivity to
    shock. This technique was used to demonstrate learning/memory deficit
    in rats exposed neonatally to chlordecone (Mactutus et al., 1982).

        Passive avoidance techniques have the advantages of speed, ease of
    performance, and low cost. They have the disadvantage of producing
    highly variable results if performed under inadequate test conditions
    or if the appropriate retention intervals are not used. Finally, it is
    imperative that the nonassociative variables mentioned above be
    measured and that alterations in motivational factors (e.g., changes
    in "pain thresholds" to footshock) be evaluated.

        One-way active avoidance tasks require the animal to respond in
    order to escape or avoid negative reinforcement. Typically, an animal
    is placed in one compartment of a shuttle box, where it can be
    shocked. Once the active avoidance or escape response to the

    neighbouring compartment has been registered, the animal is returned
    to the original compartment and the process is repeated. Using this
    type of test, Tilson et al. (1979a) reported that rats with long-term
    exposure to chlordecone learned to avoid as rapidly as controls but
    displayed a marked retention deficit when retested several days later.

        Another variant of the shock-motivated learning task is the
    two-way shuttle box paradigm. In this procedure, rats learn to shuttle
    from one compartment to another in the presence of a warning signal in
    order to escape or avoid electric footshock. Unlike one-way avoidance,
    the animals must learn to return to a compartment where they have just
    been shocked. Interesting differences in effects can be observed
    between one- and two-way procedures. For example, Sobotka et al.
    (1975) reported that rats exposed neonatally to lead performed as well
    as controls in a one-way shock avoidance test but displayed
    significant deficits in a two-way paradigm.

        In general, one- and two-way avoidance tasks entail one training
    trial or several discrete massed training trials. These are followed
    by one or more trials in which retention is assessed. Dependent
    variables for both types of avoidance tasks are avoidance latencies,
    number of correct responses, and number of trials to a predetermined
    criterion of learning. A useful measure of activity is number of
    inter-trial crossings; this information can aid in interpretation of
    treatment effects. Tilson et al. (1982a) reported a triethyl lead-
    induced facilitation of two-way shuttle-box performance that was not
    associated with altered motor activity (inter-trial crossings) or
    flinch-jump threshold.

        The symmetrical Y maze is a somewhat more complete learning task
    than either one- or two-way avoidance. In this procedure, a light or
    tone is activated in one of two arms of a maze not occupied by the
    animal. The animal is given a predetermined amount of time to run to
    the proper arm to avoid electric shock. Dependent variables include
    all those previously mentioned for avoidance procedures as well as
    number of correct choices. The Y maze has been used very successfully
    by Vorhees (1974) to study learning ability in rats exposed in utero
    to vitamin A. The paradigm involves two types of learning: when to run
    and where to run. These may be affected differently by treatment (Ray
    & Barrett, 1975). As pointed out in section 2, strain differences must
    also be considered: rats of the Fischer strain easily learn when and
    where to avoid, but Sprague Dawley rats do not readily learn when to
    avoid (Tilson & Harry, 1982).

        Experimental paradigms that involve other tasks and types of
    negative reinforcement have been used to assess learning. In the water
    maze, animals are placed in a maze filled with water and are required
    to learn a series of correct turns in order to gain access to an exit
    ramp. Learning trials are preceded by straight-channel swimming trials
    as an adaptive procedure and to determine if there are any measurable

    neuromotor deficits. Frequently, initial learning is tested in 6
    trials on day 1, retention is tested after 7 days, and then reversal-
    learning is tested by changing the position of the escape ramp.
    Performance measures are swimming time, number of errors, and number
    of correct choices. Water maze studies have revealed learning deficits
    in animals exposed developmentally to vitamin A (Vorhees et al., 1978)
    and to lead or mercury (Brady et al., 1976; Zenick et al., 1979).
    Vergieva & Zaikov (1981) studied the performance in a water maze of
    adult rats after short- and long-term inhalation of styrene.

    (b) Procedures using positive reinforcement

        Learning and memory can also be assessed in paradigms that use
    positive reinforcement. A number of techniques involve positive
    reinforcement of discrimination tasks. The type of discrimination can
    be spatial or sensory (visual, auditory). Spatial discrimination tasks
    usually involve simple T mazes or more complicated versions of the
    T maze, such as the Hebb-Williams maze, which actually is a sequence
    of successive T mazes. Dependent measures are the number of correct
    trials or the time needed to reach the goalbox. Snowdon (1973) found
    that rats exposed to lead neonatally showed impaired performance in a
    Hebb-Williams maze, whereas rats exposed post-weaning or as adults did
    not.

        Visual discrimination tasks frequently take the form of
    simultaneous two-choice pattern discrimination. Dependent variables
    usually are the number of correct discriminations or number of trials
    needed to reach a predetermined criterion. Winneke et al. (1977)
    trained food-deprived rats to make a discrimination based on either
    the size (circle size) or orientation (horizontal versus vertical
    stripes) of a cue placed on a door in a maze leading to food
    reinforcement. Lead-exposed animals took longer to acquire the more
    difficult size discrimination but resembled controls in learning the
    orientation-cued response. Since visual deficit, as assessed by visual
    evoked potentials, did not occur until blood-lead levels exceeded
    400 g/litre (Winneke, 1979), visual dysfunction could be ruled out as
    an alternative explanation for the lead-induced learning deficit,
    which occurred at blood-lead levels below 200 g/litre (Winneke et
    al., 1983).

        In a similar two-choice visual discrimination task (Tilson et al.,
    1982b), animals were trained to make a nose-poke response to the side
    of a cue panel that contained a visual cue. After an animal had
    learned the correct discrimination (which occurred over a period of
    days), the contingency was reversed, i.e., a response to the side in
    which the cue lamp remained unlit was reinforced. Rats exposed to
    chlordecone during development showed a trend towards altered
    acquisition performance and were markedly different in the way that
    they responded during the reversal phase of this test.

        Visual discrimination learning has also been used to study the
    effects on monkeys of lead exposure during the developmental period
    (Bushnell & Bowman, 1979; Rice & Willes, 1979). In these studies,
    lead-induced deficit was observed only in the more demanding
    discrimination reversal paradigm. This constitutes evidence for
    "silent" toxicities that become apparent only with tasks of increased
    complexity. Similar conclusions can be drawn from studies on rats in
    which lead-induced learning deficit was demonstrated for difficult but
    not simple discrimination tasks (Winneke et al., 1977, 1983).

        The previous measures of learning have involved between-subject
    designs that can require large numbers of animals, if there is large
    inter-animal variability. A within-subject design enables the effects
    of chemical agents to be evaluated with reference to the animal's
    stable pre-exposure baseline and so controls for inter-animal
    variability. One such paradigm is repeated chain acquisition: an
    animal is given a series of 4 buttons that must be pressed in a
    specific order. The order is changed each day, so that a new but
    generally similar repeatable task is presented to the animal each day.
    In this paradigm, carbaryl affected the rate of performance of
    monkeys, but errors were not as clearly affected (Anger & Setyes,
    1979). A similar method was used to detect alteration of response
    patterning by lead (Dietz et al., 1978) and mercuric chloride (Leander
    et al., 1977). In addition to studies of the effects of chemical
    agents on the acquisition of new behaviour, there have been studies of
    memory, and the persistence or lack of persistence of acquired
    information with the passage of time. Memory has been studied using
    between-subject designs employing a radial maze.

        The radial-arm maze (RAM) is a complex spatial learning task in
    which animals must "remember" a list of previously entered and
    unentered feeders during a free-choice test session (Olton et al.,
    1979, 1980). The RAM is believed to be useful in studying working
    (information relevant to a single trial), as well as reference memory
    (information relevant to all trials). The most commonly used RAM
    consists of a circular arena from which eight equidistant arms radiate
    like spokes from a wheel. A trial begins with all arms baited and ends
    when all pellets have been consumed or when a fixed period of time has
    elapsed. The most effective strategy for solving the maze is to enter
    and eat in each of the arms only once. Results can be expressed in
    several ways such as: (a) number of correct choices in the first eight
    selections (control rats will often obtain all eight pellets without
    an error); (b) total number of errors made in obtaining the eight
    pellets. Walsh et al. (1982b) reported that trimethyltin-exposed rats
    displayed impaired performance in this task and that the behavioural
    deficit might be due to an alteration in the integrity of limbic
    forebrain structures such as the hippocampus. The RAM has also been
    used to study loss of spatial memory as a function of isolation in
    darkness and of time elapsed since initial learning (Buresova & Bures,
    1982).

        Memory has been studied using operant discrete trial techniques,
    which enable greater control over the stimuli applied in this study
    and thus more sensitivity, but require more training. Animals are
    trained to respond in a series of trials that are separated by time
    intervals. Performance in a trial depends on information presented in
    the previous trial. These operant techniques always use within-subject
    designs and have proved useful in the study of drugs because they make
    it possible to dissociate the effects on memory from the effects on
    motivation or motor control. For example, it has been shown that
    scopolamine impairs memory but does not interfere with responding in
    delayed, go-no go, alternation (Heise & Milar, 1984). This technique
    was applied in behavioural toxicology using delayed, spatial
    alternation, which combined two toxicologically sensitive tasks,
    discrimination and reversal spatial memory (Heise, 1983). It was found
    that carbaryl decreased both memory and responding.

        Modifications of Harlow's "learning set formation paradigm" for
    use with primates also deserves mention in this context. In this task,
    the animal is given a great number of discrimination problems to be
    solved successively. As training progresses, new problems are solved
    faster and faster (i.e., the number of trials necessary to solve each
    problem decreases with increasing number of problems). Lilienthal et
    al. (1983) used this task to demonstrate lead-induced cognitive
    deficit in rhesus monkeys; simple discrimination learning was not
    affected, but transfer of learning was. Thus, acquisition of
    information was not impaired but memory for previously acquired
    information was disrupted.

    3.3.3.5  Eating and drinking behaviour

        Many of the behavioural tests described above use food as the
    primary reinforcer to get the animal to perform an instrumental
    response. Eating is thus involved in the results of the various tests
    employed. However, eating can also be used in the assessment of the
    potential behavioural toxicity of chemical compounds. Eating and
    drinking are naturally occurring behaviour that can be measured in the
    animals' laboratory environment; once a stable eating and drinking
    pattern is established, toxic agents can be introduced and the
    resulting alteration in behaviour measured (Tilson & Cabe, 1978). For
    example, since it has been shown that carbon monoxide and hypoxia
    depress food intake, Annau (1975) compared the effects of both
    conditions on food and water intake in naive rats. Two control groups
    were compared with groups exposed to 250, 500, and 1000 ppm carbon
    monoxide and 16%, 14%, and 10% oxygen.  Although both experimental
    conditions produced a decrease in body weight and in food and water
    intake, the shapes of the resulting curves were very different,
    suggesting that carbon monoxide may not act on these biological
    systems in an identical manner to hypoxia; in fact, it appears that
    hypoxia has a more severe effect on behaviour than the equivalent
    concentration of carbon monoxide. A more recent investigation (Bloom

    et al., 1983) showed that subconvulsive doses of pyrethroid
    insecticides reduced variable interval performance as well as food
    intake, indicating that further attention should be devoted to the
    relationship between eating and drinking and other behavioural
    response changes.

        Eating responses in rodents also altered when a food having a
    specific taste was paired with an illness produced either by
    irradiation or by the administration of a toxic substance, thus making
    it possible to assess the aversive properties of the agent tested.
    This conditioned taste aversion has been employed as an experimental
    preparation in the evaluation of the unconditioned stimulus functions
    of several toxic substances, such as chlordimeform (MacPhail &
    Leander, 1982), trialkyltin (MacPhail, 1982), and industrial solvents
    (Vila & Colotla, 1981).

        Schedule-induced, or adjunctive, drinking is another type of
    consummatory behaviour (Colotla, 1981) of interest to the
    neurobehavioural toxicologist for two reasons; first, it has been
    demonstrated that the procedure can generate "voluntary" consumption
    of alcohol and several other drugs, and second, it produces a
    consistent and regular post-reinforcement behaviour pattern that is
    sensitive to the effects of several drugs (Colotla, 1981). Although,
    to date, no reports have appeared in the neurotoxicological literature
    on the effects of toxic substances on adjunctive behaviour, this type
    of eating response may be useful as part of a test battery for
    neurobehavioural toxicity.

    3.3.3.6  Social behaviour

        Social behaviour implies behaviour involving two or more
    individuals (Hinde, 1974), which means that almost all activities of
    an animal are, or can be, social. Since social behaviour represents a
    complex set of interactions, its investigation in the laboratory is
    not simple: each behavioural response of an animal living in a group
    (including a pair) may be influenced by previous interactions with
    members of the group and by the behavioural characteristics of
    individuals forming the group.

        Although considerable literature exists on the social behaviour of
    laboratory animals, additional methodological development is necessary
    before it can be included in routine toxicological evaluations.
    Methodological and theoretical problems include the definition of
    individual classes of social behaviour, and the generalization of
    results obtained with different species. There are problems dealing
    with the objectivity of measurement, standardization of testing
    procedure, and statistical evaluation. Use of laboratory animals
    (mice, rats) raises the question of adequacy of the social environment
    created by the research worker because the laboratory environment is
    very different from the social environment of the species in the wild.

        Although social behaviour has not been used extensively in
    toxicology, it has found some use as a diagnostic tool in
    psychopharmacology, where various kinds of social interaction have
    differentiated between the effects of drugs of different
    pharmacological classes (Miczek & Barry, 1976). This body of
    psychopharmacological work demonstrates that the behavioural effects
    of different chemicals can be modified by the social environment and
    thus can differ from the effects in isolated animals.

        Two basic approaches have been used to study social behaviour in
    animals. The first evaluates the impact of social setting on the
    various types of behaviour in animals living for long periods in semi-
    natural or artificial conditions. Under this test situation, housing
    conditions and population density can be shown to affect various
    behavioural and physiological responses. Litter size, for example, was
    shown to influence an animal's response to various stimuli; the effect
    depended not only on nutritional status but also on social behaviour
    (Frankova, 1970). Overcrowding has been shown to affect behaviour, as
    well as endocrine and other organ systems functions (Thiessen, 1964).
    Chemicals such as amphetamine have greater toxicity when administered
    to grouped mice (Chance, 1946). There are differences between
    individually-and colony-housed rats in the oral ingestion of morphine
    (Alexander et al., 1981), aversiveness to Naloxone (Pilcher & Jones,
    1981), and ethanol consumption (Kulkosky et al., 1980).

        A second more common approach to the study of social behaviour is
    the short-term observation of an animal's behavioural response to
    another individual (e.g., cage mate, sexual partner) or group. Here
    the pair or group is observed out of the home cage, usually in a
    specially equipped box. Various chemically-induced disruptions of
    different social interactions have been reported for rodents and
    primates. Silverman (1965) developed criteria for paired interactions
    in rats (e.g., mating aggression, submission, escape) and demonstrated
    the effects of drugs on these categories of social behaviour. Various
    other studies (Frankova, 1977; Krsiak, 1979; File, 1980) have
    evaluated effects of drugs on isolated and group-bound mice and rats.
    Cutler (1977) demonstrated that lead exposure disrupts social
    behaviour in mice.

        Other studies have investigated the effects of chemicals on the
    social behaviour of non-human primates. Apfelbach & Delgado (1974)
    administered chlordiazepoxide to gibbon colonies and observed a
    decrease in mobility and aggressive acts and increased play, grooming,
    and water intake. Bushnell & Bowman (1979) showed that long-term lead
    ingestion affected play and other social behaviour in infant rhesus
    monkeys.

        The types of social behaviour most frequently examined for effects
    of chemicals (especially drugs) are dominance and submission
    (Baenninger, 1966), isolation-induced aggression (Krsiak, 1979;
    Eichelman et al., 1981), sexual behaviour, and maternal behaviour
    (Grota & Ader, 1969; Frankova, 1971, 1977).

    3.3.4  Strengths and weaknesses of various methods

        Many of the test methods described in this section have been used
    successfully to model human neurotoxicity, while the relationship of
    findings of other tests to human disease is not known. Also, of
    course, each method or group of methods has its proponents; often the
    individuals or groups who have developed and used it extensively.
    However, over the past decade, there has been a broadening in the
    exchange of test methods between research laboratories and countries.
    Some obvious examples are the use in the USSR of T-maze and shuttle-
    box avoidance testing (Kholodov & Solov'eva 1971; Asabayev et al.,
    1972) and of unique test devices for teaching simple discrimination 
    using positive reinforcement (Kotliarevski, 1957; Masterov, 1974;
    Medvedev, 1975); the use in Eastern Europe and South America of
    operant techniques (Colotla et al., 1979; Vergieva & Zaikov, 1981);
    and the use in the USA and South America of reflex conditioning to
    elucidate pharmacological and toxicological effects (Costa & Murad,
    1969; Young & Fechter, 1983).

        It is probably true that investigators who have perfected the
    technical aspects of their own paradigms "get the most out of them."
    Behavioural science, as any other science, involves a set of
    techniques of such sophistication that many are best learned in the
    laboratories in which they were developed. This presents some
    difficulties in inter-country transfer of methods and the unrestrained
    endorsement of methods developed in another country with other
    behavioural/psychological traditions. Those who adopt a method for
    which they have no ready contacts with experienced investigators may
    abandon it as "insensitive", when the only problem may be technical
    errors in application.

        Respondent behaviours have the advantage of being rapidly formed.
    In most cases, only 40-60 trials or tests in a single day are required
    for their formation, and the behaviours may be elicited periodically
    during long-term exposure studies with no diminution in their
    usefulness. Pure muscle responses or integrated motor activity (as in
    a shuttle box or startle response chamber) can be measured. These
    methods can reflect both increases (stimulation) and decreases
    (depression) in responding, and can be completely automated, thus
    eliminating subjective factors.

        Operant methods can entail the kinds of complex behaviour needed
    to evaluate such complex factors as learning. Their proponents also
    point out that they can be used to measure fundamental behavioural

    properties related to the control of behaviour. However, operant
    paradigms require several weeks for the animals to reach stability on
    a schedule of reinforcement and a great deal of time of a highly
    skilled research worker to train animals to perform complex tasks. As
    with respondent methods, operant methods can detect both increases and
    decreases in responding and can be completely automated.

        In addition, it is axiomatic that behaviour is essentially an
    integration of sensory, cognitive, and motor processes. The
    specificity of a toxic effect on any behavioural parameters must
    always be evaluated in the context of the experimental design and in
    conjunction with controls for effects on the other systems.

    3.4  Research Needs

    3.4.1  Compensatory mechanisms

        Because the functional redundancy of the nervous system can mask
    some perturbations, procedures need to be devised and used that will
    reveal toxic effects that are not apparent under normal test
    conditions. Challenges by environmental and pharmacological agents
    have proved to be useful in the search for subclinical toxic effects,
    and their use is recommended. By creating additional demands on
    behavioural integration, challenges can reveal "hidden" functional
    deficits (Hughes & Sparber, 1978; Tilson et al., 1979b; Tilson &
    Squibb, 1982). The pharmacological challenges most commonly used are
    psychoactive drugs and substances that mimic or block the actions of
    putative neurotransmitters or alter their synthesis, storage, or
    release. Some common stresses that have been used as environmental
    challenges are alterations in circadian rhythms, density of housing,
    noise level, and ambient temperature (MacPhail et al., 1983). Studies
    to increase the understanding of subclinical toxic effects and provide
    information concerning appropriate rationale for the selection of
    specific challenges represent an important research need (Tilson &
    Mitchell, 1984).

        Adaptive changes in the nervous system can also occur after
    repeated exposures to high levels of a toxic agent. There is an
    initial decrement or increment in behaviour after the first exposure,
    but then there is a recovery of function to the pre-exposure levels as
    exposure continues. This apparent return to normal function results
    from adaptive changes in the systems that control behaviour. Sometimes
    this change is due to tolerance within the affected systems, but in
    other cases, there could be a redundant system that enables the return
    to the pre-exposure state.  Research on these forms of compensatory
    mechanisms is to be encouraged.

    3.4.2  Method development and refinement

        Development of methods, evaluation, and refinement remain basic
    needs in behavioural toxicology. Many behavioural test strategies and
    test batteries have been proposed for the longitudinal assessment of
    behavioural function in animals exposed to toxic agents either during
    development or as adults (Butcher, 1976; Grant, 1976; Rodier, 1978;
    Buelke-Sam & Kimmel, 1979; Butcher & Vorhees, 1979; Tilson et al.,
    1979c; Zbinden, 1981; Mitchell et al., 1982; Vorhees & Butcher, 1982).
    Most authors agree that it is essential that a core of functions
    should be assessed, since it is unlikely that any single one will be
    sensitive to all toxic agents. The guidelines for reproductive testing
    in France, Japan, and the United Kingdom require assessment of several
    functions but do not specify exact tests. Many fear premature
    standardization of specific tests while behavioural toxicology is
    still in its initial period of growth (Weiss & Laties, 1979).
    Development, validation, and standardization of primary (screening)
    tests need to continue and could be facilitated by the use of
    reference compounds (Horvath & Frantik, 1973).

        An easily obtained source of information that deserves more
    attention, especially in long-term studies, is the uninterrupted
    behaviour of animals in their home cages. Food and water intake can
    provide information on homeostatic functioning, while a measure of
    activity can assess changes in movement patterns or in the circadian
    rhythm of activity.

        There does not yet appear to be an animal model of fatigue.
    Further elaboration of animal models to evaluate fatigue dissociated
    from motivational factors needs consideration.

        It is generally accepted that complex tasks are more sensitive to
    chemical disruption than simple tasks, and that a behaviour that is
    not fully learned or practiced is more sensitive to chemical
    disruption than one that is well learned and established. These and
    many related assumptions require testing. Publication of skilfully
    selected comparative data from a variety of laboratories will assist
    all investigators in the future selection of test methods. In
    particular, more and better methods involving complex tasks that may
    be relatable to human beings are required.

        Increased and improved automation of test methods is another need.
    Automation can eliminate observer (and even trainer) bias, and should
    be pursued where feasible. The increasing availability of inexpensive
    microprocessors continues to put more and more laboratories in a
    position to acquire the central tool for automation of more complex
    methods.

    4.  NEUROPHYSIOLOGICAL METHODS IN NEUROTOXICOLOGY

    4.1  Introduction

        The term neurophysiology may refer to all studies of the function
    of the nervous system. As such, it could include studies of
    conditioning and behaviour as well as electrical recordings obtained
    from the nervous system. However, for the purposes of this section, a
    more restricted definition of neurophysiology will be used, with few
    exceptions, to mean the study, by measurement of electrical activity,
    of nervous system activity.

        Even with this restricted definition, an enormous array of
    neurophysiological methods is available to the neurotoxicologist.
    Depending on the methods selected, neurophysiological studies can be
    used to achieve goals as diverse as detecting neurotoxicity,
    characterizing neurotoxicity (i.e., which neural systems are involved)
    and unravelling mechanisms of neurotoxicity. Each of the methods
    selected for presentation below can be used to achieve at least one of
    these goals. In addition, some of the methods can be used in human
    studies as well as those on laboratory animals, thereby providing
    ready opportunity for cross-species extrapolation.

        While the use of physiological methods to assess the impact of
    toxic chemicals on the nervous system has a long history (Citovich,
    1930; Zakusov, 1936), the development of more sophisticated recording
    devices has led to an expansion of their use. As the technology to use
    these techniques develops, it becomes progressively more affordable,
    and should therefore become even more popular in the near future.

        In this section, the neurophysiological evaluation of
    neurotoxicity will be discussed in terms of assessment of the
    peripheral nervous system, the autonomic nervous system, and the
    central nervous system. With careful reading of the different
    sections, it should become evident that some methods described in one
    place may be useful under other circumstances. In the final section, a
    few issues pertaining to the interpretation of neurophysiological data
    will be addressed.

    4.2  Methods for Evaluation of the Peripheral Nervous System

    4.2.1  Conduction velocity

        Many substances are known to produce alterations in the peripheral
    nervous system (Spencer & Schaumburg, 1980), and evaluation of the
    functional integrity of peripheral nerves is the subject of the
    clinical science known as electrodiagnosis (Johnson, B.L., 1980).
    Conduction velocity, the speed at which action potentials are
    conducted along axons and nerves, is the most widely used measure of
    peripheral nerve function. Conduction velocity is usually measured in

    such a way that the activity of the fastest conducting axons is
    assessed. Changes in conduction velocity that occur following exposure
    to toxic substances producing axonopathy are reliable, but usually not
    large, often ranging from 10 to 30% of control values (Gilliatt,
    1973). On the other hand, demyelination produces large decrements
    (50%) in conduction velocity (McDonald, 1963).

        Principles involved in performing conduction velocity studies have
    been presented by Daube (1980), and techniques using rodents have been
    described, in detail, by others; some of these toxicological studies
    have been reviewed by Johnson, B.L. (1980) and Fox et al. (1982).

        Many other techniques besides conduction velocity have been
    applied in the assessment of peripheral nerve function. They include
    assessment of the refractory period (Hopf & Eysholdt, 1978; Lowitzsch
    et al., 1981), assessment of the extent to which axons and nerves can
    follow trains of stimuli occurring at high rates (Lehmann & Tachmann,
    1974), accommodation indices (Quevedo et al., 1980), and the use of
    collision techniques for selectively blocking activity of some nerve
    axons to study others (Kimura, 1976). Some of these techniques will
    certainly provide even greater sensitivity than the simple velocity
    measurements in common use.

    4.2.2  Peripheral nerve terminal function

        Methods for evaluating function in peripheral sensory receptors
    are particularly valuable in neurotoxicology. Toxic agents that
    exhibit a preference for the distal ends of long peripheral nerves, a
    dying-back neuropathy, or distal axonopathy might be expected to alter
    or impair the sensory function of these receptors (Fox et al., 1982).
    Indeed, in the case of proprioceptors, such as muscle spindles,
    neurotoxic insult can contribute to ataxia, areflexia, and
    incoordination (Lowndes et al., 1978a,b). Detailed discussion of the
    methods can be found in Fox et al. (1982). Although measuring
    peripheral nerve terminal function is more difficult than measuring
    peripheral nerve conduction velocity, such measurements are important
    since alteration of function in the terminal portions of axons
    frequently precedes alterations in conduction velocity. For example,
    function of muscle spindles (Lowndes et al, 1978a,b), motor nerve
    terminals (Lowndes & Baker 1976), and primary afferent terminals
    (Goldstein et al., 1981) has been reported to be compromised by toxic
    agents, long before any alterations are detectable in conduction
    parameters. Another advantage of these techniques is that, not only
    can the presence of neurotoxicity be detected, but the site(s) of
    neurotoxic action can also be investigated.

    4.2.3  Electromyography (EMG)

        The recording of biopotentials from muscle (electromyography) has
    been extensively used in human clinical studies in the diagnosis of
    certain diseases of the muscle (Johnson, E.W., 1980). EMG is an
    objective and sensitive method for the detection of changes in
    neuromuscular function. Altered neuromuscular function using EMG was
    detected in organophosphorus insecticide workers who did not exhibit
    other detectable signs and symptoms of poisoning including depressed
    blood cholinesterase activity (Drenth et al., 1972; Roberts, 1977).

        EMG methods have been little used for the study of neurotoxic
    substances in experimental animals. According to Johnson, B.L. (1980),
    there are probably two reasons for this. First, few toxicologists are
    trained in EMG procedures. Second, one important component of an EMG
    examination involves the evaluation of the voluntary graded
    contraction of the muscles. This is sometimes difficult to control in
    experimental animals. Nevertheless, methods using experimental animals
    are available (Johnson, B.L., 1980), and have been used successfully
    to study the neurotoxic effects of methyl n-butyl ketone (Mendell et
    al., 1974) and manganese (Ulrich et al., 1979).

        Electromyography is used to study direct toxic effects on muscles.
    Evoked muscle responses to nerve stimulation are invaluable in
    examining the neuromuscular junction, which can be affected by various
    neurotoxic agents (e.g., botulinum and tetanus toxins and
    organophosphate insecticides).

    4.2.4  Spinal reflex excitability

        Segmental spinal monosynaptic and polysynaptic reflexes are
    relatively simple functions of the central nervous system that can be
    easily evaluated by quantitative techniques (Mikiskova & Mikiska,
    1968; Fox et al., 1982). Many of the methods used in animals are
    direct laboratory counterparts of some of the clinically used
    neurological tests in human beings. There are two basic approaches.
    One does not requires any invasive procedures and thus is most akin to
    the tests used in human beings. The other involves electrophysio-
    logical techniques for examining the effects of neurotoxic agents on
    mono- and polysynaptic reflexes.

        In the non-invasive techniques, the functional state of the reflex
    arch is inferred either from the latency and size of the reflex
    response evoked by stimuli of a predetermined intensity (Zakusov,
    1953) or from the stimulus intensity (threshold) just sufficient to
    elicit a detectable response (Mikiskova & Mikiska, 1968). The
    threshold approach has been used by Mikiskova & Mikiska (1966, 1968)
    to study a variety of volatile solvents. In their view, the procedure

    is useful both as a screening test and in estimating the relative
    toxicities of substances. It is critical that a stimulator with a
    constant current output is used so that the stimulating current does
    not depend on the resistance of the skin and electrodes.

        The time required for a stimulus to a peripheral nerve to reach
    the spinal cord and return to the site of stimulation directly
    (F response), particularly after crossing a single synapse
    (H response), can indicate the excitability of the motoneuron pool.
    Fox et al. (1982) present electrophysiological techniques for
    examining the effects of neurotoxic agents on mono- and polysynaptic
    reflexes. This approach can provide better clues than the non-invasive
    approach concerning possible site(s) of action for the neurotoxic
    agent, but is considerably more time consuming. Moreover, the manner
    in which it is generally carried out (decerebrate animals) precludes
    repeated testing on the same animal. Thus, for most types of
    investigations (screening, determining threshold concentrations for
    effect) the non-invasive approach is preferable.

    4.3  Methods for Evaluation of the Autonomic Nervous System

        Compared to other approaches, relatively little effort has been
    expended in assessing the impact of suspected toxic agents on the
    activity of the autonomic nervous system. Most neurophysiological
    methods are designed to measure relatively rapid events, and therefore
    they are not particularly well suited to the evaluation of the
    autonomic nervous system. However, a few exceptions are noteworthy.

    4.3.1  Electrocardiography (EKG)

        Electrocardiography supplements other neurophysiological methods
    by providing data on the central and peripheral nervous control of
    autonomic functions. However, the interpretation of EKG changes is
    complex and must take into account direct effects on the myocardium
    (Mikiskova & Mikiska, 1968).

    4.3.2  Blood pressure

        Concomitant recording of EKG and blood pressure using
    environmental and pharmacological challenges is an approach worthy of
    consideration for evaluating animals exposed to suspected toxic
    agents. Altered circulatory responses to these types of stimuli may
    yield important information concerning the functional status of the
    autonomic nervous system. Indeed, exaggerated responses to vasopressor
    and cardiac acceleratory stimuli would suggest that such exposure
    might be accompanied by a higher risk of cardiovascular disease.

    4.4  Methods for Evaluation of the Central Nervous System

    4.4.1  Spontaneous activity - electroencephalography (EEG)

        EEG analysis was one of the first forms of electro-diagnosis of
    nervous system dysfunction. Since its discovery by Berger (1929),
    there have been progressively more sophisticated attempts at analysis,
    and progressively greater promises for its value. At present, the EEG
    is used widely in clinical settings, and consequently a variety of
    sources are available describing the technical details of recording,
    analysis, and interpretation (Basar, 1980; Niedermeyer & Lopes da
    Silva, 1982). In clinical neurology, the EEG has been used for the
    diagnosis and description of epilepsy, localization of tumours,
    description of sleep stage, as well as many other neurological
    disorders (Niedermeyer & Lopes da Silva, 1982). It has been used less
    often for the detection of subtle toxic agent-induced dysfunction,
    though it is an integral part of Soviet and Eastern European
    neurotoxicological studies (Horvath & Frantik, 1973, 1976).

        The normal EEG, whether recorded from the scalp or with indwelling
    electrodes in specific brain regions, has an amplitude of up to about
    100 uv. The useful frequency spectrum of the EEG is below 50 Hz,
    though higher frequencies are encountered in certain brain regions
    (Johnson, B.L., 1980). It is common to analyse the scalp-recorded EEG
    according to the amount of electrical activity contained within
    specific frequency bands, specifically: delta (2-4 Hz), theta
    (4-8 Hz), alpha (8-13 Hz), 1 (13-20), and 2 (20-30 Hz) (Lindsley
    & Wicke, 1974). A variety of electronic frequency analysers, as well
    as computer procedures, are available for use in analysing the power
    spectrum of the EEG throughout the entire frequency range or within
    selected frequency bands (Lindsley & Wicke, 1974).

        When electrodes are implanted in specific brain areas, the
    resultant EEG can be used to assess the effects of toxicants on these
    different brain areas or structures. Also, specific brain regions
    (e.g., the hippocampus) have particular patterns of after-discharge
    following chemical or electrical stimulation, which can be
    quantitatively examined and used as a tool in neurotoxicology (Dyer et
    al., 1979).

        It should be pointed out, however, that disassociation between the
    EEG pattern and behaviour can occur. For example, high voltage-slow
    activity (generally associated with sleep) has been seen following
    administration of atropine in animals that seemed to be excited
    (Bradley, 1958). Also, low voltage-fast activity (generally associated
    with arousal or with REM sleep) has been reported following
    physostigmine in animals that seemed to be asleep (Bradley, 1958).
    Thus, caution must be used in the interpretation of EEG changes alone.

        Extensive discussions and references, concerning the use of the
    EEG in neurotoxicology, are presented in Horvath & Michalova (1956),
    Mikiskova & Mikiska (1968), Johnson, B.L. (1980), and Fox et al.
    (1982). The full potential of this technique in neurotoxicology has
    not yet been fully exploited. However, it is clear that modern
    electronic and computer facilities are required to maximize its use.
    Moreover, changes in the pattern of the EEG elicited by stimuli
    producing arousal (light, sound, electrical stimulation) (Desi & Sos,
    1962; 1963) or produced by sleep (Fodor, et al., 1973) will most
    probably enhance its usefulness in neurotoxicology (Zilov et al.,
    1983). It should be noted that changes in the EEG have been reported
    after treatment with organophosphate compounds before depression of
    acetylcholinesterase activity was noted in the blood or brain tissue
    (Desi et al., 1974). In a more recent study, Desi (1983) compared
    changes in EEG with a behavioural test (maze), cholinesterase enzyme
    activity, and several general toxicological tests following exposure
    to 13 different pesticides. He concluded that the EEG was the most
    sensitive test for the detection of early and mild changes caused by
    these pesticides.

    4.4.2  Sensory systems

        In studies of sensory systems, the response of the nervous system
    to a well defined, yet physiological, input stimulus can be evaluated.
    Precise specification and control of the input stimulus reduces
    variability in the measured end-points, and increases the clarity with
    which toxicant-induced alterations are detected. Certain neurotoxic
    agents seem to have a particular affinity for sensory systems (e.g.,
    methanol). Furthermore, there are few cases when toxicant-induced
    damage to a non-sensory system is not parallelled by damage in sensory
    systems.

        The overall functional integrity of a sensory system is most
    directly assessed using evoked potential techniques. These techniques
    require the application of a discrete sensory stimulus, and averaging
    of the electrical activity from an appropriate neural pathway or brain
    area during a brief (e.g., 0.3 second) post stimulus epoch over
    repeated (e.g., 100) trials. Such averaging reveals a characteristic
    waveform that is specific to a particular modality, stimulus,
    electrode location, or species. Alterations in the latency from the
    stimulus to the specific peaks in the waveform, or in the amplitude of
    specific peaks in the waveform, are diagnostic of dysfunction.

        Sensory-evoked potential techniques are widely used in
    neurological clinics (Beck et al., 1975), and are becoming
    increasingly used in neurotoxicity evaluations. Since they may be
    readily recorded from unanaesthetized unrestrained animals, sensory
    evoked potentials may be obtained during, and correlated with,
    behavioural studies. They may detect, with various levels of
    sensitivity, alterations resulting from exposure to such chemicals as

    organometals (Dyer et al., 1978) and pesticides (Boyes et al., 1985). 
    More detailed description of the rationale, methods, interpretation,
    strengths and limitations of sensory evoked potential techniques can
    be found in Dyer (1985a,b).

    4.4.3  General excitability

        A common readily observed consequence of exposure to high
    concentrations of neurotoxic agents is an alteration in behavioural
    arousal. While some compounds produce sleep and coma, others produce
    seizures (Holmstedt, 1959; Lipp, 1968; Joy et al., 1980; Joy, 1982).
    These alterations may be presumed to reflect disordered excitability
    of the brain, and have led to the assumption that quantitative
    measures of excitability would be useful for the detection of
    dysfunction in this dimension. Four main approaches have been taken in
    the assessment of general excitability; (a) convulsive phenomena;
    (b) stimulation of motor cortex; (c) recovery functions; and (d) EEG
    recordings. EEG recordings have already been discussed in section
    4.4.1.

    4.4.3.1  Convulsive phenomena

        While there are many experimental models of epilepsy (Purpura et
    al., 1972), not all of these are suitable for the detection and
    characterization of neurotoxicity. In neurotoxicology, seizure
    susceptibility has been most often assessed using either electrical
    stimulation or systemically administered drugs to produce seizures.

        Pharmacological stimulation, using agents such as pentylenete-
    trazol (Metrazol), is simple, quick, and cheap. While interpretation
    may be complicated if the toxic agent under study alters the
    metabolism of the convulsant agent, practice has failed to reveal many
    such instances. Depending on the convulsant selected, the integrity of
    selected neurochemical systems may be assessed. For example,
    picrotoxin is presumed to act by blocking activity in GABA-ergic
    systems.

        Electrically induced seizures have been widely used in
    neurotoxicology. Horvath & Frantik (1979) and Frantik & Benes (1984)
    evaluated the effects and relative potentials of a wide variety of
    organic solvents on the extensor phase of electrically induced
    seizures in rats. They compared the severity of these seizures with
    other behavioural effects obtained in both laboratory animals and man,
    thus providing a comparative measure of the relative toxicity of these
    chemicals. The seizures were suppressed by relatively low air
    concentrations of most of the solvents tested and this occurred in a
    concentration-dependent manner.

        Electrically-induced after discharges offer the opportunity to
    study seizure activity in specific brain regions with implanted
    electrodes in the absence of behavioural correlates. These tests have
    been used particularly in areas of the limbic system known to be
    sensitive to neurotoxic agents and to have low seizure thresholds
    (Dyer et al., 1979).

        Repeated elicitation of after-discharges from certain brain
    regions leads to the progressive recruitment of behavioural correlates
    by a process known as kindling. The development of kindling-induced
    seizures, while more time consuming than other methods, may provide
    more detailed information regarding the nature of alterations produced
    by the toxic agent. A discussion of kindling as a model for the study
    of neurotoxic agents can be found in Joy (1985).

    4.4.3.2  Stimulation of the cerebral motor cortex

        The motor area of the cerebral cortex is preferred because
    stimulation results in constant, clearly defined motor responses. The
    procedure is discussed in detail by Benesova et al. (1956), Mikiska
    (1960), and Mikiskova & Mikiska (1968). Basically, "excitability" is
    determined by measuring the current required to evoke minimal movement
    of the contralateral fore limb. Studies can be conducted acutely or
    with the long-term implantation in animals of electrodes (generally
    fine screws) touching the motor cortex. Depressant agents uniformly
    raise the threshold to electrical stimulation (by as much as several
    fold) whereas stimulant type agents lower it (but to a lesser extent).
    Effects of drugs and toxic agents in this procedure are cited by
    Mikiska (1960) and Mikiskova & Mikiska (1968). It should be noted that
    direct electrical stimulation of central nervous tissues has far more
    complex consequences than stimulation of peripheral nerve. The
    stimulus excites thousands of neurons mutually connected by excitatory
    and inhibitory synapses, which often form reverberating circuits
    (Mikiskova & Mikiska, 1968). Thus, interpretation concerning the
    possible mechanism of effect is difficult.

    4.4.3.3  Recovery functions

        Repeated stimulations may alter the threshold for evoking a
    response. Neuronal recovery processes are among several neurophysio-
    logical challenges reviewed for their relevance for the detection and
    characterization of neurotoxicity by Dyer & Boyes (1983). While these
    methods have only recently been used in neurotoxicology, there are a
    number of instances in which they detect neurotoxic effects earlier
    than other methods (Dyer & Boyes, 1983).

    4.4.4  Cognitive function

        Not all evoked potentials are directly related to eliciting
    sensory stimuli. Parts of some waveforms (most often the "late" or
    "slow" components) are presumed to be elicited by events that are
    internal, and may reflect cognitive activity or initiation of motor
    activity (Otto, 1978). In human beings, changes in one such potential
    have been associated with extremely low body burdens of lead (Otto et
    al., 1981). Unambiguous interpretation of such findings is not yet
    possible, and the methods have not been applied extensively to animal
    studies. However, increased research activity in this area should make
    assessing the value of these "slow potentials" more straightforward.

    4.4.5  Synaptic and membrane activity

        Using appropriate microelectrode neurophysiological techniques, it
    is possible to determine in a direct way whether exposure to a toxic
    agent impairs the response properties, synaptic function, or membrane
    properties of neurons (Fadeev & Andrianov, 1971; Andrianov et al.,
    1977; Fadeev, 1980; Barker & McKelvy, 1983; Dingledine, 1984). While
    exquisite in the precision of the information they provide, these
    techniques are technically difficult and expensive to use, and are
    most profitably used to assess mechanism of toxicity. Data and
    discussion of the value of these techniques can be found in Joy
    (1982), Narahashi (1982), and Narahashi & Haas (1967, 1968); all of
    these reports consider the effects of organochlorine insecticides and
    pyrethroid insecticides on properties of neuronal membranes.

    4.5  Interpretation Issues

        Most neurophysiological methods provide information at several
    different levels. For purposes of organization, it is convenient to
    divide the brain into discrete functional systems such as the
    somatosensory system, the visual system, the extrapyramidal system,
    etc. While these systems are clearly interrelated, they are almost
    always treated separately. Most such systems are constructed of
    neurons with similar membrane and axonal properties, but they differ
    with respect to their connections, neurotransmitters, and metabolic
    activity. Thus, evaluation of any one system provides a combination of
    information, some unique to the system, and some common to most or all
    neural systems.

        The choice of the appropriate neurophysiological methods for a
    given study depends on the questions posed. If emphasis is on
    identification of systems involved in the toxicity produced by a
    specific compound, then gross measures of the functional activity of
    whole systems are desirable (e.g., evoked potentials in the visual
    system). On the other hand, if a compound is known to be toxic for a
    particular system, for example the hippocampus, and the emphasis is on
    the mechanism of toxicity, then microelectrode techniques may be

    appropriate. Investigators about to perform neurophysiological studies
    of neurotoxicity should become familiar with the strengths and
    limitations of the methods mentioned in this review before selecting
    one. Furthermore, attention should be paid to the possibility that the
    toxicant-induced alterations observed reflect secondary dysfunction,
    which in turn is produced by primary dysfunction in another system.
    For example, a compound that produces hypothalamic dysfunction and
    compromises the body's ability to thermo-regulate will appear to
    produce changes in other systems that are really secondary to
    hypothermia.

        A major advantage of many of the neurophysiological techniques
    mentioned above (EEG and evoked potential) is that they are readily
    recorded in human beings as well as laboratory animals. This feature
    provides a ready framework for cross species extrapolation of data,
    though considerable work must still be performed before the health
    implications of alterations in some of these end-points is fully
    understood. A second important feature is that the relationships
    between behavioural and electrophysiological alterations produced by
    toxic agents can be observed in awake subjects, as can any
    dissociation between these two types of end-points.

        Finally, it cannot be emphasized too strongly that the CNS is
    constantly receiving afferent input and sending efferent output to all
    the organs of the body. This principle of self-regulation of
    physiological functions, incorporating the cybernetic principle of
    feedback gave rise to the functional system theory of Anokhin (1935).
    This theory has been described in detail by Anokhin (1968, 1974) and
    Sudakov (1982). According to this theory, functional systems are
    dynamic self-regulated organizations, the components of which
    contribute to the attainment of a useful adaptive result for an
    organism. The components of any given functional system are united or
    integrated by the afferent and efferent CNS influences impinging on
    it. The importance of this theory to toxicologists lies in its focus
    on: (a) adaptive mechanisms; and (b) the inseparable nature between
    the CNS and the other organs via the afferent input and efferent
    outflow. First, it is the interference with adaptive mechanisms that
    gives rise to the signs and symptoms produced by a toxic agent.
    Second, the inseparable nature of the CNS and other organs complicates
    localization of the primary target of a toxic agent. Thus, the
    electrophysiological changes seen in the CNS or the alteration in
    behaviour may not be due to a primary effect on the CNS, but may
    rather be the result of alterations in afferent input due to
    disturbances in peripheral systems such as the gastrointestinal tract,
    liver, or kidney. Or, the chemical could interfere with a metabolic
    process common to a number of "functional systems" subserving
    different biological needs. The lesson, as stated previously in both
    sections 2 and 3, is that neurotoxicologists must not be too hasty in
    concluding that any effect observed is due to a direct effect on the
    CNS. They should, in fact, demonstrate that it is not due to an effect
    on another organ.

    4.6  Summary and Conclusions

        Neurophysiological methods are important in animal neuro-
    toxicology. They provide direct laboratory counterparts for many of
    the tests used in human beings. They can give insight into possible
    site(s) and mechanism(s) of actions. They can be particularly useful
    when used concomitantly with behavioural methods. The use of evocative
    techniques (arousal, stimuli, work loads, interactions with
    psychopharmacological agents) can increase the sensitivity of the
    tests for detecting toxicant effects. As with behavioural methods, it
    is incumbent on the investigator to determine whether an effect on the
    CNS is due to a primary action of the toxic agent or a secondary one
    as a result of damage to some other organ, thus altering the afferent
    input into the CNS or its processing within the CNS.

        The most useful techniques for screening are those that are
    minimally invasive, relatively cheap and rapid to perform, and test
    neural function in a broad sense. Depending on the particular
    application, these may include evoked potential, EEG, EMG,
    excitability, or conduction velocity studies. More restricted use of
    these techniques, or use of microelectrode techniques, may be useful
    for further characterizing the systems affected and the mechanisms of
    action of neurotoxic agents. In some instances, neurophysiological
    techniques have been reported to be more sensitive than behavioural,
    biochemical, or neuropathological measurements. Further research using
    direct comparison with other methods is needed to determine the
    conditions under which this is so and the significance of this in
    human risk assessment.

    5.  MORPHOLOGICAL METHODS

    5.1  Introduction

    5.1.1  Role of morphology

        Just as neurobehavioural methods find their special use in the
    description and analysis of neurotoxic diseases for which there are no
    pathological correlates, the morphologist's special contribution is to
    describe neurotoxic conditions associated with structural alterations
    of nervous tissue (Table 2). Commonly, as in most chronic neurotoxic
    diseases, pathological alterations in the nervous system lead to
    changes in behaviour and function. These changes may provide important
    clues as to the site and even the nature of the underlying structural
    damage. Conversely, analysis of the location and type of pathological
    change often allows the morphologist to predict the type of
    accompanying dysfunction, the likely duration of the abnormality, and
    the degree of reversibility. This type of structural-functional
    correlation, so widely used in the evaluation of human neurological
    disorders, is especially helpful for assessing the severity of
    neurotoxic disease in experimental animals and, thereby, the
    implications for human exposure.

        A properly conducted morphological examination of experimental
    animals establishes or rules out the existence of structural damage,
    identifies the most vulnerable sites within the nervous system and
    traces the temporal evolution of pathological changes. Armed with this
    information, the morphologist often is able to advise the electro-
    physiologist and biochemist where to focus their attention for the
    early detection and detailed analysis of the underlying neurocellular
    dysfunction.

    5.1.2  Basis for morphological assessment

        An understanding of the organization, structure, and function of
    the normal nervous system is the point of departure for any assessment
    of pathological changes induced by exposure to chemical substances.
    This requires not only a working knowledge of neuroanatomy (Williams &
    Warwick, 1975; Pansky & Allen, 1980), but also an understanding of
    structural variations that may occur in relation to factors such as
    the species under study and the age of the animal. In addition, since
    the morphologist usually studies dead tissue, an understanding of
    possible post-mortem changes is required.

    5.2  The Nervous System and Toxic Injuries

    5.2.1  The nervous system

        The nervous system may be separated anatomically into central and
    peripheral divisions. The peripheral nervous system (PNS) is composed
    of nerve cells (neurons) and their processes (axons) which conduct

        Table 2.  Morphological assessment in neurotoxic injuries
                                                                                      

    Type of neurotoxic            Chemical              Pathological change
         injury
                                                                                      

    Neurons

        excitable (neuronal)      pyrethroid            none expected
           membrane

        neurotransmitter          anticholinesterase    none or terminal and
           systems                                      muscle swelling

        anabolic disturbance      doxorubicin           chromatolysis, or somal
                                                        degeneration and neuronophagia,
                                                        Wallerian degeneration,
                                                        muscle atrophy (PNS),
                                                        or transynaptic
                                                        neuronal degeneration (CNS)

        catabolic disturbance     swainsonine           increase of axonal and/or
                                                        somal lysosomes

        axonal transport          acrylamide            accumulation of cytoskeletal
                                                        elements and/or organelles,
                                                        Wallerian degeneration

        dedrite                   lathyrus toxin        swelling, variable involvement
                                                        of soma

    Special sense organs

        retina                    methanol              oedema

        inner ear                 arsenical             degeneration of stria
                                                        vascularis

    Glial and myelinating cells

        astrocyte                 6-aminonlcotinamide   swelling and degeneration

        oligodendrocyte           isoniazid             degeneration and myelin
                                                        vacuolation

        central myelin            triethyltin           myelin vacuolation and loss
                                                                                      

    Table 2 (Cont'd)
                                                                                      

    Type of neurotoxic            Chemical              Pathological change
         injury
                                                                                      


        Schwann cell              diptheria toxin       degeneration and local
                                                        demyelination

        peripheral myelin         hexachlorophene       myelin vacuolation and loss

    Blood vessels
                                  cadmium               haemorrhage and associated
                                                        neurocellular degeneration
                                                                                      
    

    information between muscles, glands, sense organs, and the spinal cord
    or brain. The PNS includes afferent (sensory) and efferent (motor)
    fibres, and both types are represented in the somatic and visceral
    (autonomic) components of the nervous system. Somatic afferent fibres
    carry information from special organs and sensory receptors in skin
    and muscles, while visceral afferents convey impulses from the gut,
    glands, and various organs. On the motor side, somatic efferents
    innervate striated muscle, while visceral efferents supply smooth
    muscles of blood vessels, glands, and gut. In toxic states involving
    the PNS, degeneration of somatic sensory-motor fibres leads to
    peripheral neuropathies associated with sensory loss (e.g., decreased
    sensitivity to vibration, touch, and position sense) and motor
    weakness in distal extremities, while dysfunction or breakdown of
    autonomic fibres may lead to abnormal sweating, cardiovascular
    changes, or gastrointestinal, urinary-tract, genital, and other types
    of dysfunction (Schaumburg et al., 1983; Dyck et al., 1984).

        Manifestations of central nervous system (CNS) disorders depend
    largely on the site and nature of the induced functional or structural
    change (Collins, 1982). The CNS consists of the parts of the nervous
    system contained within the skull and vertebral column. The spinal
    cord receives information from PNS afferents supplying skin, muscles,
    and glands, transmits signals for motor function by way of efferent
    fibres, and communicates via specific pathways with coordination
    centres within the brain. The brain is immensely complex and
    responsible for initiating, receiving, and integrating signals needed
    to maintain internal homeostasis, cognition, awareness, memory,
    language, personality, sexual behaviour, sleep and wakefulness,
    locomotion, sensation, vision, audition, balance, and many other body
    functions. Most of the available information on the structural-

    functional correlates of brain come from the study of higher mammals,
    including man (Truex & Carpenter, 1983). The brainstem, consisting of
    the midbrain, ports, and medulla oblongata, receives and processes
    information from skin, muscles, and special sense organs (e.g., inner
    ear) and, in turn, controls these muscles, as well as certain
    autonomic functions. The cerebellum and basal ganglia are required for
    the modulation and coordination of muscle movement. The diencephalon,
    including the thalamus and hypothalamus, is a relay zone for
    transmitting information about sensation and movement, and also
    contains important control mechanisms to maintain the internal
    homeostasis of the body. The hypothalamus functions as the primary
    control centre for the visceral system and serves to integrate the
    activity of the endocrine and other systems. The cerebral hemispheres,
    capped by the cerebral cortex, are concerned with perceptual,
    cognitive, motor, sensory, visual, and other functions. The optic
    nerves and their radiations conduct visual information from the
    retina, through the thalamus to the occipital cortex.

    5.2.2  Cellular structure of the nervous system

        Neurons and glial cells are the fundamental cellular elements of
    the nervous system, and these are associated with blood vessels and
    other specialized epithelial and connective tissue cells (Williams &
    Warwick, 1975). Neurons are equipped with multiple short processes
    (dendrites) that receive information from other nerve cells, and a
    single long axon that conducts electrical signals to other neurons and
    muscles, and to or from the skin, muscles, and glands (Fig. 1). The
    axon terminates at a synapse where chemically-encoded information is
    conveyed to neurons or muscle.

        Glial cells in the CNS include astrocytes, oligodendrocytes, and
    microglia (Fig. 2). Astrocytes are divisible into protoplasmic and
    fibrous forms, are closely associated with neurons and blood vessels,
    and may play a nutritive role in maintaining neurons and other cells.
    Oligodendrocytes are responsible for elaborating short lengths of
    myelin around multiple axons, and microgila have a phagocytic
    function. In the PNS, Schwann cells envelop multiple small axons
    (unmyelinated fibres) or associate with and elaborate lengths of
    myelin (internodes) around single axons. Myelinated and unmyelinated
    fibres are separated from each other by endoneurial connective tissue
    composed of fibroblasts and collagen, and these elements are bound
    together in fasciles surrounded by a fibrocellular sleeve, the
    perineurium. Bundles of fascicles held together by epineurial
    connective tissue form a peripheral nerve.

    FIGURE 1

    FIGURE 2

    5.2.3  Neurocellular reaction to injury

    5.2.3.1  Biological principles

        Neurons are highly atypical cells because their cytoplasmic
    processes often occupy a much greater volume than their cell somata.
    This unusual cellular architecture provides an enormous surface area
    for chemical attack and places a great demand on the soma, since it
    alone has the metabolic machinery required to maintain the cellular
    processes. Specialized transport systems have evolved to convey
    information to the cell body, the axon, and its terminal regions
    (Ochs, 1982). The functional problem can be illustrated by considering
    a peripheral motor neuron, located in the lumbar spinal cord and
    innervating muscle in the foot, which must maintain the structure and
    function of an axon that is about a metre in length. Failure to
    maintain the entire length of this enormous column of cytoplasm, by
    interruption of axonal transport or by some other mechanism, may
    account for the vulnerability of distal axons in many types of
    neurotoxic disease. Other types of neurotoxins (e.g., doxorubicin) may
    interfere with the metabolic machinery of the soma, thereby resulting
    in degeneration of the entire neuron. Comparable explanations can be
    offered to explain the vulnerability to chemical attack of myelinating
    cells, their cytoplasmic processes, and myelin sheath (Spencer &
    Schaumburg, 1980).

        There are two considerations when contemplating toxic damage to an
    organ or tissue. First, there are the changes that can be related to
    the action of the chemical agent; second, there is the reaction of the
    tissue to these changes. When toxic damage to neurons and myelinating
    cells is considered, it is often difficult to separate the two phases
    of the pathological process being studied. The situation is further
    complicated as the type of change observed is often dependent on dose,
    and may also vary with other factors, such as the species. In general,
    however, chemical attack leads to two types of primary change in
    neural cells:

        (a) the accumulation, proliferation, or rearrangement of
            structural elements (e.g., intermediate filaments,
            microtubules) or organelles (e.g., mitochondria, lysosomes);
            and/or

        (b) the breakdown (degeneration) of cells, in whole or in part.

    The latter is usually followed by regenerative processes, and these
    may occur during the period of intoxication. Changes in the cellular
    elements of intraneural blood vessels may occur, and secondary changes
    may develop in other organ systems, notably voluntary muscle (Walton,
    1974).

        Neural cells appear to have a limited repertoire of pathological
    responses, and the consequences of many types of toxic damage can be
    predicted from an understanding of the biology and function of the
    cells that are involved. As in any organ system, changes in one cell
    type -- the primary target -- usually lead to secondary and tertiary
    responses in related cells, so that the net effect is a predictable
    cascade of pathological responses. For the nervous system, however,
    the neuropathological changes often have to be considered with regard
    to the fact that, while cellular responses to toxic injury may occur
    locally, they may also occur at distant sites. Changes may also
    develop in different locations as a function of time, dose, and/or
    duration of intoxication.

    5.2.3.2  Neurons

        Loss of the cell body of neurons (neuronopathy) is an irreversible
    event seen in many types of intoxication. Similar tissue reactions
    occur in the CNS and PNS. Fig. 3 illustrates the changes occurring
    with loss of primary sensory neurons in dorsal root ganglia. Glial
    cells and macrophages proliferate around the dying cell (neurono-
    phagia) and axon processes under Wallerian degeneration. This involves
    breakdown of the entire length of the axon, dissolution and removal of
    the myelin sheath, and the proliferation of phagocytic and glial
    cells. Degenerated myelin is removed more slowly in the CNS than in
    the PNS and astrocytes increase the number and size of their fiberous
    processes to occupy the space left by the degenerating nerve fibre.
    Axon degeneration also produces significant secondary changes in the
    denervated cells, such as muscle (neurogenic atrophy) or other neurons
    (transynaptic and retrograde transynaptic degeneration) (Blackwood et
    al., 1971).

        Neurons survive and recover from certain types of toxic assault,
    notably those that cause structural damage to their processes.
    Abnormalities of axons may be expressed in the form of generalized
    atrophy or localized swellings containing excessive numbers of
    structural elements or organelles, with associated secondary changes
    in the myelin sheath in the form of corrugation and displacement
    (secondary demyelination), respectively. Disruption of axonal
    integrity leads to axon atrophy and/or degeneration (axonopathy) below
    the site of injury (Fig. 4), with secondary changes in target cells,
    such as muscle (neurogenic atrophy) or neuron (transynaptic
    degeneration). Oligodendrocytes and Schwann cells lose their ability
    to maintain myelin and then undergo cell division, while phagocytic
    cells remove the myelin debris. The cell bodies of affected neurons
    may also undergo responses secondary to axonal lesions, including cell
    necrosis or, more commonly, rearrangement of cellular components
    (chromatolysis) as a prelude to axon regeneration. Regrowth of the
    axon commences promptly at the position of axon interruption and, in
    the PNS, the elongating neuronal process reassociates with Schwann
    cells, which elaborate a new myelin sheath and conduct the

    FIGURE 3

    FIGURE 4

    regenerating axon to its target organ (muscle or sense organ)
    (Schaumburg et al., 1983). In the CNS, astrocytes respond to injury by
    increasing the number and size of their fibrous processes, and these
    appear to inhibit regeneration of all but the smallest (e.g.,
    monoaminergic) axons. Prominent axon swellings develop at the viable
    end of larger fibres, a process known as axonal dystrophy.

    5.2.3.3  Myelinating cells

        Some chemical compounds induce primary morphological changes in
    myelinating cells or their myelin sheaths. It is usual to separate
    agents that induce changes in the cell body of Schwann cells or
    oligodendrocytes from those that cause abnormalities selectively in
    the myelin sheath, though this distinction may be misleading as many
    compounds can produce both types of change, depending on dose and
    length of exposure. The net result of either type of CNS or PNS insult
    is primary demyelination (myelinopathy) (Fig. 5). Loss of
    oligodendrocyte somata is associated with the arrival of phagocytic
    cells which remove myelin and become filled with droplets of neutral
    fat. Astrocytes are commonly affected by chemical agents that perturb
    oligodendrocytes but, if they survive, astrocytes may also take up
    small amounts of myelin debris, divide, hypertrophy and, if
    remyelination does not occur, surround the demyelinated axons. More
    usually, however, new oligodendrocytes are recruited, their processes
    contact the demyelinated axon and these elaborate new but
    foreshortened internodal lengths of myelin (remyelination). Loss of
    Schwann cells also results in primary demyelination, cell
    multiplication, and remyelination. Another common type of myelinopathy
    involves the processes of CNS and PNS myelinating cells: these
    disorders are usually visualized by the accumulation of oedema fluid
    within the myelin sheath or its associated cellular processes (Spencer
    & Schaumburg, 1980). Vacuolation of the myelin sheath is usually a
    reversible process, though severe intramyelinic swelling may constrict
    the axon and induce Wallerian degeneration.

    5.3  Experimental Design and Execution

    5.3.1  General principles and procedure

        Morphological examination is primarily concerned with studying
    structural changes in the nervous system that might explain
    behavioural disturbances. These changes are most commonly found when
    there is repeated dosing, but permanent neurological damage may also
    occur after single doses of some compounds. On occasion, such changes
    appear to be minor but, if restricted to a few neurons of a nucleus of
    considerable importance, may result in profound functional effects.
    This is particularly true of the developing nervous system where,
    because of stepwise, interdependent development, damage induced by a
    chemical agent at one stage may have a "domino effect" on later stages
    of development. Damage to CNS structures, especially when loss of

    FIGURE 5

    nerve cells is involved, is more likely to be associated with
    permanent behavioural changes than disorders predominantly affecting
    the peripheral nervous system, where satisfactory regeneration is more
    often the rule. Rapidly reversible effects on behaviour are
    occasionally, but not often, associated with detectable structural
    damage to the nervous system.

    5.3.2  Gross morphology

        While most neurotoxic damage is most evident at the microscopic
    level, it is important to pay attention to the weight of the brain as
    well as to macroscopic manifestations such as discolouration (for
    example, due to bilirubin), discrete or massive haemorrhage, or other
    localized lesions (for example, transverse myelitis).

        The new methods of nuclear magnetic resonance (NMR) can supply
    data concerning structural changes, for example, demyelination, both
    in vivo and post-mortem.

    5.3.3  The role of histology

        There is no substitute for a thorough light-microscope examination
    of the nervous system and other organs for the initial assessment of
    animals suspected of having neurotoxic injury. The same careful
    attention to anatomical detail should be given in the examination of
    experimental animal tissues as would be given by a pathologist
    studying autopsy tissue from a patient who has suffered from a new
    type of neurological disease. The results of this type of analysis
    will provide direction for additional studies that focus on more
    specific questions such as the type and extent of structural damage,
    the location where changes first occurred, the time-to-onset of
    structural damage, the reversibility of these changes, the lowest dose
    that caused pathological changes, and the highest no-observed-adverse-
    effect dose.

    5.3.3.1  Biological principles dictating tissue response

        The study design must take into account several important
    principles of underlying cellular responses to injury. These
    principles will influence when and where tissue should be sampled for
    morphological study.

        Often, there is a delay of days or weeks between dosing and the
    first appearance of structural changes in animals treated with
    neurotoxic agents. The extent of the delay depends largely on the
    chemical used, the dose, and the species studied. Once the
    pathological process has been initiated, a chain of degenerative and
    regenerative events may be set in motion, and these may or may not
    cease to evolve after dosing has stopped. Most neurotoxic substances
    that damage neurocellular elements produce a largely symmetrical

    pattern of structural change, whether in the CNS or PNS. Some neurons
    or nerve fibres may be more vulnerable than others to chemical attack,
    so that at any point in the pathological process, the observer may be
    presented with many or all of the stages in the reaction of individual
    cellular elements. The active pathological process may also move in
    space with time, as in dying-back axonpathies in which degeneration
    proceeds retrogradely along affected nerve and fibre tracts (Cavanagh,
    1964). Careless sampling of animals with these or other types of
    neurotoxic diseases may reveal little or no pathological changes. By
    contrast, the careful and thorough investigator will be able to
    distinguish between early and later changes, and thereby develop a
    hypothesis of the spatial-temporal sequence of the pathological
    process. This can be tested in subsequent studies in which animals are
    examined at different stages in the development of the disease.
    Investigations of this type will also provide information on the
    relationship between the dose and duration of treatment required to
    induce neurotoxic injury. No-observed-adverse-effect levels can be
    established by determining doses that do not produce any
    characteristic changes after a specific period of chemical treatment.
    Quantitative estimation of structural abnormalities in treated versus
    control animals becomes increasingly important as the no-observed-
    adverse-effect dose is approached.

    5.3.4  Use of controls

        Parallel study of age-matched control animals is mandatory,
    whenever treated animals are subjected to neuropathological
    examination. Control groups should include untreated animals, vehicle-
    treated animals, and positive controls. The may receive a neurotoxic
    dose of the agent under study or of another compound with well-
    characterized neurotoxic properties that elicits a comparable type of
    neurotoxic damage. Demonstrating the development of an appropriate
    neurotoxic response in the positive control strengthens the validity
    of data obtained from animals treated with the agent under study,
    especially if these prove to lack pathological changes. The use of
    positive controls is imperative for the inexperienced investigator who
    needs to gain confidence from the self-demonstration that a known
    neurotoxic agent produces the expected changes. Inclusion of these
    results in a published report will also reassure the audience that the
    investigator has reproduced the expected pattern of structural damage
    associated with a known neurotoxic chemical.

        There are important advantages in studying the structure of
    nervous tissue without knowledge of the treatment group. The
    investigator is forced to be more critical in seeking and describing
    abnormal findings, and the final assessment is a more objective
    statement of the similarities and differences between treatment and
    control groups.

    5.3.5  Pattern of response

        It is not the purpose of this publication to provide a detailed
    analysis of the various problems of selective neuronal damage
    encountered under neurotoxic conditions. However, it is emphasised
    that occasionally the localization is very precise, limited to one or
    two nuclei in the brain stem, or to distal regions of long fibre
    tracts and peripheral nerves. Selective pathological changes are given
    in Table 3 for reference in interpreting new situations.

    5.3.6  Data acquired

        At the end of the survey, clear answers must be found to the
    following questions about the CNS:

    1.  Are there changes in nerve cells in the area of the cerebral
        cortex, basal ganglia, hypothalamus, limbic system, diencephalon,
        brain stem, cerebellum, or spinal cord?

    2.  Are there any changes in shape or number of astroglia and/or
        microglial cells?

    3.  Is there any loss or other change in oligodendroglia?

    4.  Are there any changes in the myelinated areas of the brain or
        spinal cord, such as myelin fragmentation or myelin vacuolation?
        Is this associated with axonal swellings and/or degeneration?

    5.  If 4 is positive, are there glial responses?

    6.  Are there focal, vascular, and necrotic changes? If so, where?

    7.  Are tumours present?

    8.  Are there any changes in special sense organs to account for
        functional changes?

        For the peripheral nerves, the following questions must be
    answered:

    1.  Are there any changes in peripheral nerves?

    2.  If so, is it axon degeneration or primary demyelination, or both?

    3.  What is the distribution of degeneration? Is it distal, affecting
        only long and large-diameter fibres, or are shorter fibres such as
        cranial nerves also affected?

    4.  Are sensory, motor, or both types of fibres involved?

        Table 3.  Examples of toxic effects illustrating the specific
              patterns of damage that may be found
                                                                                

    Pattern                            Cause                 Reference
                                                                                

    Neuronal changes

    laminar cortical necrosis          anoxia                Briefly (1976)

    hippocampal pyramidal cell         trimethylin           Brown et al.
      damage                                                 (1979)

    selective granule cell loss        methylmercury         Hunter & Russell
      from the cerebellum                                    (1954)

    selective loss of inferior         3-acetylpyridine      Desclin (1974)
      olivary cells

    selective degeneration of          doxorubicin           Cho et al. (1980)
      perikarya of sensory
      ganglion cells

    selective degeneration of          methylmercury         Cavanagh & Chen
      axons of sensory ganglion                              (1971)
      cells

    retrograde degeneration of         organophosphorus,     Cavanagh (1964);
      long sensory and motor           acrylamide,           Spencer &
      axons in CNS and PNS             hexanedione           Schaumburg (1977)

    retrograde degenerations of        clioquinol            Krinke et al. (1979)
      axons of long spinal cord
      tracts

    Myelin changes

    CNS myelin vacuolation             triethyltin           Aleu et al. (1963)

    CNS and PNS myelin                 hexachlorphene        Towfighi et al.
      vacuolatin                                             (1975)

    focal degeneration of PNS          diphtheria            Cavanagh & Jacobs
      myelin without axon loss                               (1954)

    focal degeneration of PNS          lead                  Fullerton (1966)
      myelin with some axon loss
                                                                                

    Table 3 (Cont'd)
                                                                                

    Pattern                            Cause                 Reference
                                                                                

    Vascular and neurotic changes

    symmetrical lesions in             misonidazole          Griffin et al.
      brain-stem nuclei                                      (1980)

    non-symmetrical brain-stem         lead                  Wells et al. (1976)
      and cortical lesions
                                                                                
    

    5.  What is the state of nerve cell bodies (i.e., sensory ganglion
        cells, anterior horn cells)?

    6.  If the cells in 5 show changes, are those in the medulla and
        trigeminal ganglia (i.e., cranial nerves) similarly affected?

    7.  Are neurons normal, chromatolytic, or degenerated?

    8.  Are automatic ganglia or nerves similarly affected or not?

    9.  Are muscles affected by denervation and/or myopathic changes?

    10. At dose levels below those that produce clinical signs, is there
        histochemical and/or morphometric evidence of metabolic neuronal
        changes that is brought out by using these more sensitive
        techniques?

    5.4.  Principles, Limitations, and Pitfalls of the Morphological
          Approach

    5.4.1  Tissue state

        Although under special circumstances the morphologist is able to
    study the reactions of living tissue to chemical exposure, either
    in vivo or in tissue culture (Yonesawa et al., 1980), the vast
    majority of pathological studies involve the assessment of dead
    tissue. The structural changes that occur in the nervous system as a
    function of time post-mortem can obscure the recognition of toxic
    injury and may confuse the inexperienced neuropathologist. Chemical,
    or rarely, physical (freezing) fixation methods are used to prevent
    the development of post-mortem changes.

    5.4.2  Principles of fixation

        The use of a reliable fixation method is mandatory for the
    morphological examination of nervous tissue (Thompson, 1963; Pease,
    1964; Hayat, 1973; Gabe, 1976; Bancroft & Stevens, 1977; Glausert,
    1980). The purpose of fixation is to preserve cellular architecture as
    closely as possible to that in the living state, to inhibit the loss
    of chemical components required for the maintenance of morphology, and
    to prepare the tissue to accept stains that enhance tissue density for
    clear resolution of cytological detail. The ideal method of fixation
    is one that instantaneously terminates life processes in the tissue of
    interest without distorting cytological detail. Rapid freezing of
    living tissue approaches this ideal. Excision and freezing of tissue
    is routinely employed for the assessment of human biopsy tissue and is
    particularly suitable for light-microscope histochemistry and
    immunocytochemistry. A few specialized laboratories use rapid freezing
    for the examination of tissue, or metal replicas thereof, by
    transmission electron microscopy.

        Chemical methods are far more commonly used in neurotoxicology for
    the preparation of nervous tissue. Although the ideal chemical
    fixative should instantaneously kill living tissue without inducing
    structural changes, all known methods fall short of this ideal.
    Instead, the morphologist must strive to induce small, consistent and
    readily recognizable changes that will not compromise tissue
    examination and assessment. The most satisfactory and widely used
    chemical fixatives are dilute, aqueous solutions of aldehydes
    (formaldehyde and/or glutaraldehyde), which rapidly cross-link
    proteins and stabilize associated lipids, prevent post-mortem changes,
    and introduce rigidity into the tissue to facilitate handling. A
    secondary phase of fixation using a dilute aqueous solution of osmium
    tetroxide to fix lipids is routinely added after glutaraldehyde
    fixation for improved preservation of tissue. Occasionally, osmium
    tetroxide is used as the primary fixative. Structural artefacts
    induced by fixation are minimized if the fixative solutions are
    buffered to match tissue pH and osmolality. Temperature should also be
    controlled; low temperatures may affect the preservation of fine
    structural elements (e.g., microtubles).

        Chemical fixatives must be delivered to the site of interest as
    rapidly as possible if post-mortem artefacts are to be minimized. Most
    effective is systemic perfusion (Pease, 1964; Hayat, 1973; Spencer et
    al., 1980; Spencer & Bischoff, 1982), in which the fixative is
    introduced via the ascending aorta of the deeply anaesthetized animal,
    and distributed throughout the vascular network under controlled
    pressure. Spaces normally occupied by blood and tissue fluid are
    rapidly replaced by the chosen fixative solution, and the animal
    expires painlessly within seconds. Perfusion of individual organs is
    an alternative, but more difficult and less satisfactory technique. An

    alternative method, suitable only for readily accessible parts of the
    nervous system (e.g., peripheral nerves), is to bathe the tissue
    in situ with a repeatedly replenished solution of fixative.

        In practical terms, the requirements for tissue sampling, organ
    weighing, etc., have to be balanced against the need to get good
    fixation, and this frequently presents conflicts of interest between
    those involved in analysing the animal tissues. Short perfusion
    fixation (section 5.5.2), with paraformaldehyde and subsequent
    secondary fixation in formalin or formal-acetic acid in the case of
    paraffin sections and light microscopy, and glutaraldehyde and osmium
    tetroxide in the case of epoxy resin, semi-thin sections, may help to
    provide a compromise. When the protocol for killing requires only
    immersion fixation, then a few animals in each group must be kept for
    perfusion fixation, if adequate tissue examination is going to be
    achieved. A study using only immersion fixation is scarcely worth the
    labour and cost involved.

        The selection of the chemical fixative and its method of delivery
    to the tissue site are chosen according to the type of information
    that is required. Systemic perfusion with the chosen fixative is
    always to be preferred. In general, formaldehyde is suitable for
    studies that require widespread tissue sampling, low resolution of
    cytological detail, or the use of special histochemical stains to
    assess the chemical content of abnormal cells and tissues. Conversely,
    in studies focusing on more selected areas of the nervous system or
    requiring resolution of fine structural detail, the use of
    glutaraldehyde and osmium tetroxide is mandatory. While the latter
    method was originally introduced for use in transmission electron
    microscopy, an increasing number of laboratories use the technique
    routinely to exploit the remarkable resolution afforded by the light
    microscope.

    5.4.3  Principles of tissue sampling

        Extensive sampling of tissue is essential for the initial
    assessment of a suspected neurotoxic injury. Organs, other than the
    nervous system, which should be examined, are listed in WHO (1978).
    Neural tissue should include: brain, including cerebellum, brain stem,
    pituitary gland, eye with occulomotor muscles and optic nerve
    attached, spinal cord, several sensory ganglia, sciatic nerve and its
    branches from its exit from the vertebral column to the level of the
    ankle, and selected muscles innervated by the sciatic nerve and its
    branches. Other regions less commonly sampled include: olfactory
    epithelium and tubercles, inner ear and labyrinths, plantar nerves and
    skin receptors, autonomic ganglia, and nerves and organs of
    innervation, such as the gut. The brain contains many sites known to
    be vulnerable to chemical agents and should be examined as thoroughly
    as possible. The use of an atlas such as that for the rat brain (Zeman
    & Innes, 1963) will be valuable. The brain stem should be sectioned
    just rostral to the cerebellum, the slice passing through the inferior

    colliculi. The cerebral hemispheres can then be sliced coronally
    (transversely). Olfactory lobes may be taken if questions of
    inhalation and olfaction are being studied. The cerebellum is removed
    by cutting through the peduncles that attach it to the brain stem.
    Sagittal sections are obtained by cutting through the midline, and
    parallel sections are also taken. Emphasis is usually placed on the
    retina and optic nerves, since these are often heavily involved in
    toxic diseases affecting neurons, axons, and myelin. Ototoxicity is
    also a common event, though the technical difficulties associated with
    the examination of the inner ear and labyrinths have prevented
    widespread study of this phenomenon. Olfactory lobes and associated
    epithelium are rarely examined, but should be taken if questions of
    inhalation and olfaction are being studied.

        The spinal cord contains ascending and descending nerve fibre
    tracts that commonly undergo changes in myelinopathies and
    axonopathies (Table 3). Since the latter are tractable diseases with
    distal accentuation, it is critical to sample various levels from the
    medulla oblongata (where the gracile tract terminates) to the sacral
    region. By taking transverse sections of the spinal cord, it is
    possible to make a simultaneous evaluation of white matter (myelinated
    tracts and glial cells), gray matter (neurons, dendrites, proximal
    axons, glial cells), blood vessels, and associated tissues.

        Dorsal root and cranial nerve ganglia contain primary sensory
    neurons that may display pathological changes in certain
    neuronopathies and axonopathies (Figs. 3, 4). Structural alterations
    in blood vessels and myelin may also occur under certain conditions.
    The sensory ganglia of the Vth cranial nerve, at least 4 sensory
    ganglia from the cervical bulb region (C5-T1 spinal levels), and 4
    from the lumbosacral regions (L3-S2 spinal levels), should be sampled
    for examination in longitudinal section. Corresponding dorsal and
    ventral spinal roots, vulnerable in toxic demyelinating diseases,
    should also be studied.

        Peripheral nerves commonly display changes in neurotoxic diseases
    affecting myelin, neurons, and axons (Table 3). As the last display
    distal, retrograde changes in dying-back axonopathies, a common
    response to chemical attack, it is critical to examine several levels
    of vulnerable nerves. Samples of the sciatic nerve and its branches
    are usually taken commencing adjacent to the vertebral column and
    terminating in the distal regions of the sural nerve, peroneal nerve,
    and/or the tibial nerve. The fine branches of the tibial nerve that
    leave the main trunk below the knee are especially vulnerable in toxic
    neuropathies, because they contain very large myelinated nerve fibres.
    Terminal regions of sensory and motor nerve fibres can be examined by
    searching for twigs supplying intrafusal (spindles) and extrafusal
    muscle fibres, respectively (Schaumburg et al., 1974). Routinely,
    anterior tibial muscles, and gastronemius with suralis muscles are

    examined. Muscle tissue is also valuable for assessing whether there
    is denervation atrophy, evidence of toxic myopathy, nerve sprouting,
    or other regenerative activity (Pearson & Mostofi, 1973).

        Tissue sampling should not only be tailored to the goal of the
    study but also to the method of fixation. In general, blocks of tissue
    several millimeters thick may be taken if the tissue has been fixed
    with formaldehyde and is destined for paraffin embedding. For tissue
    fixed with glutaraldehyde, the relatively slow rate of penetration of
    osmium tetroxide necessitates the use of thin (1 m) slices of tissue.
    It is often helpful to mark a small area of tissue (with Indian ink or
    a small cut) to maintain the orientation. Elongate structures, such as
    peripheral nerves and spinal cord, are most readily studied in
    transverse sections, though important information is acquired by
    complementary examination of longitudinal sections. Particularly
    informative for the identification of pathological changes in
    peripheral nerves is the technique of microdissection (teasing apart
    intrafascicular tissue), which yields long lengths of individual
    myelinated nerve fibres suitable for light microscope examination.
    Inspection of these preparations can demonstrate rapidly the presence
    of primary demyelination and remyelination and axonal degeneration and
    regeneration. The minor changes characteristic of early toxic
    neuropathies, readily missed in the examination of sections, can be
    easily detected by this method. Other parts of the nervous system are
    technically difficult to examine by microdissection because of the
    absence of a collagen network to support the individual nerve fibres.

        While it is prudent to embed tissues in paraffin wax or epoxy
    resin promptly, regions that will not be examined immediately can be
    stored in a solution and at a temperature appropriate to the method of
    fixation. Loss of tissue quality will occur in proportion to the
    length of storage.

    5.4.4  Preparation of tissue for examination

        While gross tissue changes can be visualized with the naked eye or
    by examination with a stereoscopic binocular microscope, and surface
    detail can be resolved with the scanning electron microscopy, routine
    examination of cellular structure requires the use of tissue sections. 
    This inevitably places severe limitations on the amount of tissue that
    is practical to study. Sections must be such that the energy beam
    (light, electrons, X-rays) of the chosen microscope can be transmitted
    through the tissue. Relatively thick sections containing several
    layers of cells can be visualized with the light microscope, but the
    ability to resolve cytological detail decreases as section thickness
    increases. The use of semi-thin sections allows the investigator to
    exploit the maximum available resolution of the light microscope. Much
    thinner sections are required for the assessment of tissue by
    transmission electron microscopy (Pease, 1964; Hayat, 1973).

        Chemically-fixed tissue must be supported by a pliable material to
    facilitate the sectioning process and to prevent disintegration of the
    sectioned tissue. Paraffin wax or epoxy resin are routinely used: the
    former for routine work, the latter for more specialized studies.
    Since both types of embedding media are water insoluble and therefore
    immiscible with the fixative solutions, water is first removed by
    stepwise dehydration in ethanol or methanol immersing the tissue in an
    aromatic hydrocarbon (clearing solution) that is miscible both with
    ethanol and the embedding medium. Dehydration, clearing, and
    infiltration with an embedding medium are procedures requiring
    considerable care and time to avoid tissue drying and distortion.
    Automatic systems are available and, for paraffin embedding, these are
    routinely used in histological laboratories. Once the tissue is
    infiltrated and embedded, the medium must be allowed to harden by
    cooling (paraffin wax) or heating (epoxy resin).

        Sections of tissue are prepared with the aid of knives mounted in
    microtomes. Steel knives are suitable to cut thick (4-15 m) sections
    of paraffin embedded tissues, whereas knives prepared by scoring and
    breaking strips of special glass are needed to prepare sections (1 m)
    of tissue embedded in the harder epoxy resin. The sections of tissue
    prepared with steel knives permits study of larger tissue areas than
    is possible with sections prepared from epoxy blocks, though the
    latter afford a higher resolution of cytological detail. With new
    developments in the preparation of glass knives and the design of
    microtomes, ever larger sections can be routinely prepared by the
    latter technique. Diamond knives mounted in ultramicrotomes are used
    to obtain the thin sections (50 nm) required for transmission electron
    microscopy.

        Contrast of cytological detail in sectioned material can be
    enhanced by the use of special light-microscope methods (phase,
    Nomarski, fluorescence) or, more commonly, by the introduction of
    stains that render structures light-dense. A wealth of cytological
    stains is available for use with paraffin sections, whereas only a few
    general purpose stains are usually used to enhance structural detail
    for the light-microscope assessment of epoxy sections. Heavy metal
    stains that are electron dense are needed to visualize fine structure
    by transmission electron microscopy.

    5.4.5  Recognition of artefact

        Some degree of artefact is unavoidable, irrespective of the care
    taken in preparing tissue for morphological examination, and
    recognition and identification of the sources of artefact are of the
    utmost importance in the morphological assessment. Possible artefacts
    are legion and stem from various sources: e.g., poor tissue
    preservation from failure of fixative to penetrate tissue; shrinkage
    induced by inappropriate dehydration or drying; traumatic changes

    associated with excision of the tissue (bending, stretching) or rough
    handling during processing; wrinkling of sections and precipitation of
    stain (Spencer & Bischoff, 1982).

    5.4.6  Recognition of normal structural variations

        While the gross and fine structure of the nervous system of any
    species and age is remarkably constant from animal to animal, it is
    critical to recognize the irregular occurrence of normal variations in
    cellular structure that may lead the inexperienced observer to an
    inappropriate conclusion. These variations include ectopic structures
    (e.g., sensory neurons misplaced in the sciatic nerve), local
    oedematous or proliferative changes (e.g., at a site of unrecognized
    trauma or infection), and those that accompany advancing age (Johnson,
    1981). The latter are especially important in long-term neurotoxicity
    studies, and there is no substitute for rigorous comparative study of
    control animals of the same age. Examples of normal age-related
    changes include neuronal pigmentation by lipofuscin, localized
    demyelination and remyelination, scattered neuronal loss, and axonal
    dystrophy and degeneration.

    5.4.7  Qualitative versus quantitative approaches

        While there can be no substitute for a thorough qualitative
    assessment of the morphological state of the nervous system,
    morphometric methods are required for a more precise description of
    these changes (Weibel, 1979; Bonashevskaya et al., 1983). This is
    especially important for the description of minor changes, or when the
    tissue reactions of different groups of animals are being compared.
    Randomization of samples is essential to minimize subjective
    influences, and a statistical approach is required both for the design
    of the morphometric study and the analysis of the data. Tissue
    sampling, group size, age, sex, body weight, and length are important
    considerations in this regard. This type of morphometric approach has
    been helpful in measuring tissue reactions to toxic chemicals in the
    form of: the numbers of fibres affected in specific pathways subject
    to degeneration, changes in cell populations in vulnerable parts of
    the brain, and the relationship between fibre diameter and intermodal
    length of peripheral myelinated nerve fibres prepared by
    microdissection (Dyck et al., 1984).

    5.5  Specific Procedures

    5.5.1  Introduction

        Strategic considerations in planning how to proceed must depend on
    the state of knowledge of the possible effects. When it is unknown
    whether damage occurs in nervous tissue, a wide-ranging examimation of
    the nervous system with the light microscope is appropriate. Most

    structural changes induced by neurotoxic chemicals can be detected
    using the primary methods given below. More specialized, secondary
    methods can be applied subsequently for more detailed studies.

    5.5.2  Primary methods

        There are strong convictions among experienced experimental
    neuropathologists as to which fixation method should be chosen for the
    primary screening. Some believe the conventional approach using
    formaldehyde-based fixatives and paraffin embedding is most
    appropriate; others prefer the more contemporary method of examining
    epoxy sections of tissue fixed in glutaraldehyde and osmium tetroxide.
    Each method has significant advantages and disadvantages and both
    complement each other. Since it is impractical to use both methods for
    the same animal, a prudent investigator will include a sufficient
    number of animals to permit the parallel use of both procedures.

    5.5.2.1  Formaldehyde/paraffin method

    (a) Tissue fixation and excision

        This technique is most suited for the study of large areas of
    tissue at low resolution and with special histochemical stains
    (section 5.5.3.3). Tissue can be removed from the carcass and fixed by
    immersion in formalin (10%) containing 2% acetic acid (added just
    before use) to improve penetration and hardening. Making up the
    solution in 80% alcohol rather than water to hasten penetration
    further has sometimes been recommended, but this is unnecessary and
    may make dissection more difficult.

        Rapid and careful transfer of tissues from the carcass to the
    fixative solution will retard artefactual cell distortion. In a
    general post-mortem dissection, the brain, including the cerebellum
    and brainstem, are removed and transferred to fixative, prior to the
    removal of other tissues. Retrieval of the spinal cord from its spinal
    column is difficult and time consuming; it is recommended that the
    whole spinal column with cord be excised quickly, trimmed of as much
    muscle as possible, and immersed in the fixative intact. Once this
    portion is fixed, the spinal cord and ganglia can be separated from
    the column with greater care. Alternatively, after approximately 24 h
    of fixation, the spinal column can be immersed in 5% formic acid
    decalcifying solution. The column can then be sliced with little
    difficulty.

        To assist with the examination of peripheral nerves, tissue is
    placed on a stiff card before immersion to retain orientation.
    Optimally, the nerve to be excised should be exposed and, prior to
    removal from the carcass, bathed with fixative for approximately
    5 min. This will begin to introduce rigidity into the tissue and
    retard decay artefact. A length of nerve can then be excised, quickly

    placed on a card and pinned at the tip of both cut ends to retain
    position after immersion. Care should be taken to avoid overstretching
    the nerve as it is pinned to the card. Samples of these tissues can
    then be sectioned in a transverse or longitudinal direction. Muscle
    can be treated in the same manner. Another technique is to take off
    the limb at the hip or shoulder joint, remove the skin and fix en bloc
    in formalacetic acid solution. The limb can then be decalcified in 5%
    formic acid solution for approximately two weeks as noted earlier for
    the spinal column. It can then either be embedded in paraffin wax as
    one large block or further dissected.

        Tissue preservation is greatly improved by perfusing animals
    systemically for 10 min with phosphate-buffered paraformaldehyde
    (section 5.5.2.2), a method that simultaneously preserves all body
    organs. Tissue can then be sampled in a more leisurely manner without
    fear of introducing post-mortem changes (section 5.5.2.2).

    (b) Dehydration and embedding

        Relatively large tissue samples can be embedded in paraffin wax,
    and this is particularly suitable for localizing and identifying
    lesions in the brain. Automated processing for paraffin embedding has
    greatly diminished the time requirements for preparing tissue samples.
    Most automated systems will conform to a variety of preparation
    schedules and process great numbers of tissue samples at one time.
    There are systems that operate under combined heat and vacuum, the
    great advantage being rapid processing. This combination is
    deleterious and may cause unnecessary shrinkage and brittleness, which
    will distort cell structure and hinder interpretation. Paraffin
    sections routinely cut 4-6 m thick, are suitable for certain
    histochemical techniques and can easily be treated with special stain
    to enhance particular cellular components.

    (c) Section staining

        Four staining methods are routinely used to enhance cellular
    detail in paraffin sections.

    Hematoxylin and eosin

        Routinely used on paraffin sections, this method stains nuclei
    blue to purple, cytoplasm and Nissl substance (and mast cell granules)
    blue, and myelin sheaths pink. Nuclear density can be readily
    assessed, as well as an increase or decrease in the size of nuclei and
    the number of cells. Neuronal damage is readily detectable. Nuclei and
    cytoplasm of other cell types, such as microgila and vascular
    endothelial cells, are also well defined, though cresyl violet may be
    preferred.

    Cresyl fast violet

        This method enhances identification of cell-population changes and
    is more aesthetically pleasing than other basic aniline dyes, such as
    thionine or gallocyanine-chrome alum. DNA stains pale blue and RNA
    purple-blue. Hypertrophy of astroglial cytoplasm and changes in
    neuronal nuclei and cytoplasm are readily observed.  Chromatolysis,
    poorly demonstrated in hematoxylin- and eosin-stained sections, is
    more apparent when stained with cresyl violet, as are most cells in
    peripheral nerves.

    Glees and Marsland's stain

        This method is easy to carry-out and selectively stains
    neurofibrillary components in axons (Marsland et al., 1954).
    Longitudinal sections demonstrate the state of terminal innervation in
    motor and sensory fibres. Empty endplates are often recognized by the
    observation of remaining clumps of nuclei on the muscle surface.

    Sudan Black B

        This method is useful to reveal nerve terminals in muscles
    (Cavanagh et al., 1964).

    5.5.2.2  Glutaraldehyde/epoxy method

    (a) Tissue fixation and excision

        This technique is best suited for the resolution of early
    pathological changes with the light microscope and is required for
    transmission electron microscopy (section 5.5.3.5). Whole body
    perfusion with buffered fixators (paraformaldehyde) followed by
    glutaraldehyde) delivered under pressure (100-150 mm Hg) is the only
    satisfactory method to prepare brain and spinal cord sections, and
    this method has the advantage of optimally fixing all other organs
    simultaneously. For this purpose, the animal is deeply anaesthetized
    with a solution (e.g., sodium pentobarbital) containing a small
    percentage of heparin to facilitate circulation. Subsequent procedures
    should be performed rapidly and smoothly to prevent the development of
    anoxic damage and to ensure global fixation. The animal is secured on
    its back and the rib cage cut bilaterally and reflected backwards to
    expose the heart. After slitting the pericardium, the right atrium is
    opened to drain circulating blood. The ventricular apex is then
    excised to provide access to the aorta, and a cannula spurting a
    column of fixative is inserted through the ventricular opening to the
    apex of the aorta and clamped in place. The circulatory system is
    cleared by a brief, initial perfusion with paraformaldehyde or saline,
    and followed by more prolonged (e.g., 10-15 min) perfusion with
    glutaraldehyde.

        Rigid tissues from a fixed carcass are less susceptible to
    handling artefact. The nervous system can be sampled in any order. In
    some laboratories, the entire nervous system, including brain with
    optic tract and nerves, spinal cord, spinal roots, ganglia, and
    peripheral nerves are exposed in the carcass, then carefully detached
    intact and removed. When this is accomplished, there is no need
    initially to mark tissues for orientation. It is imperative to avoid
    tissue drying during dissection and excision. Exposed tissues should
    be bathed with fixative and tissues should be manipulated as little as
    possible with forceps.

    (b) Dehydration and embedding

        After glutaraldehyde fixation, tissue samples are post-fixed for
    2-3 h by immersion in buffered 1-2% osmium tetroxide in the cold. The
    samples are then dehydrated stepwise in ascending concentrations of
    ethanol, immersed in acetone or toluene, infiltrated with a solution
    of epoxy resin and finally, embedded and polymerized in epoxy resin.
    For light-microscope examination, 1 m sections are cut with specially
    prepared glass knives mounted in an ultramicrotome (an instrument
    designed specifically for the purpose of obtaining the thin sections
    suitable for electron microscope examination). New techniques and
    equipment have recently become available which allow larger tissue
    samples to be prepared for sectioning in epoxy resin. Sections one
    micrometer thick can be examined by phase-contrast optics or by bright
    field after a brief staining with 1% toluidine blue.

    (c) Section staining

        Thick (1 m) sections of tissue fixed in glutaraldehyde and osmium
    tetroxide and stained with borate-buffered 1% toluidine blue allow
    resolution of structures as small as a single mitochondrion. Brain and
    spinal cord neurons display clearly defined nuclei, nucleoli, issl
    substance, and lipofuscin. Dendrites stand out against a more darkly-
    stained neuropil. Small, densely staining oligodendrocytes are readily
    distinguished from large, pale astrocytes. Myelinated nerve fibres
    contain a pale axon and darkly-stained myelin. In cross-sections of
    peripheral nerves, large and small diameter myelinated fibres, as well
    as the smallest unmyelinated axons, can be seen. Motor and sensory
    nerve terminals are detectable in muscle.

    5.5.3  Special methods

    5.5.3.1  Peripheral nerve microdissection

        Isolated fibre preparations provide information that may be missed
    in cross-sections, and it is recommended that this procedure be
    included in any study concerned with toxic neuropathy. Several
    fixation methods are available to prepare peripheral nerves for
    microdissection. Perfused tissue is ideal, though satisfactory

    preparations can be obtained by bathing the tissue with fixative prior
    to excision. Post-excision immersion in fixative can also be used,
    though the act of transecting the nerve introduces major fibre
    artefacts that are not restricted to the cut ends and are readily
    mistaken for pathological changes.

        Optimal preparations suitable for high-resolution light microscopy
    require the use of glutaraldehyde, post-fixation with osmium
    tetroxide, stepwise dehydration, and infiltration with a low-viscosity
    epoxy resin. This is an ideal medium to tease apart nerve fibres and
    is suitable for the storage of tissue at low temperature. Mounted
    fibres can be polymerized by heat to affix them to a glass slide and
    prevent squashing by a coverslip (Spencer & Thomas, 1970). Bright-
    field examination will reveal changes in overall structure, and
    Nomarski differential interference microscopy can be used to study the
    axon.

        Older techniques yield poorer preservation, low resolution of
    detail, and isolated fibres are susceptible to squashing, an important
    consideration if morphometric study is planned. One method involves
    formalin fixation, post-fixation in osmium tetroxide, and infiltration
    with a dilute solution of glycerol, which is used to support the
    fibres during microdissection. Single fibres are mounted on the
    microscope slide, cleared with cresol, blotted dry, and mounted with
    pure glycerin (Thomas, 1970). Sudan Black B can also be used to stain
    tissue prior to microdissection. Nerves are microdissected with the
    aid of a pair of mounted needles and stereoscopic dissecting
    microscope. Epineurial connective tissue is removed and a small
    fascicle selected. After splitting the perineurium surrounding the
    fascicle, the sleeve is removed to expose the intrafascicular tissue
    containing the nerve fibres. The tissue is then repeatedly split into
    progressively smaller longitudinal bundles until individual fibres are
    obtained. These are picked up on the end of sharpened wooden
    applicator sticks and transferred to a clean glass slide and provided
    with a coverslip. Alternatively, a rapid "squash" preparation can be
    obtained by teasing a small bundle of fibres loosely apart to create a
    mesh formation on a microscope slide; a coverslip is then added for
    microscope examination.

    5.5.3.2  Frozen sections

        Frozen sections lend themselves well to enzyme histochemical, 
    immunocytochemical,  and  silver-impregnation procedures, but for the
    purposes of a large-scale study, this method is impractical. Tissues
    fixed instantaneously by immersion in liquid nitrogen are particularly
    vulnerable to cell distortion and/or destruction caused by excessive
    temperature changes. Encasing tissue samples in gelatin or albumin is
    advisable, especially for brain tissues (Crane & Goldman, 1979).
    Ice-crystal formation is a disruptive artefact that produces a
    vacuolated appearance in frozen sections. To minimize crystal

    formation, it is important to freeze tissues as rapidly as possible
    and maintain critical temperatures throughout the cutting procedure to
    avoid thawing and refreezing. Typically, frozen sections are between
    10 and 15 m thick.

    5.5.3.3  Histochemical methods

        Using these procedures, it is possible to assess the activity of
    cell energy systems, the rate of use of substrates for processes of
    oxidative phosphorylation, and the functional state of intracellular
    organelles. The activities of succinate dehydrogenase and malate
    dehydrogenase, enzymes of the Krebs cycle, are well known, the former
    being closely linked with mitochondrial membranes. Enzymes of the
    hexose monophosphate shunt (glucose-6-phosphate dehyrogenase and
    gluconate-6-phosphate dehydrogenase) and those of the electron
    transport chain (NAD and NADP disphorase) should also be studied. The
    rate of exchange of nitrogenous bases can be examined by the
    determination of activity of dihydroorotate dehydrogenase for fatty
    acids; monoamine oxidase for biogenic amines; and acetylcholinesterase
    for acetylcholine.

        Histochemical enzyme assays are performed on freshly frozen
    sections cut on a cryostat (Dubowitz & Brooks, 1973). Tissues are
    quick-frozen in liquid nitrogen. Control and experimental tissues
    should be mounted on one slide and reactions carried out using the
    methods outlined in Pearse (1968).

        The following methods can be used to demonstrate the functional
    state of nerve cells (Pearse, 1968):

        (a) RNA (Brachet's method with ribonuclease control);

        (b) DNA (Feulgen's method with deoxyribonuclease control);

        (c) Total nucleic acids (Einarsen's method);

        (d) Total protein (sublimate with bromophenol blue);

        (e) Sulfhydryl groups (Barnet and Zeligman's method);

        (f) Glycogen and glycosaminoglycans (amylase control-PAS
            reaction); and

        (g) Lipids (Sudan III, Sudan Black, Nile Blue methods).

        To express histochemical reactions quantitatively, the following
    formula for Mean Histochemical Index (MHI) can be used:

                         MHI = 3a + 2b + 1c + 0d
                                        N

    where 3, 2, 1, and 0 are the degrees of intensity of colour (from 3
    to 0), and a, b, c, and d are the numbers of the cells with the given
    intensity of colour; the denominator N is the number of cells counted.
    The results allow statistical calculation of the assays and estimation
    of the reliability of the results.

    5.5.3.4  Golgi method

        This unique silver-impregnation method demonstrates the shape and
    surface characteristics of entire neurons, including cell bodies and
    their processes. Golgi preparations are particularly useful in the
    study of dendritic arborizations and synaptic associations (Santini,
    1975). However, the technique is capricious in that only a few neurons
    are unpredictably demonstrated: many preparations must therefore be
    examined to gather a representative sample.

        In the Rapid Golgi method, tissue is fixed with formaldehyde or
    glutaraldehyde and post-fixed with osmium tetroxide. The tissue is
    then impregnated with silver nitrate, dehydrated, and infiltrated with
    celloidin. Thick (120 m) sections are prepared, mounted, and examined
    by bright-field microscopy.

    5.5.3.5  Transmission electron microscopy

        This technique is laborious and requires extensive training.
    However, it has resulted in extensive advances in the understanding of
    cellular processes in neurotoxicity. There is little need for such
    exacting methods to determine the presence of changes in cell
    structure, few experimental protocols require electron microscopy. The
    transmission electron microscope is properly used to confirm and study
    further the nature of lesions already shown and mapped by light
    microscopic methods. It is all too easy to seek, and find, changes to
    which no significance will ultimately be attached. Until the pattern
    of change induced by the chemical under study has been well identified
    by light microscopy, electron microscopy should not be considered. On
    rare occasions, such as in toxic disorders of unmyelinated axons,
    examination by transmission electron microscopy is necessary to
    localize cellular changes not revealed by the methods discussed
    previously. Thin (50 nm) plastic sections are prepared with the aid of
    a diamond knife mounted in an ultramicrotome and subsequently
    impregnated with heavy metals for electron-microscope examination
    (Pease, 1964; Hayat, 1973).

    5.5.3.6  Other anatomical methods

        There are numerous additional specialized anatomical methods that
    are waiting to be applied in the study of neurotoxicological issues. 
    These include histological techniques to trace anatomical pathways
    (with horseradish peroxidase) and cellular activity. Examples of the
    latter are the use of autoradiography to localize the distribution of
    radiolabelled precursors such as thymidine and 2-dexoyglucose
    (Sokoloff et al., 1977). Other methods susceptible to light and
    electron microscope analysis include immunocytochemical studies of
    receptors, proteins, and other cell structures (Emson, 1983).
    Fluorescence microscopy, particularly for the localization and
    detection of regional concentrations of catecholamines using the
    formaldehyde vapour or glyoxylic acid technique of Falck et al.
    (1962), supplements quantitative biochemical methods in evaluating
    alterations in catecholamine levels. Fluorescence microscopy is also
    of value in antigen and antibody localization techniques. Certain
    light and electron microscope techniques may be combined, such as the
    examination of fine structure of single teased nerve fibres (Spencer &
    Lieberman, 1971; Ochoa, 1972) or of Golgi-stained preparations.
    Advanced methods in electron microscopy include scanning (for surface
    features), energy-dispersive X-ray analysis (for detection of elements
    of high relative atomic mass) and energy-loss (for detection of
    elements of low atomic mass). Freeze-fracturing tissue can reveal
    details of the internal surfaces of cellular membranes when replicas
    are examined by transmission electron microscopy (Hayat, 1973).

    5.6  Conclusions

        The information derived from morphological studies is highly
    relevant to the interpretation of biochemical, neurophysiological, and
    behavioural data. Structural changes have always been the firm
    foundation on which analysis of clinical neurological disease has been
    based. Both diagnosis and prognosis depend heavily on previous
    neuropathological experience for their accuracy. Moreover, treatment
    of neurological disease can only be satisfactorily planned by using
    the knowledge gained from experimental and morphological studies that
    have provided an understanding of the pattern by which neural cells
    are affected. Indeed, in the majority of human intoxications of the
    nervous system, knowledge of the structural changes is based almost
    exclusively on animal studies. There is no reason to believe that it
    will be less so in the future.

    6.  BIOCHEMICAL AND NEUROENDOCRINOLOGICAL METHODS

    6.1  Introduction

        Biochemical tools are valuable for the study of neurotoxicity.
    They have the potential for both identifying toxic compounds and
    delineating mechanisms of action of known toxic substances. However,
    the range of biochemical techniques is vast and careful decisions must
    be made in devising an effective research strategy. The most important
    decision is identification of the objective of the research, which may
    be conveniently categorized into three groups: determining mechanisms
    of action of neurotoxic substances, identifying exposed individuals,
    and screening for toxic conditions.

        Devising an effective biochemical screen is the most challenging
    of the three objectives.  It requires the selection of biochemical
    parameters that are general enough to indicate toxicity resulting from
    multiple sites of damage, but specific enough to prevent the
    accumulation of false positives. No single biochemical parameter is
    likely to suffice. An effective screen must include indices of each of
    the major functions of nervous tissue (e.g., cellular metabolism,
    neuronal propagation, and neurotransmission). It might include
    estimates of RNA and/or protein synthesis, lysosomal enzymes, membrane
    transport, channel and neurotransmitter receptors, or their turnover.
    The precise choice of biochemical indices must be limited by the
    expertise and equipment of individual laboratories, but every attempt
    must be made to consider a range of functional events rather than a
    single parameter.

        Considerable discussion has centred on the use of in vitro
    procedures for screening potential neurotoxic compounds. The
    advantages of an in vitro approach are obvious. It is less expensive
    and less time-consuming than whole animal study and biochemical
    techniques can be employed under precise, standard conditions.
    In vitro procedures must include both parent compounds and potential
    toxic metabolites. However, the exclusive use of in vitro methods
    would fail to detect compounds in which the neurotoxicity results from
    alterations in non-neuronal systems.

        Theoretically, the choice of biochemical tools for the
    identification of mechanisms of toxicity is the simplest of the three
    objectives. Biochemical tools are virtually limitless, but they can be
    chosen on the basis of known information about the toxicity of the
    compound. Scientists can then proceed systematically from one level of
    analysis to another. By synthesizing data from behavioural, neuro-
    physiological, neuroendocrinological, neuropharmacological, and
    neuropathological studies, the biochemist can employ successively more
    selective tools to determine the initial biochemical lesion induced by
    the toxic compound.

        The purpose of this section is to review the biochemical
    approaches that may be useful for identifying and further
    understanding the effects of toxic compounds on the nervous system.

    6.2  Fractionation Methods

        Neural tissue has several features not shared by other tissues. In
    particular, neural tissue exhibits considerable cellular,
    morphological, and chemical heterogeneity. It is often desirable to
    precede biochemical procedures with some attempt to decrease this
    complexity. One approach is to separate the tissue into discrete parts
    of brain prior to analyses. A second approach is to separate specific
    cell types. A third involves subcellular fractionation procedures. In
    many cases, it may be desirable to use a combination of these
    fractionation procedures.

    6.2.1  Brain dissection

        Brain tissue can be fractionated into any number of potential
    sections. However, the most often used scheme is based on that
    described by Glowinski & Iverson (1966). In this method the brain is
    divided into relatively discrete units of cerebellum, thalamus,
    hypothalamus, striatum, hippocampus, etc., by using visible anatomical
    landmarks. Such dissection procedures have been very useful in
    describing the neurochemistry of different brain areas and are
    essential in evaluating the effects of toxic compounds on molecules
    (such as neurotransmitters) that are not uniformally distributed in
    neural tissue. Because of the regional variability of brain chemistry,
    whole brain analyses are seldom useful for evaluating brain function. 
    More importantly, the use of the whole brain for studies of toxic
    compounds may fail to detect functionally-relevant alterations in
    specific brain areas. For example, effects of lead on GABA levels have
    been reported for cerebellar tissue (Silbergeld et al., 1980) and
    acrylamide effects have been noted in striatal dopamine receptors
    (Bondy et al., 1981).

        Differences in the response to toxic compounds would be
    anticipated from differences in neurochemistry in various regions of
    the brain. In addition, many toxic substances are not distributed
    uniformly across brain areas (e.g., chlordecone (Fujimori et al.,
    1982) and manganese (Bonilla et al., 1982). Although not predictive,
    examination of such regions can provide a biochemical clue regarding
    possible molecular sites of interaction of the toxic compounds.

        In more precise dissection procedures, a small needle, of defined
    diameter, is used to obtain samples within anatomically defined areas
    (Palkovits, 1973; O'Callaghan et al., 1983). Such finite samples have
    been used in identifying several neurochemical events,  e.g.,
    neurotransmitter concentrations (Palkovits, 1973) and neuro-

    transmitter-induced phosphorylation of specific phosphoproteins
    (Dolphin & Greengard, 1981), and they may be especially valuable in
    locating the site of neural responses to toxic compounds.

    6.2.2  Isolation of specific cell types

        Although neural tissue is recognized by a marked heterogeneity of
    cellular elements, based on function and embyronic origin, these cell
    types are conveniently categorized as neurons or glia. Each cell type
    makes a unique contribution to neural function and their sensitivity
    to toxic compounds may differ. Several methods have been described for
    the bulk separation of neuronal and glial cell populations from whole
    brain or specific brain regions (Rose & Sinha, 1970; Appel & Day,
    1976; Magata & Tsukada, 1978). A dissociated cell suspension that
    avoids disruption of the cells is prepared by forcing the tissues
    through fine sieves. Proteolytic enzymes are used to facilitate the
    dissociation. The fraction enriched in neuronal perikarya is separated
    from the smaller glial cells (also containing some synaptosomes) by
    centrifugation. The cell separations are never complete and the
    procedures generally provide low yields. However, enrichment of the
    neurons and glia is usually sufficient fox determining whether a toxic
    agent interferes primarily with neuronal or glial metabolism. Isolated
    cell populations have been used to study the lipid and protein
    composition of membranes and the several metabolic differences between
    neuron and glia. Such information may provide baseline data for future
    studies of neurotoxic chemicals. Disadvantages of this method include
    loss of cell processes from the perikarya, reduction of cell surface
    area, and damage to the cell membrane. These factors must be
    considered in any metabolic study.

        To date, there have been few attempts to use such separation
    procedures in neurotoxicity and they are not recommended as an initial
    study. The method is laborious and time consuming and must be
    performed on fresh tissue. Only a limited number of samples can be
    simultaneously processed and it is never clear if the low yields
    result from random or specific loss of cell types.

    6.2.3  Subcellular fractionation

        The neuron can be divided into relatively discrete functional
    units. The cell body contains the metabolic machinery for the
    synthesis and packaging of macromolecules and for the general
    maintenance of cellular homeostatis. The long axonal process acts as a
    communication link, propagating electrical impulses from the cell body
    to the terminal and transporting vital nutrients and cellular
    components to more distal regions including the nerve terminal.
    Finally, the synapse functions to transfer chemically encoded
    information from one cell to another. Toxic substances may disrupt
    neuronal biochemistry at any one, or all, of these cellular sites.

        The isolation of subcellular organelles and specific membrane
    fractions can be a first approximation in determining the subcellular
    sites of action of toxic agents. Differential centrifugation
    procedures provide a rapid means of obtaining fractions consisting
    predominantly of a single cellular component (Gray & Whittaker, 1962;
    Cotman & Barker, 1974; Rose and Sinha, 1970). Subcellular
    fractionation is never totally effective, but it can produce an
    enrichment of organelles and cellular subfractions. Some of the more
    commonly used biochemical markers that can help identify the degree of
    enrichment are listed in Table 4. These markers can be used as
    potential indices of toxicity. For example, the effects of triethyltin
    on myelination in the developing brain has been studied by examining
    the activity of the myelin marker, 2',3'cyclic nucleotide 3'
    phosphohydrolase (Konat & Clausen, 1977). Merkuryva & Tsapkova (1982)
    have used several marker enzymes in their study on the toxic effects
    of ethanol on the nervous system.

        When subcellular fractionation is combined with regional analysis,
    the number of potential samples requiring examination can be
    overwhelming. Therefore, careful selection must be used in deciding
    which brain area and which subcellular fraction to investigate. This
    selection can be very difficult, especially when knowledge concerning
    a particular compound is limited. The recent introduction of
    immunochemical methods promises to facilitate this task, not by
    reducing the difficulty of the selection, but by increasing the number
    of chemicals that can be subjected for analysis.

        Immunochemical methods have been used to identify molecules
    associated with various cell types, for localizing enzymes responsible
    for the synthesis of neurotransmitters and for identifying proteins
    specific to the nervous system. While the procedures initially require
    the preparation of antisera to highly purified proteins, once the
    antisera are available, they can be sensitive tools for identifying
    specific changes and for characterizing different cell populations. In
    addition, monoclonal antibodies can be used to determine the mechanism
    of action of toxins.

        These procedures are especially valuable in studying neuronal
    peptides. These peptides (e.g., substance P, encephalins, hypothalamic
    releasing factors, somatostatin, etc.) play a significant role as
    neuromodulators (Snyder & Inms, 1979). Many are potent at low
    concentrations. Prior to the use of immunological procedures, peptide
    molecules could only be examined by using bioassay techniques and
    these were not always specific for particular molecular species.
    Radioimmunoassay techniques are more specific. Assay kits for various
    peptides are now available commercially, and their number is rapidly
    increasing.

        Table 4.  Examples of biochemical markers of brain subcellular fractions
              and cell types
                                                                                

    Fraction                  Marker                       Reference
                                                                                

    nuclei                    DNA                          Steele & Busch (1963)

    nuclear membrane          RNA polymerase               Roeder & Rutter (1969)

    cytosol                   lactate dehydrogenase        Kornberg (1955)

    microsomes                NADPH; cytochrome c          Miller & Dawson (1972)
                              oxidoreductase (rote
                              none insensitive)

    Mitochondria

    inner membrane            D-3-hydroxybutyrate          Fitzgerald et al.
                              NAD+ oxidoreductase          (1974)

    outer membrane            monoamine oxidase            Schnaitman et al.
                                                           (1967)

    lysosomes                 -glucuronidase              Fishman & Bernfeld
                                                           (1955)
                              -glucosidase                Robins et al. (1968)
                              -galactosidase              Robins et al. (1968)
                              N-arylamidase                Boer (1974)

    myelin                    2',3'-cyclic nucleotide-     Konat & Calusen (1977)
                              3'-phosphohydrolase

    synaptosomes              Na+K+-activated              Verity (1972)
                              oubain-sensitive
                              ATPase

    neuronal cells            guanyl cyclase               Goridis et al. (1974)
                              tyrosine hydroxylase         Kuczenski & Mandell
                                                           (1972)

    oligodendroglial cells    glyceride galactosyl         Radin et al. (1972)
                              transferase
                                                                                
    
        Because the method is very recent, radioimmunoassay procedures are
    only beginning to be use in neurotoxicology. However, using such
    methods, Hong and his colleagues (Ali et al., 1982; Hong & All, 1982)
    have successfully identified several peptide responses following
    chlordecone treatment. The major limitation of the method for
    identifying toxic compounds is the small number of purified compounds,
    the small number of available RIA kits, and the expense of obtaining
    radioiodinated compounds. With increasing availability of RIA kits,
    the procedure could become a powerful screening tool.

    6.3  DNA, RNA, and Protein Synthesis

        Changes in the amount of DNA can be used to detect whether toxic
    agents affect cellular proliferation and cell death. Polyploidy is
    uncommon in the nervous system, thus the DNA content of brain regions
    can be taken as an index of cell number. This can be related to tissue
    weight and RNA or protein content to estimate cell size. For example,
    Krigman & Hogan (1974) reported that lead exposure during early
    development reduced brain weight and decreased total brain proteins,
    but the number of brain cells was unchanged. DNA measurements have
    been particularly useful in studying the toxic effects of drugs on
    retinal photoreceptors (Dewar et al., 1977). It is also well-known
    that the DNA content changes after nutritional restriction during
    development, or after exposure to various toxic chemicals that
    influence nutritional status.

        Although the measurement of DNA content is relatively simple and
    in the nervous system can be an index of cell number, it is probably
    not a very sensitive or early indicator of neurotoxicity. For example,
    neuronal loss might be accompanied by glial proliferation and produce
    no change in total DNA. The earliest effects of a toxic substance on
    DNA will probably involve: (a) the disturbance of DNA-repair enzymes;
    (b) the intercalation of the substance into the DNA molecule; or
    (c) the direct binding of the substance to the nucleic acid moiety or
    its associated chromosomal proteins. Two to three percent of added
    mercuric chloride was reported to bind to the chromatin in cultured
    glial cells (Ramanujam et al., 1970). Since this mode of action will
    ultimately interfere with the functional integrity of the CNS, it
    offers great promise for the identification of toxic compounds.

        The total amount of RNA or protein could also be examined after
    toxic insult. However, neither total RNA nor total protein is likely
    to provide a sensitive index of toxic damage. Only in the extreme
    stages of neurotoxicity will total or even regional levels of these
    macromolecules be likely to change. More sensitive indices of their
    metabolism rely on estimates of synthesis or degradation.

        The rate of RNA or protein synthesis can be evaluated by using
    radiolabelled precursors and measuring their rate of incorporation
    into the macromolecule in vitro or in vivo. Although this

    technique is simple and sensitive, it assumes that the rate of
    incorporation of the precursor directly reflects the rate of
    macromolecular synthesis and this may not always be true (Dunn, 1977).
    Changes in labelling could also occur as a result of changes in
    cerebral blood flow, precursor transport, precursor pool size, energy
    charge, etc. Inhibition of amino acid transport across the blood-brain
    barrier has been reported for mercury (Partridge, 1976) and lead
    (Lorenzo & Gerwitz, 1977). Thus, it is also important to consider
    whether a toxic agent can cause changes in the amino acid or nucleic-
    acid-precursor pool sizes by using one of several methods available
    for compensating for precursor pool fluctations (Munro et al., 1964;
    Dunlop et al., 1975; Dunn, 1975). Changes in brain-protein synthesis
    have been reported after exposure to methylmercury (Verity et al.,
    1977), and carbon disulfide (Savolaiene & Jarvisalo, 1977). However,
    in many cases, only the absolute levels of incorporation of
    radioactive amino acids into protein fractions were studied and it is
    difficult to understand the primary mechanisms of action.

        Although these relatively crude methods are subject to several
    criticisms, all the variations that influence the incorporation of the
    radioactive precursor can ultimately influence cellular metabolism.
    These methods can, therefore, be useful for identifying metabolic
    disturbances. The relative amount of protein synthesis can also be
    estimated simply by measuring the polyribosome to free ribosome pool.
    This method is based on the assumption that during high rates of
    protein synthesis, the greatest proportion of the total ribosomes will
    be attached to an mRNA molecule. The relative polyribosome to monosome
    ratio has the advantage of allowing for normalization across animals
    and therefore reducing the interanimal variability that often plagues
    neurochemical assessments. Furthermore, the use of this method would
    detect toxic effects not only on protein synthesis but on the
    stability or synthesis of the ribosome itself.

        The rate of RNA and protein synthesis in the brain is very high
    and nerve cell functioning is dependent on protein metabolism. The
    turnover rates of brain macromolecules may vary considerably, and it
    is always going to be difficult to interpret any data dealing with the
    turnover of total RNA or protein. However, with the availability of
    newer separation methods for proteins (detergent gel electrophoresis
    combined with chromatofocusing cellulose ion exchange chromatography,
    etc.), soluble and bound polypeptides can be separated to study the
    synthesis of individual proteins at the subcellular level, in specific
    cell types and in localized brain regions. Large classes of RNA can be
    analysed by hybridization procedures or specific mRNA molecules could
    be targeted for investigation by using specific cDNA probes.

    6.4  Lipids, Glycolipids, and Glycoproteins

        Complex lipids, glycolipids, and glycoproteins have several
    important functions in neural tissue (Zuber, 1978). They constitute
    structural elements of the plasma membrane, act as components of ion

    channels, comprise portions of neurotransmitter receptors, and are
    major constituents of myelin. Complex lipids (e.g., phospholipids,
    cerebrosides, sulfatides, and gangliosides) make up neuronal
    membranes. Phosphatides (phosphatidylethanolamine,  phosphatidyl-
    choline, phosphatidylinositol, phosphatidylserine) are the most
    important group of phospholipids. They consist of some of the most
    metabolically active of the phospholipids (Prokhorova, 1974) and
    participate in the movement of Na and K ions through the neural
    membrane. A closely related group of phospholipids, the plasmalogens,
    are concentrated in myelin and constitute 18-30% of total brain
    phospholipids.

        Ganliosides are localized primarily in the plasma membrane
    (Karpova et al., 1978), but small amounts of monosialoganglioside are
    detectable in myelin. Receptors of neurotransmitters, such as
    serotonin, and other biologically-active compounds may contain
    ganglioside (Avrona, 1971), and evidence suggests that gangliosides
    may bind biologically-active compounds, as well as various
    neurotoxins, through their terminal N-acetylneuraminic acid moiety
    (Kryzhavosky, 1973). Glycosaminoglycans are also functional and
    structural elements of the neuronal membrane and have been reported to
    exhibit disturbed metabolism in various hereditary diseases of the
    mucopolysaccharidosis type (Constantopoulos et al., 1976; Vasan &
    Chase, 1976).

        The most interesting of the functionally-active, minor components
    of neuronal membranes are sialic acids, and particularly
    N-acetylneuraminic acid. As sialo-glycoproteins and gangliosides,
    N-acetylneuraminic acid participates in carrying out specific
    neuronal functions such as the establishment of synaptic contact
    neurotransmission, and axonal propagation (Partingron & Daly, 1979).
    Gangliosides and sialo-glycoproteins bind Ca++ at their hydrophilic
    ends and thus, by influencing the concentration of this ion, affect
    depolarization and neurotransmitter release.

        Myelin metabolism may be affected by a variety of toxic agents
    (Cammer, 1980). After administration of these compounds, there is
    often a delay before the appearance of neurotoxic signs and symptoms
    such as the ataxia indicative of peripheral demyelination. Central
    demyelination often occurs to a lesser extent. Compounds that cause
    disturbances in myelin metabolism may exert their effect directly on
    myelin-forming cells, or demyelination may be a secondary response to
    a disturbance of neuronal metabolism.

        Several compounds have been reported to alter complex lipids or
    their metabolism. Neonatal exposure to lead has been reported to
    decrease the total brain content of phospholipids, galactolipids,
    plasmalogens, and cholesterol (Van Gelder, 1978). Hydrogen sulfide and
    sulfur dioxide decrease total brain lipids and/or phospholipids
    (Haider et al., 1980). Meta-SystoxR (phosphorothioic acid

    O-[2-(ethylthio)ethyl] O,O-dimethyl ester mixture with
    S-[2-(ethylthio)ethyl] O,O-dimethyl phosphorothioate)
    (O,O-dimethyl-S-2 (ethylsulfinyl) ethylthiophosphate), an
    organophosphorus pesticide, has been reported to decrease brain levels
    of total lipids, phospholipids, cholesterol, esterfied fatty acids,
    and gangliosides (Islam et al., 1983). Merkuryva et al. (1978) found
    one of the early affects of carbon disulfide intoxication to be
    changes in the enzyme substrate system of N-acetylneuraminic acid,
    N-acetylneuraminate lyase. In the olfactory bulb of the rabbit,
    carbon disulfide led to an accumulation of N-acetylneuraminic acid
    because of a reduced degradation by -acetylneuraminate lyase. In some
    cases, the carbon disulfide-induced disturbance of the metabolism of
    cerebral glycoconjugates could be correlated with the changing of
    physiological parameters (e.g., lowering of the amplitude of the
    cortical evoked potential (Bokina et al., 1976)). Long-term exposure
    to lead acetate in aging rats was reported to decrease
    N-acetylneuraminic acid content in cerebral tissue and also to
    decrease the activity of the degradative enzyme (Merkuryva &
    Bushinskaya 1982). Long-term exposure to ethanol led to an
    accumulation of N-acetylneuraminic acid in the subfornical region of
    the hypothalamus and in the midbrain reticular formation (Merkuryva et
    al., 1980).

        A recently applied approach to the study of neurotoxicology
    involves the investigation of lipid peroxidation. Peroxidation
    involves the direct reaction of oxygen and lipids to form free radical
    intermediates and semistable peroxides. Lipid peroxidation is damaging
    because of the subsequent reactivity of these free radicals (Tappel,
    1970). Biomembranes and subcellular organelles are the major cellular
    components damaged by lipid peroxidation. Increased lipid peroxidation
    has been reported after exposure to thallium, nickel, or cobalt (Hasan
    & Ali, 1981). Since the estimate of lipid perioxidation is a
    relatively simple technique, this method may prove to be an effective
    tool for screening for neurotoxic compounds.

    6.5  Neurotransmitters

        Chemical synaptic transmission involves a complex series of events
    (synthesis and storage of neurotransmitters, release of neurotrans-
    mitters, re-uptake or degradation of transmitters, interaction of
    transmitters with postsynaptic membrane), any or all of which could be
    disturbed by neurotoxins. In addition to the classical neurotrans-
    mitters (Table 5), a number of peptides have been discovered, which
    may act as chemical messengers (Burgen et al., 1980). Most
    toxicological studies have focused on the neurotransmitters listed in
    Table 5, because considerable background information on these
    neurotransmitters is already available. Nearly every class of neuron
    contains some marker enzyme (Table 5), and these have been useful in
    studies in which attempts have been made to identify neurotransmitter
    specific cells and in the determination of the regional distribution
    of neurotransmitter systems.

        Table 5.  Enzyme markers for neurotransmitter specific cells
                                                                                   

    Neurotransmitter         Marker enzyme                 Receptor types
                                                                                   

    acetylcholine            choline acetyltransferase     mucarinic, nicotinic

    dopamine                 tyrosine hydroxylase          DA1 and DA2

    noradrenaline            dopamine--hydroxylase        alpha1, alpha2, 1,
                                                           2 adrenoreceptors

    GABA(gamma-              glutamate decarboxylase       GABAA and GABAB
      aminobutyric acid)                                   (strychnine-insensitive;
                                                           picrotoxin-sensitive)

    glycine                  unknown                       strychnine-sensitive

    serotonin                1-amino acid decarboxylase    5 HT1, 5 HT2
                                                                                   
    

    6.5.1  Synthesis/degradation

        Many neurotoxic compounds have been reported to alter steady-state
    levels of neurotransmitters, but: it is difficult to draw functional
    inferences from such data since the steady-state levels of
    neurotransmitters can be influenced via multiple mechanisms (e.g.,
    rate of synthesis, rate of release, rate of degradation). More
    informative data are obtained by examination of the turnover rates,
    reflecting the metabolic half-life of the neurotransmitter (Costa,
    1970). Turnover of 5-HT, after disulfiram treatment, was examined by
    observing the accumulation of 5-HT following administration of
    pargyline (Minegishi et al., 1979). It has been speculated that carbon
    disulfide acts by altering brain catecholamine concentrations. Carbon
    disulfide increased DA and decreased NE by inhibiting dopamine
    -hydroxylase, the enzyme responsible for converting dopamine (DA)
    into norepinephrine (NE) (McKenna & DiStefano, 1975). Manganese is
    known to inhibit DA formation by blocking tyrosine hydroxylase in the
    striatum (Bonilla, 1980) and many acute effects of organophosphorus
    compounds are related to inhibition of AChE activity (O'Brien, 1976).
    Soon after their introduction, the inhibition of AChE was accepted as
    a mechanism for the acute toxic effects of organophosphorus
    insecticides (DuBois et al., 1949). However, this mechanism does not
    account for the delayed neurotoxicity seen after organophosphate
    poisoning (Abou-Donia & Preissig, 1976; Abou-Donia, 1981).

        Interference with axonal transport may also disrupt
    neurotransmitter function by altering the availability of
    neurotransmitter enzymes or by decreasing the transport of peptide
    precursors. The transport of materials along axons is bidirectional:
    anterograde and retrograde. Anterograde transport conveys materials
    from cell body to axon and terminals at slow (1-3 mm per day) and fast
    (410 mm per day) rates (Ochs, 1972; Hoffman & Lasek, 1975). The former
    contains the bulk of axoplasmic proteins (notably of neurofilaments
    and neurotubules) and the latter, membrane and other components.
    Retrograde transport, a system that conveys materials at rapid rates
    (300 mm per day) may also be affected early in toxic axonopathies.
    Retrograde transport may be a critical factor in the toxicity
    resulting from exposure to various compounds.

    6.5.2  Transport/release

        Neurotransmitters or their precursors can rapidly concentrate in
    nerve endings through specific high-affinity uptake systems associated
    with each transmitter-specific class of neurons (Iverson, 1971). The
    high-affinity, Na+-dependent transport mechanisms are specific to
    nerve cells and can be distinguished from the ubiquitous low-affinity
    transport systems that concentrate a large variety of amino acids,
    sugars, and nucleosides, or their precursors. High-affinity uptake
    phenomena may be studied in brain homogenates, brain minces, or
    synaptosomal preparations. Lead (Silbergeld & Goldberg, 1975), the
    insecticide chlordecone (Chang-tsui & Ho, 1979), and erythrosin B
    (Lafferman & Silbergeld, 1979) have been reported to interfere with
    neurotransmitter uptake processes. However, a variety of nonspecific
    events may influence apparent uptake, so caution must be exercised in
    interpreting results of these studies. For example, the effects of
    erythrosin B on uptake mechanisms were reported to be non-specific
    (Mailman et al., 1980).

        The release of neurotransmitters from synaptic endings occurs
    through an exocytotic process that is triggered by an influx of
    calcium ions on depolarization of the nerve endings (Cotman et al.,
    1976). Neurotransmitter release can be measured by preloading nerve
    endings with labelled neurotransmitters, exposing tissue slices or
    synaptosomal fractions maintained on filter beds to calcium ions, and
    measuring the appearance of labelled or unlabelled compounds in the
    supernatant medium (Bondy & Harrington, 1979). Heavy metals such as
    lead have been reported to interfere with calcium-dependent
    neurotransmitter release (Bondy et al., 1979; Ramsay et al., 1980),
    and manganese has been reported to block transmitter release (Kirpekar
    et al., 1970; Balnave & Gage, 1973; Kostial et al., 1974), possibly by
    blocking inward Ca2+ current (Baker et al., 1971). Again, caution
    must be used in comparing the results from different assay systems.

    6.5.3  Binding

        Binding of neurotransmitters to specific membrane receptors is the
    first step in a complex series of events initiated in the
    post-synaptic cell.  Such binding interactions are reversible,
    stereospecific, nonenzymatic, have equilibria with low dissociation
    contants, and involve configurational recognition (Yamamura et al,
    1981). Studies on receptor binding have been made possible by the
    availability of specific radioactive analogues of neurotransmitters.
    In general, the most potent binding ligands are the pharmacological
    antagonists or agonists of a given neurotransmitter, rather than the
    transmitter itself. Kinetic characteristics of binding and receptor
    density are estimated by incubating receptor-enriched membranes with
    radioactive ligands. Excess ligands, not bound to membranes, can be
    removed by filtration or centrifugation. Non-specific binding of the
    labelled compound is estimated by repeating the incubation in the
    presence of a high concentration (10-4 to 10-6 mol) of an
    unlabelled analogue. The solubility of ligands either in lipids or in
    water has to be considered when interpreting results of ligand-binding
    studies. This method is relatively simple and has been suggested as a
    useful screening procedure for the detection of neurotoxic compounds
    (Damstra & Bondy, 1980, 1982). If properly employed, such analyses can
    provide valuable information. However, interpretation of the data can
    be complicated by the often low degree of specificity of the ligand
    and the location of the receptor, which may be pre- or post-synaptic
    or on non-neuronal glial elements. Furthermore, for many
    neurotransmitters, several subtypes of receptors exist and may produce
    different effects on the post-synaptic membrane. Functional
    extrapolations from neurotransmitter binding studies, therefore,
    should be made with caution.

        Receptor binding techniques have only recently been applied to
    neurotoxicological studies. However, in a short time, many compounds
    have been reported to increase or decrease estimates of receptor
    density. Acrylamide (Bondy et al., 1981), chlordecone (Seth et al.,
    1981), lead and mercury (Bondy & Agrawal, 1980), and cadmium (Hedlund
    et al., 1979) have all been reported to alter putative
    neurotransmitter receptors.

    6.5.4  Ion channels

        Many neuronal functions are regulated by ion channels. The Na+
    channel is responsible for depolarizing the membrane, and K+
    channels are responsible for repolarizing the membrane. Maintenance of
    appropriate Na+ and K+ concentrations requires the classical
    Na+/K+ ATPase. Disturbed functioning of these ion channels or
    disturbance of ATP availability can severely disturb neuronal
    functioning. Several naturally-occurring toxins affect Na+ channels
    (Hille, 1976; Ritchie, 1979; Catterall, 1977). Tetrodotoxin and
    saxitoxin inactivate the channel at nanomolar concentrations and can

    be used as probes for measuring channel density or isolation of
    channel polypeptides (Agnew et al., 1978). Batrachotoxin and
    veratridine bind the channels and produce persistent activation by
    preventing channel inactivation (Albuquerque & Daly, 1976). Peptide
    toxins, such as scorpion toxin (Couraud et al., 1978), exhibit either
    voltage sensitive or insensitive binding to the extracellular surface
    of the Na+ channel (Catterall, 1977). Several potassium channels
    probably exist in neurons (Reichardt & Kelly, 1983), but purification
    and characterization of these channels has lagged behind that of the
    Na+ channel. However, a scorpion toxin that exibited K+ channel
    affinity has been described (Carbone et al., 1982).

        Na+/K+-ATPase consists of 2 polypeptide chains. The smaller
    polypeptide is a glycoprotein (Carilli et al., 1982). The larger
    polypeptide binds ATP internally and ouabain externally. Hormones and
    neurotransmitters such as catecholamines regulate the ATPase (Clausen
    & Flatman, 1977), and neurotoxicants could alter neuronal functioning
    by disturbing this control of ionic gradients. Desaiah et al. (1980)
    have reported inhibition of Na+/K+-ATPase and decreased ouabain
    binding to synaptosomes after treatment of mice with the insecticide
    chlordecone. The evaluation of ion balance and their control will
    probably be used increasingly in neurotoxicology.

        For the most part, transmitter release is regulated by Ca2+
    entry through a Ca2+-selective channels. Depolarization,
    neurotransmitters, and hormones regulate Ca2+-selective channel
    (Reichardt & Kelly, 1983). Nerve terminals contain several Ca2+
    calmodulin-dependent protein kinases and a Ca2+-phospholipid
    activated kinase, each of which has a distinct set of protein
    substrates. Cytoplasmic Ca2+ binds calmodulin and other Ca2+
    binding proteins, which directly or indirectly activate other enzymes
    via their Ca2+-dependent kinases. How such activation facilitates
    exocytosis is not known, but presumably the phosphorylation state of
    synapsin I, a phosphoprotein present primarily in nerve terminals and
    associated with synaptic vesicles, is believed to regulate transmitter
    vesicle release (Dolphin & Greengard, 1981).

        Changes in Ca2+ concentration induced by neurotoxicants could
    have significant consequences for neuronal function.  Cytoplasmic
    Ca2+ stimulates not only exocytosis, but also glucogenolysis and
    mitochondrial respiration (Landowne & Ritchie, 1976), endocytosis
    (Ceccarelli & Hurlbut, 1980), and neurotransmitter synthesis (Collier
    & Ilson, 1977). Even though many of these events are homeostatic
    responses to transmitter utilization, neurotoxic disruption of calcium
    influx or sequestering could produce disturbances not restricted to
    neurotransmitter release.

        Few neurotoxic compounds have been investigated for their
    influence on calcium channels. Manganese is reported to block
    neurotransmitter release by blocking inward Ca2+ current (Kirpekar

    et al., 1970), and other heavy metals have been reported to interfere
    with calcium-mediated neurotransmitter release (Bondy et al., 1979;
    Ramsay et al., 1980). Similar effects of heavy metals were also
    observed for the Na+-Ca2+ exchange system. Ion channels and
    Na+-Ca2+ probes may be a valuable screening approach for some
    types of neurotoxic compounds.

    6.5.5  Cyclic nucleotides

        Nerve terminal function and the effects of neurotransmitters are
    often regulated by cyclic nucleotides. Binding of agonists to
    receptors on the nerve membrane can result in activation of second
    messenger systems. Activation of different receptors results in
    specific changes in cyclic AMP and cyclic GMP levels and consequent
    alterations in protein kinases. It is thought that on phosphorylation,
    conformational changes occur in membrane proteins that may change ion
    permeabilities. Thus, the responsiveness of the nervous system to some
    toxic agents can also be measured by determining changes in the
    activities of adenylate cyclase, cyclic nucleotides, and protein
    phosphorylation. For example, lead has been shown to inhibit
    cerebellar adenylate cyclase (Nathanson & Bloom, 1975) and 
    dopamine-sensitive adenylate cyclase in the striatum (Wilson, 1982).
    Various pesticides cause increases in the levels of cyclic nucleotides
    in brain tissue (Aldridge et al., 1978). For example, tri-o-cresyl
    phosphate enhanced Ca2+-calmodulin-dependent in vivo
    phosphorylation of proteins in chicken brain (Patton et al., 1983).

    6.5.6  Summary of nerve terminal function

        The analysis of synaptic function and neurotransmitter metabolism
    is reasonably complex, and it should be borne in mind that disturbance
    at many levels of neuronal organization will ultimately alter synaptic
    function. However, examination of the nerve terminal is an excellent
    approach for the initial study of toxic chemicals. However, it is
    important to recognize the dynamic nature of the nerve terminal and to
    remain alert to the possibility of false positives.

    6.6  Energy Metabolism

        Nervous tissue, and brain tissue in particular, require
    disproportionately large amounts of energy to sustain the
    translocation of ions important for electrical activity and to
    maintain the highly-active biosynthetic machinery of the tissue.
    Since neural tissue has only limited stores of energy (e.g., glycogen
    and creatine phosphate), glucose availability and enzymes critical for
    energy production are vital to neuronal functioning. A number of
    neurotoxic chemicals have been shown to interfere with glucose and
    energy metabolism in both the central and peripheral nervous system.
    These include methylmercury (Bull & Lutkenhoff, 1975), hexachlorophene
    (Cammer & Moore, 1972), organotin compounds (Lock, 1976), and alcohol
    (Merkuryva et al., 1980). Glycolytic enzymes have been shown to be


    inhibited by alcohol (Merkuryva et al., 1980) and by chemicals known
    to cause distal axonopathy, e.g., acrylamide (Howland et al., 1980)
    carbon disulfide, and methyl-n-butyl ketone (Sabri et al., 1979).

        Active brain areas have a higher rate of glucose consumption than
    less active regions. The brain regional metabolic activity is
    determined by the 14C-2-deoxyglucose method (2-DG) of Sokoloff et
    al. (1977). This method is based on the use of 2-deoxy-D-14C-glucose
    as a tracer for the exchange of glucose between the plasma and the
    blood and its phosphorylation by hexokinase. The product,
    2-deoxyglucose-6-phosphate (2-DG-P), is trapped in the tissue and its
    accumulation in the various regions of the brain might be a measure of
    the glucose use and neuronal activity. The method provides an index of
    the in vivo glucose metabolic activity of brain regions and has been
    used in experimental animals to determine which brain regions are
    depressed or activated during acute Soman intoxication (McDonough et
    al., 1983).

        Metabolic compartmentalization in nerve tissue could provide an
    additional tool in toxicological studies. It is well established that
    neurons mainly use glucose in their energy metabolism, while glia use
    substrates other than glucose (Hertz, 1981). Thus, it may be possible
    to differentiate between the toxicological effects of chemical
    substances on neuronal and glial populations in vitro or in vivo.

    6.7  Biochemical Correlates of Axonal Degeneration

        A relatively large class of neurotoxic agents is known to produce
    axonal degeneration (Griffin & Price, 1980; Sabri & Spencer, 1980;
    Thomas, 1980). Some, such as acrylamide and organophosphorus
    compounds, produce the distal to proximal "dying back" pattern of
    degeneration, while others, such as ,-iminodipropionitrile, lead to
    proximal to distal degeneration. Since many of these compounds produce
    Wallerian degeneration, biochemical correlates of Wallerian
    degeneration have been suggested as toxicological indices (Dewar &
    Moffett, 1979). These authors tabulated the chemical changes in
    peripheral nerves that occurred during degeneration and suggested that
    -gluocorinadase and -galactosidase might be good indices of
    toxicity. Acrylamide, methylmercury, and dimethyl sulfoxide were shown
    to increase these lysosomal enzymes (Dewar & Moffett, 1979). The use
    of lysosomal enzymes involves relatively simple procedures and might,
    therefore, be appropriate for the initiation of a biochemical screen.

        Disturbance of axoplasmic transport has been suggested as an
    alternative approach to detecting compounds producing axonal
    degeneration. Several substances, e.g., organophosphates (Reichart &
    About-Donia, 1980) and acrylamide (Pleasure et al., 1969) disrupt
    axonal transport. However, these methods are more difficult, more
    expensive, and more time-consuming than the lysosomal enzyme assays.

    6.8  Neuroendocrine Assessments

        The number of toxic compounds being recognized for their
    neuroendocrine actions is increasing.  Hormonal balance results from
    the integrated action of the hypothalamus, pituitary, and endocrine
    target organ. Each site is susceptible to disruption by environmental
    toxic agents. This disruption may result from the direct interaction
    of the toxic agent with the endocrine organ, pituitary, or
    hypothalamus. Alternatively, neuroendocrine dysfunction may occur
    because of a disturbance in the regulation and/or modulatory elements
    of the complex neuroendocrine feedback systems. Any such disturbances
    would ultimately modify anterior pituitary secretions. Thus, the
    analysis of blood levels is the most appropriate starting point
    (section 6.8.2.1).

    6.8.1  Anterior pituitary hormones

        The main effector organ in the neuroendocrine system is the
    pituitary. This "master gland" consists of an anterior 
    adenohypophysis, and a posterior neurohypophysis. Based on
    histological, immunocytochemical, and electron microscopic examination,
    the following cell types are differentiated in the anterior pituitary
    (Junqueira et al., 1977); follicular cells, which do not contain any
    secretory granules but form the support stroma of the glandular cells;
    chromophobe cells, which are undifferentiated, nonsecretory and
    secretory cells without discernible granules in the light microscope;
    somatotropic cells, which are acidophilic with immunocytochemically-
    detectable growth hormone and contain granules of about 350 nm
    diameter; prolactin cells, which are acidophilic and contain prolactin
    granules of 600-900 nm diameter; gonadotropic cells, which consist of
    2 basophilic subgroups: follicle-stimulating hormone secreting and
    luteinizing-hormone secreting; follicle-stimulating hormone secreting
    cells, which are large round cells with dense 200 nm diameter
    granules; luteinizing hormone-containing cells, which are small and
    contain 200-250 nm diameter granules; thyrotropic cells, which are
    basophilic with thyroid-stimulating hormone granules of 120-200 nm
    diameter; and finally, cortico cells, which are the least abundant of
    the cell types and contain basophilic granules 100-200 nm in diameter.

        The glandular cells of the anterior pituitary secrete seven
    endocrine hormones, of which four act directly on other endocrine
    glands. Follicle-stimulating hormone (FSH) promotes spermatogenesis in
    the male testes and facilitates follicular maturation and estrogen
    secretion in the female ovary. Lutenizing hormone (LH) acts on the
    male testes to facilitate testosterone secretion from the interstitial
    cells of Leydig and is a key hormone that promotes follicular rupture
    (ovulation) in the female ovary. Thyroid-stimulating hormone (TSH)
    triggers the secretion of thyroxine from the thyroid gland, and
    adrenocorticotropic hormone (ACTH) causes the adrenal cortex to
    secrete its products, especially glucocorticoids. The secretion of

    these tropic, anterior pituitary hormones is in turn regulated by the
    hormones of the endocrine gland. In most cases, this is a negative
    feedback regulation. Estrogen decreases the output of pituitary FSH
    and LH; thyroxine decreases the secretion of TSH; and glucocorticoids
    reduce the output of ACTH.

        The three other anterior pituitary hormones are prolactin (PRL),
    melanocyte-stimulating hormone (MSH), and growth hormone (GH). A major
    target for prolactin is the mammary gland, where prolactin promotes
    milk production. PRL also plays a role in maintaining the ovarian
    corpus luteum after ovulation. The entire body is a target for GH,
    which promotes growth and maintenance of cellular integrity. GH
    facilitates growth, in part, because it increases the transport of
    amino acids into the cell, where they can be used for protein
    synthesis. MSH increases the production of melanin by the melanocytes
    of the skin.

        Hypothalamic control of anterior pituitary secretions occurs
    through the release of hypothalamic-hypophysiotropic hormones (HHH).
    The hypothalmus secretes these HHH into vessels of the hypothalamic-
    hypophyseal portal system, by which they are transported to the
    anterior pituitary. One of the major advances of the last two decades
    has been the isolation and characterization of these HHH.

    6.8.2  Disruption of neuroendocrine function

        Disruption of neuroendocrine function can occur at any one or all
    of the levels of hypothalamic-pituitary-target organ integration.
    Alternatively, neuroendocrine effects may result from modification of
    "higher" levels of neuronal processing or from alterations in
    peripheral metabolism. In the following discussion, approaches will be
    indicated for the study of each of these possible sites of neurotoxic
    action.

    6.8.2.1  Direct pituitary effects

        At the most dramatic level, massive changes in the weight or size
    of the pituitary, such as occur in the adenohypophysis after large
    doses of estrogen (Schreiber & Pribyl, 1980) might be seen. However,
    by far the most accessible method for studying pituitary function
    relies on the examination of blood levels of pituitary hormones by
    radioimmunoassay. Furthermore, the analysis of blood-hormone levels is
    one of the few techniques applicable for screening human populations.
    A large number of toxic substances have been reported to modify
    adenohypophyseal secretions. Methallibure and allied substances
    inhibit thyrotropin (Tulloch et al., 1963), and prolactin (Benson &
    Zagni, 1965) secretion. Modifications of pituitary hormone secretions
    have been reported for acrylamide (Uphouse et al., 1982), alcohol
    (Mendelson et al., 1977), 1,4-DDT (Gellert et al., 1972), and
    chlordecone (Uphouse et al., 1984). Dimethylsulfoxide may influence

    adrenal glucocorticoids by elevating pituitary secretions (Allen &
    Allen, 1975). In spite of the fact that changes in the blood levels of
    pituitary hormones may be the most sensitive index of neuroendocrine
    modification by toxic compounds, they have been examined for
    relatively few compounds. The high sensitivity of this approach
    results from the fact that the blood levels represent the final
    consequence of the complex neuroendocrine integration. Regardless of
    the actual site of modification, neuroendocrine disturbances will
    usually be revealed in modified levels of circulating hormones.
    Therefore, such changes, alone, cannot be regarded as evidence of
    direct neuroendocrine disruption.  The major disadvantage of serum-
    hormone measurements is the responsiveness of the hypothalamic-
    pituitary axis to a variety of environmental and chemical stimuli. It
    can be difficult to identify the toxic agent as the causal agent.
    Because hormones are secreted in a pulsatile manner, single point
    analyses of serum hormones must be cautiously interpreted. However,
    even with these limitations, analysis of blood levels of pituitary
    hormones remains the most appropriate starting point for evaluating
    potential neuroendocrine toxicity.

        Similar analytical methods may be used to investigate the
    pituitary content of respective pituitary secretions. The pituitary is
    removed, homogenized, and extracted for use with the appropriate RIA.
    For most pituitary secretions, neurotoxicologists have not yet studied
    pituitary tissue directly. However, pituitary endorphins has been
    reported to be modified by chlordecone (Hong & All, 1982).

    6.8.2.2  Peripheral target effects

        Most anterior pituitary hormones are subject to negative feedback
    control by peripheral endocrine glands. If the neurotoxic agent
    modifies peripheral secretions, neuroendocrine changes can result from
    this altered feedback. Modifications in the functioning of these
    endocrine secretions could occur after toxic exposure. Such approaches
    have been widely applied in the experimental and clinical literature
    and have shown that a number of compounds alter blood levels of
    glucocorticoids, thyroxine, estrogen, and testosterones (Chapman,
    1983).

        Target-tissue effects can also be evaluated by a variety of
    additional techniques. One of the simplest assessments is the change
    in the weight, morphology, or biochemistry of the target organ.
    Various environmental contaminants have been reported to produce such
    changes. Lesions of the thyroid have been reported after exposure to
    carbon disulfide (Cavalleri, 1975) and strong magnetic fields
    (Persinger et al., 1978). Marked changes occur in the adrenals during
    antimony and lead poisoning (Minkins et al., 1973) and in response to
    very long chain saturated fatty acids (Powers et al., 1980).
    Chlordecone causes adrenal, cortical, and medullary hypertrophy
    (Eroschenko & Wilson, 1975) and alters the epinephrine and

    norephinephrine content of the adrenal medulla (Baggett et al., 1980).
    DDT (McBlain et al., 1976) and other chlorinated hydrocarbons
    (Dicksith & Datta, 1972; Fellegiova et al., 1977), carbon disulfide
    (Rosewickyi et al., 1973), and cadmium salts (Parizek, 1956;
    Zylber-Haran et al., 1982) have all been described as having toxic
    effects on the gonads. Although such changes are not necessarily due
    to direct neuroendocrine effects, target organ changes can often be a
    first indication of neuroendocrine changes. In some cases, even
    relatively gross target organ events have been a first clue towards
    the mode of action of a neurotoxic agent. For example, the chlorinated
    pesticides, DDT and chlordecone, modify gonads in a manner reminiscent
    of the natural steroid, estrogen (Gellert et al., 1972; Eroschenko &
    Palmiter, 1980; Kupfer & Bulger, 1980). Uterine and/or oviduct weight
    changes in the immature animal were the first evidence that the
    pesticides might exert estrogenic action.

    6.8.2.3  Disruption of hypothalamic control of pituitary secretions

        Hypothalamic control of anterior pituitary secretions includes
    direct regulation through hypothalamic hypophysiotropic hormones and
    modulation through neurotransmitters and neurally-active peptides.
    Evaluation of the effect of a neurotoxic agent on HHH is directly
    measureable only for the HHH for which antisera are available. For
    these compounds, it is possible (though difficult) to measure release
    into the portal blood supply, to evaluate the effects of toxic agents
    on either amounts or release, and to identify pituitary responsiveness
    to the hypothalamic factors. Where identified hormones are not
    available, hypothalamic extracts may still be used with in vitro
    pituitary responsiveness as a bioassay. Details of such methods can be
    found in several sources (Burgus & Guillemin, 1970; Oliver et al.,
    1974). However, the value of such approaches for neurotoxicology has
    yet to, be tested. Such methods will be of greatest value in testing
    hypotheses regarding the mechanism of action of known neuroendocrine
    toxic agents. They are not recommended for initial screens.

        Biochemical changes in the hypothalamus can also be used as
    indices of potential neuroendocrine disruption. Such hypothalamic
    effects of neurotoxic compounds have received much less attention than
    brain areas such as the striatum, cortex, or cerebellum. Methods for
    studying the hypothalamus are the same as those used for other brain
    areas. However, the hypothalamic neurons that regulate pituitary
    function receive numerous synaptic inputs, both conventional and from
    putative neurotransmitters. Consequently, the neuroendocrine
    significance of changes in hypothalamic neurotransmitters and
    neuropeptides is usually only inferential. However, any disruption of
    hypothalamic biochemistry has the potential to alter neuroendocrine
    function. When combined with more direct measurements of
    neuroendocrine function, such studies are very important. There have
    been a few reports of biochemical changes in the hypothalamus and/or
    pituitary, correlated with neuroendocrine toxicity. One approach has

    been to examine the effects of toxic compounds on hormone-mediated
    changes. Administration of estrogens is followed by hypothalamic
    ascorbic acid depletion (Schreiber et al., 1982) and by increases in
    polyphenol oxidase (ceruloplasmin) activity in the hypothalamus and
    the blood (Schreiber & Pribyl, 1980). This can be inhibited by the
    simultaneous administration of silver nitrate. Disulfiram
    (tetraethylthiuram disulfide) inhibits the reaction of the
    adenohypophysis (Schreiber et al., 1979). Disulfiram also inhibits
    dopamine--hydroxylase (Szmigielski, 1975) and lowers the
    blood-prolactin level (Cavalleri et al., 1978). The effect of
    disulfirams on prolactin may be due to a "sparing of dopamine" through
    inhibition of dopamine--hydroxylase.

        Hypothalamic peptides have been studied most extensively after
    chlordecone treatment. Long-term exposure to the pesticide reduced
    hypothalamic -endorphin levels under conditions where substance P,
    neurotensin, and met-enkephalin were unchanged (Ali et al., 1982).

    6.8.2.4  Other sites of action

        For many neurotoxic agents, there may be no, identifiable effects
    on neuroendocrine function, but it may be altered through the indirect
    effects of the toxic compound. Such interruption could occur via
    modification of neural integration leading to an altered response of
    the organism to environmental challenge. Detection of such integrative
    action and its relevance to neuroendocrine function might include
    treatment with the toxic agent followed by environmental,
    pharmacological, or hormonal challenges known to produce a
    neuroendocrine response. Several toxic agents have been reported to
    produce behavioural changes, such as stress-induced analgesia, and
    these undoubtedly involve some aspect of neuroendocrine function. CNS
    opiate systems are important components of the CNS response to stress
    and studies of peripheral and brain endorphins and encephalins are
    potentially relevant to such disturbances. For example, neonatal
    treatment with chlordecone dissolved in dimethyl sulfoxide produces an
    elevated corticosterone response to footshock (Rosecrans et al.,
    1982), and long-term exposure of female rats to chlordecone has been
    reported to decrease hypothalamic -endorphin levels (Ali et al.,
    1982).

        A toxic agent could also indirectly affect neuroendocrine function
    by altering the peripheral metabolism of endocrine secretions. Such
    metabolic differences could indirectly influence CNS function.
    However, hormone transformation and the formation of active
    metabolites occurs in the CNS so that metabolic disturbances may even
    have a direct effect on CNS functioning. Many metabolic variables have
    been reported to change after treatment with a toxic agent.
    Aromatization of testosterone to estrogen, an important step in the
    metabolism of estrogen, took place in the brain of animals (Gallard et
    al., 1978). Its inhibition by aminoglutethimide or other blockers of

    aromatization (Morali et al., 1977) markedly altered the sexual
    behaviour of experimental animals. Aminoglutethimide also inhibited
    total adrenal steroidogenesis as well as gonadal steroidogenesis and
    thyroid function (Schreiber et al., 1969). The transformation of
    thyroxine to the more active metabolite triiodothyronine was inhibited
    by a series of toxic substances such as ethanol (Shimizu et al.,
    1978), propylthiouracil (Leonard & Rosenberg, 1980), iopanoic acid
    (Kaplan, 1980), and sodium salicylate (Chopra et al., 1980). Since the
    monodeiodination of thryoxine to triiodothyronine also takes place in
    the adenohypophysis, these factors may be an important component of
    neuroendocrine reactions to neurotoxic substances. There is evidence
    that demonstrates that the direct dopaminergic effect of estrogen on
    the pituitary may require conversion of estrogen to catechol estrogen
    (Paul et al., 1980). Such conversion occurs not only in the pituitary,
    but also in the hypothalamus and cerebral cortex and may mediate some
    of the effects of estrogen on the CNS. Study of this enzyme promises
    to be important for future research, especially for the investigation
    of estrogen-like toxic agents.

        A final way in which the toxic agent might disrupt neuroendocrine
    function is by altering the peripheral metabolism of the endocrine
    secretions within liver tissue. Although beyond the scope of this
    publication, several neurotoxic agents (chlordecone, dieldrin,
    heptachlor, lindane, p,p'-DDE, and toxaphene) stimulate the
    metabolism of estrone by liver microsomal enzymes (Welch et al.,
    1971). Biochemical approaches for identifying these metabolic
    disturbances are the same as those discussed in other sections.
    However, the precise metabolic end-point should be carefully chosen,
    when inferences are to be made about neuroendocrine function.

    6.8.3  Sex differences

        Because neurotoxic agents may produce changes in neuroendocrine
    function and, since males and females differ with regard to a variety
    of metabolic variables, a valuable approach to any neurotoxicological
    investigation is the comparison of the toxic response in the sexes.
    Whenever neuroendocrine effects are suspected, sex differences should
    be a routine aspect of the overall investigation. Sex-related
    differences have been observed for many environmental compounds.
    Often, these result from the influence of gonadal hormones on the
    metabolism of the compound. Hormonal influences on the 
    biotransformation and toxicity of DDT (Durham et al., 1956) and several
    organophosphates (Murphy & DuBois, 1958; DuBois & Puchala, 1961) have
    been demonstrated. Parathion is an organophosphate insecticide that
    exerts its toxicity through its active metabolite paraoxon phosphate.
    Agrawal et al. (1982) have recently shown that the sex difference in
    AChE inhibition after parathion treatment was not evident with
    paraoxon treatment. This suggests that the sex difference was in the
    rate of metabolism to the toxic product. The female's increased
    sensitivity to amobarbital (Castro & Gillette, 1967) may also be due
    to sex differences in hepatic metabolism. Sex differences have also

    been reported for the rate of disposition of chlordiazepoxide
    (Greenblatt et al., 1977; Roberts et al., 1979) and for the toxic
    effects of ethylmorphine, aniline, p-nitroanisole (Nicholas &
    Barron, 1932; Holck et al., 1937; Quinn et al., 1958) and
    polychlorinated hydrocarbons (Lamartiniere et al., 1979).

    6.9  Recommendations for Future Research

        Several biochemical and neuroendocrine approaches are available
    that have not yet been applied to the study of neurotoxic compounds.
    These have the advantage of increasing the sensitivity of the
    biochemical approach and furthering understanding of the ways in which
    compounds disrupt nervous system function. For example, some enzymes
    involved in neurotransmitter synthesis have been purified and used for
    the production of specific antibodies (John et al., 1973). It is now
    possible to use immunological titration to determine whether changes
    in enzyme activity result from activity alterations or variations in
    the amount of enzyme protein.

        Identification of direct pituitary modification can be
    accomplished by preparation of anterior pituitary cell cultures and
    the measurement of hormone release in response to hypothalamic
    releasing factors, neurotransmitters, or suspect toxic compounds. The
    pituitary cells are removed, dispersed, and agitated with collagenase.
    After resuspension in medium, the cells are plated and used for
    examination. Under such culture conditions, pituitary cells respond to
    releasing factors and neurotransmitter regulation (Enjalbert et al.,
    1978; Drouin & Labrie, 1981). Thus, it is possible to determine
    whether the toxic agent modifies the responsiveness of the pituitary
    to hypothalamic control. By comparing the effects of the toxic agent
    in vivo with those in vitro, direct versus indirect effects of the
    compound can be identified.

        Furthermore, for suspect steroid-like compounds, direct evaluation
    of receptor interactions is possible. Steroid receptors modify their
    target tissue by binding to intracellular receptors. Measurement of
    intracellular steroid receptors involves the preparation of a high
    speed cytosol (180 000 g) fraction from the appropriate target tissue.
    Identification of binding is accomplished by incubating the cytosol
    in vitro in the presence of radioactively-labelled hormone. Bound
    label is removed from unbound by a variety of techniques distinct for
    the particular receptor of interest. Specific binding is assessed by
    incubating the labelled hormone in the presence of excess unlabelled
    hormone as competitor. Specific binding refers to bound molecules that
    are removed in the presence of this unlabelled competitor. When a
    neurotoxic agent is suspected of interacting with hormone receptors,
    the toxic agent may be substituted for the unlabelled competitor and
    its ability to compete for binding determined. This procedure has
    demonstrated competition by both chlordecone (Palmiter & Mulvihill,
    1978) and o,p'-DDT (Kupfer & Bulger, 1976) for the estradiol
    cytosol receptor in the uterus. Using an estradiol exchange method,

    investigators have shown that these pesticides produce translocation
    of the estradiol receptor to the nucleus (Hammond et al., 1979).
    Employment of the exchange method necessitates the in vivo treatment
    of the organism with the toxic agent. Nuclei and cytosol are prepared
    from the target tissue and incubated in vitro with the
    radioactively-labelled hormone. Since a major movement of cytosol
    receptors into the nucleus occurs only after binding to the hormone,
    the presence of the receptor in the nucleus after exposure to the
    toxic agent suggests direct binding of the toxic agent to the steroid
    cytosol receptor.

        Modifications of the procedures described for neurotransmitter
    receptor binding are used in the measurement of membrane receptors.
    However, no neurotoxic agents have yet been tested for their ability
    to modify these hormone receptors. For both intracellular and
    extracellular hormone receptors, most studies have been on peripheral
    tissues. However, the brain and the pituitary contain receptors for a
    variety of steroid and peptide hormones, and these offer a number of
    targets for disruption by environmental compounds. Such analyses have
    great potential for future studies of neurotoxic compounds. However,
    they will usually be applied for the investigation of mechanisms of
    action rather than as screens for neurotoxic compounds.

        Finally, hypothalamic regulation of pituitary function should
    receive further emphasis. Neurotransmitters also regulate pituitary
    secretions via neurotransmitter receptors of the pituitary gland. The
    potential for a direct modification of these receptor interactions by
    neurotoxic compounds is only just being realized. The most thoroughly
    described pituitary neurotransmitter receptor is that of dopamine,
    which regulates the inhibition of prolactin release by dopamine.
    Steroid hormones, such as estrogen, by conversion to catechol
    estrogens, increase prolactin secretion by direct interaction with the
    pituitary dopamine receptors (Gudelsky et al., 1981; Fishman, 1982).
    Neuroleptic drugs elevate prolactin secretion by antagonistic action
    on pituitary dopamine receptors (Horowski & Graf, 1979; Besser et al.,
    1980).

    7.  CONCLUSIONS AND RECOMMENDATIONS

        There is ample evidence of real and potential hazards of
    environmental chemicals for nervous system function. Changes or
    disturbances in central nervous function, many times manifest by vague
    complaints and alterations in behaviour, reflect on the quality of
    life; however, they have not yet received attention.
    Neurotoxicological assessment is therefore an important area for
    toxicological research.

        It has become evident, particularly in the last decade, that
    low-level exposure to certain toxic agents can produce deleterious
    neural effects that may be discovered only when appropriate procedures
    are used. While there are still episodes of large-scale poisoning,
    concern has shifted to the more subtle deficits that reduce
    functioning of the nervous system in less obvious, but still important
    ways, so that intelligence, memory, emotion, and other complex neural
    functions are affected.

        Information on neurobehaviour, neurochemistry, neurophysiology,
    neuroendocrinology, and neuropathology is vital for understanding the
    mechanisms of neurotoxicity. One of the major objectives of a
    multifaceted approach to toxicological studies is to understand
    effects across all levels of neural organization. Such a multifaceted
    approach is necessary for confirmation that the nervous system is the
    target organ for the effect. Interdisciplinary studies are also
    necessary to understand the significance of any behavioural changes
    observed and thus, to aid in extrapolation to human beings by
    providing specific neurotoxic profiles. Concomitant measurements at
    different levels of neural organization can improve the validity of
    results.

        The following recommendations are made on the basis of the
    information contained in this book:

    1.  Health personnel should take into account the possible role of
        exposure to chemicals whenever a patient presents with any
        neurobehavioural complaint.

    2.  Neurotoxicological testing should be an essential consideration in
        any profile developed by the agencies responsible for the control
        of toxic chemicals.

    3.  In the determination of the potential of a chemical to produce
        neurotoxic effects, a multidisciplinary strategy should be used.

    4.  Priority should be given to obtaining clinical and epidemiological
        data when exposure occurs to chemicals suspected of being
        neurotoxic.

    5.  Test development in preclinical neurotoxicology has evolved to the
        point where interlaboratory validation of procedures using
        prototypic neurotoxic agents could be attempted.

    REFERENCES

    ABOU-DONIA, M.B. (1981) Organophosphorus ester-induced delayed
    neurotoxicity. Ann. Rev. Pharmacol. Toxicol., 21 511-548.

    ABOU-DONIA, M.B. & PREISSIG, S.H. (1976) Delayed neurotoxicity of
    leptophos: toxic effects on the nervous system of hens. Toxicol.
    appl. Pharmacol., 35 269-282.

    ABOU-DONIA, M.G., JENSEN, D.N., & LAPADULA, D.M. (1983) Neurologic
    manifestations of tri-o-cresyl phosphate delayed neurotoxicity in
    cats. Neurobehav. Toxicol. Teratol., 5: 431-442.

    AGNEW, W.S., LEVINSON, S.R., BRABSON, J.S., & RAFTERY, M.A. (1978)
    Purification of the tetrodotoxin-binding component associated with the
    voltage-sensitive sodium channel from Electrophorus electricus
    electroplax membranes. Proc. Natl Acad. Sci. (USA), 75: 2606-2610.

    AGRAWAL, A.K., SQUIBB, R.E., & BONDY, S,C. (1981) The effects of
    acrylamide treatment upon the dopamine receptor. Toxicol. appl.
    Pharmacol., 58 89-99.

    AGRAWAL, D.K., MISRA, D., AGARWAL, S,, SETH, P.K., & KOHLI, J.D.
    (1982) Influence of sex hormones on parathion toxicity in rats;
    antiacetylcholinesterase activity of parathion and paraoxon in plasma,
    erythrocytes, and brain. J. Toxicol. environ. Health, 9: 451-459.

    ALBURQUERQUE, E.X. & DALY, J.W. (1976) Steroidal alkaloid toxins and
    ion transport in electrogenic membranes. In: Cuatracasas, P., ed. The
    specificity of animal, bacterial, and plant toxins, London, Chapman
    and Hall, pp. 297-338.

    ALDER, S. & ZBINDEN, G. (1977) Methods for the evaluation of physical,
    neuromuscular, and behavioral development in rats in early postnatal
    life. In: Newbert, D., Merber, H.J., & Kwasigroch, T.E., ed. Methods
    in prenatal toxicology, Stuttgart, George Thiene, pp. 175-185.

    ALDRIDGE, W.M., CLOTHIER, B., FORSHAW, P., JOHNSON, M.K., PARKER,
    V.H., PRICE, R.J., SKILLETER, D.N., VERSCHOYLE, P.D., & STEVENS, C.
    (1978) The effect of DDT and pyrethroids cismethrin and decamethrin on
    the acetylcholine and cyclic nucliotide content of rat brain.
    Biochem. Pharmacol., 27: 1703-1706.

    ALEU, F.P., KATZMAN, R., & TERRY, R.D. (1963) Fine structures and
    electrolyte analyses of cerebral oedema induced by alkyl tin
    intoxication. J. Neuropathol. exp. Neurol., 22: 403-413.

    ALEXANDER, B.K.  BEYERSTEIN, B.L., HADAWAY, P.F., & COAMBS, R.B.
    (1981) Effect of early and later colony housing on oral ingestion of
    morphine in rats. Pharmacol. Biochem. Behav., 15: 571-576.

    ALFANO, D.P. & PETIT, T.L. (1981) Behavioral effects of postnatal lead
    exposure. Possible relationship to hippocampal dysfunction. Behav.
    neural Biol., 32: 319-333.

    ALI, S.F., HONG, J.S., WILSON, W.E., LAMB, J.C., MOORE, J.A., MASON,
    G.A., & BONDY, S.C. (1982) Subchronic dietary exposure of rats to
    chlordecone (kepone) modifies levels of hypothalamic beta-endorphin.
    Neurotoxicology, 3: 119-124.

    ALLEN, J.P. & ALLEN, C.F. (1975) The effect of dimethyl sulfoxide on
    hypothalamic-pituitary-adrenal functions in the rat. Ann. NY Acad.
    Sci., 243: 325-336.

    ALTMAN, J. & SUDARSHAN, K. (1975) Postnatal development of locomotion
    in the laboratory rat. Anim. Behav., 23: 896-920.

    AMIN-ZAKI, L., MAJEED, M.A., ELHASSANI, S.B., CLARKSON, T.W.,
    GREENWOOD, M.R.  & DOHERTY, R.A. (1979) Prenatal methylmercury
    poisoning: clinical observations over five years. J. Dis. Child,
    133: 172-177.

    ANDRIANOV, V.V., ZAKHAROV, N.C., & FADEEV, JU.A. (1977) Microiono-
    phoretic analysis of chemical properties of cortical units in freely
    moving animals. Physiol. J. (USSR), 63(4): 602-605.

    ANGELL, F. & WEISS, B. (1982) Operant behavior of rats exposed to lead
    before or after weaning. Toxicol. appl. Pharmacol., 63: 62-71.

    ANGER, W.K. (1984) Neurobehavioral testing of chemicals: impact on
    recommended standards. Neurobehav. Toxicol. Teratol., 6(2): 147-153.

    ANGER, W.K. & JOHNSON, B.L. (1985) Chemicals affecting behavior. In:
    O'Donoghue, J.L., ed. Neurotoxicity of industrial and commercial
    chemicals, Boca Raton, Florida, CRC Press.

    ANGER, W.K. & SETYES, J.V. (1979) Effects of oral and intramuscular
    carbaryl administration on repeated chain acquisition in monkeys.
    J. Toxicol. environ. Health, 5: 793-808.

    ANNAU, W. (1975) The comparative effects of carbon monoxide and
    hypoxia on behaviour. In: Weiss, B. & Laties, V.G., ed. Behavioural
    toxicology, New York, Plenum Press, pp. 105-127.

    ANOKHIN, P.K. (1935) [Problem of center and periphery in contemporary
    physiology of nervous system.] In: Anokhin, P.K., ed. [Problem of
    centre and periphery in physiology of nervous activity], Gorkiy,
    Gorkiy Publishing House, pp. 9-71 (in Russian).

    ANOKHIN, P.K. (1964) [Comparative electrophysical analysis of
    alterations of laility of the cortical potential components in
    postnatal ontogenesis.] Fiziol. Zh. SSSR, 50: 773-778 (in Russian).

    ANOKHIN, P.K. (1968) [Biology and neurophysiology of the conditioned
    reflex,] Moscow, Meditsina, pp. 547 (in Russian).

    ANOKHIN, P.K. (1974) The functional system as a basis of the
    physiological architecture of the behavioural act. In: Anokhin, P.K.,
    ed. Biology and neurophysiology of the conditioned reflex and its
    role in adaptive behaviour, New York, Pergamon Press, pp. 190-254.

    APFELBACH, R. & DELGADO, J.M.R. (1974) Social hierarchy in monkeys
    (Macaca mulatta) modified by chlordiazepoxide hydrochloride.
    Neuropharmacology, 13: 11-20.

    APPEL, S.H. & DAY, E.D. (1976) Cellular and subcellular fractionation.
    In: Seigel, G.L., Albers, R.W., Katzman, R., & Agranoff, B.W., ed.
    Basic neurochemistry, Boston, Toronto, Little, Brown and Company,
    pp. 34-57.

    ARBUKHANAOVA, M.S. (1981) Metabolism of glutamic acid in the cerebral
    mitochondria during cooling  disease.  In: Mitochondria: the
    mechanisms of conjugation and regulation, Puschino, p. 76.

    ASABAYEV, CH., BONCHKOVSKAYA, T.YU., & ZHEGALLO, I.G. (1972) [A study
    of the reaction of the animal nervous system to the effect of
    low-strength electromagnetic fields.] In: [Labour hygiene and the
    biological effect of radio frequency electromagnetic fields,]
    Moscow, pp. 48-49 (in Russian).

    AVROVA, N.F.  (1971)  Brain ganglioside patterns of vertebrates.
    J. Neurochem., 18: 667-674.

    BAENNINGER, L.P. (1966) The reliability of dominance orders in rats.
    Anim. Behav., 14: 367-371.

    BAGGETT, J.MCC., THURESON-KLEIN, A., & KLEIN, R.L. (1980) Effects of
    chlordecone on the adrenal medulla of the rat. Toxicol. appl.
    Pharmacol., 52: 313-322.

    BAKER, P.F., HODGKIN, A.L., & RIDGWAY, E.B. (1971) Depolarization and
    calcium entry in giant squid axons. J. Physiol., 218: 709-775.

    BAKIR, F., DAMLUJI, S.F., AMIN-ZAKI, L., MURTADHA, M., KHALIDI, A. 
    AL-RAWL, N.Y.  TIKRITI, S.  DHARIR, H.I., CLARKSON, T.W., SMITH, J.C.,
    & DOHERTY, R.A.  (1973) Methylmercury poisoning in Iraq: an
    inter-university report. Science, 181: 230-241.

    BALNAVE, R.J. & GAGE, P.W. (1973) The inhibitory effect of manganese
    on transmitter release at the neuromuscular junctions of the toad.
    Br. J. Pharmacol., 47: 339-352.

    BANCROFT, J.D. & STEVENS, A. (1977) Theory and practice of
    histological techniques, New York, Churchill Livingston.

    BARKER, J.L. & MCKELVY, J.F., ed. (1983) Current methods in cellular
    neurobiology: electrophysiological and optical recording techniques,
    New York, John Wiley and Sons, Vol. 3.

    BARRETT, R.J., LEITH, N.J., & RAY, O.S. (1973) A behavioral and
    pharmacological analysis of variables mediating active avoidance
    behaviour in rats. J. comp. Physiol. Psychol., 82: 489-500.

    BARRETT, R.J., LEITH, N.J., & RAY, O.S. (1974) An analysis of the
    facilitation of avoidance acquisition produced by delta-amphetamine
    and scopolamine. Behav. Biol., 11: 189-203.

    BASAR, E. (1980) EEG-brain dynamics. In: Relation between EEG and
    brain-evoked potentials, Amsterdam, Oxford, New York, Elsevier
    Science Publishers.

    BAUMGARTNER, G.  GAWEL, M.J.  KAESER, H.E., PALLIS, C.A., CLIFFORD,
    F.R., SCHAUMBURG, H.H., THOMAS, P.K., & WADIA, N.H. (1979)
    Neurotoxicity of halogenated hydroxyquinolines: clinical analysis of
    cases reported outside Japan. J. Neurol. Neurosurg. Psychiatr.,
    42: 1073-1083.

    BEARD, R.R. & GRANDSTAFF, N. (1975) Carbon monoxide and human
    function. In: Weiss, B. & Laties, V., ed. Behavioural toxicology,
    New York, Plenum Press, pp. 1-26.

    BECK, E.C., DUSTMAN, R.E., & LEWIS, E.G. (1975) The use of the
    averaged evoked potential in the evaluation of central nervous system
    disorders. Int. J. Neurol., 9: 211-232.

    BENESOVA, O., HORVATH, M., & MIKISKA, A. (1956) [Assessment of the
    depth and duration of narcosis according to the electrophysiological
    characteristics of the cerebral cortex.] Physiol. bohemoslov.,
    5: 188-194 (in German).

    BENNETT, S. & BOWERS, D. (1976) An introduction to multivariate
    techniques for social and behavioral sciences, New York, Halsted
    Press.

    BENSON, C.K. & ZAGNI, P.A. (1965) The effect of 1-alpha-
    methylallylthiocarbamyl-2-methylthiocarbamylhydrazine (ICI 33828) on
    lactation in the rat. J. Endocr., 32: 275-285.

    BERGER, H. (1929) [On the electroencephalogram in man.] Arch.
    Psychiat. Nervenkr., 87: 527-550 (in German).

    BERNARD, C. (1927) An introduction to the study of experimental
    medicine, New York, MacMillan Inc., p. 127.

    BESSER, G.M., DELITALA, G., GROSSMAN, A., STUBBS, W.A., & YEO, T.
    (1980) Chlorpromazine, haloperidol, metoclopramide, and domperidone
    release prolactin through dopamine antagonism at low concentrations
    but paradoxically inhibit prolactin release at high concentrations.
    Br. J. Pharmacol., 71: 569-573.

    BHAGAT, B. & WHEELER, M. (1973) Effect of amphetamine on the swimming
    endurance of rats. Neuropharmacology, 12: 1161-1165.

    BIGNAMI, G., ROSIC, N., MICHALEK, H., MILOSEVIC, M., & GATTI, G.L.
    (1975) Behavioral toxicity of anticholinesterase agents:
    methodological, neurochemical and neuropsychological aspects. In:
    Weiss, B. & Laties, V.G., ed. Behavioral toxicology, New York,
    Plenum Press, pp. 155-215.

    BLACKWOOD, W., MCMENEMEY, W.H., MEYER, A., NORMAN, R.M., & RUSSELL,
    D.S. (1971) Greenfield's neuropathology, London, Edward Arnold
    Publishing, Ltd.

    BLOOM, A.S., STOOTZ, C.G., & DIERINGER, T. (1983) Pyrethroid effects
    on operant responding and feeding. Neurobehav. Toxicol. Teratol.,
    5: 321-324.

    BOER, G.J. (1974) A microassay for arylamidase activity. Anal.
    Biochem., 59: 410-411.

    BOKINA, A.I., EKSLER, N.D., SEMENENKO, A.D., & MERKURYVA, R.V. (1976)
    Investigation of the mechanism of action of atmospheric pollutants on
    the central nervous system and comparative evaluation of methods of
    study. Environ. Health Perspect., 13: 37-42.

    BOKINS, A.I., KAN, T.M., BEREZINA, O.V., & PINIGINA, I.A. (1979)
    [Study of complex forms of animal behavior in a hygenic evaluation of
    toxic substances.] Gig. i Sanit., 7: 49-51 (in Russian).

    BONASHEVSKAYA, T.I., SHAMARIN, A.A., KESAREV, V.S., & YURASOVA, O.I.
    (1983) [Morphological criteria and a system of indexes for the
    evaluation of neurotoxic effects produced by environmental chemical
    factors (on the basis of a carbon monoxide model).] Gig. i Sanit.,
    10: 31-33 (in Russian, with English summary).

    BONDY, S.C. & AGRAWAL, A.K. (1980) The inhibition of cerebral high
    affinity receptor sites by lead and mercury compounds. Arch.
    Toxicol., 46: 249-256.

    BONDY, S.C. & HARRINGTON, M.L. (1979) Calcium-dependent release of
    putative neurotransmitters in the chick visual system. Neuroscience,
    4: 1521-1527.

    BONDY, S.C., HARRINGTON, M.L., ANDERSON, C.L., & PRASAD, K.N. (1979)
    Low concentration of an organic lead compound alters transport and
    release of putative neurotransmitters. Toxicol. Lett., 3: 35-41.

    BONDY, S.C., TILSON, H.A., & AGRAWAL, A.K. (1981) Neurotransmitter
    receptors in brains of acrylamide-treated rats. II. Effects of
    extended exposure to acrylamide. Pharmacol. Biochem. Behav.,
    14: 533-537.

    BONILLA, E. (1980) L-tyrosine hydroxylase activity in the rat brain
    after chronic oral administration of manganese chloride. Neurobehav.
    Toxicol., 2: 37-41.

    BONILLA, E.  SALAZAR, E.  VILLASMIL, J.J., & VILLALOBOS, R. (1982) The
    regional distribution of manganese in the normal human brain.
    Neurochem. Res., 7: 221-227.

    BOYES, W.K., JENKINS, D.E., & DYER, R.S. (1985) Chlordimeform produces
    contrast-dependent changes in visual-evoked potentials in hooded rats.
    Exp. Neurol., 89: 391-407.

    BRACELAND, F.J. (1942) Mental symptoms following carbon disulfide
    absorption and intoxication. Ann. intern. Med., 16: 246-261.

    BRADLEY, P.B. (1958) The central action of certain drugs in relation
    to the reticular formation of the brain. In: Jasper, H.H., Proctor,
    L.D., Knighton, R.S., Noshay, W.C., & Costello, R.T., ed. Reticular
    formation of the brain, Boston, Toronto, Little, Brown and Company,
    pp. 123-149.

    BRADY, K., HERRERA, Y., & ZENICK, H. (1976) Influence of parental lead
    exposure on subsequent learning ability of offspring. Pharmacol.
    Biochem. Behav., 3: 561-565.

    BRIERLEY, J.B. (1976) Cerebral hypoxia. In: Blackwood, W. & Corsellis,
    J., ed. Greenfield's neuropathology, London, Edward Arnold
    Publishing, Ltd., pp. 43-85.

    BROWN, A.W., ALDRIDGE, W.N., STREET, B.W., & VERSCHOYLE, R.D. (1979)
    The behavioural and neuropathological sequelae of intoxication by
    trimethyltin compounds in the rat. Am. J. Pathol., 97: 59-82.

    BUELKE-SAM, J. & KIMMEL, C.A. (1979) Development and standardization
    of screening methods for behavioral teratology. Teratology,
    20: 17-29.

    BULL, R.J. & LUTKENHOFF, S.D. (1975) Changes in the metabolic
    responses of brain tissue to stimulation, in vitro, produced by in
    vivo administration of methylmercury. Neuropharmacology,
    14: 351-359.

    BURES, J. & BURESOVA, O. (1979) New methods in memory research. In:
    Proceedings of the 21st Annual Meeting of the Czechoslovak
    Psychopharmacological Society, Jesenik Spa, 9-14 January, 1979,
    pp. 7-12.

    BURESOVA, O. & BURES, J. (1982) Capacity of working memory in rats as
    determined by performance on a radial maze. Behav. Processes,
    7: 63-72.

    BURGEN, A.S.V., KOSTERLITZ, H.W., & IVERSEN, L.L., ed. (1980)
    Neuroactive peptides, London, Royal Society.

    BURGUS, R. & GUILLEMIN, R. (1970) Chemistry of thyrotropin releasing
    factor (TRF). In: Meites, J., ed. Hypophysiotropic hormones of the
    hypothalamus: assays and chemistry, Baltimore, Maryland, Williams
    and Wilkins Company, pp. 227-241.

    BUSHNELL, P.J. & BOWMAN, R.E. (1979) Effects of chronic lead ingestion
    on social development in infant rhesus monkeys. Neurobehav.
    Toxicol., 1: 207-219.

    BUTCHER, R.E. (1976) Behavioral testing as a method for assessing
    risk. Environ. Health Perspect., 18: 75-81.

    BUTCHER, R.E. & VORHEES, C.V. (1979) A preliminary test battery for
    the investigation of the behavioral teratology of selected
    psychotropic drugs.  Neurobehav. Toxicol., 1(Suppl. 1): 207-212.

    CABE, P.A. & TILSON, H.A. (1978) Hind limb extensor response: a method
    for assessing motor dysfunction in rats. Pharmacol. Biochem. Behav.,
    9: 133-136.

    CAMMER, W. (1980) Toxic demyelination: biochemical studies and
    hypothetical mechanisms. In: Spencer, P.S. & Schaumburg, H.H., ed.
    Experimental and clinical neurotoxicity, Baltimore, Maryland,
    Williams and Wilkins Company, pp. 239-256.

    CAMMER, W. & MOORE, C.L. (1972) The effect of hexachlorophene on the
    respiration of brain and liver mitochondria. Biochem. biophys. Res.
    Comm., 46: 1887-1894.

    CANNON, S.B., VEAZEY, J.M., Jr, JACKSON, R.S., BURSE, V.W., HAYES, C.,
    STRAUB, W.E., LANDRIGAM, P.J., & LIDDLE, J.A. (1978) Epidemic kepone
    poisoning in chemical workers. Am. J. Epidemiol., 17: 529.

    CARBONE, E., WANKE, E., PRESTIPINO, G., POSSANI, L.D.  & MAELICKE, A.
    (1982) Selective blockage of voltage-dependent K+ channels by a
    novel scorpion toxin. Nature (Lond.), 296: 90-91.

    CARILLI, C.T., FARLEY, R.A., PERLMAN, D.M., & GANTHEY, L.C. (1982) 
    The active site structure of Na+- and K+- stimulated ATPase. J.
    biol. Chem., 257: 5601-5606.

    CASARETT, L.J. (1975) Toxicologic evaluation. In: Casarett, L.J. &
    Doull, J., ed. Toxicology; the basic science of poisons, New York,
    MacMillan Inc., pp. 11-25.

    CASTRO, J.A. & GILLETTE, J.R. (1967) Species and sex differences in
    the kinetic constants for the N-demethylation of ethyl-morphine by
    liver microsomes. Biochem. biophys. Res. Comm., 28: 426-430.

    CATTERALL, W.A. (1977) Activation of the action potential Na+
    ionophore by neurotoxins. J. biol. Chem., 252: 8669-8676.

    CAVALLERI, A. (1975) Serum thyroxine in the early diagnosis of carbon
    disulfide poisoning. Arch. environ. Health, 30: 85-87.

    CAVALLERI, A., POLATTI, F., & BOLIS, P.F. (1978) Acute effects of
    tetramethylthiuram disulfide on serum levels of hypophyseal hormones
    in humans. Scand. Work environ. Health, 4: 66-72.

    CAVANAGH, J.B. (1964) The significance of the "dying-back" process in
    experimental and human neurological disease. Int. Rev. exp. Pathol.,
    3: 219-267.

    CAVANAGH, J.B. & CHEN, F.C.-K. (1971) The effects of methyl-mercury-
    dicyandiamide on the peripheral nerves and spinal cord of rats. Acta
    neuropathol., 19: 208-215.

    CAVANAGH, J.B. & JACOBS, J.M. (1954) Some quantitative aspects of
    diphtheritic neuropathy. Br. J. exp. Pathol., 45: 309-327.

    CAVANAGH, J.B., PASSINGHAM, R.J., & VOGT, J.A. (1964) Staining of
    sensory and motor nerves in muscles with Sudan Black B. J. Pathol.
    Bacteriol., 88: 89-92.

    CECCARELLI, B. & HURLBUT, W.P. (1980) Ca-dependent recycling of
    synaptic vesicles at the frog neuromuscular junction. J. cell Biol.,
    87: 297-303

    CHANCE, M.R.A. (1946) Aggregation as a factor influencing the toxicity
    of sympathomimetic amines in mice. J. Pharmacol. exp. Ther.,
    87: 214-219.

    CHANG, L. (1977) Neurotoxic effects of mercury -- a review. Environ.
    Res., 14: 329-373.

    CHANG, L. (1980) Mercury. In: Spencer, P.S. and Schaumburg, H.H., ed.
    Experimental and clinical neurotoxicology, Baltimore, Maryland,
    Williams and Wilkins Company, pp. 508-526.

    CHANG-TSUI, Y.H. & HO, I.K. (1979) Effects of kepone (chlordecone) on
    synaptosomal gamma-aminobutryic acid uptake in the mouse.
    Neurotoxicology, 1: 357-361.

    CHAPMAN, R.M. (1983) Gonadal injury resulting from chemotherapy. Am.
    J. ind. Med., 4: 149-161.

    CHIBA, S. & ANDO, K. (1976) Effects of chronic administration of
    kanamycin on conditioned suppression to auditory stimulus in rats.
    Jpn. J. Pharmacol., 26: 419-426.

    CHO, E.S., SPENCER, P.S., JORTNER, B.S., & SCHAUMBURG, H.H. (1980) A
    single intravenous injection of doxorubicin (AdriamycinR) induces
    sensory neuropathy in rats. Neurotoxicology, 1: 583-591.

    CHOPRA, I.J., SOLOMON, D.J., TECO, G.N.C., & NGUYEN, A.H. (1980)
    Inhibition of hepatic outer ring monodeiodination of thyroxine and
    3,3',5'-triiodothyronine by sodium salicylate. Endocrinology,
    106: 1728-1734.

    CIOMS (1983) Safety requirements for the first use of new drugs and
    diagnostic agents in man, Geneva, Council for International
    Organizations of Medical Sciences.

    CITOVIC, I.S. (1930) [The procedure of studying the effects of oil
    products on the organism of animals.] Inst. Work Saf., 1: 92
    (in Russian).

    CLAUSEN, T. & FLATMAN, J.A. (1977) The effects of catecholamines on
    Na-K transport and membrane potential in rat soleus muscle.
    J. Physiol., 270: 383-414.

    COLLIER, B. & ILSON, D. (1977) The effect of preganglionic nerve
    stimulation on the accumulation of certain analogues of choline by a
    sympathetic ganglion. J. Physiol., 264: 489-509.

    COLLINS, R.D. (1982) Illustrated manual of neurologic diagnosis, 2nd
    ed., Philadelphia, Pennsylvania, J.B. Lippincott Company.

    COLOTLA, V.A. (1981) Adjunctive polydipsia as a model of alcoholism.
    Neurosci. Biobehav. Rev., 5: 335-342.

    COLOTLA, A., BAUTISTA, M., LORENZANA-JIMENEZ, M., & RODRIGUEZ, R.
    (1979) Effects of solvents on schedule-controlled behavior.
    Neurobehav. Toxicol., 1(Suppl. 1): 113-118.

    CONSTANTOPOULOS, G., MCCOMB, R.D., & DECABAN, A. (1976) Neurochemistry
    of the mucopolysaccharidoses: brain glycosaminoglycans in normal and
    four types of mucopolysaccharidoses. J. Neurochem., 26: 901-908.

    CORY-SLECHTA, D.A. & THOMPSON, T. (1979) Behavioral toxicity of
    chronic postweaning lead exposure in the rat. Toxicol. appl.
    Pharmacol., 47: 151-159.

    CORY-SLECHTA, D.A., BISSEN, T., YOUNG, M., & THOMPSON, T. (1981)
    Chronic postweaning lead exposure and response duration performance.
    Toxicol. appl. Pharmacol., 60: 78-84.

    COSTA, A. & MURAD, J.E. (1969) Study of the association between
    chlorpromazine and analgesics through the method of electrical
    stimulation of the rabbit dental pulp. Rev. Bras de Pesgnisa Med.
    Biol., 2: 235-240.

    COSTA, E. (1970) Simple neuronal models to estimate turnover rate of
    noradrenergic transmitters in vivo. In: Costa, E. & Giabocini, E.,
    ed. Advances in biochemical psychopharmacology. II. Biochemistry of
    simple neuronal models, New York, Raven Press.

    COTMAN, C.W. & BARKER, G.A. (1974) The making of a synapse. Rev.
    Neurosci., 1: 1-62.

    COTMAN, C.W., HAYCOCK, J.W., & WHITE, W.F. (1976) Stimulus secretion
    coupling processes in brain: analysis of noradrenaline and gamma-
    aminobutyric acid release. J. Physiol., 254: 475-505.

    COURAUD, F.C., ROCHAT, H., & LISSITSKY, S. (1978) Binding of scorpion
    and sea anemone neurotoxins to a common site related to the action
    potential Na+ ionophore in neuroblastoma cells. Biochem. biophys.
    Res. Comm., 83: 1525-1530.

    CRANE, A.M. & GOLDMAN, P.S. (1979) An improved method for embedding
    brain tissue in albumin-gelatin. Stain Technol., 54: 71-75.

    CRAPPER, D.R. & DE BONI, U. (1980) Aluminum. In: Spencer, P.S. &
    Schaumburg, H.H., ed. Experimental and clinical neurotoxicology,
    Baltimore, Maryland, Williams and Wilkins Company, pp. 326-335.

    CUTLER, M.G. (1977) Effects of exposure to lead on social behaviour in
    the laboratory mouse. Psychopharmacology, 52: 279-282.

    DAMSTRA, T. (1978) Environmental chemicals and nervous system
    dysfunction. Yale J. Biol. Med., 51: 457-468.

    DAMSTRA, T. & BONDY, S.C. (1980) The current status and future of
    biochemical assays for neurotoxicity. In: Spencer, P.S. & Schaumburg,
    H.H., ed. Experimental and clinical neurotoxicology, Baltimore,
    Maryland, Williams and Wilkins Company, pp. 820-833.

    DAMSTRA, T. & BONDY, S.C. (1982) Neurochemical approaches to the
    detection of neurotoxicity. In: Mitchell, C.L., ed. Nervous system
    toxicology, New York, Raven Press, pp. 349-373.

    DAUBE, J. (1980) Nerve conduction studies. In: Aminoff, J., ed.
    Electrodiagnosis in clinical neurology, New York, Churchill
    Livingstone, pp. 229-264.

    DAVIS, M. (1980) Neurochemical modulation of sensory-motor reactivity:
    acoustic and tactile startle reflexes. Neurosci. Biobehav. Rev.,
    4: 241-263.

    DE JESUS, C.P.V., TOCOFIGHI, J., & SNYDER, D.R. (1978) Sural nerve
    conduction study in the rat: a new technique for studying experimental
    neuropathies. Muscle Nerve, 1: 162-167.

    DESAIAH, D., GILLILAND, T., HO, I.K., & MEHENDALE, H.M. (1980)
    Inhibition of mouse brain synaptosomal ATPases and ouabain binding by
    chlordecone. Toxicol. Lett., 6: 275-285.

    DESCLIN, J.C. (1974) Histological evidence supporting the inferior
    olive as the major source of cerebellar climbing fibres in the rat.
    Brain Res., 77: 365-384.

    DESI, I. (1983)  Neurotoxicological investigation of pesticides in
    animal experiments. Neurobehav. Toxicol. Teratol., 5: 503-515.

    DESI, I. & SOS, J. (1962) Central nervous injury by a chemical
    herbicide. Acta Med. Acad. Sci. Hung., 18: 429-433.

    DESI, I. & SOS, J. (1963) Experimental lesions of the central nervous
    system induced by triorthocresyl phosphate. Acta Physiol. Acad. Sci.
    Hung., 23: 63-68.

    DESI, I., GONCZI, L., SIMON, G., FARKAS, I., & KNEFFEL, Z. (1974)
    Neurotoxicologic studies of two carbamate pesticides in subacute
    animal experiments. Toxicol. appl. Pharmacol., 27: 465-476.

    DEWAR, A.J. & MOFFETT, B.J. (1979) Biochemical methods for detecting
    neurotoxicity. A short review. Pharmacol. Ther., 5: 545-562.

    DEWAR, A.J., MOFFETT, B.J., & SITTON, M.F. (1977) The effect of
    anti-inflammatory drugs on retinal dystrophy in the rat. Toxicol.
    appl. Pharmacol., 42: 65-74.

    DEWS, P.B. (1982) Epistemology of screening for behavioral toxicity.
    In: Mitchell, C.L., ed. Nervous system toxicology, New York, Raven
    Press, pp. 229-236.

    DEWS, P.B. & WENGER, G.R. (1979) Testing for behavioral effects of
    agents. Neurobehav. Toxicol., 1(Suppl. 1): 119-127.

    DIAMOND, S.S. & SLEIGHT, S.D. (1972) Acute and subchronic
    methylmercury toxicosis in the rat. Toxicol. appl. Pharmacol.,
    23: 197-207.

    DICKSITH, T.S.S. & DATTA, K.K. (1972) Effect of intratesticular
    injection of lindane and endrine on the testes of rats. Acta
    pharmacol., 31: 1-10.

    DIETZ, D.D., MCMILLAN, D.E., GRANT, L.D., & KIMMEL, C.A. (1978)
    Effects of lead on temporally-spaced responding in rats. Drug chem.
    Toxicol., 1(4): 401-419.

    DINGLEDINE, R., ed. (1984) Brain slices, New York, Plenum Press,
    442 pp.

    DIXON, R.L. (1976) Problems in extrapolating toxicity data for
    laboratory animals to man. Environ. Health Perspect., 13: 43-50.

    DOLPHIN, A.C. & GREENGARD, P. (1981) Neurotransmitter- and
    neuromodulator-independent alterations in phosphorylation of protein I
    in slices of rat facial nucleus. J. Neurosci., 1: 192-203.

    DRAPER, W.A. (1967) A behavioural study of the home-cage activity of
    the white rat. Behaviour, 28: 280-306.

    DRENTH, H.J., ENSBERG, I.F.G., ROBERTS, D.V., & WILSON, A. (1972)
    Neuromuscular function in agriculture workers using pesticides. Arch.
    environ. Health, 25: 395-398.

    DROUIN, J. & LABRIE, F. (1981) Interactions between 17 beta-estradiol
    and progesterone in the control of luteinizing hormone and follicle
    stimulating hormone release in rat pituitary cells in culture.
    Endocrinology, 108: 52-57.

    DUBOIS, K.P. & PUCHALA, E. (1961) Studies on the sex differences in
    the toxicity of cholinergic phosphorothioate (26791). Proc. Soc. Exp.
    Biol. Med., 107: 908-911.

    DUBOIS, K.P., DOULL, J., SALERNO, P.R., & COON, J.M. (1949) Studies on
    the toxicity and mechanisms of action of p-nitro-phenyl diethyl-
    thionosphosphate (parathion). J. Pharmacol. exp. Ther., 95: 79-91.

    DUBOWITZ, V. & BROOKE, M.H. (1973) Muscle biopsy: a modern approach,
    London, W.B. Saunders.

    DUNLOP, D.S., VAN ELDEN, W., & LAJTHA, A. (1975) Optimal conditions
    for protein synthesis in incubated slices of rat brain. Brain Res.,
    99: 303-318.

    DUNN, A.J. (1975) Intercerebral injections inhibit amino acid
    incorporation into brain protein. Brain Res., 99: 405-409.

    DUNN, A.J. (1977) Measurements of the rate of brain protein synthesis.
    In: Roberts, S., Lajtha, A., & Gispen, W.H., ed. Mechanisms,
    regulation, and special function of protein synthesis in the brain,
    Amsterdam, Oxford, New York, Elsevier Science Publishers, pp. 97-105.

    DURHAM, W.F., CUETO, C., Jr, & HAYES, W.J., Jr (1956) Hormonal
    influence on DDT metabolism in the white rat. Am. J. Physiol.,
    187: 373-377.

    DYCK, P.J., THOMAS, P.K., LAMBERT, E.H., & BUNGE, R. (1984)
    Peripheral neuropathy, 2nd ed., Philadelphia, Pennsylvania, W.B.
    Saunders, Vol. 1.

    DYER, R.S. (1980) Effects of prenatal and postnatal exposure to carbon
    monoxide on visually-evoked responses in rats. In: Weiss, B. &
    Merigan, W., ed. Neurotoxicity of the visual system, New York, Raven
    Press, pp. 17-33.

    DYER, R.S. (1985a) The use of sensory-evoked potentials in toxicology.
    Fundam. appl. Toxicol., 5: 24-40.

    DYER, R.S. (1985b) Neurophysiological measures following exposure to
    toxic substances. In: Annau, Z., ed. Behavioral toxicology,
    Baltimore, Maryland, Johns Hopkins.

    DYER, R.S. & BOYES, W.K. (1983) Use of neurophysiological challenges
    for the detection of neurotoxicity. Fed. Proc., 42: 3201-3206.

    DYER, R.S., ECCLES, C.U., & ANNAU, Z. (1978) Evoked potential
    alterations following prenatal methylmercury exposure. Pharmacol.
    Biochem. Behav., 8: 137-141.

    DYER, R.S., SWARTZWELDER, H.S., ECCLES, C.W., & ANNAU, Z. (1979)
    Hippocampal after discharges and their post-ictal sequelae in rats: a
    potential tool for assessment of CNS neurotoxicity. Neurobehav.
    Toxicol., 1: 5-19.

    ECKERMAN, D.A., GORDON, W.A., EDWARDS, J.D., MACPHAIL, R.C., & GAGE,
    M.I. (1980) Effects of scopolamine, pentabarbital, and amphetamine on
    radial arm maze performance in the rat. Pharmacol. Biochem. Behav.,
    12: 595-602.

    EDWARDS, P.M. & PARKER, V.H. (1977) A simple, sensitive, and objective
    method for early assessment of acrylamide neuropathy in rats.
    Toxicol. appl. Pharmacol., 40: 589-591.

    EGAN, G.F., LEWIS, S.C., & SCALA, R.A. (1980) Experimental design for
    animal toxicity studies. In: Spencer, P.S. & Schaumburg, H.H., ed.
    Experimental and clinical neurotoxicology, Baltimore, Maryland,
    Williams & Wilkins Company, pp. 708-725.

    EICHELMAN, B., HEGSTRAND, L.R., MCMURRAY, T., & KANTAX, K.M. (1981)
    Cyclic nucleotides and aggressive behaviour. Psychopharmacology of
    aggression and social behaviour. Pharmacol. Biochem. Behav.,
    14(Suppl. 1): 7-12.

    ELSNER, J., LOOSER, R., & ZBINDEN, G. (1979) Quantitative analysis of
    rat behaviour patterns in a residential maze. Neurobehav. Toxicol.,
    1(Suppl. 1): 163-174.

    EMSON, P.C. (1983) Chemical neuroanatomy, New York, Raven Press.

    ENJALBERT, A., RUBERG, M., & KORDON, C.  (1978) Neuroendocrine control
    of prolactin secretion. In: Robyn, C. & Harter, M., ed. Progress in
    prolactin physiology and pathology, Amsterdam, Oxford, New York,
    Elsevier Science Publishers, pp. 83-94.

    EROSCHENKO, V.P. & PALMITER, R.D. (1980) Estrogenicity of kepone in
    birds and mammals. In: McLachlan, J., ed. Estrogens in the
    environment, Amsterdam, Oxford, New York, Elsevier Science
    Publishers, pp. 305-325.

    EROSCHENKO, V.O. & WILSON, W.O. (1975) Cellular changes in the gonads,
    liver, and adrenal glands of Japanese quail as affected by the
    insecticide kepone. Toxicol. appl. Pharmacol., 31: 491-504.

    EVANS, H.L. (1978) Behavioral assessment of visual toxicity. Environ.
    Health Perspect., 26: 53-58.

    EVANS, H.L. & WEISS, B. (1978) Behavioral toxicology. In: Blackman,
    D.E. & Sanger, D.J., ed. Contemporary research in behavioral
    pharmacology, New York, Plenum Press, pp. 449-487.

    EVANS, H.L., LATIES, V.G., & WEISS, B. (1975) Behavioral effects of
    mercury and methylmercury. Fed. Proc., 34: 1858-1867.

    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.

    EVANS, W.O. (1962) A comparison of the analgetic potency of some
    analgesics as measured by the "flinch-jump" procedure.
    Psychopharmacologia, 3: 51-54.

    FADEEV, JU.A. (1980) [Unit activity of the brain cortex in realization
    of goal-directed behaviour.] Usp. fiziol. nauk., 11(3): 12-46 (in
    Russian).

    FADEEV, JU.A. & ANDRIANOV, V.V. (1971) Methods of registration of
    neuronal activity in freely-moving animals. Physiol. J. (USSR),
    57(8): 1217-1219.

    FALCK, B., HILLARP, N.A., THIEME, G., & TORP, A. (1962) Fluorescence
    of catecholamines and related compounds condensed with formaldehyde.
    J. Histochem. Cytochem., 10: 348-354.

    FALK, J.L. (1970) Readings in behavioral pharmacology, New York,
    Appleton-Century-Crofts.

    FELLEGIOVA, M., ADAMEC, O., & DAHKOVA, A. (1977) [The effect of
    lindane on the metabolism of testosterone in the rat.] Cs. Hyg.,
    22: 115-120 (in Slovak).

    FERSTER, C.B. & SKINNER, B.F. (1957) Schedules of reinforcement, New
    York, Appleton-Century-Crofts.

    FESTING, M.F. (1979) Properties of inbred strains and outbred stocks,
    with special reference to toxicity testing. J. environ. Health
    Toxicol., 5: 53-68.

    FILE, S. (1980) The use of social interaction as a method for
    detecting anxiolytic activity of chlordiarepoxide-like drugs.
    J. Neurosci. Methods, 2: 219-238.

    FINK, G. (1979) Feedback actions of target hormones on hypothalamus
    and pituitary with special reference to gonadal steroids. Ann. Rev.
    Physiol., 41: 571-585.

    FISHMAN, J. (1982) Catechol estrogens and pituitary function. In:
    Muldoon, T.G., Mahesh, V.B., & Perez-Ballester, B., ed. Recent
    advances in fertility research. Part A. Developments in reproductive
    endocrinology, New York, Alan R. Lyss, pp. 183-189.

    FISHMAN, W.H. & BERNFELD, P. (1955) Glucuronidases. In: Colowick, S.P.
    & Kaplan, N.O., ed. Methods in enzymology, New York, Academic Press,
    Vol. 1, p. 262.

    FITZGERALD, G.J., KAUFMAN, E.E., SOKOLOFF, L., & SHEIN, H.M. (1974)
    D(-)-beta-hydroxybutyrote dehydrogenase activity in closed cell lines
    of glial and neuronal origin. J. Neurochem., 22: 1163-1166.

    FODOR, G.G., SCHLIPKOTER, H.W., & ZIMMERMAN, M. (1973) The objective
    study of sleeping behaviour in animals as a test of behaviour
    toxicity. In: Horvath, M. & Frantik, M., ed. Adverse effects of
    environmental chemicals and psychotropic drugs, Amsterdam, London,
    New York, Elsevier Science Publishers, Vol. 1, pp. 115-122.

    FOLCH, J., ASCOLI, I., LEES, M., MEATH, J.A., & LEBARON, F.N. (1951)
    Preparation of lipid extracts from brain tissue. J. biol. Chem.,
    191: 833-841.

    FOWLER, S.C. & PRICE, A.W. (1978) Some effects of chlordiaz-epoxide
    and delta-amphetamine on response force during punished responding in
    rats. Psychopharmacology, 56: 211-215.

    FOX, D.A., LOWNDES, H.E., & BIERKAMPER, G.G. (1982) Electro-
    physiological techniques in neurotoxicology. In: Mitchell, C.L., ed.
    Nervous system toxicology, New York, Raven Press, pp. 299-335.

    FOX, W.M. (1965) Reflex-ontogeny and behavioral development of the
    mouse. Anim. Behav., 13: 234-241.

    FRANKOVA, S. (1970) Behavioral responses of rats to early
    overnutrition. Nutr. Metab., 12: 228-230.

    FRANKOVA, S. (1971) Relationship between nutrition during lactation
    and maternal behavior of rats. Act. Nerv. Super. (Praha), 13: 1-8.

    FRANKOVA, S. (1977) Drug-induced changes in the maternal behavior of
    rats. Psychopharmacology, 3: 83-87.

    FRANTIK, E. (1970) The development of motor disturbances in
    experimental chronic carbon disulfide intoxication. Med. Lav.,
    61: 309-313.

    FRANTIK, E. & BENES, V. (1984) Central nervous effect and blood level
    regressions on exposure time paralleled in solvents (toluene, carbon
    tetrachloride, and chloroform). Act. Nerv. super. (Praha), 26: 131.

    FUJIMORI, K., BENET, H., MEHENDALE, H.M., & HO, I.K. (1982) Comparison
    of brain discrete area distribution of chlordecone and mirex in the
    mouse. Neurotoxicology, 3: 125-130.

    FULLERTON, P.M. (1966) Chronic peripheral neuropathy produced by lead
    poisoning in guinea-pigs. J. Neuropathol. exp. Neurol., 25: 214-236.

    GABE, M. (1976) Histological techniques, New York, Paris,
    Springer-Verlag, Masson.

    GAD, S.C. (1982) Statistical analysis of behavioral toxicology data
    and studies. Arch. Toxicol., Suppl. 5: 256-266.

    GAD, S.C. & WEIL, C.S. (1982) Statistics for toxicologists. In: Hayes,
    A.W., ed. Principles and methods of toxicology, New York, Raven
    Press, pp. 270-320.

    GALLARD, G.V., PETRO, Z., & RYAN, K.J. (1978) Phylogenetic
    distribution of aromatase and other androgen-converting enzymes in the
    central nervous system. Endocrinology, 10: 2283-2290.

    GELLER, I., MENDEZ, V., HAMILTON, M., HARTMANN, R.J., & GAUSE, E.
    (1979) Effects of carbon monoxide on operant behavior of laboratory
    rats and baboons. Neurobehav. Toxicol., 1(Suppl. 1): 179-184.

    GELLERT, R.J., HEINRICHS, W.L., & SWERDLOFF, R.S. (1972) DDT
    homologues: estrogen-like effects on the vagina, uterus, and pituitary
    of the rat. Endocrinology, 91: 1095-1100.

    GERHART, J.M., HONG. J.S., UPHOUSE, L.L., & TILSON, H.A. (1982)
    Chlordecone-induced tremor: quantification and pharmacological
    analysis. Toxicol. appl. Pharmacol., 66(2): 234-243.

    GILLIATT, R.W. (1973) Recent advances in the pathophysiology of nerve
    conduction. In: Desmedt, J.E., ed. New developments in
    electromyography and clinical neurophysiology, Basel, Karger,
    Vol. 2, pp. 2-18.

    GLATT, A.F., TALAAT, H.N., & KOELLA, W.P. (1979) Testing of peripheral
    nerve function in chronic experiments with rats. Pharmacol. Ther.,
    5: 539-543.

    GLAUSERT, A.M. (1980) Fixation, dehydration, and embedding of
    biological specimens, Amsterdam, Oxford, New York, Elsevier Science
    Publishers.

    GLOWINSKI, J. & IVERSEN, L.L. (1966) Regional studies of
    catecholamines in the rat brain. I. The disposition of 3H-norephine-
    phrine, 3H-dopamine, and 3H-DOPA in various regions of the brain. J.
    Neurochem., 13: 655-699.

    GOLDSTEIN, B.D., LOWNDES, H.E., & CHO, E.S. (1981) Neurotoxicology of
    vincristine in the cat: electrophysiological studies. Arch.
    Toxicol., 48: 253-264.

    GONCHARUK, V.D. (1981) Alteration of activity of glutamate
    dehydrogenase in ganglionated neurons of rabbits under acute
    experimental emotional stress. Bull. exp. Biol. Med., 10: 494-496.

    GOOD, P. (1979) Detection of a treatment effect when not all
    experimental subjects will respond to treatment. Biometrics,
    35: 483-489

    GORIDIS, C., MASSARELLI, R., SERSENBRENNER, M., & MANDEL, P. (1974)
    Guanyl cyclase in chick embryo brain cell cultures: evidence of
    neuronal localization. J. Neurochem., 23: 135-139.

    GOYER, R.A. & RHYNE, B.C. (1973) Pathological effects of lead. Int.
    Rev. exp. Pathol., 12: 1-7.

    GRANT, L.D. (1976) Research strategies for behavioral toxicology
    studies. Environ. Health Perspect., 18: 85-94.

    GRAY, E.G. & WHITTAKER, V.P. (1962) The isolation of nerve endings
    from brain: an electron microscopic study of the cell fragments of
    homogenation and centrifugation. J. Anat., 96: 79-87.

    GREENBLATT, D.J., SHADER, R.I., FRANKE, K.K., MACLAUGHLIN, E.D.,
    RONSIL, G.J., & KOCH-WESER, J. (1977) Kinetics of intravenous
    chlordiazepoxide: sex differences in drug metabolism. Clin.
    Pharmacol. Ther., 22: 893-903.

    GREENGARD, P. (1979) Cyclic nucleotides, phosphorylated proteins and
    the neuronal system. Fed. Proc., 38: 2208-2217.

    GREGORY, E.H. & PFAFF, D.W. (1971) Development of olfactory-guided
    behavior in infant rats. Physiol. Behav., 6: 573-576.

    GRIFFIN, J.W. & PRICE, D.L. (1980) Proximal axonopathies induced by
    toxic chemicals. In: Spencer, P.S. & Schaumburg, H.H., ed.
    Experimental and clinical neurotoxicology, Baltimore, Maryland,
    Williams and Wilkins Company, pp. 161-178.

    GRIFFIN, J.W., PRICE, D.L., KLUTHE, D.O., & GOLDBERG, M.A. (1980)
    Neurotoxicity of misonidazole in rats. Neuropathol. appl.
    Neurobiol., 6: 656-666.

    GROSS-SELBECK, E. & GROSS-SELBECK, M. (1981) Changes in operant
    behavior of rats exposed to lead at the accepted no-effect level.
    Clin. Toxicol., 18: 1247-1256.

    GROTA, L.S. & ADER, R. (1969) Continuous recording of maternal
    behavior in the rat. Anim. Behav., 17: 72-729.

    GUDELSKY, G.A., NANSEL, D.D., & PORTER, J.C. (1981) Role of estrogen
    in the dopaminergic control of prolactin secretion. Endocrinology,
    108: 440-444.

    HAIDER, S.S., HANSAN, M., & ISLAM, F. (1980) Effect of air pollutant
    hydrogen sulfide on the levels of total lipids, phospholipids, and
    cholesterol in different regions of the guinea-pig brain. Ind. J.
    exp. Biol., 18: 418-420.

    HAMMOND, B., KATZENELLENBOGEN, B.S., KRAUTHAMMER, N., & MCCONNEL, J.
    (1979) Estrogenic activity of the insecticide chlordecone (kepone) and
    interaction with uterine estrogen receptors. Proc. Natl Acad. Sci.
    (USA), 76: 6641-6645.

    HARPUR, E.S. & D'ARCY, P.F. (1975) The quantification of kanamycin
    ototoxicity in the rat using conditioned tone discrimination.
    J. Pharm. Pharmacol., 27: 907-913.

    HARRY, G.J. & TILSON, H.A. (1982) Postpartum exposure to triethyltin
    produces long-term alterations in responsiveness to apomorphine.
    Neurotoxicology, 3: 64-71.

    HASAN, M. & ALI, S.F. (1981) Effects of thallium, nickel, and cobalt
    administration on the lipid peroxidation in different regions of the
    rat brain. Toxicol. appl. Pharmacol., 57: 8-13.

    HAY, A. (1978) Neurotoxins may go unrecognized. Nature (Lond.),
    274: 206.

    HAYAT, M.A. (1973) Principles and techniques of electron microscopy,
    New York, Van Rostrand Reinhold Company, Vols. 1 and 2.

    HEDLUND, B., GAMARA, M., & BARTFAI, I. (1979) Inhibition of striatal
    muscarinic receptors in vivo by cadium. Brain Res., 168: 216-218.

    HEISE, G.A. (1983) Toward a behavioural toxicology of learning and
    memory. In: Zbinden, G., Cuomo, V., Racagni, G., & Weiss, B., ed.
    Application of behaviour pharmacology and toxicology, New York,
    Raven Press, pp. 27-38.

    HEISE, G.A. & MILAR, C. (1984) Drugs and stimulus control. In:
    Iverson, S.D. & Snyder, S.H., ed. Handbook of psychopharmacology,
    New York, Plenum Press, pp. 129-190.

    HOFFMAN, H.S. & FLESHLER, M. (1963) Startle reaction: modification by
    background acoustical stimulation. Science, 141: 928-930.

    HERMAN, S.P., KLEIN, R., TALLEY, F.A., & KRIGMAN, M. (1973) An
    ultrastructural study of methylmercury-induced primary sensory
    neuropathy in the rat. Lab. Inv., 28: 104-118.

    HERTZ, L. (1981) Gamma amino butyric-acid glutamate pathway in glia.
    5th European Neuroscience Congress, Liege, Belgium, 14-18 September,
    1981. Neurosci. Lett, Suppl. 7: 172.

    HETZLER, B.E., HEILBRONNER, R.L., GRIFFIN, J., & GRIFFIN, G. (1981)
    Acute effects of alcohol on evoked potentials in visual cortex and
    superior colliculus of the rat. Electroencephalogr. clin.
    Neurophysiol., 51 69-79.

    HILLE, B. (1976) Gating in sodium channels of nerve. Ann. Rev.
    Physiol., 38 139-152.

    HINDE, R. (1974) Biological basis of human social behaviour, New
    York, McGraw Hill, Inc.

    HOFFMAN, H.S. & FLESHLER, M. (1963) Startle reaction: modification by
    background acoustical stimulation. Science, 141: 928-930

    HOFFMAN, P.N. & LASEK, R.J. (1975) The slow component of axonal
    transport. Identification of major structural polypeptides of the axon
    and their generality among mammalian neurons. J. cell Biol.,
    66: 351.

    HOLCK, H.G.O., MUNIR, A.K., MILLS, L.M., & SMITH, E.L. (1937) Studies
    upon the sex differences in rats in tolerance to certain barbiturates
    and to nicotine. J. Pharmacol. exp. Ther., 60: 323-346.

    HOLMSTEDT, B. (1959) Pharmacology of organophosphorus cholinesterase
    inhibitors. Pharmacol. Rev., 11: 567-688.

    HONG, J.S. & ALI, S.F. (1982) Chlordecone (kepone) exposure in the
    neonate selectively alters brain and pituitary endorphin levels in
    prepuberal and adult rats. Neurotoxicology, 3: 111-118.

    HOPF, H.C. & EYSHOLDT, M. (1978) Impaired refractory periods of
    peripheral sensory nerves in multiple sclerosis. Ann. Neurol.,
    4: 499-501.

    HOROWSKI, D. & GRAF, K.J. (1979) Neuroendocrine effects of
    neuropsychotropic drugs and their possible influence on toxic
    reactions in animals and man: the role of dopamine-prolactin system.
    Arch. Toxicol. Suppl. 2: 93-104.

    HORVATH, M., ed. (1973) Adverse effects of environmental chemicals
    and psychotropic drugs. Quantitative interpretation of functional
    tests, Amsterdam, Oxford, New York, Elsevier Science Publishers,
    Vol. 1, 292 pp.

    HORVATH, M. & FRANTIK, E. (1973) Quantitative interpretation of
    experimental toxicological data: the use of reference substances. In:
    Horvath, M., ed. Adverse effects of environmental chemicals and
    psychotropic drugs, Amsterdam, Oxford, New York, Elsevier Scientific
    Publishing Co., Vol. 1, pp. 11-21.

    HORVATH, M. & FRANTIK, M., ed. (1976) Adverse effects of
    environmental chemicals and psychotropic drugs: neurophysiological
    and behavioural tests, Amsterdam, New York, Oxford, Elsevier,
    Vol. 2, 334 pp.

    HORVATH, M. & FRANTIK, E. (1979) Industrial chemicals and drugs
    lowering central nervous systems activation level: quantitative
    assessment of hazard in man and animals. Act. Nerv. super. (Praha),
    21: 269-272.

    HORVATH, M. & MICHALOVA, C. (1956) Changes of EEG in experimental
    exposure to carbon disulfide in the light of Vvedenskys theory of
    parabiosis. Arch. Gewerbepathol., 15: 131-136.

    HORVATH, M., FRANTIK, E., & KREKULE, P. (1981) Diazepam impairs
    alertness and potentiates a similar effect of toluene. Act. Nerv.
    super. (Praha), 22: 17-19.

    HOWLAND, R.D., VYAS, I.L., LOWNDES, H.E., & ARGENTIERI, T.M. (1980)
    The etiology of toxic peripheral neuropathies; in vitro effects of
    acrylamide and 2,5-hexanedione on brain enolase and other glycolytic
    enzymes. Brain Res., 202: 131-142.

    HUGHES, J.A. & SPARBER, S.B. (1978) Delta-amphetamine unmasks
    postnatal consequences of exposure to methylmercury in utero:
    methods for studying behavioral teratogenesis. Pharmacol. Biochem.
    Behav., 8: 365-375.

    HUNT, V.R., SMITH, M.K., & WORK, D., ed. (1982) Banbury Report II -
    Environmental factors in human growth and development, New York,
    Cold Spring Harbor Laboratory.

    HUNTER, D. & RUSSELL, D.S. (1954) Focal cerebral and cerebellar
    atrophy in human subjects due to organic mercury compounds.
    J. Neurol. Neurosurg. Psychiatr., 17: 235-241.

    HUNTER, D., BORNFORD, R., & RUSSELL, D. (1940) Poisoning by
    methylmercury compounds. Q. S. Med., 9: 193-213.

    HUTCHINGS, D.E. (1978) Behavioral teratology: embryopathic and
    behavioural effects of drugs during pregnancy. In: Gottlieb, G., ed.
    Studies on the development of behaviour and nervous system: early
    influences, New York, Academic Press, pp. 7-34.

    IRWIN, S. (1968) Comprehensive observational assessment: a systematic
    quantitative procedure for assessing the behavioral and psychologic
    state of the mouse. Psychopharmacologia, 13: 222-257.

    ISLAM, F., TAYYABA, K., & HASAN, M. (1983) Organophosphate metasystox-
    induced increment of lipase activity and lipid peroxidation in
    cerebral hemisphere: diminution of lipids in discrete areas of the
    brain. Acta pharmacol. toxicol., 53: 121-124.

    IVERSEN, L.L. (1971) Role of transmitter uptake mechanisms in synaptic
    neurotransmission. Br. J. Pharmacol., 44: 571-591.

    JACOBWITZ, D.M. (1974) Removal of discrete fresh regions of the rat
    brain. Brain Res., 80: 111-115.

    JOHN, T.H., GEGHMAN, C., & REIS, D.J. (1973) Immunochemical
    demonstration of increased accumulation of tyrosine hydroxylase
    protein in sympathetic ganglia and adrenal medulla elicited by
    reserpine. Proc. Natl Acad. Sci. (USA), 70: 2767-2771.

    JOHNSON, B.L. (1980) Electrophysiological methods in neurotoxicity
    testing. In: Spencer, P.S. & Schaumburg, H.H., ed. Experimental and
    clinical neurotoxicology, Baltimore, Maryland, Williams and Wilkins
    Company, pp. 726-742.

    JOHNSON, B.L. & ANGER, W.K. (1983) Behavioral toxicology. In: Rom,
    W.N., ed. Environmental and occupational medicine, Boston, Little
    Brown and Co., pp. 329-350.

    JOHNSON, B.L., ANGER, W.K., SETZIER, J.V., & XINTARAS, C. (1975) The
    application of a computer-controlled time discrimination performance
    to problems in behavioral toxicology. In: Weiss, B. & Laties, V.G.,
    ed. Behavioral toxicology, New York, London, Plenum Press,
    pp. 129-153.

    JOHNSON, E.W. (1980) Practical electromyography, Baltimore,
    Maryland, Williams and Wilkins Company.

    JOHNSON, J.E., Jr (1981) Aging and cell structure, New York, Plenum
    Press, Vol. 1.

    JOY, R.M. (1982) Chlorinated hydrocarbon insecticides. In: Ecobichon,
    D.J. & Joy, R.M., ed. Pesticides and neurological diseases, Boca
    Raton, Florida, CRC Press, pp. 91-150.

    JOY, R.M. (1985) Kindling as a tool in neurotoxicology. Fund. appl.
    Pharmacol., 5: 41-65.

    JOY, R.M., STARK, L.G., PETERSON, S.L., BOWYER, J.T., & ALBERTSON,
    T.E. (1980) The kindled seizure; production of and modification by
    dieldrin in rats. Neurobehav. Toxicol., 2: 117-124.

    JUNQUEIRA, L.C., CARNEIRO, J., & CONTOPOULOS, A. (1977) Basic
    Histology, 2nd ed., Los Altos, Lange.

    KAPLAN, K.L. (1980) Thyroxine 5'-monodeiodination in rat pituitary
    homogenates. Endocrinology, 106: 567-576.

    KARPOVA, O.B., OBEHOVA, E.L., AVIOVA, N.F., & SCHVARTS, E.G. (1978)
    [Content and composition of cerebral gangliosides during Down's
    Syndrome.] Vopros. med. Xhim., 4: 524-527 (in Russian).

    KELLEHER, R.T. & MORSE, W.H. (1968) Determinations of the specificity
    of behavioral effects of drugs. Ergeb. Physiol., 60: 1-56.

    KHOLODOV, Ju.A. & SOLOV'EVA, G.R. (1971) [Occasional publication:
    news on medical equipment manufacture,] fasc. 3, pp. 69-72
    (in Russian).

    KIMURA, J. (1976) Collision techniques: physiological block of nerve
    impulses in studies of motor nerve conduction velocity. Neurology,
    26: 680-682.

    KING, J.A. & VESTAL, B.M. (1974) Visual acuity of peromyscus.
    J. Mammal., 55: 238-243.

    KIRPEKAR, S.M., DIXON, W., & PRATT, J.C. (1970) Inhibitory effect of
    manganese on norephinephrine release from the splenic nerves of cat.
    J. Pharmacol. exp. Ther., 174: 72-76.

    KLEIN, R., HERMAN, S., BRUBAKER, P.E., & LUCIER, G.W. (1972) A model
    of acute methylmercury intoxication in rats. Arch. Path.,
    93: 404-418.

    KODAMA, J., FUKUSHIMA, M., & SAKATA, T. (1978) Impaired taste
    discrimination against quinine following chronic administration of
    theophylline in rats. Physiol. Behav., 20: 151-155.

    KOHLER, W. (1976) The mentality of ages, New York, Liveright,
    336 pp.

    KONAT, G. & CLAUSEN, J. (1977) Triethyl lead-induced intoxication as
    an experimental model of hypomyelination. Proc. Int. Soc.
    Neurochem., 6: 215.

    KORNBERG, A. (1955) Lactic dehydrogenase in muscle. In: Colowick, S.P.
    & Kaplan, N.O., ed. Methods in enzymology, New York, Academic Press,
    Vol. 1, p. 441.

    KOSTIAL, K., LANDEKA, M., & SLAT, B. (1974) Manganese ions and
    synaptic transmission in the superior cervical ganglion of the cat.
    Br. J. Pharmacol., 51: 231-235.

    KOTLIAREVSKI, L.I. (1951) [Methodology for studying motor conditioned
    reflexes in some small animals (white rats and guinea-pigs.]
    Zh. Vyssh. nerv. deyat. imeni, 1(5): 753-761 (in Russian).

    KOTLIAREVSKI, L.I. (1957) [Procedure for the investigation of motor-
    food conditioned reflexes in small animals.] Tr. Inst. Vyssh. nerv.
    deyat. AN SSR, 3: 23-28 (in Russian).

    KRIGMAN, M.R. & HOGAN, E.L.  (1974)  Effect of lead intoxication on
    the postnatal growth of the rat nervous system. Environ. Health
    Perspect., 7: 187-199.

    KRINKE, G., SCHAUMBURG, H.H., SPENCER, P.S., SUTER, Y., THOMANN, G., &
    HESS, R. (1981) Pyridoxine megavitaminosis produces degeneration of
    peripheral sensory neurons (sensory neuronopathy) in the dog.
    Neurotoxicology, 2: 13-24.

    KRINKE, G., SCHAUMBURG, H.H., SPENCER, P.S., THOMANA, P., & HESS, R.
    (1979) Clinoquinol and 2,5-hexanedione induce different types of
    distal axonopathy in the dog. Acta neuropathol., 47: 213.

    KRSIAK, M. (1979) Effects of drugs on behavior of aggressive mice.
    Br. J. Pharmacol., 65: 525-533.

    KRUSIUS, T. & FINNE, J. (1977) Structural features of tissue
    glycoproteins. Fractionation and methylation analysis of gylcopeptides
    derived from rat brain, kidney, and liver. Eur. J. Biochem.,
    78: 369-379.

    KRYZHANOVSKY, G.N. (1973) Mechanism of action of tetanus toxin. Effect
    on synaptic processes and some particular features of toxin binding by
    the nervous tissue. Jaunyn-Schmiedeberg's Arch., 276: 247-270.

    KUCZENSKI, R.T. & MANDELL, A.J. (1972) Regulatory properties of
    soluble and particulate rat brain tyrosine hydroxylase. J. biol.
    Chem., 247: 3114-3116.

    KULKOSKY, P.J., ZELLNER, D.A., HYSON, R.L., & RILEY, A.L. (1980)
    Ethanol consumption of rats in individual, group, and colonial housing
    conditions. Physiol. Psychol., 8: 56-60.

    KUPFER, D. & BULGER, W.H. (1976) Studies on the mechanism of
    estrogenic actions of o-p'-DDT; interactions with the estrogen
    receptor. Pestic. Biochem. Physiol., 6: 561-570.

    KUPFER, D. & BULGER, W.H. (1980) Estrogenic properties of DDT and its
    analogues. In: McLachlan, J., ed. Estrogens in the environment,
    Amsterdam, Oxford, New York, Elsevier Science Publishers, pp. 239-263.

    LAFFERMAN, J.A. & SILBERGELD, E.K. (1979) Erythrosin B inhibition of
    neurotransmitter accumulation by rat brain homogenate. Science,
    206: 363-365.

    LAJTHA, A. & MARKS, N. (1971) Protein turnover. In: Lajtha, A., ed.
    Handbook Neurochemistry, New York, Plenum Press, Vol. 5, pp. 551-629

    LAMARTINIERE, C.A., DIERINGER, C.S., & LUCIER, G.W. (1979) Altered
    ontogeny of glutathione S-transferases by 2,4,5-2',4',5'-
    hexachlorobiphenyl. Toxicol. appl. Pharmacol., 51: 233-238.

    LANDOWNE, D. & RITCHIE, J.M. (1976) On the control of glycogenolysis
    in mammalian nervous tissue by calcium. J. Physiol., 212: 503-517.

    LATIES, V.G. (1982) Contributions of operant conditioning to
    behavioral toxicology. In: Mitchell, C.L., ed. Nervous system
    toxicology, New York, Raven Press, pp. 67-80.

    LAZAROVA, M., BENDOTTI, C., & SAMANIN, R. (1984) Evidence that the
    dorsal raphe area is involved in the effect of clonidine against
    pentylenetetrazol-induced seizures in rats. Naunyn-Schmiedebergs
    Arch. Pharmacol., 325: 12-16.

    LEANDER, J.D. & MACPHAIL, R.C. (1980) Effect of chlordimeform
    (a formamidine pesticide) on schedule-controlled responding of
    pigeons. Neurobehav. Toxicol., 2: 315-321.

    LEANDER, J.D., MCMILLAN, D.E., & BARLOW, T.S. (1977) Chronic mercuric
    chloride: behavioural effects in pigeons. Environ. Res.,
    14: 424-435.

    LEHMANN, H.J. & TACKMANN, E. (1974) Neurographic analysis of trains of
    frequent electrical stimuli in the diagnosis of peripheral nerve
    disease. Eur. Neurol., 12: 293-308.

    LEHRER, G.M. (1974) Measurement of minimal brain dysfunction. In:
    Xintaras, C., Johnson, B.L., & DeGroot, I., ed. Behavioral
    toxicology, Cincinnati, Ohio, US Department of Health, Education and
    Welfare, pp. 450-454 (US DREW (NIOSH) Publication No. 74-126).

    LEONARD, J.L. & ROSENBERG, L.H. (1980) Characterization of essential
    enzyme sulfhydryl groups of thryoxine 5'deiodinase from rat kidney.
    Endocrinology, 108: 444-451.

    LEVINE, T.E. (1976) Effects of carbon disulfide and FLA-63 on operant
    behavior in pigeons. J. Pharmacol. exp. Ther., 199: 669-678.

    LEWEY, F.H. (1941) Experimental chronic carbon disulfide poisoning in
    dogs. J. ind. Hyg. Toxicol., 23: 415-535.

    LILIENTHAL, H., WINNEKE, G., BROCKHAUS, A., MOLIK, B., & SCHLIPKOTER,
    H.-W. (1983) Learning set-formation in rhesus monkeys pre- and
    postnatally exposed to lead. In: Proceedings of the International
    Conference on Heavy metals in the Environment, Heidelberg, September,
    1983, Edinburgh, CEP Consultants, p. 901.

    LINDSLEY, D.B. & WICKE, J.D. (1974) The electroencephalogram:
    autonomous electrical activity in man and animals. In: Thompson, R. &
    Patterson, M., ed. Bioelectric recording techniques, Part B, New
    York, Academic Press.

    LIPP, J.A. (1968) Cerebral electrical activity following soman
    administration. Arch. int. Pharmacodyn., 175: 161-169.

    LOCK, E.A. (1976) Increase in cerebral fluids in rats after treatment
    with hexachlorophene or triethyltin. Biochem. Pharmacol.,
    25: 1455-1458.

    LORENZO, A.V. & GERWITZ, M. (1977) Inhibition of 14C-tryptophane
    transport into brain of lead-exposed neonatal rabbits. Brain Res.,
    132: 386-392.

    LOWITZSCH, K., GOHRING, U., HECKING, E., & KOHLER, H. (1981)
    Refractory period, sensory conduction velocity, and visual-evoked
    potentials before and after haemodialysis. J. Neurol. Neurosurg.
    Psychiatr., 44: 121-128.

    LOWNDES, H.E. & BAKER, T. (1976) Studies on drug-induced neuropathies.
    III. Motor nerve deficit in cats with acrylamide neuropathy. Eur. J.
    Pharmacol., 35: 177-184.

    LOWNDES, H.E., BAKER, T., CHO, E.S., & JORTNER, B.S. (1978a) Position
    sensitivity of deafferented muscle spindles in experimental acrylamide
    neuropathy. J. Pharmacol. exp. Ther., 205: 40-48.

    LOWNDES, H.E., BAKER, T., MICHAELSON, L.P., & VINCENT-ABLAZEY, M.
    (1978b) Attenuated dynamic responses of primary endings of muscle
    spindles; a basis for depressed tendon responses in acrylamide
    neuropathy. Ann. Neurol., 3: 435-437.

    LUKAS, E. (1970) Stimulation electrohypography in experimental
    toxicology (carbon disulfide neuropathy in rats). Med. Lav.,
    61: 302-308.

    MCBLAIN, W.A., LEWIN, V., & WOLFE, F.H. (1976) Differing estrogenic
    activities for the enatiomers of o,p'-DDT in immature female rat.
    Can. J. Physiol., 54: 629-632.

    MACDONALD, J.S. & HARBISON, R.D.  (1977) Methylmercury induced
    encephalopathy in mice. Toxicol. appl. Pharmacol., 39: 195-205.

    MCDONALD, W.I. (1963) The effects of experimental demyelination on
    conduction in peripheral nerve. II. Electrophysical observations.
    Brain, 86: 501-524.

    MCDONOUGH, J.H., Jr, HACKLEY, B.E., Jr, CROSS, R., SAMSON, F., &
    NELSON, S. (1983) Brain regional glucose use during soman-induced
    seizures. Neurotoxicology, 4: 203-210.

    MCKENNA, M.J. & DESTEFANO, V. (1975) A proposed mechanism of action of
    carbon disulfide on carbon disulfide on dopamine-beta-hydroxylase.
    Toxicol. appl. Pharmacol., 33: 137.

    MCMILLAM, D.E. (1982) Effects of chronic administration of pesticides
    on schedule-controlled responding by rats and pigeons. In: Chambers,
    J.E. & Yarborough, J.D., ed. Effects of chronic exposures to
    pesticides on animal systems, New York, Raven Press, pp. 211-226.

    MACPHAIL, R.C. (1982) Studies on the flavour aversions induced by
    trialkyltin compounds. Neurobehav. Toxicol. Teratol., 4: 225-230.

    MACPHAIL, R.C. & LEANDER, J.D. (1980) Flavour aversions induced by
    chlordimeform. Neurobehav. Toxicol., 2: 363-365.

    MACPHAIL, R.C., CROFTON, K.M., & REITER, L.W. (1983) Use of
    environmental challenges in behavioral toxicology. Fed. Proc.,
    42: 3196-3200.

    MACTUTUS, C.F., UNGER, K.L., & TILSON, H.A. (1982) Neonatal
    chlordecone exposure impairs early learning and memory in the rat on a
    multiple measure passive avoidance task. Neurotoxicology, 3: 24-44.

    MAGATA, Y. & TSUKADA, W. (1978) Bulk separation of neuronal cell
    bodies and glial cells from mammalian brain and some of their
    biochemical properties. Rev. Neurosci., 3: 18-27.

    MAILMAN, R.B., FERRIS, R.M., VOGEL, R.A., KILTS, C.D., LIPTOM, M.A.,
    TANG, F.L., SMITH, D.A., MUELLER, R.A., & BREESE, G.R. (1980)
    Non-specificity of the biochemical actions of erythrosine (Red No. 3):
    what relation to behavioural changes. Science, 207: 535-537.

    MARES, P. & VELISEK, L. (1983) Influence of ethosuximide on metrazol-
    induced seizures during ontogenesis in rats. Act. Nerv. super
    (Praha), 25: 292-298.

    MARSHALL, J.F. (1975) Increased orientation to sensory stimuli
    following lateral hypothalamic damage in rats. Brain Res.,
    86: 373-387.

    MARSHALL, J.F. & TEITELBAUM, P. (1974) Further analysis of sensory
    inattention following lateral hypothalamic damage in rats. J. comp.
    physiol. Psychol., 86: 375-395.

    MARSHALL, J.F., TURNER, D.H., & TEITELBAUM, P. (1971) Sensory neglect
    produced by lateral hypothaLamic damage. Science, 174: 523-525.

    MARSHALL, J.F., RICHARDSON, J.S., & TEITELBAUM, P. (1974)
    Nigrostriatal bundle damage and the lateral hypothalamic syndrome.
    J. comp. Physiol. Psychol., 87: 808-830.

    MARSLAND, T.A., GLEES, P., & ERIKSON, L.B. (1954) Modification of
    Glee's silver impregnation for paraffin sections. J. Neuropathol.
    exp. Neurol., 13: 587-591.

    MASTEROV, A.V. (1974) [Features of the interaction of rats in an
    electromagnetic field following extinction of conditioned food
    reactions.] In: [Problems of clinical and experimental medicine,]
    Moscow, pp. 61-63 (in Russian).

    MAURISSEN, J.P.J. (1979) Effects of toxicants on the somatosensory
    system. Neurobehav. Toxicol., 1(Suppl. 1): 23-32.

    MEDVEDEV, B.M. (1975) [The effect of a static electric field on
    conditioned reflex activity.] Fiziol. Zh., 3: 320-325 (in Russian).

    MELLO, N.K. (1975) Behavioral toxicology; a developing discipline.
    Fed. Proc., 34: 1832-1834.

    MENDELL, J.R., SAIDA, K., GANANSIA, M.F., JACKSON, D.B., WEISS, H.,
    GARDIER, R.S., CHRISMAN, C., ALLEN, N., COURI, D., O'NEILL, J., MARKS,
    B., & HETLAND, L. (1974) Toxic polyneuropathy produced by methyl
    n-butyl ketone. Science, 185: 787-789.

    MENDELSON, J.H., MELLO, N.K., & ELLINGBOE, J. (1977) Effects of acute
    alcohol intake on pituitary-gonadal hormones in normal human males.
    J. Pharmacol. exp. Ther., 202: 675-682.

    MERIGAN, W.H. (1979) Effects of toxicants on visual systems.
    Neurobehav. Toxicol., 1: 15-22.

    MERIGAN, W.H. (1980) Visual fields and flicker thresholds in
    methylmercury poisoned monkeys. In; Merigan, W.H. & Weiss, B., ed.
    Neurotoxicity of the visual system, New York, Raven Press,
    pp. 149-163.

    MERKURYVA, R.V. & BUSHINSKAYA, L.I. (1982) [Metabolism of
    sialo-containing glycoproteins and lysosomal enzyme activity under
    neurotropic effect of chemical environmental factors.] Ukranian
    Biochem. J., 53: 31-35 (in Russian).

    MERKURYVA, R.V. & TSAPKOVA, N.N. (1982) Hygienic significance of a
    system of biochemical criteria for the evaluation of hepatotoxic and
    neurotoxic effects of some chemical factors in the environment.
    J. Hyg. Epidemiol. Microbiol. Immunol., 26: 336-341.

    MERKURYVA, R.V., BUSHINSKAIA, L.J., AULIKA, B.V., SHATERNIKOVA, J.S.,
    PROTSENKO, S.J., KULYGINA, A.A., & EKSLER, N.D. (1978) The activity of
    lyosomal and cytoplasmic enzymes under the experimental influence of
    bisulfide of carbon. Vop. Med. Xhim., 2: 151-156.

    MERKURYVA, R.V., KOTOV, A.V., SILOV, V.G., BUSHINSKAIA, L.J.,
    MUCHAMBETOVA, L.KH., & TSAPKOVA, N.N. (1980) Experimental study of
    biochemical and physiological mechanisms of cerebral activity
    influenced by alcohol, Nauka, Minsk Publishing House, pp. 93-94.

    MEYER, O.A., TILSON, H.A., BYRD, W.C., & RILEY, H.T. (1979) A method
    for the routine assessment of fore- and hindlimb grip strength of rats
    and mice. Neurobehav. Toxicol., 1: 233-236.

    MICZEK, K.A. & BARRY, H. (1976) Pharmacology of sex and aggression.
    In: Glick, S.D. & Goldforb, J., ed. Behavioral pharmacology,
    St. Louis, Missouri, C.V. Mosby Co., pp. 176-257.

    MIKISKA, A. (1960) [Determination of the susceptibility to electrical
    activity of the cerebral motor cortex, and its pharmacological and
    toxicological applications. I. Methodology, controlled trials and
    physiological basis.] Arch. Gewerbepathol. Gewerbehyg., 18: 286-299
    (in German).

    MIKISKOVA, H. & MIKISKA, A. (1962) [A contribution to the application
    of defensive skin stimulation in the toxicity of substances acting on
    the CNS. II. Measuring stimulus threshold intensity in guinea-pigs.]
    Int. Arch. Gewerbepathol. Gewerbehyg., 19: 68-75 (in German).

    MIKISKOVA, H. & MIKISKA, A. (1966) Trichloroethanol in
    trichloroethylene poisoning. Br. J. ind. Med., 23: 116-125.

    MIKISKOVA, H. & MIKISKA, A. (1968) Some electrophysiological methods
    for studying the action of narcotic agents in animals, with special
    reference to industrial solvents: a review. Br. J. ind. Med.,
    25: 81-105.

    MILLER, D.B. (1983) Neurotoxicity of the pesticidal carbamates.
    Neurobehav. Toxicol. Teratol., 4: 779-788.

    MILLER, E.K. & DAWSON, R.M.C. (1972) Can mitochondria and synaptosomes
    of guinea-pig brain synthesize phospholipids? Biochem. J.,
    126: 805-807.

    MINEGISHI, A., FUJUMORI, R., SATOH, T., KILAGAWA, H., & YANAURA, S.
    (1979) Changes in serotonin turnover and the brain sensitivity to
    barbituates by disulfiram treatment in rats. Res. Comm. Chem. Pathol.
    Pharmacol., 24: 273-285.

    MINKINS, C., CHEKUNOVA, M.P., & LEVINA, E.N. (1973) [Effect of
    antimony and lead on the adrenal glands and biogenic amines.]
    Gig. Tr. Prof. Zabol., 17: 21-24 (in Russian).

    MITCHELL, C.L. & TILSON, H.A. (1982) Behavioral toxicology in risk
    assessment: problems and research needs, Boca Raton, Florida, CRC
    Press, pp. 265-274 (CRC Critical Reviews in Toxicology 10).

    MITCHELL, C.L., TILSON, H.A., & CABE, P.A. (1982) Screening for
    neurobehavioral toxicity: factors to consider. In: Mitchell, C.L., ed.
    Nervous system toxicology, New York, Raven Press, pp. 237-245.

    MIYOSHI, T. & GOTO, I. (1973) Serial in vivo determinations of nerve
    conduction velocity in rat tails: physiological and pathological
    changes. Electroencephal. clin. Neurophysiol., 35: 125-131.

    MOLINENGO, L. & ORSETTI, M. (1976) Drug action on the "grasping"
    reflex and on swimming endurance: an attempt to characterize
    experimentally anti-depressant drugs. Neuropharmacology,
    15: 257-260.

    MORALI, G., LARSSON, K., & BEYER, C. (1977) Inhibition of
    testosterone-induced sexual behaviour in the castrated male rat by
    aromatization blockers. Horm. Behav., 9: 203-213.

    MORGANTI, J.B., LOWNE, B.A., SALVATERRA, P., & MASSARO, E.J. (1976)
    Effects of open-field behaviour of mice exposed to multiple doses of
    methylmercury. Gen. Pharmacol., 7(1): 41-44.

    MUNRO, A.J., JACKSON, R.J., & KORNER, A. (1964) Disaggregation of
    brain polysomes. Biochem. J., 92: 289-294.

    MURPHY, S.D. & DUBOIS, K.P. (1958) The influence of various factors on
    the enzymatic conversion of organic thiophosphates to
    anticholinesterase agents. J. Pharmacol. exp. Ther., 124: 194-202.

    NARAHASHI, T. (1982) Cellular and molecular mechanisms of action of
    insecticides: neurophysiological approach. Neurobehav. Toxicol.
    Teratol., 4: 753-758.

    NARAHASHI, T. & HAAS, H.G. (1967) DDT: interaction with nerve membrane
    conductance changes. Science, 157: 1438-1440.

    NARAHASHI, T. & HAAS, H.G. (1968) Interaction with DDT with the
    components of lobster nerve membrane conductance. J. gen. Physiol.,
    51: 177-198.

    NATHANSON, J.A. & BLOOM, F.E. (1975) Lead-induced inhibition of brain
    adenylcyclase. Nature (Lond.), 255: 419-420.

    NICHOLAS, G.S. & BARRON, D.H. (1932) The use of sodium amytal in the
    production of anaesthesia in the rat. J. Pharmacol. exp. Ther., 46;
    125-129.

    NIEDERMEYER, E. & LOPES DA SILVA, F. (1982) Electroencephalography:
    basic principles, clinical applications, and related fields, 
    Baltimore, Maryland, Munich, Urban Schwartzenberg.

    NIOKESARYISKY, A.A. & TUROVSKY, V.S. (1980) N-acetylneuraminic acid
    content in water-soluble and membranous tissue fractions of various
    analysers of cerebral cortex. Biochimia, 45: 1829-1932.

    NORTON, S. (1973) Amphetamine as a model for hyperactivity in the rat.
    Physiol. Behav., 11: 181-186.

    O'BRIEN, R.D. (1976) Acetylcholinesterase and its inhibition. In:
    Wilkerson, C.F., ed. Insecticide biochemistry and physiology, New
    York, Plenum Press, pp. 271-296.

    O'CALLAGHAN, J.P., LAVIN, K.L., CHESS, Q., & CLOUET, D.H. (1983) A
    method for dissection of discrete regions of rat brain following
    microwave irradiation. Brain Res., 11: 31-42.

    O'DONOGHUE, J.L., ed. (1985) Neurotoxicity of industrial and
    commercial chemicals, Boca Raton, Florida, CRC Press.

    OCHOA, J. (1972) Ultrathin longitudinal sections of single myelinated
    fibres for electron microscopy. J. neurol. Sci., 17: 103.

    OCHS, S. (1972) Fast transport of materials in mammalian nerve fibres.
    Science, 176: 252.

    OCHS, S. (1982) Axoplasmic transport and its relation to other nerve
    functions, New York, John Wiley and Sons.

    OHI, G., NISHIGAKI, S., SEKI, H., TAMURA, Y., MIZOGUCHI, I., YAGYA,
    H., & NAGASHIMA, K. (1978) Tail rotation, an early neurological sign
    of methylmercury poisoning in the rat. Environ. Res., 16: 353-359.

    OLIVER, C., ESKAY, R.L., BEN-JONATHAN, N., & PORTER, J.C. (1974)
    Distribution and concentration of TRH in rat brain. Endocrinology,
    96: 540-546.

    OLTON, D.S., BECKER, J.T., & HANDELMANN, G.E. (1979) Hippocampus
    space, and memory. Behav. Brain Sci., 2: 313-365.

    OLTON, D.S., BECKER, J.T., & HANDELMANN, G.E. (1980) Hippocampal
    function: working memory or cognitive mapping? Physiol. Psychol.,
    8: 239-246.

    OTTO, D.A. (1977) Neurobehavioral toxicology: problems and methods in
    human research. In: Zenick, H. & Reiter, L.W., ed. Behavioral
    toxicology: an emerging discipline, Research Triangle Park, North
    Carolina, US Environmental Protection Agency, pp. 10/1-10/31
    (EPA 600/9-77-042).

    OTTO, D.A. (1978) Multidisciplinary perspectives in event-related
    brain potential research. In: Proceedings of the Fourth International
    Congress on Event-Related Slow Potentials of the Brain, Washington
    DC, US Environmental Protection Agency (EPA 600/9-77-043).

    OTTO, D.A., BENIGNUS, V.A., MULLER, K.E., & BARTON, C.N. (1981)
    Effects of age and body lead burden on CNS function in young children.
    I. Slow cortical potentials. Electroencephalogr. clin.
    Neurophysiol., 52: 229-239.

    OVERMANN, S.R. (1977) Behavioral effects of asymptomatic lead exposure
    during neonatal development in rats. Toxicol. appl. Pharmacol.,
    41: 459-471.

    PADICH, R. & ZENICK, M. (1977) The effects of developmental and/or
    direct lead exposure on FR behavior in the rat. Pharmacol. Biochem.
    Behav., 6: 371-375.

    PALKOVITS, M. (1973) Isolated removal of hypothalamic or other brain
    nuclei of the rat. Brain Res., 59: 449-450.

    PALMITER, R.D. & MULVIHILL, E.D. (1978) Estrogenic activity of the
    insecticide kepone on the chick oviduct. Science, 201: 356-358.

    PANCHENKO, L.F., DUDCHENKO, A.M., & SHPAKOV, A.A. (1975) The influence
    of acute hypoxic hypoxia on the activity of rats cerebral cortex:
    mitochondria enzymes possessing different sensibility to oxygen
    deficiency. In: Gorky, ed. Biochemistry of hypoxia, pp. 59-64.

    PANSKY, B. & ALLEN, D.J. (1980) Review of neuroscience, New York,
    McMillan Inc.

    PARIZEK, J. (1956) Effects of cadmium salts on testicular tissue.
    Nature (Lond.), 177: 1036-1037.

    PARTINGRON, C.R. & DALY, J.W. (1979) Effect of gangliosides on
    adenylate cyclase activity in rat cerebral cortical membranes. Mol.
    Pharmacol., 15: 484-491.

    PARTRIDGE, W.M. (1976) Inorganic mercury: selective effects on
    blood-brain barrier transport systems. J. Neurochem., 27: 333-335.

    PATTON, S.E., O'CALLAGHAN, J.P., MILLER, D.B., & ABOU-DONIA, M.B.
    (1983) Effect of oral adminstration of tri-o-cresyl phosphate on in
    vitro phosphorylation of membrane and cytosolic proteins from
    chicken brain. J. Neurochem., 41: 897-901.

    PAUL, S.M., HOFFMAN, A.R., & AXELROD, A. (1980) Catechol estrogens:
    synthesis and metabolism in brain and other endocrine tissues. In:
    Martini, L. & Ganong, W.F., ed. Frontiers in neuroendocrinology, New
    York, Raven Press, pp. 203-217.

    PAVLENKO, S.M. (1975) Methods for the study of the central nervous
    system in toxicological tests. In: Methods used in the USSR for
    establishing biologically safe levels of toxic substances, Geneva,
    World Health Organization, pp. 86-108.

    PAVLOV, I.P (1927) Conditioned reflexes, Oxford, Oxford University
    Press.

    PEARSE, A.G.E. (1968) Histochemistry: theoretical and applied, 3rd
    ed., Baltimore, Maryland, Williams and Wilkins Company.

    PEARSON, C.M. & MOSTOFI, F.K. (1973) The striated muscle, Baltimore,
    Maryland, Williams and Wilkins Company.

    PEASE, D.E. (1964) Histological techniques for electron microscopy,
    2nd ed., New York, Academic Press.

    PERSINGER, M.A., LAFREINIERE, G.P., & CARREY, N.V. (1978) Thyroid
    morphology and wet organ weight changes in rats exposed to different
    low intensity 0.5-Hz magnetic fields and pre-experimental caging
    conditions. Biometeorology, 22: 67-73.

    PILCHER, C.W.T. & JONES, S.M. (1981) Social crowding enhances
    aversiveness of naloxone in rats. Pharmacol. Biochem. Behav.,
    14: 299-303.

    PIRCH, J.H. & RECH, R.H. (1968) Effect of isolation on alpha-
    methyltyrosine-induced behavioral depression. Life Sci., 7: 173-182.

    PLEASURE, D.E., MISHLER, K.C., & ENGEL, W.K. (1969) Axonal transport
    of proteins in experimental neuropathies. Science, 166: 524-525.

    POST, E.M., YANG, M.G., KING, J.A., & SANGER, V.L. (1973) Behavioral
    changes of young rats force-fed methylmercury chloride. Proc. Soc.
    Exp. Biol. Med., 143: 1113-1116.

    POWERS, J.M., SCHAUMBURG, H.H., JOHNSON, A.E., & RAINE, C.S. (1980) A
    correlative study of the adrenal cortex in adrenoleukodystrophy:
    evidence for a fatal intoxication with very long-chain saturated fatty
    acids. Invest. cell Pathol., 3: 353-376.

    PRABHU, V.G. (1972) Amphetamine aggregation effect in mice under
    conditions of altered microsomal enzymes. Arch. int. Pharmacodyn.
    Ther., 195: 81-86.

    PROKHOROVA, M.J. (1974) Structure and function of phosphornositides
    and gangliosides of the brain. In: Progress of neurochemistry,
    Leningrad, Nauka, pp. 61-73.

    PRYOR, G.T., UYENO, E.T., TILSON, H.A., & MITCHELL, C.L. (1983)
    Assessment of chemicals using a battery of neurobehavioral tests: a
    comparative study. Neurobehav. Toxicol. Teratol., 5: 91-117.

    PURPURA, D.P., PENRY, J.K, TOWER, P., WOODBURY, D.M., & WALTER, R.
    (1972) Experimental models of epilepsy, New York, Raven Press.

    QUEVEDO, L., CONCHA, J., MARCHANT, Z., ESKUCHE, W., & MADARIAGA, J. 
    (1980) Nerve accommodation in chronic alcoholic subjects.
    Pharmacology, 21: 229-232.

    QUINN, G.P., AXELROD, J., & BRODIE, B.B. (1958) Species, strain, and
    sex differences in metabolism of hexobarbitone, amidopyrine,
    antipyrine, and aniline. Biochem. Pharmacol., 1: 152-159.

    RADIN, N.S., BRENHERT, A., ARORA, R.C., SELLINGER, O.Z., & FLANGAS,
    A.L. (1972) Glial and neuronal localization of cerebroside-
    metabolizing enzymes. Brain Res., 39: 163-166.

    RAFALES, L.S., BORNSCHEIN, R.L., MICHAELSON, A., LOCH, R.K., & BARKER,
    G.F. (1979) Drug-induced activity in lead-exposed mice. Pharmacol.
    Biochem. Behav., 10: 95-104.

    RAMSAY, R.B., KRIGMAN, M.R., & MORELL, P. (1980) Developmental studies
    of the uptake of choline, GABA, and dopamine by crude synaptosomal
    preparations after in vivo or in vitro lead treatment. Brain
    Res., 187: 383-387.

    RAY, O.S. & BARRETT, R.J. (1975) Behavioral, pharmacological, and
    biochemical analysis of genetic differences in rats. Behav. Biol.,
    15: 391-417.

    REICHARDT, L.F. & KELLY, R.B. (1983) A molecular description of nerve
    terminal function. Ann. Rev. Biochem., 52: 871-926.

    REICHERT, B.L. & ABOU-DONIA, M.B. (1980) Inhibition of fast axoplasmic
    transport by delayed neurotoxic organophosphorus esters: a possible
    mode of action. Mol. Pharmacol., 17: 56-60.

    REITER, L.W. (1978) Use of activity measures in behavioral toxicology.
    Environ. Health Perspect., 26: 9-20.

    REITER, L.W. (1980) Neurotoxicology -- meet the real world.
    Neurobehav. Toxicol., 2: 73-74.

    REITER, L.W. (1983) Chemical exposure and animal activity: utility of
    the Figure-Eight Maze. In: Hayes, A.H., Schnell, R.C., & Miya, T.S.,
    ed. Developments in the science and practice of toxicology,
    Amsterdam, Oxford, New York, Elsevier Science Publishers, pp. 78-84.

    REITER, L.W. & MACPHAIL, R.C. (1979) Motor activity: a survey of
    methods with potential use in toxicity testing. Neurobehav.
    Toxicol., 1(Suppl. 1): 53-66.

    REITER, L.W. & MACPHAIL, R.C. (1982) Factors influencing motor
    activity measurements in neurotoxicology. In: Mitchell, C.L., ed.
    Nervous system toxicology, New York, Raven Press, pp. 45-65.

    REITER, L.W., KIDD, K., HEAVNER, G., & RUPPERT, P. (1980) Behavioral
    toxicity of acute and subacute exposure to triethyltin in the rat.
    Neurotoxicology, 2: 97-112.

    REUHL, K. & CHANG, L. (1979) Effects of methylmercury on the
    development of the nervous system: a review. Neurotoxicology,
    1: 21-55.

    REYNOLDS, G.S. (1958) A primer of operant conditioning, Glenview,
    Illinois, Scott, Foresman and Co.

    RICE, D.C. & GILBERT, S.G. (1982) Early chronic low-level
    methylmercury poisoning in monkeys impairs spatial vision. Science,
    216: 759-761.

    RICE, D.C. & WILLES, R.F. (1979) Neonatal low-level lead exposure in
    monkeys (Macaca fascicularis): effects on two choice non-spatial
    form discrimination. J. environ. Pathol. Toxicol., 2: 1195-1203.

    RICE, D.C., GILBERT, S.G., & WILLES, R.F. (1979) Neonatal low-level
    lead exposure in monkeys: locomotor activity, schedule-control
    behavior, and the affects of amphetamine. Toxicol. appl. Pharmacol.,
    51: 503-513.

    RICE, D.F. (1985) Central nervous system effects of perinatal exposure
    to lead or methylmercury in the monkey. In: Clarkson, T.W. & Nordberg,
    G., ed. Reproductive developmental toxicity of metals, New York,
    Plenum Press.

    RITCHIE, J.M. (1979) A pharmacological approach to the structure of
    sodium channels in myelinated axons. Ann. Rev. Neurosci.,
    2: 341-362.

    ROBERTS, D.V. (1977) A longitudinal electromyographic study of six men
    occupationally exposed to organophosphorus compounds. Int. Arch.
    occup. environ. Health, 38: 221-229.

    ROBERTS, R.K., DESMOND, P.V., WILKINSON, G.R., & SCHENKER, S. (1979)
    Disposition of chlordiazepoxide; sex differences and effects of oral
    contraceptives. Clin. Pharmacol. Ther., 25: 826-831.

    ROBINS, E., HIRSCH, H.E., & EMMONS, S.S. (1968) Glycosidases in the
    nervous system. I. Assay, some properties, and distribution of
    beta-galactosidase, beta-glucoronidase, and beta-glucosidase.
    J. biol. Chem., 243: 4246-4249.

    RODIER, P. (1978) Behavioral teratology. In: Wilson, J.G. & Fraser,
    F.C., ed. Handbook of teratology, New York, Plenum Press, Vol. 4,
    pp. 397-428.

    ROEDER, R.J. & RUTTER, W.J. (1969) Multiple forms of DNA-dependent RNA
    polymerase in eukaryotic organisms. Nature (Lond.), 224: 234-236.

    ROSE, S.P.R. & SINHA, A.K. (1970) Separation of neuronal and neuropil
    cell fractions: a modified procedure. Life Sci., 9: 907-912.

    ROSECRANS, J.A., HONG, J.S., SQUIBB, R.E., JOHNSON, J.H., WILSON,
    W.E., & TILSON, H.A. (1982) Effects of perinatal exposure to
    chlordecone (kepone) on neuroendocrine and neurochemical
    responsiveness of rats to environmental challenges. Neurotoxicology,
    3: 131-142.

    ROSEWICKYI, S., STANOSZ, S., & SAMOCHOWIEC, L. (1973) [The effect of
    protracted exposure to carbon disulfide on the hormonal activity of
    the ovaries in rats.] Med. Pracy., 24: 133-136 (in Polish).

    ROZIER, L., SHIRAKI, H., & GRCEVIC, N., ed. (1977) Neurotoxicology,
    New York, Raven Press, Vol. 1, 658 pp.

    RUDNEV, M., BOKINA, A., ESKLER, N., & NAVAKATIKYAN, M. (1978) The use
    of evoked potential and behavioural measures in the assessment of
    environmental insult. In: Proceedings of the Fourth International
    Congress on Event-Related Slow Potentials of the Brain, Henderson,
    North Carolina, 4-10 April, 1976 (EPIC IV), Washington DC, US
    Environmental Protection Agency, pp. 444-447 (EPA 600/9-77-0443).

    RUSSELL, R.W., WARBURTON, D.M., & SEGAL, D.S. (1969) Behavioral
    tolerance during chronic changes in the cholinergic system. Comm.
    Behav. Biol., 4: 121-128.

    RUSAK, B. & ZUCKER, I. (1975) Biological rhythms and animal behavior.
    Ann. Rev. Psychol., 27: 137-171.

    SABRI, M.I. & SPENCER, P.S. (1980) Toxic distal axonopathy:
    biochemical studies and hypothetical mechanisms. In: Spencer, P.S. &
    Schaumburg, H.H., ed. Experimental and clinical neurotoxicology,
    Baltimore, Maryland, Williams and Wilkins Company, pp. 206-219.

    SABRI, M.I., MOORE, C.L., & SPENCER, P.S. (1979) Studies on the
    biochemical basis of distal axonopathies. Inhibition of glycolysis
    produced by neurotoxic hexacarbon compounds. J. Neurochem.,
    32: 683-690.

    SAKMANN, B. & NEHER, E. (1983) Single channel recording, New York,
    Plenum Press, 503 pp.

    SALVATERRA, P., LOWN, B., MORGANTI, J., & MARSARO, E.J. (1973) 
    Alterations in neurochemical and behavioural parameters in the mouse
    induced by low doses of methylmercury. Acta pharmacol. toxicol.,
    33(3): 177-190.

    SANTINI, M. (1975) Golgi Centennial Symposium: Perspectives in
    Neurobiology, New York, Raven Press.

    SAVOLAINEN, H. (1983) Trends and prospects in experimental
    neurotoxicology. Scand. J. Work environ. Health, 9: 214-218.

    SAVOLAINEN, H. & JARVISALO, J. (1977) Effects of acute CS2
    intoxication on protein metabolism in rat brain. Chem.-biol.
    Interact., 17: 51-59.

    SAVOLAINEN, H., PFAFFLI, P., TENGEN, M., & VAINO, H. (1977)
    Biochemical and behavioural effects of inhalation exposure to
    tetrachloroethylene and dichloromethane. J. Neuropathol. exp.
    Neurol., 36: 941-949.

    SCHALLERT, T., WHISHAW, I.Q., RAMIREZ, V.D., & TEITELBAUM, P. (1978)
    Compulsive, abnormal walking caused by anticholinergies in akinetic, 6
    hydroxydopamine-treated rats. Science, 199: 1461-1463.

    SCHAUMBURG, H.H. (1982) Pyridoxine megavitaminosis produces sensory
    neuropathy in humans. Ann. Neurol., 12: 107-108.

    SCHAUMBURG, H.H., SPENCER, P.S., & THOMAS, P.K. (1983) Disorders of
    peripheral nerves, Philadelphia, Pennsylvania, F.A. Davis.

    SCHAUMBURG, H.H., WISNIEWSKI, H., & SPENCER, P.S. (1974)
    Ultrastructural studies of the dying-back process. I. Peripheral nerve
    terminal and axon degeneration in systemic acrylamide intoxication.
    J. Neuropathol. exp. Neurol., 33: 260-284.

    SCHECHTER, M.D. & WINTER, J.C. (1971) Effect of mescaline and lysergic
    acid diethylamide on flicker discrimination in the rat. J. Pharmacol.
    exp. Ther., 177: 461-467

    SCHEUHAMMER, A.M. & CHERIAN, M.G. (1982) The regional distribution of
    lead in normal rat brain. Neurotoxicology, 3: 85-92.

    SCHEVING, L.E., HALBERG, F., & PAULY, J.E., ed. (1974)
    Chronobiology, Tokyo, Igaku Shoin Ltd.

    SCHNAITMAN, C., ERWIN, V.G., & GREENSWALT, J.W. (1967) The
    submitochondrial localization of monoamine oxidase. An enzymatic
    marker for the outer membrane of rat liver mitochondria. J. cell
    Biol., 32: 719-720.

    SCHOENFELD, W.N., ed., (1970) The theory of reinforcement schedules,
    New York, Appleton-Century-Crofts.

    SCHOTMAN, P., GIPON, L., JENNEKENS, F.G.I., & GISPEN, W.H. (1978)
    Polyneuropathies and CNS protein metabolism. III. Changes in protein
    synthesis induced by acrylamide intoxication. J. Neuropathol. exp.
    Neurol., 37: 820-837.

    SCHREIBER, V. & PRIBYL, T. (1980) Possible role of ceruloplasmin in
    endocrine regulation. In: Ish, S., Hirano, T., & Wade, M., ed.
    Hormones, adaptation, and evolution, Tokyo, Japan Science Society
    Press/Berlin, Springer-Verlag.

    SCHREIBER, V., PRIBYL, T., & JAHODOVA, J. (1982) Inhibition by an
    ergoline derivative (dopaminergic agonist) of estradiol-induced
    adenohypophyseal growth and of the decrease in the hypothalamic
    ascorbic acid (HA) concentration in rats. Physiol. Bohemoslov.,
    31: 97-100.

    SCHREIBER, V., PRIBYL, T., & JAHODOVA, J. (1979) Effect of a dopamine
    p-hydroxylase inhibitor (disulfiram) on the response of
    adenohypophysis, serum ceruloplasmin, and hypothalamic ascorbic acid
    to estradiol treatment. Endoc. Exper. (Bratislava), 13: 131-138.

    SCHREIBER, V., ZBUZEK, V., & ZBUZKOVA-KMENTOVA, V. (1969) Reactions of
    the rat and hamster endocrine system to aminoglutethimide. Physiol.
    Bohemoslov., 18: 249-261.

    SEPPALAINEN, A.M. & LINNOILA, I. (1975) Electrophysiological studies
    on rabbits in long-term exposure to carbon disulphide. Scand. J. Work
    environ. Health, 1: 178-183.

    SEPPALAINEN, A.M., TOLONEN, M., KARLI, P., HANNINEN, H., & HEINBERG,
    S. (1972) Neurophysiological findings in chronic carbon disulfide
    poisoning: a descriptive study. Work Environ. Health, 9: 71-75.

    SETH, P.K., AGRAWAL, A.K., & BONDY, S.C. (1981) Biochemical changes in
    the brain consequent to dietary exposure of developing and mature rats
    to chlordecone (kepone). Toxicol. appl. Pharmacol., 26: 262-267.

    SHANDALA, M.G., RUDNEV, M.I., & NAVAKATIKYAN, M.A. (1980) [Behavioural
    reactions in experimental hygienic research.] Gig. i Sanit.,
    6: 43-48 (in Russian).

    SHAW, C.M., MOTTET, N.K., & CHEN, W.J. (1980) Effects of methylmercury
    on the visual system of rhesus macaque (Macaca mulatta). II.
    Neuropathological findings (with emphasis on vascular lesions in the
    brain). In: Merigan, W.H. & Weiss, B., ed. Neurotoxicity of the
    visual system, New York, Raven Press, pp. 123-134.

    SHAW, C.M., MOTTET, N.K., BODY, R.L., & LUSCHEI, E.S. (1975)
    Variability of neuropathological lesions in experimental methylmercury
    encephalopathy in primates. Am. J. Pathol., 80: 451-470.

    SHIBUSAWA, K., YAMAMOTO, T., NISHI, K., ABE, C., & TOMIE, S. (1959)
    Urinary secretion of TRF in various functional states of the thyroid.
    Endocrinol. Jpn., 6: 131-136.

    SHILJAGINA, N.N. (1971) [Evoked potentials in the projection of sign
    stimulus during conditional reflex formation in ontogenesis.]
    Zh. Vyssh. nerv. deyat. imeni, 21: 492-500 (in Russian).

    SHIMIZU, T., TAKAISHI, M., & SHISHIBE, Y. (1978) Effect of ethanol on
    the peripheral metabolism of thyroxine. Endocrinol. Jpn,
    25: 201-204.

    SHUGALOV, N.P. (1969) Primary and associative evoked responses to
    optic irritants during various stages of conditioned reflex formation
    in cats. In: Summary of Reports of the 22nd Meeting on the Problems
    of Higher Nervous Activity, Ryazan, USSR, p. 268.

    SHUMILINA, A.I. (1971) Characteristics of summary electric activity
    and evoked potentials under motivating stimulation of various
    biological significance. In: Proceedings of the Sixth All-Union
    Conference on Electrophysiology of the Central Nervous System,
    pp. 301-302.

    SILBERGELD, E.K. & GOLDBERG, A.M. (1974a) Hyperactivity: a
    lead-induced behavior disorder. Environ. Health Perspect.,
    7: 227-232.

    SILBERGELD, E.K. & GOLDBERG, A.M. (1974b) Lead-induced behavioral
    dysfunction: an animal model of hyperactivity. Exp. Neurol.,
    42: 146-157.

    SILBERGELD, E.K. & GOLDBERG, A.M. (1975) Pharmacological and
    neurochemical investigations of lead-induced hyperactivity.
    Neuropharmacology, 14: 431-444.

    SILBERGELD, E.K., HRUSKA, R.E., MILLER, L.P., & ENG, N. (1980) Effects
    of lead in vivo and in vitro on GABA energic neurochemistry.
    J. Neurochem., 34: 1711-1718.

    SILVERMAN, A.P. (1965) Ethological and statistical analysis of drug
    effects on the social behaviour of laboratory rats. Br. J.
    Pharmacol., 24: 579-590.

    SLOANE, S.A., SHEA, S.L., PROCTER, M.M., & DEWSBURY, D. (1978) Visual
    cliff performance in 10 species of muroid rodents. Anim. Learn.
    Behav., 6: 244-248.

    SNOWDON, C.T. (1973) Learning deficits in lead injected rats.
    Pharmacol. Biochem. Behav., 1: 599-603.

    SNYDER, D.R. & BRAUN, J.J. (1977) Dissociation between behavioral and
    physiological indices of organomercurial ingestion. Toxicol. appl.
    Pharmacol., 41: 277-284.

    SNYDER, S.H. & INMS, R.B. (1979) Peptide neurotransmitters. Ann. Rev.
    Biochem., 48: 755-782.

    SOBOTKA, T.J., BRODIE, R.E., & COOK, M.P. (1975) Psychophysiologic
    effects of early lead exposure. Toxicology, 5: 175-191.

    SOKOLOFF, L., REIVICH, M., KENNEDY, C., DES ROSIERS, M.H., PATLAK,
    C.S., PETTIGREW, K.D., SAKURADA, O., & SHINOHARA, M. (1977) The
    14C-deoxyglucose method for the measurement of local cerebral glucose
    utilization: theory, procedure, and normal values in the conscious and
    anaesthetized albino rat. J. Neurochem., 28: 897-916.

    SPENCER, P.S. & BISCHOFF, M.C. (1982) Contemporary neuropathological
    methods in toxicology. In: Mitchell, C.L., ed. Nervous system
    toxicology, New York, Raven Press, pp. 259-275.

    SPENCER, P.S. & LIEBERMAN, A.R. (1971) Scanning electron microscopy of
    isolated peripheral nerve fibres. Normal surface structure and
    alterations proximal to neuromas. Z. Zellforsch. Mikrosk. Anat.,
    119: 534-551.

    SPENCER, P.S. & SCHAUMBURG, H.H. (1980) Experimental and clinical
    neurotoxicology, Baltimore, Maryland, Williams and Wilkins Company,
    929 pp.

    SPENCER, P.S. & THOMAS, P.K. (1970) The examination of isolated nerve
    fibres by light and electron microscopy, with observations on
    demyelination proximal to neuromas. Acta neuropathol., 16: 177-186.

    SPENCER, P.S., BISCHOFF, M., & SCHAUMBURG, H.H. (1980)
    Neuropathological methods for the detection of neurotoxic disease. In:
    Spencer, P.S. & Schaumburg, H.H., ed. Experimental and clinical
    neurotoxicology, Baltimore, Maryland, Williams and Wilkins Company,
    pp. 743-757.

    SPERANSKIJ, S.V. (1965) [Advantages incident to the use of mounting
    current in analysing the ability of Albino mice to summate subliminal
    impulses.] Farmakol. i Toksikol., 1: 123-124 (in Russian).

    SPYKER, J. (1975) Behavioral teratology and toxicology. In: Weiss, B.
    & Laties, V.G., ed. Behavioral toxicology, New York, Plenum Press,
    pp. 311-349.

    SQUIBB, R.E. & TILSON, H.A. (1982) Neurobehavioural changes in adult
    Fisher 344 rats exposed to dietary levels of chlordecone (KeponeR):
    a 90-day chronic dosing study. Neurotoxicology, 3: 59-66.

    STEBBINS, W.C. (1970) Animal psychophysics: the design and conduct of
    animal experiments, New York, Appleton-Century-Crofts.

    STEEL, R.G.D. & TORRIE, J.H. (1960) Principles and procedures of
    statistics, New York, McGraw-Hill, pp. 83-347.

    STEELE, W.J. & BUSCH, H. (1963) Studies on acidic nuclear proteins of
    the Walker tumour and liver. Cancer Res., 23: 1153-1157.

    STOCKINGER, H.E. (1974) Behavioral toxicology in the development of
    threshold limit values. In: Xintaras, C., Johnson, B.L., & DeGroot,
    I., ed. Behavioral toxicology, Cincinnati, Ohio, US Department of
    Health, Education and Welfare, pp. 18-19 (US DHEW (NIOSH) Publication
    No. 74-126).

    SUDAKOV, K.V. (1982) Functional system theory in physiology.
    Agressologie, 23: 167-176.

    SZMIGIELSKI, A. (1975) Effect of disulfiram and sodium
    diethyldithiocarbamate on thyroxine and dopamine hydroxylases in rat
    brain and breast. Pol. J. Pharmacol. Pharm., 27: 153-159.

    TAKEUCHI, T. (1968) Pathology of Minamata disease. In: Katsuma, M.,
    ed. Minamata disease. Study Group of Minamata Disease, Japan,
    Kumamoto University, pp. 141-228.

    TAKEUCHI, T., ETO, N., & ETO, K. (1979) Neuropathology of childhood
    cases of methylmercury poisoning (Minamata disease) with prolonged
    symptoms with particular reference to the decortication syndrome.
    Neurotoxicology, 1: 1-20.

    TALKE, T., HIRABAGASHI, J., KONDO, R., MATSUMOTO, M., & KOJIMA, K.
    (1979) Effect of butyrate on glycolipid metabolism of two cell types
    of rat ascites hepatomes with different ganglioside biosynthesis.
    J. Biochem., 96: 1139-1402.

    TAPPEL, A.L. (1970) Lipid peroxidation damage to cell components.
    Fed. Proc., 29: 239.

    TAYLOR, J.R., SELHORST, J.B., & CALABRESE, V.P. (1980) Chlordecone.
    In: Spencer, P. & Schaumburg, H.H., ed. Experimental and clinical
    neurotoxicology, Baltimore, Maryland, Williams and Wilkins Company,
    pp. 407-421.

    THIESSEN, D.D. (1964) Population density and behavior: a review of
    theoretical and physiological contributions. Texas Rep. Biol. Med.,
    22: 266-314.

    THOMAS, P.K. (1970) The quantitation of nerve biopsy findings.
    J. neurol. Sci., 11: 285-295.

    THOMAS, P.K. (1980) The peripheral nervous system as a target for
    toxic substances. In: Spencer, P.S. & Schaumburg, H.H., ed.
    Experimental and clinical neurotoxicity, Baltimore, Maryland,
    Williams and Wilkins Company, pp. 35-47.

    THOMPSON, S.W. (1963) Selected histochemical and histopathological
    methods, Springfield, Illinois, Charles C. Thomas.

    THORNDIKE, E.L. (1932) Reward and punishment in animal learning.
    Comp. Psychol. Monogr., 39: 8.

    TILSON, H.A. & CABE, P.A. (1978) Strategy for the assessment of
    neurobehavioral consequences of environmental factors. Environ.
    Health Perspect., 26: 287-299.

    TILSON, H.A. & HARRY, G.J. (1982) Behavioral principles for use in
    behavioral toxicology and pharmacology. In: Mitchell, C.L., ed.
    Nervous system toxicology, New York, Raven Press, pp. 1-27.

    TILSON, H.A. & MITCHELL, C.L. (1984) Neurobehavioral techniques to
    assess the effects of chemicals on the nervous system. Ann. Rev.
    Pharmacol. Toxicol., 24: 425-450.

    TILSON, H.A. & SQUIBB, R.D. (1982) The effects of acrylamide on the
    behavioral suppression produced by psychoactive agents.
    Neurotoxicology, 3: 113-120.

    TILSON, H.A., BYRD, N., & RILEY, M. (1979a) Neurobehavioral effects of
    exposing rats to kepone via the diet. Environ. Health Perspect.,
    33: 321.

    TILSON, H.A., CABE, P.A., ELLINWOOD, E.H., Jr, & GONZALEZ, L.P.
    (1979b) Effects of carbon disulfide on motor function and
    responsiveness to delta-amphetamine in rats. Neurobehav. Toxicol.,
    1: 57-63.

    TILSON, H.A., MITCHELL, C.L., & CABE, P.A. (1979c) Screening for
    neurobehavioral toxicity: the need for and examples of validation of
    testing procedures. Neurobehav. Toxicol., 1(Suppl. 1): 137-148.

    TILSON, H.A., CABE, P.A., & BURNE, T.A. (1980) Behavioral procedures
    used in the assessment of neurotoxicity. In: Spencer, P.S. &
    Schaumburg, H.H., ed. Experimental and clinical neurotoxicology,
    Baltimore, Maryland, Williams and Wilkins Company, pp. 758-766.

    TILSON, H.A., MACTUTUS, C.F., MCLAMB, R.L., & BURNE, T.A. (1982a)
    Characterization of triethyl lead chloride neurotoxicity in adult
    rats. Neurobehav. Toxicol. Teratol., 4: 671-681.

    TILSON, H.A., SQUIBB, R.E., & BURNE, T.A. (1982b) Neurobehavioral
    effects following a single dose of chlordecone (kepone) administered
    neonatally to rats. Neurotoxicology, 3: 45-52.

    TOLMAN, E.C. (1948) Cognitive maps in rats and men. Psychol. Rev.,
    55: 189-208.

    TOWFIGHI, J., GONATAS, N.K., & MCCREE, L. (1975) Hexachlorophene
    retinopathy in rats. Lab. Invest., 32: 330-338.

    TROPKINA, J.K., SELZNEVA, S.N., & POGODAEV, K.V. (1981) Prenatal
    influence of hypoxia on metabolic structure of cerebral mitochondria
    in the oncogenesis with epileptic superactivity. In: Mitochondria:
    mechanisms of conjugation and regulation, Puscino, p. 39.

    TRUEX, R.C. & CARPENTER, M.B. (1983) Human neuroanatomy, 8th ed.,
    Baltimore, Maryland, Williams and Wilkins Company.

    TSUJI, S. & FUJIHSIMA H. (1975) Paraplegias: clinical statistics,
    Fukuoka, Kyushu Rosai Hospital, Department of Orthopedics and
    Neurology.

    TULLOCH, M.L., CROOKS, J., & BROWN, P.S. (1963]) Inhibition of thyroid
    function by a dithiocarbamoylhydrazine. Nature (Lond.),
    199: 288-289.

    ULRICH, C.E., RINEHART, W., & BRANDT, M. (1979) Evaluation of the
    chronic inhalation toxicity of a manganese oxide aerosol. III.
    Pulmonary function, electromyograms, limb tremor, and tissue manganese
    data. Am. Ind. Hyg. Assoc. J., 40: 349-353.

    UPHOUSE, L., NEMEROFF, C.B., MASON, G., PRANGE, A.J., & BONDY, S.C.
    (1982) Interactions between "handling" and acrylamide on endocrine
    responses in rats. Neurotoxicology, 3: 121-125.

    UPHOUSE, L., MASON, G., & HUNTER, V. (1984) Persistent vaginal estrus
    and serum hormones after chlordecone (kepone) treatment of adult rats.
    Toxicol. appl. Pharmacol., 72: 177-186.

    US FDA (1978) Nonclinical laboratory studies. Good laboratory
    practice regulation, Washington DC, US Food and Drug Administration,
    Department of Health, Education and Welfare (Federal Register 43:
    59986-60025).

    US NAS (1975) Principles for evaluating chemicals in the
    environment, Washington DC, National Academy of Science (A Report of
    the Committee for the Working Conference on Principles of Protocols
    for Evaluating Chemicals in the Environment).

    US NAS/NRC (1970) Evaluating the safety of food chemicals,
    Washington DC, National Academy of Sciences, National Research
    Council, Food Protection Committee, Food Nutrition Branch, Division of
    Biology and Agriculture.

    VAN GELDER, G.A. (1978) Lead and the nervous system. In: Oehme, F.W.,
    ed. Toxicity of heavy metals in the environment, Part 1, Basel, New
    York, Marcel Dekker, Inc., pp. 101-121.

    VANDERWOLF, C.H. & LEUNG, L.-W.S. (1983) Hippocampal rhythmic slow
    activity: a brief history and the effects of entorhinal lesions and
    phencyclidine. In: Seifert, W., ed. Neurobiology of the hippocampus,
    New York, Academic Press, pp. 275-302.

    VASAN, N.S. & CHASE, H.P. (1976) Brain glucosaminoglycans
    (mucopolysaccharides) following intrauterine growth retardation.
    Biol. Neonate, 28: 196-206.

    VEKSLER, M.S. & ARBUKHANOVA, M.S. (1981) Cerebral and spinal cord
    mitochondria under strenuous physical work. In: Mitochondria:
    mechanism of conjugation and regulation, Puscino, p. 81.

    VERGIEVA, T. & ZAIKOV, H.R. (1981) Behavioral changes in rats with
    inhalation of styrene. Hig. Zdraveopazvane, 24: 242-247.

    VERITY, M.A. (1972) Cation modulation of synaptosomal respiration.
    J. Neurochem., 19: 1305-1309.

    VERITY, M.A., BROWN, W.J., CHEUNG, M., & DZER, G. (1977) Methylmercury
    inhibition of synaptosomes and brain slices protein synthesis: in
    vivo and in vitro studies. J. Neurochem., 29: 673-679.

    VILA, J. & COLOTLA, V.A. (1981) Some stimulus properties of inhalants:
    preliminary findings. Neurobehav. Toxicol., 3: 447-480.

    VINAR, O., FRANTIK, E., & HORVATH, M. (1981) Subjectively-perceived
    effects of diazepam, toluene, and their combination. Act. Nerv.
    super. (Praha), 23: 179-182.

    VORHEES, C.V. (1974) Some behavioral effects of maternal
    hypervitaminosis A in rats. Teratology, 17: 269-273.

    VORHEES, C.V. & BUTCHER, R.E. (1982) Behavioral teratogenicity. In:
    Snell, L., ed. Developmental toxicology, London, Croom Helm Press,
    pp. 247-298.

    VORHEES, C.V., BRUNNER, R.L., MCDANIEL, C.R., & BUTCHER, R.E. (1978)
    The relationship of gestational age to vitamin A induced postnatal
    dysfunction. Teratology, 17: 271-276.

    WALLMAN, J. (1975) A simple technique using an optomotor response for
    visual psychophysical measurements in animals. Vision Res., 15: 3-8.

    WALSH, J.M., CURLEY, M.D., BURCH, L.S., & KURLANSIK, L. (1982) The
    behavioral toxicity of a tributyltin ester in the rat. Neurobehav.
    Toxicol. Teratol., 4: 241-246.

    WALSH, R.N. (1980) On the necessity for a shift in emphasis from
    means-oriented to problems-oriented research in developmental
    psychobiology. Dev. Psychobiol., 13: 229-231.

    WALSH, T.J., MILLER, D.B., & DYER, R.S. (1982b) Trimethyltin, a
    selective limbic system neurotoxicant, impairs radial-arm maze
    performance. Neurobehav. Toxicol. Teratol., 4: 177-184.

    WALTON, J.N. (1974) Disorders of voluntary muscle, 3rd ed.,
    Edinburg, London, Churchill Livingston.

    WANG, G.H. (1932) Relation between spontaneous activity and oestrous
    cycle in the white rat. Comp. Psychol. Monogr., 6: 1-32.

    WEIBEL, E.R. (1979) Stereological methods. In: Practical methods for
    biological morphometry, New York, Academic Press, Vol. 1.

    WEISS, B., FERRIN, J., MERIGAN, M., & COX, C. (1981) Modification of
    rat operant behaviour by ozone exposure. Toxicol. appl. Pharmacol.,
    58: 244-251.

    WEISS, B. & LATIES, V. (1961) Changes in pain tolerance and other
    behavior produced by salicylates. J. Pharmacol. exp. Ther.,
    131: 120-129.

    WEISS, B. & LATIES, V. (1969) Behavioral pharmacology and toxicology.
    Ann. Rev. Pharmacol., 9: 297-326.

    WEISS, B. & LATIES, V.G., ed. (1975) Behavioral toxicology, New
    York, Plenum Press, 469 pp.

    WEISS, B. & LATIES, V.G. (1979) Assays fox behavioural toxicity: a
    strategy for the Environmental Protection Agency. Neurobehav.
    Toxicol., 1(Suppl. 1): 213-215.

    WELCH, R.M., LEVIN, W., KUNTZMAN, R., JACOBSON, J., CONNEY, A.H.
    (1971) Effect of halogenated hydrocarbon insecticides on the
    metabolism and uterotropic action of estrogens in rats and mice.
    Toxicol. appl. Pharmacol., 19: 234-246.

    WELLS, G.A.H., HOWELL, J.M., & GOPINATH, C. (1976) Experimental lead
    encephalopathy of calves: histological observations on the nature and
    distribution of the lesions. Neuropathol. appl. Neurobiol.,
    2: 175-190.

    WHEATLEY, B., BARBEAU, A., CLARKSON, T.W., & LAPHAM, L.W. (1979)
    Methylmercury poisoning in Canadian indians -- the elusive diagnosis.
    Can. J. neurol. Sci., 6: 417-422.

    WHO (1975) Methods used in the USSR for establishing biologically
    safe levels of toxic substances, Geneva, World Health Organization,
    171 pp.

    WHO (1978) Principles and methods of evaluating the toxicity of
    chemicals. Part 1, Geneva, World Health Organization (Environmental
    Health Criteria Series No. 6, Part 1).

    WHO (1984) Principles for evaluating health risks to progeny
    associated with exposure to chemicals during pregnancy, Geneva,
    World Health Organization (Environmental Health Criteria Series
    No. 30).

    WHO (in preparation) Principles and methods of evaluating the
    toxicity of chemicals. Part 2, Geneva, World Health Organization
    (Environmental Health Criteria Series No. 6, Part 2).

    WILLIAMS, P.L. & WARWICK, R. (1975) Functional neuroanatomy of man,
    Philadelphia, Pennsylvania, W.B. Saunders Company.

    WILSON, W.E. (1982) Dopamine-sensitive adenylate cyclase inactivation
    by organolead compounds. Neurotoxicology, 3: 100-107.

    WINNEKE, G. (1979) Modification of visual evoked potentials in rats
    after long-term blood lead elevation. Act. Nerv. super. (Praha),
    21: 282-284.

    WINNEKE, G., BROCKHAUS, A., & BALTISSEN, R. (1977) Neurobehavioral and
    systematic effects of longterm blood lead-elevation in rats. I.
    Discrimination learning and open field behavior. Arch. Toxicol.,
    37: 247-263.

    WINNEKE, G., FODOR, G., & SCHLIPKOTER, H. (1978) Carbon monoxide,
    trichloroethylene, and alcohol: reliability and validity of
    neurobehavioural effects. Multidisciplinary perspectives in event-
    related brain potentials. In: Proceedings of the Fourth International
    Congress on Event-Related Slow Potentials of the Brain, Henderson,
    North Carolina, 4-10 April, 1976 (EPIC IV), Washington DC, US
    Environmental Protection Agency (EPA 600/9-77-0443).

    WINNEKE, G., LILIENTHAL, H., & ZIMMERMANN, U. (1984) Neurobehavioural
    effects of lead and cadmium. In: Proceedings of the Third
    International Congress on Toxicology, San Diego, 28 August-2
    September 1983, Amsterdam, Oxford, New York, Elsevier Science
    Publishers.

    WOLF, D.L. (1976) Rotating rod, spontaneous activity, and passive
    avoidance conditioning: their suitability as functional tests in
    industrial toxicology. In: Horvath, M. & Frantik, E., ed. Adverse
    effects of environmental chemicals and psycho- tropic drugs,
    Amsterdam, Oxford, New York, Elsevier Science Publishers, pp. 293-303.

    WOOD, R.W. (1978) Stimulus properties of inhaled substances. Environ.
    Health Perspect., 26: 69-76.

    WOOD, R.W. (1979) Reinforcing properties of inhaled substances.
    Neurobehav. Toxicol., 1(Suppl. 1): 67-72.

    WOOD, R.W. (1981) Neurobehavioral toxicity of carbon disulfide.
    Neurobehav. Toxicol. Teratol., 3: 397-405.

    WOOLLEY, D.E. (1977) Electrophysiological techniques in
    neurotoxicology. In: Zenick, H. & Reiter, L.W., ed. Behavioral
    toxicology: an emerging discipline, Research Triangle Park, North
    Carolina, US Environmental Protection Agency, pp. 9-1 -- 9-25 (EPA
    600/9-77-042).

    XINTARAS, C. & BURG, J.R. (1981) Neurotoxicity studies of workers
    exposed to carbon disulfide. G. Ital. Med. Lav., 3: 83-86.

    XINTARAS, C., JOHNSON, B.L., & DEGROOT, I., ed. (1974) Behavioral
    toxicology, Cincinnati, Ohio, US Department of Health, Education and
    Welfare, 507 pp (US DHEW (NIOSH) Publication No. 74-126).

    YAMAMURA, H.I., ENNA, S.J., & KUHAR, M.J. (1981) Neurotransmitter
    receptor binding, New York, Raven Press.

    YONESAWA, T., BORNSTEIN, M.B., & PETERSON, E.R. (1980) Organotypic
    cultures of nerve tissue as a model for neurotoxicity investigation
    and screening. In: Spencer, P.S. & Schaumburg, H.H., ed. Experimental
    and clinical neurotoxicology, Baltimore, Maryland, Williams and
    Wilkins Company, pp. 788-802.

    YOUNG, J.S. & FECHTER, L.D. (1983) Reflex inhibition procedures for
    animal audiometry: a technique for assessing ototoxicity. J. Acoust.
    Soc. Am., 73: 1686-1693.

    ZAKUSOV, V.V. (1936) [Estimation of the reflex time under the
    influence of some narcotic substances.] Fiziol. Zh. USSR,
    23: 276-278 (in Russian).

    ZAKUSOV, V.V. (1953) [Pharmacology of the nervous system,]
    Leningrad, Medgiz, 258 pp (in Russian).

    ZBINDEN, G. (1981) Experimental methods in behavioral teratology.
    Arch. Toxicol., 48: 69-88.

    ZEMAN, W. & INNES, J.R.M. (1963) Craigie's neuroanatomy of the rat,
    New York, Academic Press.

    ZENICK, H., RODRIQUEZ, W., WARD, J., & ELKINGTON, B. (1979) Deficits
    in fixed-interval performance following prenatal and postnatal lead
    exposure. Dev. Psychobiol., 12(5): 509-514.

    ZILOV, V.G., ROGACHEVA, S.K., & IVANOVA, L.I. (1983)
    [Electroencephalographic analysis of limbic-reticular interrelations
    in various motivational reactions of rabbits under ethanol.] Bull.
    exp. Biol. Med., 11: 74-77 (in Russian).

    ZUBER, V.L. (1978) Intensity of metabolism of various fractions of
    cerebral phospholipids under hyperphenyl-alaninaemia. Vop. med.
    Xhim., 2: 163-166.

    ZYLBER-HARAN, E.A., GERSHMAN, H., ROSENMANN, E., & SPITZ, I.M. (1982)
    Gonadotrophin, testosterone, and prolactin interrelationships in
    cadmium-treated rats. J. Endocrinol., 92: 123-130.
    


    See Also:
       Toxicological Abbreviations